STUDIES ON CYTOKININ BINDING IN TOBACCO CELLS Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY MICHAELRICHARD SUSSMAN 1976 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 1293 10516 7567 L .i R Y Michigan State University I n" 3‘ a ‘5 This is to certify that the thesis entitled Studies on cytokinin binding in tobacco cells. presented by Xiohael Richard Sussman has been accepted towards fulfillment of the requirements for Ph 0D ' degree in P0138 ny Ila, (Leggy—— Major professor Date 3a” Lr g/I??7 0-7639 ABSTRACT STUDIES ON CYTOKININ BINDING IN TOBACCO CELLS BY Michael Richard Sussman For the detection of a small number of higher-affinity cytokinin binding sites, high specific activity Ne-benzyladenine p-[3H]-benzyl (p—[3H1-BA) was synthesized by a simple procedure. p-[3H1-BA at 10 Ci/mmol was obtained by catalytic dehalogenation of p—bromo-N6-benzyl- adenine with carrier-free tritium gas. A radiochemical purity of 97% was obtained by a single TLC purification step. Using this p-[3H1-BA and a particulate fraction from cytokinin- dependent tobacco cells grown in suspension culture, high- and low— affinity cytokinin binding sites were observed. The low-affinity binding site was stable to a heat pre-treatment and showed some speci- ficity for cytokinin-like compounds (e.g., kinetin, 21F and BA bind well, while adenine and the cytokinin ribosides do not). However, when the specificity was examined in finer detail with a series of chemically similar cytokinin analogues possessing different degrees of biological activity, no correlation was observed between binding and biological activities of the analogues. In contrast to these results with the low-affinity binding site, the high—affinity binding site was heat—labile and the ability of cytokinin analogues to compete with BA correlated well with their Michael Richard Sussman biological activity. Thus, on the basis of heat denaturability and binding specificity, it was concluded that only the high-affinity binding site is specific, and possesses the characteristics expected for a cytokinin receptor protein. This specific binding site in a particulate reaction of tobacco cells is probably not identical to the cytokinin-binding protein isolated from wheat germ ribosomes by Fox and Brion (1975). In con— trast to their findings, 1 M KCl did not solubilize the specific cytokinin binding sites. I also observed high- and low-affinity, saturable binding sites for p-[3H]-BA using the non-biological material, talc. However, in neither case was the binding specificity with talc similar to that observed in any cytokinin bioassay. The analogues which were synthesized and tested for cytokinin specificity studies were the following BA derivatives possessing a single halogen atom in the N6 benzyl side-group: ortho-F-BA, meta-F- BA, para-F-BA, ortho-Cl-BA, meta-Cl-BA, para-Cl-BA, ortho-Br-BA and para-Br-BA. When the biological activity of the cytokinin analogues was compared in several bioassays, an interesting difference was noted. In a tobacco cell division (suspension culture) assay, the order of biological activity was BA > ortho-Cl-BA > meta-Cl-BA = para-Cl-BA (inactive), but in the moss bud formation assay, the observed order of activity was meta-Cl-BA 3 BA > ortho-Cl-BA = para-Cl-BA. Models emphasizing either steric or electronic properties of the N6 benzyl side-group for binding to a cytokinin receptor were proposed to explain these results in the tobacco and moss systems, respectively. Michael Richard Sussman In addition, in a tobacco callus shoot formation assay, the order of activity of the analogues paralleled their activity in the cell division assay with the same tissue, though one compound, meta- Cl-BA, stimulated shoot formation with little or no discernible stimulation of cell division. Efforts to determine the precise cellular location of the spe- cific cytokinin binding site were greatly hampered by the large back— ground of non-specific binding. Thus, for a more quantitative and detailed characterization of specific cytokinin binding sites, an alternative technique, photoaffinity labelling, was explored. A potential cytokinin photoaffinity reagent, 8—azido-N6-benzyladenine (8-N3-BA) was synthesized from 8-bromo-adenosine in a four-step reaction sequence with 8-azido-adenosine, l-benzyl-B-azido-adenosine and 8—azido—N6-benzyladenosine as intermediates. 8-N3-BA was found to be a fully active cytokinin in the moss bud formation bioassay and slightly more active than BA in a tobacco cell division (suspension culture) bioassay. In addition, at high concentrations, where BA inhibits cell division, 8-N3-BA did not. 8-N3-BA was readily photolyzed by long- and short-wavelength UV light yielding product(s) which were inactive in the moss bioassay but may have had slight activity in the tobacco bioassay. These results therefore justify the future synthesis and use of radioactive 8-N —BA 3 to covalently label specific cytokinin binding sites. STUDIES ON CYTOKININ BINDING IN TOBACCO CELLS BY Michael Richard Sussman A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1976 To my parents, Stella and Dan ii ACKNOWLEDGEMENTS The author wishes to express appreciation to all of his many friends in the PRL community who helped in his education, both personal and scientific. In particular, the author thanks his thesis advisor, Dr. Hans Kende, for allowing the freedom to pursue interests, for the encouragement to strive for precision in thought and writing and for his patience. Help from the members of his guidance committee, and in par- ticular that of Dr. Anton Lang, is acknowledged with warm appreciation. The author wishes to thank Dr. Renata L. DeZacks for her assist- ance in the bioassays, and members of the mass spectrometry laboratory of Dr. C. Sweeley for all of the mass spectral analyses. Finally, the author thanks Nancy Jo Gaedke for her loving, gentle nature and for assistance in the delivery of this thesis. The research reported here was supported by the U.S. Atomic Energy Commission and the 0.8. Energy Research and Deve10pment Administration under Contract E(ll-l)-1338 and by graduate fellow- ships awarded to the author by the Woodrow Wilson Foundation (1971- 1972) and by the National Science Foundation (1972-1975). iii TABLE OF CONTENTS GENEML INTRODUCTION . C O O O O O O C C O O O O O O C I O O O 0 SECTION 1 - THE SYNTHESIS OF A RADIOACTIVE CYTOKININ WITH HIGH SPECIFIC AmIVITY C O O C O O O C O O O O O O O O 0 Introduction . . . . . . . . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . . . . Synthesis of p-Br-BA. . . . . . . . . . . . . . . Dehalogenation reaction with 1H2 and 3H2 . . . . . Purity and stability of p-[3HI-BA . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . SECTION II - BIOLOGICAL ACTIVITY OF CYTOKININ ANALOGUES . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . . . . Synthesis of cytokinin analogues. . . . . . . . . Tobacco cell suspension: growth characteristics and bioassay procedure . . . . . . . . . . . . Moss bioassay . . . . . . . . . . . . . . . . . Agrostemma bioassay . . . . . . . . . . . . . . . Tobacco callus differentiation bioassay . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . . Physical and chemical properties of ortho-, meta-, and para-halogenated BA derivatives. . . Biological properties of ortho-, meta- and para-halogenated BA derivatives. . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . Speculations on the distribution in nature of cytokinin receptors and the topography of their cytokinin binding sites. . . . . . . . . SECTION III - IDENTIFICATION OF A SPECIFIC CYTOKININ BINDING SITE IN A PARTICULATE FRACTION FROM TOBACCO CELLS. . . Introduction . . . . . . . . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . . . . Tobacco cell culture. . . . . . . . . . . . . . . Preparation of cell-free particulate fractions. . Cytokinin-binding assay . . . . . . . . . . . . . Binding terminology . . . . . . . . . . . . . . Protein and RNA determinations. . . . . . . . . . Enzyme assays . . . . . . . . . . . . . . . . . . Media and radiochemicals. . . . . . . . . . . . . iv Page 24 24 25 25 37 40 4O 4O 44 44 45 45 56 62 62 64 64 65 65 67 67 68 69 Results. . . . . . . . . . . . . . . . . . . . . . Comparison of the binding of 8-[14C]-BA and p-[3H1-BA. . . . . . . . . . . . . . . . . High- and low-affinity saturable BA binding . . Structure—activity requirements of in Vitro p-[3HI-BA binding. . . . . . . . . . . . Localization of the high-affinity cytokinin binding site from tobacco cells. . . . . . . Discussion . . . . . . . . . . . . . . . . Low-affinity binding as measured with 8-[14C]-BA and p-[3Hl-BA. . . . . . . . . . . . . . . . Specific versus non-specific BA binding . . . . Localization of specific cytokinin binding site Where do we go from here? . . . . . . . . . . . SECTION IV - CHEMICAL SYNTHESIS AND BIOLOGICAL ACTIVITY OF 8-AZIDO-BENZYLADENINE, A POTENTIAL CYTOKININ PHOTO- AFFINITY REAGENT . