THE CHEMISTRY AND MODE OF ACTION OF — PLANT GROWTH SUBSTANCES WAm & WIGHTMAN THE CHEMISTRY AND MODE OF ACTION OF PLANT GROWTH SUBSTANCES W '3 The Chemistry and Mode of Action of Plant Growth Substances Proceedings of a Symposium held at Wye College {University of London) July 1955 Edited by R. L. WAIN and F. WIGHTMAN LONDON BUTTERWORTHS SCIENTIFIC PUBLICATIONS V_ 1956 BUTTERWORTHS PUBLICATIONS LTD. 88 KINGSWAY, LONDON, W.C.2 BUTTERWORTH & CO. (AFRICA) LTD. DURBAN: 33/35 Beach Grove BUTTERWORTH & CO. (AUSTRALIA) LTD. .SYDNEY: 8 O'Connell Street MELBOURNE: 430 Bourke Street BRISBANE: 240 Queen Street BUTTERWORTH & CO. (CANADA) LTD. TORONTO: 1367 Danforth Avenue NEW ZEALAND: BUTTERWORTH & CO. (AUSTRALIA) LTD. WELLINGTON: 49/51 Ballance Street AUCKLAND: 35 High Street U.S.A. Edition published by ACADEMIC PRESS INC., PUBLISHERS 125 EAST 23rd STREET NEW YORK 10, NEW YORK AFRICA: AUSTRALIA: CANADA: Printed at The Universities Press , Belfast by Sir Isaac Pitman & Sons, Ltd. PREFACE During the last twenty years, developments in the field of plant growth substances have been rapid and impressive. As is well known, some of these materials have found important and widespread uses in agriculture and horticulture and are playing a vital role in increasing the yield of crops. At the same time, research along more fundamental lines has not been neglected, and, although the precise mechanism's which underlie the growth response are not yet vniderstood, much progress has been made in studies on physiological, chemical, and other aspects, and these investigations are being vigorously pursued in many countries. Furthermore, the development of paper chromatography has enabled critical investigations to be made on the auxin and auxin-inhibitor status of different plant tissues. Already the literature on all these fundamental aspects is very extensive and such is the rate of progress that it has become essential for specialized workers within these fields to have the opportimity of meeting from time to time to present and discuss their results. In September 1953, the writer attended an International Conference on Plant Growth Substances at the University of Lund, Sweden, organized by Professor H. Burstrom. All those present were working within this field. The high level and great value of the papers and discussions led, two years later, to the idea of holding a similar Conference in England. This suggestion was well received, and, as a result, workers from seven countries met at Wye College from 17-22 July 1955. The present volume contains the papers read at this Conference. For the purposes of historical record it is perhaps worth while to refer briefly to previous meetings of this kind. In this connection, the writer is indebted to Dr. P. Larsen for the reminder that the First International Conference on Growth Substances was held in Paris in 1937. This meeting was organized under the auspices of the League of Nations and the papers were published as a special volume. Etudes et recherches sur les phytohormones (Paris, 1938). Professor P. Boysen-Jensen was the President at this meeting and Professor V. J. Koningsberger and Dr. G. S. Avery were amongst those attending. The Second Growth Substance Conference, held in Wisconsin in 1949, was almost entirely American, Professor H. Burstrom being the only representative from Europe to present a paper. The contributions, edited by Professor F. Skoog and published in 1951 under the title Plant Growth Substances, have undoubtedly proved of great value to all workers in the field. Although only an abstract of the proceedings of the Lund Conference (1953) was published, the great success of this meeting must be recorded. The Wye Conference was organized along similar lines with the help of a grant kindly provided by the Governing Body of Wye College. Their generosity in this respect, and that of Sir William Slater, Secretary of the Agricultural Research Council, is gratefully acknowledged. The following acted as Chairmen at the various sessions : Prof H. Veldstra, Preface Dr. G. S. Avery, Prof. J. Bonner, Prof. T. A. Bennet-Clark, Prof. H. Burstrom, Dr. P. Larsen, and Prof. L.J. Audus. Their assistance in diis connection was greatly appreciated. It is also a pleasure to record the help provided by all members of the Chemistry Department and Agricultural Research Council Growth Substance Unit at Wye in preparing for this meeting. Finally, warmest thanks are due to Dr. Frank Wightman, the Organizing Secretary, for all his efforts to ensure the smooth running of the Conference. R. L. Wain Wye College, University of London 16 September 1955 VI CONTENTS PACE Preface ........... v Conference Photograph . . . . . . . . x Members of the Conference ....... xi I. NATURAL AUXINS Methods for the investigation of natural auxins and growth inhibitors. J. P. Nitsch {Harvard) .... 3 The distribution of natural hormones in germinating seeds and seedling plants. Phyllis M. Cartwright, J. T. Sykes, and R. L. Wain ( Wye) 32 Hormones and hormone precursors in leaves, roots, and seeds. Joyce A. Bentley, S. Housley, and G. Britton {Manchester) . 40 Chromatographische Untersuchungen iiber die Wuchsstoffe und Hemmstoffe der Haferkoleoptile. H. Soding und Edith Raadts {Hamburg) ....... 52 Indole compounds in photoinduced plants. A. J. Vlitos {New York) 57 The biogenesis of natural auxins, .S*. A. Gordon {Lemont) . 65 Geotropic responses in roots. Some theoretical and technical problems. P. Larsen {Bergen) ..... 76 II. CHEMICAL STRUCTURE AND BIOLOGICAL ACTIVITY On the effects of /;ara-substitution in some plant growth regulators with phenyl nuclei. B. Aberg {Uppsala) . . 93 On form and function of plant growth substances. H. Veldstra {Amsterdam) . . . . . . . • 117 The mode of growth action of some naphthoxy compounds. H. Burstrom and Berit A. M. Hansen {Lund) . . ■ 134 VII '5'3135 Contents Chemical configuration and action of different growth sub- stances and growth inhibitors: new experiments with the paste method. H. Linser {Linz) ■ ■ ■ • • 141 The action-concentration curves of mixtures of growth- promoting and growth-inhibiting substances. K. Kaindl {Linz) 159 The chemical induction of growth in plant tissue cultures. F. C. Steward and E. M. Shantz {Cornell) . . . . 165 The degradation of certain phenoxy acids, amides, and nitriles within plant tissues. C. H.Fawcett, H.F. Taylor, R. L. Wain, and F. Wightman {Wye) . . . . . . 187 Relationship of molecular structure of some naphthyloxy compounds and their biological activity as plant growth regulating substances. L. C. Luckwill and D. Woodcock {Long Ashton) 195 Studies on the relation between molecular structure and pene- tration of growth regulators into plants. J. van Overbeek {Modesto) 205 The resolution of phyto-hormone acids by means of their optically active phenyhjopropylamides. Resolution of a- (1-naphthyl) -propionic acid. J. M. F. Leaper and J. R. Bishop {Ambler) 211 III. METABOLISM AND MODE OF ACTION Some metabolic consequences of the administration of indole- acetic acid to plant cells. A. W. Galston {Pasadena) . . 219 Chromatographic investigations on the metabolism of certain indole derivatives in plant tissues. R. C. Seeley, C. H. Fawcett, R. L. Wain, and F. Wightman {Wye) . . . . 234 The effects of synthetic growth substances in the level of endogenous auxins in plants. L. J. Aitdns and Ruth Thresh {London) 248 Interrelationships between the uptake of 2 :4-dichlorophenoxy- acetic acid, growth, and ion absorption. G. E. Blackman {Oxford) 253 Auxin-induced water uptake. J. Bonner, L. Ordin, and R. Cleland {Pasadena). 260 viii Contents The influence of growth substances upon sulphydryl compounds. A. C. Leopold and C. A. Price {Purdue) . . . . 271 Salt accumulation and mode of action of auxin. A preliminary hypothesis. T. A. Bennet-Clark {London) . . . 284 IV. APPLICATIONS OF KINETICS TO AUXIN-INDUCED GROWTH The kinetics of auxin-induced growth. J. Bonner and R. J. Foster {Pasadena) 295 The kinetics of auxin-induced growth. T. A. Bennet-Clark {London) 310 IX "^ Members of the Conference D. L. Abbott, Long Ashton B. Aberg, Uppsala, Sweden L. J. AuDUS, London G. A. Avery, Brooklyn, New York H. W. B. Barlow, East Mailing G. E. Barnsley, Woodstock J. A. Bentley, Manchester T. A. Bennet-Clark, London G. E. Blackman, Oxford J. Bonner, Pasadena, California R. C. Brian, Jealott's Hill R. Brown, Oxford H. Burstrom, Lund, Sweden P. M. Cartwright, Wye E. E. Cheesman, London P. Button, Manchester C. H. Fawcett, Wye A. Fredga, Uppsala, Sweden A. W. Galston, Pasadena, California D. J. Galston, Pasadena, California S. A. Gordon, Lemont, Illinois B. Hansen, Lund, Sweden C. R. Hancock, East Mailing E. S. J. Hatcher, East Mailing N. He YES, Oxford B. J. Heywood, Dagenham S. HousLEY, Manchester A. Jonsson, Stockholm K. Kaindl, Linz, Austria V. J. Koningsberger, Utrecht P. Larsen, Bergen, Norway J. M. F. Leaper, Ambler, Pennsylvania A. C. Leopold, Lafayette, Indiana H. Linser, Linz, Austria E. J. Maskell, Birmingham H. Mayr, Linz, Austria C. C. McCready, Oxford D. G. Morgan, Hamburg J. P. NiTSCH, Harvard, Massachusetts D. J. Osborne, Oxford J. VAN Overbeek, Modesto, California R. PoHL, Freiburg, Germany R. D. Preston, Leeds E. Raadts, Hamburg L. Reinhold, Oxford G. G. Samuel, London W. W. ScHWABE, Rothamsted R. C. Seeley, Wye Sir William Slater, London H. Soding, Hamburg M. S. Smith, Wye D. M. Spencer, Wye F. C. Steward, Ithaca, New York N. Sunderland, Oxford J. F. Sutcliffe, London J. T. Sykes, Wye F. Taylor, Wye R. Thresh, London A. J. Vlitos, New York H. Veldstra, Amsterdam D. Vince, Reading R. L. Wain, Wye R. L. Weintraub, Camp Detrick, Maryland F. Wightman, Wye XI NATURAL AUXINS METHODS FOR THE INVESTIGATION OF NATURAL AUXINS AND GROWTH INHIBITORS! J. P. NitschJ Biological Laboratories, Harvard University The field of knowledge of plant growth substances has developed tremend- ously, but unequally. Thus, the knowledge of synthetic compounds capable of modifying the growth of plants in a manner similar to native substances has accumulated at a prodigious pace, as have the horticultural and agricultural applications of this knowledge. One must admit, however, that the science of auxins rests upon the small and shaky foundation of the meagre and in- adequate knowledge we have of the identity of the native substances at work in the plant. Moreover, the numerous attempts which have been made to explain how 'auxins' in general regulate growth may be futile unless we actually know the nature of all the auxins that occur in the normal develop- ment of plant tissues. It is high time that physiologists and chemists made a really comprehensive survey of the native growth-promoting and growth- inhibiting substances present in plants. The main difficulty in such work arises from the extraordinarily low concentration of these substances in plants. Fortunately, the new techniques of paper chromatography and paper electrophoresis, recently applied to auxin problems by many different workers, show great promise as tools in this task. While trying to proceed in this direction, however, the plant scientist encounters many technical problems which have to be solved before any real advance can be made. Among these, one may list: a reliable extraction method, a good chromatographic technique, and a suitable bio-assay. Like other workers, we have been faced with these questions, and have tried to study the matter systematically. The results of our effort will be presented in this paper. We will consider successively the extraction of the active substances from the plant material, the purification of the extract, the chromatographic techniques, the bio-assay, and, finally, a technique for the chemical identification of the substances on the chromatograms. THE EXTRACTION OF AUXINS FROM PLANT MATERIAL When we extract something from the plant, we want the extraction to be complete; we want no more and no less than what is actually present in the tissues. This means that we do not want any synthesis of auxin to take place i" This work was supported in part by the American Cancer Society, Inc.. as recommended by the Committee on Growth of the National Research Council, and in part by the National Science Foundation through grants-in-aid made to Professors R. H. Wetmore and K. \'. Thimann of Harvard University. The author expresses his deep gratitude to Professor Wetmore and Professor Thimann for the facilities, help, and inspiration which they so generously provided, and to his wife, who performed a large number of the experiments. J Present address: Department of Floriculture and Ornamental Horticulture, Cornell University, Ithaca, New York, U.S.A. Natural auxins during the extraction procedure, nor do we want any destruction of these substances. It has been shown by many authors (Thimann et al., 1940, 1942; Gustafson, 1941 ; van Overbeek et al., 1947) that when ether is used, auxin can be generated from plant material over periods as long as a year. Temperature stimvdates this generation (Wildman and Muir, 1949). That boiling stops the production indicates the enzymatic nature of the process. On the other hand, especially when tissues are ground or cut, the destruction of auxins during extraction may be extensive, depending on the kind of material used. This destruction is also an enzymatic phenomenon of the oxidative type (Tang and Bonner, 1947; Wagenknecht and Burris, 1950; Steeves et al., 1953; Briggs et al., 1955a, b). To obtain a reliable picture of the auxins present in the plant at the time of extraction, one has to stop this oxidative destruction of auxins. Table 1 gives the results of a series of tests performed on the tuber tissue of Jerusalem artichoke. The extent of browning was taken as an index of the oxidative activity at the cut surfaces. Tissues in ether and ethyl acetate turned dark brown, indicating that these solvents do not stop the activity Table 1 The prevention of tissue browning during auxin extraction Jerusalem artichoke tuber tissues left overnight in the ice-box at about 5°C. Extracting solvent Browning after 18 hours Absolute methanol Absolute ethanolf + Acetone + Petroleum ether (boiling point: 30-65°C) + + + Chloroform + + + + + Chloroform + 8-hydroxyquinoline (2 mg/c.c.) + + + + Ethylacetate +++++++ Ether +++++++ Ether (95%) + absolute methanol (5%) + + + + Ether (70%) + absolute methanol (30%,) + + + Ether (50%) + absolute methanol (50%) + + Ether (70%) + absolute ethanol (30%o) + + + Ether + NajS204 (1 mg/c.c. undissolved J) Ethylacetate (50%) + absolute methanol (50%,) ++ t The tissues are shrivelled after extraction. J Dissolves Vi'hen the fresh tissue is put in the mixture. of at least the polyphenoloxidase system. It is not surprising then, that previous workers have found that enzymes which produce auxins also are not inactivated by ether. Unless special precautions are taken, such as short time extraction and low temperatures, ether is not a reliable extracting solvent for auxins. According to Table 1, chloroform is slightly better than ether in preventing browning, but petroleum ether, acetone, and absolute ethanol are even better. The very best solvent, however, seems to be 4 Investigation of natural auxins and growth inhibitors methanol, which yielded beautiful white tissues without the slightest trace of browning. To improve the ether solvent, several addenda were tried. The best one was Na2S204, which is practically insoluble in ether, but which dissolves in the water coming out from the tissues. This powerful reducing agent kept the tissues perfectly white, but the effect it might have on the auxins themselves has not yet been investigated. The addition of methanol to ether improved the latter solvent so much that it was not necessary to look further into the matter. It was inferred, as a working hypothesis, that since either methanol alone or a mixture of ether and methanol would prevent browning, they would perhaps also stop the enzymatic activity of plant tissues which have caused in the past so many artefacts during auxin extrac- tion. If enzymatic activity is prevented (at least by the 0°C temperature and the short duration of the extraction, such as 1-3 hours), then there should be no danger of other artefacts, such as those arising from an enzymatic esterifica- tion of indole-3-acetic acid (lAA) with methanol or ethanol. Such an esterification does not seem to proceed at an appreciable rate in vitro. We have left lAA in ethanol (1 mg/c.c.) for three months at 25°C in the dark, and after that time we have chromatographed it and detected only one spot on the paper, that of lAA. Of course, we not only need a solvent which does not cause artefacts to develop, but also one which has a good extractive power. In order to deter- mine how several organic solvents compare, as far as the actual extraction of auxins is concerned, we have performed with different solvents parallel extractions of aliquots of the same material. The extracts were chromato- graphed, so that not only the total ciuantity, but also the nature of the extracted substances would become apparent. Three examples of this study will be given. The first one {Figure 1) concerns tomato fruits, harvested 20 days after pollination, which have been immediately frozen, lyophilized, and ground to a fine powder. Aliquots of 100 mg (dry weight) of this powder were extracted for 3 hours at 0°C (ice-bath) with 20 c.c. of the chosen solvent, plus two rinsings. The extracts were filtered over cotton, concentrated under reduced pressure, and transferred to the paper strips by means of a tuberculin syringe, as previously described (Nitsch and Nitsch, 1955). The chromatograms were run in the dark and at room temperature in the ?>opropanol (80)-f 28 per cent ammonia (10)+H2O (10) (v/v) solvent recom- mended by Stowe and Thimann ( 1 954) . When the solvent front had travelled 20 cm past the initial spot, the paper strips were taken out of the tubes, dried in a stream of air, and cut transversely into twenty segments 1 cm wide. Each of these segments was incubated with 0-5 c.c. of a buffer at pH 5-0 (see below) containing 2 per cent sucrose and with ten 4 mm oat coleoptile sections. The diagrams o^ Figure 1 show the location on the chromatograms and the activity in the coleoptile sections of the growth substances extracted with (1) anhydrous acetone, (2) ether, (3) ethyl acetate, and (4) absolute methanol. In all cases, two groups of active compounds are apparent, one around Rf 0-4, the other around RfO-7. The size of the growth peaks varies with the solvent, however. In addition, the number of the peaks also varies with the solvent. For example, acetone seems to extract substances which remain near the origin and which are not visible in the diagram corresponding to the ether Natural auxins extract. Indications of the presence of growth inhibitors can be seen, especially in the ether extract. When the amount of tissue extracted with ether is increased, the inhibitory regions around Rf 0-5 and Rf 0-9 become very marked. Thus, this first example shows that different extracting solvents may give different auxin pictures, both quantitatively and qualita- tive'ly. As far as quantities are concerned, with the tissue used in this 3-5 - 2-0- '111 f-5 2-5- 2-0 1-5 - Acetone i ^1 Ether ^It^i li rtil I ! I I Ethyl acetate i 111' Figure 1. Histograms showing the biological activity of substances extracted ivith different solvents from tomato ovary tissues. The unripe tomato fruits (male sterile mutant of the var. John Baer) were harvested 20 days after hand- pollination and lyophilized. 100 mg {dry weight) of the lyophilized powder were extracted for 3 hours at C with the indicated solvent. The crude extracts were concentrated under reduced pressure and chromatographed in KOpropanol +2?>°o ammonia + water (80; 10; 10) {vjv). When the solvent front had travelled 20 cm past the initial spot, the paper strips ( Whatman No. 1 paper) were air-dried and cut into twenty 1 cm segments. Each of these paper segments was incubated with 0-5 c.c. of solution and ten oat coleoptile sections. The elongation of these sections after about 20 hours is plotted along the orditmte axis of the diagrams. The position of the paper segment on the original chromatogram is plotted in Rf units along the other axis. particular instance, ethyl acetate and especially methanol were better than ether. A second example is taken from another fruit, the bean. Lyophilized bean seeds, dissected out of the pods about 8 days after pollination, were extracted for 5 hours at 0°C in the dark with 10 c.c. each of absolute methanol, absolute ethanol, or anhydrous ethyl acetate. The aliquot extracted with ethyl acetate was washed at the end with absolute ethanol and the two extracts combined. The chromatograms were developed in the same solvent Investigation of natural auxins and growth inhibitors as above and assayed with oat coleoptile segments. The results shown in Figure 2 were obtained. It is clear from Figure 2 that for bean seed tissue methanol seems to be a better solvent than ethanol, since the peaks of growth, especially the one in the indole-3-acetonitrile (IAN) region, are higher. It seems also that ethyl acetate extracts a still higher amount of auxins, although the peak in the IAN region is less marked ; in addition, ethyl acetate appears to extract a substance having a /?^ higher than lAA but lower than IAN. To find out whether ethyl acetate was a better solvent than methanol, the following experiment was performed. Lyophilized immature bean seeds were first extracted for one and a half hours with absolute methanol at 0°C, then with anhydrous ethyl acetate for another one and a half hours, also at 0°C. Figure 2. Histograms made under conditions similar to those of Figure 1 . except that the three extracts are of 15 mg {dry weight) each of lyophilized bean seeds {var. Dwarf Shell Red Kidney), dissected out of the immature pods about one week after /lowering. The extractions were made at C during 5 hours with absolute methanol, absolute ethanol. and ethyl acetate respectively. The material extracted with ethyl acetate ivas washed with absolute ethanol and the two extracts were combined. The two extracts were chromatographed separately in our new /j-obutanol (80)+methanol (5) +H2O (15) (v/v) solvent which has been substituted for the previous one for reasons which will be discussed below. The results obtained when the chromatograms were assayed with oat coleoptile segments (see Figure 3) show clearly that after the methanol extraction, which yields large quantities of auxins, ethyl acetate extracts only very small amounts. In addition, no indication of auxins different from those extracted with meth- anol can be seen on the diagram corresponding to ethyl acetate, since the small peaks obtained correspond in general with those of the other diagram. It may be noted in passing that the chromatographic solvent used in the present instance reveals the existence of substances which were not found with the fj^opropanol-ammonia solvent (compare with Figure 2). These substances either are not separated or are transformed into other substances during Natural auxins chromatography in the latter solvent. As far as the identity of the growth peak found in the lAA region is concerned, it is most probably lAA since the characteristic colour reaction of lAA was obtained on the paper in the lAA position with both the Salkowski and the Ehrlich reagents. Figure 3. Histograms showing the growth- promoting activity of 1 cm sections of chromato- grams run in isobutanol-methanol-water (80 : 5; 15). Lyophilized immature bean seeds {as in Figure 2, but 30 mg dry weight) were first extracted at C for 1-5 hours with 10 + 5+5 c.c. of cold absolute methanol. This extract gave the upper histogram. The same bean material was then extracted for another 1-5 hours at C with 10 + 5 + 5 c.c. of cold ethyl acetate and chromatographed in the same solvent {lower histogram) . THE PURIFICATION OF THE EXTRACTS To obtain a clean chromatogram with well-separated spots, one has generally to purify the crude plant extract. Such a purification can be achieved in several ways, among which we have used the following ones : 1. Bicarbonate purification — This well-known technique, already used by Boysen-Jensen (1941), purifies the acid auxins, especially lAA, very well. They enter the aqueous phase as salts, whereas most of the coloured impurities remain in the ether phase. Upon acidification of the aqueous fraction with dilute HCl (Larsen, 1955), it is possible to re-extract the acid auxins with fresh ether. Clean chromatograms of lAA which can be used for colour reactions are obtained in this way. 2. Acetonitrile purification — The bicarbonate technique, unfortunately, does not work for neutral auxins. Adapting a technique devised originally for insecticides by Jones and Riddick (1952), it was possible to eliminate at least the fatty substances and some of the carotenoids from the extract (Nitsch, 1955). The original extract is evaporated down to dryness, then shaken with 8 Investigation of natural auxins and growth inhibitors about equal parts of anhydrous acetonitrile and hexane or heptane. Under these conditions, lAA and IAN remain in the acetonitrile phase, as well as ethylindole acetate (lAE), although there might be, possibly, some losses of the latter compound. 3. Chromatographic purification — It was found that some of the coloured impurities of plant extracts are strongly absorbed on paper, so that when a chroraatogram is eluted with ether, some of the coloured material remains on the paper. Two types of chromatographic purifications have been used : (a) The crude extract was first chromatographed in distilled HoO and dried. The initial spot was then cut off and the rest of the paper eluted with ether or absolute ethanol. (b) The crude extract was first chromatographed in ziopropanol (80) -|- ammonia (10) -{-water (10) (v/v) and dried. Then the initial spot was cut off and discarded, and the rest of the paper cut transversely in the middle. The lower portion then contains mainly the acid auxins and the upper one the neutral auxins which can be eluted and rechromatographed separately. During the purification procedure, there is always a possibility of loss of active material and alteration of labile compounds. It has been possible, however, to avoid any purification by using crude extracts of small quantities of plant material and a very sensitive bio-assay test. The procedure adopted will be described below. THE CHROMATOGRAPHIC PROCEDURE In order to obtain an initial spot as small as possible, the plant extract is applied on the starting line of the paper strip by means of a tuberculin syringe {Figure 4) in a stream of air to allow the rapid evaporation of the Figure 4- Tectinique used to obtain small initial spots on cliromatograms. The extract (£) was taken up in a tuberculin syringe (.S) and deposited on the starting point of a paper chromatogram (C) laid down on a perforated paper disk (P). A stream of air, Jittered through cotton, was blown through the glass tubing at the right and was sucked into the suction flask connected with a vacuum pump, thus activating the evaporation of the extract on the paper strip {cf> Nitsch and Nitsch, 1955). solvent (Nitsch and Nitsch, 1955). After complete drying, the paper strip is equilibrated overnight over the solvent in a large glass tube {Figure 5), a technique originally used by Stowe and Thimann ( 1 954) . All the experiments reported were performed at room temperature in a dark cabinet. The finding of a suitable solvent is the key to successful chromatographic separation of a mixture of compounds. More or less empirically, many Natural auxins workers in different laboratories arrived at various recipes for solvent mixtures that would separate certain auxins, especially the acid ones. In general, a mixture of an alcohol, ammonia, and water has been adopted. Stowe and Thimann (1954) substantiated this through systematic studies IV- ^C(^T h fi Figure 5. (a) The chromatoamm is first equilibrated over the solvent overnight. A paper wick (W) keeps the upper part of the tube saturated with the vapour of the solvent. A ring of rubber tubing (R) supports the inner glass rod. (b) After equilibration, the inner glass rod is pushed down so that the lower 5 mm oj the paper strip will penetrate into the solvent. {c) When the solvent has travelled up a height (H) of 20 cm. the paper is air-dried, and sprayed with a suitable reagent. Both the position (h) and length (I) of the coloured spots are noted. The technique illustrated under (a) and (b) is essentially that of Stowe and Thimann, 1954. {After Nitsch and Nitsch, 1955.) and recommended a mixture of zVopropanol (80) +28 per cent ammonia ( 1 0) + water ( 1 0) (v/v) which we have used on many occasions with good results. We have tried to go further in this investigation by asking this question: 'What is the simplest mixture which can separate auxins on paper?' The results of Table 2 speak for themselves : first of all, the presence or the absence of ammonia does not appreciably affect the separation of the three auxins Table 2 Search for the simplest chromatographic solvent All paper strips were equilibrated overnight over the solvent. Solvent RA IB AX lAE woPropanol (80%) + water (10%) f.roPropanol (80%) + water (10%) woPropanol (100%) Water (100%) 28% ammonia (10%) 0-35 0-48 0-76 0-40 0-50 0-75 lost? 0-65 0-64 0-32 0-83 0-83 0-83 0-47 t The symbol Rf ('ratio front') designates the ratio of the movement of a given conipouml to the movement of the front of the solvent. Rf — hjH {see Figure 5). } IBA = indole-3-butyric acid. 10 Investigation of natural auxins and growth inhibitors considered, nor their R/s (see also Figure 6(a)); secondly, the removal of uopropanol changes the R/s but does not impair the separation of the compounds — in fact, it improves it; thirdly, the removal of water definitely ruins the value of the solvent. Thus, from all components of the mixture only one is really indispensable, namely water. This is true even in the case of water-insoluble compounds such as hexane, as demonstrated by Nitsch and Nitsch (1955). It is probable that water, at least in the vapour state, has to impregnate the paper fibres in order to give a good chromatogram. These investigations give support to the use of water alone previously reported by Bitancourt, Schwartz, and Dierberger (1954) and by Sen and Leopold (1954). These authors, however, did not equilibrate their paper strips over the water (^) I/^ ^lAE o IAN 05 1 2 3315 678 3 10 Normal ify of NH^OH -* p H — Figure 6. (a) R/s of I A A, lAE, and IAN in water to which various concentrations of ammonia have been added, {b) The effect of pH on the R/s oflAA, lAE, and IAN in water. pH 3-0 has been obtained both with citrate-phosphate buffer (about 0-03 M) and HCl (0-001 M) ; pH'5 4-0-7-0 with citrate- phosphate buffers [about 0-03 M) ; pH 8-0 with citrate-phosphate buffer and with borate-KC\ buffer {about 0-05 M) ; pHV 9-0- 10-0 with borate-KCl buffers {about 0-05 M) {cf. Nitsch and Nitsch, 1955). solvent before chromatography. It has been shown (Nitsch and Nitsch, 1955) that a preliminary equilibration improves markedly the quality of the chromatography in water by giving compact spots. The use of water alone in the chromatographic separation of auxins has its limitations. First of all, it does not separate the acid auxins one from another (for example, lAA and indole-3-butyric acid (IBA) practically run together, as shown in Table 2). In the second place, when fatty material is present in the plant extract, the auxins, more soluble in fats than in water, may not move readily out of the initial spot, and streaking may occur. For these reasons, it was necessary to improve the water solvent. Changing the pH did not help at all {Figure 6(b)), but the addition of an organic solvent such as acetonitrile, or an alcohol, allowed a good separation of several acid auxins (Figure 7) , whereas the neutral ones, such as IAN and I AE, ran together. It was then concluded that no one solvent could be ideal for compounds as widely different in their physico-chemical properties as acids, esters, and nitriles, and that it was better to look (a) for a solvent separating the acid compounds, (b) for another one separating the neutral substances. For acid auxins there exists already an excellent solvent, the one developed by Stowe and Thimann (1954). The need for a different mixture was not felt until it was discovered, in the course of the search for a solvent applicable 11 Natural auxins to neutral auxins, that ammonia, under certain circumstances, leads to the hydrolysis of lAE ( Table 3). Such an hydrolysis was not apparent (using only a colorimetric detecdon of auxins) with the ?\yopropanol solvent. However, lOr Acefonifnile ■j/\E' \zoPropano/ oio~ZO 30 W 50 eO 70 80 90 WO 10 80 30 W 50 BO 70 80 SO 100 IVafer — » % Figure 7. Effect on the R/s of I A A, IBA, IAN, andlAE of various percentages of acetonitrile {left) and isopropanol (right) in water. the possibility that ammonia might affect natural auxins more labile than lAE led us to start a new search for a good solvent to separate the acid auxins. Knowing that ammonia did not change the R/s appreciably, we simply omitted it and used Z5opropanol+ water (80;20) (Nitsch and Nitsch, Table 3 Hydrolysis of ethylindole acetate by ammonia on paper chromatograms All the chromatograms were equilibrated overnight, then run for about 8 hours at room temperature in the dark. Only ethylindole acetate (4 y) was applied to each chromatogram initially. The spots were revealed with either the Gordon and Weber or the Ehrlich reagent. Intensity of colour Solvent mixture viv) {Proportions in per cent lAE spot lAA spot Hexane H2O 28°o ammonia 80 20 _ + + + + 80 18 2 + + + + 80 16 4 + + + + + 80 10 10 + + + + + 80 4 16 — + + + + 80 — - 20 _!_ ^ + + woPropanol H,0 28% ammonia 80 20 _ + + + + — 80 10 10 + + + + — 80 5 15 + + + + — 80 ■ 20 + + 12 Investigation of natural auxins and growth inhibitors 1955). Later, however, with a new supply of /iopropanol, we occasionally ran into two difficulties : ( 1 ) the I AA spot was somewhat larger than when ammonia was present, (2) the R^ of lAA was sometimes too high. Several different alcohols were then tried systematically {Table 4). In general, Table 4 Effect of various alcohol isomers on the size of the lAA spot The chromatograms were equilibrated overnight, then run for 8-10 hours at room temperature in the dark. 2 y of the indole-3-acetic acid were applied to each strip. .\11 proportions are v/v. Solvent mixture Length of spot in mm Propanols K-propanol (86%) + H20 (14%) 22 uopropanol (86%) + H20 (14%) 24 Butanols «-butanol (86%) + H.,0 (14%) 17 /iobutanol (86%) + H20 (14%) 15 sec.huianol (86%) + H20 (14%) 27 tert.huianol (86°;,)-H,0 (14%) 20 compact lAA spots were obtained with A?-butanol, but /Vobutanol was even better, yielding consistently small, rounded spots with lAA and indole-3- butyric acid (IBA). However, with zj^obutanol, the Rf of the lAA spot was too low. It is known (Stowe and Thimann, 1954; Nitsch and Nitsch, 1955) that increasing the percentage of water in an alcohol-water mixture brings the Rf of I AA up [Figure 7) . Unfortunately, water is not very soluble in /jobutanol, so that a one-phase mixtiu-e cannot be achieved with more than about 14 per cent water at room temperature. This difficulty was overcome by adding to the uobutanol-water mixture either tertiary butanol (in which water is soluble in all proportions) or methanol. Thus, the following- solvents gave good results: (1) zVobutanol {p(i°/Q)-\- tertiary butanol (30%) +water (20%), and (2) /^obutanol (80%)+methanol (5%)+water (15%). The latter solvent was preferred because it ascends somewhat faster than the first one, which moves very slowly. In conclusion, two good chromato- graphic solvents for auxins are as follows : lAA IBA IAN lAE Stowe and Thimann's uopropanol (80%) +28% ammonia (10%) + water (10%) ziobutanol (80%)+methanol (5%) +water(15%) The numbers in parentheses give the length of the spots in mm for 2 y of lAA, 1 y of IBA, 10 y of IAN, and 5 y of lAE when the solvent front has moved 20 cm. It can be judged by comparing the diagrams of Figures 2 and 3 that ^-- 13 0-37 0-50 0-75 0-82 (18) (17) (17) (24) 0-24 0-45 0-76 0-85 (16) (18) (25) (16) Natural auxins the ?Vobiitanol solvent proposed here separates a greater number of different compounds in bean extracts than does the zi-opropanol-ammonia-water solvent. In attempting to separate the neutral auxins such as IAN and lAE, it was observed that they generally have high R/a in the alcohol-water or acetonitrile-water mixtures. We looked, therefore, for solvents in which these substances would be very slightly soluJDle. Thus hexane and heptane were tried, together with carbon tetrachloride, jjenzene, carbon disulphide, toluene, etc. It was soon found that success or failure in these attempts depended on the presence of a very small amount of water, even though these solvents were water-insoluble. In the case of carbon tetrachloride and carbon disulphide, which are heavier than water, the paper strip was lowered in a beaker containing the organic solvent with water surrounding the beaker. When water was present, IAN and lAE were usually separated, except in the case of CCI4, CS2, and benzene, which are slightly soluble in water. The best results were obtained with hexane {Table 5) and the 'practical' grade Table 5 The use of water-insoluble compounds in the chromatographv of auxins Solvents R, Comments lAA IAN lAE Hexane (Eastman, practical) (90%) + water (10%) U 0-50 0-98 Excellent separation Hexane (100%) 0-57 O'c/oHexane (90%,)+ water (10%) 0-39 0-97 «-Heptane (90%) + water (10°o) 0-26 0-99 Trimethylpentane (90%) + water (10%o) 0-14 0-78 Benzene (95%) + water (5%) 0-81 0-93 IAN and lAE spots touching Benzene (100%) lost 0-63 Carbon tetrachloride (90%) + water (10%) 0-89 0-93 IAN and lAE spots touching Carbon disulphide (90%) + water (10%) 0-80 0-87 IAN and lAE spots touching (Eastman P 1 135) was better than w-hexane. Although water is insoluble in hexane, it was noted that the amount of water in the tube modifies the position of the IAN spot on the chromatogram (Nitsch and Nitsch, 1955). This peculiarity seems to depend on how well the paper wick plunges into the water phase to provide enough water vapour above the solvent; indeed, when the paper wick does not touch the water, the chromatograms may behave as if they were run in anhydrous hexane. The fact that two types of chromatographic solvents are now available, one for the acid auxins, the other for some neutral ones, does not, unfortunately, solve everybody's problems. Rather, it seems that each worker may have to devise his own solvent to fit his particular needs, though it is hoped, at least, that the principles studied above may be useful in this task. 14 Investigation of natural auxins and growth inhibitors THE BIO-ASSAY Since the primary aim of our study is to separate and identify biologically active substances, it is important to have a sensitive, reliable, and easy-to- perform bio-assay. The results of a search for a suitable plant material! l^d us to conclude that the coleoptile and the first internode of the oat seedling were suitable test objects. The variety Brighton:|:, a hulless oat, was used throughout these studies. When coleoptiles were wanted, the seeds were exposed to red light after soaking to prevent the growth of the first internode. Figure 8. The physiological 'merry-go-round'' used to rotate the oat first internodes. The 125 c.c. Erlenmeyer flasks are ready to receive the 13 X 100 mm Pyrex tubes containing the sections plus 0-5 c.c. of solution. If the atmosphere of the room is dry, the tubes should be closed with vaccine stoppers. The machine rotates at 1 r.p.m. When first internodes were wanted, the seeds were grown in total darkness. All manipulations were performed under green light, at a wavelength (about 546 m^) which has practically no inhibitory effect on the growth of first internodes. A difficulty encountered early in this work was that the first internodes curved so much in the auxin solutions that they were difficult to measure. Rotating the sections around a horizontal axis solved this problem, and the sections coming out of the physiological 'merry-go-round' [Figure 8) grew straight. The coleoptiles, on the other hand, were gently shaken on a "j" The details of this work can be found in a paper by Nitsch and Nitsch, to appear in Plant Physiol. 31 (1956). % Kindly supplied by the Canadian Department of Agriculture, Central Experimental Farm, Ottawa, Canada. 15 Natural auxins Figure 9. Thimanns guillotine, which allows exact sectioning of coleoptiles and first iniernodes. Avena coleopfile Sorghum first infernode Avena first internode 13 3 f 5 Tn.iri. yj" W 1 2 3 5 Distance from tip or node Figure 10. Effect of the location o/4 mm section on the growth in sucrose+ buffer without added I AA {black circles, dotted lines) and with lAA {white circles, solid lines), and on the difference of these two elongations {curves {b), {d), and (/)). Left ordinate: scale in mm. Right ordinate: scale inper centof the original 4 mm length. On the abscissa of the curves (c). {d). {e), and (/). 'xV means that the section includes the coleoptilar node. Each point is the mean of ten replicates. 16 Investigation of natural auxins and growth inhibitors horizontal shaker. To be able to section the coleoptiles and internodes accurately, we used Thimann's 'guillotine' {Figure 9), an improvement over the original Van der Weij's 'coleoptile microtome' (1932). The primary leaf was always left included in the coleoptile. This procedure simplifies the test without impairing its sensitivity or reliability. The first question to be decided was: 'Where should we cut the sections?' A series of experiments in which 4 mm sections were cut, first including the tip or the node, then starting 1 mm, etc., below them, gave the results shown in Figure 10. It should be noted that what is important here is not so much Coleoptile First internode Figure 11. Effect of the initial length oj oat sections on their growth in buffer + sucrose without added 1.4.4 (white circles, dotted lines) and with I.AA {white circles, solid lines) : graphs (a) and {d) . The other curves represent this effect on the difference Gix^ — G^ expressed in mm {curves {b) and {e)) or in per cent of the original lengths {curves {c) and if)). 2 3 1 S 10 2 3 15 Initial lengtt) of section 10 mm the maximum growth in lAA as the maximum differential growth between the lAA-treated sections and the controls. This difference, which we can represent by Gj^^ — Gq (growth in lAA minus growth without lAA), is maximum under our conditions when the 4 mm section is cut 3 mm below the tip or 3 mm below the node of the oat seedling. Since the growth of internode sections taken 3 mm below the node is very low without auxin (which makes it difficult to detect any inhibitor on chromatograms), we decided always to cut the internode sections only 2 mm below the node. The second question was: 'What length should we cut the sections?' Some workers use 3 mm sections, others prefer them 10 mm long. An 17 Natural auxins experiment designed to clear up this point gave the results shown in Figure 11. It was evident that, as the initial length of the section increased, the absolute amount of elongation increased. However, the growth of the controls also increased, so that the difference Giaa" ^o "^^Y ^'^^ benefit from longer initial lengths. This is the case for internodes {Figure 1 l{e)) in which the zone which responds to the addition of lAA seems more restricted than in the case of coleoptiles. If one looks for maximum absolute response to lAA over that of controls, then a 10 mm coleoptile section or a 5 mm internode section is better than shorter ones. But if one considers this response expressed as percentage of the initial length, then a 4 mm section gives about the largest Coleoptile First internode 15 20 30 35 V(7 15 Length of organ 30 Figure 12. Effect of the physiological age of the oat seedling {represented by the length of the coleop- tile or of the first internode at the time of sectioning) on the growth of the sections in buffer+ sucrose without lAA {white circles, dotted lines) and with lAA {black circles, solid lines), and on the difference between these growths {curves b and d). Scales at the left ordinate are in mm, on the right ordinate in per cent of the initial 4 mm length. diflferential response {Figure 11, curves c and/). To be able to use small volumes of solution and avoid the lack of straightness inherent to long sections, an initial 4 mm length was adopted for both coleoptile and internode sections. An important factor in the sensitivity of oat plants is their age. In general, the younger they are, the more they grow, but they do so also without added auxin. The optimal size at the time of sectioning is when the coleoptile and the mesocotyl have reached about respectively 20 and 15 mm in length {Figure 12). Longer internodes rapidly lose their sensitivity, whereas coleoptiles up to 35 mm in length (in the variety Brighton) still respond 18 Investigation of natural auxins and growth inhibitors relatively well to added lAA. In oiu- experiments, the coleoptiles were sectioned when they had reached about 25 mm in length and the internodes about 20 mm. If one wants to employ small volumes of test solutions, it is important to wash the sections before use. The length and the nature of this pretreatment was investigated. It was found that washing both coleoptiles and mesocotyls for one hour in glass-distilled water increased the lAA response and the reproducibility of the test results ( Table 6) . If the presoaking time is increased, the response to lAA decreases. It also decreases if the sections are immersed. Table 6 Effect of various pretreatments on the response to lAA Ten 4 mm sections were used in each case. The sections were ahvays breaking the surface of the sohitions unless otherwise indicated. Growth Plant part Pretreatment Time of Growth in GlAA — Go {oat) {hrs) controls {mm) lAA {mm) {mm) First internodes None 1-27 10}'/1. :2-95 1-68 HoO 1 1-09 10}'/1. :3-34 2-25 First internodes HjO, floating 3 1-30 10)'/!.: 3-25 1-95 H2O, immersed 3 l-77t 10}'/!.: 2-28 0-51 2% sucrose, floating 3 1-40 10y/l.:3-24 1-84 2% sucrose, immersed 3 1-58 10}'/1. :3-43 1-85 Coleoptiles None - 1-40 50)'/!. : 2-49 1-09 H.,0 1 1 -69 50}'/].: 3-38 1-69 Coleoptiles H,,0 5 1-88 50y/l. :2-72 0-84 H,0+ MnSO^HaO at 0- 1 mg/1. 5 3-00 50}'/!. : 3-98 0-98 H20^MnS04H20 at 1 mg/1. 5 2-54 5O7/I. : 4-64 2-10 HaO+MnSO^HjO at 10 mg/1. 5 1-93 50-/1. : 3-46 1-53 t Immersion in water increases somewhat the length of the controls, but a close examination shows that the sections become more slender than when they are floated. This decrease, however, can be prevented by the addition of sucrose (2 per cent). Immersion in the liquid presents no problem for coleoptiles which naturally float on the solutions, but, in the case of first internodes which are of solid stem tissue, it was necessary to lay them on stretched cheese-cloth to keep them breaking the surface of the solution. It was thought that the increase in response to lAA caused by presoaking in water might be due to the washing away of an lAA-destroying system. To test this idea, presoaking in reducing agents such as ascorbic acid, glutathione, Na2S204 was tried, but with no significant results. Nor did presoaking in Ca pantothenate, folic acid, arginine, or cobalt chloride (10-1,000 y/1) have any appreciable effect. On the contrary, manganese, which is well known to stimulate growth when used together with lAA, was found to increase the growth of coleoptiles but not of internodes when supplied in the pretreatment water at the concentration of 5-9 X 10~^ M. However, we did not add MnS04 to the solutions during the whole test period for fear that it might help to destroy labile substances present in the plant extracts. In summary, a 3-hour 19 Natural auxins soaking period in MnS04-|-H20 (1 mg/l.) was adopted for coleoptile sections and a 1-hour presoaking period in glass-distilled water for the inter- node sections, the latter being maintained breaking the surface on stretched cheese-cloth. In a given plant extract, the quantity of auxin is generally small. The problem is how to make the best use of it to get the maximum growth response. If we have, let us say, 0-01 y of lAA, is it better to dissolve it in 1 c.c. of assay solution, or in 10 c.c. ? It is known (cf. Schneider, 1938) that ^ ^ 1^0 1 u me c.c. M C 0-5 -K^- 0-5 1-0 — ■>--i-o Z-0 — ■— S-0 180 leo mo 120 r 100 -^ 80 GO Avena f/nst internode \\ I I \ \ \ L 'I I I 100 76 S ^ ■^ :^ 60 -c ■^ %. -o^ 2d o^ 50 --3 ^ '^ J 1 5 ' 2 3 % of M or OSS 5 1 Figure 14. Effect q/" sucrose concentration on the growth of sections in buffer without added I AA [white circles, dotted lines) and with lAA (black circles, solid lines), and on the difference between these growths (curves (b), (d), and (/))■ Scales are in mm (left ordinate) and in per cent of the initial ^r mm length (right ordinate). Figure 15. Effect of pH on the growth of oat sections in buffer A- sucrose without lAA (white circles, dotted lines) and with lAA (black circles, solid lines) , and on the differ- ence between the two (curves (b) and id)). Coleoptile First inter node ^ ^ I J (b) \ \ V ^ Con. "° J l_J _l I L 7 8 J L_l_ 361/. pH units 21 Natural auxins Similar investigations concerning the effect of pH and buffer concentration gave the results shown in Figures 15 and 16. Accordingly, a pH of 5-0 was used in all our experiments and a concentration of 10~^ M for KgHPO^ in the buffer was adopted. This is not the optimal concentration as far as the i ^■0- 3-0 I 80 10 Ai^ena coleoptile 0-1 (b) Sorjhum first infernode Alpena f/nsf inhnnode 10 0-1 1 10 0-1 Buffer concentration ^ fO'^ M Figure 16. Effect of concentration of citrate-K phosphate buffer at pH 5-0 in the presence of sucrose on the growth of 4 mm sections. The lAA concentration was 50 yjl. in the case of oat coleoptiles and 10 yjl. for sorghum and oat first internodes. The concentration given along the abscissa is that of the K2HPO4 component; the concentration of citric acid is one half of it. coleoptile sections are concerned, but it was feared that lowering the con- centration would diminish the buffering capacity to such an extent that certain compounds separated on the chromatograms of plant extracts would change the pH considerably. -^3-0 E-0 ^ 1-0 I ^1-0 Coleoptile First internode (a) lAASoJl. / ^ (1)) Cc) IAAiOy/l. Con. 15 18^21 Bf 39 15 18 ^y 39 Time Hrs Figure 17. Effect o/^length of incubation on growth of oat sections in buffer -\- sucrose without added lAA {white circles, dotted lines) and with lAA {black circles, solid lines), and on the difference between these growths {curves {b) and {d)). Before time 0, the coleoptile sections had been presoaked for 3-5 hours in MnS04 ■ H2O (1 mg\l.) and the internode sections for 1 hour in glass-distilled water. To exclude the effect of light during the measurements, 5 replicates of 10 sections each were started at the same time ; at each time interval, one of these dishes was measured and discarded. The relationship between elongation and time of growth in the solution was also investigated. By starting with a certain number of replicate dishes and measuring one of them at various time intervals, the curves o^ Figure 17 22 Investigation of natural auxins and growth inhibitors were obtained. They show that coleoptiles in 50 y/l. of lAA (phis buffer and sucrose) continue to grow for a long time after they have been put in the solution, whereas the first internodes in the auxin solution stop growing after 18 hours. A short incubation time has the great advantage of minimizing the danger of bacterial contamination in the solutions. Therefore, an incubation period of about 20 hours was adopted and was found adecjuate in subsequent work. We can now attempt to obtain a comparison between the response of coleoptiles and that of internodes to increasing concentrations of lAA. As 1 s 10 so too 300 -JOOO sooo Concenfraf/on of IA4(x5x70 ' 9 Y\) Figure 18. Growthcurvesofoatcoleoptile (C) and first internode {M ) sections in buffer^ sucrose+ Tiveen 80 (0-1 per cent) with various concentrations of I A A. The scale on the left ordinate gives the growth in mm over the initial 4 mm; the scale on the right ordinate gives it in per cent of the origiiial length. The scale giving the LAA concentration is logarithmic. The coleoptile sections had been floated 5 hours on MnS04 • HjO (1 mgjl.) and the internode sections 1 hour on HjO prior to being put in the auxin solutions. Figure 18 shows, the amplitude of the response is larger in internodes tlian in coleoptiles. A clearly detectable increase over the growth of the controls can be obtained with the first internodes at a concentration of 10~^ M lAA, or 1-75 }'/l. The lowest amount of lAA detectable is approximately 2 y/1. in 0-5 c.c, or one thousandth of a gamma. The lowest amount of lAA detect- able in the Avena curvature test, for which about 0-3 c.c. of agar has to be used in order to make 12 small blocks, is 10 y/1. in 0-3 c.c, or 3 X 10~^ y. The new internode test can thus be 3 times more sensitive than the Avetia curvature test. Moreover, as shown on Figure 18, its proportionality range is wider than that of the curvature test (about a hundred fold, from 2 to 200 y/1.), higher concentrations giving irregular results. The accuracy, however, is less than that obtained with the Avena test in which the curvature is directly propor- tional to the auxin concentration, whereas here, the elongation is pro- portional to the logarithm of this. 23 Natural auxins So far, we have studied the effect of lAA only. But, in extracts of plants there are other auxins, such as IAN or lAE. The effect of these substances on the growth of coleoptiles and internodes of Brighton oats under the conditions defined above (presoaking, buffer, 2 per cent sucrose, 0-5 ex. volumes, etc., plus 0-1 per cent Tween 80 to help dissolve substances which are poorly soluble in water), can be seen in Figures 19 and 20. On the whole. i 5-0 ~ f-0 I ^ 3-0 8-0 120 WO t 80 V 60 1 3000 Concentration of auxins (xSxIO V\) Figure 19. Growth curves of oat coleoptile sections {var. Brighton) in buffer '\- sucrose + Tween 80 (0-1 per cent) and various molar concentrations of I A A. IAN. and lAE. Pretreatment of sections and scales are as in Fimre 18. ^ 70- e-0 5-0 lAE h /lAN/ \ i^-O 3-0 2-0 1-0 m 160 110 m 100 80 60 VO Figure 20. Growth curves of oat internode sections in conditions similar to those of Figure 19. 