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . . . Synthesis of non-radioactive 8-N3-BA. . . . . . Cytokinin bioassays . . . . . . . . . . . . . . Photolysis of 8-N3-BA . . . . . . . . . . . . . Spectroscopy. . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . Chemical properties and photolysis of 8-N3-BA . Biological activity of 8—N3—BA. . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . GENERAL DISCUSSION. . . . . . . . . . . . . . . . . . . . . . REFERENCES 0 O O 0 O O O O O O O O O O O O O O O O O O O O O O Page 69 69 72 74 83 91 91 91 92 93 95 95 96 96 100 103 103 104 104 104 109 116 119 Table 10 11 LIST OF TABLES Rf values of compounds used in the synthesis of p-[3H]-BA in different thin layer chromatography systems. . . . . . . . . . . . . . . . . . . . . . . . . Sephadex LH-20 chromatographic mobility of cytokinin analogues. . . . . . . . . . . . . . . . . . . . . . . . Effect of cytokinin analogues on shoot formation in tobacco callus . . . . . . . . . . . . . . . . . . . . . . l4 3 . . . Comparison of 8-[ Cl-BA and p-[ H]-BA in binding assays using an 80,000 x g particulate fraction from tobacco cells. . . . . . . . . . . . . . . . . . . . . . Concentration dependence of heat-labile, high-affinity cytokinin binding to an 80,000 x g particulate fraction from tobacco cells . . . . . . . . . . . . . . . . . . . Analogue specificity of high-affinity saturable cyto- kinin binding in an 80,000 x g particulate fraction from tobacco cells . . . . . . . . . . . . . . . . . . . Analogue specificity of low—affinity saturable cyto- kinin binding in an 80,000 x g particulate fraction from tobacco cells . . . . . . . . . . . . . . . . . . . Analogue specificity of high- and low-affinity saturable cytokinin binding to talc (Mallinkrodt, U.S.P.). . . . . Comparison of cytokinin-binding activity, RNA and pro- tein content in various tobacco cell particulate fractions obtained by differential centrifugation. . . . UV spectra of intermediates and related compounds used in the synthesis of 8-N3-BA. . . . . . . . . . . . . . . Thin layer chromatography on silica gel of inter- mediates and related compounds used in the synthesis Of 8-N3-BA . o o a o o o o o o o o o o o o o o o o o o 0 vi Page 11 36 52 71 73 78 79 82 84 98 99 11 ((1)1 II II." I: i/ Figure 15 16 17 18 19 20 21 LIST OF FIGURES Page Mass spectra of benzyladenine (BA) and para-bromo- benzyladenine (p-Br—BA). . . . . . . . . . . . . . . . . . 10 Time course of dehalogenation reaction of p-Br-BA with non-radioactive H2 gas, as measured by GLC. . . . . . 14 Mass spectrum after GLC of the trimethylsilyl deriva- tive of the dehalogenation product with p-Br-BA. . . . . . l6 Thin-layer chromatography of p—[3HI-BA in three systems. . l9 Chromatography of BA, p-Br-BA and p-[BHJ-BA on Sephadex LH-20. . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Mass spectra of cytokinin analogues: o-F-BA, m-F-BA, p-F—BA, o-Cl-BA, m-Cl-BA, p-Cl-BA, o-Br-BA, N-methyl- m-Br-BA’ and m-Br-BA o o o o o o o o o o o o o o o o o o 0 27-35 Growth curve for cytokinin-dependent tobacco cells grown in liquid medium . . . . . . . . . . . . . . . . . . 39 Determination of the optimum auxin (2,4-D) and cyto- kinin (BA) concentrations for callus and for shoot formation in tobacco callus grown on solid medium. . . . . 43 Cytokinin activity of halogenated-BA derivatives in a tobacco suspension culture cell division bioassay. . . . . 47 Cytokinin activity of halogenated BA derivatives in the moss bud formation bioassay. . . . . . . . . . . . . . . . 49 Cytokinin activity of halogenated-BA derivatives in the bioassay measuring the induction of nitrate reductase in excised Agrostemma githago embryos. . . . . . . . . . . 51 Pictorial comparison of the effects of BA, o—Cl-BA, m-Cl-BA and p-Cl-BA on shoot formation in solid-grown tobacco callus . . . . . . . . . . . . . . . . . . . . . . 54 A "steric" model of the cytokinin binding site of a cytokinin receptor in tobacco cells. . . . . . . . . . . . 58 vii Figure Page 22 Scatchard plot of p-[BHJ-BA binding at different BA concentrations to show high-affinity and low-affinity cytokinin binding sites in an 80,000 x g particulate fraction from tobacco cells and their contrasting sensitivity to a heat pre-treatment (100°C, 15 min). . . . 76 23 Scatchard plot of p-[3HJ-BA binding to talc. . . . . . . . 81 24 Visually observed bands in linear sucrose density gradients after centrifugation of an homogenate from tobacco cells. . . . . . . . . . . . . . . . . . . . . . . 87 25 Fractionation of membrane marker enzymes and BA binding activity after centrifugation of a tobacco cell homogen- ate on linear sucrose density gradients. . . . . . . . . . 9O 26 Mass spectrum of 8-N3-BA . . . . . . . . . . . . . . . . . 102 27 UV spectrum of 8-N3—BA before and after various times of photolysis with a long-wavelength UV lamp (900 uW/Crnz) o o o o o o o o o o o o o o o o o o o o o o o o o o 106 28 Time course of 8-N3-BA photolysis comparing the effectiveness of long and short wavelength UV light, with and without a polystyrene petri dish cover as short wavelength filter. . . . . . . . . . . . . . . . . . 108 29 Comparison of the cytokinin activity of 8-N3-BA, BA and pre—photolyzed 8-N3-BA in the tobacco cell division bioassay (suspension culture), in the absence of actinic light . . . .-. . . . . . . . . . . . . . . . . 111 30 Comparison of cytokinin activity of 8-N3-BA, BA and pre-photolyzed 8-N3-BA in the moss bud formation bio- assay in the absence of actinic light. . . . . . . . . . . ll3 viii LIST OF ABBREVIATIONS N6-Benzyladenine 8—Azido-N6-benzy1adenine 2,4-Dichlorophenoxyacetic acid Equilibrium dissociation constant Total concentration of binding sites Thin-layer chromatography Gas-liquid chromatography Ultraviolet light Curie Standard error of the mean; calculated accord- ing to the equation: standard error = standard deviation/“n , where n = number of replicate samples ix GENERAL INTRODUCTION Despite much research and speculation, we still have no knowledge of the early (0-5 min) biochemical reactions which occur in a plant cell in response to the application of a hormone. One approach to this problem is to identify and localize the cell component(s) which binds the hormone and initiates the bio- chemical reactions leading to the physiological response. This cell component is commonly referred to as the hormone "receptor" and, in all cases known to date, has been found to be a protein the binding activity of which, like that of enzymes, is sensitive to denaturants which destroy protein conformation. For example, protein receptors have been isolated which bind to and mediate the effects of the following animal hormones and drugs: steroids (Jensen and DeSombre, 1972), peptide hormones (Cuatrecasas and Hollenberg, 1976), acetyl- choline (Hall, 1972), opium (Sharma et al., 1975), and LSD (Nathanson and Greengard, 1974). In bacterial systems, analogous effector- receptor interactions occur, for which functional binding proteins have been isolated. Such proteins include: plasma membrane-bound chemotactic receptors (Spudich and Koshland, 1975), a soluble CAMP receptor (Zubay et al., 1970), and the numerous transcriptional repressors (cf. Riggs and Bourgeois, 1968). In addition to the precedents mentioned above, a theoretical argument can be made as to why hormone and drug receptors must be proteins. It is generally accepted that only proteins are capable of 1 2 forming the wide diversity of specific and high-affinity binding sites which are required to explain these common characteristics of hormone and drug responses: (1) stereospecificity of hormones and drugs and (2) the very low concentrations at which the responses are elicited and also saturated. With regard to plant hormones, the above theoretical argument is equally valid and has been critically discussed (Kende and Gardner, 1976). Though no receptor protein for a plant hormone has yet been isolated (Kende and Gardner, 1976), pioneering work in this field by Hertel and his co-workers has resulted in the identifica- tion of a specific binding site for the auxin transport inhibitors, l-N-naphthylphthalamic acid (Thomson, 1972; Thomson et al., 1973) and morphactins (Thomson and Leopold, 1974), which appears to be localized in the plasmalemma. In addition, the interest which their studies on the specific binding of auxin to membrane fractions (Hertel et al., 1972: Hertel, 1974; Dohrmann, 1975) has generated, as evidenced by the increasing number of reports from other laboratories (Oostrom et al., 1975; Batt et al., 1976; Batt and Venis, 1976; Dollstadt et al., 1976; Kasamo and Yamaki, 1976) suggests that a breakthrough in this field may be near. The main object of the following study has been to detect and characterize in vitro the cytokinin binding protein(s) which may act in vivo as the receptor(s) for this class of plant hormones. All active cytokinins are adenine derivatives substituted at the N6 position with a hydrophobic side-group (see Leonard, l974,for the most recent review on cytokinin chemistry). Cis- and trans-N6- (4-hydroxy-3-methy1but-2-enyl)adenine (cis— and trans-zeatin) and N6-(3-methylbut-2-enyl)adenine (2iP) are the most abundant natural 3 cytokinins, while NG-furfuryladenine (kinetin) and N6-benzy1adenine (BA) are the most commonly used synthetic ones. These compounds were discovered on the basis of their remarkable stimulation of cell division in cultured plant cells (Miller et al., 1955). They are now known to effect many diverse physiological processes, including bud formation in mosses, shoot formation in tobacco callus, and senescence and directed transport of assimilates in higher plants (Kende, 1971). The discovery that cytokinins were located adjacent to the 3' end of the anticodon of certain tRNA's led to much speculation as to whether this was causally related to the biological activity of this plant hormone. Kende and Tavares (1968) were the first to provide evidence that this was not so, and their conclusion has since been confirmed by many others (this evidence is summarized by Kende and Gardner, 1976). Recently, Hecht, Skoog and their co-workers have described an exhaustive and systematic study comparing the biological activities of purine substituted and deaza analogues of cytokinins in a tobacco callus bioassay. The results of their experiments, in which the likely sites of metabolism of the purine ring of cytokinins have been modified to preclude such metabolism, has led to the conclusion that metabolic modification is not a prerequisite for cytokinin activity, "i.e., that the active form of the cytokinins can be the exogenous species themselves" (Hecht et al., 1975). These studies suggest, but do not prove, that a non-covalent rather than covalent interaction of the cytokinin molecule with some cell constituent(s) is required for the biological response. 4 Direct, but preliminary, evidence for the non—covalent binding of cytokinins to a receptor has been obtained by autoradiography using 14C-BA in moss protonemata (Brandes and Kende, 1968). In addition, results on cytokinin binding to plant ribosomes have been reported from two laboratories (Berridge et al., 1970; Fox and Erion, 1975): Though there was no effect of cytokinins on in vitro protein synthesis (Berridge et al., 1972), in vitro inhibition of a ribosomal protein kinase was observed (Ralph et al., 1972). This result was confirmed by Keates and Trewavas (1974), but was inter- preted by these authors as a non-physiological effect of high cyto- kinin concentrations due to competition with ATP at the substrate binding site. This conclusion has since been refuted (Ralph et al., 1976). Takegami and Yoshida (1975) have also recently described a soluble cytokinin binding protein from tobacco leaves. Unfortunately, it is not possible to correctly evaluate the physiological significance of the cytokinin binding observed in these studies since no adequate tests of specificity have been per- formed. As will be shown in the present study (Section III), there arenon-specificrsaturable binding sites for cytokinins, even in non- biological material such as talc, which superficially mimic specific binding sites. As was discussed in greater detail by Kende and Gardner (1976), in order to assign physiological significance to a binding site, it is essential that the specificity of binding be critically tested with analogues that are chemically very similar but have widely diverging degrees of biological activity. In the present study, my goal has been to differentiate between specific, high—affinity and non-specific, low—affinity cytokinin binding sites in homogenates of tobacco cells. The report of this 5 study is divided into four sections. In Section I, I describe the chemical synthesis of a high specific activity, tritiated cytokinin. This compound was deemed necessary for the detection of specific, high-affinity binding sites which are present at low concentrations in the plant cell. In Section II, the biological activities of a series of cytokinin analogues are given; these data are needed to relate the in vitro binding results to the in vivo physiological responses. In preliminary experiments with 14C-BA, a large amount of non- specific binding to the particulate fraction of tobacco cells was observed. Therefore, the emphasis for designing binding experiments with p-[3fll—BA (Section III) was placed on deciding with confidence whether a specific binding site was present in the particulate frac- tion of tobacco cells. In order to allow in future studies a more detailed characterization of the observed, specific cytokinin binding site, I have described in Section IV the synthesis and biological activity of a cytokinin photo—affinity reagent, 8-azido-benzyladenine. SECTION I THE SYNTHESIS OF A RADIOACTIVE CYTOKININ WITH HIGH SPECIFIC ACTIVITY Introduction Receptors for animal hormones (Cuatrecasas et al., 1974) and drugs (Goldstein, 1974; Farrow and Vunakis, 1972; Nathanson and Greengard, 1974) are present in tissues at low concentrations, and the binding of ligands to these receptors is more readily detected if a ligand of high specific radioactivity is available. This is especially true if one wishes to detect a small number of higher-affinity binding Sites in the presence of a large excess of lower-affinity sites. Ligands labelled with 14C cannot be obtained at sufficiently high specific activity, hence most studies employ tritiated or iodinated ligands. Previously published attempts to introduce a stable tritium label into biologically active cytokinins by the Wilzbach method or by cata- lytic exchange in an aqueous solution (Berridge et al., 1970; Letham and Young, 1971; Elliott and Murray, 1972; Walker et al., 1974) have failed to produce a specific radioactivity greater than 231 mCi/mmol. In this section, I describe a new and simple method for synthesizing a radioactive cytokinin, N6-benzyladenine[p-BH-benzyl] (p-BH-benzyl) with tritium in a stable, known position at near maximum specific radio- activity. The method employs catalytic dehalogenation (hydrogenation) of a halogenated precursor, p-bromo-N6—benzylaminopurine (p—Br-BA). 