1 3 10 30 100 300 JOOO 3000 Concentration of auxins (x5xf0'^V\) at equal molar concentrations, the three substances have about the same activity, especially at low concentrations. At higher concentrations, IAN and especially lAE tend to be more active than lAA, but the difference is not a ten-fold one such as can be obtained in the case of IAN with Victory oats under different experimental conditions. The following examples show how this more sensitive first internode test 24 Investigation of natural auxins and growth inhibitors can be used in the detection of auxins on paper chromatograms. The first example is taken from our work on fruits and concerns the auxins in the seeds of Tokay grapes. The ether extract (2 hours at 0°C) of 3 g (fresh weight) of the seeds of nearly ripe grapes was purified through hexane and Figure 21. Histograms showing the biological activity of 1 (7/( sections of chromatograms when incubated for about 20 hours at 25"C in the dark with 0-5 c.c. buffer -\- sucrose -{- Tween 80 {0- 1 per cent). The first histogram shows the activity of an ether extract of 1-5 g (fresh weight) of seeds from nearly ripe Tokay grapes {after purification through hexane and acetonitrile) when assayed with coleoptiles. The middle histogram shows the activity of the same extract when assayed with first inter- nodes. The bottom histogram shows the activity obtained with 18-0 g {fresh weight) of the same material when assayed with coleoptiles. The chromato- grams were run in hexane -\- water. e , III III 1 I I ni— TT Coleopflles[l-5(^) 4 20': 1 1 1 L ' 60 50 1-5': 1 1 1 - 1 1 1 W 10'- IAN\ \IAE\ •^ 1 1 1 30 20 0-5'- 1 ' First internoc/es(T5o^, ) 20-. Jl 1 ' - 50 t < 1 1 1 Coleoptiles (180 Q^) ii ; : ' 30 -^ 20 >: 100^ 3-5': i SO 30': ij^ 80 70 25': 20': 1 1 60 50 1-5 \ 1 1 1 1 1 - W 10': 1 1 1 - III - 30 20 0-&r . J ' ' 1 I 7 0-2 O-ii- 06 0-8 10 ^f- acetonitrile, and divided into two equal portions which were chromato- graphed in hexane (90) -f- water (10) iyj^)- One chromatogram (correspond- ing to 1 -5 g of grape seeds) was assayed with coleoptile sections, the other with first internodes. The results with coleoptiles (^Figure 21, top diagram) were inconclusive, except that a marked inhibition was obtained at the initial spot. The first internodes, on the other hand, gave a clear-cut result (middle diagram) with the same amount of grape extract, with a growth peak in the IAN and another in the lAE position. No inhibitor was detected, for first internodes are less sensitive to inhibitors than coleoptiles, their endogenous growth being less. The growth peaks detected with internodes can, however, also be detected with coleoptiles, but it is necessary to extract ten times more material (bottom diagram). Thus, this first example shows that the use of internodes instead of coleoptiles allows one to obtain a good picture of the auxins present in an extract with at least ten times less plant material. A second example of the use of the internode test can be found in the 25 Natural auxins field of plant tissue cultures. It is well known that certain tissues can be grown in vitro if the medium contains an auxin, whereas others can grow on sugar and mineral salts alone (see Gautheret, 1954). It has also been shown that the extract of the tissues which can grow without added auxin in the medium are active in the Avena curvature test (Kulescha, 1951), but it is not verv clear if the difference in the auxin requirement is due to an increased Figure 22. Histograms showing the bio- logical activity of chromatograms of tissue culture extracts run in hopropanol + water (80;20) and assayed with first internode sections. PTN = crude ether extract (3 hours at 0°C) of 6 g { fresh weight) of Partheno- cissus tricuspidata, normal tissue, grown in vitro for 4-5 months on a medium containing sucrose, mineral salts, and lAA (50 yjl.). After this length of time, the added lAA had been completely exhausted. PTH = crude ether extract (2-5 hours at "C) of 4-5 g {fresh weight) of Parthenocissus tricuspidata 'habituated' tissue, grown for 4-5 months on a medium containing only sucrose and mineral salts. PTG = crude ether extract (2 hours at 0°C) of 5 mg (fresh weight) of Partheno- cissus tricuspidata crown-gall tissue, grown for 2-5 months on a medium containing ordy sucrose and mineral salts. 08 0¥ 0-6 OS 10 R _* / endogenous synthesis (Henderson and Bonner, 1952), or to a reduced destruc- tion (Piatt, 1954). Using three types of tissues from the same species, Parthenocissus tricuspidata (Sied. and Zucc.) Planch., namely the 'normal' tissue (requiring added auxin), the tissue 'habituated to auxin' (derived from the normal strain, but capable of growing without added auxin), and the crown-gall tissue (growing without added auxin) isolated by Morel (1948), we have studied the auxins which can be detected after a 3-hour ether extraction at 0°C. About 5 mg (fresh weight) of tissue were generally used and the extracts were not purified. The chromatograms were assayed with 26 Investigation of natural auxins and growth inhibitors first internodes. When wopropanol (80) -|- water (20) was used as a solvent, a very interesting picture was obtained (^Figure 22). The normal tissues, which require added lAA to grow, did not contain appreciable quantities of lAA, whereas a growth peak coidd be detected at the lAA position in the extracts of the two other types of tissues. On the other hand, a very large growth promotion was detected in the zone of neutral auxins such as IAN and I AE in all cases. Thus the hypothesis was formed, that perhaps the normal tissues manufacture appreciable amounts of a substance like IAN, but are unable to convert it into lAA, whereas the other tissues can. Other tissues grown in vitro, such as the vascular parenchyma of the tuber of Jerusalem artichoke, are incapable of transforming IAN into lAA, as Figure 23. Effect of various concentrations of lAA, IAN, and lAE on the growth of small cylinders of the tuber of Jerusalem artichoke, var. Pi^dallu 17. The initial explants each weighed 22 rng {fresh weight); they were harvested after 2 1 days of growth on media containing sucrose, mineral salts, and the added auxin {sterilized by filtration). Note that concentrations are given on a molar basis. 70-'' 10'^ 10- Concenfration, M shown in Figure 23. Nevertheless, from the results of Figure 22 alone we cannot be certain that the growth peaks found in all cases in the IAN region are actually due to IAN, for the solvent used cannot separate IAN from other auxins such as lAE. Another set of chroma tograms was developed, therefore, in the hexane-water solvent {Figure 24), but unfortunately from a different batch of cultures. In all three cases, a marked growth promo- tion was detected in the IAN region, but in addition, another growth peak in the lAE region was also apparent. Thus, at this point, it is difficult to say if the hypothesis presented above is correct, for lAE is roughly as active, on a molar basis, as lAA in tissue culture, at least in the case of Jerusalem artichoke tissues. A last example will illustrate even better the usefulness of the first internode assay technique. It concerns the examination of a chromatogram obtained with the extract of about 150 apices of rhizomes of the fern Adiantum pedatum dissected under the binocular microscope by Professor Wetmore, so that only the top millimetre or less of tissue was cut off. Altogether, these meristems weighed about 300 mg (fresh weight). Upon dissection, each of them was immediately plunged into a mixture of 2/3 ether and 1/3 absolute ethanol (v/v) kept cold in brine. The total extract was chromatographed in uopropanol-ammonia-water (80; 10; 10) (v/v) and assayed with first 27 Natural auxins Figure 25. Histogram showing the biological activity of a chromatogram run in hopropanol + ammonia-]- water (80; 10; 10) of an extract of 300 mg {fresh weight) of apical meristems of Adiatum pedatum, assayed with oat first internodes. Figure 24. Histograms showing the biological activity of chromatograms of tissue culture extracts run in he x ane -\- water and assayed with first internode sections. PTN = extract with 2/3 ether+ 1/3 ethanol (3-5 hours at 0°C) of 4-5 g (fresh weight) of Parthenocissus tricuspidata normal tissue grown in vitro for 5 months on a medium containing sucrose, mineral salts, and lAA (50 yjL). PTH = extract with 2/3 ether + 1/3 ethanol (2 hours at 0°C) of A g (fresh weight) of Parthenocissus tricuspidata habituated' tissue, grown for 2 months on a medium containing ordy sucrose and mineral salts. PTG = extract with IjZ etker+lj'i ethanol (2 hours at O'C) of 3-75 g {fresh weight of Parthenocissus tricuspidata crown-gall tissue grown for three months on a medium containing only sucrose and mineral salts. internodes. Figure 25 gives the results which demonstrate well that a good picture of the auxins present in plant tissues can be obtained with a crude extract, and with as little as 300 mg fresh weight of material. THE CHEMICAL IDENTIFICATION OF AUXINS ON THE CHROMATOGRAMS From the results presented so far, it is clear that one can detect numerous biologically active substances on the chromatograms of plant extracts. This, however, does not conclude our task. We want to know also the chemical 28 Investigation of natural auxins and growth inhibitors identity of these substances. This is much more difficult because, for analytical investigations, the chemist generally needs larger quantities of material than the biologist. However, if we have a set of synthetic auxins, we can try to see if some of the natural compounds are identical with them. For example, we always run control chromatograms with synthetic lAA, IAN, and lAE. If a given substance in an extract runs at the exact location of, let us say, synthetic IAN, we have a first presumption that it could be IAN. If the same substance, chromatographed in a totally different solvent, runs again at the location of synthetic IAN, then the presumption is strengthened. This is not enough to identify a compound, however. We have attempted to identify compounds on the paper by determining their absorption spectra. on 0-12 Figure 26. Absorption spectrum in ultra-violet light of A y of lAA on unwashed Whatman J\fo. 1 paper observed in a Beckman spectrophoto- meter. The determination was made 20 minutes after spraying the chromato- gram with paraffin oil. I I OfO - aos 006- 00¥ 002 260 280 SOO 320 Wai^elen^fh 3f0 This can be done in two ways. First, if we know the location of the compound on the paper, we can try to obtain its absorption spectrum in the u.v. This cannot be done readily because the paper absorbs the ultra-violet light. To render the paper translucent, we have simply sprayed it with paraffin oil. The readings, taken 20 minutes after spraying, in a Beckman spectro- photometer. Model DU, gave the absorption curve oi Figure 26^ with 4 y of lAA. The fine details of the lAA absorption curve are lost, of course, owing to the paper (unwashed Whatman No. 1), but the general shape and the region of maximum absorption fit reasonably well with those given by lAA in solution. On the other hand, if we do not know where the substance is, but if it appears as a coloured spot upon spraying with a suitable reagent, then we can determine the absorption spectrum of the coloured product (in visible light, this time) and compare it with that of a synthetic auxin. The curves of Figure 27 have been obtained in this manner. The coloured product obtained after spraying 1 y of lAA with the ferric-perchloric acid reagent of Gordon and Weber (1951) (100 c.c. of 35 per cent perchloric acid+270 mg of FeCla . 6H2O+ 100 c.c. of 95 per cent ethanol) had, on the paper, an absorption maximum at 550 m/^ 45 minutes after spraying, while the maximum was at 580 m/i in the case of IAN and around 460 mfx for 29 Natural auxins lAE. In the latter case, a 'bump' in the curve around 550 m^ probably indicated the presence of lAA, liberated by hydrolysis under the influence of perchloric acid. 0-30 WO 600 600 VOO 500 600 Wavelength 700 900 500 600 Figure 27. Absorption spectra, in visible light, of the coloured products formed on unwashed Whatman No. 1 p'aper after spraying the spots of synthetic lAA, lAE, and IAN with the ferric-perchloride reagent described in the text. These spectra change with time. The absorption maxima can be used to determine quantitatively, and directly on the paper, amounts of lAA, IAN, and lAE as low as 0-5 y {Figure 28) . This procedure avoids the losses which occur when one tries to 0-300 0-250 '- lAA I _(j5(nin ,,5Wm^) ^ UiiUU X 0-150 § X 0-050 lAE \l00m\'n,130m}x) Z 18 3 1 Gammas of auxin Figure 28. Quantitative relationships between absorbency and amounts of lAA, lAE, and IAN when measured directly on the paper after spraying with the ferric-perchloride reagent. eluate a substance from the paper. It requires that the spot be kept small and that no interfering substances be present. Crude as these techniques may still be, they nevertheless can assist in the difficult task of identifying chemically the biologically active growth substances. In conclusion, it is hoped that this study of the extraction, chromatography, bio-assay, and chemical identification techniques may be a help to those engaged in the auxin field. REFERENCES BiTANCOURT, A. A., Schwartz, K., and Dierberger, R. (1954). La Nature des auxines des tumeurs vegetales. C. R. Soc. Biol, Paris, 148, 822. Boysen-Jensen, p. (1941). Quantitative Bestimmung der beschleunigenden Streckungswuchsstoffe in der sauren Fraktion der Atherextrakte aus hoheren Pflanzen. Planta, 31, 653. 30 Investigation of natural auxins and growth inhibitors Briggs, W. R., Morel, G., Steeves, T. A., Sussex, I. M., and Wetmore, R. H. ( 1955a). Enzymatic auxin inactivation by extracts of the fern, Osmimda cinnamomea L. Plant Physiol. 30, 143. Briggs, W. R., Steeves, T. A., Sussex, I. M., and Wetmore, R. H. (1955b). Comparison of auxin destruction by tissue extracts and intact tissues of the fern, Osmunda cinnamomea L. Plant Physiol. 30, 148. Gautheret, R. J. (1954). Catalogue des cultures de tissus vegetaux. Rev. gen. Bot. 61, 672. Gordon, S. A., and Weber, R. P. (1951). Colorimetric estimation of indoleacetic acid. Plant Physiol. 26, 192. GusTAFSON, F. (1941). The extraction of growth hormones from plants. Amer.J. Bot. 28, 947. Henderson, J. H. M., and Bonner, J. (1952). Auxin metabolism in normal and crown gall tissue of sunflower. Amer. J. Bot. 39, 444. Jones, L. R., and Riddick, J. A. (1952). Separation of organic insecticides from plant and animal tissues. Analyt. Chem. 24, 569. KuLESCHA, Z. (1951). Recherches sur I'elaboration de substance de croissance par les tissus vegetaux. Thesis, Paris. Larsen, P. (1955). On the separation of acidic and non-acidic auxins. Physiol. Plant. 8, 343. Morel, G. (1948). Recherches sur la cuhure associee de parasites obligatoires et de tissus vegetaux. Ann. Epiphyt. Ser. Path. Veg. 14 (N.S.), 123. NiTSCH, J. P. (1955). Free auxins and free tryptophane in the strawberry. Plant Physiol. 30, 33. NiTSCH, J. P., and Nitsch, C. (1955). The separation of natural plant growth substances by paper chromatography. Beitr. Biol. Pfl. 31, 387. Overbeek, J. VAN, Vazquez, E. S. de, and Gordon, S. A. (1947j. Free and bound auxin in the vegetative pineapple plant. Amer. J. Bot. 34, 266. Platt, R. S., Jr. (1954). The inactivation of auxin in normal and tumorous tissues. Annee biol. 30, 349. Schneider, C. L. (1938). The interdependence of auxin and sugar for growth. Amer. J. Bot. 25, 258. Schneider, C. L. (1941). The effect of red light on growth of the ylrwi« seedling with special reference to the first internode. Amer. J. Bot. 28, 878. Sen, S. p., and Leopold, A. C. (1954). Paper chromatography of plant growth regulators and allied compounds. Physiol. Plant. 7, 98. Steeves, T. A., Morel, G., and Wetmore, R. H. (1953) . A technique for preventing inactivation at the cut surface in auxin diffusion studies. Amer. J. Bot. 40, 534. Stowe, B. B., and Thimann, K. V. (1954). The paper chromatography of indole compounds and some indole containing auxins of plant tissues. Arch. Biochem. Biophys. 51, 499. Tang, Y. W., and Bonner, J. (1947). The enzymatic inactivation of indoleacetic acid. I. Some characteristics of the enzyme contained in pea seedlings. Arch. Biochem. 13, 1 1. Thimann, K. V., and Skoog, F. (1940); The extraction of auxin from plant tissues. Amer.J. Bot. 27, 951. Thimann, K. V., Skoog, F., and Bver, A. C. (1942). The extraction of auxin from plant tissues. II. Amer. J. Bot. 29, 598. Wagenknecht, a. C, and Burris, R. H. (1950). Indoleacetic acid inactivating enzymes from bean roots and pea seedlings. Arch. Biochem. 25, 30. Weij, H. G. van der (1932). Der Mechanismus des Wuchsstofftransportes. Rec. Trav. bot. neerl. 29, 379. Wildman, S. G., and Muir, R. M. (1949). Observations on the mechanism of auxin formation in plant tissues. Plant Physiol. 24, 84. 31 THE DISTRIBUTION OF NATURAL HORMONES IN GERMINATING SEEDS AND SEEDLING PLANTSj Phyllis M. Cartwright:]:, J. T. Sykes, and R. L, Wain Wye College, University of London Many investigators have reported on the occurrence of auxins and inhibitors in seeds. Germination inhibitors or 'blastocholines' are of widespread occurrence in fruits and seeds, and seem to be important in those which require special environmental conditions for germination. The purpose of the present investigation was to identify the natural auxins and inhibitors in maize and pea seeds and to follow their changes in concentration and distribution during germination and subsequent seedUng growth. It was hoped that by using paper chromatography all the active components could be isolated and their concentrations estimated by bio- logical and perhaps by chemical methods. Ether extracts were made from seeds at different stages of germination and these extracts were separated into acid and non-acid fractions by a method developed from that of Boysen- Jensen (1941). The components of the acid fraction were isolated by paper chromatography and identified by biological assay and chemical colour reactions. The non-acid fractions, prepared by a variety of methods, always contained large amounts of oily substances which interfered with the chroma- tography. Those examined gave unsatisfactory results and little biological activity was found, though in each case all regions of the paper were assayed. The importance of obtaining a comprehensive picture of hormone and inhibitor status in plant tissues is, of course, recognized, and has been stressed by a number of workers. However, the quantitative estimation of such activity in the neutral fraction of tissue extracts has not been successfuUy accomplished here or elsewhere. Because of these limitations, only the results from acid fractions will be considered in this paper, and in adopting this procedure the work is in line with similar investigations on natural auxins carried out by other workers. It will be realized that any hormones of a non-acidic nature, such as indolylacetonitrile, would be present mainly in the non-acid fractions. In carrying out the experiments, large numbers of seeds were germinated under uniform conditions, samples taken at suitable intervals of time, and their auxin and inhibitor content estimated. For the extraction procedure, the seeds were cut into slices and extracted at 2°C with 3 changes of wet peroxide-free ether for a total period of 24 hours. It is agreed with other workers that this does not give a maximum yield of auxin, and with maize, for example, further small amounts have been obtained by prolonged extrac- tions for 2 weeks. However, since the largest part of the auxin is extracted in the first 24 hours, it was decided to limit the analysis to this fraction in the first instance. For separation of acidic and non-acidic fractions, a technique t This paper was read at the conference by Phyllis M. Cartwright. + Present address: Department of Agricultural Botany, University of Reading. 32 Distribution of natural hormones developed from that of Boysen-Jensen (1941) and similar to that of Bennet- Clark and Kefford (1953) was used. The acidic components were taken out from the ether extract by shaking repeatedly with 1 per cent sodium bicarbonate solution. The combined aqueous layers were titrated to pH 4 with X sulphuric acid and re-extracted with ether. This ether extract was dried over sodium sulphate in a refrigerator overnight and then concentrated to small bulk in a water bath not exceeding 45°C. The concentrated acid fraction so obtained was spotted on to the starting line of a Whatman No. 1 chromatogram paper, usually in 3 small spots about half an inch apart. Extract spotted - here Storting line Mixture of authentic lAAand IAN spotted here 10 lAA reference spot IAN reference spot -Solvent front 0*- Ri r -f-O Figure. 1. Division of chromatogram for biological assay {see text). The chromatograms were developed in the descending manner in aqueous butanol/ammonia solvent (200 ml «-butanol : 6 ml 0-880 ammonia : 36 ml water) at a controlled temperature of 20°C for about 14 hours, by which time the solvent front had travelled about 10 inches. The paper was then removed from the tank and quickly dried by hanging in a fume cupboard. It was then cut transversely into 10 strips of equal size, representing R^ values of 0-0-1, 0- 1-0-2, etc. These will be referred to as strip 1, strip 2, etc. (see Figure 1). The biological activity of these strips was determined by a special modifica- tion of the wheat cyhnder test of Smith, Wain, and Wightman (1952). Each chromatogram strip was immersed in 3 ml 1 per cent sucrose solution in a small glass dish. Ten 5 mm segments from wheat coleoptiles, threaded on glass capillaries, were floated on top of the solution and the dishes placed in a water-saturated atmosphere at 25°C. The lengths of the coleoptile segments were measured after incubation for 24 hours. In all experiments a mixture of known indole compounds was run simultaneously as authentic spots and their positions were detected by spraying with Ehrlich rej^gent. Under these conditions lAA has RfO-\7 and runs in strip 2 of the chromatogram, and IAN has R^ 0-82 and runs mainly in strips 8 and 9. EXPERIMENTAL RESULTS The results from an experiment in which a solution containing 1 mg lAA and 1 mg IAN was separated into acid and non-acid fractions, both of which were assayed by the methods described above, are shown in Figure 2. All the lAA activity is in strip 2 of the chromatogram. The IAN activity is divided between the fractions, the larger part being in the non-acid fraction. Since equal amounts were taken, it is clear that IAN is several times as active as lAA in this particular test. 33 Natural auxins :^ zoo Si "I I 100 50 Acid fraction lAA IAN \ ^ 250 % -^^ ZOO ^ 150 Hon -acid fraction % 100 50 Figure _. uiuiugitui ujjuj uj chromalograms of acid and non- acid fractions from a solution con- taining 1 jig lAA and 1 /(g IAN. Experiments with maize A preliminary experiment was carried out with maize, variety 'White Horse Tooth.' Samples of 200 seeds were extracted and assayed at intervals of 12 hours during 60 hours germination at 25°C in the dark. The results are presented in Figure 3. It can be seen that most of the auxin extracted by this <^irs 1-0 2 Rf~ Figures. 'White Horse Tooth' maize. Preliminary experiment. procedure is in strip 2 of the chromatogram and it has been confirmed that this is indolylacetic acid by colour reactions and by measurements of 7?^ in different solvents. It seems that the amount of I AA extracted by this pro- cedure increases rapidly in the early stages ofgermination, reaches a maximum after about 24 hours, and then begins to decrease. 34 Distribution of natural hormones Further experiments were carried out in order to relate biological activity of lAA to the actual amounts present in the extracts. Although this auxin is concentrated mainly in strip 2 of the chromatogram, it occasionally spreads into strip 1 and strip 3. In these experiments, therefore, the chromatograms were divided up in a different manner. The zone of Rf 0-0-3 was cut longitudinally into 3 similar strips, which were assayed separately, and the lAA content of the sample was calculated from their mean value. By -using known amounts of lAA it was possible to construct a calibration graph in which the wheat cylinder extension growth, expressed as a percentage of the control growth, was shown to give a linear relationship with the log of the lAA concentration. Samples of germinating maize grains were taken at intervals of 12 hours as in the preliminary experiment and biological assays of I AA were carried out on extracts from different numbers of seeds at each sampling time. The results of these experiments are presented in Table 1 and Figure 4. These show Table 1 Extracts from whole seeds Time of Independent estimates of 95'^',, confidence germination, lAA content per 100 seeds. Mean Standard limits of hours y error real mean 3 18, 10, 30, 48, 32 27 6-5 27±17-9 12 75, 45. 32, 49, 63 53 7-4 53 + 19-8 24 233, 122,352,228, 159, 154 223 33-6 223 + 76-9 36 221, 185, 339, 336, 333, 300, 345 294 27-6 294 + 59-1 48 238. 107,96, 128 142 33-8 142 + 110-7 Changes in lAA content of seeds during germination Time interval, hours Change in lAA content, y per 100 seeds Least significant difference, to 5% 3-12 12-24 24-36 36-48 + 26 + 170 + 71t -151 22-9 77-5 127-0 100-1 t Not significant. a rapid increase in the lAA content of the seed during the first 24 hours of germination. The mean values obtained for 24 and 36 hour extracts are not significantly different but a significant decrease is indicated between 36 and 48 hours. The results so far reported deal with total changes in the whole seed. An experiment was next carried out in order to follow changes in the embryo and endosperm separately. At each sampling time, the seeds were separated into 35 Natural auxins 350 Figure 4. lAA content of maize grains at different stages of germination. 12 2t 36 Time of germination ¥8 hrj embryo (including scutellum) and endosperm and the lAA contents deter- mined independently. The results are summarized in Table 2. It is clear that the estimated lAA contents of the endosperm extracts are of the same order as those of the whole seed extracts, and there is practically no lAA in the endosperm and scutellum at any stage of germination. Table 2 Extracts from separated endosperms and embryos (a) Endosperms Time of germination, hours Independent estimates of lAA content, y per 1 00 endosperms Mean 12 15 24 36 48 18,22 95, 232, 58, 49 342, 657, 377, 197, 308, 123 143, 285, 462, 624, 300 84, 154, 50, 52, 360 108 334 363 140 {b) Embryos -\-scutellums Time of germination, Independent estimates oflAA content. hours y per 100 embryos -^scutellums 10,2,0,0 12 0, 13 15 0, 24, 7, 4 24 22,7 36 0, 5, 0, 6 48 0, 5, 7, 36 Distribution of natural hormones In chromatograms sprayed with Ehrlich reagent, an intense bkie spot due to lAA always developed in strip 2. Immediately behind this there were often two more faint and diffuse spots, a pinkish region of i?^ approx. 0-14 and a greenish blue one of R^ 0-10. It was difficult to separate tliese completely from the lAA but our results would suggest that they were either inactive or only weakly active in the wheat cylinder test. SO i t i 1 i Experiment A Feb. 1955 Experiment B May 1955 2\\n 2V\\rh 72\\r% -Rf i/alue- lAA zone Inhibitor zone 1-0 120 hn 168 hrs Figure 5. Biological assay of hormones and inhibitors in extracts from germinating peas, var. Alaska. {Acid fractions.) Experiments with peas Similar experiments have been carried out with pea seeds — a typical non-endospermic dicotyledon. In these experiments the variety 'Alaska' was used and the methods were similar to those described for maize. The results of two experiments with peas are presented in Figure 5. From the results the following observations can be made: (a) At all stages the lAA content is much lower than in maize. In the early stages of germination it is extremely low and it increases slowly as germination proceeds. 37 Natural auxins {b) There is at the earher sampHng times an inhibitor, an acidic substance with R^ 0-3-0-4, whose concentration decreases during the first 3 days of germination. None was detected in the 72- and 96-hour extracts. In the next few days of seedhng growth, however, this inhibitor re-appears. (c) In the 7-day-old etiolated pea seedling an appreciable amount of lAA has been built up, and there is also a significant concentration of inhibitor. (d) In many instances, there is some stimulatory activity in strips 7 and 8 of the chromatogram. This, however, does not appear consistently and it cannot with certainty be attributed to any known hormone. ^200 s> k^so May 1955 300 Shoots 300 Roofs 300 Cof/ledons k ^100 b°^ 50 ^ June 1955 200 Shoots 200 Roofs 200 Cotjiedons - Rf yalue •- Figure 6. Distribution of hormones and inhibitors in ^-day-old etiolated pea seedlings. A series of six experiments with peas indicated similar changes in auxin and inhibitor content during germination. There was, however, a considerable variability in the amounts of activity on different occasions and no attempt has been made to relate the results to absolute amounts of lAA present at each sampling time. The results of two experiments carried out to show the distribution of auxin and inhibitor in 8-day-old etiolated seedlings are presented in Figure 6. The shoots contain inhibitor and a relatively large amount of lAA. In the root extracts there is a lower auxin and inhibitor activity. Again, calculation of the absolute amounts of lAA in the tissues has not been attempted, but the histograms suggest that in these etiolated seedlings the shoot contains of the order of 10 times as much lAA as the root system. Since the fresh weight of the shoot is only about 25 per cent greater than that of the root, a higher internal I AA concentration is indicated. In the cotyledons there is stimulatory activity in the TAA zone' and in strips 7 and 8 of the chromatogram. Chromatograms similar to those used for biological assay were sprayed with Ehrlich reagent and besides the blue spot due to lAA a pink spot of Rf 0- 1 developed which has the same Rf and colour reactions as indole-3-carboxylic acid. The possible significance of this is discussed in other papers of this Symposium (Seeley el at., 1956; Fawcett et at., 1956). From these investigations on germinating peas it would seem that the disappearance of an inhibitor during the early stages of germination may be of physiological significance. This inhibitor has completely disappeared after 48 hours when the radicles have emerged. The auxin content, however, increases very slowly during germination proper and only builds up to an 38 Distribution of natural hormones appreciable concentration in the shoot of the young seedHng. It might there- fore be tentatively suggested that auxin has little physiological significance during the actual germination process but may be important in the subsequent growth of the seedling shoot. In maize, the auxin content of the seed is considerably higher, and there is a rapid increase in extractable auxin during the first 12 hours of germina- tion. After about 48 hours when the radicles emerge, the auxin content is decreasing. Our results, however, show that these changes are in the endo- sperm only and that there is practically no lAA in the embryo and scutellum at any stage of germination. The maize experiments provide no evidence of a growth inhibitor, although such a substance may be present in the non- acid fraction. Voss (1939) and Pohl and Tegethoff (1949) reported the presence of an 'auxin inactivator' in the scutellum of maize, and since in this investigation little auxin was found in the scutellum and embryo, it is possible that indolylacetic acid moving out of the endosperm into the scutellum is immediately inactivated by the scutellum 'Hemmstoff.' Skoog (1937) and Voss (1939) obtained evidence that auxin precursor or inactivated auxin was translocated to the coleoptile tip and was there re-converted to the auxin which controls extension growth in the coleoptile. On the basis of all results reported in this paper, it would seem justified to suggest, as a general hypothesis, that auxin extracted from seeds by the procedure we have adopted is not of physiological importance during germination. The growth of the seedling shoot, however, is clearly associated with auxin production. REFERENCES Bennet-Clark, T. a., and Kefford, N. P. (1953). Chromatography of the growth substances in plant extracts. Nature, 171, 645. Boysen-Jensen, p. (1941). Quantitative Bestimmung der beschleunigenden Streckungswuchsstoffe in der sauren Fraktion der Atherextrakte aus hoheren Pflanzen. Planta, 31, 653. Fawcett, C. H., Taylor, H. F., Wain, R. L., and Wightman, F. (1956). The degradation of certain phenoxy acids, amides, and nitriles within plant tissues. This volume, p. 187. PoHL, R., and Tegethoff, B. (1949). Der Hemmstoff des Maisscutellums — ein Wuchsstoffinaktivator. Naiurwissenschaften, 36, 319. Seeley, R. C, Fawcett, C. H., Wain, R. L., and Wightman, F. (1956). Chromato- graphic investigations on the metabolism of certain indole derivatives in plant tissues. This volume, p. 234. Skoog, F. (1937). A deseeded Arena test method for small amounts of auxin and auxin precursors. J. gen. Phvsiol. 20, 311. Smith, M. S., Wain, R. L., and Wightman, F. (1952). Studies on plant growth regulating substances. V. Steric factors in relation to mode of action of certain aryloxyalkylcarboxylic acids. Ann. appl. Biol. 39, 295. Voss, H. (1939). Nachweis des inaktiven WuchsstofTes, eines Wuchsstoffantagoniste und deren Wachsumsregulatorische. Planta, 30, 261. 39 HORMONES AND HORMONE PRECURSORS IN LEAVES, ROOTS, AND SEEDSt Joyce A. Bentley,J S. Housley, and G. Brixton Department of Botany, Manchester University This work is concerned with the naturally occurrinsr hormones and their precursors in several plant tissues. The substances studied have been located in the main by paper chromatography of plant extracts and bio-assay of the separated fractions for physiological activity. So far, workers have concerned themselves almost entirely with the acid ether-soluble constituents of plant extracts, and evidence has accumulated that there are several active sub- stances in this fraction. 3-Indolylacetic acid (lAA), 3-indolylacetonitrile (IAN), 3-indolylpyruvic acid (IPyA), accelerator-a, and an inhibitor have all been demonstrated or postulated by various workers on the evidence of zones of activity obtained at various R/s on paper chromatograms run under specified conditions. In this paper other fractions of plant extracts, in particular the aqueous, non-ether-soluble fraction, have been examined and have led to evidence of other active substances and hormone precursors. The presence of IAN in an acid ether fraction, reported in particular by Bennet-Clark and Kefford (1953) in their extracts oi Aegopodium rhizomes and potato tubers, was of interest to us, since we have obtained evidence of a neutral, ether-soluble substance with R^ and chromogenic reactions of the nitrile being released from a water-soluble precursor. An acidic water-soluble precursor of a neutral growth substance in cabbage has already been reported by Bonde (1953). In view of these reports it was decided to examine the aqueous fractions of our extracts. In addition, work on synthetic IPyA led us to reorganize our views on the role of this substance as a plant hormone and the possibility of its being identical with Bennet-Clark's accelerator-a. TECHNIQUES All commercially-supplied cheinicals were of the pvu^est grade obtainable. Anaesthetic ether was rendered and maintained peroxide-free by standing over sodium wire in darkness at 0°C. Glass-distilled water was used for preparing all solutions. Oats (var. Victory) of the 1951 harvest were used in the bio-assays. Extraction techniques (a) Cabbage — A single plant was collected in early October 1953, the inner etiolated leaves frozen and ground, and extracted with 95 per cent ethyl alcohol acidified to pH 3-2 with sulphuric acid for 40 hours at — 10°C. The pH of the tissue-free extract was raised to 5 with barium hydroxide, the extract filtered, and the alcohol removed at 35°C by distillation under f This paper was read at the conference by Joyce A. Bentley. J Present address: Marine Laboratory, Torry, Aberdeen. 40 Hormones and hormone precursors reduced pressure. The aqueous extract remaining was extracted with ether and the ethereal sohition separated into acid and neutral fractions with sodium bicarbonate. Some of the aqueous fraction was distilled under reduced pressure until a thick syrup remained. (b) Excised and seedling roots of tomato [var. Best-of-All) — Excised roots were obtained by growth of 10 mm clonal tips, derived from clonal sector cultures, in a modified White's medium for 10-12 days at 27°C. The method of culture has been descriljed by Street (1954). The average yield of material was 18 g fresh weight per 100 roots. Seedling roots were grown in sand in a greenhouse, watered daily with the culture medium (less sucrose, vitamins, and Fe2(S04)3), and harvested after 4-5 weeks. Harvested roots were stored at — 15°C until sufficient quantities had been collected. The extraction technicjue was essentially the same as for the cabbage, except that sometimes, where stated, ether was used for extraction instead of alcohol. {c) Maize roots and seeds — Seed of maize (var. Great White Horse Tooth) was germinated in sand and the roots and seeds collected when the coleoptiles were approximately 2 cm long. The material was well washed, frozen, and extracted with ether. Paper chromatography technique The descending method was used with Whatman No. 1 papers. Loading was usually in the form of a strip, and controls of lAA and IAN (approx. 40 //g) were run simultaneously with each extract. Chromatograms were developed in the dark at room temperature without prior equilibration, in either 4 vol. zVopropanol : 1 vol. 0-15 N ammonia, or «-butanol saturated with 1-5 N ammonia. They were developed for about 20 hours, by which time the solvent front normally reached a distance of 20-30 cm, dried in a current of air at room temperature for 20 min and examined in filtered u.v. light (2537 A transmitted) for fluorescence. Chromatograms were usually divided for bio-assay according to the fluorescence pattern. Each division was eluted in 12 ml water. Chromatographic solvents were removed in vacuo with air drawn through the solution at a bath temperature of 35-40°C. Spray reagents Chromogenic reactions were developed by drying the papers at 35°C, and then heating at 50-60°C for 1-5 min. The following sprays were used: ferric chloride/perchloric acid, 2 ml 0-05M FeClg plus 100 ml 5 per cent HCIO4; nitrous/nitric acid, 1 g KNOg in 200 ml HNO3 (sp. gr. 1-42 diluted XlO); /j-dimethylaminobenzaldehyde, 2 g in a mixture of 20 ml HCl (sp. gr. 1-18) and 80 ml absolute ethanol. Bio-assay techniques The method of Avena straight-growth assay has been reported (Bentley, 1950), and the technicjue described in detail (Bentley and Housley, 1954). 10 mm sections were cut from coleoptiles of either 1-4-1 -6 cm, 1-6-1 -8 cm, or 1-8-2-0 cm, floated on 10 ml test solution in 2 in. Petri dishes, and their length measured after approximately 1 7 hours. Ten replicates per treatment were used except when otherwise stated. V^ 41 Natural auxins The cress assay method was a modification of that used by Moewus (1949a and b). Seeds, placed 1 cm apart, were germinated in Petri dishes lined with wet filter paper. 24-28 hours later, 6 seedlings with roots of 5 mm length were placed into a 2 in. Petri dish containing 10 ml 1-5 per cent agar into which the test solution had been previously incorporated. Approxi- mately 1 7 hours later the root lengths were determined to the nearest mm by measuring on graph paper after removal from the agar medium. The assay was carried out at 27°C in phototropically-inactive light, and subsequent root growth was at 25°C in the dark. i" 77 1^ 13 10 - (a) lAA IAN 1 — 1 1 — 1 ~ ^ fc4 H,0^ ^ ^M^ ^^^^H 1 II 1 i 1 1 / i 1 1 1 2 3 ; : 1 1 1 ! Response to lAA (Tng/L) 0-2 0-f 0-e 0-8 18 n I w lAA 1-0 IAN V 10 \ <5> w?fi 18 3 [ZZ] Response to lAA (mg/l.) 0-2 0-f 0-e 0-8 1-0 Figure 1. Chromatography in ammoniacal isopropanol (a) and n-butanol (b) of the aqueous fraction of an extract of cabbage leaves. U.v. fluorescence on the chromatograms is shown below, (a) ^one 1 yellow pigment on blue background, 2 blue with faint purple, 3 bluish-purple, (b) ^one 1 light blue, 2 purple, 3 bluish-purple, 4 bluish-purple. RESULTS Aqueous fraction of cabbage Chromatography of the syrup (approximately 50 mg) obtained from distillation of the aqueous fraction was carried out initially in /jopropanol/ ammonia; a typical result is shown in Figure J (a). Peaks of growth promo- tion were obtained at the lAA and IAN zones; chromogenic sprays gave shades of yellow with ferric chloride/perchloric acid and with nitrous/nitric acid between R^ 01 and 0-5. Near the IAN region nitrous/nitric acid produced a pale mauve colour resembling that of IAN at low concentration. The demonstration of a growth promoter in the I AA zone but not giving 42 Hormones and hormone precursors a red chromogenic reaction typical of lAA, prompted the examination of the syrup in another solvent system, n-butanol/ammonia. The result {Figure l{b)) differed from that shown in Figure l{a) in that the first peak of activity no longer corresponded to lAA though activity again occurred in the IAN zone. Results with chromogenic sprays gave yellow at Rf 0-0-2 and mauve at IAN. These results, in which the pattern of growth promotion varied with the solvent system used, suggested that the growth-active zones at lAA [Figure l{a)) and at the starting line [Figure 1(b)) should be examined further to see whether they are identical. Two 28 cm chromatograms were each loaded with 200 mg syrup and chromatographed in wopropanol/ammonia. U. v. -fluorescent zones corre- sponding to 1 and 2 in Figure l{a) were removed and eluted at 5°C with 4 ml water. 0-75 ml of this eluate was pipetted on each of two papers, and chromatographed in ammoniacal z.s'opropanol or w-butanol (Figures 2(a) and (b)). The histogram of the ?.ropropanol chromatogram is essentially the same as that of Figure 1(a), while that in ?i-butanol differs but slightly from that of Figure 1(b). In Figure 2(b), the chromatogram division at Rf 0-0-08 was according to u.v. fluoresence, but it was too uneven to warrant the conclusion that two growth promoters had been separated. Before discussing this experiment further, data of the complementary experiment are presented. Two chromatogram papers were prepared as in the preceding experiment and chromatographed in «-butanol/ammonia. Two zones were removed, viz., the area of initial loading and the following zone to Rf 0-18 (corre- sponding to u.v. -fluorescent zones 1-3 in Figure 1(b)). These were eluted with 1 ml and 2 ml water respectively. 0-4 ml of each eluate was rechromato- graphed in ammoniacal isopropanol or «-butanol. The histogram of the starting line eluate run in uopropanol/ammonia (Figure 2(c)) resembles Figure 1(a), except for differences in relative position of IAN and the growth promoter near the solvent front. In «-butanol/ammonia (Figure 2(d)) the histogram is basically similar to Figure 1(b), but there is an ill-defined peak at Rf 0-8 and continuous tailing behind it. It is probable that growth pro- motion at Rf 0-06-0-19 does not actually overlap the lAA control and that division of the chromatogram has caused this effect. Histograms of the remaining eluate differ slightly from those described above. Chromatography in uopropanol/ammonia (Figure 2(e)) gave an active region at Rf 0-12-0-25, which may correspond with the Rf 0-0-02 zone on the «-butanol chromatogram (Figure 2(f)). Uneven chromatogram division, however, prevents one from stating this with certainty. The results shown in Figures 2(a)-(f) indicate that the substance at the lAA zone on /j^opropanol chromatograms is identical with that near the starting line on «-butanol chromatograms. Thus, differences in relative position of this substance and lAA result from chromatographic factors, and not from molecular change under the influence of these different solvent systems. From Figures 1(a) and (b), it is clear that chromatography in w-butanol gives better separation of active zones, which in this solvent may be complete. One may deduce from Figures 2(c)-(f) that promotion at the IAN zone originates from a precursor on and near the starting line of n-butanol chromatograms; in wopropanol this precursor occurs at the lAA zone and/or in a zone of smaller Rf range (see Figures 2(a) and (b)). 43 Natural auxins ^2C- ♦ 13- % (a) xhoPropancI/ ammonia 12 10 (t) '^-Bufanol/ammon/a lAA IAN Response to I AA ■rag/l.) 0-^ O-t OS 0-8 1-0 Response toIAA J I L _L J I I L 0-2 04 0-6 0-8 1-0 ^ 20 78 I 76 7V- u; 72 70 (c) mPropanol/ammonia lAA IAN Response foIAA (mg/U J I 1 L J I I I L n, -Bufanol/ammonia lAA IAN 0-2 OV 0-6 0-8 1-0 Response to I AA ^ _l(mg/L.) J I I I J 1 1 L_L 0-2 OV- OS OS 7-0 % ^ 76 7¥ 12 10 - ( e) mPropanoI/Ammonia - lAA IAN 1— 1 ■ 1 1 1 1 1 1 1 1 4. 1 1 1 (f) u -Butanol/ammonio lAA IAN Response to lAA o (lag/l) 0-2 04 0-6 0-8 IS Response toIAA o A. — * (lag/l.) _L J L X 0-2 04 OS OS 70 Figure 2. Chromatography of the aqueous fraction of an extract of cabbage leaves, (a), (b) Chromato- graphy was first carried out in ammoniacal isopropanol. The u. v. -fluorescent zones corresponding to 1 and 2 in Figure 1 (a) were eluted with water and rechromatographed in ammoniacal isopropanol (a) and n-butanol {b). {c)-{f) Chromatography was first carried out in ammoniacal n-butanol. The starting line was eluted in water and rechromatographed in ammoniacal isopropanol (c) and n-butanol (d). Following this zone, u.v. -fluorescent zones corresponding to 2 and 3 in Figure l{b) were treated in a similar manner {{e) and (f) respectively). 44 Hormones and hormone precursors Comparison of Figures 2{c), 2{e), and l{a) suggests that there is a growth promoter at Rf 0- 1-0-2, which is probably being formed from another substance, and the lack of the peak in Figure 2{a) is in agreement with this. It is also possible that unstable compounds are involved, and small differences in technique are influencing their reactions. Aqueous fraction of tomato roots The hormone content of ethereal extracts of excised and seedling roots is low. Attention was therefore turned to the aqueous fraction, which, like that of cabbage, showed high activity. 130 g fresh weight of excised roots were extracted with ether and the aqueous fraction prepared as in the cabbage experiments. Approximately ^ s /7 76 (a) //M I 1 ;/ HjO^c. HNO3/HNO2 Me Yelloiv Response to lAA (rng/l.) 0-2 04 0-e 0-8 1-0 'f- Figure 3{a). Aqueous fraction of approximately \'i g excised tomato roots developed in ammoniacal hopropanol. one-tenth of the aqueous fraction was chromatographed in /j^opropanol/ ammonia. On initial chromatography separation was poor, so the chromato- gram was eluted and redeveloped, when good separation was obtained. The results show [Figure 3(a)) that there is considerable activity in the aqueous fraction, with optimum stimulation at R^ 0-36 to 0-48 almost as great as the 1 mg/1. lAA control. The u.v. fluorescence at the position of optimum stimulation was deep purple tailing to pale blue. With HNO2/HNO3 a pale yellow colour was obtained at R^ 0-36-0-54, but no . colour appeared with FeCl3/HC104 or with /?-dimethvlaminobenzaldehvde (MeAB). A further one-tenth of the aqueous fraction was chromatographed and gave the same fluorescence as in the previous experiment. The zones of Rf 0-32 to 0-76, corresponding to the stimulatory zone of the previous experiment, were isolated, and re-eluted in a small quantity of water, which was then reduced to 2 ml by distillation under reduced pressure. This extract was chromatographed in /Vopropanol/ammonia, examined in u.v. light and cut into 2-5 cm strips for bio-assay. The results are shown in Figure 3{b). The greatest activity, in the region R^ 0-72 to 0-84, was almost equal to that of 1-0 mg/1. lAA and corresponded in position with the IAN control, though it could not be associated with any fluorescence. Clearly, 45 Natural auxins these experiments show that an active substance which runs at Rf 0-4-0-5 in wopropanol/ammonia Uberates a further active substance which runs at the same R^ as IAN on further chromatography. The re-chromatography involves heating to about 60°C and further treatment with weak alkali. 78 -<5i 1 H n . lAA 1 1 Z4/V 1 1 ^^ 1 ^ ^ r-^ 1 1 1 1 7£ "i!- 1 ' Response to lAA (mgAJ 0-2 O-f OB 0-3 7-0 'f Figure 3{b). Active area from chromatogram of aqueous fraction of approximately 13 g excised roots re-chromatographed in ammoniacal isopropanol. The experiment was repeated to examine the chromatogram with colour sprays. No colour reactions could be obtained in the IAN region with HNO2/HNO3, FeCl3/HC104, or with MeAB. There is thus no evidence, apart from R^ value, that the activity near the solvent front is due to IAN. 78 16 ^ -¥ ■^ 7f I 72 - - v^ 1 X 2 Y 1 H,n . ^ ^ llgU - ■ 1 1 1 I 1 Response fo lAA (mg/l.: 0-2 04 0-e 70 Figure 4. Aqueous fraction of approximately 13 g excised tomato roots developed in ammoniacal n-butanol. On spraying with bromo-cresol green, a yellow area in the zone of activity corresponding to the lAA region was obtained, suggesting that the second active substance near the solvent front comes from an acidic precursor. Because of the poor separation in ijopropanol/ammonia, the experiment was repeated using n-butanol/ammonia {Figure 4). Again separation was poor, so the chromatogram was eluted and re-developed, when a good 46 Hormones and hormone precursors separation was achieved. Peaks of activity were obtained at R. 0- 1-0-3 (zone X), Rf 04-0-6 (zone Y), and R^ 0- 7-0-9 (zone Z). Stimulation at Z was greater than with 1 -0 mg/1. lAA. Stimulation at X, which closelv followed the lAA control, could be associated with a pale blue and purple fluorescence. Zone Y had a pale pink fluorescence and zone Z no fluores- cence. With the colour sprays no reactions similar to those of lAA and IAN could be obtained. This experiment was repeated several times, with the same results. It was decided to try to isolate these three peaks to study their behaviour and to determine any relationships between them. 76 n lAA 1 — 1 (a) .^_ IAN 1 — 1 III -_J-y- 1 11 \ i 1 1 1 1 Response to lAA (mg/l.) t I rf! ^ g n IS lAA IAN (b) H.O — J L J L Response foIAA (mg/V) 16 - lAA 1 1 ic) IAN 1 — 1 n - r- -^Y . 1 r — . YT H30-^L_ 1 r 1 1 1 t 1 1 1 1 1 J Response to lAA (rag/l.) OS 0-f 0-6 0-8 1-0 Figure 5. [a) Aqueous fraction of excised roots developed in ammoniacal n-butanol. (b) ^otie X of aqueous fraction developed in ammoniacal n-butanol. (c) ^one Y of aqueous fraction developed in ammoniacal n-butanol. Two further chromatograms, each loaded with one-tenth of the aqueous fraction, were developed in the same tank. One chromatogram was bio- assayed to locate zones X, Y, and Z (Figure 5(a)). The zones of the second chromatogram corresponding to A' and Y were re-chromatographed in «-butanol/ammonia (Figures 5(b) and (c) respectively). In each case the running of zone A' or zone Y had produced three peaks. A pale purple fluorescence was obtained at zone Z in each chromatogram. It is possible that separation of A^ and Y may not have been complete before re-chromatography. However, it is clear that Z was formed from either X or Y , though it is not possible to state which of the two zones was responsible. It would appear that A' and Y are interconvertible, behaving like tautomers 47 Natural auxins of a substance which is the precursor of Z. This precursor is either active itself or is converted to an active substance by the action of the coleoptile tissues during bio-assay. Zone Z was eluted from Figure 5{a) and attempts were made to extract it with ether. It remained almost entirely in the aqueous fraction, however, and this result, together with the lack of reaction with the colour sprays, is evidence against Z being IAN. Further experiments similar to the ones reported above were carried out with seedling roots, and essentially the same results were obtained, with two peaks of activity in ^jopropanol/ammonia and three peaks, X, Y , and Z, in i" >< 72 lAA IAN ^ ¥- Response to lAA (mg/l.) 0-2 o-v- 0-e OS 10 'f Figure 6. Aqueous fraction of germinating maize seeds developed in ammoniacal isopropanol. w-butanol/ammonia. This fact is important, in view of the large amount of work now performed on excised tissues, and in particular on excised roots. If the results of work on excised tissues are to be of any real value in plant physiology, then we must have a comparison between excised and intact tissues. Aqueous fraction of maize roots and seeds In these experiments there was again considerable activity in the aqueous fractions, just as was found in both cabbage and tomato. The maize roots showed a pattern of hormone activity similar in many respects to that obtained with tomato roots, in both ammoniacal uopropanol and ammoniacal ;?-butanol. Full details of these results will be presented elsewhere as they do not contribute anything further to the present discussion. Chromatography of the aqueous extract from maize seeds, however, is worth considering in detail at this stage as it led to an investigation of 3-indolylpyruvic acid (IPyA), a hormone reported to be present in maize seeds (Stowe and Thimann, 1953). The results of chromatography of an aqueous extract are shown in Figure 6. Again there is considerable activity in the lAA region and some near the solvent front, but the region we wish to consider in particular is the zone of activity approximately half-way between the starting line and the lAA zone. This corresponds to the zone of activity found by Stowe and Thimann in extracts of dormant maize seeds, and identified by them as IPyA. Slight activity in a similar position was found in cabbage extracts in the aqueous 48 Hormones and hormone precursors fraction (e.g. Figures l{a), 2{c), 2{e)) and also in the acid and neutral ethereal fractions (results of these fractions are published in detail elsewhere). It was therefore decided to examine synthetic IPyA to determine whether, in fact, it could be responsible for the activity in this region. 