6 7 The procedure used to synthesize p-Br-BA was that of Elion et a1. (1952) as described by Okumura et a1. (1959). Though the synthe- sis of the 0-, m-, and p-Cl analogues of 6-benzylaminopurine (BA) had already been reported (Okumura et al., 1959), we chose to synthesize the previously unreported Br analogues since exchange with tritium is considered easier with this halogen. The p-position was chosen to reduce any possibilities of steric hindrance in the synthesis and dehalogenation reactions. A condensed report of this work has appeared (Sussman and Firn, 1976). Materials and Methods Synthesis of p-Br-BA 6-(Methylmercapto)purine (0.3 g) was added to p-bromo-benzylamine (1.25 g) and heated at 120-130°C in a closed tube for 18 hr with intermittent agitation. p-Br-BA was precipitated by the addition of methylene chloride and collected by filtration on Whatman GF/C filter paper, followed by washing with methylene chloride. After recrystal- lization from dimethylformamide, the product was collected by filtra- tion, washed with methylene chloride and dried in an oven at 80°C overnight. The yield was 28%. In a separate experiment, recrystal- lization from ethanol gave a similar yield. A preliminary Beilstein test of the recrystallized product indicated the presence of halogen, which was confirmed by obtaining a green color after heating a sample on a copper wire over a gas flame. The irregularly shaped crystals changed to needles before melting at 288-289°C. Elemental analysis found: C, 47.34; H, 3.38; N, 22.96; Br, 26.34%. Br requires: C, 47.39; H, 3.31; C12H10N5 8 N, 23.03; Br, 26.27%. Identity of the compound was further con- firmed by its UV, IR and mass spectra. The UV spectrum is indistinguishable from BA, with the following maxima at acid, neutral and basic pH: 1:2: EtOHnm(log 2): pH 2, 275.5 (4.26); pH 7, 269.5 (4.28); pH 12, 275.5 (4.28) with a shoulder at 280-285 nm. The IR spectrum (in KBr discs) is also similar to BA in showing (1) an unsubstituted purine N-9 position (typical weak absorption peaks at 2400-2800 cm-l), and (2) the absence of the strong primary amine N-H stretching absorption peak at 3380 cm-1 seen in p-bromo- benzylamine. Finally, the assigned structure was confirmed by mass spectrometry (Figure 1). The natural abundance of the two stable isotopes of Br, Br79 and Br81, in the ratio 100:97.5, permits facile identification of Br-containing fragments. The fragmentation pattern of p-Br-BA is similar to that of BA. The only peaks showing the presence of Br are the molecular ion, at m/e 303, 305 and fragments at m/e 169, 171 and 184, 186 identified respectively as the Br sub- stituted tropylium ion (C H Br+; analogous to m/e 91 in BA) and a 7 5 + species (CGH BrCH N H; m/e 106 in BA) formed by the loss of the 5 2 purine ring system. In addition, intense peaks are observed at m/e 89 and m/e 90 corresponding to the loss of HBr and Br, respec- tively, from the tropylium ion. All other major peaks were accounted for as arising from the purine ring. The purity of p-Br-BA was checked by TLC on silica gel, cellulose and aluminum oxide (Table 1). In all instances, only one compound was detected, using several visualization techniques. In particular, the absence of potentially reducible unreacted p—bromobenzylamine was verified by the absence of ninhydrin positive spots in the above TLC .uouoeouuoomm mmmz ooom Homo: man no momma >0 on us onoum uooufio we omcaouno mums ouuoomm .Aamlumumv ocacoomeuconoeounlmqu one Admv ocflcoomamucon mo muuoomm mmmz .H ousmflm 10 H shaman 00m 0mm OON Omfl OOH om . . I t . . . a BB 4 .._ .om TA“ mm. .225 .3 2.. r: a J 3.4mm_u...:..z::---l E s. 1:. 10¢ .......u....:-.mw-:.:sm a. a :RmImWVWIme a. .0m 1 z : 1AVMW Humane: to <95.— (3an Pi. OOH 00m 0mm OON omH OOH om a _.. . :om TA“ 02 «3...; i encufliunzwvuiiik. 10¢ IIIIITIIIWnI. :. .5 a N-----m@.m-1mk =. row a 1 i made: a __ 3. (a 0m .3. 02 N I.\' ll . 3 Table l. Rf values of compounds used in the syntheSis of p-[ Hl-BA in different thin layer chromatography systems TLC System Silicaa Aluminum? Compound gel Cellulose ' oxide BAl'2 0.36 0.64 0.72 p-Br-BA1'2'5 0.36 0.54 0.72 . 1,3 6-(Methylmercapto)purine 0.36 0.44 - p—Bromobenzylamine4 0.09 0.50 - Benzylamine4 0.07 0.80 - Visualization methods: 1quenching of fluorescence induced by excitation at 254 nm after spraying with 0.1% fluorescein in ethanol; 2purple fluorescence after excitation at 254 nm; 3yellow fluorescence after excitation at 254 nm; 4ninhydrin stain; 5silver chromate stain for purines (Whitfield, 1969) using commercial plastic-backed silica gel and cellulose TLC plates. Solvent systems: a10% methanol in methylene chloride; b0.01 N HCl; Ccellulose plates were pre-washed with 0.1 N HCl and dried just prior to use; dH20 saturated ethyl acetate. 12 systems. Absence of impurities was also demonstrated by GLC (Figure 2) on a 3% 83-30 column in which only a single peak was observed. As expected from comparison with the effect of bromine substitution on the retention time of benzylamine derivatives, p-Br-BA had a retention time 2-3 times that of BA. Dehalogenation reactionWith 1H2 and 3H2_ The dehalogenization reaction withJIEBgas was performed by us on the same small scale as the later reaction with 3H2 gas, in order to allow more direct comparison between the two reactions. This entire procedure was performed at room temperature. Ten percent palladium on carbon (3 mg) was added to dimethylformamide (2 ml) containing 10 ul triethylamine in a 5 ml reaction vessel containing a Viton stopper and a Teflon magnetic stirring bar. The vessel was evacuated, lHZ gas added and kept at a slight pressure (approximately 0.7 kg/cm2 above atmospheric pressure) for 30 min to saturate the Pd. The vial was then opened, p-Br-BA (10 mg) added and allowed to dissolve by stirring for 3 min. The vessel was closed, evacuated and 1H2 applied again at slight pressure for the time denoted. For measuring the time course of the dehalogenation reaction by GLC, 50 ul samples were withdrawn periodically, filtered to remove charcoal and silylated just prior to injection. The time course (Figure 2) showed that after 1 hr less than 10% of the original p~Br-BA was left, and the reaction was essentially completed after 2 hr. This concluSion was confirmed by the absence of a p-Br-BA zone in TLC on cellulose of the Z-hr reaction product. The identity of the product was verified by co-chromatography with authentic BA on cellulose TLC (Table l), GLC and by GLC-MS (Figure 3). 13 Figure 2. Time course of dehalogenation reaction of p-Br—BA with non-radioactive H2 gas, as measured by GLC. Samples were silylated in dimethylformamide by adding a 3- to 5-fold excess volume of N,0-bis-(trimethylsilyl)-trif1uoroacetamide containing 1% trimethylchlorosilane. GLC was performed using a 1.8 m x 1.83 mm glass column packed with 3% 83-30; column temperature, 205°C isothermal, injection port temperature, 225°C, flame detector temperature, 250°C, carrier gas He at 87.5 ml/min. 14 37. 93-30 momentum. 205 ° Figure 2 15 .Hmzaudmiumlm How cHE NH ocm Hmzelmm you CHE m mums oo>uomno moEHu coflucouou map was GHE\Uov um ooEEmumoum was ououmuomfiou cEdHoo one .GHE\HE on no mom uoaunmo on» sows Hovmlmm wm.m spas ooxomm menace mmoam ES mm.H x E m.a m 0cm >o on on HouoEouuoomm mmmz ooom Homo: mMA m moan: poEMOMHom mos szan .m ousmwm CH mm mums mcofluaocoo cowumaaaam .mm mo o>aum>fiuoo Ahaamamzuoefiuu owucosusm on» sums oo>ummno more on Hmowucoofi who cocflmuno Eduuoomm pom mafia cowucoumu use .dmuumlm sows uosooum cow» locomonnop onu mo o>flum>fluoo Haafimflmnuoefluu on“ no Ugo noumm Ezuuommm mmmz .m madman l6 2R 0mm com 030 0.04 nzuiw use .34 _ . 2 #4i I § --.A.....- 0.. an... so. .72 30.4 2...- I- I. ......u III»? sh = z z = I 2.6.: m onsmwm u\£ omfi OOH om Mo. 30. .r p 31713 52.0.... 20.53.. 3.3. its OOH N ii. 1 III | . 17 For dehalogenation with 3H gas, a sample (10 mg) of p—Br-BA 2 was sent to Mallinkrodt Chemical Works (Box 5439, St. Louis, MO 63160) with instructions to perform the reaction exactly like the reaction . . . . 3 . . . . . with non-radioactive H2 gas but With H at maXimum speCific actiVity 2 (59 Ci/mmol). Volatile 3H was removed from the radioactive product, and the sample was purified on preparative cellulose TLC as described for the non-radioactive reaction. Again, only a single fluorescent zone at the Rf of BA was observed. The sample was eluted from the cellulose with 50% ethanol and subsequently stored in this solvent at -80°C at a concentration of 3—30 x 109 dpm/ml. The chemical con- centration of this sample was measured by UV absorbance at 270 nm and by comparison with the absorbance of standard authentic lH-BA solu- tions. The specific radioactivity was calculated to be 10 Ci/mmol. The UV spectrum for the radioactive sample was found to be identical to BA at acid, neutral and basic pH. In addition, the radioactivity co-chromatographed with BA in 3 TLC systems and on Sephadex LH-20 column chromatography, as described below. Purity and stability of p-[3al—BA The radiochemical purity of this compound was shown to be at least 97% by chromatography using three TLC systems (Figure 4) and by column chromatography on Sephadex LH-20 (Figure 5). Re—chromatography on silica gel TLC raised this value to 99%. With some commercial silica gel plates (Merck pre-coated glass TLC plates), 5-10% of the radio- activity stayed at the origin. Since re—chromatography of the BA zone in the same system gave the same percentage of non-moving radioactivity, and since it was not observed on other silica gel TLC preparations (Brinkman Polygram Sil G pre-coated plastic sheets), it was considered 18 Figure 4. Thin—layer chromatography of p-[3H]-BA in three systems. 