3-Indolylpyriivic acid IPyA has been synthesized and fully characterized by Dr. G. F. Smith, Dr. K. R. Farrar, and Mr. W. Taylor of the Chemistry Department, Manchester University. Its behaviour under the conditions of chromato- graphy used by Stowe and Thimann to demonstrate its presence in maize endosperms, and also its growth-promoting activity, have been investigated. In order to permit a detailed account of this work to appear elsewhere (Bentley, Farrar, Housley, Smith, and Taylor, 1956), the remainder of the lecture which dealt with the results of these investigations is not reported here. It may be briefly stated, however, that work on synthetic IPyA has shown that it breaks down completely to form lAA, together with other substances, under the conditions of ammoniacal chromotography used in the studies of Stowe and Thimann. The presence of IPyA itself cannot be demonstrated using the solvent system employed by these authors. Chromogenic studies of chromatograms of the aqueous extract of maize {Figure 6) have shown that lAA is not present on these chromatograms. It appears improbable, therefore, that the activity near the starting line in Figure 6 is due to IPyA, suggesting that there is yet another unknown auxin in this region. It has also been suggested by Stowe and Thimann that the ether-soluble acidic substance called accelerator-a (Bennet-Clark and Kefford, 1953) which runs between lAA and the starting line under conditions of ammoniacal chromatography is probably IPyA. Bennet-Clark and Kefford have demonstrated that accelerator-a promotes root-growth, whereas we find that synthetic IPyA causes only inhibition of root-growth (cress assay method), with an activity approximately one-tenth that of lAA. The zone of activity between the starting line and lAA in Figure 6 has not yet been tested for its effect on root growth. DISCUSSION Examination of the aqueovis, ether-insohible portion of cabbage, tomato, and maize extracts has shown that this fraction has auxin activity which is usually considerably greater than that obtained in ether extracts. Only low activity is obtained in acid ether extracts, and this largely follows the same pattern as the aqvieous extracts. Two clear-cut peaks are obtained in cabbage with both ammoniacal /^opropanol and «-bvitanol ; in tomato, there are three clear-cut peaks with ammoniacal «-butanol. The growth promoter near the solvent front in cabbage is neutral, ether- soluble, and has the Rp fluorescence, and chromogenic reactions of IAN, with which it is believed to be identical. On the other hand, the hormone near the solvent front in tomato has been similarly examined, and, apart from Rf value, no convincing evidence has been obtained that this zone is due to the nitrile. It has not been possible to obtain even a transient reaction with HNO2/HNO3 at this position. 49 Natural auxins Clearly, both these zones arise from precursors which travel on the chro- matograms nearer to the starting line. In tomato there is evidence of two precursors which are interconvertible {Figure 5). It is not possible to say from the evidence so far obtained whether these precursors are themselves physiologically active or whether they are converted to an active substance or substances during bio-assay. The first step to answering this question is obviously chemical identification. Earlier workers have constructed theories (e.g. the Cholodny-Went theory) to explain the growth of roots in terms of lAA. It is evident that the hormone situation in roots is far more complex than had hitherto been suspected. Obviously, more work needs to be done on the inter-relationships between the water-soluble hormones and hormone precursors demonstrated in the present work, and on their chemical identity, before it is possible to construct satisfactory theories of hormone-controlled root growth. Yet another zone of activity nearer to the starting line than lAA has been demonstrated in maize seeds, both in the aqueous fraction in the present work, and in the ethereal fraction by earlier workers. It has been suggested that activity in this region of the chromatogram is due to IPyA. Work on synthetic IPyA has shown that it breaks down to form lAA, together with other substances, tmder the conditions of chromatography used in these studies. On chromatograms of maize extracts, FeCIg/HClO^ gives a yellow colour at the position of I AA while other chromogenic reagents clearly show that physiological activity at this position cannot be due to lAA. Therefore, quite apart from the fact that IPyA does not survive under conditions of ammoniacal chromatography, it cannot be responsible for the activity near the starting line in the maize extracts, as there is no production of lAA. This zone is therefore due to yet another unidentified auxin. We regard the presence of IPyA in plants as yet unproved, but even if it does exist, it cannot be identical with the accelerator-a of Bennet-Clark and Kefford (1953) as postulated by Stowe and Thimann (1953): the former authors have shown that a promotes root growth, whereas the present work shows that IPyA causes only inhibition. SUMMARY 1 . A report in the literature indicates that an acidic precursor of a neutral growth substance exists in cabbage. An alcoholic extract of mature cabbage has been prepared, and from this an aqueous, ether-insoluble fraction has been obtained. This fraction has been examined by chromatography in ii^opropanol/ammonia and n-butanol/ammonia. It contains a substance which is physiologically active on bio-assay with Avena coleoptiles. This substance runs differently relative to 3-indolylacetic acid (lAA) controls in the two solvent systems used, and gives a yellow colour with ferric chloride/ perchloric acid and nitrous/nitric acid sprays. It gives rise to a further substance by the action of mild alkali, including ammoniacal chromato- graphy, and heat. This second substance is believed to be 3-indolylacetonitrile (IAN). 2. Aqueous fractions of seedling and excised tomato roots and of maize roots gave a similar result, except that no evidence could be obtained that the substance liberated from the precursor was IAN. The precursor could 50 Hormones and hormone precursors be resolved into two zones of activity when chromatographed in n-butanol/ ammonia. These zones appear to be interconvertible. 3. A further zone of activity was obtained between the starting line and lAA controls in aqueous fractions of maize seeds. Synthetic 3-indolyl- pyruvic acid (IPyA) was examined to determine whether it could be responsible for this activity, but results showed that it appears to be destroyed under conditions of ammoniacal chromatography and cannot therefore be demonstrated in plant extracts by this technique. From an examination of its effect on root growth, IPyA cannot be identical with the accelerator-a of Bennet-Clark and Kefford. ACKNOWLEDGEMENTS The authors wish to thank Professor S. C. Harland, F.R.S., and Professor E. R. H. Jones, F.R.S., for facilities for this work in the Departments of Botany and Chemistry, the Agricultural Research Council for grants which hav'e assisted the botanical work, and Mrs. V. Shaw for technical assistance. The work described in this lecture will be reported in greater detail in the Journal of Experimental Botany (cabbage and tomato) and the Biochemical Journal (IPyA). We are indebted to Professor H. E. Street for sugsfesting the examination of hormones in excised roots and for providing help and facilities for growing this material. REFERENCES Bennet-Clark, T. A., and Kefford, N. P. (1953). Chromatography of the growth substances in plant extracts. Nature, 171, 645. Bentley, J. A. (1950). An examination of a method of auxin assay using the growth of isolated sections oi Avena coleoptiles in test solutions. J. exp. Bot. 1, 201. Bentley, J. A., Farrar, K.R., Housley, S., Smith, G. F., and Taylor, W. (1956). Some chemical and physiological properties of indole-3-pyruvic acid. Biochem. J. (in the press). Bentley, J. A., and Housley, S. (1954). Bio-assay of plant growth hormones. Physiol. Plant. 7, 405. Bonde, E. K. (1953). Auxins and auxin precursors in acid and nonacidic fractions of plant extracts. Bot. Gaz. 115, 1. MoEwus, F. (1949a). Die Wirkung von Wuchs- und Hemmstoffen auf die Kresse- wurzel. Biol. Zbl- 68, 58. MoEwus, F. (1949b). Der Kressewurzeltest, ein neuer quantitativer Wuchsstofftest. Biol. Zbl. 68, 118. Street, H. E. (1954). Growing roots without plants. Discovery, 15, 7. Stowe, B. B., and Thimann, K. V. (1953). Indolepyruvic acid in maize. Mature, 172, 764. Stowe, B. B., and Thimann, K. V. (1954). The paper chromatography of indole compounds and some indole-containing auxins of plant tissues. Arch. Biochem. Biophys. 51, 499. 51 CHROMATOGRAPHISCHE UNTERSUCHUNGEN UBER DIE WUCHSSTOFFE UND HEMMSTOFFE DER HAFERKOLEOPTILEt H. SoDiNG und Edith Raadts Staatsinstitut fiir Allgemeine Botanik, Hamburg Die chemische Natur des WuchsstofFes der Haferkoleoptile ist noch nicht restlos klar. Als sicher erscheint, daB es sich um einen indolartigen Wuchsstoff handelt, Indolessigsaure (lES) oder einen nahe verwandten Stoff. Wildman und Bonner (1948) konnten mit der Salkowskireaktion den extrahierbaren Wuchsstoff der Koleoptile quantitativ bestimmen und hielten ihn ftir lES. Zum selben Ergebnis kommt Reinert (1950) auf Grund der iibereinstimmenden Wirkungskurven und der Empfindlichkeit gegen Erbsenenzym. In eigenen Untersuchungen erwies sich der aus der Spitze der Koleoptile dijfimdierende Wuchsstoff als saureempfindlich und laugenstabil in Ubereinstimmung mit lES. Aus weiteren Befunden schlossen wir aber, daB es sich jedenfalls nicht um reine und freie lES handeln konne. Im Maisendosperm konnten auBer der lES verschiedene andere Indol- derivate gefunden werden. Stowe und Thimann (1954) trennten chromato- graphisch vier Stoffe, lES, Indolbrenztraubensaure, und zwei weitere noch nicht identifizierte Wuchsstoffe, wahrend Yamaki und Nakamura (1952) im Maisendosperm lES und Indolacetaldehyd fanden. Es scheint jedoch, daB ein Unterschied im Gehalt an Wuchsstoffen bei den einzelnen Maissorten besteht. lES konnte in alien Extrakten als einziger gemeinsamer Wuchsstoff festgestellt werden. Bei der Chromatographic des Haferkoleoptilenwuchsstoffes fand Terpstra (1953) nur lES. Bei Extrakten von zerriebenen und gefrorenen Spitzen treten aber im Chromatogramm Schwanze auf. Diese sollen durch einen Komplex entstehen, der auf dem Papier langsam lES freigibt. Im Dijfusat aus intakten Koleoptilenspitzen war nur freie lES enthalten ohne Schwanz- bildung im Chromatogramm. Unsere eigenen Versuche wurden ausschlicBlich mit Diffusionswuchsstoff gemacht. Im Gegensatz zu Terpstra (1953) erhielten wir aber bei der Chromatographic zwei Wuchsstoffe (Haferkriimmungstest, Zylindertest). Der cine stets anzutreffende Wuchsstoff ist lES, der zweite Wuchsstoff ist eine sehr labile Substanz, die nicht in alien Fallen nachweisbar war; alle Versuche, durch Anderung der Methodik die Ausbeute zu erhohen, waren vergeblich (Diffusion oder Extraktion mit Alkohol oder heiBem \Vasser, Abfangen in gepufferter Losung, Einengen des Wuchsstoffes oder Abdampfen bei 100°C oder niedriger, verschiedene Chromatographierfliissigkeiten, Anzucht der Pflanzen bei Licht oder Dunkelheit). Dieser Wuchsstoff konnte noch nicht identifiziert werden. Er ist aber nach dem R^AMert in 70 prozent. Athanol nicht identisch mit Indolbrenztraubensaure. Wenn dieser zweite Wuchsstoff fehlte, so konnte meist durch Auswaschen ■f This paper was read at the Conference by H. Soding. 52 Wuchsstoffe und Hemmstoffe der Haferkoleoptile der betreffenden Stelle eiii Eluat gewonnen werden, das nach Stehen iiber Nacht bei pH 3 einen WuchsstofFenthielt, der mit Chloroform ausgeschtittelt werden konnte und im Haferkriimmungstest wirksam war. Wenn das Eluat aber selbst eine geringe Wirksamkeit hatte, wurde diese durch Saurebehandlung und Ausschiitteln mit Chloroform verstarkt. (In diesen Versuchen war 70 prozent. Athanol als Chromatographierflussigkeit benutzt worden.) Nach diesen Ergebnissen ist also in der Pflanze eine inaktive WuchsstofT- vorstufe vorhanden, die bei Saurebehandlung leicht in einen aktiven Wuchsstoff iibergeht. Wie es aber kommt, daB dieser Ubergang in der Pflanze im Zylindertest, selbst bei 20-stundiger Testzeit, nicht geschieht, konnen wir nicht sagen. Vielleicht handelt es sich hier um einen ahnlichen Voi'gang, wie ihn Larsen (1944) bei seinem neutralen Wuchsstoff aus Erbsen beschixibt, der teils in einer sofort wirksamen, teils in einer in- aktiven Form extrahiert wurde. Auch Bonde (1953) gibt an, daB in der neutralen Fraktion von Erbsenstengeln und im Erbsensaft eine inaktive Vorstufe erst durch Behandeln mit Koleoptilensaft aktiviert und im Hafertest nachweisbar wird. Es ist also auBer der lES noch ein zweiter aktiver Wuchsstoff oder wenigstens seine Vorstufe im Spitzendiffusat vorhanden. Ein weiterer Hinweis auf einen inaktiven Wuchsstoff im Spitzendiffusat ergibt sich aus folgenden Versuchen. Nach Alkalisieren (pH 8) laBt sich aus dem DifTusat kein im Hafertest wirksamer Wuchsstoff ausschiitteln. Chromatographiert man aber diese ausgeschiittelte Substanz, so erhalt man wieder lES vmd den zweiten W^uchsstofT. Im SpitzendifTusat muB also ein nichtsaurer inaktiver Wuchsstoff vorhanden sein, der beim Chromato- graphieren aktiviert wird. Ob diese nichtsaure Vorstufe identisch ist mit dem inaktiven WuchsstofT, der durch Saurebehandlimg aktiviert wird, konnten wir noch nicht feststellen. Es gelang uns jedenfalls nicht, den aus der alkalischen Losung ausgeschiittelten Stoff durch Saure zu aktivieren. Auch konnten wir nach Saurebehandlung keine groBere Aktivitat im Gesamtdiffusionswuchsstoff feststellen. Allerdings miissen diese Versuche noch wiederholt werden. Wir finden also in unseren Versuchen folgende Stoffe : 1. lES, 2. den „2. Wuchsstoff", 3. eine inaktive Vorstufe, die durch Saurebehandlung leicht aktivierbar ist, aber mit dem Zylindertest auch bei langer Reaktionszeit nicht nach- weisbar ist, 4. eine nichtsaure Vorstufe, die aus alkalischer Losung ausgeschiittelt werden kann und beim Chromatographieren die beiden aktiven Wuchsstoffe liefert,f 5. zwei Hemmstoffe. Es ist noch nicht restlos klar, ob die beiden inaktiven WuchstofTe vielleicht doch identisch sind, und wie die genetischen Beziehungen zwischen diesen Stoffen sind. Alle papierchromatographischen Ergebnisse sind also mit grosser Vorsicht t Weitere Versuche machen es zweifelhaft, ob der 2. Wuchsstoff des Rohdiffusates und der 2. Wuchsstoff der alkalischen Ausschiittelung identisch sind. 53 Natural auxins zu beurteilen. Noch klarer zeigt das die Papierchromatographie des aus dekapitierten Stumpfen abzufangenden aiifsteigenden Wuchsstoffstromes. Nach zahlreichen alteren Versuchen handelt es sich bei diesem Wuchsstoff um einen im Haferkriimmungstest unwirksamen inaktiven Stoff (precursor oder WuchsstofTvorstufe), der iu der Koleoptilenspitze in den absteigenden Wuchsstoff umgewandelt wird. Der inaktive Wuchsstoff laBt sich aus saurer und aus alkaUscher Losung ausschtittehi. Er muB also eine neutrale Substanz sein (Raadts, 1952). Von Koleoptilzylindern wird er im Gegensatz zu dem durch Saurebehandlung aktivierbaren inaktiven Spitzenwuchsstoff leicht aktiviert (Zylindertest) ; beides miissen also verschiedene inaktive Wuchsstoffe sein. Auch ist es sehr zweifelhaft, ob er mit der nichtsauren Vorstufe identisch ist, die bei alkalischer Reaktion aus Spitzendiffusat ausgeschtittelt werden kann. Die Papierchromatographie ergab, mit oder ohne vorhergehendes Ausschiitteln (mit Chloroform), bishev nur lES als Wuchsstoff. Durch die Manipulationen muB sich also der aufsteigende inaktive Wuchsstoff in lES umgewandelt haben. Weiter findet man noch einen Hemmstoff, der vielleicht mit dem 1 . Hemmstoff des Spitzendiffusates identisch ist. Der aufsteigende Wuchsstoff enthalt also : 1. einen neutralen inaktiven Wuchsstoff, der in vitro und in vivo sehr leicht in lES umgewandelt werden kann; 2. einen Hemmstoff. Der Deutschen Forschungsgemeinschaft und der Joachim-Junguis- Gesellschaft in Hamburg danken wir fur die Untersttitzung dieser Unter- suchungen. SUMMARY 1. The growth substances occurring in Avena coleoptiles were studied, using a chromatographic technique. Table 1 summarizes the experiments carried out and the results obtained. Table 1 Pari of coleoptile studied Method of extraction Substances found to occur on paper chromatogram prepared from extract {Assay Methods, Avena Curvature Test, Avena Section Test) Apices 1 . Diffusion into water 1. 2. 3. 4. lAA A second growth-promoting substance Two growth-inhibiting substances One inactive precursor (becomes active when treated with acid) 2. Diffusion into water, with chloroform at pH shaken 8 1. 2. lAA A second growth-promoting substance (other substances not tested) Coleoptile bases + seeds 1. Diffusion from base into water 1. 2. lAA One growth-inhibiting substance 54 Wuchsstoffe und Hemmstoffe der Haferkoleoptile 2. The water difTusate from the apices contained the active substance indoleacetic acid, a second growth-promoting substance and two growth- inhibiting substances. Table 2 shows the i^^^-vahies of these substances (descending method). An inactive precursor was also found in this diffusate. This became active when treated with acid. Table 2 Second First Second Solvent mixtures lAA growth- promoting growth- inhibiting growth- inhibiting substance substance substance Water-saturated buryl alcohol + ammonia 0-36 0-56 0-82 — woPropyl alcohol + water + ammonia. 80 : 1 5 :5 047 0-25 ~0-90 ~0-07 Ethyl alcohol + water + ammonia, 70:30(25) :(5) 0-74 0-14 0-94 — Water+ ammonia, 95:5 ~0-86 ~0-50 — — Proportions: v/v. Ammonia = 25 per cent NHj. 3. The diffusate, when made alkaUne and shaken with chloroform, yielded a possibly different inactive precursor. During the development of the chromatogram this gave rise to the same two growth-promoting substances as were obtained in the water diffusate. f 4. The chromatogram developed from the inactive growth substances moving from the seed to the apex of the coleoptile showed two active substances : indoleacetic acid and a growth-inhibiting substance. The lAA must be formed during the chromatographic preparations. LITERATUR BoNDE, E. K. (1953). Auxin and auxin precursors in acid and nonacid fractions of plant extracts. Bot. Gaz. 115, 1. Larsen, p. (1944). 3-indole acetaldehyde as a growth hormone in higher plants. Dansk bot. Ark. 11, 1. Raadts, E. (1952). tJber den inaktiven Wuchsstoff der Haferkoleoptile. Platita, 40,419. Reinert, J. (1950). tJber den Wuchstoffgehalt der Avena-Koleoptilen-Spitze und die chemischen Natur des extrahierbaren Auxins. <^. JVaturf. 5B, 374. SoDiNG, H., and Raadts, E. (1953). Uber das Verhalten des Wuchstoffes der Koleoptilenspitze gegen Saure und Lauge. Planta, 43, 25. Stowe, B. B., and Thimann, K. V. (1954). The paper chromatography of indole compounds and some indole-containing auxins of plant tissues. Arch. Biochem. Biophys. 51, 499. t From the results of further experiments, it is doubtful whether the second growth substance of the raw diffusate and the second growth substance of the alkaline shaking are identical. 55 Natural auxins Terpstra, W. (1953). Chromatographic identification of the growth substance extracted from Avena coleoptile tips. Proc. K. Ned. Akad. Wetensch. Ser.C,56, 206. WiLDMAN, S. G., and Bonner, J. (1948). Observations on the chemical nature and formation of auxin in the Avena coleoptile. Amer. J. Bot. 35, 740. Yamaki, T., and Nakamura, K. (1952). Formation of indoleacetic acid in maize embryo. Sci. Pap. Coll. gen. Education, Univ. Tokyo, 2, 81. 56 INDOLE COMPOUNDS IN PHOTOINDUCED PLANTS A. J. Vlitos| Boyce Thompson Institute for Plant Research, Yonkers, New York INTRODUCTION Indole compounds are known to regulate vegetative growth of plants but their influence on flowering has not been extensively studied. There is considerable evidence that the response to photoperiodic treatment can be modified by 3-indoleacetic acid (lAA). Application to short-day plants may inhibit flowering (Bonner and Liverman, 1953; Fisher and Loomis, 1954); an effect which can be counteracted by subsequent treatment with antiauxin. Treatment of long-day plants with auxin may reverse the photoperiodic response since it prevents their flowering when held under long-day condi- tions and induces floral initiation under short-day conditions. Because of these observations Fisher and Loomis (1954) have postulated that flowering of Lincoln soybeans may depend upon a proper balance between auxin and a flowering hormone. Further evidence in support of this hypothesis might be obtained by determining effects of photoinduction on the auxin content of plants. Studies were undertaken, therefore, to determine whether there was a decrease in the amount of indole compounds present in typical short-day plants such as Maryland Mammoth tobacco and Biloxi and Lincoln soybeans when they were photoinduced. The methods of extracting, identifying, and measuring quantitatively the various indole compounds before and after photoinduction are described in this paper. PAPER CHROMATOGRAPHY OF INDOLE DERIVATIVES Most of the work with quantitative auxin assays has been based upon nonspecific, biological assays (i.e. Avena coleoptile curvatures or Avena straight growth tests). Therefore, the precursors of lAA in plant tissues, examined by means of biological assays, have not been adequately considered in photoperiodic studies. With the advent of paper chromatography, a technique is now available to study, individually, the role of free lAA, lAA bound to proteins, and various precursors which can be converted into lAA in flowering. In addition, paper chromatography permits a direct study of the effect of naturally-occurring inhibitors on an auxin of known chemical composition. In 1953 we described a method for the quantitative determination of lAA and other indole compounds in plant tissues (Vlitos and Meudt, 1953). The technique is based upon paper chromatographic methods and upon the measurement of the maximum density of spots developed with /7-dimethyl- aminobenzaldehyde. A typical standard curve for lAA is reproduced in t Holder of the Carbide and Carbon Chemicals Company Fellowship at the Boyce Thompson Institute. 57 Natural auxins Figure 1. If the concentration of lAA applied to the chromatogram is plotted against per cent of light transmissions, a straight-line relationship exists over the range 0-1 and 10 micrograms of lAA. The calculated error inherent in the quantitative paper chromatography of lAA is within 3 per cent. This value is considerably below the values given by Block, LeStrange, and Zweig (1952) for the determination of amino acids by the maximum density method. 0-05 D1 0-2 05 f-0 S-'mdo/eacef/c acid 50 /^5 Figure 1. A calibration curve for 3-indoleacetic acid based on densitometer readings of spots developed with p-dimethylaminobenzaldehyde on paper chromatographs. EXTRACTION OF INDOLE COMPOUNDS FROM PLANT TISSUES It became apparent early in our study that a suitable extraction method for lAA and other indole compounds from plant tissues was necessary before quantitative paper chromatography could be utilized effectively (Vlitos and Meudt, 1954a). An extraction method was needed which would (1) preclude the conversion of tryptophan and other lAA precursors to lAA during extraction, and (2) extract free lAA with a single extraction over a short period of time. While our work (Vlitos and Meudt, 1954a) was still in progress, a paper appeared by Terpstra (1953) which adequately reviewed the problems involved in the extraction or diffusion of auxins from plant tissues. Terpstra has described a method for extracting auxin from Avena coleoptiles, and from some green plant tissues, based on the water extraction of frozen material and subsequent dissociation of the extracted auxin-complex with ethyl ether. Kefford (1953) at about the same time described a method for the extraction of free indole compounds from plant tissues using absolute alcohol as the solvent. He found that all of the free auxin was obtained with one extraction at — 12°C. We proceeded to evaluate the methods of Terpstra and Kefford. Absolute ethanol was found to be a more efficient solvent than water for the extraction of lAA and tryptophan from tomato or spinach tissue {Tables 1, 2, and 3). The extraction of lAA from frozen plant tissue with absolute ethanol at — 10°C was found to fulfil the requirements for an extraction method set 58 Indole compounds in photoinduced plants Table 1 Percentage recovery of added lAA and tryptophan from control and treated tomato tissue with water at ^\Q'C on paper partition chromatograms. Results of two experiments (A and B) Treatment lAA A B 0/ - ,0 , Tryptophatfl A B 0/ 0' ,0 , Other indole A 0, ,0 compounds B 0/ /o Tissue+BOmglAA Tissue + 80 mg tryptophan Tissue (control) 80 mg lAA 80 mg tryptophan 67 75 62 75 80 71 85 100 6: lo: t Tryptophan was recovered from the aqueous fractions. { The Rf vahie of this compound corresponds to that for 3-indole-pyruvic acid and agrees with that reported by Stowe and Thimann (195-1). Table 2 Percentage recovery of JAA and tryptophan from treated tomato tissue with absolute ethanol and with mixtures of ethanol and water at — 10 C Treatment Percent of lAA recovered with Absolute ethanol Ethanol-water (4:1) Tissue + 80 mg lAA Tissue + 80 mg tryptophan Tissue (control) 80 mg lAA 80 mg tryptophan 100 100 100 2-lt 100 13 t Micrograms of lAA recovered from control tissue. Table 3 Percentage recovery of lAA and tryptophan from spinach tissue with water and with absolute ethanol at — lO'C Percent recoi 'ered with Treatment Water lAA Ethanol Water Ethanol Tryptophan Tissue + 80mgIAA 10 90 Tissue + 80 mg tryptophan — — 10 85 Tissue (control) 80 mg lAA 25 80 80 mg tryptophan 40 95 59 Natural auxins forth above (i.e. the free lAA was extracted with a single extraction and tryptophan was not converted to lAA during extraction). Therefore, the following extraction and chromatographic method was adopted for routine use in determining indole compounds in photoinduced tissues: I. Extraction 1. 300 g plant tissue harvested and frozen instantly. 2. Frozen tissue ground to fine powder without thawing. 3. Frozen ground tissue extracted with absolute ethanol for 12 to 24 hours at -10°C. 4. Ethanol evaporated off in vacuo leaving aqueous fraction, 5. Aqueous fraction acidified to pH 4-0 with 0-1 N H3PO4. 6. Acidified aqueous fraction saturated with glucose and extracted three times with ethyl ether. 7. Ether extract concentrated to 2-5 ml or less. II. QjLiantitative Paper Chromatography 8. 2-5 jul of extract were removed and applied to Whatman No. 1 paper as described previously (Vlitos and Meudt, 1953). 9. 0-1, 0-25, 0-5, 1-0, 2-5, and 5-0 micrograms of lAA (or any other indole compound) were applied to Whatman No. 1 paper. 10. After development of papers, density measurements of spots were used to construct standard calibration curve (Vlitos and Meudt, 1953). 1 1 . The amount of indole compound in extract was calculated by referring to a standard calibration curve. All the operations pertaining to extraction were performed in a cold room ( — 10°C). The standard calibration curve is based on five replicate paper chromatographs run simultaneously. If smaller quantities of tissue are extracted it is of course possible to rely upon qualitative paper chromatography followed by elution of the substance from the paper 'and subsequent biological assays {Avena coleoptile tests, root-growth measurements, etc.). The latter procedure introduces additional steps in the method and, no doubt, reduces the quantitative precision of the technique. Therefore it is desirable to use at least 300 g of tissue per extraction and maintain the quantitative aspect of the method. INFLUENCE OF ENVIRONMENTAL VARIABLES ENCOUNTERED DURING EXTRACTION ON THE STABILITY OF lAA In addition to avoiding the conversion of precursors to I A A during extrac- tion, it was necessary to assay the influence of the environmental variables which might affect the recovery of free lAA from plant tissues. The effects of pH, light, and temperature on the stability of lAA were evaluated only in so far as they might influence the recovery of lAA for the short periods of time encountered during extraction and chromatography of the extracts. The influence of pH on the stability of lAA in a series of phthalate buffers adjusted to pH 1-5 to 9-0 with HCl and NaOH is recorded in Table 4. If an extraction of free lAA from plant tissues is not completed within a few hours there is a possibility of inactivation of the compound at low pH values. However, in experiments of this type an acidified extract is usually 60 Indole compounds in photoinduced plants Table 4 Stability of lAA in a series of buffers ranging from pH 1-5 to 9-Ot Incubation {hours ) Buffer +IA A 10 fig pH 1 3 76 Amount lAA recovered i/ig) Phthalate-HCl 1-5 10 10 <1 Phthalate-HCl 2-5 10 10 <3 Phthalate-HCl 3-0 10 10 >3 Phthalate-NaOH 4-0 10 10 >8 Phthaiate-NaOH 6-0 10 10 >8 Phosphate-NaOH 90 10 10 >8 t Similar results were obtained when the experiments were done in the presence of plant tissue extracts in the buffers. not kept for more than a few hours so that changes in the structure of lAA would not be expected to occur. The influence of infra-red and fluorescent radiation on the stability of lAA both in the presence and absence of plant brei is recorded in Table 5. Table 5 Influence of infra-red and fluorescent radiation on the stability of lAA in the presence and absence of soybean leaf tissue Optical density of spots developed with p-dimethylaminobenzaldehyde on paper Incubation period Soybean chromatographs^ {minutes) leaf tissue Fluorescent Infra-red Q-2 g caljcm'^-jmin 1-15 ^ caljcm-jmin 5 Present 0-28 0-26 5 Absent 0-28 0-28 10 Present 0-23 0-26 10 Absent 0-29 0-28 30 Present 0-20 0-30 30 Absent 0-30 0-26 60 Present 0-24 0-18 60 Absent 0-27 0-26 180 Present 0-23 0-16 180 Absent 0-24 0-32 t Theoretical value for total recovery of lAA is 0-28 to 0-30. The results indicate that there is little or no inactivation of lAA as a result of short-term exposures to either infra-red or flviorescent lighting. The intensities used in these experiments are much higher than those which are usually encountered under laboratory conditions. It is often necessary to inactivate the enzyme systems which are responsible for catalysing the synthesis of lAA from tryptophan. Heating plant tissue to destroy these enzymes was first proposed by Gustafson (1941), but later discarded by other workers who found that boiling water released inhibitors of auxin from the tissue (Terpstra, 1953). Since the Arena coleoptile test was 61 Natural auxins used for auxin determinations, the inhibitors may have had no direct effect on auxin but instead may have been inhibitory in themselves to the growth of Avena sections. Paper chromatography, however, permits a study of the effect of boiling directly on the stability of I AA. The results of experiments involving boiling tissue and added lAA are given in Table 6. Table 6 Stability of lAA to heat in the presence and absence of plant tissue as determined by paper chromatography Treatment Optical density\ Tissue boiled 2 minutes ; lAA added to boiled extract 0-34 Tissue + lAA boiled 2 minutes 0-32 lAA — no tissue — boiled 1 minute 0-30 lAA — no tissue — boiled 2 minutes 0-29 lAA — no tissue — boiled 10 minutes 0-30 lAA — no tissue — boiled 30 minutes 0-32 t Theoretical value for total recovery of lAA is 0-33. There is no discernible release of inhibitors of lAA from plant tissue which is boiled. The results suggest that boiling plant tissue for short periods may be used to inactivate the enzymes which catalyse the conversion of tryptophan to lAA, providing that lAA content is determined chemically and not by means of biological assays. FREE INDOLE ACIDS IN SHORT-DAY PLANTS GROWN UNDER PHOTOINDUCTIVE AND NON-PHOTOINDUCTIVE DAYLENGTHS The development of the extraction procedure and the quantitative paper chromatographic technique was a pre-requisite to assays of indole compounds in photoperiodically sensitive plants (Vlitos and Meudt, 1954b). The above techniques were applied to answer the question : are free I AA levels decreased in short-day plants as a result of photoinduction? Such information is essential if plant physiologists are to examine the validity of the hypotheses which have been proposed to account for the influence of auxin in the flowering process (Bonner and Liverman, 1953; Fisher and Loomis, 1954). Soybean {Glycine max Mer. varieties Biloxi and Lincoln) and tobacco {Micotiana tabacum L. var. Maryland Mammoth) were the short-day plants used in the study. One group of plants was grown under greenhouse conditions in an environment providing a total of 18 hours light in a 24-hour cycle. The prevailing natural daylength was extended by light emanating from 500-watt Mazda tungsten lamps. The second group of plants was grown under short-day conditions (8-hour daylengths) attained by covering the plants with light-proof, aluminium-foil shields. The number of photo- inductive cycles necessary for floral initiation was determined from previous experiments under similar conditions. In every case the plants which were harvested for extraction were photoinduced or were completely vegetative. Photoinduced plants were harvested before macroscopic flower parts had formed. Vegetadve plants were of the same age as photoinduced ones at 62 Indole compounds in photoinduced plants harvest. Employing the extraction method and quantitative paper chromato- graphy as described above the types and amounts of indole compounds recovered from extracts were recorded. Results of a typical experiment appear in Table 7. Table 7 Amounts of Z-indoleacetic add {lAA), 3-indolepyruvic acid (IPA), and tryptophan {Tryp) in short-day plants grown under 8- and \8-hour photoperiods as determined by quantitative paper chromatography Tissue Photoperiod (hr) lAA gl^OO g of tissue IPA TrypX (R,0-5l) (/?rO-21)t {Rf 0-46) Biloxi soybean 8 125-0 + + + Not determined Biloxi soybean 18 <0-l + Not determined Lincoln soybean 8 0-1-0-25 + + + Not determined Lincoln soybean 18 O-l Not determined Maryland Mammoth tobacco leaves 8 <0-l 1 1 + + Maryland Mammoth tobacco leav es 18 <0-l + + Maryland Mammoth tobacco apices 8 <0-l Maryland Mammoth tobacco apices 18 <0-l + + t It was not possible to express the amounts of 3-indolepyruvic acid quantitatively since synthetic samples of the compound were too unstable to chromatograph quantitatively. J Tryptophan was contained in the aqueous fractions of tobacco extracts. Soybean extracts were not assayed for tryptophan content. A compound possessing an R^ value identical to 3-indolepyruvic acid (IPA) was detected in the three short-day plants which were studied. Strongly coloured spots were obtained on chromatograms for this compound in extracts from tissues grown under photoinductive cycles. Free lAA was present in highest concentration in extracts of tissues of Biloxi soybean plants grown under short-day conditions. Lincoln soybeans which were grown under these short-day conditions also contained more lAA than comparable plants grown under long-day cycles. The results of this study indicate that free lAA and IPA levels are increased under light conditions which favour the flowering of short-day plants, Biloxi and Lincoln soybean. IPA levels in Maryland Mammoth tobacco leaves are highest in photoinduced tissue. These results are at variance with the hypothesis that photoinduction and reduction in endogenous lAA levels are related processes. The recent work of Cooke (1954) substantiates most of the findings outlined above. Cooke found changes in auxin content in vegetative Maryland Mammoth tobacco {Nicotiana tahaatm L.) and Biloxi soybean to be correlated with the daylength in which the plants were growing. In contrast to our work he found more auxin in plants growing under long-day than in plants growing under short-day conditions. Plants which were suddenly shifted from long-day to short-day conditions showed a rise in auxin content for the next few days. Similar increases in auxin content have been noted in plants which are transferred to total darkness (Gustafson, 1946). According to Cooke (1954) increased auxin concentration occurs during the period of floral induction and it is not until several days later that the concentration of auxin begins to decrease. A decrease in auxin is not the cause of flowering as this decrease does not occur until after the induction period. Cooke in his 63 Natural auxins study measured auxin concentrations using the standard Avena method; therefore it is difficuk to make direct comparison of lAA or I PA content which might have influenced the curvature of the Avena sections. In summary, the evidence (Cooke, 1954; Vlitos and Meudt, 1954b) to date does not support the hypotheses advanced for the role of auxin or of indole derivatives in floral initiation. In the studies mentioned above where a thorough attempt has been made to follow the levels of either indole derivatives or unidentified auxins using two distinct techniques, the levels of indole compounds in photoinduced plants appear to be related with length of day rather than with floral induction. It is possible that the inhibi- tion of flowering by exogenously applied auxins may be due to a general inhibition of growth rather than any direct effect on floral tissue. In any event, as newer experimental techniques become av^ailable workers in this field will be in a better position to evaluate critically the many observational and arbitrary hypotheses which have been advanced to account for photo- periodism in plants, particularly those hypotheses which require a quantitative interpretation. REFERENCES Block, R. J., LeStrange, R., and Zweig, G. (1952). Paper Chromatography , Academic Press, New York, p. 195. Bonner, J., and Liverman, J. (1953). Hormonal control of flower initiation, Growth and differentiation in plants, Iowa State College Press, p. 283. Cooke, A. R. (1954). Auxin content during the photoinduction of short-day plants. Plant Physiol. 29, 440. Fisher, J. E., and Loomis, W. E. (1954). Auxin-florigen balance in flowering of soybean. Science, 119, 71. GusTAFSON, F. G. (1941). The extraction of growth hormones from plants. Amer.J. Bot. 28, 947. GusTAFSON, F. G. (1946). Influence of external and internal factors on growth hormone in green plants. Plant Physiol. 21, 49. Kefford, N. p. (1953). Properties of growth substances separated from plants by chromatography. Thesis, University of London. Overbeek, J. VAN, Olivo, G. D., and Vazquez, E. S. de (1945). A rapid extraction method for free auxin and its application in geotropic reactions of bean seedlings and sugar cane nodes. Bot. Gaz. 106, 440. Stowe, B. B., and Thimann, K. V. (1954). The paper chromatography of indole compounds and some indole containing auxins of plant tissues. Arch. Biochem. Biophys. 51, 499. Terpstra, W. (1953). Extraction and identification of growth substances. Meded. bot. Lab. Rijksuniv., Gent, 4, 64. Vlitos, A. J., and Meudt, W. (1953). The role of auxin in plant flowering. I. A quantitative method based on paper chromatography for the determination of indole compounds and of 3-indoleacetic acid in plant tissues. Contr. Boyce Thompson Inst. 17, 197. Vlitos, A. J., and Meudt, W. (1954a). The role of auxin in plant flowering. II. Methods for the extraction and quantitative chemical determination of free 3-indoleacetic acid and other indole compounds from plant tissues. Contr. Boyce Thompson Inst. 17, 401. Vlitos, A. J., and Meudt, W. (1954b). The role of auxin in plant flowering. III. Free indole acids in short-day plants grown under photoinductive and non- photoinductive daylengths. Contr. Boyce Thompson Inst. 17, 413. 64 THE BIOGENESIS OF NATURAL AUXINS S. A. Gordon Division of Biological and Medical Research, Argonne National Laboratory, Lemont, Illinois In attempting to organize material for this symposium, I realised that extremely little is known about how auxin is synthesized in the plant. Biochemical pathways and intermediates are easily formulated, and, unfortunately, much too easily accepted. It is indeed fitting to bring to an institution such as Wye College, whose tradition verges into the mythological past, a topic so rich in mythology. In this paper I will attempt to summarize current ideas of auxin biosynthesis and point out a few of the uncertainties. Most of us accept the pragmatic definition of auxins : growth substances that, though of multiple effect, have the specific ability to stimidate cell enlargement in shoots. Essentially it is a definition resting ultimately on bio-assay. With this biological criterion, what is the native auxin of higher plants ? Two decades ago this question could be readily answered — auxins a and b. Since then a number of attempts to repeat the original classical work have not succeeded, and many experiments indicated that the auxin in plant extracts and diffusates was indoleacetic acid (lAA). Several workers then took the position that lAA was the only auxin of higher plants. This position was equally unwarranted, since it was based almost entirely on indirect tests of doubtful validity: stability of activity in acid and alkali, molecular weight by diffusion velocity, dose-response curves, biological responses elicited by natural extracts as compared to pure auxins, colour tests, and destruction by TAA oxidase'. Auxin is not a chemically specific term; it encompasses a large number of compounds having certain structural similarities. In the past few years less equivocal techniques have identified chemically, or indicated chromato- graphically, the presence in plants of ethyl indoleacetate, indoleacetonitrile, indoleacetaldehyde, and indolepyruvic acid. These compounds are all biologically active and may be classified as auxins since at low concentrations they stimulate cell elongation. I believe it is profitable, however, to consider that they are inactive per se, and exhibit activity only after being converted to lAA. Though this generalization cannot be rigorously defended, it seems particvdarly appropriate to the identified indolyl substances. Many of the responses these substances elicit are concomitantly accompanied by the formadon of lAA. Indeed, indoleacetonitrile appears active only in those plant species or organs possessing an active mechanism for hydrolysing this nitrile to the acid. Thus, it does not seem unreasonable, at the moment, to assume that the native auxin of hormonal function in the plant is lAA. Pure lAA has the characteristics of polar transport, correlative action, and histo- genicity as the native auxin formed at production loci. On this basis, our title of natural auxin biogenesis may be altered to the more specific topic of lAA biosynthesis. 65 Natural auxins Schematic pathways of lAA biogenesis are familiar to most of you. Figure 1 indicates a recent version. The concept that tryptophan was the primary precursor of lAA in plants arose from the works of Dolk, Nielson, and Thimann in the early nineteen- thirties, culminating in the isolation and characterization of lAA from Rhizopus cultures by Thimann (1935). Since the yield of auxin in cultures of ^^^"^ -CH2-CH-COOH NH, CH, — C — COOH ^ II NH Indolylimmopnopionic acid CH2-C — COOH II NH'"" Indolylpyruvic acid ^^ NH Indolylethylidenimine CH, — C .^ =N NH Indoljlacetaldehyde NH Indolylacetonitrile f?^^>^ ^CH2 — CNHz ,^ NH' Indoljlacetamide NH' Indolylacetic acid Figure 1. Potential pathways of tryptophan conversion to indoleacetic acid {after Reinert, 1954). the micro-organism was dependent upon the amount of tryptophan present in the medium, and roughly proportional to the extent of aeration, the oxidative deamination of tryptophan to indolepyruvic acid (IPyA) followed by oxidative decarboxylation of the keto acid to lAA was postulated. This picture was strengthened subsequently by Wildman, Ferri, and Bonner ( 1 947) , who showed that leaf enzyme preparations could convert tryptophan to auxin or lAA, and that IPyA could also function as a source of auxin. Since Skoog (1937) found that tryptamine as well as tryptophan was converted slowly to auxin when applied to coleoptiles, the alternative pathway of tryptamine oxidation to lAA via indoleacetaldehyde was proposed. The existence of this dual pathway was also suggested by the results with whole tissue and enzyme preparations of the pineapple leaf. With these materials, the tryptophan to indoleacetaldehyde to lAA reactions apparently took 66 The biogenesis of natural auxins place; either the amine or keto-acid could act as potential precursors of the aldehyde and auxin. Many experimental results indirectly support the concept of tryptophan as the primary precursor. There is the almost ubiquitous occurrence of the tryptophan-IAA enzyme system in metabolically active tissues. There is also a parallelism in the relative enzyme activity and substrate concentration with known auxin production loci, and a parallelism between endogenous auxin level and tryptophan availability. Further support for the thesis that native auxin or lAA is formed from tryptophan can be adduced from experiments dealing with the responses of Figure 2. Relative free auxin levels in kidney bean plants follow- ing X-irradiation. The values shown were derived from the ^ig free auxin found per plant [bio-assay). plants to ionizing radiation (Gordon, 1955). Low doses of ionizing radiation cause an immediate depression of free auxin levels in various plant species. For example, Figure 2 indicates the relative auxin levels in kidney bean plants at various intervals after exposure to X-i-adiation. It may be observed that all the dosages employed, beginning at 25 r, caused an immediate depression of free auxin level. With the lower radiation dosages apparent recovery to control level was attained within one or two weeks. Inhibition and no recovery resulted after irradiations of 5 and 10 kr. The assumption that the lowering of auxin levels was caused bv reduced rates of auxin production was substantiated by a number of studies on the effect of radia- tion on morphological phenomena known to depend on continued auxin production. This sensitivity to irradiation and the pattern of temporary inhibition and recovery is closely paralleled by the effect of radiation on the enzymes converting tryptophan to lAA. Figures 3, 4, and 5 indicate that the enzyme system in mung bean seedlings may be examined under conditions where the conversion of tryptophan to auxin as a function of time is approximately linear. Substrate saturation and linear conversion may be attained in vivo by 67 V Natural auxins § f t 0-1 - 001 _ Conf_ro/ infiltrated with H2O only J 1 L Figure 3. Auxin production in mung bean seedlings as a function of the weight of tryptophan incorporated into 5 plants by vacuum infiltration. OS 1-0 Trypt option infiltrated IS lug Figure 4. Auxin production in mung bean seedlings as a function of time after vacuum infiltration of 1 per cent tryptophan. Figure 5. The in vitro conversion of tryptophan to auxin {cell-free homo- genates of mung bean seedlings) . mm 68 The biogenesis of natural auxins vacuum infiltration of tryptophan {Figures 3 and 4) or in vitro by the use of cell-free enzyme preparations {Figure 5). The vertical broken lines indicate the incubation times used in examining the effect of X-radiation on the activity of the tryptophan to lAA enzyme system. Figure 6. The in vitro conversion of tryptophan to auxin by cell-free homogenates made from rnung bean seedlings immediately after X-irradiation. The lower curve gives the relative amounts of lAA formed, the upper curve indoleacetaldehyde . ¥ 5 rbntgens xlO^ The radiation sensitivity in vivo of the tryptophan-I AA enzyme is shown in Figure 6. Unusually low doses of X-rays inhibited the conversion of tryptophan to lAA. An almost identical response to irradiation was shown using the 6 8 10 Time after irradiation 12 IV- days Figure 7. Relative ability of irradiated mung bean seedlings to convert infiltrated tryptophan to auxin at various times after irradiation. vacuum-infiltration technique. The order of dosage required for inhibition of enzyme activity is similar to that shown in Figure 2 for the reduction in vivo of free auxin level by X-radiation. Figure 7 describes the post-irradiation responses of the enzyme system. Represented are the inhibitions and recoveries in the rates of tryptophan conversion relative to non-irradiated controls. It can be seen that the extent of inhibition following irradiation increases as the dose becomes larger. A 69 Natural auxins recovery of biosynthetic ability is manifest up to about 1 kr. Recovery is virtually complete at either one or two weeks with the lowest dosages; partial recovery has been attained in two weeks by the plants receiving 1 kr; an inhibition with no recovery is shown by the seedlings receiving 5 kr. Again the similarity to Figure 2 is clear. The parallel effects of radiation on auxin production (as indicated by free auxin assays and morphological studies) and on tryptophan conversion to lAA offers strong supporting evidence that the two processes are synonymous. The effect of ionizing radiation on auxin formation may be tied a little closer, albeit less rigorously, to the accepted pathway of indoleacetate formation. Figure 6 indicates that the presumed immediate precursor of lAA, Figure 8. The relative conversion of crude indoleacetaldehyde to auxin by cell-free homogenates of X-irradiated mung bean seedlings. 5x10^ rontgcns indoleacetaldehyde, piles up during the conversion of tryptophan to lAA by homogenates of irradiated tissues. The same phenomenon was observed in the conversion of infiltrated tryptophan by irradiated plants. The evidence that we are really dealing with indoleacetaldehyde in these experiments is not unequivocal. The substance is neutral, possesses no or little activity in the Avena curvature test when tested directly, but is converted to lAA by coleoptile preparations and forms adducts with dimedone and bisulphite. We assumed that this substance was indoleacetaldehyde. Its accumulation suggested that the radiation block occurred at the enzyme which oxidizes the aldehyde to the acid, analogous to the biochemical blocks in Neurospera mutants. It appeared desirable, therefore, to examine more directly the effect of irradiation on the activity of the terminal enzyme. Figure 8 shows the relative conversion of crude preparations of indoleacetaldehyde to lAA by homogenates of irradiated tissues. Here again there is apparent the unusual in vivo sensitivity and dose response to ionizing radiation. It may be con- cluded that the radiation damage to lAA biosynthesis from tryptophan occurs at the oxidation of indoleacetaldehyde to lAA. If this is true, these experiments support the view that the aldehyde is an intermediate in auxin biogenesis. We might at this point begin to examine more critically certain portions of the pathways of indoleacetate formation represented in Figure 1. We men- tioned that the coleoptile can convert tryptamine to auxin and that leaves and leaf preparations of the pineapple plant form both the aldehyde and lAA from the amine. Even though active amine oxidases which could produce the 70 The biogenesis of natural auxins corresponding aldehyde have been shown to occur in plants and animals^ certain plants cannot utilize tryptamine in the formation of auxin or lAA. No auxin is formed from tryptamine by the spinach leaf. Neither mung bean seedlings nor cultures of U. zea will produce lAA from tryptamine. Perhaps more conclusive evidence for the non-participation of the amine is available from recent experiments with amine-oxidase inhibitors by Dr. R. Moss in our laboratory. In Table 1 are shown the effects of both enzyme inhibitors and Table 1 Amine-oxidase inhibitors and auxin formation % Inhibition Inhibitor f Tryptamine oxidation [O.^ uptake) Tryplamine to lAA Tryptophan lO A lAA i Marsalid IQ-^ M aa-Dipyridyl 10"^ M «-Butylamine 10-- M Spermine IQ-^ M l:3-Diaminopropane 10"^ M 70 100 + + + 100 90 31 24 13 1! \ X In all experiments, tryptamine and tryptophan 2-5 x- 10 M. competitive substrates on tryptamine oxidation and auxin formation. It may be seen that the first two compounds, marsalid and dipyridyl, inhibited almost completely the oxidation of tryptamine and the conversion of tryptamine to lAA. However, these compounds had no observable effect on the enzymatic conversion of tryptophan to lAA. Similarly, the mono- amines and the diamine inhibited both tryptamine oxidation and conversion, but had no effect on the tryptophan reaction. If one may extrapolate to other plant materials from these experiments, it appears highly unlikely that trypta- mine functions as an intermediate in the formation of lAA from tryptophan. On the basis of comparative biochemistry, one might anticipate that indolepyruvate would be the more probable intermediate. Here again a scepticism is warranted on the basis of the experimental evidence available. Experiments with the keto-acid are complicated by its high degree of spon- taneous breakdown to both lAA and the aldehyde as well as to other indole compounds. If corrections are made for spontaneous breakdown, or if enzymatic examination is facilitated by removal of the unreacted keto-acid as the phenylhydrazone after incubation, certain generalizations may be made. Several plant species can accelerate enzymatically the conversion of indolepyruvate to lAA. In this conversion, it has been shown with a number of plant tissues or plant preparations that no significant rise occurs in the level of the neutral component possessing the characteristics of the aldehyde. On the other hand, a similar treatment of tryptophan results in quite significant rises in level of the neutral component considered to be indole- acetaldehyde. At first glance this would suggest that the keto-acid is not an intermediate in the conversion of tryptophan to lAA. 71 Natural auxins This inference tends to be supported by the studies with ionizing radiation. The enzymatic conversion of indolepyruvate to lAA is not appreciably affected by ionizing radiation. It will be recalled from Figures 2, 6, and 7 that low X-ray dosages will result in the inhibition of both native auxin formation and the formation of lAA from tryptophan. Conversely, the enzymatic transformation of indolepyruvate to lAA is not significantly affected by irradiation of mung bean plants with more than 10 kr X-ray doses. Thus, if we accept the view that indoleacetaldehyde is the immediate precursor of lAA in tryptophan conversion, a view supported by the radiation studies, the lack of inhibition of IPyA conversion to lAA by radiation makes it difficult to accept the function of IPyA in the sequence. The conversion of IPyA to lAA without apparent formation of free aldehyde in most plant materials or preparations raises the interesting question of how the keto-acid is oxidized. By-pass of the aldehyde suggests the ad hoc explanation of 'simultaneous oxidative decarboxylation'. A mechanism whereby this may be accomplished was suggested by Krebs ( 1 936) from his work on amino-acid oxidases. For example, kidney preparations will catalyse in vitro the oxidation of a-amino acids to yield the corresponding keto-acid and hydrogen peroxide. In the presence of catalase or a coupled peroxidative action to reduce the peroxide, the keto-acid accumulates. In the absence of a mechanism for peroxide utilization, the keto-acid undergoes further non-enzymatic oxidative decarboxylation to yield the next lower saturated acid. In terms of tryptophan conversion to lAA, this sequence can be represented as follows: R_C— C— C00H + 02^^--R— G— C— COOH+HoOo .... (1) NH2 NH R— C— C— COOH + HgO-^^^^R— C— C— COOH + NH3 .... (2 NH O ) R— C— C— COOH+H2O2 -R— C— COOH + CO,+HoO .... (3) I II I O While it appears likely that reactions (2) and (3) may occur in the plant, evidence so far is against the occurrence of reaction (1). Plant amino-acid oxidases active on tryptophan are notable by their absence, or, if indicated, are of so low activity as to be of questionable significance. Of course, the low concentration of lAA in tissues and the low yields of lAA obtained from tryptophan might be considered evidence for the function of an enzyme in low concentration. At the moment, however, I wovild tend to discount the idea of an active tryptophan oxidase participating in IPyA synthesis. If IPyA does function in lAA formation from tryptophan, a transa.mma.s,e active on tryptophan might be considered as a mechanism whereby the keto-acid is formed. This seems plausible, since IPyA will replace tryptophan 72 The biogenesis of natural auxins as an essential amino acid in animals. It is of interest in this connection that no enhancement whatsoever of tryptophan conversion to lAA is obtained when a-keto-glutarate is added to suitable enzyme preparations, even with the amino acid in relatively high concentration. Hence one would also tend to discount a transaminase reaction on tryptophan as a primary reaction in the biogenesis of auxin. Verification that IPyA occurs as such in plants is needed. If we accept the chromatographic evidence that it does occur, there appears to be no experimental basis for assuming that it is formed from tryptophan. I would like to consider at this point a mechanism of pyruvate dehydro- genase action that may tie together a number of observations. Let us assume h" COz + Aldehyde DPT Acetaldehyde Acetaldehj'de Pyruvate — ►Acetom f-COj Acetolactate Diacetyl " Acetj'l acetoin 2Fe(CN)e 2:e-Cl2-<;)indo<(i Oj* particles Lipoic acid -►Acetate OX DPNH + H^ DPN* -* FP — *HA Lipoic acid red Acetate Pyruvate Lactate Acyl PO^ Figure 9. Pyruvate dehydrogenase mechanism (adapted from Gunsalus, 1954; 1955). that IPyA occurs in plants and that it is derived from tryptophan. It may be suggested that the subsequent steps of auxin formation proceed by a sequence analogous to the dehydrogenase mechanism proposed for pyruvic acid oxidation {Figure 9). The 'core' of these reactions is the formation of an intermediate 'aldehyde-diphosphothiamine' or 'aldehyde-DPT-enzyme' compound, followed by donor reactions for the aldehyde. Oxidation to the acyl-CoA would occur via reduction of lipoic acid without the liberation of free aldehyde or acetate. Hydrolysis of the CoA ester would then yield the free acid. Alternatively, the free acid may arise directly from the aldehyde- DPT either enzymatically or non-enzymatically in the presence of suitable electron acceptors. The above sequence would account for the conversion of IPyA to lAA by some tissues and tissue preparations either with or without the concomitant formation of free aldehyde. It is pertinent that free aldehyde also may be liberated from the DPT-complex in the presence of proton donors; possibly 73 Natural auxins pertinent also is the observation by Nance that acetaldehyde is liberated from some tissues by the addition of chlorophenol or indolephenol. This suggests that free aldehyde liberation depends both upon the particular redox system as well as upon the aldehyde diversion reactions present in tissues or heterogeneous tissue preparations. It would be of considerable interest to establish whether the mechanism indicated in Figure 9 operates with the indolyl analogue of pyruvic acid, and whether the sequence can be blocked by dimedone or bisulphite. Either of these carbonyl reagents will completely inhibit the conversion of tryptophan to lAA. The recent work of Leopold and Guernsey (1953), supported in part by that of Siegel and Galston (1953), is appropriate in this connection. It may be recalled from the former work that auxins enhanced sulphydryl oxidation in CoA mixtures; the latter work indicated that esterification of auxin with CoA does occur. Moreover, the results of Leopold and Guernsey show a fascinating correlation between the physiological activity of various auxins and their ability to enhance what is presumably CoA esterification in vitro. Additional significance may therefore be attached to the hypothetical scheme. It implies that (1) the process of auxin formation may be channeled directly into the auxin action mechanism without necessarily going to free lAA; (2) administered lAA and other auxins function by backing into the mechan- isms of auxin action (as well as auxin 'inactivation') as the acyl ester; (3) the auxin which functions as a hormone, i.e. as a translocated free acid, is the residual of concomitant biosynthetic and depletion mechanisms. Attractive as such a hypothesis may be, the operation of the pyruvate dehydrogenase mechanism upon which it is based is difficult to reconcile with the inhibitions occurring after irradiation by X-rays. The enzymatic conversions of trypto- phan and indoleacetaldehyde are highly radiosensitive, whereas very large doses of radiation have no observable effect on the enzymatic transformation of indolepyruvate to lAA in the same plant material. Finally, we may glance again at Figure 1 and comment briefly about the nitrile (IAN). Although the existence of IAN in one family of plants may be considered as established, the mechanisms of nitrile biosynthesis and of hydrolysis to lAA indicated in Figure 1 are very hypothetical. There is no evidence that the nitrile arises from tryptophan in vivo, or even that a mechanism for the transformation of tryptophan to IAN exists in plant tissue. While hydrolysis of the nitrile to lAA in a number of plant tissues does occur, apparently the amide is not an intermediate as suggested by Jones et al. (1952). In the chromatographic examination of IAN hydrolysed by plant preparations, Stowe and Thimann (1954) could find no trace of the amide, though they did identify I AA. The lack or low activity of IAN as an auxin in a number of plant tissues where lAA is active, and the absence of a mechanism converting the nitrile to an acid in many tissues able to utilize tryptophan, makes it rather unlikely that IAN is a normal intermediate in the conversion of tryptophan to lAA. In brief, at this time it can be said that tryptophan quite likely is the primary precursor of lAA, that indoleacetaldehyde probably is involved either as a free or complexed carbonyl, that indolepyruvate may be involved, and that tryptamine and indoleacetonitrile probably are not involved in normal auxin production (see also Gordon, 1954). If the keto-acid does participate, 74 The biogenesis of natural auxins the possibiUty of a pyruvate dehydrogenase mechanism has interesting imphcations. In conchision, I hope I have left the impression that pathways of auxin biosynthesis and interconversion are by no means clearly understood. In a very real sense the work is just beginning, because we have only now reached the stage of being able to define the problem. The field of auxin anabolism merits a concerted attack — not only as a provocative problem in biochemistry, but also as a physiological sequence functioning as a morpho- genetic determinant. REFERENCES Gordon, S. A. (1954). Occurrence, formation, and inactivation of auxins. Anriu. Rev. PL Physiol. 5, 341. Gordon, S. A. (1955). Studies on the mechanism of phytohormone damage by ionizing radiation. Proc. int. Conf. Atomic Energy, A/8/P/97, Geneva, Switzerland. GuNSALUS, I. C. (1954). Oxidative and transfer reaction of lipoic acid. Fed. Proc. 13, 715. GuNSALUS, I. C., HoRECKER, B. L., and Wood, W. A. (1955). Pathways of carbo- hydrate metabolism in microorganisms. Bact. Rev. 19, 79. Jones, E. R. H., Henbest, H. B., Smith, G. F., and Bentley, J. A. (1952). 3-indolyl- acetonitrile: a naturally occurring plant growth hormone. Nature, 169, 485. Krebs, H. a. (1936). Metabolism of amino acids and related substances. Annu. Rev. Biochem. 5, 247. Leopold, A. C., and Guernsey, F. S. (1953). A theory of auxin action involving coenzyme A. Proc. Nat. Acad. Sci., Wash. 39, 1 105. Reinert, J. (1954). Wachstum. Fortschr. Bot. 16, 330. SiEGEL, S. M., and Galston, A. W. (1953). Experimental coupling of indoleacetic acid to pea root protein in vivo and in vitro. Proc. Nat. Acad. Sci., Wash. 39, 1111. Skoog, F. (1937). A deseeded Avena test method for small amounts of auxin and auxin precursors. J. gen. Physiol. 20, 311. Stowe, B., and Thimann, K. V. (1954). The paper chromatography of indole compounds and some indole-containing auxins of plant tissues. Arch. Biochem. Biophys. 51, 499. Thimann, K. V. (1935). On the plant growth hormone produced by Rhizopus suinus. J. biol. Chem. 109, 279. WiLDMAN, S. G., Ferri, M. G., and Bonner, J. (1947). The enzymatic conversion of tryptophan to auxin by spinach leaves. Arch. Biochem. 13, 131. 75 GEOTROPIC RESPONSES IN ROOTS. SOME THEORETICAL AND TECHNICAL PROBLEMS p. Larsen Botanical Laboratory, University of Bergen THE AUXIN THEORY OF GEOTROPISM The auxin theory of geotropism offers a logical explanation of the upward curvature of stems and coleoptiles. The theory has been extended to explain the downward curvature of main roots as well. At least in the textbooks, the existence of a supra-optimal auxin content in roots is generally accepted. The well-known schematic curve {Figure 1), representing the reladonship between growth rate and internal auxin concentration in the elongation zone of roots, is the basis of the explanation of the downward curvature. We assume that the internal auxin concentration is supra-optimal for growth (point A on the curve of^ Figure 1). An addidonal supply of auxin to the Figure 1. Schematic representation of the relationship between growth rate and internal auxin concentration in roots. Intepnal concentration of auxin lower side of a main root placed in the horizontal position will thus retard the growth rate of that side, and consequently the root will curve down- ward. The theory seems to be well founded since both the higher auxin concentration in the lower half of the root and the retarding effect of added auxin on the rate of root elongation have been demonstrated experimentally. There are, however, various points in this theory which need clarification. 1 . Firstly, is the internal auxin concentration really supra-optimal ? This problem has been approached by observation of the effect of decapitation on the rate of elongation of roots. In several cases, decapitation results in an increase of the rate of elongation and this has been interpreted as a con- sequence of reducing the supra-optimal, internal auxin concentration to values closer to the optimum. The operation itself may have an influence on the growth rate, but one would expect that the effect of wounding would most likely be a retardation of growth. In various cases, decapitation actually decreases the growth rate; in those instances, however, in which a growth acceleration was noted, the result favours the concept that the normal, internal auxin concentration is supra-optimal. 76 Geotropic responses in roots Another method of approach has involved the external application of auxins to the roots. In this connection, only normal intact roots are of interest. The response of isolated or decapitated roots, in which the auxin level has been reduced artificially, permits no conclusions as to the auxin level in intact roots. A mostly slight, but significant acceleration of root elongation following the addition of auxin has been demonstrated in a number of roots {Table 1). These results seem to indicate a sub-optimal auxin concentration in the roots (point B in Figure 1). It would still be possible to apply the classical auxin theory of geotropism to such roots. The change in auxin concentration would only have to be large enough to bring the concentration in the lower half of the root well above the optimum. Theoretically, one would expect that very slight geotropic stimulations would result in negative curvatures, but these might be so small and transitory that they would escape observation. Table 1 Examples of stimulation of elongation in intact roots after application of lAA Reference Plant Time of observation'] Concentration, molsjl. Elongation, %of control Macht and Grumbein, 1937 Lupinus 24 hoursj 24 hours § 5-7x10-8 1-1x10-9 115 112 Thimann and Lane, 1938 Avena 24 hours 48 hours 5-7x10-11 5-7 X 10-11 136 117 Naundorf, 1940 Helianthus 5 hours 10-8 220 Lundegardh, 1942 Triticum 6 days 10-8 120 Moewus, 1949 Lepidium 17 hours 5-7x10-11 115 Linser, 1949 Lepidium 17 hours 5-7x10-11 110 Pohl and Ochs, 1953 Lepidium 17 hours 5-7 x 10-13 114 Pilet, 1951, Fig. 19 Fig. 20 Fig. 21 Lensll 1 day old 6 days old 12 days old 24 hours 24 hours 24 hours 10-' 10-8 10-11 909 185 •151 Ashby, 1951 Artemisia 24 hours 5-7x10-10 114 t After application of lAA. J § Auxin treatment 15 and 40 minutes, respectively. II Treated with K-salt of lAA. It can be questioned, however, whether the reported results really indicate sub-optimal auxin concentrations. As far as can be seen from the papers cited [Table 1), the positive response to added auxin was noted only after several hours, generally 17 hours or more. Only Naundorf (1940) reported a considerable increase in root length (in Helianthus) 5 hours after the addition of auxin. But even 5 hours is a very long time as compared with the time 77 Natural auxins required for a geotropic response to become manifest. In Artemisia roots, the geotropic reaction time is less than 8 minutes. Thus in order to test the auxin theory of geotropism in roots we need information on the response of the roots within, say, less than half an hour after the application of auxin. There are but few studies on the effect of auxin on the elongation of roots within such a short time, and all of these show retardation rather than acceleration of growth. Some observations showed that a growth accelera- tion, if found, was preceded by a retardation of elongation (e.g. Thimann and Lane, 1938), but there are several cases in which retardation only, with no acceleration of root growth, was found. Thus, for example, Seller (1951) measured the elongation of maize roots at 20-minute intervals and found that the application of lAA for 30 seconds resulted in a transitory retardation of elongation, no acceleration of growth being observed up to 4 hours after treatment with lAA. Lundegardh (1949) was able to measure root elonga- tion at 4-minute intervals, and although he had previously reported accelera- tion of the elongation of wheat roots 6 days after the addition of auxin (Lundegardh, 1942), he found only retardation when the response was recorded after a few hours (1942) or after less than one hour (1949). Similarly Ashby (1951) showed that auxin increased the growth rate of Artemisia roots when measured 24 hours after the application of auxin, but found no effect, or even a slight retardation, after 4 hours (Table 2). Table 2 The tirtie factor in auxin-induced stimulation of root elongation Time of observation'^ Concentration Elongation, Reference Plant oflAA. molsjl /o 0-7. Curvatures were measured directly on the negatives at a magnification of X 35. Figure 3. Continuous, unilateral exposure to gravity. Plants not rotated. Photographed at 0, 32, 64, 128, 192 minutes, and 10 hours. Figure 4. Rotational 1 revolution per O-b minutes {RjO-b) . Upper series: horizontal exposure {E) for 32 mimdes before rotation. Photographed at 0, 32, 64, 128, 192 minutes, and 23-5 hours. Lower series: no horizontal exposure. Photographed at 0, 64, 128, 192 minutes, and 8-5 hours. Figure 5. Rotation at 1 revolution per 32 minutes (/?/32). Upper series: horizontal exposure (E) for 32 minutes before rotation. Photographed at 0, 32, 64, 128, 192 minutes, and 22 hours. Lower series: no horizontal exposure. Photographed at 0, 32, 64, 224 minutes, and 22 hours. Figures. Development of geotropic curvatures. Ordinate: total curvature (C^^j,). Abscissa: sum of exposure (E) in the horizontal position and the time {T)of rotation at 1 revolution per 0-5 minutes{RI0-5). Upper curve: continuous stimulation, T = 0. 82 Geotropic responses in roots line. If the position of the plant were constantly adjusted in such a manner as to maintain the root tip in the horizontal position, one might expect the curvature to continue at the same constant rate for a considerable length of time. It is clear from the upper curve in Figure 6 that the reaction time is surpris- ingly short, at the most about 8 minutes. Definite curvatures could be observed after 10 minutes. Whether the curve should be extrapolated back as a straight line so as to cut the abscissa at 8 minutes, or whether it should be extrapolated as a smooth curve approaching a point closer to the origin, cannot be decided on the basis of these experiments. The latter procedure has, however, been chosen in the present work. The approximately linear course of the curve is followed until the curva- ture has reached a value of about 35°. At this point the sine of the angle of attack of gravity is still large, about 0-91 . After 20 hours the mean curvature is ca. 87°. The effect of klinostat rotation on the development of curvatures Plant chambers which were to be rotated were placed on the klinostat in such a position that the slides and agar slices were initially vertical, and the roots were parallel to the horizontal axis of the klinostat. Geotropic curva- tures would thus develop in the plane between the agar slices. Rotation was begun after given periods of exposure {E) in the horizontal position (mainly 0, 0-5, 4, 16, or 32 minutes). Curvatures were recorded after rotation for various lengths of time {T). Only two rates of rotation were used: 1 revolution in 0-5 minutes {RIO-5), and 1 revolution in 32 minutes (i?/32). The centrifugal forces developed by rotation were 1-3 X 10~^ Kg and 3-2 X 10^^ X^, respectively, i.e. they are entirely negligible as compared with gravitation. Rotation at RjO-5 — The results of rotation at 1 revolution in 0-5 minutes are shown in Figures 4 and 6. If roots are first exposed for 32 minutes in the horizontal position and then rotated at 7?/0-5, they will have developed a curvature of about 12° at the end of the exposure. They gradually enlarge their curvature during the rotation until they reach a maximum value of about 30° after some 100 minutes of rotation. The course of the curvature after that time is rather uncertain because the variability increases as a linear function of time, a consequence of the spontaneous movements (see later). After an initial exposure of 16 minutes the curvatures follow a similar, but lower course. If rotation is begun after 4 minutes of exposure, no measurable curvature has developed at the start of the rotation. Otherwise, the development of the response follows the same type of curve as in the case of longer exposures. An exposure for 0-5 minutes yields positive mean values, whereas the curvatures recorded during rotation without a preceding exposure are, on average, practically zero. After about 32 minutes of rotation, the curvatures in roots stimulated for 0-5 minutes are significantly higher than in the rotated, unstimulated roots (cf. Figure 11). Since the variability increases with time, a statistically significant difference could not be demonstrated after longer periods of rotation using the present set of data for exposures of and 0-5 minutes. Nevertheless, we may conclude that the presentation time is shorter than 0-5 minutes. 83 Natural auxins It is now of interest to compare the course of the rotation curves from the beginning of rotation and onwards, i.e. after they leave the curve for con- tinuous exposure, and this has been done in Figure 7. In the case of exposures for 16 and 32 minutes, the curvature present at the end of the exposure was subtracted from all the later values. The abscissa is the time {T) after E-16 to 32 min = 0-SSvn\n £=0-5m'\r\ £L. Figure 7. Development of geo tropic curvatures during rotation at i?/0-5 on the klinostat. Ordinate: increase in curvature {C^+ t~'^e) during the period of rotation. Abscissa: time {T) after beginning of rotation. Duration of preceding horizontal exposure (E) indicated for each curve. Data from Figure 6. 20 eo Time — •■ 100 no min beginning of rotation. When it comes to exposures of 4 and 0-5 minutes, the graphic transposition requires an assumption as to the course of the lower part of the reaction curve for non-rotated roots. If this curve were to Figure 8. Attempt at an estimation of the presentation time. Ordinate: increase in curvature during the rotation (C^+y — C^), readon curves inFigure? at times [T) indicated for each curve. Abscissa: logarithm of the duration of horizontal exposure {E) . The log- arithm of the presentation time {the shortest exposure which will produce a response later) estimated by extra- polation to —0-5, corresponding to a presentation time of 0-3 minutes. be extrapolated back as a straight line, a fictitious negative an'gle would be present at the beginning of rotation, and a corresponding angle should be added as a correction. In making this set of curves, however, it was assumed that a positive angle of about 1° was present after 4 minutes of exposure and that practically no curvature was present after 0-5 minutes. It is seen that the development of curvatures during the rotation is practically the same after 16 and 32 minutes of exposure. Roots exposed for 0-5 and 4 minutes, on the other hand, increase their curvature at a lower rate, and they reach lower maximum values. The initial velocity of each of the individual curves probably indicates the 84 Geotropic responses in roots velocity of geotropic curvature at the end of the horizontal exposure. This is the reason for assuming an accelerated course of the lower portion of the curve for continuous exposure. It is not certain, however, whether the course of the curves in Figure 7 reflects the velocity of the actual curvature or the velocity of some limiting process controlling the rate of geotropic bending, a process which is accelerated from almost the beginning of exposure, but which reaches a constant rate after 16 minutes, or perhaps sooner. It may now be possible to estimate the length of the presentation time by plotting the angles of (C^_^y— Cj,.) {Figure 7) after various times of rotation and extrapolating the resulting curves. In an attempt to do this, the angles at T = 32, 64, and 96 minutes were plotted against the logarithm of E (see Figure 8). Extrapolation yields a presentation time of 0-3 minutes. ■i E-32m\n + £ =16 min ° E-V- min o E =0 min V \ / \ /\V/ \/ w \ / \ A m 2w Time — ► mm Figure 9. Development of geotropic curvatures during rotation at 1 revolution per 32 minutes (/?/32). Ordinate: total curvature (C'^+y)- Abscissa: time (T) after begirming of rotation. Duration of preceding horizontal exposure {E) indicated. Curvature at the end of exposure can be read at T = 0. Symbols indicate actual determinations. Other maxima and minima were interpolated. The procedure applied here was chosen arbitrarily and other methods of graphic presentation are possible. At any rate, however, the true presenta- tion time must be somewhat shorter than 0-5 minutes, though it is probably of this order of magnitude. Rotation at /?/32 — If unstimulated roots are rotated at /?/32, the root tips bend back and forth with an amplitude very close to 5° {Figure 9). This result was to be expected on the basis of the known velocity of curvature during continuous exposure to 1 g and the size of the opposing stimuli received during the rotation. Measurements were made at 16-minute intervals, approximately when maxima and minima were to be expected. It was mentioned in a previous publication (Larsen, 1953) that pre- stimulated roots tend to straighten out again if they are rotated slowly, whereas their curvatures will increase if the plants are rotated at a higher velocity. This has been confirmed in the present experiments {Figure 5). The roots do not straighten out simply by curving in the direction opposite to the initial one; they follow the motion of the klinostat just as unstimulated roots do. The spontaneous movements which develop during rotation at /?/0-5 seem to occur only in the plane of the interface between the two layers 85 Natural auxins of agar, probably because the resistance is lower there than in other direc- tions. The geotropic movements induced during rotation at /?/32, on the other hand, most likely take place in all directions, so that the root tip follows a spiral. Deviations as small as ±2-b° would not be hindered much by the agar. The upper curve in Figure 9 shows the course of the curvatures of roots which have been stimulated in the horizontal position for 32 minutes. The average curvature at the end of the exposure was about 13°. The curva- tures increase during the first 16 minutes of rotation at i?/32 and reach a maximum of about 21°. During the second halfofthe revolution the curvature "^ fs - Figure 10. Variability of curvatures, a measure of the extent of the spontaneous move- ments. Ordinate: standard deviation of changes in curvature font zero time to the time of each of the subsequent readings. Abscissa : time after first read- ing. Most points based on 25- 35 degrees of freedom {exception: 6 upper xV; /= 17 or 8). Time is reduced by about 5°. During subsequent revolutions the amplitude is approximately constant, but both the maxima and the minima become lower and lower, so that after about 10 hours the root tip is pendling between 3 and 8 degrees. This pendling, probably a spiral movement, goes on for at least 12 more hours. Spontaneous movements. Variability When unstimulated roots are rotated at /?/32, they remain approxi- mately straight, apart from the spiral pendling of their tips. At i?/0-5 on the other hand, the roots will carry out spontaneous movements, resulting in large irregular curvatures. A measure of these spontaneous movements may be obtained by taking the standard deviation of the changes in curvature which take place after the initial measurement. 86 Geotropic responses in roots Figure 10 shows the magnitude of the standard deviation of the curvatures of roots under various conditions. The squares represent roots kept in the normal, vertical position. Their spontaneous movements are small, since they are constantly being corrected by gravitational stimulation. The upright crosses represent roots which were rotated at /?/32 without preceding stimulation. Their variability is somewhat greater than that of vertical roots, but still quite low. If roots are placed in the horizontal position, they start carrying out geotropic movements, and this is attended by a sudden increase in their variability (filled-in circles). Part of this variability is probably due to differences in the geotropic sensitivity and reactivity of the roots. After about 100 minutes, when the roots have reached an average curvature of about 45°, the variability starts decreasing again, and at 5 hours the variability of the previously horizontal roots is the same as in roots rotated at 7?/32. When roots are rotated at RjO-o, the picture is entirely different. They are subjected only to small and insignificant geotropic stimuli on the klino- stat, and are free to perform spontaneous movements which are only slightly, if at all, modified by gravitational influences. The open circles represent roots which were rotated without preceding stimulation. Their variability increases as an almost straight line for about 200 minutes and then levels off at a value (about 22°) much higher than in roots rotated at i?/32. It should be borne in mind that the mean curvature of these roots is very close to zero. When roots were first stimulated in the horizontal position for 0-5 minutes and then rotated at RjO-D their variability followed a similar course (x's), but did not level off at 22°. After 10 hours, the variability still seemed to be increasing. The standard deviation was 46°, and the mean curvature 2-9°. When roots are geotropically stimulated for longer periods before rotation, the spontaneous movements are superimposed on the actual response to the stimulation. DISCUSSION AND CONCLUSIONS The klinostat is an interesting instrument, and the present series of results do contribute to the discussion of the theory of the klinostat effect. In this communication, however, only a few points which may be considered of practical importance for the carrying out of geotropic experiments will be considered. 1 . Measurement of the response — As pointed out previously, the practice of recording the percentage of curving roots is inconvenient for various reasons. Since the technique used in the present study permitted the detection of very small curvatures, few roots could be considered absolutely straight, even when kept in the vertical position. This is not surprising in view of the spontaneous movements which have been demonstrated. In individual roots, the curvature present at the start rarely exceeded ±5°. A deviation of as much as ±8°, however, will have no measurable influence on the early response when a group of such roots is placed in the horizontal position, since the sine of 82'' is still 0-99. On the other hand, roots deviating as little as 3° from the vertical line will be stimulated by 5 per cent of the force of gravity and thereby induced to straighten out (sin 3^ = 0-05). 87 Natural auxins In the present study, the change in curvature from the start of exposure to the time of subsequent recordings was regarded as a response. Figure 11 shows an example of the distribution of curvatures in samples of individual roots under the influence of two different treatments. If a sufficiently fine technique of measuring could be applied, nearly 50 per cent positive curvatures would be present at any time in unstimulated roots. In general practice, however, workers using 50 per cent positive curvatures as a criterion for the development of a minimum response have fixed a certain angle as a limit below which all curvatures are neglected. The size of this minimum angle depends mainly on the technique of observation. It is thus clear that, whenever such a limiting angle has been chosen, the recorded result, 20 -Cl I. £=0\ £=0-5 __ 1 -1^-5 -1-6 1-5 'i-S -1^-5 -1-5 Cupvatune 1-5 i^-S 7-5 10-5 degrees Figure 11. Distribution of changes in curvature from to 32 minutes in 30 unstimulated roots [E = 0) and in 29 roots exposed for 0-5 minutes to \ g [E = 0-5). Curvatures measured after 32 minutes of rotation at RjO-5. Note increase in variability, cf Figure 10. expressed as per cent positive curvatures, is not a minimum response but a reaction of a certain magnitude. Possible negative curvatures, and positive curvatures smaller than the fixed minimum, are not considered by this method. On the other hand, the recorded result (per cent) may include a number of much greater curvatures which are not evaluated quantitatively. In order to detect a minimum response, therefore, the writer would strongly recommend determination of the mean curvature. The presentation time may then be determined by simple or logarithmic extrapolation, as shown for example in Figure 8. This does not mean extrapolating back until the first root has curved, but until the mean value about which the roots perform their spontaneous movements is just barely positive. If the chosen technique is such that curvatures smaller than, for example, 10° are neglected, 50 per cent positive curvatures will correspond to a mean curvature of roughly 10°. The exposure, E^q, which yields this mean value at the time of recording (e.g. 30 minutes after the end of the exposure) is longer than the presentation time. If we find that £50 is prolonged, for example by the addition of auxin, this does not necessarily mean that the presentation time has been prolonged. The result would be the same if the presentation time were unchanged and only the velocity of bending had been reduced. At the time of recording, a positive mean curvature, although 88 Geotropic responses in roots smaller than 10°, may actually have developed after an exposure of the same duration as in the auxin-free control. Similar considerations can be applied to determinations of the reaction time. 2. Klinostat rotation — If the curvatures are to be recorded after rotation on a klinostat, the rate of rotation should be fast enough to permit the curva- tures to develop freely, and the measurements should be made so early that the spontaneous movements do not create an excessive variability. If klinostat rotation is too slow, the roots will follow the motion of the klinostat, and the presence of maxima and minima of curvature will create difficulties in measuring. 3. Time of recording — Wlien studying the development of curvatures, particularly after a stimulation of such duration that measurable angles are present before the end of the exposure, the problem arises as to the correct time of recording the response. It seems impossible a priori to fix any particular time at which the observed curvatures may be taken as an adequate measure of the response to the stimulus applied. If one wants to fix a definite time at which to record the curvature, it is difficult to decide whether the starting point should be the beginning, the middle, or the end of the exposure. Another possibility would be to take the maximum curvature as a measure of the response, regardless of when it occurs. A way out of this difficulty, especially when studying the influence of external auxin on the geotropic responses, would be to determine the velocity of curvature under the influence of a stimulation of constant intensity, for example 1 g. Changes in this velocity will pi'obably be a satisfactory criterion for the efiect of changes in the experimental conditions. SUMMARY 1 . The acceleration of root elongation by added auxin reported in the literature does not necessarily indicate a sub-optimal auxin concentration in the elongation zone of normal intact roots. Such acceleration has not been recorded until several hours after the addition of auxin, whereas the immediate response, the one which has any bearing on normal geotropic reactions, is a retardation of elongation. There is, therefore, at present no necessity for changing the classical auxin theory of geotropism, which is based on the assumption of a supra-optimal auxin concentration in the root. 2. The behaviour of stimulated and unstimulated roots of Artemisia absinthium (wormwood) when rotated parallel to the horizontal axis of a synchronous klinostat at two velocities of rotation is described. At a velocity of 1 revolution per 0-5 minutes, geotropically induced curvatures are gradually enlarged during rotation. Large spontaneous movements, however, are superimposed on the actual geotropic response. The variability, therefore, increases enormously with the length of time of rotation. At 1 revolution in 32 minutes the variability is much lower, but the root tips follow the motion of the klinostat, i.e. they perform spiral movements with an amplitude of 5°. 3. Certain precautions to be taken in the recording of geotropic responses are pointed out. It is recommended that the mean curvature rather than the percentage of positive curvatures should be used as the measure of a geotropic response. 89 Natural auxins REFERENCES AsHBY, W. C. (1951). Effects of certain acid growth substances and their corre- sponding aldehydes on the growth of roots. Bot. Gaz. 112, 237. Brain, E. D. (1942). Studies in the effects of prolonged rotation of plants on a horizontal klinostat. III. Physiological reactions in the hypocotyl o{ Lupinus albus. New Phytol. 41, 81. BiJNNiNG, E. (1948). Entwicklungs- und Bewegungsphysiologie der Pflanze, Heidelberg (Springer). BuRSTROM, H. (1942). The influence of heteroauxin on cell growth and root develop- ment. Ann. agric. Coll. Sweden, 10, 209. BuRSTROM, H. (1950). Studies on growth and metabolism of roots. I\'. Positive and negative auxin effects on cell elongation. Physiol. Plant. 3, 277. BuRSTROM, H. (1954). Studies on growth and metabolism of roots. XI. The influence of auxin and coumarin derivatives on the cell wall. Physiol. Plant. 7, 548. Cholodny, N. (1932). 1st die Wachstumsgeschwindigkeit der Wurzel von deren Lage abhangig? Planta, 17, 794. CzAjA, A. Th. (1935). Polaritat und Wuchsstoff. Ber. dtsch. bot. Ges. 53, 197. Larsen, p. (1953). Influence of gravity on rate of elongation and on geotropic and autotropic reactions in roots. Physiol. Plant. 6, 735. LiNSER, H. (1949). Die Wuchsstoffwirksamkeit von 2,4-Dichlorphenoxyessigsaure und Phenoxyessigsaure. PJlSchBer. 3, 131. LuNDEGARDH, H. (1942). The growth of roots as influenced by pH and salt content of the medium. Ann. agric. Coll. Sweden, 10, 31. LuNDEGARDH, H. (1949). The influence of auxin anions on the growth of wheat roots. Arkivfor Bot. 1, 289. Macht, D. J., and Grumbein, M. L. (1937). Influence of indole acetic, indole butyric, and naphthalene acetic acids on roots oiLiipinus albus seedlings. Amer.J. Bot. 24, 457. MoEWUS, F. (1949). Der Kressewurzeltest, ein neuer quantitativer Wuchsstofftest. Biol. Zbl- 68, 118. Naundorf, G. (1940). Untersuchungen iiber den Phototropismus der Keimwurzel von Helianthus annuus. Planta, 30, 639. Navez, a. E. (1933). 'Geo-growth' reactions of roots of Lupinus. Bot. Gaz. 94, 616. OvERBEEK, J. VAN, Davila Olivo, G., and Santiago de Vazques, E. M. (1945). A rapid extraction method for free auxin and its application in geotropic reactions of bean seedlings and sugar cane nodes. Bot. Gaz. 106, 440. PiLET, P. E. (1951). Contribution a 1' etude des hormones de croissance (auxines) dans la racine de Lens culinaris Medikus. Alem. Soc. vaud. Sci. nat. 10, 137. PiLET, P. E. (1953). Auxines et amidon. IV. Essais d'interpretation du geotropisme des racines de Lens cidinaris Medikus. Bull. Soc. vaud. Sci. nat. 65, 409. PoHL, R., and Ochs, G. (1953). Uber die Wuchsstoffwirkung beim Streckungs- wachstum der Wurzel. Naturwissenschaften, 40, 24. ScHMiTZ,H. (1933). tjber Wuchsstoff und GeotropismusbeiGrasern. Planta, 19, &\'\. Seiler, L. (1951). Uber das Wurzelwachstum und eine Methode zur quantitativen Untersuchung des Einflusses von Wirkstoffen. Ber. schweiz. bot. Ges. 61, 622. Thimann, K. v., and Lane, R. H. (1938). After-effect of the treatment of seed with auxin. Amer. J. Bot. 25, 535. 90 II CHEMICAL STRUCTURE AND BIOLOGICAL ACTIVITY ON THE EFFECTS OF P^T^^-SUBSTITUTION IN SOME PLANT GROWTH REGULATORS WITH PHENYL NUCLEI B. Aberg Institute of Plant Physiology, Royal Agricultural College, Uppsala, Sweden INTRODUCTION The remarkable effect of introducing 3. para chlorine atom into the phenoxy- acetic acid molecule, thereby changing the physiological character of this substance from weakly anti-auxinic (or better: intermediate) to strongly auxinic, has also been found to have its counterpart among the more typical auxins and anti-auxins. The strong increase in the auxin activity from 2-chloro- to 2:4-dichlorophenoxyacetic acid is well known, and among the anti-auxins such a para chlorine effect upon the activity has been found for the phenoxy?jobutyric acids (Burstrom, 1951a) as well as for different types of substituted phenoxyacetic acids (see Aberg, 1954). Under the circumstances it was thought pertinent to make a more extensive study of several pairs of substances differing only with respect to a para chlorine substituent. Some insight into the nature of its function could be expected from studies on the effects of other para substituents. A number of phenoxyacetic acids with such substituents have therefore been included in the present connection. Most of the substances used have been synthesized by Prof. A. Fredga and Dr. M. Matell of Uppsala, to whom I tender my sincere thanks for their generous and unfailing co-operation. The following abbreviations will be used: POA: phenoxyacetic acid, CgHg-O-CHg-COOH; substituents in the phenyl nuclevis are indicated in the usual manner, Me signifying a methyl group, Et an ethyl group, iP an iso-propyl group, and /B a terf.-hutyl group. Some phenoxyacetic derivatives have been further abbreviated, according to common usage: 2:4-D = 2:4-Cl2POA, 2:6-D = 2:6-Cl2POA, 2:4:6-T = 2:4:6-Cl3POA. Th : thymoxyacetic acid, 2-?P-5-MePOA. ClTh: chlorothymoxyacetic acid, 2-fP-4-Cl-5-MePOA. POP: a-phenoxy-propionic acid, C6H50CH(CH3)COOH. POi'B: a-phenoxy?>obutyric acid. The 4-chloro derivative, 4-ClPOiB, was originally designated PCIB by Burstrom (1950). Tr PF lAA 1-NMSA transcinnamic acid, CgH5CH:CHCOOH. phenyl-formic or benzoic acid, CgH^COOH. 3-indolylacetic acid. 1 -naphthylmethyl-sulphide-acetic acid, C2qH7-CH2'S-CH2"COOH. 1-NMSP: the corresponding a-propionic acid. 93 Chemical structure and biological activity METHODS In order to obtain a reasonably complete picture of the growth effects of the different substances, three methods of testing have been used : (i) The Jiax root test {S-test, Aherg, 1950 ; 1953b; 1955) showing preferen- tially the inhibition of root growth by auxins, and also the restorative effects of anti-auxins upon 2:4-D or lAA inhibited roots. The activity of a certain substance in this test will sometimes be given as the molar concentration (C^q) needed for 50 per cent inhibition, or as the negative logarithm of this value (pCso). {ii) A wheat root test closely corresponding to the flax root test and showing much more pronounced root-growth stimulations at application of anti- auxins than the latter one (cf. Burstrom, 1950; 1951a,b; 1955; Hansen, 1954; Aberg, 1952; 1955). The wheat seedlings, var. 'Diamant IT, are used when the median root has reached a length of 10 mm, the test period (18 hours), the temperature (25°C), and the basic solution (pH 5-9) being the same as in the flax root test. During the germination period the seedlings grow on moist filter paper in large Petri dishes, and during the test period they are placed in perforated cork disks floating on the solution. Only the median root is measured. The average dispersion of the G values (growth as per cent of control) from independent experiments is 7-9J^0-5 per cent (of G) as judged from 25 series, each comprising 7-9 experiments. No clear correlation between the inter-experimental dispersion and the magnitude of the G value is apparent for this material, which fact may be connected with the occurrence of steep parts of the action curves within the whole of the G range covered (63-197) (cf. Aberg, 1955). (Hi) An Avena cylinder test performed mainly as described by Aberg and Khalil (1953), the seedlings being reared, however, on a net of stainless steel placed in a large glass dish immediately over the water surface, and the test period being 18 hours. The average growth of the control sections during this period amounts to 2-833b0-04 mm {a = 0-41, N = 120) or 28 per cent of the initial length. The average inter-experimental dispersion of the G values (expressed in per cent of control growth) is 10-7j^O-7 per cent (of G) as judged from 11 series, each comprising 7-10 experiments. WORKING HYPOTHESIS There is at present no generally accepted theory as to the mechanism of the physiological action of auxins and related substances. As a guide for our orientation among the multitude of experimental data we will, however, use the following working hypothesis which seems to give a coherent picture of a major part of them. The auxins are thought to interact with some 'receptor' of protein nature, and to exert their effects only when bound to these 'active sites' or 'growth centres' of the protoplasm. The binding is assumed to be effected by manv weak bonds ('multipoint attachment'), thus allowing for a high degree of specificity and taking advantage of the analogy with some better known enzyme-substrate complexes (Veldstra, 1953; Aberg, 1953b). The question .of the function of the hypothetical auxin-receptor complex 94 Pflra-substitution in regulators with phenyl nuclei is a very difficult one, and the solution will certainly depend on the attain- ment of a better insight into the complex chain of processes connecting the genotype of an organism with the result of its growth processes. Meanwhile we may correlate empirically the action of the auxins with different growth results. The fundamental outcome of such studies is that a substance showing auxin activity in a certain growth process will probably do it also in other connections (Went and Thimann, 1937; Audus, 1953), This could mean that there is a common receptor, but the possibility of several receptors with a similar affinity for different auxins must not be overlooked. It would also be reasonable to expect some variations in the type of the receptor (or group of receptors) for different plant species. In this connection well-documented exceptions from the rule of coupled and general activity may be of considerable interest. A typical competitive auxin antagonist, or anti-auxin, wovild now be a substance which is bound to the auxin receptor in the same place as the auxins, but which gives a complex unable to initiate the usual growth responses. There is at present a rather extensive series of substances, for example 1-naphthylmethylsulphideacetic acid and the corresponding a-propionic acid (1-NMSA and 1-NMSP), a-(4-chlorophenoxy)-?^o-butyric acid, l( — )-a-(2-naphthoxy)-«-butyric acid, 4-chloro/ra/2^cinnamic acid, and 4-z.5'o-propyl-phenoxyacetic acid, which may be assumed to behave in this way. All substances mentioned above : (a) counteract the inhibiting effects of externally applied auxin upon flax root growth, even if applied in con- centrations which are without effect on or slightly depress the growth of control roots, {b) stimulate the growth of intact flax roots slightly and that of wheat roots strongly, the latter ones presumably containing a more highly supra-optimal content of native auxin than do flax roots (cf. Aberg, 1955), (c) inhibit the growth of Avena coleoptile sections. As far as tested, these substances also counteract the effect of externally applied auxins on Avena coleoptile sections in the manner expected (McRae and Bonner, 1953; Aberg and KhaHl, 1953). Many other anti-auxins are known which are not yet tested by all of the different methods, or which show slight deviations from the pattern mentioned (e.g. no stimulation of flax roots). Such deviations are natural with respect to the occurrence of various complications in the physiological system (competition during uptake and transport, synergism, effects upon auxin metabolism, non-specific toxicity). No alternative to the hypothesis of a competitive auxin antagonism which is able to connect and explain the growth phenomena mentioned above has as yet been proposed. The counteraction of the effects of externally applied auxins could naturally be partly or wholly related to a hypothetical antagonism during the uptake (see Aberg, 1953b). Owing to the exceedingly low concentrations of the active substances which must be used, such a question is not easily attacked by direct methods, and the possibility of surface effects complicates the situation considerably. A comparison between the antagonistic effects obtained in different tests may, however, give some clues. Preliminary results for a series of homologous l-( — )-2-naphthoxy-alkylcarboxylic acids do indicate that a 'bulky' substituent in the side chain of an anti-auxin may possibly interfere to some extent with its uptake (gradually decreased stimulation of wheat root growth, decreased inhibition of Avena cylinder 95 Chemical structure and biological activity growth), while leaving its antagonistic effects against externally applied auxin intact. However, other explanations may also be possible. Once the possibility of a competitive inhibition of the auxin effects is accepted, the next step is to inquire into the existence of substances inter- mediate between the typical auxins and the typical anti-auxins (cf Aberg, 1952). The molecules of such a substance should have a conspicuously decreased, though not totally eliminated activity in initiating the auxin responses when JDOund at the proper sites. For the sake of brevity we will denote an activity defined in this way as 'intrinsic auxin activity'. We then have to differentiate between three different ways of defining the activity of a certain regulator as related to a specified growth result : (i) The 'gross activity' (i.e. secondary activity of Went and Thimann, 1937), expressing the relation between growth result and external amount or concentration of applied regulator; it is conditioned by the 'primary activity' of the regulator and also by such factors as penetration, transport, and destruction. (ii) The 'primary activity' (Went and Thimann, 1937), defined in relation to the amount of regulator immediately available at the active sites. {Hi) The 'intrinsic activity (McRae, Foster, and Bonner, 1953), defined in relation to the number of regulator molecules bound to the active sites, and thus indicating the specific growth activity of these molecules when properly situated (or of the receptor-regulator complex). Apparently the 'intrinsic activity is synonymous with the 'capacity' as used by Veldstra (1953) and by Jonsson (1955). In a system previously devoid of auxin, or with a low degree of auxin saturation, a regulator of the intermediate type can apparently be expected to give positive auxin effects. In a system already saturated, or nearly so, with auxin molecules of high intrinsic activity, it must on the other hand be able to replace so many of these that the sum of intrinsic activities decreases, that is, a competitive anti-auxin effect will occur. Substances which behave in this way are already known. As a good example we will mention 4-methyl-phenoxyacetic acid (4-MePOA) (Aberg, 1954). It strikingly increases the growth of 2:4-D-inhibited flax roots, but on the other hand, its own inhibiting effects may be effectively counteracted by an anti-auxin like 1-NMSP. In the Avena cyhnder test it gives a peculiar action curve {Figure 1), an initial inhibition being followed by a growth restoration which, however, does not reach the control level before the ulti- mate non-specific toxicity sets in. In the wheat root test an action curve of the inverse type is obtained, while the fiax roots with their higher auxin sensitivity and lower response to anti-auxins give a simple curve of the 'auxin type'. The ultimate inhibition of the wheat roots may possibly be of complex nature, auxin effects being mixed with non-specific toxicity. Curves similar to that of 4-MePOA have been found in the Avena cylinder test for 1-naphthoxyacetic acid and 2-naphthoxy/jobutyric acid (Aberg and Khahl, 1953), for 2-naphthylacetic acid, 2:6-dichlorophenoxyacetic acid, 3-methyl-phenoxyacetic acid, etc. Other substances, e.g. 2-methylphenoxy- acetic acid and 2-naphthoxyacetic acid, give curves where the initial inhibi- tion is followed by a stimulation up to nearly twice that of the control growth. The action curve of phenoxyacetic acid also seems to belong to this main 96 Para-substitution in regulators with phenyl nuclei class, though both the initial inhibition and the final stimulation are comparatively slight. The very strong difference in the inhibiting activities of 2-naphthoxyacetic acid (2-NOA) upon flax and wheat roots {Figure 1) is paralleled by a similar shift in the anti-auxin direction for several other 2-naphthoxy deriva- tives. It seems probable that here we meet with a case of species difference in the auxin system. For wheat roots and Avena coleoptiles, however, the sensitivity might be of the same type, the action curves indicating at first an 'anti-auxin component' of the activity which is, at higher concentrations, superseded by the auxin effects. It is highly interesting to find that 10~^ M 2-NOA has a conspicuous restoration effect upon wheat roots inhibited by SxlO^'^M I-naphthylacetic acid (Burstrom, 1955). Mofar concn. — •■ Figure 1. Theej]ectsof-\-n:ethylphenoxyaceticacid{-\-MePOA) and 2-naphthoxyacetk acid {2-NOA) upon the growth of fax roots (F), cress roots (C), wheat roots (W), and Avena coleoptile cylinders (A). Growth (G) in per cent of control, plotted against a logarithmic concentration scale. N is the number of independent experiments represented by the points. Further data, especially on the interaction of 2-NOA with other growth regulators in the growth of coleoptile cylinders, might be necessary for the final elucidation of these 'anti-auxin' effects (action II according to Burstrom, 1955). The quantitative treatment of the auxin-anti-auxin interactions can only be made in a tentative way at present, but analogies from simple model systems are certainly very useful in making more precise hypotheses as to the mechanisms vmderlying the growth results. It seems reasonable to start from the general laws governing chemical and adsorption equilibria, making as few additional assumptions as possible. This has been done by McRae et al. (1952, 1953) in applying the enzyme kinetics of Michaelis and Menten, and by Hellstrom (1953) in a more general approach. In both cases it must be assumed that a relatively slight proportion of the regulator molecules is 97 Chemical structure and biological activity actually bound, an assumption that finds good support in the studies of auxin uptake in plant tissues made by Sutter (1944) and by Reinhold (1954). Also, experiments with labelled 2:4-D (Hanson and Bonner, 1955) do seemingly indicate that the concentration of readily exchangeable regulator in the tissue is comparable to that of the surrounding solution. Qjuite generally we then find : M MIK a = -— , or a — where a is the fraction of places in an adsorption system which are occupied by the regulator, M the molar concentration of this regulator, and K the dissociation constant of the receptor-regulator complex. For a ^ 0-5 we get K = M. Upon addition of another regulator, for example an anti-auxin with the constant K' at the concentration M', we find the a-value for the first substance diminished to MIK Cf. l-^MjK + M'jK' Making further the rather bold assumption that the growth results, for example root-growth inhibition, are proportional to a, action curves may be constructed which resemble the empirical cvirves very closely. The effect of an antagonist in displacing the root-growth inhibition curve to higher auxin concentrations without any fundamental change in its form, is also in good agreement with expectation (Aberg, 1951; Aberg and Jonsson, 1955; Hellstrom, 1953). Some complications arise, however, from the presence of native auxin in the roots and from the non-specific toxic effects of higher concentrations of the antagonist. The action curve of a pure anti-auxin may be deduced as resulting from an adsorption in two patterns (competition with the native auxin, toxic effect at higher concentrations), the actual growth being determined by the product of both influences (Hellstrom, 1953). The Michaelis-Menten formula may be written as follows: V [^] M V K+[S] K-\-M where v is the reaction velocity, V the maximum reaction velocity, \S'\ or M the substrate concentration at equilibrium, and K the dissociation constant of the enzyme-substrate complex. Now, v is thought to be proportional to the amount of this complex, and vjV \% thus equal to a. If we now speak of the plasmatic receptor instead of the enzyme and the growth regulator instead of the enzyme substrate, this type of treatment appears to be identical with the former one. The assumption that a very small proportion of the regulator is actually bound, is made by McRae et al. (1952, 1953) by putting [S^ equal to the concentration of added regulator. Though calculations with help of the methods indicated above may often give results which are in good agreement with a limited series of experimental data, difficulties certainly arise when a more extensive set of results have to be treated. This is perhaps not surprising when due respect is paid to the complexity of the situation. The proportionality between the amount of 98 Pflra-substitution in regulators with phenyl nuclei receptor-regulator complex (corrected for variations in intrinsic activity) and the growth result may be restricted to a rather limited range. Time and penetration factors have to be considered (see Housley et al., 1954), effects upon the metabolic systems regulating the synthesis and destruction of native auxin are likely to occur and may be especially troublesome in systems involving externally applied indoleacetic acid (see Aberg, 1953a; Aberg and Jonsson, 1955), binding and competition at 'inactive sites' may be of import- ance, many substances certainly exert a non-specific toxicity at higher concentrations, and so on. In view of these difficulties it is natural that the attempts at quantitative precision must remain tentative. On the other hand, even limited success in such attempts is valuable as a starting point for further comprehensive quantitative treatment. PHENOXYACETIC ACID AND SOME OF ITS MAIN AUXINIC DERIVATIVES Phenoxyacetic acid (POA) has often been characterized as wholly inactive as a growth regulator. Some slight positive auxin effects have, however, been reported: e.g. negative stem curvatures induced by highly concentrated lanolin solutions (Zimmerman and Hitchcock, 1942) and positive effects in the Avena cylinder test at high concentrations (Muir et a/., 1949). On the other hand, the conspicuous restorative effects of POA on flax or cress roots inhibited by 2:4-D (Aberg, 1952; Audus and Shipton, 1952) and the positive effects upon wheat root growth and cell length (Hansen, 1954) might indicate a prevailing anti-auxin activity. This anti-auxin activity is also clearly shown by the inhibitive action of POA on Avena coleoptile growth when using cylinders with a fairly high residual growth (20 per cent) in absence of added auxin (Ingestad, 1953). In our experiments the residual growth has been even higher (28 per cent) and the inhibition by POA appears at about ten times higher concentrations than in Ingestad's experiments. This may be due to the higher pH used (5-9 and 4-5 respectively). It is highly interesting that at strongly increased concen- tions there is a significant decrease in the inhibition, and that positive effects occur in the range above 10~^ M. For example G(10-^) = 98-l±4-2; G(10-4) = 83-0±4-4; G(10-3) = 103-9±4-3; G(3xl0-3) = 117-0±14-0; A[G(10-3)_G(10-4)] = 20-9±4-9 {N = 1). These results, together with the positive effects observed by Muir et al. (1949) at lower concentrations (residual control growth 8 per cent and pH probably lower than in the present tests), suggest that the physiological activity of the POA-molecule is of the intermediate type, fairly low affinity for the growth centres being combined with very low intrinsic auxin activity. As regards the conspicuous auxin activity of 4-chlorophenoxyacedc acid (4-ClPOA) there can be little doubt. In the pea test and the Avena cylinder test, its activity is about 5-10 times less than that of 2:4-D (Fawcett et al., 1953; Wain and Wightman, 1953; Muir and Hansch, 1953) which correlates well with its inhibiting effects on wheat (Hansen, 1954) and flax roots {Figure 2). When comparing the POA and 4-ClPOA curves of Figure 2 we must remember that the inhibiting effects may be of different types. At the 99 Chemical structure and biological activity relatively high concentrations where POA begins to inhibit root growth, non-specific toxic effects are certainly possible, and even if the inhibition should be caused by the residual auxin effect of the POA molecules, the affinity of these molecules for the growth centres must be higher than indicated by the CgQ-value (p. 94). A better measure for this affinity may perhaps be obtained from combined experiments with 2:4-D (Aberg, 1952). We will use the concentration of the antagonist which, combined with 10~'^ M 2:4-D, gives a doubled growth rate as compared to that in 2:4-D alone (C = 200), and this concentration will be denoted C^ (negative logarithm = pC^). A calculation from a simple competitive model system shows that the C j- values of pure antagonists may be roughly comparable to the CjQ-values of Mo/ar concn. — ► Figure 2. The effects of phenoxyacetic acid {POA) and some of its derivatives upon the root growth of flax seedlings. The derivatives are indicated by means of the substituents : 2-Cl = 2-chloro-phenoxy- acetic acid {2-ClPOA), and so on. Values presented as in Figure 1. typical auxins as affinity measures, but because of the complexity of the physio- logical situation and the complications caused by the variations in intrinsic activity we prefer to use them in a purely provisional and exploratory manner in order to see if both sets of values may fit the same general picture. Even if we use the pC .