19 mm we: cunounooumv of p-[°H]-I5A CPM,°/o MAXIMUM SILICA Oil 15% mow In men, n “ugh-inc '3'“ H i—-—I IOO .. _ . . F“ Origin ‘ alufl 0L CELLULOSI 00' N MCI . - Quin, ~@‘ 'M93 35 IOO— _ II Origin 1 Solve»? o - _ L M AlUMINUM OXIDE ,6.» H20 Iot'd ethyl acetate I00- Origin solved OI. 20 IO onsnuc: (cm... Figure 4 [III] II, II. 'llll i IN I I .. 1 I! (illil 20 .vezasubcchCuzc can ASL: :3. x a. «mummmaum m0 unsoso HHoEm c £ua3 poxfimm mo3 _:n weaned» one .ooms mm: A>uo m mmv omimq xooonaom no urmdes vs: Eu em b:_:_er:ec csrficr nnc_: A.©.flv Eu m.m m .omuma xoponmom co Id6 on um ooawmuno muo3.ouuoomm Had .uouweouuoomm mmmz mumu Home: cmfluo> m .mmuumns new .amuumIEIHEuwsuoz .«muumno .amuaoum .amnaote .mmuauuo .dmumum .«mlmiE .mmlmlo "mosmonam sacwxouho mo muuoomm mmmz .va smacks» o newsman 27 . @ ousmam oxE 0mm 00m 00. ~bP~.P+——.—I-—.—._F_-—,_ \ : 1.: .. I... I j; If) g. 01 I 00. on mcEmuoincom -802“. no Mgsuaiul uol annolaa 0\E ODN CON 00— ._P_-___+kar—_P.p_b_.L-L M¢N +2 00— OD _. F —.. bLL _.h.._.‘blP P:P-_ w _- p F _ __ 1: ._....._ _.__ .. .__... ._._._ _._. 05:30:38 533... I E Mgsuagul uoI anyway 29 m musmfim a): OnN OON on. P — .P ‘m b —4~cmm u 202 u. nu ‘ ll. Ally i. .Ill‘ [ill III {I 30 m musmwm o): CON CON 4?. . :— T oo. mchUUENcmm .6320 no on Mgsuaw; uol amulet] 31 OH mugmflm a): 00m Cow CON .b._._k_bPP.b-—.r—PrqpbP—-P‘n—L :1 1 _—1 1 ‘ ‘ mom T +2 on. 00. on mEcmuoiNcmm no.2: o .. E Mgsumm uog annolaa 32 HH ousmfim o): mom +2 I .p- 00. on J i ___— Lr~b_‘.—‘p;- —-—,L.—p p—p— ...___-:..._.._._...____ 44—474.? 1.3. #3:. 4 d I Mgsuaw. u0| ammea oc_couo;~c¢m soSEOL NH musmflm 3E 0.0m OON . CON \bbbhb_.—q_?—fl—b—1ab—1¢Pp— mOm.mOm +2 LT .. On. 00. On \1 _ ééjgfip Mgsuam u0| anyway mcEmuoigom .0695 ..o 34 MH musmflm 0\E 00¢ 00m con 5. P u — — h L b p P b _ r P ~ — u k . P p _ P Ll. \ - mfififi .12 Dow OON P — n p b b p h _ L b —| b — Up; h - h- _I n. h! b P .i ._.. i . .1 . .1. . .._.—i 8. oo_ 8 1 _ . _ P p b — - _ P . . _ .,_- - \1-:-_...-._.: 1: flu:— mchngficmm ..oEEm IE u .285 u z Mgsuawl uol ammaa 35 «a musmflm 0}: 00¢ 00m 00m \F~.__b._p_L_P~._-hPHPL.—r_b_ T non.mom.. .ve‘ ODN CON . _ . _ .r-_ .. _ . _ . _ .. _ . _ . _ . _ . _ - _ _+;- . - \J J 441 u q u 1.1 a ‘ ‘ 1.— 14 1414 4 «fi_ 111—4 4—.— 1: on. 00. On ascmugicmm .0505 ..E Mgsuawn uoI ammaa 36 Table 2. Sephadex LH-20 chromatographic mobility of cytokinin analogues Compounda Relative Mobilityb'c'd BA 2.9 o-F-BA 3.3 m-F-BA 3.5 p-F—BA 3.5 o-Cl—BA 4.2 m-Cl—BA 4.7 p-Cl-BA 4.9 o-Br-BA 5.0 p-Br-BA 5.6 BA-riboside 2.1 8—N3-BA 4.1 aA list of the mobilities of additional cytokinin analogues is given by Armstrong et al. (1969). bA relative mobility of 1.0 is given by an elution volume equal to one bed volume. CThe excluded volume is at a relative mobility of 0.3. dColumn dimensions and solvent as in Figure 5. 37 Mass spectroscopic analysis proved most valuable in showing that a sample of "m-Br—BA" which had appeared homogeneous on Sephadex LH-20 chromatography was in fact mostly another compound which sublimed in the mass spectrometer at a slightly lower temperature than m-Br-BA and which gave an unexpected molecular ion at m/e 317, 319. On the basis of its mass spectrum (Figure 13), this compound was identified as NG-methyl-m-Br-BA. The formation of this N6—methyl derivative was found to be caused by contamination of the commercially obtained m-Br—benzylamine with N-methyl-m-Br—benzylamine (K and K Rare and Fine Chemicals, Plainview, NY). Because of this, the "m-Br-BA" deriva— tive was not further studied in the bioassays. Tobacco cell suspension: growth char- acteristics and bioassay procedure The cytokinin-dependent cell suspension used in this study was kindly supplied to us in June 1974 by Dr. C. Réaud—Lenoél (Centre Universitaire de Marseille—Luminy, Biochimie Fonctionnelle des Plantes, F-13288 Marseille, Cedex 2, France). It was strain #21 (Tandeau de Marsac and Jouanneau, 1972), an isolate originally obtained from tobacco pith (Nicotiana tabacum L.cv. Wisconsin 38). These cells were normally subcultured biweekly using a defined liquid medium (Jouanneau and Réaud-Lenoél, 1967) containing 2 x 10*7 M 2,4-D and 5 x 10-7 M kinetin. To obtain tissue routinely in sufficient quantity for in vitro binding studies (see Section III), 480 ml medium in 1-1iter Erlenmeyer flasks was inoculated with approximately 2‘ ml of a 2-week-old culture. A growth curve under these conditions (Figure 15) showed a lag period followed by an exponential phase. By using larger inocula, the lag period could be shortened without a change in doubling time or final fresh weight yield. 38 Figure 15. Growth curve for cytokinin-dependent tobacco cells grown jg) liquid medium. Results are expressed as the mean fresh or dry weight :_S.E. of the tissue from triplicate flasks. Dry weight was measured after the cells were dried at 60°C for 2-3 days. 39 mm“ 28: .8 3 Emma .55 4 3 - q .- u l5 l0 DAYS AFTER SUBCULTURE .mm. b b .7. o a -2 8: .8 3 tom; 18¢... Figure 15 40 For bioassay, 250 mu Erlenmeyer flasks containing 60 ml medium were used. An inoculum free of the exogenously supplied kinetin was commonly obtained by aseptically collecting the cells from a 2-week- old culture on a Miracloth filter. The cells were then rinsed and resuspended in cytokinin-free medium and aliquots used as inocula for the bioassay. Culture flasks were agitated on a circular shaker at 125 rpm, under normal room lights, at 28°C :_2°C. After 11 days (early stationary phase), the cells were collected by filtration on coarse fritted glass filters. Before weighing, the cells were rinsed with a solution of 40 mM sucrose, 2 mM Na -EDTA, 5 mM KCl, 2 adjusted with NaOH to pH 5.7 at room temperature,in order to solu- bilize and remove accumulated extracellular polysaccharides (A. Hanson, personal communication). Moss bioassay Cytokinin-induced bud formation in protonemata of the moss, Funaria hygrometrica (L.) Sibth. was measured as described by Whitaker and Kende (1974). Buds were counted 48 hr after transfer of protonemata to cytokinin-containing medium. Agrostemma bioassay Enhancement of nitrate reductase activity by cytokinins in excised embryos of Agrostemma githago was measured 5 hr after start of the hormone treatment. The procedure of Kende et a1. (1971) was followed. Tobacco callus differentiation bioassay Induction of shoot formation in tobacco callus was examined using callus which had been grown from pith (Nicotiana tabacum L.cv. 41 Wisconsin 38) isolated no longer than 9 months previously. This callus was subcultured every 6—7 weeks on standard Linsmaier-Skoog medium (Linsmaier and Skoog, 1965) with indoleacetic acid (10'.5 M) and kinetin (10-7 M). To determine the optimal auxin and cytokinin concentrations for the shoot formation assay, explants were placed on the same minimal medium but containing various concentrations of BA and 2,4-D. The results (Figure 16) showed that callus formation alone was maximal on medium with 5 x 10..7 to l x 10“6 M BA and 10-6 M 2,4—D but shoot formation was maximal on medium containing 5 x 10—6 to 1 x 10"5 M BA in the absence of 2,4-D. Thus, for the differentiation assay with cytokinin analogues, 3 callus explants (approximately 0.4 g each) were placed in each of five 125-ml Erlenmeyer flasks containing 50 ml medium with various concentrations of the test substance, but no auxin. Tissue was grown for 6 weeks in the dark, followed by 1 week under room lights. I attempted to quantitate shoot formation (Table 3) by two methods: (1) physically separating undifferentiated callus tissue from shoot and leaf tissue and weighing each separately and (2) measuring the chlorophyll concentration in the entire explant by extracting with a three-fold (ml/g F.W.) volume of 95% ethanol. The second method was feasible because only the shoots and leaves greened up appreciably. The two methods gave qualitatively similar results concerning the relative amounts of shoots and undifferentiated callus within each treatment. 42 Figure 16. Determination of the optimum auxin (2.4-0) and ibis (SA) concentrations for callus and for shoot forzation recto c llus grown on solid medium. Results are expressed the me‘n fresh weight of the tissue from quintruplicate flasks, ‘ ter a cu1ture period of 3-1/2 weeks. 43 Optimal callus alone Shogflggl'afion ”um” "I . 3 callus Md” mm" - II M 2 lllll"Ill Ill" Ill“). )1) I “I" I "In" "M“ \ ° ill"): IIIIIII "ii" 3 .. "" \ 4" HII'I/lml "I'm" .i""' 2 F (“i “priiwfi lwil..ll 'i‘ilii'...-"‘"" - . . 3 M W'fl‘llllw/ " '0 6 5x9; 6 I" ’l """"" 5xl0'7 B A ‘ w (M) 5XIO'6 ' :UI'IH '0- -T shoots clone '65 5x|0 8 2,4-0 (erratic) 5 z’ . (M) 5xIO' 0 Figure 16 44 Results thsical and chemical properties of ortho—, meta-, and para-halogenated BA derivatives The observed elution order of the cytokinin analogues in adsorp- tion chromatography on an alkylated Sephadex support, Sephadex LH—20, in 35% ethanol was (Table 2), BA > o-F-BA > m-F-BA > p-F-BA > o-Cl—BA > o-Cl-BA > m-Cl-BA > p-Cl-BA > o-Br—BA > p—Br-BA. This parallels their expected order of lipophilicity (i.e., para > meta > ortho, according to Kuraishi, 1959). There is apparently little separation of these BA derivatives on the basis of molecular size since this would cause an elution order of Br < Cl < F, which is exactly reverse of the observed one. Although the overall fragmentation pattern of the derivatives (Figures 1 and 6-14) was very similar to that obtained with BA (Figure 1), there was a systematic difference in the intensities of certain ions relative to each other. The most notable difference was an increase in the relative intensity of the following ions in the spectra of the ortho-substituted derivatives (X = halogen): (1) M+ - X; (2) C7H5X+, the substituted tropylium ion; (3) C6H4X+; and (4) M+ - X + 1. The increased intensitives of the first three ions can possibly be regarded as being due to an increase in lability of the respective bonds because of a 6-membered ring formed through 6 . hydrogen bonding of the ortho-substituted halogen to the N -amino hydrogen (Figure 21). 4S Biological properties of ortho-, meta- and para-halogenated BA derivatives In the tobacco cell suspension bioassay the relative activities of the halogenated BA derivatives were (Figure 17): o-F—BA = m-F-BA = p-F-BA = BA > o-Cl-BA Z o—Br-BA > m-Cl-BA 2 p-Cl-BA ; p-Br-BA (inactive). In the moss bioassay, the order of activity was (Figure 18): m—Cl—BA 2 o-F-BA = m-F-BA = p-F-BA = BA > p-Cl-BA 2 o-Cl-BA = o-Br-BA > p—Br-BA (inactive). In the Agrostemma bioassay, a limited number of analogues was tested, and their observed order of activity was (Figure 19): BA = m—Cl-BA > o-Cl-BA > p-Cl-BA. The other deriva- tives were less extensively tested in this bioassay, but in general, they were found to follow the same order as in the moss and tobacco assays (i.e., o- and p—F-derivatives completely active and p-Br-BA completely inactive). In an experiment (Table 3) designed to measure the activities of cytokinin analogues in inducing differentiation of shoots in tobacco callus, the results showed that BA and o-Cl-BA were both slightly more active than m-Cl-BA. However, a striking feature of this experiment was the ability of m-Cl-BA (at 5 x 10'.6 M and 10-5 M) to cause shoot formation in almost the complete absence of callus formation.' With this compound, the shoots appeared to arise from a brown, necrotic piece of callus (Figure 20). This result was consistently obtained in each of the three repeats of this experiment. A similar, but more erratic, response was obtained at the lower (10'.6 M) concentrations of BA and o-Cl-BA (Table 3). Discussion A result which was common to all three bioassays was that in any one position of the benzene ring, the biological activity of the :3 lb. erivatives d oioassa IV]. 3101']. t“re cell d v‘. C FRESH WEIGHT (9 per flask)tS.E. I0.0 - CI‘BA "o m\ p Br-BA 0\ C) lltln2iin 1 4L%t*i, 1 n KINETIN ZEATIN I0.0- . \CIs o".'05 .5 5 0'705 .5 5 6"?05 .5 5 CYTOKININ CONCENTRATION mm Figure 17 48 Figure 18. Cytokinin activity of halogenated BA derivatives in the moss bud formation bioassay. Results are expressed as the mean :_S.E.of three replicate samples. 49 BA F-BA 0.1 2,000r m l l" +l T g \” LIJ I .. CZ) ,000 l i '— O 0: 0. 2i; _____;v\}1 I 1 1 Q- I CI-BA m 2,000L c3 :3 m ”5 f“ 0: 1,000 - #114 LIJ p s i \ D i i i i0 z \o :5. 0w ¢ 5 5 ____L¥~il 1 L 21 11411 L 0 '01 l0 l0 IOO 0"‘0.I |.O I0 l00 CYTOKININ CONCENTRATION (11M) Figure 18 Figure 19. Cytokinin activity of halogenated-BA derivatives in the bioassay measuring the induction of nitrate reductase in excised Agrostemma githago embryos. Results are expressed as the mean :_S.E. of three replicate samples. 51 IO TIOITDITOI. I. BA Oh- ; mCl-BA oCl-BA pCI-BA ' . i0 CYTOKININ CONCENTRATION (#M) m. 0 o ficxoanQWmEtoe .Noz $62. 5 >._._>_._.o< mmqhoaomm mkqmtz Figure 19 52 .00c0num 0mm mo 06 m 009 00000» m0 0 0 00 mmo. H om0.o m0 Edmmoa :0 0>0o 0000000 0:0 0:000 0500000 .0.z.0 0x05 0 .0o00000 0000 000000 000 mo uo0uux0 000050000 00 mo mmod 0:» 000000006 >0 ©0u050000 003 «000m 00% c0000uucwocoo 00>£m0000£on .Amoonumz 000 000000002 0000 00000m 0050050 x003 h 0 00000 .m.m.H x000m 00m 050000 00 000003 ammnw c005 0:0 00 Ummmwumxm 000 0005000 .C0x00 0:0x000 850005 00000 0E om m00 I200ucoo 0x000m 000052000m 0E mm0 0>0m no 3000 00 00050050 0003 00000 0 0.00 0u000mx0 050000 000:90 I I 0.0.0 0.0 0.0.0 0.0 0.0 2 0I00 .00I00Ie 000. m 000.0 0.0 m 0.00 0.0 m 0.00 0.0 m 0.0 0.0 m 0.00 2 0I00 000010 000. + 000.0 0.0 + 0.00 0.0.0 0.00 0.0.0 0.0 0.0 + 0.00 z 0I00 x 0 .00I00Ie I I 0.0 + 0.0 0.0 + 0.0 0.0 2 0I00 .00I00Is 000. m 30.0 0.0 m 0.0 0.0 m 0.00 0.0 m 0.0 0.0 m 0.00 z 0I00 .00I00Io 000..H 000.0 0.0.H 0.0 0.0.H 0.00 0.0 0.0.0 0.0 H 0.00 z 0I00 x m .00I00Io 000. + 000.0 0.0 + 0.00 0.0 + 0.00 0.0 + 0.0 0.0 + 0.00 z 0I00 .00I00Io 000. m 00.0 0.0 m 0.0 0.0 m 0.00 0.0 m 0.0 0.0 m 0.00 z 0I00 .00 000..H 000.0 0.0.” 0.0 0.0.H 0.00 0.0.H 0.0 0.0.” 0.00 z 0I00 x 0 .00 000. + 000.0 0.0 + 0.0 0.0 + 0.00 0.0 + 0.0 0.0 + 0.00 z 0I00 .00 I I 0.0.H 0.0 0.0.H 0.0 0.0 000 . uo0uuxm Am\mv 00009 000000 muoonm ochQEOU o 0 mo mmod 0:000U\0000£m 0Amv x0000 00m u£m003 £0000 050000 0000300 :0 COHumfiuom 00030 :0 00:0000c0 00:0xoumo mo uummwm .m 00Q0B .0000000000000 mio0 00 nadomeoo 000» 0:» 0000000000 800008 :0 0x003 b you suzoum umumn x000m 000 no 00000000 0>0u0ucmmmumwn 000 030:0 0005000 039 .050000 0000000 nzoumic0000 :0 c00p0Euow 00050 no 00I00Im 000 00I00Is .00I00Io .00 00 0000000 000 00 0000000500 000000000 .00 000000 54 Figure 20 55 halogenated BA derivatives followed this order: F = H > C1 > Br. This relationship correlates with the order of their atomic size (H > F > C1 > Br) but not their electronegativity (F > C1 > Br > H). The major difference which was observed between the bioassays concerned the relative activities of the o- and m—Cl—BA derivatives. Thus, in the tobacco cell division assay, o—Cl-BA is active (but less so than BA) and m—Cl—BA is almost inactive. In the moss bud formation assay, however, the reverse is true; i.e., m-Cl—BA is fully active and o—Cl-BA is much less active. In the Agrostemma assay, m-Cl—BA is as.active as BA, as in the moss bud formation assay; how- ever, o—Cl—BA is also active, though significantly'less active than BA, exactly as in the tobacco cell division assay. Thus, apparently the specificity of the Agrostemma response‘can be construed as a summation of_the activities observed in the tobacco cell division and moss bud formation assays.- * In the discussion to follow, I assumed.as the simplest and most likely situation that the structural requirements for biological activity observed in these bioassays is solely a reflection of dif- ferences in the respective cytokinin binding.requirements at a receptor site,0and that metabolism and uptake do not contribute to the differences in biological activities of these analogues. There are no.data in any plant system to confirm or refute this assumption. In light of this, I have clearly labelled each of the following headings and conclusions as "speculative." Regardless of whether or not a receptor is the only protein involved, the reader should be aware that the differences in biological activities of cytokinins in different bioassays are most likely due to differences in a specific cytokinin binding site(s) of a protein(s) since it is highly unlikely that a S6 purely physical process, such as uptake through membrane diffusion, could be different in moss, as compared to tobacco. §peculations on the distribution in nature of cytokinin receptors and the topography of their cytokinin binding sites (i) Tobacco cell division receptor. Kuraishi (1959), using BA and kinetin analogues with substituted N6 side-groups, and the radish leaf expansion bioassay, interpreted differences in biological activity of cytokinin analogues on the