^-value for POA (4-0) instead of the pCjo-value (3-4), the comparison gives an increase in affinity of 4-ClPOA (pCgQ = 6-8) of about six hundred times. For 2-chlorophenoxyacetic acid (2-ClPOA) there are clear indications of a weak auxin activity in the Avena cylinder test and in the pea test (see, for example, Muir et al., 1949; Wain and Wightman, 1953). Also the experi- ments with wheat roots (Hansen, 1954) and flax roots {Figure 2) indicate weak auxin activity. The growth inhibition of flax roots, however, is strongly counteracted by the anti-auxin 1-NMSP. In 2:4-dichlorophenoxyacetic acid (2:4-D) this weak auxin activity is strongly augmented. As indicated by the Avena cylinder test, the increase may be as high as four hundred times (Muir and Hansch, 1953) though the data of Wain and Wightman (1953) give an increase of only about fifty times. From experiments with flax roots 100 Para-substitution in regulators with phenyl nuclei {Figure 2) we get an increase of about two hundred times, and from the cell length data of Hansen (1954) about twenty times. For 2-methylphenoxyacetic acid (2-MePOA), the weak but conspicuous auxin activity is clear from the results of the Avena cylinder test (Muir and Hansch, 1953; cf. also above p. 96), the wheat root test (Hansen, 1954), and the flax root test {Figure 2; the inhibition is strongly counteracted by 1-NMSP). With this type of compound, the introduction oi a. para chlorine atom to give 2-methyl-4-chlorophenoxyacetic acid strongly augments the auxin activity. Both the flax root test and the cell length data from wheat roots indicate an increase of about one hundred times. Thus, with the above three pairs of substances, all having at least one ortho position free and with acetic acid as a side chain, we have found a very conspicuous increase in the gross auxin activity accompanying the introduction of a para chlorine atom into the nucleus. In root growth tests the transport factors are to a large extent eliminated, and it also seems probable that the effects of possible diflferences in the penetration rates are not very pronounced (see the Concluding Remarks, p. 112). Preliminarily, we may therefore assume a real difference \n primary activity. Much of this difference is probably due to an increased affinity to the growth centres, but as is evident from the behaviour of the pair POA-4-C1POA, we must also reckon with an influence upon intrinsic activity. That there may be an anti-auxin component also in the effect of 2-MePOA upon some tissues is suggested by its peculiar type of action curve in the Avena cylinder test (p. 96). For 2-ClPOA, an intrinsic activity lower than that of 2:4-D is suggested by the difference in the maximum stimulations obtained in the Avena cylinder test with these two substances {Mxxir et al., 1949; Wain and Wightman, 1953). It is thus possible that the true estimate of the affinities of 2-MePOA and 2-ClPOA might be somewhat higher than those indicated by the pCjQ values, and that the para chlorine atom influences intrinsic activity as well as affinity. The effect of such a substitution must certainly be expected to vary to some extent with the type of the parent substance, and it would not be suiprising if the affinity of POA is more strongly influenced than that of 2-MePOA or 2-ClPOA. In order to demonstrate the generality of the para chlorine effect among the auxinic phenoxyacetic acids, the following substances may be cited Irom Wain and Wightman (1953): 3-chloro-, 2:3-dichloro-, and 2:5-dichloro- phenoxyacetic acid. With all of these compounds the introduction o( a. para chlorine atom conspicuously increases the auxin activity as judged from the Avena cylinder and pea tests. In the phenylsulphideacetic (or phenylthioglycollic) acid series, the para chlorine effect seems to be of exactly the same type as for the phenoxy- acetic acids. This is true for unsubstituted phenylsulphideacetic acid and for its 2-chloro-, 3-chloro-, 3-methyl-, and 2 : 5-dichloro-derivatives (Sugii and Sugii, 1953; Kato, 1954). There is little doubt that /^ara-chlorination will turn out to have a similar effect also in the anihno-acetic acids (CeH^-XH-CHa-COOH). The 4-chloro- and 2:4-dichloro-derivatives show conspicuous activity in the Ave?ia cylinder and pea tests (Muir and Hansch, 1953; Veldstra and Booij, 1949), while no auxin activity has been reported for anilinoacetic acid itself 101 Chemical structure and biological activity With phenylacetic acid, jfeara-chlorination also increases auxin activity considerably, though less than o?//io-chlorination (Melnikov et al., 1953). The reported decrease in activity from 2-chloro- to 2:4-dichloro-phenylacetic acid makes a more detailed comparison between these substances highly desirable. SOME a-PHENOXYPROPIONIC ACIDS The racemic forms of a-phenoxypropionic acid (POP) and its para-ch\oro- derivative (4-ClPOP) have been compared in several tests by Fawcett et al. (1953), and their effects upon wheat roots have been studied by Burstrom (1951b) and Hansen (1954). While decidedly more active in the pea test, for example, racemic 4-ClPOP is only about twice as active as POP in inhibiting the growth of wheat roots. The cell length of wheat roots growing in a solution of 4-ClPOP may even be slightly greater than those in roots srowinsf in an equallv concentrated solution of POP. After resolution of both substances into their optically active forms (Fredga and Matell, 1952; Matell, 1954), the flax root test indicates a five-fold increase in the activity of D-( + )-POP upon/?ara-chloro-subsdtution, while the physiologically inactive l-( — )-POP is changed into a weak auxin antagonist. Detailed data from these experiments will be published in a later communication. The increased activity after /jara-chlorination both in the auxinic D-series and the anti-auxinic L-series might well be caused by a corresponding increase in the affinity of the chlorinated molecules to the growth centres. For the racemic a- (2-methylphenoxy) propionic acid, Synerholm and Zimmerman (1945) found an approximately tenfold increase in cell- elongation activity (tomato shoot) upon/jara-chlorination. With a-(3-methyl- phenoxy) propionic acid, however, the introduction of a para chlorine atom resulted in a decrease in activity. The reproducibility of these results is not easily judged, but irrespective of this, the latter pair of substances should not be cited as a case of absent or reversed para chlorine effect. We must remember that in this instance we are dealing with a mixture of two optically active forms which are probably of opposite physiological character. In spite of a possible increase in the activity of both forms when tested separately, the effect produced by the racemic mixture may be an unchanged or even decreased activity. In the present connection, the conspicuous rise in root-growth-inhibidng activity from a-anilinopropionic acid to its 2:4-dichloro-derivative (Aberg, 1953b) is also of some interest as it may be caused to a considerable extent by the para chlorine atom. SOME ACIDS WITH A STRONG ANTI-AUXIN COMPONENT IN THEIR ACTIVITY The anti-auxin effects of a-(4-chlorophenoxy)wobutyric acid (4-ClPO«B) on wheat roots were first described by Burstrom (1950, 1951a), who also found that they occurred at about ten times lower concentrations than the corre- sponding efTects of the unchlorinated a-phenoxywobutyric acid (PO?B). Quite similar results have been obtained by Hansen (1954) and the present author {Figure 3). This pair of substances have also been studied in the flax root test. The action curves of the pure substances show the expected course 102 Pflra-substitution in regulators with phenyl nuclei with much smaller stimulations than in the case of wheat roots (for PO/B see Aberg, 1952; 3 X lO"" M 4-ClPO/B gives a G value of 1 1 l-3±l-7;. The G'-curves from the combination experiments {Figure 3) show about a six-fold increase in the antagonistic activity of 4-C1P02B as compared to POzB. As both substances seem to be completely or almost completely devoid of intrinsic auxin activity, and assuming in the first instance a negligible influence of the possible difference in their penetration rates (pp. 101-112), this increase may apparently be related to a corresponding increase in their receptor affinity. 4-ClPOfB is known to counteract the effect of externally applied lAA or 2:4-D on the growth ofAvena coleoptile cylinders (McRae and Bonner, 1953) ; in the concentration range used (about 5> 10~' to 5 <10~^M) the effect 200 180 160 1 ^100 I 80 60 fO 20 "I 1 1 r yPOiB POiB _L _L "1 r f-ZlTr WO 300 200 ■s. I 100 % w'' w 10~ 10'"^ 10'^ 10''^ Molar concn. 10'^ 10'^ W^ Figure 3. The effects ofphenoxy'isobutyric acid {POiB) , iranscimamic acid { Tr) . and their A-chloro- derivatives {4-ClPOiB and 4-ClTr, respectively) upon the growth of wheat roots and 2-A-D-inhibited flax roots. The left scale {G) refers to the growth of wheat roots in per cent of control without added regulator (continuous lines), the right one [G') to the growth of the 2 : -i-D-inhibitedflax roots in per cent of control with 10"' M 2:4-Z) only (broken lines). The growth values are plotted against a logarithmic concentration scale. X is the number of independent experiments represented by the points. upon control sections was negligible. When tested by the present method (p. 94),4-ClPOiB gives a conspicuous inhibition of coleoptile cyhnder growth at 10-^ M and higher concentrations. Up to 5 X 10-^ M, 4-ClPO/B is a stronger inhibitor than FOiB, but at higher concentrations the curves tend to cross each other. This phenomenon will be further studied as it could possibly indicate a very faint residual intrinsic auxin activity of 4-ClPO?B molecules. As in the case of wobutyric acids, there are good reasons to assume that in transc'mna.mic acids the side chain is of a type normally excluding any 103 Chemical structure and biological activity conspicuous intrinsic auxin activity, while the affinity to the growth centres remains fairly high. The anti-auxin effects of /ra/zj'cinnamic acid (Tr) was first demonstrated by Overbeek et al. (1951) in experiments with pea stem sections. They have been confirmed here in experiments with flax roots {Figure 3). The peculiar form of the action curve of the pure substance (a 'plateau' at G = 90 in the range 3x10"^ to SxlQ-'^M) does, however, indicate that complications, in the form of a slight synergistic activity, may perhaps occur. In the wheat root test {Figure 3) we found a conspicuous stimulation preceded by faint and uncertain signs of an inhibition at lower concentrations which might again be related to the possible 'synergistic' component of its activity; such a synergistic effect might easily become wholly concealed by the stronger anti-auxin effects in this test. The weak action of Tr in relation to 2:4-D introduces some doubt to such a hypothesis (cf. below), but the synergistic phenomena are not fully under- stood and may comprise mechanisms of different types (see Aberg and Jonsson, 1955). In 4-chloro-ira«5cinnamic acid (4-ClTr) the anti-auxin activity is increased by nearly one hundred times in relation to Tr as judged from the effects on wheat roots and on 2:4-D-inhibited flax roots {Figure 3). It will be an interest- ing problem in further studies to decide if the stronger para chlorine effect observed in this instance, as compared to that shown by the pair POzB-4-ClPOzB, may be related to an elimination of the possible 'synergistic' component of the activity of Tr, or whether in fact it is due to an appreciable difference in the effect upon affinity. With 2-chloro-/rawj'cinnamic acid (2-ClTr) , the optimum stimulating concentration for wheat roots is 3 X 1 0~^ M, as compared to 3 X 10~^ M in the case of Tr. This may be due to increased toxicity (or the induction of some faint intrinsic auxin activity) as a result of ori^o-chlorination, again perhaps in combination with the elimination of the hypothetical synergistic component in the activity of the unsubstituted acid. Para-chlorination of 2-ClTr to give 2:4-dichloro-^ra«j^cinnamic acid (2:4-Cl2Tr) strongly increases anti-auxin activity {Figure ■^), though it does not exceed that shown by 4-ClTr. The absence of any strong effect of ori^o-chlorination on anti-auxin activity is in good agreement with the conditions in the isohutyv'ic acid series described by Burstrom (1951a). It is interesting to note that the increase in 'toxicity' from Tr to 2-ClTr strongly diminishes the apparent anti-auxin effects of the latter substance. Such a phenomenon well illustrates the unreliability of judging anti-auxin effects from one test or one set of conditions only. Both in the pair PO/B-4-ClPO?B and in Tr-4-ClTr, the anti-auxin effects on wheat roots become perceptible at considerably lower concentrations than those producing similar effects on 2:4-D-inhibited flax roots, which is only to be expected when one considers the higher total auxin concentration in the last case. With benzoic acid (phenyl-formic acid, PF), however, we meet an instance where the strength of the effects is reversed. The compara- tively slight stimulation of wheat root growth {Figure 4) has its counterpart in the much lower restorative effect exerted upon lAA-inhibited flax roots (unpublished experiments) than upon the 2:4-D-inhibited ones. The action curve of PF in the flax root test shows some resemblance to that of 2:3:5-tri- iodobenzoic acid (TIB) (cf. Aberg, 1953a; 1955; Aberg and Jonsson, 1955). 104 Para-substitution in regulators with ptienyl nuclei Further studies will also be made in order to test the possibility of a 'syner- gistic' component in the activity of PF. When tested for auxin effects upon shoot parts, PF has invariably been found inactive. In the Avena cylinder test 5^- 10""*M causes inhibition, while lower concentrations are without effect (Muir and Hansch, 1951). The introduction of a para chlorine atom in PF to give 4-chloro-benzoic acid (4-chloro-phenylformic acid, 4-ClPF) strongly increases the growth- stimulating activity both in respect to wheat roots and to 2:4-D-inhibited flax roots {Figi/re 4). When tested on Avena coleoptile sections with the same methods as used for PF, positive auxin effects are still found to be absent, while the inhibiting effects begin to appear at somewhat lower concentrations (10-^ M, Muir and Hansch, 1951). Afo/ar concn. Figure 4. The effects of some chlorinated tnLnscinnnmic acids (Tr) on the growth of wheat roots, and the effects of benzoic acid (PF) and its 4-chloro-derivative {4-ClPF) upon the growth of wheat roots and 2-A-D-inhibited flax roots. Values presented as in Figure 3. Both thymoxyacetic acid (Th) and its/;ara-chloro-derivative (ClTh) have previously been found to exert conspicuous anti-auxin effects in the flax root test when combined with 2:4-D or lAA (Aberg, 1954). Some further data have now been collected, and the G' curves o^ Figure 5 clearly show about a five-fold increase in the anti-auxin activity in respect to 2:4-D from Th to ClTh. Also the positive effects on wheat root growth begin at lower con- centrations for ClTh than for Th. The action curve of the latter substance also indicates some complications in the form of a possible 'synergistic' component in relation to the natural auxin system, which may be absent in respect to 2:4-D. This could possibly cause the depression of the G curves compared to the G' curves. For 2:6-dichlorophenoxyacetic acid (2:6-D) both positive auxin effects and anti-auxin effects are well documented. Though this compound exerts a V 105 Chemical strvicture and biological activity fairly strong antagonistic activity against externally applied 2:4-D and lAA in the Avena cyHnder test (McRae and Bonner, 1952) and in the flax root test (Aberg, 1954), and is therefore assumed to be a substance of mainly anti- auxin character, it is clear that at high concentrations it may also exert some auxin activity. This was first shown with the pea test by Thimann (1952) and has been corroborated by other workers (Wain and Wightman, 1 953 ; Osborne et al., 1954). In the Avena cylinder test it has been found inactive, or at fairly high concentrations, inhibiting (McRae and Bonner, 1952; McRae, personal communication ; Muir and Hansch, 1953; Wain and Wightman, zzo Molar concn. Figure 5. The effects of thymoxyacetic acid ( Th), 2 : 6-dichloro-pheno.xyacetic acid (2 : 6-D). and their A-chloro-derivatives {ClTh and 2 : 4: 6-T, respectively) upon the growth of wheat roots and 2 : 4-D-inhibited flax roots {broken curves). Values presented as in Figure 3. 1953). When the concentration is fvu'ther increased, the growth may, how- ever, be significantly restored, though it usually does not exceed the control growth. The action curve obtained with the present method almost coincides with that of 4-MePO A given in Figure 1. When using shorter test periods stimu- lations over the control growth may even be obtained (Osborne et al., 1954). The existence of a synergistic component in the activity of 2: 6-D has also been suggested (Thimann, 1952). The absence of conspicuous stimulations in the wheat root test {Figure 5) may under the circumstances be interpreted as the complex result of several different actions; thus the effect of compedtion with the native auxin is compensated by synergistic effects, the slight intrinsic auxin effects, and the toxic effects of 2: 6-D molecules. For 2:4:6-trichlorophenoxyacetic acid (2:4:6-T), no corresponding positive auxin effects on coleoptile sections at high concentrations have been reported in the literature, nor have indications of such effects been found during the present investigation. A very slight activity is indicated in the 106 Pflra-substitution in regulators with phenyl nuclei pea test (Thimann, 1952; Wain and Wightman, 1953; Osborne g/ a/., 1954), which may, however, possibly be of a synergistic nature (Veldstra, 1953) or due to the combined effect of very weak auxin activity and synergistic activity. The anti-auxin effects of 2:4:6-T are clearly shown by its inhibiting action on Avena coleoptile section growth induced by simultaneously applied 2:4-D or lAA (McRae and Bonner, 1952) and from its growth-restoring effects on 2:4-D-inhibited flax roots (Aberg, 1954; see aho Figure 5). From the experi- ments made by Hoffmann (1953) with tomato plants it seems clear that 2:4:6-T not only counteracts externally applied auxins of different types, but also interferes with plant responses controlled by native auxins, such as geotropic curvatures and rooting of cuttings. In the wheat root test used in the present investigation, 2:4:6-T gives comparatively slight but highly significant stimulations [Figure 5). For the cell length, however, Hansen (1954) did not find any corresponding increase. As in the case of 2:6-D we must expect the 2:4:6-T curve to be the complex resultant of counterbalancing effects, and it is quite natural that some differences should occur when using different materials and methods. On the whole it seems safe to conclude from experiments with wheat roots, with 2:4-D-inhibited flax roots {Figure 5), and with Avena coleoptile sections (McRae and Bonner, 1952), that the introduction of a para chlorine atom in 2:6-D to give 2:4:6-T increases the anti-auxin component of its activity. This increase is probably related to the higher affinity of 2:4:6-T molecules for the growth centres, but may also be partly related to a decreased intrinsic auxin activity of these molecules. In the clearly anti-auxinic 3:5-dichlorophenoxyacetic acid (Hansen, 1954), /)flra-chlorination to give 3:4:5-trichlorophenoxyacetic acid induces some slight positive auxin activity (e.g. in Avena cyUnder test and pea test. Wain and Wightman, 1953). Further comparative studies of this highly interesting pair of substances in different types of growth tests, especially those able to show the anti-auxin components of their activity, seem necessary before this phenomenon can be more closely discussed. THE EFFECTS OF para SUBSTITUENTS OTHER THAN CHLORINE In phenoxyacetic acid the introduction of a para chlorine atom results in maximum auxin activity of the new compound. Fluorine or bromine, on the other hand, gives a compound with considerably lower gross auxin activity than 4-chlorophenoxyacetic acid, but upon closer inspection the two compounds are found to be of rather different character. In the flax root test, the growth-inhibiting activity of 4-fluorophenoxy- acetic acid (4-FPOA) is five to six times lower than that of 4-ClPOA {Figure 6) ; further, the inhibition caused by 4-FPOA is strongly counteracted by the auxin antagonist 1-NMSP. In the wheat root test, 4-FPOA gives a rather slowly declining curve {Figure 6), and no stimulation is apparent at low concentrations, whereas in the Avena cylinder test, it causes strong growth stimulations. It seems fairly probable, therefore, that 4-FPOA should be characterized as an auxin with high intrinsic activity but with an affinity to the active sites which is well below that of 4-ClPOA. The action curve of this latter substance in the wheat root test {Figure 6) shows a slight but highly 107 Chemical structure and biological activity significant stimulation in the range 10"^ to SxlQ-^M (G = 105-8^ 1-4, iV = 10, and G = 104-9^ 1-6, iV = 11, respectively). Apparently we may assume that even in this fairly strong auxin the intrinsic activity is somewhat lower than that of the native auxin of wheat roots, and therefore a weak anti-auxin effect may be demonstrated with this material. Consistent with such an assumption is the lower maximum stimulation obtained in the Avena cylinder test with 4-ClPOA than with lAA or 2:4-D (Muir, Hansch, and Gallup, 1949; Wain and Wightman, 1953). In 4-bromophenoxyacetic acid (4-BrPOA) the intrinsic auxin activity is probably further lowered, resulting in a lower root growth-inhibiting activity in the flax root test and stronger stimulations in the wheat root test {Figure 6). In the Avena cylinder test, 4-BrPOA shows about 30 per cent of the gross activity of 4-ClPOA (Muir and Hansch, 1953). Figure 6. The effects of the para- halogenated phenoxyacetic acids on the growth of flax and wheat roots. F = 4-fluorophenoxyacetic acid (4- FPOA),andso on. Growth values (G) in per cent of control, presented as in Figure I. Molar con en. — " In 4-iodophenoxyacetic acid (4-IPOA) the intrinsic auxin activity seems to be very weak or absent; Muir and Hansch (1953) report that it is inactive in the Avena cylinder test. Its affinity to the active sites must be quite consider- able, however, as judged from the strong stimulations obtained with this substance in the wheat root test [Figure 6). In the flax root test it strongly reduces the inhibition caused by 10"'^ M 2:4-D when added in a concentra- tion of 10"" M. It is interesting to compare the halogenated phenoxyacetic acids with the corresponding a-phenoxyzj-obutyric acids, which have been studied by Hansen (1954). The anti-auxin activity of the latter substances increases with the size of the halogen atom. If intrinsic auxin activity is wholly absent from the 4-ClPOzB molecules, this would suggest that affinity progressively increases from 4-ClPO/B to 4-BrPO/B and 4-IPO/B. Alternadvely, it could mean that the weak residual intrinsic auxin activity is more completely eliminated in the last two substances than in 4-ClPO/'B. For 4-methyl- and 4-ethyl-phenoxyacetic acids (4-MePOA and 4-EtPOA), it has earlier (Aberg, 1954) been shown that they are able to counteract the inhibiting effects of 2:4-D upon flax root growth, and also that they produce root inhibitions which may be alleviated by addition of an anti-auxin like 1-NMSP. The anti-auxin component of their actixity is 108 Para-substitution in regulators with phenyl nuclei most pronounced with 4-EtPOA, and correspondingly we find stronger stimulations in the wheat root test with this substance than with 4-MePOA {Figure 7). The intermediate character of 4-MePOA is also apparent from the type of action curve obtained in the Avena cylinder test {Figure 1). In the flax root test with externally applied 2:4-D the antagonistic effect of the 4-alkyl-phenoxyacetic acids culminates with 4-?.fopropyl-phenoxyacetic acid (4-/P-POA), which also gives a significant stimulation of flax root growth in the range 10~'^ to 10~^M. With 4-?^ri.-butylphenoxyacetic acid (4-/B-POA) both this stimulation and the restorative effect upon 2:4-D- inhibited flax roots are slightly weakened (Aberg, 1954). For wheat roots, on the other hand, growth stimulations have already Figure 7. The effects of some pzn-a- alkyl-phenoxyacetic acids on the growth of flax and wheat roots. Me = 4- methjlphenoxyacetic acid {4-MePOA), and so on. Et = ethyl, iP = isopropyl, tB = tert.-butyl, Growth values (G) in per cent of control, presented as in Figure 1. The iP and XB effects on wheat roots do not differ significantly, and the values for both substances have therefore been jointly plotted. 10 ' 10" 10 ^ 10 10 ^ 10'^ 10 "' Molar concn. — - reached their maximum with 4-EtPOA {Figure 7). This is probably partly related to the lower auxin sensitivity of this material as compared with flax roots. To some extent it may perhaps also be caused by a decreased 'uptake' of the substances with 'bulky' alkyl-substituents (cf. p. 95). Summarizing, it may be said that the highest intrinsic auxin activity is possibly attained in the 4-FPOA molecules which, however, show a low affinity to the active sites. In 4-ClPOA this affinity is much increased, and the intrinsic activity is still high, but not high enough to exclude a very faint anti-auxin component of the activity. With further increase in the size of the para- substituent we get substances of clearly intermediate character (4-BrPOA, 4-MePOA, 4-EtPOA, and probably 4-1 POA), while the introduction of still larger substituents results in regulators of an almost pure anti-auxin type (4-?P-POA, 4-/B-POA). The lowering of intrinsic auxin activity seems to parallel fairly closely the increase in the van der Waals radius of the para- substituent (Pauling, 1948): e.g. /^ara-substituent : F CI Br CHg I van der Waals radius : 1-35 1-80 1-95 2-0 2-15 A 109 Chemical structure and biological activity THE TOXICITY OF PLANT GROWTH REGULATORS It seems to be a general rule that highly active anti-auxins which stimulate the growth of wheat or flax roots at very low concentrations also show inhibition at lower concentrations than weak anti-auxins. Such a phe- nomenon could lead to the assumption that the inhibitions exerted by sufficiently high concentrations of all anti-auxins were in some way related to the auxin system. A decrease in the level of active auxin below its optimum value would be a possible explanation, but this is ruled out by the fact that such inhibitions are not relieved by the addition of auxin (Aberg, 1951; 1953a). Instead, it has been assumed that the inhibitions are due to accessory toxic effects which have no connection with the auxin system proper (Aberg, 1951, 1953a; Burstrom, 1951a,b; Hansen, 1954). The nature of such a 'toxicity' is, of course, a problem in itself. It is tempting to assume some connection with a general ability of regulator substances to become bound to enzymes and other proteins, thereby upsetting their physiological functions when present in high concentrations. A certain parallelism between the affinity of the regulators for the specific growth centres and their general 'toxicity' would then be quite natural. In this connection, it is very interesting to find that there are indications of a parallelism between the bactericidal effect of many phenols and the growth- regulating activity of the corresponding phenoxyacetic acids. As a measure of the bactericidal activity, the 'phenol coefficient', giving the relation between the concentration of unsubstituted phenol and the equi-effective concentration of the substance to be characterized, may be used (see Reddish, 1954). A rise in the phenol coefficient upon /jara-chlorination is apparent for several pairs of phenols related both to auxinic and to anti-auxinic phenoxy- acetic acids. We thus find about a four-fold increase from phenol to 4-Cl- phenol, from 2-Cl- to 2:4-Cl2-phenol, and from thymol to chlorothymol; from 2:6-Cl2- to 2:4:6-Cl3-phenol the increase is approximately two- to three-fold, and from 2-Me- to 2-Me-4-Cl-phenol about five-fold (data obtained from Suter, 1941; McCuUoch, 1945; Wolf and Westveer, 1952; Reddish, 1954). Such a para chlorine effect is also apparent for many pairs of phenols studied by Blackman et al. (1955) in respect to their capacity to induce chlorosis in Lemna minor. In this case the rise in activity from phenol to 4-Cl-phenol is 7-3, from 2-Me- to 2-Me-4-Cl-phenol 10-6, and from the 3:5-Me.2-, 2:5-Me2-, 2:6-Me2-, and 3-Me-5-Et-phenols to the corresponding /?ara-chlorinated compounds 7-3, 9-5, 9-2, and 9-0 respectively. Hypothetically we may thus assume that /^ara-chlorination of phenols and phenol derivatives increases the affinity of the substance for receptor places in the cytoplasm, thereby conditioning in part the increased activity of several auxins and also leading to increased activity of many anti-auxins, to higher 'toxicity' (root-growth-inhibiting activity) of anti-auxins and to increased chlorose-inducing and bactericidal effects of many phenols. The strong increase in the rate of hydrolysis of certain chymotrypsin substrates upon /)flra-chlorination of a phenylalanine residue (see Neurath and Schwert, 1950; Aberg, 1953b) is relevant in the present connection in that it indicates a corresponding increase in enzyme-substrate affinity. The presence of a 110 Para-substitution in regulators with phenyl nuclei /)flrfl-chlorophenyl group in the systemic herbicide 3-(4-chIorophenyI)- 1 : 1 -dimethylurea (CMU) is of some interest too. Its action is certainly different from that of the auxins (Christoph and Fisk, 1954; Muzik et al., 1954), but nevertheless the general affinity of the /7«ra-chlorophenyl group to the plasma proteins may be of importance in conditioning it. Upon closer inspection deviations are likely to occur within the series of phenomena now compared. Though perhaps similar to a certain extent, the receptor sites for such widely different processes may certainly be expected to differ in other respects. For the mainly auxinic substances, the magnitude of the para chlorine effect upon affinity may be difficult to separate quantita- tively from simultaneous effects on intrinsic activity, and other difficulties are encountered if regulator substances of clearly intermediate character are included in the comparisons. For an anti-auxin with conspicuous residual auxin activity, the inhibiting effect on root growth exerted by fairly high concentrations may thus be of auxin character, or it may be a mixed effect comprising both an auxin component and a component of 'nonspecific toxicity'. For weak auxins the root inhibitions at high concentrations may likewise be mixed. CONCLUDING REMARKS There still remains to be discussed the possible effects of varying penetration rates of the different regulators to the active sites. The exact location of these sites offers immediate difficulties. In the case of surface-controlled reactions of growing cells, the penetration factor may probably be regarded as rela- tively unimportant when working with such objects as roots. At least in the outer tissues, transport through the cell walls will hardly differ very much for the various substances; when of a purely diffusive character it will naturally be somewhat slower for substances of higher molecular weight. If, however, it is assumed that a main part of the regulating action takes place after penetration of the plasma membrane, the factors determining the rate of this process may become important, provided that the rate is too slow to permit the equilibrium concentration to be reached under the dynamic conditions of growth. How far this situation is realized cannot be judged at present, and certainly it must differ for different types of material and even for different rates of growth. Nevertheless, it may be of value to consider briefly two possible sources of variations in the uptake of different regulators. It is fairly certain that most of these substances penetrate into the cell mainly in the form of undissociated, lipophilic molecules, though some uptake in the form of anions may also occur (cf. Simon and Beevers, 1952). A difference in the dissociation constant [K) of two regulators tested at the same pH might thus possibly cause a difference in their uptake rates. The effect of j&ara-chlorination on the dissociation of some of the substances used in the present study is shown by the following /v-values : POA: 6-8x10-4, 4-ClPOA: 7-9 10-4 (Muir f/«/., 1949). PO/B: 3-8 xlO-^ 4-ClPOfB: 3-3x10-4 (Burstrom, 1951a). Tr: 3-65x10-5, 4-ClTr: 3-86x10-5 (Nivard, 1951). PF: 6-27 xlO-^ 4-ClPF: 10-6x10-5 (Nivard, 1951). v^ Ill Chemical structure and biological activity Apparently there is no consistent effect of such magnitude that it can be related to the observed differences in gross activity. The situation might be influenced, however, by the degree of lipophily of the undissociated molecules, which, together with the molecular size, is a very important factor determining the permeation rate (see Collander, 1954). For several aliphatic substances, Collander (1949) found about a five-fold rise in the partition coefficient C'ethei/Qater upon the introduction of a chlorine atom, and the same may be true for aromatic substances. Thus, the water solubility of iV-phenylmaleimide is about twenty times higher than that of the 4-chloro- derivative, and this has been suggested as explaining, in part at least, the higher physiological activity of the chlorinated compound (Overbeek et al., 1955). As pointed out by these authors, however, such a mechanism cannot fully explain the specific effect ois. para chlorine atom, and such specific effects are clearly apparent with the chloro-phenoxyacetic acids, the chloro-phenoxyuobutyric acids and the chloro-irawi-cinnamic acids. It seems, therefore, that we must turn from the less specific solubility effects to the much higher specificity possible in a multipoint adsorption system in order to get a plausible hypothesis as to the nature of the para chlorine effect. As a reminder against over-estimating the lipoid solubility factor, the parti- tion coefficients Qther/^water fo^" phcnylacctic acid and indoleacetic acid may be cited: PA 37, lAA 20 (Collander, 1949). Additional evidence for the probable absence of conspicuous penetration factors related to undissociated molecules in a root growth test comes from Burstrom's (1951a; 1952) studies on the effects of PO/B and 4-ClPOeB on wheat roots at different pH values. An increase in the pH value from 4 to 7 has, in the presence of a normal calcium supply, no significant effects upon the growth stimulation caused by these substances, in spite of the strongly reduced amount of undissociated molecules present at higher pH values. The non-auxinic inhibitions, on the other hand, are markedly shifted to higher concentrations with increasing pH. Though /?ara-chlorination may certainly influence the permeation rate of a regulator substance, and though this rate may be of importance in condi- tioning the magnitude of the gross activity in some types of tests, nevertheless we may conclude that such a factor is likely to be of little significance in root tests. This may perhaps be connected with the localization of the growth centres at the cell surfaces and with the rapid penetration of the regulators on to these surfaces. Further experimental studies are needed, however, for clarification of the situation, but provisionally the permeation factor may be neglected when considering the results of such tests and the emphasis laid upon the probably more important effects of variations in affinity and in intrinsic activity. The effects of/xzra-chlorination may then be summarized as follows: (?) It generally increases the auxin activity of weak auxins. In the phen- oxyacetic acid series this increase may amount to one hundred times or more, and it appears probable that it is mainly conditioned by an increase in the affinity to the growth centres and in the intrinsic activity of the molecules. (n) In the typical anti-auxins, the increase in the growth-stimulating activity upon wheat roots or in the growth-restoring activity upon 2:4-D- inhibited flax roots usually is of the order of about five to ten times or more. 112 Pflra-substitution in regulators with phenyl nuclei As the intrinsic activity of these molecules is thought to be completely or almost completely blocked, this increase may preliminarily be taken as related mainly to the affinity for the growth centres. {Hi) The affinity effect must certainly be expected to vary within fairly wide limits for different types of substances. Tentatively, the estimate obtained from anti-auxins of the phenoxy type may, however, be applied to substances like POA, 2-ClPOA, and 2-MePOA, for which the para chlorine effect could then be divided into about a five- to ten-fold increase in receptor affinity and about an equal or stronger increase in intrinsic auxin activity. The present hypothesis evidently postulates the existence of a graded series of intermediate growth regulators with intrinsic activities between those of the typical auxins and the very low or absent activities of the typical anti-auxins. The interactions of substances like y-phenylbutyric, 4-methylphenoxyacetic 4-ethylphenoxyacetic, and carvacroxyacetic acid (2-Me-5-iP-POA) with 2:4-D and 1-NMSP strongly indicate that they occupy such an intermediate position (Aberg, 1952; 1954). The auxin and anti-auxin effects shown by substances like 2:6-dichlorophenoxyacetic acid or 4-bromophenoxyacetic acid in different types of tests, and the peculiar compound action curves of such substances in the Avena cylinder test (p. 96) also point to the same conclusion. That even fairly strong auxins may have an intrinsic activity below the possible maximum is evident from the quantitative treatment of the IAA-2:4-D-interactions in the Avena cylinder test (McRae, Foster, and Bonner, 1953) and from the slight stimulations of wheat roots by low concentrations of 4-chlorophenoxyacetic acid (Figure 6). The maximum stimulations obtainable under certain conditions with Avena coleoptile sections may be directly related to intrinsic activity, the maximum stimulations of wheat roots may be inversely related to the same measure, and data for judging the position of a certain substance may be obtained from combination experiments like those made with flax roots. In all these different methods disturbances may be caused in several different ways (uptake and transport factors, non-specific toxicity, synergism, and so on). Therefore, the safest way of approach to a comprehensive quantitative treatment of the primary reaction, or reactions, underlying the diverse growth responses seems to be the parallel use of different test methods and the attempted synthesis of the results into a coherent picture. REFERENCES Aberg, B. (1950). On auxin antagonists and synergists in root growth. Physiol. Plant. 3, 447. Aberg, B. (1951). The interaction of some auxin antagonists and 2:4-D in root growth. Physiol. Plant. 4, 627. Aberg, B. (1952). On the effects of weak auxins and antiauxins upon root growth. Physiol. Plant. 5, 305. Aberg, B. (1953a). On the interaction of 2 : 3 : 5-triiodobenzoic acid and maleic hydrazide with auxins. Physiol. Plant. 6, 277. Aberg, B. (1953b). On optically active plant growth regulators. LantbrHogsk. Ann. 20,241. Aberg, B. (1954). Studies on plant growth regulators. IX. Para-alkyl-phenoxy- acetic and -propionic acids, and some related derivatives of naturally occurring phenols. Physiol. Plant. 7, 241. 113 Chemical structure and biological activity Aberg, B. (1955). Studies on plant growth regulators. X. On the reproducibility of root growth results. LantbrHogsk. Ann. 21 (1954) 197. Aberg, B., and Jonsson, E. (1955). Studies on plant growth regulators. XI. Experi- ments with pea roots, including some observations on the destruction of indole- acid acid by different types of roots. LantbrHogsk. Ann. 21 (1954) 401. Aberg, B., and Khalil, A. (1953). Some notes on the effect of auxin antagonists and synergists upon coleoptile growth. LantbrHogsk. Ann. 20, 81. AuDUS, L.J. (1953). Plant growth substances, Leonard Hill Ltd., London. AuDUS, L. J., and Shipton, M. E. (1952). 2 :4-Dichloranisole-auxin interactions in root growth. Physiol. Plant. 5, 430. Blackman, G. E., Parke, M. H., and Garton, G. (1955). The physiological activity of substituted phenols. I, IL Arch. Biochem. Biophys. 54, 45. Burstrom, H. (1950). Studies on growth and metabolism of roots. IV. Positive and negative auxin effects on cell elongation. Physiol. Plant. 3, 277. Burstrom, H. (1951a). Studies on growth and metabolism of roots. VI. The relative growth action of different i^obutyric acid derivatives. Physiol. Plant. 4, 470. Burstrom, H. (1951b). Studies on growth and metabolism of roots. VII. Thegrowth action of a- (phenoxy) propionic acids. Physiol. Plant. 4, 641. Burstrom, H. (1952). Studies on growth and metabolism of roots. VIII. Calcium as a growth factor. Physiol. Plant. 5, 391. Burstrom, H. (1955). Evaluation of the growth activity of naphthalene derivatives. Physiol. Plant. 8, 174. Christoph, R. J., and Fisk, E. L. (1954). Responses of plants to the herbicide 3-(/?-chlorophenyl)-l:l-dimethylurea (CMU). Bot. Gaz. 116, 1. Collander, R. (1949). Die Verteilung organischer Verbindungen zwischen Ather und Wasser. Acta chern. scand. 3, 717. Collander, R. (1954). The permeability of Nitella cells to non-electrolytes. Physiol. Plant. 7, 420. Fawcett, C. H., Osborne, D. J., Wain, R. L., and Walker, R. D. (1953). Studies on plant growth-regulating substances. VI. Side-chain structure in relation to growth-regulating activity in the aryloxyalkylcarboxylic acids. Ann. appl. Biol. 40,231. Fredga, a., and Matell, M. (1952). Studies on synthetic growth substances. III. The steric relations of the optically active a-phenoxy-propionic acid. Ark. Kemi, 4, 325. Hansen, B. (1954). A physiological classification of 'shoot auxins' and 'root auxins'. I, II. Bot. Notiser, 230, 318. Hanson, J. B., and Bonner, J. (1955). The natttre of the lag period in auxin-induced water uptake. Amer. J. Bot. 42, 41 1. Hellstrom, N. (1953). An attempt to explain the interaction of auxin and anti- auxin in root growth by an adsorption mechanism. Acta chem. scand. 7, 46 1 . Hoffmann, O. L. (1953). Inhibition of auxin effects by 2:4:6-trichlorophenoxy- acetic acid. Plant Physiol. 28, 622. HousLEY, S., Bentley, J. A., and Bickle, A. S. (1954). Studies on plant growth hormones. III. Application of enzyme reaction kinetics to cell elongation in the Avena coleoptile. J. exp. Bot. 5, 373. Ingestad, T. (1953). Kinetic aspects on the growth-regulating effect of some phenoxy acids. Physiol. Plant. 6, 796. Jonsson, A. (1955) . Synthetic plant hormones. VIII. Relationship between chemical structure and plant growth activity in the arylalkyl-, aryloxyalkyl- and indole- alkylcarboxylic acid series. Svensk kem. Tidskr., 67, 166. Kato, J. (1954). Studies on the relation between auxin activity and chemical struc- ture. I. On phenyl- and naphthyl-thioglycolic acid derivatives and related compounds. Mem. Coll. Sci. Kyoto, B21, 77. 114 PflrN=C\ , which mav be considered to take the place of the ring-system in the well-known growth substances. By kindly putting a number of these compounds at our disposal Dr. van der Kerk enabled us to test them under our conditions. Fioure 8 and Table 1 show that the compounds XXIII and [a) ib) (c) Figure 8. Pea test. (a) XXIII : 500, 300, 100, 25, 10, 4.10"=^ /«o///. (b) XXIV : 500. 300, 100, 25, 10, 4.10-^ ?«o///. (c) XXVI : 500, 300, 100, 25, 10, 4.10'^ moljl. XXIV are distinctly active, XXIV being the most active one. When N(CH3)2 is replaced by N(C2H5)2 (XXVI), or C=S by C=0 (XXV), very weak activity or inactivity results, to be expected according to the view expressed by van der Kerk et al., that in these cases the formation of the flat structure with a double bond is favoured less or not at all. In the zj-obutyric acid derivative (XXVII) indications for antagonistic activity were found in the suppression of the normal geotropic behaviour of the roots in the cress-root test. 128 On form and function of plant growth substances The fact that the activity of these compounds is found to be higher in the straight growth test than in the pea test is rather exceptional. It would be worth while investigating whether this might be caused by a higher degree of decomposition in the pea tissue. Van der Kerk et al. have thus shown for the first time that a ring-system is not an absolute requirement for a compound to be active as a growth substance. The polar group apparently may be 'presented' in quite a number of ways, provided that the molecule can assume (or already possesses) the form required for an 'active' fitting on to the receptor. This again suggests to us that the ring-system or its equivalent as a whole is functioning as an attaching unit and that no localized reactive function can be attributed plausibly to certain 'points' of it. GROWTH SUBSTANCES WITH A PYRIDINE NUCLEUS Pyridine compounds have hardly been investigated as growth substances until now. We reported pyridyl acetic acid to be inactive in the pea test (Veldstra, 1952), and supposed that this might be explained by too low a surface activity of the pyridine ring-system (Veldstra, 1944). That this type of compound merits further consideration, however, is shown by recent work of den Hertog et al. on pyridoxyacetic acids and related compounds, part of which has been published (Maas, de GraafT, and den Hertog, 1955). Figure 9. Pea test. Top XXIX: 500, 250, 100, 50.10-5 mo///. Bottom XXX : 100, 50, 25, 10, 4, l.lO-^wo///. 129 Chemical structure and biological activity Prof, den Hertog kindly afforded us an opportunity to test a number of his compounds and with his consent a preliminary account of the results is given (see Figure 9 and Table 1). Cl- -Cl 0=xN/ ^j^/-OCH.3COOH ^j^/-OGH,COOH I CH2COOH XXVIII XXIX XXX Whereas iV-(2-pyridonyl) acetic acid (XXVIII) is inactive, 2-pyridoxy- acetic acid (XXIX) and its 3 : 5-dichloro-derivative (XXX) are active both in the pea test and in the straight growth test, the activity of the latter approaching the level of lAA. It is noteworthy that, as compared with 2:4-D, the pyridine analogue shows a far lower phyto-toxicity, at least towards these tissues. After the other isomeric pyridoxyacetic acids and their derivatives have been investigated, this group of compounds, possibly in connection with the dithiocarbamates, might well enable us to analyse further the details of structure-activity relationships. In these compounds we are dealing with more precise variations in the ring-system and its influencing by the location of the side chain than have so far been studied with naphthalene and indole ring-systems. Their effect on activity may provide a possibility to 'reconnoitre' some aspects of the receptor surface. ON THE FUNCTION OF GROW^TH SUBSTANCES As stated above, for fundamental reasons we cannot expect to solve the problem of auxin function by exclusively studying their form (structure) and its influence on activity. Such work may certainly provide information, however, as to dimensions and nature of the receptor(s) as well as to the type of auxin action. The various interpretations of results obtained in the structure-activity domain have led to different attempts to solve the primary question. We have to admit, however, that despite all these efforts, our joint explora- tions have so far not yielded real basic information as to the mode of action (cf. the reviews by Bonner and Bandurski, 1952, and by Gordon, 1953) ; for, as Gordon has expressed it, 'the biochemical mechanism of auxin action still remains as one of the pressing and challenging problems in plant physiology'. So apparently more data are needed and a constant re-evaluation of those already obtained (even if affecting beloved theories) will be compulsory too. When deciding to shift the accent from an analysis of structure-activity to a biochemical one (in the cell), our way of analysing will obviously be determined to a large extent by the conclusions about the type of action referred to above. Thus, when trying to 'visualize' auxin function as a physico-chemical action of an amphipatic organic anion with a particular spatial pattern, an 130 On form and function of plant growth substances interaction at a functioning macromolecular surface and/or at interfaces in the cell, in our opinion is the most plausible starting point. We had previously, therefore, considered the possibility that auxins might 'start' activity of potential, but 'masked' enzymes (Veldstra and Booij, 1949) (cf. also the discussion by Soding (1953), a theme further elaborated by Booij (1953)). While at the moment no direct evidence in support of these views can be given and work in this domain is still in a preliminary stage, it may be useful to focus attention on other possibilities of a physico-chemical action. Since auxins, in addition to promoting growth, are acting also in differen- tiation (initiation of root formation, control of flower initiation) and are morphological determinants too, it may well be that, apart from actions at protein surfaces, the site of action will have to be looked for 'higher up' in the hierarchy of the cell, viz. with the nucleic acids. In fact, Skoog (1950; 1953) already concluded that the action of auxins in growth is intimately concerned with nucleic acid metabolism (DNA/RNA balance), on account of evidence for lAA-adenine interactions in growth and in organ formation and of lAA effects on nucleic acid content of the cells (cf. also Galston et al., 1953, and Ber, 1953). In this connection we should like to point out the well-known facts that in certain cases the deformations in plants, caused by an over-dosage of growth substances (e.g. by 2:4-D), hardly can be distinguished from the symptoms of a virus infection and that growth substances are able to influence tiie course of virus infection in plants. Here one gets the impression that influencing of synthesis (and/or functioning) of nucleic acids by growth substances are involved or, perhaps, nucleic acid dynamics are interwoven with auxin balances. Moreover, in recent work on animal cell physiology and embryonic development, the possibility emerges that the action of oestrogens may be directly connected with nucleic acid synthesis (Tondury, 1955). Since there are connections between oestrogens and plant growth substances, with regard to type of compounds and importance of spatial structure (cf. the doisynolic acids (Miescher, 1946)) as well as to action (see, for example, Helmkamp and Bonner, 1953), all possibilities of putting the pieces of this jigsaw puzzle together should be sought. While expecting that they may be put together, we are convinced that progress in our understanding of auxin action not only will imply a contri- budon to the solution of one of the main problems of plant physiology, but also to our insight into mechanisms which govern development and differentiation in general. REFERENCES Ber, A. (1953). Studien iiber Auxinwirkung. Endokrinologie, 30, 329. Bonner, J. (1954). The hormonol control of plant growth. The Harvey Lectures, Series XLVIII, I. Bonner, J., and Bandurski, R. S. (1952). Studies of the physiology, pharmacology, and biochemistry of the auxins. Annu. Rev. PL Physiol. 3, 59. Booij, H. L. (1953). Lecture given at Plant Growth Substance Conference, Lund; see review by R. L. Wain, Nature, 172, 710. 131 Chemical structure and biological activity BuRSTROM, H. (1951). Studies on growth and metabolism of roots. VI. The relative growth action of different wobutyric acid derivatives. Physiol. Plant. 4, 470. Galston, a. W., Baker, R. S., and King, J. W. (1953). Benzimidazole and the geometry of cell growth. Physiol. Plant. 6, 863. Gordon, S. A. (1953). Physiology of hormone action, Growth and Differentiation in Plants, ed. W. E. Loomis, p. 253. Greenham, C. G. (1953). Phosphonic acids as auxins and substances affecting growth. Aiistr. J. Sci. 16, 66: see also Maguire, M. H., and Shaw, G. J. chem. Sac. (1953)1479; (1955)1756; Craniades, P., and Rumpf, P. Bull. Sac. chim.Fr. (1952) 1063; (1954) 719; Pastac, J., and Craniades, P. C.R. Soc. Biol, Paris, 148 (1954) 35; Bull. Soc. Chim. biol., Paris, 36, (1954) 675. Helmkamp, G., and Bonner, J. (1953). Some relationships of sterols to plant growth. Plant Physiol. 28, 428. Jonsson, a. (1955). Synthetic plant hormones. VIII. Relationship between chemical structure and plant growth activity in the arylalkyl-, aryloxyalkyl-, and indolealkylcarboxylic acid series. Svensk kern. Tidskr. 67, 166. Kerk, G. J. M. VAN DER, Raalte, M. H. van, Kaars Sijpesteijn, a., and Veen, R. van DER ( 1955) . A new type of plant growth-regulating substances. Nature, 176, 308. (First communication at S.E.B. Conference, Groningen, April, 1955). Maas, J., Graaff, G. B. R. de, and Hertog, H. J. den (1955). The action of diazoacetic ester on pyridone-2. A new synthesis of 2-pyridoxyacetic acid. Rec. Trav. chim. Pays-Bas, 74, 175. MiESCHER, K. (1946). Recherches recentes en Suisse dans le domaine des hormones. Experientia, 2, 237. MiESCHER, K. (1948). On the relation of activity to constitution in the sexogens, with special reference to the doisynolic acids. Recent Progr. Hormone Res. 3, 47. MiNARiK, C. E., Ready, D., Norman, A. G., Thompson, H. E., and Owings. J. F. (1951). New growth-regulating compounds. II. Substituted benzoic acids. Bat. Gaz. 113, 135. MuiR, R. M., and Hansch, C. (1951). The relationship of structure and plant-growth activity of substituted benzoic and phenoxyacetic acids. Plant Physiol. 26, 369. See also Hansch, C, Muir, R. M., and Metzenberg, R. L. Further evidence for a chemical reaction between plant growth regulators and a plant substrate. Plant Physiol. 26 (1951) 812. Skoog, F. (1950). Chemical control of growth and organ formation in plant tissues. Annie biol. 26, 545. Skoog, F. (1953). Substances involved in normal growth and differentiation of plants. Brookhaven Symp. Biol. 6, 1 . Soding, H. (1953). Die Wuchsstoffe und ihre Bedeutung im der hoheren Pflanze. Ber. dtsch. hot. Ges. 66, 383. Thimann, K. V. (1952). The role of ortho substitution in the synthetic auxins. Plant Physiol. 27, 392. ToNDURY, G. (1955). Entwicklungsstorungen durch chemische Faktoren und \ iren. Naturwissenschaften, 42, 312. See also Duspiva, F. ibid. 42, (1955) 305. Zur Biochemie der normalen Wirbeltierentwicklung. Veldstra, H. (1944). Researches on plant-growth substances. V. Relation between chemical structure and physiological activity. II. Contemplations on place and mechanism of the action of the growth substances. Enzymologia, 11, 137. Veldstra, H. (1952). Plant growth regulators. XXI. Structure and activity. VI. Halogenated benzoic acids and related compounds. Rec. Trav. chim. Pays-Bas, 71, 15. Veldstra, H. (1953). The relation of chemical structure to biological activity in growth substances. Anna. Rev. PI. Physiol. 4, 151. Veldstra. H. (1955). Stereochemie in de levende eel. Chem. Weekbl. 51, 158. 132 On form and function of plant growth substances Veldstra, H., and Aberg, B. (1953). Aryl- and aryloxy-acetic acids with a 'bulky' a-substituent. Biochun. biophys. Acta, 12, 593. Veldstra, H., and Booij, H. L. (1949). Researches on plant growth regulators. XVII. Structure and activity. On the mechanism of the action. III. Biochim. biophys. Acta, 3, 278. Veldstra, H., Kruyt, W., Steen, E.J. der, and Aberg, B. (1954). Researches on plant growth regulators. XXII. Structure and activity. VII. Sulphonic acids and related compounds. Rec. Trav. chim. Pays-Bas, 73, 23. Veldstra, H., and Westeringh, C. van de (1952). On the growth substance activity of substituted benzoic acids. Rec. Trav. chim. Pays-Bas, 71, 318. Wain, R. L. (1953). Plant growth substances. The Royal Institute of Chemistryy Monograph No. 2. ^ 133 10 THE MODE OF GROWTH ACTION OF SOME NAPHTHOXY COMPOUNDSf H, BuRSTROM and Berit A. M. Hansen Botany Laboratory, University of Lund INTRODUCTION It has been shown in a previous communication (Burstrom, 1955) that naphthalene acetic and naphthoxyacetic acids increase and decrease root elongation according to a fairly complicated pattern. By testing combinations of these acids it was also shown that some effects of low concentrations of the 2-substituted acids were physiologically independent of the others, and that, at high concentrations, the acids can exert toxic effects which complicate the assessment of real growth activities. Furthermore, 1 -naphthalene acetic and 2-naphthoxyacetic acids inhibited root elongation and were classified as auxins, whereas 2-naphthalene acetic and 1 -naphthoxyacetic acids increased root elongation and antagonized the two others. They should be classified as root auxins according to Hansen's (1954) terminology; they cannot a priori be called anti-auxins in the sense of the semi-official nomenclature (Tukey et al., 1954) because it has not been shown conclusively that they act as competitive inhibitors of auxin. It is also convenient for other reasons to use a term which does not anticipate a special mode of action. Of the two supposed auxins, 2-naphthoxyacetic acid has unanimously been classified as an auxin in the literature (see Hansen, 1954); Street (1955), however, has questioned whether the root-growth effect of 1 -naphthalene acetic acid is physiologically of the same kind. A tentative explanation of why these two acids are auxins and the other two are not has been given by Jonsson (1955). This implies that they are auxins because the side-chain can form a pseudo-ring in the plane of the naphthalene ring with the carboxyl near the centre of the compound ring system. For sterical reasons this is impossible with the other two acids. This provides an amendment to the commonly accepted structural requirements of auxins, giving a reason for the different activities of these four acids. In order to test this explanation, experiments have been conducted with a number of chlorine derivatives of the two auxins, all of which were prepared and kindly suppHed by Dr. Jonsson. The tests with the derivatives of 2-naphthoxyacetic acid have led to some rather unexpected results, which may be of general importance for elucidating the mode of action of these compounds. t This paper was read at the Conference by H. Burstrom. 134 The mode of growth action of some naphthoxy compounds ACTIVITY OF PURE 2-NAPHTHOXYACETIC ACID DERIVATIVES The following substances have been tested in addition to the unchlorinated acid (I): l-chloro-2-naphthoxyacetic acid (II), 3-chloro-2-naphthoxyacetic acid (III), and 1 : 3-dichloro-2-naphthoxyacetic acid (IV). HO I C Crl2 o ^o CI o HO I Cj — CHo II \ o ^o o \ CI I o CHfl C CH2 OH O OH CI CI II III IV According to Jonsson's explanation, III should be an auxin like the unchlorinated acid I, whereas II and IV should lack auxin activity owing to steric hindrance caused by the insertion of chlorine in the 1 -position. 300 Figure 1. The irifliierice on the root cell elongation of chlorinated 2-naphlhoxyacetic acids (2-NOAA). On the ordinate cell length in /i. 200 100- o E-NOAA X 7-Cl- • i-Cl- *1:3-diZ\ t Roofs dj/ng 10-^ 10-^ 10- 10-'* M The effects on cell elongation are shown in Figure 1. It appears that with the exception of the 3-chloro derivative, the acids slightly increase cell elongation up to 3.10~^M, which is a non-auxinic effect shown by the unchlorinated acids (Burstrom, 1955). Their behaviour, however, differs at higher concentrations; thus, the 1 -chlorinated acid gives a considerable increase in elongation, giving an activity curve resembling that of the root auxin 1-naphthoxyacetic acid. The dichloro acid is practically inactive, whereas the 3-chloro compound shows all usual signs of being an auxin. These observations are in good agreement with Jonsson's rules, under which the 1- and 1 : 3-chlorinated acids should lack auxin activity. However, it is impossible from this or any other rule to predict whether an acid lacking auxin activity should possess some kind of antagonistic activity. In this 135 Chemical structure and biological activity instance the dichloro acid is inactive and behaves similarly to that of di- or^/?osubstituted phenoxyacetic acids, to which acids it bears structural resemblances. This seems to indicate that dior//2osubstitution can annihilate all kinds of root-growth activity in a specific manner. In order to corroborate the above classification of the acids, they were investigated in the presence of a supposed true anti-auxin, 3-indolei.yobutyric acid. t^OO 300 ZOO 100 Z-NOAA /-Cl- ' ^J-diCl- -+ro~^^IiBA — Control 10'^ 3x10 -s 10 -5 3x10'^ M Figure 2. The interaction between chlorinated naphthoxyacetic acids and 3-indoleisobutyric acid {liBA). Dotted curves controls without HBA; full-drawn curves with addition of I'lBA. THE INTERACTION WITH 3-INDOLEUOBUTYRIC ACID The results are shown in Figure 2 and can be summarized in the following points. 1. Up to 3.10""^ M neither of the acids has any effect on the growth increase caused by the uobutyric acid ; on the contrary, its growth-promoting effect and the weak one of the naphthoxy derivatives are obviously additive. This is in agreement with the earlier conclusion (Burstrom, 1954) that these effects are of non-auxinic nature. 2. With increasing concentrations of the 1- and 1 : 3-chlorinated acids the effect caused by the addition of the isohutyric acid decreases and becomes nil at 3.10-^ M. 3. For the unchlorinated and 3-chloro acids, the graphs with and without wobutyric acid run parallel within reasonable experimental errors. This means that the activities of these two acids over the whole range of concentra- tions are additive to that of the anti-auxin as nearly as can be expected from two wholly independent kinds of action. The z'j-obutyric acid exerts the same growth-promoting effect in absence of auxin, as when auxin suppresses the elongation by 50 per cent. These results are surprising and contrary to what had been expected. They imply that the supposed auxins are not antagonists of the supposed anti- auxin, but that the root auxin and the inactive acid are antagonistic. 136 The mode of growth action of some naphthoxy compounds This prompted an investigation of the interaction between 3-indoleacetic acid and 3-indolewobutyric acid, the resuk of which is presented in Figure 3. In the best agreement with the last-mentioned resuks it shows that the actions of the two acids are strictly additive. The f^obutyric acid causes a constant promotion of elongation even when the auxin addition has reduced elongation to practically nil. If any conclusions at all can be drawn on the mode of interaction of the compounds from the shape of activity curves, the logical conclusion in the present instance must be that the root-growth-inhibiting auxins and the root-growth promoters of the zjobutyric acid type act in two physiologically different systems, and that they do not simply compete in one single reaction Figure 3. The interaction between 3-indoleacetic acid (lAA) and 3-indole- isobutyric acid (liBA) on cell elon- sation. mediating the cell elongation. Thus 3-indole/5obutyric acid should not be an anti-auxin according to the definition, but a root-growth promoter, i.e. a root auxin, with an action of some other kind. THE LOCATION OF THE ACIDS IN THE CELL-ELONGATION MECHANISM In order to find an explanation of this unexpected result we must turn to the mechanism of cell elongation itself, and see if it affords some ground for an assumption of independent modes of action of these compounds. This mechanism has been studied extensively on roots (cf Burstrom, 1942 ; 1954), and the results have led to a hypothesis summarized in Figure 4. It implies that the elongation takes place in two steps, the first phase involving a passive, plastic stretching of the cell wall, the second phase an active growth of the wall. The first one may be regarded as a preparatory reaction, necessary for the ensuing active growth. This picture is not founded on preconceived opinions of certain actions of auxins, but on observations of cell wall properties during the elongation, the osmotic conditions in elongat- ing cells, their reaction against calcium ions, and so on. There are some reasons to assume, however, that avLxin promotes the first and inhibits the second growth phase, the latter effect usually dominating in roots (Burstrom, 1942). This picture has also been supported by studies of Rufelt (1954) on the geotropic response of roots. V 137 Chemical structure and biological activity This model contains the required two sites where auxin and other regulators could influence elongation independently of each other. Attempts should therefore be made to locate the action of the compounds specifically studied, though such efforts must necessarily be very hypothetical. Assuming that auxin-inhibiting growth acts on phase II, then the physio- logically independent promoting action of root auxins should be located to phase I. This leads, however, to the complication that auxin promotes this part, whilst the root auxin promotes the growth too; consequently they should have the same effect in this respect. This possibility must be seriously considered. A number of observations have been made which indicate a similar action of auxins and root auxins. Maximum elongation / / / Elonootion normally limited tty unknown factor Active watt growth PhaseH {requires calcium and a hypotonic nutrient solution) Growth bf plastic stretching of the ivait (does not require calcium; majgo on until isofonj is neactied) Phase I 4,im/f to which growth isjnhibited ^ non killing^ _ ^^hMlTo/^^yelFng^dJ^ Time Figure 4. Diagram of the suggested mechanism of cell elongation of roots. If isohutyric acids influence the cell-wall tensibility at all, it is in the same direction as auxin, causing an increase, not a decrease as had been expected (Burstrom, 1954). One way of locating the action of the compounds is by means of differential elongation. It has been shown that in cells and tissues of many kinds the elongation is not uniformly distributed in the cells (Burstrom, 1942; Meeuse, 1942). This means that cells elongate only at restricted points, a behaviour which can be demonstrated on roots using the insertion of the root hairs as fixed marking points. Normally, root hair free cells and the basal parts of root hair cells grow at the same rate, but the apical parts much slower or not at all. Indoleacetic acid in high concentrations causes an accentuation of this differentiation (Burstrom, 1942). Indolez^obutyric acid, however, has the surprising effect of increasing the elongation of the basal portions yet relatively retarding the apical parts, as auxin does ( Table 1). With regard to the elongation in a strict sense, auxin and root auxin have the opposite effects, but with regard to the differentiation both behave as auxins. These are two modes of action which can be simply demonstrated under the microscope. What is observed in recording the average cell elongation, not 138 The mode of growth action of some naphthoxy compounds Table 1 The influence of 3-indoleacetic acid {lAA) and oi-{3-indole) isobutyric acid {liBA) on the differential elongation in wheat roots Cell lengths in fi. The concentrations of lAA and I;BA have been chosen so that the combination should give no change in the average elongation. Additions Root hair cells Root hair free cells Average apical part basal part Control 60±1 125±3 258-5 222 Action of lAA without liBA -38 -60 -142 -120 with LB A -30 -53 -157 -120 L /o -57 -45 -58] Action of Ii'BA without lAA -2 + 58 + 163 + 109 with lAA + 6 + 65 + 148 + 109 L /o + 3 + 49 + 60] to mention the bulk growth of an organ, is, of course, the sum of these two different actions. It is not known, however, how these are related to the elongation mechanisms. The ideal additive effect of the two compounds is well shown in Table 1. That the inhibiting action of auxin is physiologically separated from the positive actions of root auxins has been clearly supported by Rufelt (1954), Free (inactive) auxin Phasel /lux in promoting Phases Auxin inlii biting Auxins 3-NOAA 3-Z\,8-N0AA Roof auxins {'anti -auxins') liBA t-Z\2-N0AA and fnacf/ve onfoffonisfs 1:3-diZ\^,E-N0AA (A = auxin i KB A = 3 -indole \i^butyric acid; 2-N0AA=2-napfithox^acetic acid with Z\-derivativesj Tentative diagram of the action of 2-naphthoxy derivatives. Figure 5. Suggested points of action of 3-indole and 2-naphthoxy compounds in the cell elongation process proceeding in two phases {cf Figure 4). who has found that auxin and root auxin act on two different, previously known and described geotropic systems in roots, which may or may not be identical with the two modes of elongation. It can be tentatively assumed, at least as a basis for further discussions, that the actions of the mentioned compounds are located as shown in Figure 5. The root auxins of the z^fobutyric acid type exert growth action on the first phase of the elongation, counteracted by their antagonists, which may be 139 Chemical structure and biological activity inactive in themselves, as is the 1 : 3-dichloronaphthoxyacetic acid, or belong to the same type of growth compounds as the 1-chloro acid. The growth- decreasing effect of auxins should be physiologically independent and located in the second growth phase. This is supported by the observation that the 1-chloronaphthoxyacetic acid does not influence the behaviour of the unchlorinated acid. On the other hand, it is clear that anti-auxins can be exchanged for the native auxin and act as growth regulators in this way. This has been deduced by means of the kinetic methods introduced by McRae and Bonner (1952), by Ingestad (1953), and shown analytically by Fransson and Ingestad (1955) on coleoptiles. A way out of this dilemma is to make the rather natural assumption that these compounds all act on both phases of the cell elongation, either by competing with the native auxin or by virtue of their own inherent activity. Owing perhaps to minor details in structure, the effect in one or the other system dominates, making one substance an auxin assumed to inhibit predominately growth phase II, another a root auxin when the action on phase I dominates, or a third substance an anti-auxin if it acts mainly by competing with natural auxin. Such details in the activity do not show up in the net elongation of cells or organs, which is the usual observation recorded, but for correlating structure with activity it may be necessary to pay due regard to the different ways the compounds can act on the growth mechanism. REFERENCES BuRSTROM, H. (1942). The influence of heteroauxin on cell growth and root develop- ment. Ann. agric. Coll. Sweden, 10, 209. BuRSTROM, H. (1954). Studies on growth and metabolism of roots. XI. The influence of auxin and coumarin derivatives on the cell wall. Physiol. Plant. 7, 548. BuRSTROM, H. (1955). Evaluadon of the growth activity of naphthalene derivatives. Physiol. Plant. 8, 174. Fransson, P., and Ingestad, T. (1955). The effect of an antiauxin on the indoleacetic acid content in Avena coleoptiles. Physiol. Plant. 8, 336. Hansen, B. A. M. (1954). A physiological classification of 'shoot auxins' and 'root auxins'. I-II. Bot. Notiser 230, 318. Ingestad, T. (1953) . Kinetic aspects on the growth regulating eff'ect of some phenoxy acids. Physiol. Plant. 6, 796. Jonsson, a. (1955). Synthetic plant hormones. VIII. Relationship between chemical structure and plant growth activity in the arylalkyl, aryloxyalkyl, and indolealkylcarboxylic acid series. Svensk kern. Tidskr. 67, 166. McRae, D. H., and Bonner, J. (1952). Diortho substituted phenoxyacetic acids as antiauxins. Plant Physiol. 27, 834. Meeuse, a. D. J. (1942). A study of intercellular relationships among vegetable cells with special reference to sliding growth and to cell shape. Rec. Trav. bot. need. 38, 18. RuFELT, H. (1954) . Influence of growth substances on the geotropic response of roots. Physiol. Plant. 7, 141. Street, H. E. (1955). Factors controlling meristematic activity m excised roots. VI. Effects of various 'antiauxins' on the growth and survival of Lycopersicum esculentum, Mill. Physiol. Plant. 8, 48. TuKEY, H. B., Went, F. W., Muir, R. M., and Overbeek, J. van (1954). Nomen- clature of chemical plant regulators. Plant. Physiol. 29, 307. 140 CHEMICAL CONFIGURATION AND ACTION OF DIFFERENT GROWTH SUBSTANCES AND GROWTH INHIBITORS: NEW EXPERIMENTS WITH THE PASTE METHOD H. LiNSER Biologisches Laboratorium der Osterreichische Stickstoffwerke A.G., Linz, Austria In 1953 at Lund, Sweden, I reported on my paste method, and on the difference between the types of concentration-action curves obtained with two types of chemicals, which may be called growth substances and growth inhibitors. Although we commonly use these two words, we are not certain if all the compounds we usually call growth substances would give the typical curve for growth substances in the paste test. There may be some compounds, classed as growth substances, which give a concentration-action curve of the growth-inhibitor type. We could show that every concentration- action curve of a cell-elongating growth substance is composed of two components, one promoting, and one inhibiting. Our paste method, which uses an intact coleoptile, is able to reveal both growth-promoting and growth- inhibiting activity in the same experiment. The two components, the pro- moting one and the inhibiting one, can be of different magnitudes. In some special cases, the promoting activity of a growth substance can be reduced to zero. Therefore we may say that a growth inhibitor is a growth substance with a promoting component of zero. In the same way we can define a growth substance as a growth inhibitor with a growth-promoting component (larger than zero). Some experiments with mixtures of a promoting and an inhibiting substance were described at Lund, and I reported in Paris (Linser, 1954a) on further similar experiments. Dr. Kaindl's paper at this Symposium is concerned with the results of these experiments and a mathematical examina- tion of them. Experiments have recently been carried out using mixtures of different growth substances with one another and different inhibitors with one another. The results obtained will be published at a later date. The earlier experiments provided evidence that there is competition between growth substances and growth inhibitors in filling a space, which we call the 'gap', which exists in the living system. The probability of filling this gap is determined by the affinity between the molecule of the active substance and the gap (Linser, 1954b,c). We know that this affinity depends mainly on the chemical nature of the compound, or on the physical properties of the atoms of which the molecule is composed and on its electronic configuration. It also depends, though to a lesser extent, on the size and form of the molecule, i.e. its space-filling capacity. In view of the above considerations, we compared our results with the 141 Chemical structure and biological activity chemical structure of the substances we found to be active as growth pro- moters and inhibitors. Although we know that the form of a molecule is only one of the different properties which must be taken into consideration if we want to get a complete knowledge of the relation between activity and chemical structure of different substances, nevertheless, we think that this point of view has so far been too much neglected. It will be seen that there are some surprising similarities in the form of substances of similar physiological action which are very different chemically. We cannot believe that the factor 'form' is of no importance when a molecule enters a living system; indeed, we have good reason to assume that the analogy of key and lock is Figure I. The concentration-action curves of indole-3-acetic acid (lAA) and indole-i- acetonitrile {IAN) in the paste test. log concn. very relevant. We consider that each molecule must be well matched with its corresponding gap by its form, and by its electrical and chemical configuration. In this communication, I will first present some recent results we obtained by investigating the activity of different synthetic substances using our paste method. Secondly, I will give a comparison of some values for the affinities and other significant constants which we evaluated using Kaindl's formula for the concentration-action curves of growth substances and growth inhibitors (Kaindl, 1954). In addition, photographs are included showing similarities of the molecular form and size of different growth substances and growth inhibitors. The first important growth substance examined was indole-3-acetonitrile, a compound first isolated from natural sources by Jones and his co-workers (1952). It was tested in comparison with indole-3-acetic acid and showed a similar concentration-action curve, though it was found to be about ten times more active than indole-3-acetic acid {Figure 1). I have to thank Prof. Jones of Manchester and Dr. Thoma of Linz for the two samples of 142 Chemical configuration and action. Paste method nitrile tested, both of which gave the same resuks. I also wish to thank Dr. J. A. Bentley of Manchester, who sent samples of 2:3:6-trichlorobenzoic acid and 2:3:6-trichlorobenzaldehyde, and, for comparison with these, the Figure 2. The concentration-action curves of some halogen-substituted benzoic acids and aldehydes in com- parison with indole-3-acetic acid. TBA = trichlorobenzaldehyde ; TBS = trichlorobenzoic acid; lAA = indole-3-acetic acid. V 6 'TBA 2 9 5 'TBA 3 9 5-TBA log concn. 2:4:6-, the 2:4:5-, and the 3:4:5-trichlorobenzaldehyde. The first two substances, as observed by Bentley (1950) showed growth-promoting activity, but nearly ten times less than indole-3-acetic acid, whereas the other three trichloroaldehydes showed a small activity as growth inhibitors [Figure 2). +eo \¥0 I S: to '20 /"^ / \ 1 \ 1 \ 1 \ 1 \ 1 1 1 V \ 1 ^^\ / /^\\ V / 1 \\ i-ai ^ \\ ^ ^^mrf 1 1 1 1 VDTP Figure 3. The concentration-action curves of 'N-dimethylthiuramacetic acid (ND TE) , NN-dimethylthiurampro- pionic acid (NDTP), and indole-3- acetic acid {I A A) in the paste test. log concn. 10" The newly discovered growth substance iViV-dimethylthiuramacetic acid (Kerk, 1955) was kindly provided by Dr. van der Veen of Eindhoven. This gave a small growth-promoting activity in the paste test [Figure 3) nearly 143 Chemical structure and biological activity 1/100 of that of indole-3-acetic acid. Weaver (1952; 1954) observed that benzothiazol-2-oxyacetic acid applied to grape plants at flowering resulted in a high percentage set of rather small fruits. We tested this substance with Figure 4. The concentration-action curves of 2-oxybenzthiazoleacetic acid (BTOES) and indole-'i-acetic acid {lAA). 10-" 10" log con en. the paste test and found slight growth-promoting activity, about 1/100 that of indole-3-acetic acid {Figure 4). In all, we have tested about one hundred different synthetic substances which seemed to be of some interest in comparing the form of the molecule with its activity. Figure 5. The concentration-action curves of 3-arnino-triazole (3 AT), imidazole-4-acetic acid (Im-4-£), and indole-'i-acetic acid {lAA). + (?^ *eo 9^ *¥0 % :S ^20 I 5 -20 lAA 10 -w --2 10-'- 10" log concn. In further experiments we found 3-aminotriazole to possess a very small activity not more than 1/1000 of indole-3-acetic acid {Figured). iViV-phenyl- methylglycine ethyl ester showed a good activity as a growth promoter, whereas that of 2-methylbenzimidazole and 2:4-dichlorophenoxyacetic acid-bisoxyethylamide was slight. In this group of 2:4-D derivatives, in 144 Chemical configuration and action. Paste method addition to all the different salts and esters of the 2:4-dichlorophenoxy acetic acid, the following substances were active as growth promoters: 2:4-dichlorophenoxyacetamide, 2:4-dichlorophenoxyacetic acid hydrazide, 2:4-dichlorophenoxyacetonitrile, and 2:4-dichlorophenoxymaleic acid hydrazide. The group of 2:4:5-trichlorophenoxy derivatives which showed promoting activity included 2:4:5-trichlorophenoxyacetamide, 2:4:5- trichlorophenoxyacetic acid hydrazide, 2 :4:5-trichlorophenoxyacetonitrile, 2:4:5-trichlorophenoxyacetaldehyde, and sodium 2:4:5-trichlorophenoxy- acetamidoethylsulphate showed slight activity. Many other substances showed strong inhibiting activity, like indazole, 3-methylindazole, 1-methyl- benzotriazole, 1-naphthoxyacetic acid, 4-oxycoumarin-3-butyric acid, salts of the 3:6-endoxotetrahydro-o-phthalic acid, the zj^opropyl ester of carbamilic acid, cinnamic aldehyde, 3:5-dinitro-2-methylbenzoic acid, 4:5-dinitro- 2-methylbenzoic acid, the lactone of the /j-oxy-/3-o-carboxyphenylpropionic acid, aa-dichloropropionic acid, allyl cyanoacetate, sodium fiopropylxan- thogenate, naphthoquinone, sodium 2:5-dichlorophenylacetate, sodium 2:4: 5-trichlorophenylacetate, ethyl 2:4: 5-trichlorothiophenoxyacetate, pentachlorophenoxyacetic acid and its esters. A tetrachlorophenoxyacetic acid product whose identity is not known exactly showed a small promoting activity with a strong inhibiting component. It is of some interest, that 4:6-dichlororesorcin-bis(carboxymethyl) ether and phenylimidodiacetic acid {Figure 6) were found to be practically inactive, whereas bis-2:4-dichlorophenoxyacetic acid {Figure 7) proved to be a strong inhibitor. This latter substance is of special interest, because it is somewhat like a Siamese twin, the two 2:4-dichlorophenoxy rings being connected by one acetic acid. In respect of the two-point attachment theory, the assump- tion could be made that this substance must be active as promoter, because the acid group is free as is an ortho position in one or both rings. No promoting activity, however, nor any promoting component within the concentration- action curve could be discovered; indeed, the curve we found was a purely inhibiting one. The two-point attachment theory is here forced to make an additional assumption: a steric hindrance. In other words, it must be assumed that the steric form of other parts of the molecule are as much involved in its activity as the 'two points' which are supposed to be necessary for the growth-promoting action. In our view, it is more useful to consider the form of the molecule in its entirety than only in terms of two separate points. Nevertheless, I do consider it a good idea to nominate two distinct points of the molecular structure as responsible for initiating the growth- promoting action, provided that we assume that the association which takes place between each of these points and the reacting living system occurs by chemical reactions establishing real chemical bonds. But even in the case of chlorinated phenoxyacetic acids, it is not obvious that the free 6-position will give a true chemical reaction with another reacting group of the living system. It therefore seems much better to assume that the 6-position is involved in mode of action in a way other than by chemical reaction. The terminal group of the side-chain of a growth substance can be changed within a wide range without losing activity. Thus we find the nitrile group to be more active than the acid group, the latter, however, having a higher activity than the aldehyde group. If we assume a chemical reaction to be V 145 Chemical structure and biological activity involved in the mode of action of these substances, it is very improbable that such action can be accomplished by several different chemical reactions; therefore we are forced to conclude that the different active substances are converted within the plant into one certain form. Thus, either the aldehyde CI 0-CH2-C00H 0-CHaCOOH CI fi^^O-CH^COOH 0-CH2-C00H < CHz-COOH CH2-C00H / n m Figure 6. The chemical formulae of dichloropyrocatechol-acetic acid (/), 4:6-dichlororesorcin- biscarboxymethyl ether (II), and phenylimidodiacetic acid {III). These substances were found to be practically inactive in the paste test. CI :CH-COOH CI Bis-2:'(-dichlorophenoxyacetic acid I I I I L 10'* 10^ 10^ 10^ 10° lo" Figure 7. The molecular model of bis-2 : A-dichlorophenoxyacetic acid with the concentration-action curve and formula. Abscissa: \o^ concn. Ordinate: change in length of coleoptiles as per cent of control. % and nitrile are converted to the acid, or more probably, the aldehyde goes to the acid and the acid into the nitrile, the latter being the only reacting form which acts directly. There is, however, another possible explanation of the mode of action; we can assume, as we did in earlier publications, that the action of a growth 146 Chemical configuration and action. Paste method Table 1 The Xy^ and Kg values of different growth substances and growth inhibitors Group Substance A„. X IQi* kg X 1015 Indole derivates Indole-3-acetic acid 7-9 0-0051 Indole-3-propionic-acid 0-525 0-0204 Indole-3-butyric-acid 0-86 0-016 Indole-3-acetonitriie 70-7 0-042 Indole 0-0379 0-00226 a-Methylindole 0-000 0-0414 /5-Methylindole (skatole) 0-0273 0-00291 Tryptophan 0-000 0-000 Naphthalene a-Naphthaleneacetic acid 1-36 0-0215 derivates a-Naphthaleneacetic acid (sodium salt) 2-08 0-0594 a-Naphthaleneacetic acid methyl ester 2-31 0-0098 Naphthoxy a-Naphthoxyacetic acid 0-000 0-0342 derivates /3-Naphthoxyacetic acid 0-000 0-144 Phenoxyacetic acid Phenoxyacetic acid 0-000 0-0948 derivates /)flraChlorophenoxyacetic acid 0-571 0-0136 2:4-Dichlorophenoxyacetic acid 2-05 0-102 2 :4-Dichlorophenoxyacetic acid(sodium salt) 3-4 0-00187 2:4-Dichlorophenoxyacetic acid butyl ester 10-7 0-0940 2:5-Dichlorophenoxyacetic acid 5-95 0-034 2:5-Dichlorophenoxyacetic acid(sodium salt) 1-67 0-00169 2:5-Dichlorophenoxyacetic acid butyl ester 7-03 0-0185 2:4: 5-Trichlorophenoxyacetic acid 23-6 0-0256 2:4:5-Trichlorophenoxyacetic acid (sodium salt) 9-2 0-199 2:4:5-Trichlorophenoxyacetic acid butyl ester 9-84 0-144 2 :4:6-Trichlorophenoxyacetic acid 0-000 0-1764 2:4:6-Trichlorophenoxyacetic acid (sodium salt) 0-000 0-00384 2:4:6-Trichlorophenoxyacetic acid butyl ester 0-000 0-139 Tetrachlorophenoxyacetic acid ethyl ester 1-35 0-132 Pentachlorophenoxyacetic acid 0-000 1-47 Bis-2 :4-dichlorophenoxyacetic acid 0-000 2-49 Other growth 2:3: 6-Trichlorobenzaldehyde 7-4 0-0435 promoters 2:3:6-Trichlorobenzoic acid 2-6 0-0139 A'-Dimethylthiuramacetic acid 0-145 0-00542 Benzthiazole-2-oxy-acetic acid 0-209 0-00322 3-Amino-triazole 0-00337 0-000 Growth inhibitors 2:3:5-Triiodobenzoic acid — 3-09 2 :4:6-Trichlorobenzaldehyde — 0-0290 2:4: 5-Trichlorobenzaldehyde — 0-0515 3:4: 5-Trichlorobenzaldehyde — 0-161 4:6-Dinitro-o-cresol 806-0 1-2 Dinitro-i^c.butylphenol — 1-174 Pentachlorophenol 5-13 0-451 uoPropylphenylcarbamate — 0-0304 Fluorescein Sodium fluorescein 0-000 0-000 derivates Tetrachloro-fluorescein 0-0577 6-13 Tetrabromo-fluorescein (cosine) — 49-5 Tetraiodo-fluorescein (erythrosine) 0-99 23-8 147 Chemical structure and biological activity substance occurs by filling in a gap in the structure of the appropriate part of the growing system, not by means of chemical reactions and the formation of new chemical bonds, but only by the action of intermolecular forces. The intensity of these forces, which depends on the gap as well as on the molecule fitting in, is called 'affinity'. This affinity will be high, if the form of the gap agrees with the form of the molecule, or, to be more precise, if the form of the gap agrees with the form of the active group of the molecule with which it is to be filled. The molecule is able to enter into the gap only when its form corresponds to that of the gap. It follows that it must be of some importance to measure the affinity (which we accept as hypothesis for our heuristic purpose) of different growth substances and inhibitors, as well as of different chemically related inactive substances. The mathematical formulation of the concentration-action curves evaluated by Kaindl (1954) enables us to obtain values from the experimental data of these curves, which are called X and are proportional to the size of the affinity between gap and molecule. These values for A are designated A„- for a growth-promoting part of the molecule and as Ijj for the growth-inhibiting part of the same molecule. We do not stipulate that these two parts must be separated. They may be partly or entirely the same. These values "k^r and Ij^ have been calculated for the main types of growth substances and growth inhibitors and are shown in Table 1. The most active growth substances show the highest affinities in the order shown in Table 2. Apart from the differences in the side chains with the different salts and esters of acid products, which may be altered by the plant metabolism and which lead to differences in ease of penetration into the cells, the most active promoters show the following order: 2:4:5-trichlorophenoxyacetic acid, indole-3-acetic acid, 2 : 5-dichlorophenoxyacetic acid, 2:3:6-trichlorobenzoic acid, 2:4-dichlorophenoxyacetic acid, a-naphthaleneacetic acid, tetrachlorophenoxyacetic acid. It is surprising that the 2: 5-dichlorophenoxyacetic acid, which has no strong weed-controlling properties, shows a higher affinity Ajj- than the very active weed-killer 2:4-D. The same is the case with the inactive weed killer indole-3-acetic acid, which shows a higher Ay,- value than 2:4-D. The most active growth inhibitors can be ranged as shown in Table 3. We see that the highest affinity of a growth inhibitor is only about half the highest affinity of a growth promoter. The affinities of the inhibiting group of growth substances (A^^) range between the affinities of the different growth substances as shown in Table 4. This table shows that the most active growth promoters possess A^ values which are very small, compared with those of most of the inhibitors, but high enough to show effects as the less active inhibitors do. It is not easy to correlate these values for the affinities and the steric forms of different active molecules, because of the difficulties in comparing different molecular shapes. The main difficulty is that the organic molecules with a molecular weight of 100-500 have no fixed form. Free rotation is possible about certain bonds and their actual forms depend on the relative positions 148 Chemical configuration and action. Paste method Table 2 The A„- values of the most active growth promoters in decreasing order Substance X,y X 1015 Indole-3-acetonitrile 70-7 2:4:5-Trichlorophenoxyacetic acid ethyl ester 46-3 2:4:5-Trichlorophenoxyacetic acid 23-6 2:4-Dichlorophenoxyacetic acid butyl ester 10-7 2:4:5-Trichlorophenoxyacetic acid butyl ester 9-8 2:4:5-Trichlorophenoxyacetic acid (sodium salt) 9-2 2:4-Dichlorophenoxyacetic acid methyl ester 8-5 Indole-3-acetic acid 7-9 2:3: 6-Trichlorobenzaldehyde 7-4 2 :5-Dichlorophenoxy acetic acid butyl ester 7-0 2:5-Dichlorophenoxyacetic acid 5-0 2:4-Dichlorophenoxyacetic acid (sodium salt) 3-4 2:3:6-Trichlorobenzoic acid 2-6 2:4-DichIorophenoxyacetic acid 2-0 a-Naphthalene-acetic acid (sodium salt) 2-0 a-Naphthalene-acetic acid 1-4 Tetrachlorophenoxyacetic acid 1-3 Indole-3-butvric acid 0-9 Indole-3-propionic acid 0-5 2-Oxybenzthiazole-acetic acid 0-2 A'^-Dimethylthiuramacetic acid 0-1 Table 3 The Kjj values of the most active growth inhibitors in decreasing order Substance Ig X 1015 Tetrabromofluorescein (cosine) 49-5 Tetraiodofluorescein (erythrosine) 23-8 Tetrachlorofluorescein 6-13 2:3:5-Triiodobenzoic acid 3-0 Bis-2 :4-Dichlorophenoxyacetic acid 2-5 Pentachlorophenoxyacetic acid 1-5 Dinitro-o-jw.butylphenol 1-2 Pentachlorophenol 0-45 2:4:6-Trichlorophenoxyacetic acid 0-18 /S-Naphthoxyacetic acid 0-15 a-Naphthoxyacetic acid 0-03 Phenoxyacetic acid 0-009 149 11 Chemical structure and biological activity Table 4 The Xu values of the most active growth promoters between the A^ values of the most active growth inhibitors Substance ?.„ X 101-^ Tetrabromofluorescein 49-5 2:3:5-Triiodobenzoic acid 3-0 /^-Naphthoxyacetic acid 0-15 2:4:5-Trichlorophenoxyacetic acid ethyl ester 0-27 2 :4:6-Trichlorophenoxyacetic acid 0-18 Indole-3-acetonitrile 0-042 2 :4:5-Trichlorophenoxyacetic acid 0-025 a-Naphthalene-acetic acid 0-021 Phenoxyacetic acid 0-009 Indole-3-acetic acid 0-005 Growth substances Indole-3-acetonitrile CHjCN Indole-3-acetic acid CHjCOOH 2:'»-Dichlorophen- oxyacetic acid CI 2:t:5-Trichlorophen- oxyacetic acid CI ■ Cl< >0CH2C00H CI oi-Naphthylacetic acid CHjCOOH 2:3:6-Trichlorobenzoic acid CI OCH2COOH CI a a COOH Figure 8. Molecular models of six of the most important growth-promoting substances. 150 Chemical configuration and action. Paste method of the revolving groups to one another. These positions, however, are determined by the environs of the molecule, and the nature of these, at site of action, is not known. We also have no information on the required position of the molecule at site of action, nor do we know how to find this active position, which could be supposed to correspond with the gaps of the reacting system. It seems reasonable to assume that different growth substances probably act in the same way on the same cell-elongating svstem, and must therefore be able to assume the same active form. Figure 9. Molecular models of some newly discovered growth promoters. Top: 3-amino-triacole. Middle: X- dimethyl-thiwamacetic acid (van der Veen). Bottom: 2-oxy benzthiazole- acetic acid. In view of the above considerations, it would seem important to try to find the positions of the more or less free rotating side-chains of different growth substances which show similar activity to one another. A long series of trials with the molecular models constructed by Stuart and Briegleb showing the space occupied by every single atom within the molecule magnified 1:1-5.108, revealed that the main growth substances with acetic acid as a side-chain are able to assume very similar forms. These similarities are shown in Figure 8. For the purpose of this comparison, we selected a standard position of the side-chain relative to the ring- (or body-) system of the molecule. The criterion of this position is that the axis of the end group of the side chain hes at an angle of 90° to the longer axis of the ring-system of the molecule {Figure 13). In the case of 2 : 4 : 5-trichloro- phenoxyacetic acid this position is possible and the COOH-group remains free to rotate even in this position ; in the case of 2 : 4 : 6-trichlorophenoxy- acetic acid, however, this is impossible. We do not know if this selected 151 Chemical structure and biological activity position is the same as that which is necessary for activity and we use it only hypothetically as the standard position for the purpose of comparison. We see {Figures 8, 9) that some other growth-promoting substances like 2:3:6- trichlorobenzoic acid or 3-aminotriazole, are not able to fall into this position, yet a chemical substance without any ring system, such as A'A'-dimethyl- thiuramacetic acid {Figure 9), in spite of the completely different chemical Id' id' z % 20 - '^ -20 -to - \ a I y^ °>CH-COOH \ CI -60 - \ Bis-2:'t-dichlorophen- \ oxyacetic acid ■SO \ 1 i._J ► W" 10^ 10^ f^ 10° 10^ % Fioiire 10. The molecular model of the 'twin' bis-2 ■A-dichlowphenoxyacetic acid in comparison with the molecular model of tetrachlorofluorescein. Abscissae: log concn. Ordinates : change in length of coleoptiles as per cent oj control. Structure, is able to fall into the standard position. A molecule like the Siamese twin (bis-2 : 4-dichlorophenoxyacetic acid) is, in spite of its chemically related structure, not able to fall into the standard position and shows no promoting action. On the other hand, this twin molecule is able to fall into another position which is quite similar to that of the substituted fluorescein molecule {Figure 10) and we know that substituted fluoresceins, such as tetrachloro-, tetrabromo-, and tetraiodofluorescein, all show growth- inhibiting actions {Figure 11) which are stronger than those shown by the twin molecule. Comparing the molecular forms of the unsuJDStituted and 152 Chemical configuration and action. Paste method N cz \ CO ^-)-D^{\-e-h<^2)-D^{\-e<''2) gives the function of an inhibiting substance. When the two substances are simultaneously applied to the coleoptile the following competing actions take place : 1. The probability that the 11^-molecule enters into a w-gap and that an //-molecule does not enter into this gap is given by P^ := (1 — e~^'«'''i)e'^«'''2, assuming that the probability of replacement 4, valid for the ousting process of a natural molecule holds approximately true for this corresponding ousting process too. 2. There is competition between H'- and //-molecules for the ousting of a natural molecule from a w-gap. The probability of the occupation of the !:f;-gap by a M'^-molecule is given then by Pg = (\—e-^-:r'^i)e-'>2 and for an //-molecule by 2^= (1— ^-/;/2)^-^->i. 3. For the IT-molecule competition of the two partners for an A-gap leads to the probability P3 = (^\—e-^-n<'i)e-hc2; for the //-molecule to 4. Finally, competition regarding the ousting process of a natural molecule from an A-gap leads to the two probabilities P^ = (^l—e-K<-i)e-^'>.'^2 for the py-molecule and 4P = (1— h;^2)^-^-a'-i for the //-molecule. The function of the mixture curve, therefore, has to be ^u,+^H = A,P,+A,P,-B,P,-B,P,^C, . ,P-D, . ,P-D., . ,P. The results calculated by this formula show a good agreement with the experimental data obtained in our biological laboratory. For Figures 4-6, I have singled out one curve of the mixture series indoleacetic acid (10~^) 162 Action-concentration curves I '^100 -20 -¥0 S -60 2:3:5— friiodobenzoic acid 70"* w-3 /^-2 \T/(?-i 70° log concn. Figure 4. The course of the theoretical concentration-action curve of the mixture of lAA (10^^ per cent concn.) and 2:3: b-triiodobenzoic acid compared with the experimental data (T). Abscissa and ordinate as in Figure 1. _L ■2:3:5-friiodobenzoic acid Figure 5. The course of the theoretical concentration-action curve of the mixture of oc-naphthvlacetic acid (I0~^ per cent concn.) and 2:3:5-triiodobenzoic acid compared with the experimental data (T). Abscissa and ordirmte as in Figure 1. i. I y20 1 t- -20 ^ -30 cc- naphthylacefic acid 10' T 10-'* it)-^ log concn. -^ Z:ii:e-fp/'ch/onophenoxy- acefic acid Figure 6. The course of the theoretical concentration-action curve of the mixture of on-naphthylacetic acid {\0~^ per cent concn.) and 2:4:6-trichlorophenox)iacetic acid compared with the experimental data (T). Abscissa and ordinate as in Figure 1. j_ 163 Chemical structure and biological activity and 2:3: 5-triiodobenzoic acid, one curve of the series a-naphthylacetic acid (10~^) and 2 : 3 : 5-triiodobenzoic acid and one curve of the series a-naphthyl- acetic acid (10~^) and 2:4:6-trichlorophenoxyacetic acid. In spite of the good agreement between the theoretical function and the experimental data, there are few deviations, especially when higher concentrations are applied. Figure 7 shows such a case (2:4:5-trichlorophenoxyacetic acid (10") and 2:4:6-trichlorophenoxyacetic acid). This deviation seems to be based on the fact that either in the mixture itself, or in the interior of the coleoptile, chemical reactions or intermolecular associations take place. The latter may 1 ^20 log Doncn. ro~^ 70'^ 7^-2 /■^-l _l I L 2:^^:6- fp/c/)/orophenoxj'aceh'c acid 2--1-5-fnichlorophenoxyocefic acid 10° Figure 7. The course of the theoretical concentration-action curve of the mixture of 'I'A'.b-trichloro- phenoxyacetic acid (10" per cent concn.) and 2:4: 6-trichlorophenoxj>acetic acid compared with the experi- mental data (T)- Abscissa and ordinate as in Figure 1. ± be more probable, and means formally that both acting molecules enter into a gap simultaneously. Even in the case shown in Figure 7 molecidar associa- tion seems to be present because the two molecules are chemically similar. The effect brought about by the simultaneous entry is not to be calculated from the fundamental functions of the single partners of the mixture, but one can obtain four correction terms. I will not occupy space here to show this in detail. Further investigations will reveal whether the conclusion is reversible; a deviation from the calculated values would mean that inter- molecular powers influence the result. In terms of the probability calculus, the two collectives are not independent of each other when experimental results deviate from theoretical results. What I want to emphasize, however, is that the instances in which such discrepancies between theory and experiment appear are very few in the cases where I have had the opportunity of calculating mixture series. Further calculations will have to support the considerations of probability applied here, but the results obtained so far indicate the usefulness of the theory based on the fundamental ideas of H, Linser (1954). REFERENCES Kaindl, K. (1954). Biophysikalische Analyse der Konzentrations-Wirkungskurven von Wirkstoffen (Insbesondere Zellstreckungswuchsstoffen). Alh. Chem. 85, 985. Linser, H. (1954). Chemische Konstitution und zellstreckungswirkung. Versuch einer allgemeinen Wirkstoffhypothese. Mh. Chem. 85, 196. 164 THE CHEMICAL INDUCTION OF GROWTH IN PLANT TISSUE CULTURES! F. C. Steward and E. M. Shantz Department of Botany, Cornell University Part I: Methods of Tissue Culture and the Analysis of Growth introduction: normal proliferative growth in the plant body Before taking note of various chemical substances and extracts that are capable of inducing rapid, proliferated growth in otherwise mature plant tissues, it is well to take note of a few salient examples in which this occurs normally in the plant body. The classical example, noted by Fitting (1910), in which pollination in the orchid stimulates the growth of the pistil or ovai-y wall, is well known. In the development of the coco-nut fruit the embryo is usually immature when the fruit falls to the ground. When the embryo eventually germinates it sends out into the central cavity containing liquid endosperm a cotyledon which acts as an absorbing organ and rapidly produces a cellular tissue capable of filling the entire cavity of the fruit. The stimulation which pollination and fertilization gives to the development of the fleshy tissue of pome fruits is well known. There are also cases, e.g. the edible banana, in which, under certain genetical situations, a fleshy layer develops without this stimulus and grows parthenocarpically by cell division and proliferation from a well-defined loculus in the fruit. The numerous cases now known of parthenocarpically induced development of fruits give point to the chemical basis of this stimulus to growth. While the formation of tubers and tuberization and the development of other storage organs, whether roots or bulbs, is often a response to length of day, this stimulus must be communicated to the cells in question by some as yet unknown chemical means. Galls, whether activated by bacteria, by insects, or by viruses, again induce in otherwise mature cells a return to active growth and proliferation. The formation of nodules, which grow on legumes, is yet another familiar example of chemically induced growth in the tissues of the host. In other words, although the following account will concentrate upon certain phenomena in which this general type of process may be studied in plant tissue cultures, there are many examples that can be cited in which similar phenomena occur normally in the development of plants. The roison d'etre of the investigations to be described was, however, not wholly the attempt to understand the chemical basis for growth induction. It was rather to use the chemical means by which growth may be so induced to contrast the behaviour of rapidly growing and non-growing or resting cells. For this reason attention was turned to the use of tissue culture methods t This paper was read at the Conference by F. C. Steward. 