basis of the lipophilicity of the N6 side—group. However, in order to explain a decrease in bio- logical activity when the lipophilic properties were increased (such as when there were 6 or more C atoms in a linear saturated hydro- carbon N6 side-group), he postulated that too high a lipophilic value is as bad as too low a value. However, using the R on Sephadex f LH-20 (Table 2) as a relative measure of lipophilicity, Kuraishi's hypothesis does not explain why o-Br-BA is more active than p-Cl-BA in the tobacco cell division assay (Figure 17). An alternative explanation of the results of the tobacco cell division bioassay is summarized in a "steric" model illustrated in Figure 21. According to this hypothesis, reduced activity with the bulky meta- and para-derivatives is caused by the inability of these derivatives to bind to a receptor, because the N6 side-group does not fit into a hydrophobic cleft of the protein. Ortho-substitutions are tolerated because substitution at this position would not be required to penetrate the cleft. The ortho—halogenated BA derivatives may also have a more favored conformation for biological activity because of hydrogen bonding between the halogen atom and the N6-amino hydrogen. Soriano-Garcia and Parthasarthy (1975) reported that the crystal 57 Figure 21. A "steric" model of the cytokinin binding site of a cytokinin receptor in tobacco cells. This model explains the results observed for the structure-activity relationships with halogenated-BA derivatives in a tobacco cell division bioassay. For simplicity, only the N6 side-group binding site is shown, though obviously the purine moiety is also bound since it too is required for activity (cf. Hecht et al., 1975). 58 I m /N. m P z .. /A\ . 0/ :0 0m 000000 .7} _ :20 0300096.»: £00.05 touamomm 59 structure of cytokinins such as BA is best described as: (1) two planes at approximately right angles to each other, one plane being defined as passing through the 10 atoms of the adenine moiety and the other plane passing through the benzyl group; (2) the N6-benzyl group is distal to the imidazole ring. Thus, the internal hydrogen bonding might allow a more favored conformation for biological activity by "cementing" the benzene ring into a plane more perpen- dicular to the purine moiety. A similar interpretation was also proposed by Dyson et al. (1972) to explain the 1,000-fold increase in activity of o-chlorophenylureidopurine over phenylureidopurine, observed in a soybean callus assay. (ii) Cytokinin receptor in moss protonemata. The high activity of m-Cl—BA coupled with the low activity of both 0- and p-Cl—BA indicates that binding to the receptor in moss protonemata involves the electron resonance properties of the Né benzyl side-group. The ability to accommodate a bulky chlorine group in the meta-position demonstrates the absence of steric restraints on binding, in contrast to the model for tobacco cells. Advantage can be taken of this observation in the design of active-site directed irreversible reagents for the moss receptor; likely candidates are the meta-substituted mercapto-, mercuro-w diazo-, or azido-BA derivatives. (iii) Are there separate cytokinin receptors for stimulating cell division and differentiation (shoot formation) in tobacco cells? As discussed above, the specificity observed in the Agrostemma assay could be explained by assuming that both a moss bud formation-like receptor and a tobacco cell division-like receptor were operating simultaneously in this tissue. 60 In the same context, I hypothesized that the tobacco cell division "receptor" and moss bud formation "receptor" were mediators of two fundamental, independent cytokinin reactions, one leading to a stimulation of cell division (tobacco cell division response) and one leading to differentiation (moss bud formation response). An attempt to test this hypothesis was made by comparing the differentiation- stimulating activity (shoot formation) of cytokinin analogues with their cell division-stimulating activity in the same tissue. At face value, the results indicate that the order of activities of the analogues is the same for both responses. However, the ability of m-Cl-BA to more consistently induce shoot formation in the complete absence of callus formation at all concentrations also suggests that the original hypothesis may be correct, but that the two responses were not adequately separated in the tobacco shoot formation assay. Such a difficulty is not unexpected since the process of shoot forma- tion(differentiation)Mull clearly require cell division and, therefore, a compound such as o-Cl-BA could have its differentiation-stimulating activity overestimated, by virtue of its strong cell division stimu- lating activity. This hypothesis is testable in a more definitive experiment which could be performed if a compound were found which was active in the tobacco cell division assay but inactive in the tobacco shoot forma- tion assay. The activity of cytokinin analogues in stimulating tobacco shoot formation could then be measured independently of cell division simply by testing cytokinin derivatives in the shoot forma- tion assay in the presence of this cell division-active, shoot formation-inactive compound. Thus, by pre-saturating the cell division response with this compound, the only additional effect 61 which would be noticed is the stimulation of shoot formation due to the differentiation response. If the two-receptor scheme I pro- posed is correct, potential candidates for this compound are meta- bolically stable, Ng-substituted derivatives since such compounds are active in tobacco and soybean cell division assays (Fox et al., 1973) but markedly inactive in the moss bud formation assay (Whitaker and Kende, 1974). SECTION III IDENTIFICATION OF A SPECIFIC CYTOKININ BINDING SITE IN A PARTICULATE FRACTION FROM TOBACCO CELLS Introduction As discussed earlier, one approach to the study of the mechanism of cytokinin action, which has not yet been successfully pursued, is the identification of the receptor(s) which binds the hormone. The results of Kende (Brandes and Kende, 1968),Skoog and co-workers (Hecht et al., 1975) suggest that this binding is non-covalent and should be amenable to identification in vitro. I thus undertook studies using radioactive benzyladenine to identify and localize in vitro the specific binding site(s) which may act as the cytokinin receptor(s). Because of current interest in membranes as sensory transducers and the availability of a suitable binding assay, we have focused our attention on membranous cell fractions. In addition, preliminary experiments with a soluble cytosol fraction using equilibrium dialysis and Sephadex G—25 chromatography were unsuccessful in detecting sig- nificant amounts of specific binding. The centrifugation binding assay as a technique for detecting particulate binding sites for plant hormones was originally pointed out by Hertel et a1. (1972) in their in vitro auxin and l-N—naphthylph- thalamic acid binding studies as the technique of choice because it posesses the unique characteristic of great sensitivity and low back- ground noise. It is also most useful for measuring under equilibrium 62 63 conditions, binding which is easily dissociated and in which the concentration of binding sites is very low. Thus, Hertel et al. (1972) were able to detect and quantitate specific binding which only represented 0.1-0.5% of the total radioactivity (i.e., bound/ free ligand = 0.001-0.005). Because of anticipated problems caused by large amounts of non-specific binding (Bezemer-Sybrandy and Veldstra, 1971), we have paid special attention to the structural specificity required for binding, as it compares to the structural specificity of the bio- logical response in the same tissue. This has been made possible by the use of the same cytokinin-dependent tobacco suspension culture for both binding assays and bioassays (see Section II). In my initial binding studies in 1972 I used l4C-BA of low spe- cific radioactivity since this was the most readily available radio- active cytokinin at that time. The preliminary results looked promising, showing the existence of pronase-sensitive binding of 14 6 to 10.5 M) and which was C-BA which was saturable (Kd N 10- specific for biologically active hormones (i.e., kinetin, BA and ZiP competed for binding, while adenine and the cytokinin ribosides did not). However, it