165 12 Chemical structure and biological activity and these were adapted by means that have been described to experiments that can be carried out in a quantitative manner. The essential outlines of these experiments only will be described here because the methods have been published elsewhere (Caplin and Steward, 1949; Steward, Caplin, and Millar, 1952). In making this summary of the technique, however, an opportunity will be taken to mention certain refinements that have not hitherto been published in full. TISSUE CULTURE METHODS The tissue of which the greatest use is made in these investigations is the secondary phloem tissue of the carrot root. Essentially the same methods have, however, been applied to the culture of artichoke tuber tissue and of potato tuber tissue. All of these tissues are essentially secondary in origin but the extent to which they may embark upon rapid, external, proliferative growth is determined by the conditions to be described: these include the use of certain isolated and now-known substances or the use of extracts that contain growth-promoting substances. The tissue of the storage organ is isolated aseptically in the form of small pieces, or cylindrical 'plugs', which weigh approximately 2 mg. Special steps are taken to ensure that the tissue explants are removed from as closely identical positions in the organ as possible. An essential feature of the tech- nique is that, in given experiments, explants are all removed from the same individual organ, be it root or tuber. In this way variability is kept to a minimum. In our experience there is no method by which already cultured tissues can be subdivided in a random fashion to produce populations of explants that are as uniform in their subsequent growth as the tissues isolated from the intact organ by the means which have been described (Caplin, 1 947) . It is not necessary to enlarge here upon the accuracy with which these techniques can be performed although one may refer to the discussion that has taken place regarding the use of relatively large, or relatively small, explants (Heller, 1953). The use of small explants is in these experiments essential to exclude those minute centres of growth that would permit the tissue to respond without the addition of the particular formative substances that are here under investigation. We can merely say that in our hands the use of small explants, and even transfers of isolated floating cells from a suspension, is capable of giving a greater degree of uniformity than any other means known to us. The tissue culture procedure which is preferred uses growth of the tissue explants in a liquid medium contained in special tubes arranged around a horizontal shaft revolving about one revolution a minute, so that the tissue spends part of its time in air and part of its time bathed in water. The type of growth curve obtained under these conditions with carrot phloem explants has been described (Caplin and Steward, 1949). The familiar criterion of growth in these experiments has been the change in fresh weight of the individual explant. The medium in which the best growth has occurred has been a standard tissue culture medium (White, 1943), supplemented by whole coco-nut milk, the liquid endosperm from the relatively mature nut (Caplin and Steward, 1948). With suitable modifications of the medium, potato tissue can be successfully cultured in a similar way (Steward and 166 Chemical induction of growth [a) Culture fask for large numbers of carrot explants : free floating cells in the ambient fluid. ^U''-/C^ , ^-*^v*' V- i" CJ* (6) Tissue culture from small explant organized growing centre. with ^ %>^/.^ {c) Tissue cidture from small explant showing root formation. {d) Carrot roots developed in cultures grown from free floating cells. Figure 1. 167 Chemical structure and biological activity Caplin, 1951), and artichoke tuber tissue can be grown in almost exactly the same fashion as carrot tissue. Use is now being made of another technique which was devised primarily to permit relatively large amounts of cultured tissue to be obtained for the purposes of biochemical examination. Here the tube, normally containing one or at most two or three individual carrot explants, is replaced by a flask {see Figure l{a)). Around the periphery of the flask, nipples are blown in such a way that the tissue explants distribute themselves in these projections and again are alternately exposed to air and bathed in liquid. 100 explants from an individual carrot root can be successfully grown in a culture of this sort containing 250 ml of nutrient fluid, and in this fashion weights up to 10 g of tissue can be grown in about 28 days. The standard deviation of such a population of explants is very small indeed (zb^-lO per cent of the mean). However, an unexpected sequel to this technique is of interest. Explants taken from the mature carrot root grown by this 'flask technique' grow in much the same fashion as individual explants in a normal culture tube, that is, they produce more or less ellipsoidal masses of a relatively undifferentiated parenchymatous type of tissue. However, when concentrated in the flask the supernatant becomes somewhat turbid, owing to the development of free floating cells which grow in the medium. Such floating cells can be used to make liquid inocula into other flasks. This technique has been in continuous use now for some two years and it has become a common experience that a large number of small tissue cultures can be grown from these liquid inocula which contain only free floating cells, either singly or in very small groups. Also, quite contrary to expectations it was found that cultures grown from these free floating cells are much more prone to organize and form roots than they are when grown in the normal medium from explants newly isolated from the carrot root. Reference is made to this technique here, however, merely to emphasize that it is now quite clear that the tissue culture technique can be extended downward so that inocula can be made, not only from small tissue explants from the whole organ, but also from a stock of free growing cells. GROWTH REqUIREMENTS OF DIFFERENT TISSUES Carrot and artichoke tissue grow adequately when a normal basal medium is supplemented by the addition of whole coco-nut milk. The condition, however, is quite different in the case of the potato tuber. Depending somewhat on the individual tuber and the strain from which it was derived, virtually no external growth occurs when minute pieces of potato tissue are exposed to the conditions under which carrot tissue grows apace. To produce an actively growing tissue culture of potato tuber, a synergistic mixture is necessary. This consists of 2 parts: (a) the coco-nut milk complex, and (b) 2:4-D or an equivalent substance present in minute amount in the solu- tion. For 2:4-D itself, the most effective concentration is of the order of 6 p. p.m. in a medium containing about 10 per cent by volume of whole coco-nut milk (Steward and Caplin, 1951). In this curious twofold require- ment for substances that induce proliferative growth in the cells of the potato tuber, it is now clear that both substances are continuously necessary. This can be shown by the experiments illustrated in Figure 2, in which the tissue 168 Chemical induction of growth was exposed to coco-nut milk alone after growing for varying lengths of time in the synergistic mixture of coco-nut milk and 2:4-D. The data clearly show that maintained rapid growth requires the continuous presence of both the 2:4-D and the coco-nut milk complex. An even greater stimulus to growth can be obtained by further interaction with casein hydrolysate. As a natural outcome of the development of this technique and its use to assay the naturally occurring growth-promoting substances and growth- inhibiting substances in plant extracts, the following general view has been developed. Plant cells, fully furnished with nutrients, salts, organic substrates, water, etc., may still fail to grow for different reasons. In some cases there is an Figure 2. Growth of potato tuber tissue explants started in basal medium with coco-nut milk and 2 :4-Z) , followed by transfers at varying time intervals to medium without 2 :4-Z). evident lack of the factors that will permit the cells actively to divide. This lack may often be met by the use of coco-nut milk or its equivalent. Even so, this may not alone suffice with many plant tissues from which extracts may be shown to contain substances that are inhibitory even when added to the carrot-coco-nut milk system. The potato tissue, shown above to require 2:4-D or its equivalent, is a case in point because it contains a sufficiently strong inhibitory substance, or mechanism, so that small amounts of potato tissue extract will antagonize the normal effect of coco-nut milk on carrot tissue. Various other storage organs, dormant resting buds, and endosperms like those of castor bean, have all been shown to contain such inhibitory 169 Chemical structure and biological activity substances (Steward and Caplin, 1952). This leads to the general philosophy that in the organized growth of the plant body one factor that suppresses the indefinite growth by cell division is the accumulation of inhibitory substances that poison or regulate the system tending to cause cell multiplication. Conversely, it also leads to the idea that examples in which cells indulge in such proliferative growth, where they would otherwise be expected to remain quiescent, furnish instances where these active growth-promoting systems, uninhibited by antagonists, can be detected. Two prominent examples are furnished by (i) the tumors which may be induced by the crown gall organism on plants such as Bryophyllum (Steward, Caplin, and Shantz, 1955) and (ii) the parthenocarpically generated fleshy tissue of the banana fruit (Steward and Simmonds, 1954). In both these cases extracts of the organ in question can be the cause of a growth induction in the carrot tissue explants which is very similar to that normally produced by coco-nut milk. Thus we are led to the idea, not by any means unique to this system, that one of the factors involved in differentiation is that the mechanism which would normally lead to indefinite proliferating growth by cell division is brought under control, either by the entire lack of the growth- promoting substances for this type of growth, or much more probably by the accumulation of inhibitory substances. In short, in the balance between the growth promoters for cell division and their inhibitors lies the clue to the regulatory control of growth in this type of system. GROWTH BY CELL DIVISION AND CELL EXPANSION The tissue culture system, therefore, presents a powerful tool in the investiga- tion of growth regulation and of the substances that both induce and antagonize growth of an undifferentiated kind. There is, however, a further development necessary to give the technique its full range and usefulness. In much of the work, previously published, the main criterion of growth was the total increment of weight in the growing culture. For a great many general purposes, this suffices. However, for the understanding of the mechanism and for the full exploitation of the technique, it is necessary to differentiate between growth in cell number and growth in cell size. This has long been recognized but the opportunity only presented itself recently for a full investigation of the growth of these tissues at the cellular level. The basis of this will now be described. Maceration techniques have been familiar to the plant anatomist for some time. It was, however. Brown and Rickless (1949) who first exploited the maceration technique to trace the course of growth in this way. Following this lead, we have adopted similar procedures to study the growth of these tissue cultures at the cellular level. Since considerable reference will be made to results obtained by this technique, its essential basis is here presented. The tissue explants, whether the original explants from the whole organ or those grown under the conditions described, are macerated in a suitable volume of a mixture of 5 per cent chromic and 5 per cent hydrochloric acids. This procedure is normally done in small vials. At the appropriate time, which experience indicates is approximately 5-10 hours at room temperature, the tissue is fragmented as much as possible by agitating the liquid. Following Brown's procedure, a hypodermic needle, through which the fluid is taken 170 Chemical induction of growth up and forcibly ejected, is used to reduce the macerate as much as possible to single cells. The total number of cells is calculated from counts made on small aliquots placed in a haemocytometer. In principle this same method is still being used but the following refinements have been added. The haemocytometer is placed upside down on the stage of a Zeiss Winkel inverted microscope fitted with a camera attachment. A sufficient field (approximately 12 mm-) on the haemocytometer is viewed with a low power objective (4x). With a suitable eyepiece (6x), photographs are taken of representative fields. The negatives are projected on a screen to count the number of cells per unit area on the haemocytometer, and filed for a perma- nent record. This method is very much more useful for routine application because a large number of such macerates can be prepared and photographed and the counting done subsequently at leisure, whereas actual microscopical counting is arduous and severely limits the number of observations that are possible on a given day. There is no reason to believe that the photographic technique is any different in its accuracy from the visual technique, whose accuracy is discussed below. The volume of liquid placed on the haemocytometer slide is a very small part of the total macerate; therefore, the error is largely one of sampling. To reduce this error, several aliquots are counted. By suitable precautions, a relatively accurate estimate can be made of the total number of cells in a tissue explant of this sort. The method also measures the average cell size. If one records the weight of the tissue explant in milligrams and knows the number of cells per culture, it is readily possible to determine cell size, either in the units micrograms per cell, or cell number per microgram of tissue. Obviously this is a measure of average cell size, but nevertheless it permits interesting conclusions to be drawn when the growth passes predominantly from growth by cell division into growth predominanth" by cell expansion. With these aids we now know and may summarize much more about the growth of these tissue cultures under controlled conditions than has previously been published. THE GROWTH CURVE OF TISSUE EXPLANTS UNDER CONTROLLED CONDITIONS The growth curve in terms of fresh weight follows the familiar sigmoid pattern (Caplin and Steward, 1949). In terms of cell number, growth may be somewhat similarly represented and changes in average cell size as the growth proceeds may also be followed. In the initial phases of the growth there is first a lag period during which little external growth occurs though preformed cells may enlarge. This is followed by a more or less prolonged period in which cell numbers rise exponentially with time and in which the growth is predominandy by cell division. During this period the average cell size of the culture as a whole steadily decreases so that the effect of any already mature cells is counteracted by the large increment of cell number and of cells that do not conspicuously enlarge. It is worth emphasizing how small are the cells typically produced in coco-nut milk (see Table 1). Even so, however, as the culture enlarges, some cells, particularly those internally, do enlarge and some areas occur in the tissue in which relatively large air spaces may be formed. (See Steward, 171 Chemical structure and biological activity Caplin, and Shantz, 1955, Plate I). As Figure 3 shows, it often happens that after about 10 days the average cell size begins to change in a way that clearly indicates that vacuolation and growth by cell enlargement now makes an appreciable contribution to the total growth of the tissue system. In this system it is found that in the most rapid phase of growth the cell divisions must occur so that successive cell layers are formed approximately every 4 hours. This rapid rate is, however, not inconceivable, and it is quite clear from the large number of cells produced in these cultures that divisions must occur with this order of rapidity. (b) Ai^erage cell size/ lime 8 12 ie ^0 Time days 8 12 IS 20 Time days Figure 3. Growth of carrot explants in medium containing coco-nut milk. It is, however, abundantly clear that the growth surrounding small inocula, producing successive layers at regular intervals of time, cannot proceed indefinitely or even for very long. Vacuolation and the formation of large air spaces certainly occurs in the larger cultures. It may be seen from Figure 3 that in the example to which this relates, growth by cell enlargement also became evident after 10 days. Visual examination of the anatomy of the cultures clearly shows that as they become larger the familiar local centres of growth, or 'buds', appear on the surface as minute proliferations. Around the familiar lignified elements in these tissue cultures, local centres of growth organize that eventually may lead to roots [Figure l{b), (c), and (d)). Thus, although the concept indicated above is clearly useful for the analysis of the earlier stages of growth, it is quite clear that it cannot be projected into the later periods, in which some parts of the culture may still indulge in random proliferation, while others are growing in a more organized fashion. Most of the work now to be described will, however, have concentrated upon the more early stages in the growth that is capable of more exact analysis in terms of cell number and of cell size. CHEMICAL FACTORS IN GROWTH INDUCTION The analysis, by these various means, of chemical factors involved in the induction and regulation of growth in these tissue cultures can best be developed in two parts. 172 Chemical induction of growth In the first part it will be assumed that the cells are adequately furnished with the media which endow them, to their maximum extent, with the ability to grow by cell division. These media consist either of coco-nut milk or of some equivalent extract obtained from immature corn (-^^fl) or Aescuhis, etc., supplemented as necessary by casein hydrolysate. The investigations will then concern the extent to which still further added substances either stimulate or depress the ability of the tissue to respond to these materials. In this connection the use of the potato tuber tissue is most effective because it does not grow to any appreciable extent in the basal medium, even though supplemented by the coco-nut milk medium, but it requires the further addition of substances like 2:4-D that are synergistically active with, and enable the tissue to respond to, coco-nut milk. In the second part of this more chemical discussion some observations will be made on the known chemical nature of those fractions of the coco-nut milk or similar extracts that have proved susceptible of isolation and, to some extent, of chemical identification. Part II: The Chemical Nature of the Growth-Promoting Substances in Coco-nut Milk and Similar Fluids : The Present Position SUBSTANCES INVOLVED SYNERGISTICALLY WITH COCO-NUT MILK IN A GROWTH RESPONSE OF POTATO TISSUE Much work has been done up to the present time on the effects of 2:4-D or similar compounds in their synergistic effect, with coco-nut milk, on potato tuber tissue. First, to summarize briefly the work that has been published. A number of other substances have been tested and some of these have been found to possess greater ability than 2:4-D to stimulate the growth of potato tissue which is normally induced by the combination of 2:4-D and coco-nut milk. This investigation was carried out jointly with the laboratory of Dr. R. L. Wain. Though maximum activity occurs at different concentra- tions for various substances, it is quite clear that some of these sub,stances are more effective than is 2:4-D at its optimum concentration. This was particu- larly true in the case of 1 :2: 3: 4-tetrahydro-l -naphthoic acid, as shown in Table 1 of the paper on this subject (Shantz, Steward, Smith, and Wain, 1955). However, from the standpoint of the chemistry of this phenomenon, the greatest interest attaches to investigations of those substances which are active in this synergistic combination and which are optically active. It was found in two examples of this kind, namely, a-(2-naphthoxy)propionic acid and a-(2:4:5-trichlorophenoxy)propionic acid, that the ( + )-forms of these compounds were active in this growth stimulation, while their respective ( — )- forms were not. Furthermore it was found, in striking confirmation of work previously published from Wain's laboratory (Smith, Wain, and Wightman, 1952), that the ( — )-form was not merely inactive but tended to suppress the activity of the ( + )-form in mixtures of the two. The contrast between the role of the optical enantiomorphs of the same substance clearly indicates that these chemically induced growth phenomena require to be explained by some very subtle properties of the causal molecules. 173 Chemical structure and biological activity EFFECT OF ADDED SUBSTANCES ON GROWTH BY CELL DIVISION AND CELL ENLARGEMENT It will be recalled that Wain and Wightman (1954) had found that the effect of the length of the aliphatic side-chain in a series of aryloxyalkylcarboxylic acids could be interpreted to mean that those substances that were active in certain growth tests also contained an even number of carbon atoms in the side-chain thus permitting the chain to be degraded by progressive beta- oxidation to a 2 carbon side-chain. In the attempt to determine the best series of substituted aryloxyalkylcarboxylic acids with which to carry out a similar trial on potato tissue in presence of coco-nut milk as an assay system, a considerable number of variously substituted phenoxyacetic acids were tested. The first result, however, was of a different and somewhat unexpected kind. Relative increase in: Fresh wt. = weight of cukures under treatment divided by weight of coco-nut milk controls. Cell No. = Cell No. of cultures under treatment divided by Cell No. of coco-nut milk controls. Cell size = Cell size of cultures under treatment divided by cell size of coco-nut milk controls. CI Treatment : Rel. increase in fresh wt. x6-9 10-2 5-0 Rel. increase in Cell No. x2-9 2-8 1-8 Rel. increase in cell size x2-3 3-8 2-7 Weight Cell No. Cell size Controls per culture per culture Cells per /tig fig per cell Basal medium 2-7 mg l-9xl0« 0-70 1-4 Basal medium + coco-nut milk 5-7 mg 34-6xl0» 6-1 0-16 Figure 4. The effects of various phenoxyacetic acids on growth of potato tuber tissue in media containing coco-nut milk The tissues in question were submitted to the analysis of growth both in terms of fresh weight and in cell number, and this data furnished the first fairly extensive information relating the effect of chemical constitution to the stimulation by the synergist of growth by cell division and/or by cell enlargement. In Figure 4 the effect of the various substances, as used at their 174 Chemical induction of growth best concentration of approximately 1-5 p. p.m., is shown in such a way that the structure may be related both to the increment of cell number in the presence of coco-nut milk and also to the change in average cell size, obtained by dividing the weight of the tissue by the number of cells it contained. The comparisons between structure of the added substance and growth are most easily made as follows. By comparison with the initial tissue, or the controls in a medium free of the added substance, the relative increase due to the substance in either cell number or cell size may be simply calculated. The first important conclusion, evident from an inspection o^ Figure 4, is that the differing structures of the substituted phenoxyacetic acids markedly determine whether the substance acts predominantly on cell number, or on cell size. This is an important result because it immediately establishes the usefulness of this general tech- nique in further work of this sort. It is obviously advantageous to be able to distinguish between the substances that stimvdate cells to grow mainly because they stimulate division, and the substances which stimulate cells to grow mainly by causing them to enlarge. Furthermore, some substances, for example 2:4-D itself, have inherent in their molecules the ability to stimulate, in balanced degree, both of these processes. Figure 4 shows clearly that to a greater or lesser degree all of the substances stimulated growth by cell enlargement when added to potato explants growing in basal medium plus coco-nut milk. However, the ability to cause the tissue to respond by cell division was conspicuously lacking in the case of that compound in which the number 4 ring-position remained unoccupied. Scanning these data, preliminary as they may well appear to be, leads to the following general conclusions : 1 . For maximum ability to stimulate the growth of potato cells, which require the presence in the medium of the growth factors present in coco-nut milk, the aryloxyacetic acid used as a synergist shoidd have the number 4 position occupied by chlorine. This being so, other groups may be added in the ring but cell division will still be stimulated. While there may be effects due to the proximity of other groups in the 3 or 5 position, the main effect is clear; namely, that it is the number 4 position that seems to hold the key to the stimulation of growth by cell division in this system. 2. In the absence of the 4 substituent, marked stimulation of growth by cell enlargement may occur, but in all these cases the position that seems to hold the key to the stimulation of growth by cell enlargement is the number 2 position. If the number 2 position is occupied, other groups may be inserted in the nucleus, leaving number 4 unoccupied, and growth by enlargement will still occur. 3. Clearly then, the prevalent role of 2:4-D as a growth-stimulating substance is susceptible to the following interpretation: One may assume that the tissue has inherent factors essential for growth by cell division or indeed cell enlargement, factors which are represented by the chemical substances present in coco-nut milk. The action of such a substance as 2:4-D can be explained through its ability to accentuate markedly the tendency to cell enlargement by virtue of the substituent in the 2-position and it may also accentuate cell division, to which the tissue may itself be some- what prone, by virtue of substitution in the 4-position. These results are 175 Chemical structure and biological activity quoted because they show the pattern of the kind of investigations that can be made, and are being made in this laboratory, rather than that they necessarily represent the ultimate and final conclusion for this type of study. From this preliminary work, a given substituted aryl nucleus (2-methyl- 4-chloro-) was selected so that a series of phenoxy acids covdd be studied to determine the effect of side chain length, following the type of investigation carried out by Wain and Wightman (1954) using entirely different kinds of assay systems. It must be confessed that there was some prior thought that again it might be found that there were parallels between the potato tuber system and the assay systems of Wain and Wightman. However, as the data show, marked ability of this series of substituted phenoxy acids to function in the potato tuber system by acting synergistically with coco-nut milk was only revealed by the first member of the series. Subsequent members of the series (see Figure 5) were but slightly active or actually toxic in their effects. Even here, however, there is some indication that the compounds with 4 and 6-carbon side chains were more effective than those with 3, 5, or 7 carbon atoms. The much greater effectiveness of the substituted caproic acid, with 5 methylene groups, finds a parallel with some recent work at Wye (Wightman, 1955). Average final fresh wt. of cultures Tuber A Tuber B Controls : Basal medium + coco-nut milk 3-2 mg 7-8 mg Basal medium + CM + 2 n (number 4-D of 21-4 mg 30-9 mg methylene groups) Treatments : Basal medium 1 21-0 31-6 + 2 2-7 4-4 coco-nut milk 3 5-6 8-9 + 4 3-0 5-6 5 11-8 11-7 0-(CH2)^CO0H 6 6-6 7-8 AfCHs Figure 5. Effect of side chain length on the ability of various oj-{A:-chloro-2-methylphenoxy)alkyl- carboxylic acids to induce growth in potato tuber explants cultured in a medium containing coco-nut milk. This type of investigation clearly needs to be extended. As this communication is being written, certain other examples of substituted phenoxy compounds are being tried to obtain further cases in which the 2 position is occupied and the 4 position is vacant. Similarly, investigations are being carried out with a considerable number of substituted aryloxy ethanols. The results of these investigations, which will be recorded in terms of fresh weight, cell number, and cell size, are not yet available but when they are they will add to the evidence documented above. The main point to be 176 Chemical induction of growth made, however, is this: The tissue cuhure method described is a precise system in which, under controlled conditions, growth may be regulated and subjected to quantitative study. It represents a useful assay tool to be applied in almost all cases where the chemical induction of growth in relation to molecular structure is in question. One of the challenges in this field is that different biological materials respond in different ways. There are few systems that cannot be reduced to a tissue culture type of growth with sufficient perseverance. When supplemented by the analysis of growth in terms of cell number and cell size, the system described is capable of yielding much information of a kind that would otherwise be difficult to obtain. THE CHEMICAL NATURE OF SOME SUBSTANCES THAT INDUCE GROWTH BY CELL DIVISION The final question to be asked and answered, so far as possible, is the chemical nature of the stimulus to growth which the coco-nut milk and similar extracts furnish. It will be recalled that growth in the otherwise mature cells of carrot phloem, artichoke tissue, potato tuber tissue, etc., has been induced by the Hquid endosperm that nourishes the immature embryo of the coco-nut; by the endosperm from immature (milk stage) fruits of corn {^ea) ; from such a nutrient material as the female gametophyte of Ginkgo ; by extracts of the proliferating tissue of tumours induced by the crown gall organism; and by extracts of the loculus of the banana fruit in which, for genetic reasons, the cells do not remain quiescent but grow into a fleshy tissue. Doubtless these various extracts contain variants on the same essential theme. While the type of response induced may be the same, there is no reason to expect a priori that identical molecules are the causal agents. In only two types of material has enough investigation proceeded as yet to tell whether the same or different substances may be involved. These materials are the liquid endosperm of the coco-nut (Cocos) and the liquid contents of the immature fruit of the horse-chestnut (Aesculus). A brief summary will now be given of the present status of these investigations. SUMMARY OF PREVIOUS WORK Previous work had shown that the most useful way of isolating the growth- promoting activity of the coco-nut milk from the large number of other unessential constituents was a method based on precipitation by mercuric acetate from alcoholic solution. By this procedure a bulk isolation was carried out from a very large volume of original coco-nut milk. From this concentrate 3 crystalline active substances were obtained which could induce growth, almost as actively as whole coco-nut milk, at concentrations of the order of a few parts per million, though they required the simultaneous presence of casein hydrolysate. Thus the growth-promoting effects of the coco-nut milk consists of at least two essential parts: (i) A non-specific part replaceable by casein hydrolysate. (ii) A specific part which is replaceable, wholly or in part, by the isolated materials referred to. These materials were designated in 1952 as Compounds A, B, and C (Shantz and Steward, 1952). Since then a further compound has been isolated in 177 Chemical structure and biological activity pure form and its activity similarly demonstrated. This substance is known as Compound F (Steward and Shantz, 1954). It should be mentioned here that a very large number of inactive substances have been fractionated and some of them identified in the coco-nut milk. A variety of amino-acids, including the recently discovered pipecolic acid, were isolated in pure form. None of these functions is the specific response in question. The main new result to be reported is that Compound A has been re- isolated (Shantz and Steward, 1955) and definitely identified as the quite simple but rather unexpected substance, 1 :3-diphenylurea. This has been established beyond any doubt by analysis, mixed melting point, completely matching infra-red [Figure 6) and ultra-violet absorption spectra, and all essential criteria for fastidious chemical identification. What is now quite clear is that the response of various strains and varieties of carrot tissue to the h/ave/ength Figure 6. diphenylurea is very variable. In some cases diphenylurea in presence of casein hydrolysate may replace, in large part, the activity of whole coco-nut milk, approaching at least 50 per cent of that activity. In other cases the activity is small by comparison. The total concentration of diphenylurea that could at its maximum be present in whole coco-nut milk is not large (order of 01 p. p.m.) and it is quite clear that the coco-nut milk owes its characteristic properties much more to the other substances that are present therein than to the diphenylurea. In fact the paper in which this new discovery has been communicated still makes cautious reservations because the isolation of a relatively small amount of a substance like diphenylurea from a very large amount of biological material requiring the use of large amounts of reagents does raise the question whether the substance was present as such in the coco-nut milk of the intact nut. While there is every reason to believe that diphenylurea did so occur, the fact still remains that this constitutes the first known case of the isolation from plants of a carbanilide. Be that as it may, it is still a matter of great interest that here we have the first case of the chemical identification of an active substance from coco-nut milk, even though the total growth response produced by that substance may not be as dramatic as that due to the coco-nut milk as a whole. 178 Chemical induction of growth The remaining substances, Compounds B, C, and F, still await complete chemical identification. It may, however, be stated that they are all cyclic nitrogen compounds though none of them contain the requisite amount of nitrogen to resemble even remotely the substance kinetin, recently described by Skoog and his associates, and which appears to originate when nucleic acids are autoclaved, or aged (Miller et al., 1955a), The structure of kinetin, now known to be 6-furfurylaminopurine (Miller et al., 1955b) does not remotelv resemble diphenylurea, nor can it remotely resemble the other substances B, C, and F. The substance kinetin is active in causing cell division in tobacco callus tissue at very great dilutions. Direct comparisons of the activity of kinetin! with coco-nut milk and some fractions thereof and with other substances known to stimulate cell division in the tissue culture system are being made. Preliminary results show that kinetin in presence of indoleacetic acid unquestionably has activity toward carrot tissue under our conditions. However, the growth so induced is much less than that caused by coco-nut milk whether this is measured by fresh weight or cell division. The visible effect of kinetin is accentuated by its ability to form small internal cells in the otherwise large cells of tobacco callus. Though definitely active toward carrot tissue in culture, it has proved almost inactive toward a strain of artichoke tuber tissue. These results are shown by the following extract from the data available. Carrot explants {\^daysat2^-5°C) Artichoke explants [19 days at 24-5°C) Wt. (mg) Cell No. [units,-: 10») Cell size i/iiglcell) Wt. (mg) Cell No. {units X 103) Cell size i/iglcell) Basal medium Basal medium + coco-nut milk Basal medium+L\A+kinetint 8-2 89-6 12-5 331 973 222 0-25 0-09 0-06 4-2 53-6 8-0 11-7 290 32-4 0-36 0-19 0-24 t Indoleacetic acid and kinetin applied separately do not give this characteristic effect. There is obviously, therefore, no single molecule that unlocks the door of cell division for all cells. This is hardly to be expected. Many such molecules have this property. Some may be of natural origin and these are of the greatest interest; many more may be found as a result of tests on synthetic substances. Indeed, by chance, a very potent substance capable of inducing growth by cell division in carrot and artichoke tissue has been found in 2-benzthiazolyloxyacetic acid. The effects of diphenylurea in stimulating growth by cell division in carrot, artichoke, and potato tissue are illustrated in Tables 1 and 2. The effect of 2-benzthiazolyloxyacetic acid on both carrot and artichoke tissue are shown in Figure 7. It can be seen that the 2-benzthiazolyloxyacetic acid even exceeded the activity of whole coco-nut milk when applied to a given population of artichoke explants grown on agar medium. It is also evident t The kinetin was kindly supplied by Dr. F. Skoog and his colleagues at the University of Wisconsin. 179 Chemical structure and biological activity Table 1 Effect of 1 : 3-diphenyl urea [DPU) on growth of plant tissue cultures (a) Carrot tissue — grown 1 7 days at 24°C Treatment Fresh wt. B in mg C Cell M. B (X103) C Cells B per fig C Basal medium 7-2 4-6 45-2 28-8 6-3 6-4 Basal medium + 2-0 p. p.m. DPU 9-4 7-5 50-2 49-0 5-4 6-5 Basal medium + casein hydrolysate 29-4 64-8 201-4 536-1 6-8 8-8 Basal medium + casein hydrolysate + 2-0p.p.m. DPU 56-8 108-5 508-6 1042-7 8-9 9-7 Basal medium + casein hydrolysate + 10% coco-nut milk 294-4 323-1 2662-8 2692-5 9-1 8-3 {b) Artichoke tissue — grown 29 days at 24°C Treatment Fresh wt. in mg Cell M. ( X 10=») Cells per fig Basal medium + casein hydrolysate Basal medium + 0-08 p.p.m. DPU 6-0 12-0 10-7 63-2 1-9 4-8 Table 2 Effect of 1 : 3-diphenyl urea (DPU) on growth of potato explants in liquid media containing 2 : 4-Z) and casein hydrolysate Explants from potato tuber — 33 days growth at 24°C 2:4-D applied at 1-0 p.p.m. Casein hydrolysate applied at 0-05%. DPU applied at 2-0 p.p.m. Coco-nut milk (CM) applied at 10% by vol. Treatment Fresh wt. per culture No casein hydrolysate Plus casein hydrolysate Basal medium Basal medium + 2 : 4-D Basal medium + 2 : 4-D + DPU Basal medium + 2:4-D+CM 1-92 1-72 1-80 20-28 2-2 2-64 8-55 46-80 ( Table 3) that the predominant effect of this substance is to stimulate cell division with the formation of a very large number of very small cells. By comparison, the effect of the naturally occurring substance diphenylurea is not conspicuously large {Table 1) though it is of a similar kind. The most dramatic development in the current work, however, was the recording of the very first experiment in which an extract from a dicoty- ledonous fruit proved effective in the carrot assay system. The fluid used was 180 Chemical induction of growth B D E • • • Mi i * "9' % » • ■* Figure 7(a). Effects of 2-benzthiazoIyloxyacetic acid iBTOA) on growth of carrot root explants. A = Basal medium — casein hydrolysate (0-05 per cent) B = Basal medium-^ casein hydrolysate^ coco-nut milk (10 per cent) C = Basal medium^ casein hydrolysate + BTO A at \0-0 p. p.m. D = Basal medium- casein hydrolysate^ BTO A at 2-{) p.p.m. E = Basal medlum^casein hydrolysate +BTO A at 0-4 p.p.m. Figure 7{b). Effects of 2-benzthiazolyloxyacetic acid [BTOA) on growth of explants from Jerusalem artichoke tuber. A = Basal medium + casein hydrolysate (0-05 per cent). B = Basal medium + casein hydrolysate + coco-nut milk {\0 per cent). C = Basal 7?iedium + casein hydrolysate + BTOA at \0-0 p.p.m. D = Basal medium + casein hydrolysate-^ BTOA at 2-0 p.p.m. E = Basal medium + casein hydrolysate + BTOA at 0-4: p.p.m. 181 13 Chemical structure and biological activity Table 3 Effects of 2-benzthiazolyloxyacetic acid [BTOA) on the growth of plant tissue cultures (a) Carrot root tissue Treatment Weight per culture in rng Cells per culture X 10' Cells per jiig Original explants Basal medium (BM) + casein hydrolysate (CH) BM + CH + coco-nut milk (10%) BM + CH + BTOAat 10-0 p.p.m. BM+CH + BTOA at 20 p.p.m. BM+CH + BTOA at 0-4 p.p.m. 3-3 10-3 1046-3 81-1 59-1 19-5 34-7 33-2 4590-4 548-2 225-5 66-1 10-5 3-2 4-4 6-8 3-8 3-4 (b) Jerusalem artichoke tuber tissue Treatment Weight per culture in mg Cells per culture X 10' Cells per jug Original explants BM + CH BM + CH + coco-nut milk (10%) BM+CH + BTOA at 10-0 p.p.m. BM+CH + BTOA at 2-0 p.p.m. BM + CH+BTOA at 0-4 p.p.m. 3-2 6-5 22-1 52-2 114-5 6-5 10-0 7-1 70-5 432-7 1084-1 16-2 3-2 1-1 3-2 8-3 9-5 2-5 obtained from the immature fruit of the walnut. By analogy with this system an examination was made of the immature fruit of the horse-chestnut. This data was reported first to the American Society of Plant Physiologists at the Gainesville meeting in 1954. It is now apparent that in the immature fruit of the horse-chestnut there occurs an extremely active growth-promoting effect comparable to that produced by coco-nut milk. Figure 8 shows that carrot explants can be made to grow in this way in a fashion comparable to that which occurs with coco-nut milk. This material has been collected on a relatively large scale and has been fractionated by methods somewhat different from the first ones used in the isolations of active substances from coco-nut milk. The present trend in this laboratory has been to use techniques in which heavy metal precipitation could be avoided in the attempt to isolate substances which are active in the carrot tissue assay system and which have some of the ultra-violet absorption properties that were thought, from the examinations of Compounds A, B, C, etc., to be a guide to the active growth-promoting principles. Bv proceeding in this general way and making use of an automatic Craig-Post liquid-liquid separator, much progress has been made. The important developments from all this work may be briefly summarized as follows: We now know conclusively that active growth-promoting qualities reside in substances present in coco-nut milk and in Aesciibis extracts, 182 Chemical induction of growth and presumably occur in all such similar materials, which 'are nitrogen free. This surprising result has led to the recognition that the general class of substances to which some of these materials belong is the leucoanthocyanins. en 60- W~ Carrot A □ Minus casein hydrolysafe ^ Plus casein liydrolysate i 03 5i 20 •^ fK ; J i i Corrof B T ", i ■: Basal medium Basal t C-lli. O-BVc OIBVc 0032''Io Basal f Aesculus endosperm Figure 8. Effect of liquid endosperm from immature fruits o/" Aesculus woerlitzensis on the growth of carrot phloem explants, compared with growth in basal medium and in basal medium plus 1 per cent coco-nut milk. The reasons for arriving at this conclusion are briefly summarized, although the full data cannot be recorded but will be reserved for the time at which one or more of these substances is unequivocally and finally identified. It may, however, be stated clearly that in the immature Aesculus fruit a large part of the growth-promoting activity can be ascribed to a leucoanthocyanin which may be a complex monoglucoside with one of the general chemical formulae indicated below, based on structures proposed by (a) Robinson and Robinson (1933), and {b) Swain HO OH H OH O c- .OH — OH HO CH-O-CsH^O^ OH (a) OH OH I -OH H /CH — O — CjHijOs OH (b) Figure 9. Structure of Leucocyanidin monoglucoside suggested by {a) Robinson and Robinson (1933) {b) Swain (1954). 183 Chemical structure and biological activity Reliance has to be placed on the known structure of certain related com- pounds to assign the substituent groups to appropriate positions, but with this latitude one can say definitely that the chemical analysis of the product as isolated, the presence of the monoglucoside group, and the analysis of the pigment which results when the glucose is split off and the cyanidin-like pigment is produced in the form of its oxonium chloride, are all consistent with this type of structure [Figure 9). This general discovery seems to open a new door in the study of the chemical induction of growth. Hitherto, the physiological role of these complex biochemical constituents, the anthocyanins and their precursors the leucoanthocyanins, has been somewhat obscure. When such a function as the ability to induce growth by cell division when added to otherwise complete nutrient media to which mature cells are exposed is attributed to a type of compound such as this, it is obvious that they may have important roles to play. Examinations have therefore been made of a number of other extracted materials (such as anthocyanins and flavanol glucosides) furnished to us by various investigators but having chemical properties somewhat similar to the structural formulae indicated above. Sufficient activity has been detected in compounds of this kind to indicate that we are proceeding along the right general track. The conclusion of this general account must therefore be as follows : To undergo growth by cell division and cell enlargement, plant cells obviously require a variety of chemical substances that perform quite different roles: water itself for enlargement, nutrients both inorganic and organic, vitamins and the essential co-factors of vitamin-like systems. Many or all of these may be specified. In addition, however, there still seem to be requirements for additional substances which determine whether the cell can fully exploit its ability to enlarge without division or to divide without enlargement. In balanced degree both processes may occur simultaneously. The presence of one substance may induce one type of growth to the exclu- sion of another, as for example the predominant effect of indoleacetic acid on cell enlargement. The effect of other types of substances, such as those found in coco-nut milk, may induce cell division to the exclusion of cell enlargement. Almost certainly both types of growth-promoting materials, the auxin-like substances of which indoleacetic acid is the predominant natural example, or the cell division-promoting substances such as those found in coco-nut milk and related materials, represent a group or family of substances rather than single unique individuals. The baffling thing is that the chemical properties which determine the ability of the substance to promote growth by cell division are so obscure that it is difficult to see what substances like diphenyl- urea, the quite different molecules of Compounds B, C, and F from coco-nut milk, and 2-benzthiazolyloxyacetic acid, with the further addition of the substance kinetin, all may have in common. This becomes even more difficult, however, when this activity is extended in its range by the fact that leucoanthocyanin-like substances, nitrogen-free, are found to be active and even very potent. It is obviously too soon to state what any or all of these molecules may have in common. The only idea that has already occurred is that, in greater or lesser degree, many of these substances might exert their activity by their 184 Chemical induction of growth abiUty to chelate or affect chelation of heavy metals. Be that as it may, the main point of this communication is achieved when it is recognized, first of all, that these properties reside in a variety of naturally-occurring substances which may have many synthetic parallels, and, second, that in the further elucidation of the relationship between structure and these functions the use of the tissue culture systems, combined with methods that evaluate growth in terms of both cell division and cell enlargement, offer a most fruitful approach to these interesting problems. ACKNOWLEDGEMENTS The authors wish to acknowledge the help of Miss K. Mears with the sterile cultures and of Miss J. Smith with the cell counts. The work has been supported by grants from the National Cancer Institute of the U.S. Department of Health, Education, and Welfare. REFERENCES Brown, R., and Rickless, P. (1949). A new method for the study of cell division and cell extension with some preliminary observations on the effect of temperature and of nutrients. Proc. Roy. Soc. B136, 110. Caplin, S. M. (1947). Growth and morphology of tobacco tissue cultures in vitro. Bot. Gaz. 108, 379. Caplin, S. M., and Steward, F. C. (1948). Effect of coconut milk on the growth of explants from carrot root. Science, 108, 655. Caplin, S. M., and Steward, F. C. (1949). A technique for the controlled growth of excised plant tissue in liquid media under aseptic conditions. Nature, 163, 920. Fawcett, C. H., Ingram, J. M. A., and Wain, R. L. (1952). /^-oxidation ol w-phenoxyalkylcarboxylic acid in the flax plant. Nature, 170, 887. Fitting, H. (1910). Weitere entwicklungsphysiologische Untersuchungen an Orchideenbluten. Z- Bot. 2, 225. Heller, R. (1953). Recherches sur la nutrition minerale des tissus vegetaux cultlves in vitro. Masson, Paris (Thesis, Faculty of Science, Paris). Miller, C. O., Skoog, F., Saltza, M. H. von, and Strong, F. M. (1955a). Kinetin, a cell division factor from DNA. J. Amer. chem. Soc. 77, 1392. Miller, C. O., Skoog, F., Okumura, F. S., Saltza, M. H. von, and Strong, F. M. (1955b). Structure and synthesis of kinetin. J. Amer. chem. Soc. 77, 2662. Robinson, G. M., and Robinson, R. (1933). A survey of anthocyanins. III. Distribution of leuco-anthocyanins. Biochem. J. 27, 206. Shantz, E. M., and Steward, F. C. (1952). Coconut milk factor: the growth- promoting substances in coconut milk. J. Amer. chem. Soc. 74, 6133. Shantz, E. M., and Steward, F. C. (1955). The identification of compound A from coconut milk as 1 : 3-diphenylurea. J. Amer. chem. Soc. 77, 6351. Shantz, E. M., Steward, F. C, Smith, M. S., and Wain, R. L. (1955). Investiga- tions on the growth and metabolism of plant cells. VI. Growth of potato tuber tissue in culture: the synergistic action of coconut milk and some synthetic growth-regulating substances. Ann. Bot., Lond. (N.S.), 19, 49. Smith, M. S., Wain, R. L., and Wightman, F. (1952). Studies on plant growth- regulating substances. V. Steric factors in relation to mode of action of certain aryloxyalkylcarboxylic acids. Ann. appl. Biol. 39, 295. Steward, F. C, and Caplin, S. M. (1951). A tissue culture from potato tuber: the synergistic action of 2,4-D and coconut milk. Science, 113, 518. 185 Chemical structure and biological activity Steward, F. C, and Caplin, S. M. (1952). Investigations on growth and metabolism of plant cells. III. Evidence for growth inhibitors in certain mature tissues. Ann. BoL, Lond. (N.S.), 16, 477. Steward, F. C, Caplin, S. M., and Millar, F. K. (1952). Investigations on growth and metabolism of plant cells. I. New techniques for the investigation of meta- bolism, nutrition, and growth in undifferentiated cells. Ann. Bot., Lond. (N.S.), 16, 58. Steward, F. C, Caplin, S. M., and Shantz, E. M. (1955). Investigations on growth and metabolism of plant cells. V. Tumorous growth in relation to growth factors of the type found in coconut. Ann. Bot., Lond. (N.S.), 19, 29. Steward, F. C, and Shantz, E. M. (1954). The growth of carrot tissue explants and its relation to the growth factors in coconut milk. II. The growth-promoting properties of coconut milk for plant tissue cultures. Annee biol. 30, 399. •Steward, F. C, and Simmonds, N. W. (1954). Growth-promoting substances in the ovary and immature fruit of the banana. Nature, 173, 1083. Swain, T. (1954). Leucocyanidin. Chem. & hid. 1144. Wain, R. L. (1953). Plant growth substances. Roy. Inst. Chem. Monograph No. 2. Wain, R. L., and Wightman, F. (1954). The growth-regulating activity of certain fo-substituted alkyl carboxylic acids in relation to their /5-oxidation within the plant. Proc. Roy. Soc. B142, 525. White, P. R. (1943). A handbook of plant tissue culture. Jaques Cattell Press, Lancaster, Pa. Wightman, F. (1955). Private communication. 186 THE DEGRADATION OF CERTAIN PHENOXY ACIDS, AMIDES, AND NITRILES WITHIN PLANT TISSUESf C. H. Fawcett, H. F. Taylor, R. L. Wain, and F, Wightman Agricultural Research Council Unit on Plant Growth Substances and Systemic Fungicides, Wye College, University of London Recent work at Wye has been concerned with the /3-oxidation of oj-phenoxy- alkanecarboxyhc acids within plant tissues. It had previously been shown that an alternation in growth-regulating activity might operate in an homo- logous series of oj-aryl- or oj-aryloxy-alkanecarboxylic acids as the side-chain length increased (Grace, 1939; Synerholm and Zimmerman, 1947). Such behaviour was ascribed by Synerholm and Zimmerman to the degradation of the side-chain by /j-oxidation. On this basis, only the acetic derivative and its alternate homologues would be expected to show activity. Further biological evidence for the /i-oxidation hypothesis was provided by Fawcett, Ingram, and Wain (1954) and by Wain and Wightman (1954), who showed that only alternate members of the oj-(4-chlorophenoxy)alkanecarboxylic acids were active in the pea curvature and wheat cylinder tests. Again, Luckwill and Woodcock (1955) have recently shown that a similar alterna- tion operates in regard to the capacity of ft>-(2-naphthoxy)alkanecarboxylic acids to induce parthenocarpic development of tomato ovaries. Chemical evidence that ^-oxidation can occur within plant tissues has been obtained at Wye. In our first experiments using flax seedlings, members of the homologous series of a>-phenoxyalkanecarboxylic acids, CgH50(CH2) „COOH with n = \ to 10, were supplied to flax seedlings through their roots. After a suitable interval to allow breakdown of the acids to occur, the plants were steam distilled and phenol in the steam distillate estimated colorimetrically. It was found that those acids with an even number of side-chain methylene groups {n = 2, 4, 6, 8, and 10) gave rise to appreciable quantities of phenol, e.g. C6H50CH,CH2COOH->[C6H50COCH2COOHl-> «"=2 [CeHsOCOOHJ^CeH^OH phenol The alternate higher homologues (n = 4, 6, 8, 10) could also yield phenol by repeated /^-oxidation. The acids with n = 1,3,5, and 7, on the other hand, yielded only traces of phenol, which is in agreement with the fact that phenoxyacetic acid is the expected end product from these compounds (Fawcett et al., 1954), e.g. C6H50CH2CHoCH2COOH->[C6H50CH2COCH2COOH]^ n=^S C6H5OCH2COOH phenoxyacetic acid ■[" This paper was read at the Conference by R. L. Wain. 187 Chemical structure and biological activity These results, then, are fully consistent with the view that the side-chain of these acids is degraded within the flax plant by ^-oxidation. The 9-phenoxy- nonane-1-carboxylic acid {n = 9) showed exceptional behaviour in yielding phenol, a result which can be explained by side-chain degradation involving oj-oxidation as follows: C6H50CH2(CH2)8COOH^[C6H50CO(CH2)8COOH]^^-^C6H50H fi — 9 phenol This findinor for the nonane acid correlates with the results of animal meta- holism studies (Verkade and Lee, 1934; Breusch, 1948). Further chemical evidence that /i-oxidation can occur in plants has been provided by recent work on the metabolism of aryloxy-acids in pea-stem tissue. We have found that phenoxyheptanoic acid {n = 6) is degraded more readily than phenoxyvaleric acid {n = 4) to phenoxypropionic acid (n = 2), a result which correlates with the growth-regulating activity shown by these com- pounds in the pea test (Fawcett et. al., 1954). Furthermore, in these experi- ments the heptanoic and valeric derivatives were converted to phenol more readily than the propionic analogue. In another aspect of this work, three homologous series of chloro-substituted phenoxy acids with chlorine substituents in the 4-, 2:4-, and 2:4:5- positions respectively, and each consisting of the first six members, were examined in the wheat cylinder and pea curvature tests (Wain and Wightman, 1954). In the former test, all series showed a typical alternation in activity which was consistent with the /i-oxidation hypothesis. In the pea test, however, although this alternation was shown in two series, in the 2:4:5-trichloro- phenoxy series only the acetic derivative was active. Also in this investiga- tion, by using chromatographic and biological methods, clear evidence was obtained that 2:4:5-trichlorophenoxyacetic acid was produced by treating the corresponding butyric and caproic acids with wheat coleoptile tissue. The most important conclusion from this work (Wain and Wightman, 1954) was that whereas different types of plant tissue may all possess /^-oxidase enzyme systems, whether or not these are able to degrade the side chain of specific fo-phenoxyalkanecarboxylic acids depends on the nature and posi- tion of the nuclear substituents present. Such considerations have led logically to developments in the field of selective toxicity (Wain, 1955). Evidence that /5-oxidation of the side chain of co-(l-naphthyl)alkanecarb- oxylic acids can occur in plant tissues has been obtained at Wye. The effect on activity of substituting alkyl groups into the a- or /3- positions of the side chain of certain of these compounds was also investigated and found to be consistent with the concept of /i-oxidation (Wain and Wightman, 1954). More recently, a detailed study has been undertaken of the influence of ring substitution in to-phenoxyalkanecarboxyhc acids on the /5-oxidation of these compounds in different plant tissues. This work, to which Miss M. B. Pybus and Miss R. M. Pascal have also contributed, has involved the synthesis of the first six members of each of fifteen homologous series of phenoxy acids. The results, which will be published in full elsewhere, provide further evidence that /i-oxidase enzyme systems are present in wheat coleoptile and pea stem tissues and that these enzymes can show considerable substrate specificity. In this connection, they confirm and 188 Degradation within plant tissues extend earlier findings (Wain, 1955) that substituents in certain positions of the nucleus of co-phenoxyalkanecarboxylic acids can hinder /^-oxidation within specific plant tissues (see Table 1). Table 1 Alternation in activity shown by homologous series of substituted co-phenoxyalkanecarboxylic acids >0(CH2)„C00H in three biological tests X Test Ring substituents Wheat Pea Tomato cylinder curvature epinasty 4-Chloro- Yes Yes Yes 2:4-Dichloro- Yes Yes Yes 2-Methyl-4-chloro- Yes Yes Yes 3:4-Dichloro- Yes Yes Yes 3-Methyl-4-chloro- Yes Yes Yes 2:4:5-Trichloro- Yes No No 2 :4-Dichloro-5-methyl- Yes No No 2:5-Dichloro- Yes No No 2-Methyl-5-chloro- Yes No No In continuing this work on the /5-oxidation of co-phenoxyalkanecarboxylic acid in plants, it was logical to study growth-regulating activity in the corresponding homologous series of amides and nitriles. Accord- ingly, the oj-(2:4-dichlorophenoxy)-series Cl2CgH30(CH2)„CONH2 and Cl2CgH30(CHo) „CN, with n ^ I to 6, were synthesized and examined in the wheat cylinder and pea curvature tests, the homologous series of acids being also included. The typical alternation in the acid series which we have already recorded (Wain and Wightman, 1954) was again evident in both tests. The results are recorded in Figures 1 and 2. Alternation was also shown in both tests by the amide series {Figures 3 and 2) and this is explicable in terms of hydrolysis of amide to acid ( — CONHg— > — COOH) followed by /i-oxidation. The inactivity of all nitriles except the acetonitrile in the pea test {Figure 4), however, indicates that a similar hydrolysis of the — CN grouping to — COOH does not readily occur with these higher homologues in pea tissue. Furthermore, whereas such hydrolysis followed by ^-oxidation would explain the activity of alternate homologvies of 2 :4-dichlorophenoxy- acetonitrile in the wheat test, no such explanation is valid for the activity shown by other members of this' series {Figure 2) . To elucidate these problems it was decided to expose solutions of all these acids, amides, and nitriles separately to wheat and pea tissues and then study the compounds present by paper chromatography. As these results will be presented in detail elsewhere, they will be given only briefly here. The procedure adopted was to run ether extracts of the treated solutions in the descending manner with «-butanol/ammonia/water ( 1 00 : 3 : 1 8) as solvent for 15 hours at 20°C. After drying, the papers were divided horizon- tally into ten equal segments representing Rf 0-0-1, 0- 1-0-2, 0-2-0-3, etc. 189 Chemical structure and biological activity 0-}p.j)m. 7p.p.Ta. Wp.p.ia. Ace/ic Propionic Bufyric Valeric Capro/c Hepfanolc Figure 1. Activities ofu)-{2-A-dichlorophenoxy)alkanecarboxylic acids in the pea curvature test. Acetic Propionic Butyric Valeric Capro/c Heptonoic Acids Amides Nitriles lU*Li»L» ■p-fXTri. — ► Figure 2. Histograms showing activities of io-{2-A-dichlorophenoxy)alkanecarboxylic acids and their corresponding amides and nitrites in the wheat cylinder test. 190 Degradation within plant tissues £'-7p.p.Tn, /p.p.m. /n -v^ ^/' \/- /T -\r~ 1 T % 1 T Y Y \ Y ^ Y \ Y \ Y q? Y . r -r ? T Y i T T Y Y ? T" J T T '^ ^ ■J^l Y ' <■ ry 1 nr T T Y Y T T Control T Y t Y T Y ? Y ^ ' 1 / Y T -Y T T" V ' Y T Y" f ~r Y" Y ^ V Y ■' -Y 1 Y /Cp.p.TTl. Acef amide Propionamide Butyramide Valeramide Capro amide Hepfanoamide Figure 3. Activities ofo)-(2-A-dichlorophenoxy)alkanecarbonamides in the pea curvature test. (7-/p.p.Ta. /p.pm. r^p.p.m. Acefonifri/e Propionitrile Bufyronitrile Valeronitrile Copronifrile Heptanonifrile Figure 4. Activities ofio-{2-A-dichlorophenoxy)alkanenitriles in the pea curvature test. 191 Chemical structure and biological activity These were placed in water and a biological assay was carried out using either wheat coleoptile sections or split pea stem segments. Authentic compounds were run for comparison in all experiments. o)- (2 : A-dichlorophenoxy) alkanecarboxylic acids Whether pretreated with wheat or with pea tissue, assessment of the chromatogram segments in both the wheat and pea tests revealed the formation of 2:4-dichlorophenoxyacetic acid (2:4-D) (i?^ 0-35-0-45) from the butyric and caproic acids (n = 3 and 5), but not from the propionic, valeric, and heptanoic acids (ra = 2, 4, and 6). These results are readily explicable on the basis of /^-oxidation. CO- (2 : A-dichlorophenoxy) alkanecarbonamides The chromatograms obtained from the first, third, and fifth members of the series {n ^ 1, 3, 5) pretreated with wheat or pea tissue showed two peaks of activity both with the wheat and the pea bio-assay methods. The first of these peaks (/?^ 0-35-0-45) represented 2:4-D, and the second, residual amide (i?^ 0-75-0-85), present as such on the paper but presumably converted to 2:4-D in the bio-assay. Since other homologues {n = 2, 4, 6) yielded no 2:4-D in these experiments, the results indicate strongly that amide to acid hydrolysis can occur in both the wheat and pea tissues and that the acid produced then undergoes /j-oxidation. CO- (2 : A-dichlorophenoxy) alkanenitriles When these compounds were pretreated with pea tissue, only the first member (w = 1) yielded 2:4-D. Two peaks of activity were revealed on this chromatogram when both the wheat and pea tests were used for bio-assay and these peaks corresponded with the R^ values of 2:4-D {R^ 0-35-0-45) and 2:4-dichlorophenoxyacetonitrile (/?^ 0-8-0-9). With other members of the series, the chromatograms assayed in the pea test showed uniform inactivity over the whole paper, thus confirming that hydrolysis of these higher nitriles to the acids cannot occur in pea tissue. As already noted, all the nitriles showed activity in the wheat test {Figure 2). The chromatographic and wheat cylinder bio-assay investigations of nitrile solutions pretreated with wheat tissue showed that all members of the series yielded 2:4-D in amounts which are related to their activity in the wheat cylinder test. In all these chromato- grams there was also a second peak of activity representing residual nitrile (i?^0-8-0-9). With a pea test bio-assay, however, these chromatograms showed no second peak of activity in the higher R^ regions since the higher nitriles are inactive in the pea test. The methods employed in this work, then, provide a means whereby the growth-regulating activity of compounds can be related to their conversion to known active auxins within the tissues of the biological material employed. It is significant that in all cases where activity is shown the active acetic acid (2:4-D) is produced (cf. Seeley et al., 1956). One of the most interesting features of this work is the activity of all the nitriles in the wheat cylinder test {Figure 2). The greatest activity is again shown by the acetic, butyric, and caproic derivatives {n = \, 3, 5) and this is explicable on the basis of hydrolysis of — CN-> — COOH followed by 192 Degradation within plant tissues /5-oxidation to yield 2:4-D. The production of 2:4-D from the propionic, valeric, and heptanoic nitriles {n ■= 2, 4, 6), however, is not explicable by this two-stage reaction mechanism. At the same time, there is no reason to suppose that in wheat tissue these three compounds are not subject to hydrolysis followed by /3-oxidation, which in their case would lead to 2:4- dichlorophenol as the end product. On the other hand, evidence of their conversion to 2:4-D in wheat tissue indicates that some other type of break- down must also be involved. To produce 2:4-D from the valeronitrile, for example, it is clear that a single carbon atom must have been lost from the side chain at some stage. This may occur as follows : OCH„CH,CR,CH„ CI CN OCH„CH,CHX;OOH OCHoCOOH CI /3-oxiclation > CI CI CI (valeronitrile) « = 4 CI 2:4-D Now there is already evidence that the — CH2CN grouping can be degraded to carboxyl in the animal body. Thus, for example, /?-chlorobenzylnitrile fed to dogs is excreted in the urine as a derivative of /?-chlorobenzoic acid (Adeline et al., 1926). Further evidence of the loss of a one- carbon fragment from the — CHgCN grouping arises from the work of Pattison (1953) in his studies on the toxicity of co-fluoroalkanenitriles to rats. A similar biochemical degradation would explain our results with oj-(2:4- dichlorophenoxy)alkanenitriles in wheat coleoptile tissue. In this connection, it is of considerable interest that a separate investiga- tion which was proceeding here with indole compounds has provided strong confirmatory evidence that nitriles can be so degraded in plant tissue with the loss of one carbon atom. In this work, which is referred to in one of our other contributions to this Conference (Seeley et al., 1956), indole-3-aceto- nitrile has been shown to be converted to indole-3-carboxylic acid in presence of wheat coleoptile or pea stem tissues, i.e. \/\NH -CH2CN / \/\NH/ COOH This clearly provides another example of the type of nitrile degradation which we suggest occurs in the phenoxy nitriles. Since, in this breakdown, the a-methylene grouping becomes oxidized, the process may be conveniently referred to as a-oxidation of nitriles. Finally, since we have shown that pea tissue can convert indole-3-aceto- nitrile to indole-3-carboxylic acid, and since there is evidence of the presence in germinating peas of a compound with properties similar to that of indole-3-acetonitrile (Cartwright et al., 1956), it was logical to look for the carboxylic acid as a natural constituent of this tissue. Dr. Cartwright has in fact found this acid to occur in young pea plants (Cartwright et al., 1956). 193 Chemical structure and biological activity The evidence provided by these separate investigations thus indicates strongly that a-oxidation of nitriles, like /3 -oxidation of acids, is a degradation mechanism which occurs not only in animal but also in plant biochemistry. REFERENCES Adeline, M., Cerecedo, L. R., and Sherwin, C. P. (1926). Detoxication of nitriles. J. biol. Chem. 70, 461. Breusch, F. L. (1948). The biochemistry of fatty acid metabolism. Advanc. Enzymol. 8, 347. Cartw RIGHT, P. M. (1955). Private communication. Cartwright, P. M., Sykes, J. T., and Wain, R. L. (1956). The distribution of natural auxins in germinating seeds and seedling plants. This volume, p. 32. Fawcett, C. H., Ingram, J. M. A., 'and Wain, R. L. (1954). The /J-oxidation of (o-phenoxyalkylcarboxylic acids in the flax plant in relation to their plant growth- regulating activity. Proc. Roy. Soc. B142, 60. Grace, N. H. (1939). Physiological activity of a series of naphthyl acids. Canad. J. Res. 17, 247. LucKwiLL, L. C, and Woodcock, D. (1955). Plant growth regulators. I. The influence of side-chain length on the activity of oj-(2-naphthyloxy)-K-alkylcarb- oxylic acids for the induction of parthenocarpy in tomatoes. J. hort. Sci. 30, 109. Pattison, F. L. M. (1953). Toxic fluorine compounds. Nature, 172, 1139. Seeley, R. C, Fawcett, C. H., Wain, R. L., and Wightman, F. (1956). Chromato- graphic investigations on the metabolism of certain indole derivatives in plant tissues. This volume, p. 234. Synerholm, M. E., and Zimmerman, P. W. (1947). Preparation of a series of ft»-(2:4-dichlorophenoxy)-aliphatic acids and some related compounds with a consideration of their biochemical role as plant growth-regulators. Contr. Boyce Thompson Inst. 14, 369. Verkade, p. E., and Lee, J. van der (1934). Researches on fat metabolism. II. Biochem.J. 28, 31. Wain, R. L. (1955). A new approach to selective weed control. Ann. appl. Biol. 42, 151. Wain, R. L., and Wightman, F. (1954). The growth-regulating activity of certain w-substituted alkyl carboxylic acids in relation to their /5-oxidation within the plant. Proc. Roy. Soc. B142, 525. 194 RELATIONSHIP OF MOLECULAR STRUCTURE OF SOME NAPHTHYLOXY COMPOUNDS AND THEIR BIOLOGICAL ACTIVITY AS PLANT GROWTH REGULATING SUBSTANCES f L. C LucKwiLL and D. Woodcock Long Ashton Research Station, University of Bristol. During recent years, despite the considerable amount of work which has been done, particularly with indolyl and phenoxy compounds as growth substances, no systematic investigation of naphthyloxy compounds as growth regulators has yet been reported. One reason for this is doubtless the low activity of this series of compounds in the two tests which have been most widely used for measuring growth substance activity, viz. the Arena curvature test and the coleoptile cylinder test. However, 2-naphthyloxyacetic acid is active in delaying the abscission of petioles (e.g. ofColeus) and in stimulating partheno- carpy in tomatoes, and in our investigations the latter property was utilized MEASUREMENT OF BIOLOGICAL ACTIVITY This was carried out by observing the rate of growth of unpoUinated tomato ovaries following treatment with known amounts of each compound. The details of the method have been described in a previous publication by Luckwill (1948). Synthesis of these compounds is represented by James and Woodcock (1951, 1953) and Pope and Woodcock (1954, 1955). MATERIALS The compounds discussed in this paper fall into three main groups, viz. those which have involved an alteration in the side-chain of 2-naphthyloxy- acetic acid, nuclear substituted acids with an altered side-chain, and nuclear substituted 2-naphthyloxyacetic acids. CO- (2-NAPHTHYLOXY) -W-ALKYLCARBOXYLIC ACIDS Data for ten members of the homologous series of a)-(2-naphthyloxy)-n- alkylcarboxylic acids has appeared in a previous publication by the authors (Luckwill and Woodcock, 1955) and is reproduced in Table 1 by permission of the Editors. The outstanding feature of these results is the complete lack of activity of those acids having an odd number of carbon atoms in the side-chain. By contrast )'-(2-naphthyloxy)-n-butyric and f-(2-naphthyloxy)-n-caproic acids with four and six carbon atoms respectively in the chain proved as active in this test as 2-naphthyloxyacetic acid itself, whilst >/-(2-naphthyloxy)- octanoic acid showed definite though reduced activity. Owing to the very low biological activity and low solubility of the two highest members of the series, (-(2-naphthyloxy)-decanoic and A-(2-naphthyloxy)-dodecanoic acids, ■]■ This paper was read at the Conference by Mr. D. L. Abbott. 195 Chemical structure and biological activity the ED 50's were well above the highest concentrations of these compounds it was possible to test, and were estimated directly by extrapolation of the probit lines. This phenomenon of alternation in growth-regulating activity has been previously reported for other homologous series of growth substances. Thus Thimann and Bonner (1938) noticed an alternation in activity (as measured Table 1 Relative molar activities of a homologous series of oi-{2-naphthvloxy)-n-alkylcarboxylic acids in the tomato ovary test (All compounds tested as sodium salts) Formula Solubility R _ ofKa Relative XV r^ M.p. salt in molar Name y\/\ o of water activity acid at at \/\/ 25-27 C p. p.m. ED 50 2-Naphthyloxyacetic acid R • CHoCOOH 155-156^ 6,550 1.000 /?-(2-Naphthyloxy) -propionic acid R(CH,).,COOH 143-144^ 7,650 y-(2-Naphthyloxy)-butyric acid R(CH,)3COOH 122-123" 5.900 1,000 (5- (2-Naphthyloxy) -valeric acid R(CH2)4COOH 111-112° 5.300 (126-127't 3.750 1.000 £-(2-Naphthyloxy)-caproic acid R(CH2)5COOH 1 94-95= 45 1,000 ^-(2-Naphthyloxy)-heptanoic acid R(CH,)6COOH 96-97= 6.100 ?;- (2-Naphthyloxy) -octanoic acid R(CH2)7COOH 91-92" 6.650 180 6-(2-Naphthyloxy)-nonanoic acid R(CH,)8COOH 98-99° 2.250 t-(2-Naphthyloxy)-decanoic acid R(CH2)9COOH 114-115- 500 n k- (2-Naphthyloxy) -dodecanoic acid R(CH.,)iiCOOH 96-97= 70 n + Two dimorphic foims of this compound were tested. X ED 50 estimated by extrapolation of probit line. by the Avena test) in the indolyl series, C8HgN(CH2) „COOH (where n varies from to 4) and Grace (1939) investigadng the naphthylacetic acids CioH7(CH2) „COOH found greater root inducing activity with the acids having an even number of carbon atoms in the side-chain, though a general decrease in activity with increasing chain length was apparent. More recently, Synerholm and Zimmerman (1947) reported a similar phenomenon in the first seven members of the oj-(2:4-dichlorophenoxy) aliphatic series. These authors also tested the first three members of the co- (2-naphthyloxy) series and found /5- (2-naphthyloxy) -propionic acid to be inactive for cell elongadon in the tomato leaf epinasty test. Synerholm and Zimmerman (1947) put forward an hypothesis to account for these results based on an analogy with the oxidation of fatty acids in animal metabolism, where it is known that oxidation normally occurs at the /3-carbon atom, leading to the formation of an acid with two fewer carbon atoms. Thus it was postulated that a compound such as e-(2:4-dichlorophenoxy)-caproic acid (with 6 carbon atoms in the side-chain) was active as a growth regulator because by 196 Naphthyloxy compounds as plant growth regulators successive /^-oxidations it could be converted by the plant into the biologically active 2:4-dichlorophenoxy acetic acid, thus: CV OCH2CH2C4H2CH2 CI a CH2COOH CI /5 • a OCH2CH2 CH2COOH CI CI OCH2COOH CI Those members of the series with an odd number of carbon atoms, on the other hand, would be oxidized via the hypothetical carbonate to 2:4- dichlorophenol, which is inactive as a growth regulator. •CHoCOOH r ^x .OCOOH1 ^v ^OH CI OCR CI LCI CI CI CI This hypothesis has received strong support from the work of Fawcett, Ingram, and Wain (1952), who produced evidence for the production of phenol in flax plants treated with oj-phenoxyalkyl carboxylic acids with an odd number of carbon atoms in the side-chain. Wain and Wightman (1954) have also demonstrated indirectly the /5-oxidation within the plant of similar compounds with an even number of carbon atoms in the side-chain. The results presented in this paper for the fo-(2-naphthyloxy)-tt-alkyl carboxylic acids are in good agreement with those of Synerholm and Zimmerman (1947) and Wain and Wightman (1954) and it seems likely that a similar mechanism is operative. The reduced activity of the 8, 10, and 12 carbon atom members compared with 2-naphthyloxyacetic acid requires special comment, since if the /j-oxidation hypothesis is correct these compounds, like the 4 and 6 carbon members, would be expected to show a relative molar activity of 1000. As can be seen from the table, there is a marked falling off in the water-solubility of the compounds having more than 8 carbon atoms in the side-chain. It seems unlikely, however, that this can be considered as a contributory cause of their lower activity for even at concentrations at which they formed true solutions in water, their activity was much less than would have been expected on the assumption of complete oxidation to 2-naphthyloxyacetic acid. Further confirmation of the fact that biological activity and water solubility are not directly related is provided by the two dimorphic forms of e-(2-naphthyloxy)-n-caproic acid which in spite of a very great difference in solubility both gave an activity of 1000 in the 197 14 Chemical structure and biological activity tomato parthenocaipy test (see Table 1). The solubility of the lower melting point dimorph of this compound is, in fact, somewhat less than that of the almost inactive dodecanoic acid. In compounds with eight and ten carbon atoms in the side-chain a certain amount of oxidation may take place at the naphthyloxy end of the chain (i.e. on the co-carbon atom). This phenomenon of (o-oxidation is known from studies on animal metabolism to occur with those fatty acids having 8 to 1 1 carbon atoms, and has also been postulated as possibly occurring in plant tissues by Fawcett et al. (1952). It would thus be possible, on the assumption of co-oxidation, to account for the reduced activity of the octanoic, decanoic, and dodecanoic acids. It seems more probable, however, that the falling-off in activity of the 8, 10, and 12 carbon acids is due primarily to restrictions on the penetration of these molecules through the plasma membranes imposed by the length of their fatty side-chains, rather than to any differences in their Table 2 Relative molar activities of certain o)-{'i-chloro-2-naphthyloxy)-rv-alkylcarboxylic acids in the tomato ovary test (All compounds tested as sodium salts.) Formula Ri — Relative M.p. molar Name yxyv o of activity acid at Cl ED 50 3-Chloro-2-naphthyloxyacetic acid RICH2COOH 182-183° 1,000 /?-(3-Chloro-2-naphthyloxy) -propionic acid Ri(CH2)2COOH 173-174° y-(3-Chloro-2-naphthyloxy)-butyric acid Ri(CH,)3COOH 174-175° £-(3-Chloro-2-naphthyloxy)-caproic acid RifCHaijCOOH 114-115° j/-(3-Chloro-2-naphthyloxy)-octanoic acid Ri(CH2)7COOH 74-75° <-(3-Chloro-2-naphthyloxy)-decanoic acid RVCHOsCOOH 90-91° primary activity at the site of action in the cytoplasm. If this is so, the very sudden drop in activity between the 6 and 8 carbon acids would imply that there is some critical side-chain length below which penetration is not affected and above which it becomes the factor limiting the over-all biological activity of the molecule. Studies on the effect of nuclear substitution in 2-naphthyloxyacetic acid have shown that the biological activity of this molecule in the tomato parthenocarpy test is unimpaired by the introduction of a chlorine atom in position 3. In the higher homologues of this series, however, the activity of the butyric, caproic, octanoic, and decanoic acids is completely destroyed by a chlorine atom in this position (see Table 2), which must presumably block the ^-oxidation process. A similar phenomenon has previously been reported, though without comment, by Synerholm and Zimmerman (1947). These authors found that y- (2 :4-dichlorophenoxy) -butyric acid, as would be 198 Naphthyloxy compounds as plant growth regulators expected, was active as a growth regulator, as was also 2:4:5-trichloro- phenoxy acetic acid, but y-(2:4:5-trichlorophenoxy) -butyric acid, which should be active, was not. Wain and Wightman (1954) studying the same compounds also found y-(2:4:5-trichlorophenoxy) butyric acid and higher homologues to be inactive in the pea and tomato leaf epinasty tests, though the homologues with an even number of carbon atoms in the side-chain were active in the wheat coleoptile cylinder test. This would imply that the /5-oxidase capable of degrading the higher homologues of 2:4:5-trichlorophenoxyacetic acid is absent from pea and tomato tissue but present in wheat. The absence of the appropriate /j-oxidase from tomato ovaries might therefore provide an alternative explanation for the lack of activity of the homologues of 3-chloro- 2-naphthyloxyacetic acid in the present investigation. a- (2-naphthyloxy) -h-alkylcarboxylic acids The introduction of alkyl groups of gradually increasing chain length on the a-carbon atom of 2-naphthyloxyacetic acid causes a rapid decrease in activity, the first homologue to be completely inactive being a- (2-naphthyl- oxy) -;2-valeric acid. In a similar series starting with 3-chloro-2-naphthyloxy- acetic acid activity is again rapidly lost {Table 3). Resolution of a-(3-chloro-2-naphthyloxy)-propionic acid using cinchonine (Pope and Woodcock, 1955) has given a highly active (-f ) form and an inactive optical antipode. Examination of this compound by Matell (1955) has shown that it most probably has the D-configuration, thus supporting the view that optically active plant growth regulators of the a-aryloxyalkyl carboxylic acid type which have greater auxin activity than their respective antipodes belong to the D-series. It is also noteworthy that the ( — ) form here shows competitive antagonism with the (-f ) form, resulting in an activity of only 60 for the {^) mixture instead of an expected value of 125. This agrees with the findings of Aberg (1951) with a- (2-naphthyloxy) -propionic acid in root inhibition, and of Smith, Wain and Wightman (1952) with a-(2:4-dichloro-phenoxy)-, a-(2 :4:5-trichlorophenoxy)- and a- (2-naph- thyloxy) -propionic acids in cell elongation promoting activity. In general the activity of these a-substituted acids is much lower than the corresponding 2-naphthyloxyacetic acid and the fairly rapid decrease with increasing chain-length would seem to indicate a size or, rather less likely, a solubility effect. The close relationship between activity and the position of nuclear substituents will be more obvious after the results in the next section have been reviewed. SUBSTITUTED 2-NAPHTHYLOXYACETIC ACIDS In one of the most recent and critical i^eviews of structure-activity relation- ships in the field of plant growth regulating substances, Jonsson (1955) has reformulated the requirements for appreciable auxin activity as follows: (i) An unsaturated ring nucleus. (ii) A side-chain carrying a — COOH group which must not be situated at a quaternary carbon atom (or otherwise highly sterically hindered). (iii) A high interface activity of the extended ring system. (iv) The side-chain must be able to assume a spatial orientation in which 199 Chemical structure and biological activity t2 :§ I c I I o lO Q W •<» a •^ C' CT-; O O inooooomcMOOiooo | ^ lO Xi lO -. CM a CM CM _« ~5 g ■u .a ■*^ ^ 1 1 ^ Q C J ■^ a; -^ o o o o o oooooooo lO I^ 00 00 (O OOCOCMCOCM— 'r~~o o • — ' CM — O r^ LOUDCMt^COCOCIjr^CMCO ■^ ^^ *— 1 ^^ «— • ^^ ^^ ^^ .—1 CO o^ ■^^ 1 1 1 1 T 1 1 1 1 1 1 1 1 1 1 ■< rf U3 t^ r^ m CTiCTir^.— 'CM — iO<,D^^r^ — ' CN — ' O r^ '^lOCMr^COCOtOI^COCJ) 1 a -naphthylo. ^xxx §000 §000 5oo5:xoOooo oooOO§ooqq OOUOOuOOOO Formul ubstituted 2 y — V ift r- 0> hH p^ P^ X HH W rt T o o u o X XXX o u u o u IT K K ^^P^ S r^ ^ ^ ^ 1— ( 2 > 1— 1 1— ( II p!ipiPip< p^ p:;p:;piOipi;o::(:<>$>, >,>,>,>.>>>^>.>>>H>> a .2 ^ fc ^ X X X X X -A jz x. j: X. jz ^. 9- 3 "w rt _0 ^^^^^x^.S.S^ P -D > 3 ">- >^>,>,>N>>Q.Q^£S^Q_fl. a i i i 1 1 1 1 ji: XXX-C^rtrtrtK!c« -t-* ^ -— — H -^-1 -^- ..^- 1 1 1 1 1 >^ >> >• >> a O-O-DhCXD-cmcmc^icmcm X X X X _o _o _o _o 03 cccccooggg >. >^ >> >^ 1 C--I CMCM(>J(>JC^^^^^^ ^ ^ ^ X 1 o o 6 6 f, -f f , f , f X -S _H -H a D. a a Ul l,t^UUuOOOUO _0 ^^°^^^^T3^^ c^ c^ cd c^ C C C C 1 1 1 1 CM CM CM CM r-; O — — — — ^ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 « » 8 8 1 8 1 1 1 1 1 1 1 1 1 1 200 Naphthyloxy compounds as plant growth regulators the atoms joining the ring and the — COOH group, including the C atom of the latter, are situated almost in the plane of the ring forming a pseudo-ring with the — COOH group close to the original ring and near the centre of the extended ring system thus formed. (v) At least one side of the plane of the extended ring system must be free of atoms other than hydrogen and one of the O atoms of the — COOH group projecting out from it. (vi) An optical configuration corresponding to the D-series of the amino- acids. This formulation is essentially an elaboration of the views of Veldstra (1944a,b) in which his 'certain spatial relationship' between the ring and the side-chain now appears as an 'extended ring system.' It will be convenient to view the results given in Tables 3 and 4 against this background. Table 4 Substituted 2-naphthyloxyacetic acids Relative molar Substituent M.p. activity at ED 50 1-chloro 169-170^ 1 3-chloro 182-183^ 1.000 4-chloro les-ieg^" 2 5-chloro 202-203^ 6-chloro 208-209 72 7-chloro 168-169° 19 8-chIoro 131-132° 280 5-nitro 211-212° 8-nitro 179-180° 29 1 : 3-dichIoro 179-180° 1 :4-dichloro 188-189^ 3:4-dichloro 182-184 20 l:3:4-trichIoro 207-208^ In general it may be said that nuclear chlorine substitution increases the activity of phenoxyacetic and benzoic acid growth substances and in certain positions, especially 4, 5, and 6, with 3-indolylacetic acid activity is only slightly diminished or is even enhanced (Veldstra, 1953). By contrast no increase in activity of 2-naphthyloxyacetic acid can be obtained by nuclear substitution of chlorine. On Veldstra's hypothesis (Veldstra and Booij, 1949) this could be explained by assuming that in 2-naphthyloxyacetic acid the lipophilic/hydrophilic character (L/// ratio) of the molecule is at its optimum for biological activity, chlorination merely serving to increase the LjH ratio beyond the optimum, resulting in a reduction in activity. The same argu- ment would apply to the effect of lengthening the side-chain in the a-acids, but not to the co-acids where the side-chain can be broken by /:?-oxidation. Considering the effect of chlorine substitution at specific positions in the ring (see Table 4) it is immediately obvious that the important ring positions are 1, 4, and 5. When any of these are blocked with CI there is a complete or 201 Chemical structure and biological activity almost complete loss of activity, both in mono and in poly-substituted compounds. Matell (1953) has also reported that the activity of (4-)-a-(2- naphthyloxy) propionic acid in the flax root growth test is lowered about 600 times when chlorinated in position 1. He also makes the significant comment that the strong antagonistic action of the ( — ) form is very little affected by chlorination. Positions 6, 7, and 8 are also important for activity, though blocking of these points with chlorine causes a reduction but not complete loss of activity. It is interesting to note that replacement of the chlorine in position 8 by a nitro-group causes a further ten-fold reduction in activity. In the first results of substitution in the 3-position which we now report, it is seen that a chlorine atom here has no detrimental effect on the activity in the acetic acid series, though some reduction in the a-propionic acid series takes place, being part of the gradual decrease in the activity of a-substituted acids already mentioned. An examination of Fischer-Hirschfelder molecular models of chlorinated 2-naphthyloxyacetic acids shows that chlorine atoms in positions 1 and 3 interfere to some extent with the free rotation of the — O • CHgCOOH side- chain, and thus influence the final spatial arrangement of atoms. Although the naphthalene nucleus is planar, large substituent groups such as chlorine may cause a buckling effect to relieve strain, which would destroy planarity, and in certain cases possibly affect the compactness of the molecule. It is unfortunate that the models give little help on this point. Jonsson (1955) does however stress that a planar structure is not the only requirement for bicyclic molecules to possess high activity. The electronic effects of nuclear chlorine atoms may also be important. Reaction with other molecules such as coenzyme A forming an acetyl-thiol ester as has been suggested by Leopold and Guernsey (1953) would appear to be encouraged by chlorine atoms in positions 1, 3, 6, and 8 but not in positions 4, 5, and 7 as shown in the following diagrams: /V V-O • CH2 • COOH y\XV%LO . CH2 • COOH CI . CHo • COOH (fS^O . CH. • COOH The inactivity of the 1 -chloro-compound would then have to be explained by a superimposed steric effect involving the side-chain (Jonsson's activity requirements (iv) and (v)). Two other hypotheses concerning mode of action have suggested inter- action of growth substances with a hypothetical plant protein. The papers of 202 Naphthyloxy compounds as plant growth regulators Hansch and Muir (1950) who postulated a two-point attachment involving the carboxyl group and one or/Ao-position, and Wain (1953) who presented much evidence in favour of a three-point attachment, an essential feature of which was a hydrogen atom on the a-carbon atom, are well known. In order to explain the importance of certain positions in 2-naphthyloxy- acetic acid for growth substance activity, it is necessary to postulate a four- point, or in order to meet Wain's a-hydrogen requirement, a five-point attachment theory. This would involve in addition to one hydrogen atom on the a-carbon atom of the side-chain, and the carboxyl group, positions 1 , 4, and 5 in the nucleus. Substitution in these positions leads to inactivity, whilst substitution in positions 6, 7, and 8 could reduce activity by hindering close interlocking with the substrate molecule. This hindrance would be greater the larger the substituent atom or group of atoms. Thus 8-nitro- 2-naphthyloxyacetic acid would be expected to be less active than the corresponding 8-chloro compound, a fact which has already been noted. Although it is possible on the basis of such a five-point attachment theory, to explain all the known results so far obtained relating to 2-naphthyloxy- acetic acid, several criticisms may be made not least of which is the fact that an entirely hypothetical substrate, different from that postulated by Muir and others for phenoxy growth substances is involved. Although different types of growth substances may each have their own substrates, this seems unlikely in view of the great similarity in the growth responses they produce. REFERENCES Aberg, B. (1951). On the growth substance activity of enantiomorphic a-(naph- thoxy) -propionic acids. Ark. Kemi. 3, 549. Fawcett, C. H., Ingram, J. M. A., and Wain, R. L. (1952). /5-oxidation of w-phen- oxyalkylcarboxylic acids in the flax plant. Nature, 170, 887. Grace, N. H. (1939). Physiological activity of a series of naphthylacetic acids. Canad. J. Res. (C) 17, 247. Hansch, C, and Muir, R. M. (1950). The ortho effect in plant growth regulators. Plant Physiol. 25, 389. James, P. M., and Woodcock, D. (1951). Synthesis of plant growth regulators. Part I. Substituted /3-naphthyIoxyacetic acids. J. chem. Soc. 3418. James, P. M., and Woodcock, D. (1953). Synthesis of plant growth regulators. Part II. Dichloro-/?-naphthyloxyacetic acids. J. chem. Soc. 2089. JoNSSON, A. (1955). Synthetic plant hormones. VIII. Relationship between chemical structure and plant growth activity in the arylalkyi-, aryloxyalkyl-, and indole- alkylcarboxylic acid series. Svensk. kem. Tidskr. 67, 166. Leopold, A. C, and Guernsey, F. S. (1953). A theory of auxin action involving coenzyme A. Proc. nat. Acad. Sci. Wash., 39, 1105. LucKWiLL, L. C. (1948). A method for the quantitative estimation of growth substances based on the response of tomato ovaries to known amounts of 2-naph- thoxy-acetic acid. J. hort. Sci. 24, 19. LucKWiLL, L. C, and Woodcock, D. (1955). Plant growth regulators. I. The influence of side-chain length on the activity of cj-(2-naphthyloxy)-n-alkylcarboxylic acids for the induction of parthenocarpy in tomatoes. J. hort. Sci. 30, 109. Maxell, M. (1953). Stereochemical studies on plant growth substances. LantbrHogsk. Ann. 20, 233. Maxell, M. (1955). Private communication. V_ 203 Chemical structure and biological activity Pope, P. M., and Woodcock, D. (1954). Synthesis of plant growth regulators. Part III. ('j-2-naphthyloxyalkanecarboxylic acids. J. Chern. Soc. 1721. Pope, P. M., and Woodcock, D. (1955). Synthesis of plant growth regulators. Part IV. Substituted 2-naphthyloxy propionic acids. J. chem. Soc. 577-579. Smith, M. S., Wain, R. L., and Wightman, F. (1952). Studies on plant growth regulating substances. V. Steric factors in relation to mode of action of certain .aryloxyalkylcarboxylic acids. Ann. appl. Biol. 39, 295. Synerholm, M. E., and Zimmerman, P. W. (1947). Preparation of a series of co-(2:4-dichlorophenoxy)-aliphatIc acids and some related compounds with a consideration of their biochemical role as plant growth regulators. Contr. Boyce Thompson Inst. 14, 369. Thimann, K. v., and Bonner, J. (1938). Plant growth hormones. Physiol. Rev. 18, 524. Veldstra, H. (1944a). Researches on plant growth substances. IV. Relation between chemical structure and physiological activity. I. Enzymologia, 11, 97. Veldstra, H. (1944b). Researches on plant growth substances. V. Relation between chemical structure and physiological activity. II. Enzymologia, 11, 137. Veldstra, H. (1953). The relation of chemical structure to biological activity in growth substances. Ann. Rev. PI. Physiol. 4, 164. Veldstra, H., and Booij, H. L. (1949). Researches on plant growth regulators. XVII. Structure and activity. On the mechanism of the action. III. Biochem. biophys. Acta, 3, 278. Wain, R. L. (1953). Plant growth substances. Royal Inst. Chem. Monograph, No. 2. Wain, R. L., and Wightman, F. (1954). The growth-regulating activity of certain oj-substituted alkylcarboxyhc acids in relation to their /jf-oxidation within the plant. Proc. Roy. Soc. B142, 525. 204 STUDIES ON THE RELATION BETWEEN MOLECULAR STRUCTURE AND PENETRATION OF GROWTH REGULATORS INTO PLANTS J. VAN OVERBEEK Agricultural Research Division, Shell Development Company. Modesto. California INTRODUCTION The penetration of auxins into plants is a property which so far has not been given much attention. One generally accepted reason for this is that it has not been possible to locate a specific site in the molecule which can be held responsible for auxin activity. It is therefore impossible to distinguish between those properties of the auxin molecule which cause it to penetrate and those which cause it to react inside the cell as an auxin. It thus becomes necessary to reason by analogy, that is to study the relation between physiological activity and molecular structure in compounds which possess something that resembles an active grouping in the molecule, even though this activity is not auxin activity. An example of a compound of this type is maleimide. There is actually a whole group of maleimides (all having the maleimide ring in common), differing from one another by the substituent on the nitrogen atom. It is reasonably certain that the anti-auxin properties of the maleimides reside in the maleimide ring (more precisely in the sulphydryl reactivity of its double bond) and that the substituents on the nitrogen atom merely function as a 'lipophilic tail' to carry the molecule into the cell. THE PROBLEM In Figure 1 has been presented a number of comparisons in which the increased auxin activity was caused by a modification of the molecular structure. In the writer's opinion the resulting increases in physiological activity of these compounds can be satisfactorily explained by increased penetration into the cells. One can distinguish three types : ( 1 ) alpha substitution of phenyl- and phenoxyacetic acids, (2) phenyl- versus phenoxyacetic acids, and (3) chlorina- tion of the benzene ring. The relatively more active member of each pair of compounds appears on the right, and is indicated by -(-. We will now scruti- nize the evidence for this contention on the basis of the study of the maleimides and later also on another pertinent example, that of the amino phosphine oxides. The maleimides The maleimides are anti-auxins in the sense that they inhibit the growth rate of coleoptile sections growing under the influence of applied indoleacetic acid. The inhibition of the growth rate is restored by increasing the con- centration of the auxin. These relations are shown on maize coleoptile sections in Fimre 2. 'to' 205 Chemical structure and biological activity V^ ® I ■0-C-COOH H (Vluir and Hansch,1953) 0-C-COOH I + aC-COOH (Veldstra,l953) V^ -C — COOH I H [;J=^^^^>p0-CH2C00H Thimann, 1951 Wahr//soOdane partition (Van Ovenbeek et al.,1355) c^^^^CHjCOOH 270 15 aO-CHgCOOH (Thimann, i95l) CI f;==^0CH2C00H CI ^ V^ CI — C — COOH I H CL ^^ ■0 — C — COOH I CI H (Osborne etal,l954) Figure 1. Comparison between pairs of regulators. The member on the right has the higher physio- logical {auxin) activity. It is argued that this is so because the more active member penetrates the cell better. Figure 2. Reversible inhibition of indoleacetic acid (lAA) induced growth of maize coleoptile sections by 2:4:-dichlorophenylmaleimide {MAL). {After van Overbeek et al., 1955.) 206 Molecular structure and penetration of growth regulators One of the more striking manifestations of the anti-auxin reaction is defoliation. A quantity of iV-1-naphthylmaleimide as small as 0-35 gamma per cm''^ leaf area will cause the complete defoliation of peach branches in three days. The leaves drop entirely green, without any evidence of chemical burn. Simultaneous application of the auxin naphthalene acetic acid (at O-I of the concentration of the maleimide) prevents defoliation completely {Figure 3). This again shows the anti-auxin nature of the maleimides. Since the maleimides react with SH compounds, and in addition lose their anti-auxin activity when the double bond is saturated, it is reasonably certain that the biochemical activitv of the maleimide molecule resides in the CONTROL EMULSION ONLY NAR MAL. NAP. MAL 2XI0-3M eXIO'^M + NA2XI0-'*M Figure 3. Defoliation of peach branches by naphthylmaleimide. When naphthalene acetic acid was applied simultaneously with the maleimide, the defoliant effect of the latter was prevented. {From van Overbeek et al., 1955, by courtesy of the American Journal oj Botany.) maleimide ring rather than in the substituents on the nitrogen atom.There- fore, if there was no penetration factor involved, all maleimides could be expected to give equal defoliation on the peach branches. This is clearly not so. In fact, /^opropylmaleimide was not active at all within the range of concentrations used in these tests [Figure 4). Phenylmaleimide has inter- mediate activity, while 2 :4-dichlorophenylmaleimide has high activity. Naphthylmaleimide has an equally high activity. Originally we were inclined to conclude that high anti-auxin activity is obtained when the maleimide possesses an 'auxin ring' from a highly active auxin. We soon learned, however, that no ring is necessary for anti-auxin activity and that the substitution on the nitrogen atom may be an ordinary aliphatic radical. Thus it was found that octylmaleimide has considerable activity [Figure 4). It appears, therefore, that the degree of physiological activity of the maleimides is determined by the character of their 'lipophilic tail'. The latter serves to carry the molecule into and perhaps through the lipoidal plasma membrane into other lipophilic bodies of the cytoplasm. The water solubilities of the maleimides [Figure 4) are low in the physiologically active materials and high in the inactive compounds of the group. From this and 207 Chemical structure and biological activity the structure of the compounds it follows that the active materials have a greater tendency to go into an oil phase than the physiologically inactive ones. Especially striking is the difference between chlorinated and un- chlorinated phenylmaleimides. This points to the conclusion that one large factor in the increased physiological effect of benzene chlorination is the increased fat solubility that goes with this type of chlorination. HC- HC- / \ N — R / R Water solubility p.p.m. % defoliation H\CH3 5000 5 ^\ 100 20 -/> 10 55 CL 5 70 <> 5 75 (ti) 10 35 Figure 4. Structures, water solubility, and physiological activity of maleimides. {After van Overbeek et al., 1955.) The amino phosphine oxides While in the maleimides we can identify the reactive grouping in the molecule, we do not know with what active site in the cell this grouping reacts. In the amino phosphine oxides, on the other hand, we have a rare example in which we can identify both the active grouping in the molecule and the site in the cell with which it interacts. The structure of a reactive amino phosphine oxide is shown in Figure 5. The reactive grouping in the molecule is the amino phosphine oxide grouping, while the butyl radicals serve to give the molecule the proper lipophilic characteristics. The reactive site in the cell is the chloroplast, and more specifically chlorophyll molecules. When a benzene solution of chlorophyll is exposed to light, or exposed to the action of other destructive agents such as acids, its colour will soon turn from green to brown because the Mg atom is lost. The presence of equimolar quantities of amino phosphine oxides will greatly retard this chlorophyll degradation. Chromatographic studies have shown that chlorophyll and amino phosphine oxides form a complex. Polar solvents such as alcohols separate the components again. In the test tube all amino phosphine oxides of the type considered here are capable of complexing with chlorophyll, whether the radicals on the P atom are methyl, ethyl, propyl, allyl, butyl, or amyl. Biochemically speaking, all these compounds are active, and if there were no penetration problem all these materials should be equally active when applied to living cells. This is not so. 208 Molecular structure and penetration of growth regulators When the amino phosphine oxides of this series are put in aqueous solutions in which Elodea leaves are present, only the higher homologues show evidence of physiological activity. This is manifested under lOOOx magnification by blue green droplets in the chloroplasts. These droplets are quite character- istic for the presence of amino phosphine oxides in the chloroplasts, and thus serve as an index for penetration. The lower homologues of the series, by contrast, do not penetrate into the cell and the chloroplasts, or only do so when supplied at a much higher concentration. H3C HaC- \H N- H3C- / /H H3C R Water/oil pant.coeff Relative penetration into cell -CH3 4 ,. / -C2H5 >0-1 I 25 ' -C3H7 0-1 90 ^ -C4H9 0-OOS 100 -C5H11 0-0007 110 Figure 5. Structures, partition coefficient, and physiological activity of dii^opropylamino phosphine oxides. When partition coefficients (water/olive oil) were made for this series of amino phosphine oxides, it was found that the more the compound partitions into the oil phase the more it penetrated into the chloroplasts. This is not surprising because chloroplasts are highly fatty bodies (Chibnall, 1939). CONCLUSION It has thus been demonstrated on these two examples, maleimides and amino phosphine oxides, that physiological activity of the molecule (as contrasted with biochemical activity inside the cell) depends on the presence of a lipoid radical whose function it is to carry the molecule into the lipoid phases of the cell. It has further been shown that simple aliphatic chains will often perform this function. By analogy it is thus concluded that the increased auxin activity of the alpha substituted growth regulators oi" Figure 1 is also due to increased oil solubility. It has further been shown that chlorination of the benzene ring caused increased lipophilic properties in the maleimide molecule. By analogy it is concluded that at least part of the increased auxin activity caused by a similar chlorination in phenyl- and phenoxy acetic acids is simply due to increased lipoid solubility. We compared the water/oil partition coefficients of phenoxy- and phenyl- acetic acids and found that the phenyl partitioned far more into the oil {Figure 1). It would seem, therefore, that the long-known difference in auxin activity between these two acids is also a logical result of their difference in capacity to penetrate into the lipoid phases of the plant. R. C. Brian substantiated this conclusion (during the discussion of another 209 Chemical structure and biological activity paper in this Symposium) from his studies on' the adsorption and penetration of regulators on and into monolayers of plant lipoids. He found that two chlorinated phenylacetic acids penetrated the film up to 5 times as much as the corresponding phenoxyacetic acids. REFERENCES Chibnall, a. C. (1939). Protein Metabolism in the Plant. Yale University Press. MuiR, R. M., and Hansch, C. (1953). On the mechanism of action of growth regulators. Plant Physiol. 28, 218. Osborne, D. J., Blackman, G. E., Powell, R. G., Sudzuki, F., and Novoa, S. (1954). Growth regulating activity of certain 2 : 6-substituted phenoxyacetic acids. Nature, 174, 742. OvERBEEK, J. VAN, Blondeau, R., and HoRNE, V. (1955). Maleimides as auxin antagonists. Amer. J. Bot. 42, 205. Thimann, K. V. (1951). The synthetic auxins: relation between structure and activity. Plant Growth Substances, ed. F. Skoog, University of Wisconsin Press, p. 21. Veldstra, H. (1953). The relation of chemical structure to biological activity in growth substances. Ann. Rev. PI. Physiol. 4, 151. 210 THE RESOLUTION OF PHYTO-HORMONE ACIDS BY MEANS OF THEIR OPTICALLY ACTIVE PHENYL/^OPROPYLAMIDES. RESOLUTION OF a-(l-NAPHTHYL)PROPIONIC ACIDt J. M. F, Leaper and J. R. Bishop American Chemical Paint Company, Ambler, Pa. HISTORY OF THE RESOLUTION OF PHYTO-HORMONE ACIDS The term 'hormone acids' will be used here to include two main groups: (a) Acids actually derived from plant sources, or closely related chemically to these; e.g. indole-3-acetic acid, a-(indole-3)propionic acid, and a-(indole-3)«-butyric acid. (b) Acids not related to these and arising solely from synthetic sources, e.g. the substituted and unsubstituted phenoxy- and naphthoxyacetic acids and homologues and their related imino and thio derivatives. Also in this category will be the a- and /:?- naphthylacetic acids and their homologues. Some of both these classes have an asymmetric carbon atom and are capable of being resolved into their optically active enantiomorphs. It was fairly obvious from the start and in fact might be considered axio- matic that when such resolutions were brought about, the resulting two isomers would produce different physiological responses in the appropriate plant tissue. I say appropriate advisedly since some plant response can often be found that does not differentiate between the two isomers, e.g. the variations observed by Kogl (1944) in the pea test and the Avena test using indole-a-propionic acid isomers. However, these are comparatively rare exceptions. As has often been pointed out, the parallel cases in animal physiology involving the optical isomers of such compounds, both naturally occurring and synthetic, as thyroxine, hyocyamine, adrenaline, ascorbic acid, amphetamine, and many others, where very striking differences in biological reactions are found as between the isomers, would seem to make it obvious that similar differenti- ations would be the case with compounds having phytobiological activity. Actually, the first resolution of what we are calling 'hormone acids' was made as far back as 1925 by Fourneau and his co-workers (Fourneau and Balaceano, 1925) at the Pasteur Institute, who resolved a-(2-naphthoxy)- propionic acid. However, this was many years before the phytobiological activity of 2-naphthoxyacetic acid was disclosed by Irvine (1938) and the relative activity of the isomers of a-(2-naphthoxy)propionic acids studied by Smith and Wain (1951). Fourneau was primarily interested in producing a cheap and readily available optically active acid for use in resolving synthetic adrenaline and related vasoconstrictors. It is worthy of note that whereas most of the useful medicinal compounds found naturally are optically active, the only well-defined phyto-hormone ■fThis paper was read at the Conference by J. M. F. Leaper. 211 Chemical structure and biological activity acid so far isolated is the, of course, optically inactive indoleacetic acid. It seems to me a reasonable speculation that the latter is perhaps the biological ultimate result of breakdown in the plant of similar acids of longer and partially branched side chains, just as in the explanation of the breakdown in the carbon pairs of the simpler a>-acids, postulated by Synerholm and Zimmerman (1947), except that in the latter case the starting acids were not potentially optically active. The first conscious effort then to resolve any hormone acid with the idea of finding out something of the contrasted activities of the enantiomorphs was made by Kogl and Verkaaik in 1944 in synthetic a- (indole-3) propionic acid (1944). It is notable, as pointed out above, that they found that the (+) isomer had thirty times the activity of the ( — ) in one test and that in another test the differences were negligible. At about the same time our group in the Agricultural Chemicals Division of the American Chemical Paint Company in Ambler, after launching the 2:4-D type of compounds for the use of American agriculture, turned our attention to homologues of this subst