was observed that (1) this binding increased almost two-fold after a heat pre-treatment (100°C, 15 min), (2) this binding could not be enriched in any fraction obtained by differen- tial centrifugation, (3) treatments which increased non-saturable binding (pre-heating,KCl) also increased saturable binding and treat- ments which lowered non-saturable binding (pronase, Triton X-100) also lowered saturable binding, and (4) biologically inactive analogues (e.g., p-Cl-BA and p-Br-BA) which were chemically similar to active 64 cytokinins were active in the binding assay. I therefore con- cluded that it was probably not the result of cytokinin binding to a specific cytokinin receptor protein, but instead was a non-specific and physiologically irrelevant retention of BA. Since it has been observed with animal hormones and with drugs that the physiologically important high-affinity binding of the ligand to a small number of receptor proteins may be undetectable because of the low specific activity of the radioactive ligand (Cuatrecasas et al., 1974) and because of an excess of lower-affinity, non-specific binding, I decided to test whether a similar situation existed with cytokinin binding in tobacco cells. Thus, a tritiated cytokinin, p-[3H]—BA, of high specific activity (SOO—fold higher than that of the 14 C—BA previously used) was synthesized (Section I). The following is a report of the results which were obtained using 3 this p—[ H]-BA to detect specific high-affinity cytokinin binding sites in cell-free particulate fractions of tobacco cells. Materials and Methods Tobacco cell culture Cytokinin-dependent tobacco cells were grown in suspension cul- ture and harvested at mid- to late exponential phase (9—10 days of culture), as described earlier (Section II). Control experiments verified that rinsing the cells to remove extracellular polysac- charides prior to homogenization did not affect their subsequent growth characteristics or cytokinin requirements. 65 Preparation of cell-freepparticulate fractions Cells were broken by homogenization in grinding medium (2 ml medium per g F.W. tissue) with a Super Dispax Tissumizer Model SD45 (Tekmar Co., P.O. Box 37202, Cincinnati, OH 45222) using a G454 generator at approximately 8,000 rpm for 30-45 sec. At this and all subsequent steps, solutions were maintained at 0-4°C. Intact cells were removed by filtration through 2 layers of Miracloth. An 80,000 x g particulate fraction was prepared by centrifuging the Miracloth filtrate at 30,000 rpm in a Beckman Preparative Ultra- centrifuge (Type 30 rotor) or at 32,000 rpm in an International Preparative Ultracentrifuge (Type A-l47 rotor). The pellets were resuspended in binding assay medium by two up and down strokes with a glass homogenizer and mechanically driven, tight-fitting Teflon pestle. For the differential centrifugation experiment (Table 8), the 3 particulate fractions were obtained by centrifugation of each lower—speed supernatant at 13,000 x g (Sorvall Centrifuge, 88-34 rotor, 10,500 rpm), 80,000 x g (as above) and 170,000 x g (Beckman Preparative Ultracentrifuge, TiSO rotor, 50,000 rpm, and International Preparative Ultracentrifuge, A-321 rotor, 55,000 rpm). The pellets obtained after each centrifugation were resuspended in binding assay medium as above. Cytokinin-binding assay Particulate fractions suspended in binding assay medium were incubated with radioactive BA plus or minus non-radioactive BA or cytokinin analogue for 1 hr at 0-4°C to reach equilibrium prior to centrifugation. Non-radioactive cytokinin analogues were added from 66 concentrated (up to 10..1 M) stock solutions in dimethylformamide. The total dimethylformamide concentration in the binding assays was kept at or below 0.1% (v/v). Dimethylformamide at these concentra- tions did not affect the binding assay results. Bound radioactivity was measured in an 80,000 x g resuspended pellet by centrifugation at 15,000 rpm (Beckman Preparative Ultra- centrifuge, Type 19 rotor) for 3 hr, using 4 m1 capacity polypropylene tubes containing 3.2 ml solution. Adapters made from Delrin were used to accommodate 72 tubes per centrifuge run (12 tubes per adapter, 6 adapters per rotor). To measure binding in a 170,000 x g particulate fraction (Table 8), the assay solutions (3.2 ml) were centrifuged at 50,000 rpm for 30 min in a Beckman Preparative Ultracentrifuge (SW-56 rotor) using 4 ml polyallomer tubes. Centrifuge tubes made of cellulose nitrate and Millipore filters were found unsuitable for binding assays because of a high background of BA binding to the tubes and filters. Bound radioactivity was determined by pouring off the super- natants, draining the tubes for 30 min, cutting off the bottom of the centrifuge tube containing the pellet and placing it into 10 m1 scintillation fluid (4-0 g/l PPO(2,5-diphenyloxazole), 0.1 g/l dimethyl-POP(1,4-bis[2-(4-methyl-5-phenyloxazolyl)]-benzene) in toluene, with 10% (w/v) BBS-3 Beckman Biosolv). The vials were agitated overnight on a horizontal shaker in the dark at room temperature. The next day, radioactivity was measured in a Packard Tri-Carb Scintillation Counter (Model 3375). This procedure resulted in the complete recovery of radioactivity bound in the pellets, as determined by comparison in one experiment with pellets in which the radioactivity was measured by combustion (Packard Tri-Carb Sample 67 Oxidizer Model 360). Counting efficiency was typically 30-35% for 4 H and 80-85% for 1 C and was monitored with each vial by automatic external standardization. Binding terminology "Saturable" binding is defined as the decrease in bound radio- activity caused by the addition of unlabelled cytokinin to the assay mixture. The assay tube with p-[BHI-BA or 8-[14CJ-BA without unlabelled cytokinin is called the "A" tube. "B" tubes are in all respects identical to "A" tubes except for containing an excess of non-radioactive cytokinin. Thus, "saturable binding" is calculated by subtracting the radioactivity found in the pellet of the "B" tube from that in the pellet of the "A" tube. "Non-saturable binding" is defined as the radioactivity in the "B" tube pellet. p-[3H]-BA binding which is reduced by the presence of low con- centrations (<5 x 10.7 M) of non-radioactive BA is called "high"- affinity saturable binding. This is in contrast to "low"-affinity saturable binding of p-[3H]—BA or 8-[14C]-BA, in which a reduction of the bound radioactivity requires the presence of still higher concentrations (>5 x 10-7 M) of non-radioactive BA. The terms "high"- and "low"-affinity binding are meant to dis- tinguish binding sites with different K 's, which usually reflect d differences in binding affinity. Protein and RNA determinations Protein was measured according to Lowry et al. (1951), and RNA by the Schmidt-Thannhauser procedure, according to Fleck and Munro (1962). 68 Enzyme assays To measure the activity of cytochrome c oxidase, a mitochondrial marker enzyme, sample aliquots (5-50 ul) were added to 0.4 ml reaction buffer (50 mM tris-acetate, pH 7.5) containing 0.1% Triton X-100. After 5 min at room temperature, the reaction was started by adding 50 pl from a stock solution of bovine heart cytochrome c (5.4 mg/ml in reaction buffer) which had been reduced with sodium dithionite (5 < ASSO/AS65 < 12). Enzyme activity was calculated from the initial rate of increase in A550 using 1 ml cuvettes (1 cm pathlength) in a Gilford Model 240 Spectrophotometer. The activities of glucan synthetase I (UDP—glucose-l,4-glucan- glucosyl transferase) and glucan synthetase II (UDP-glucose-l,3— glucanglucosyl transferase) were measured using a procedure developed by Ray, Hertel and their co-workers and described by Dohrmann (1975). According to these authors, glucan synthetase I activity in corn coleoptile homogenates is associated with a membrane fraction which bands at 30% sucrose (w/w) and which probably consists mainly of the Golgi apparatus, while glucan synthetase II activity peaks at 35—38% sucrose (w/w) and probably is associated with a plasma membrane frac- tion. The technique for determining glucan synthetase activities is based on measuring the incorporation of [14C1-UDPG into alcohol- insoluble material. For the experiment with tobacco cell homogenates (Figure 25), 50 ul sample aliquots were added to 50 ul reaction buffer (containing 60 mM tris-acetate, pH 8.0 with either 130 mM MgCl2 for glucan synthetase I assays or 2 mM non-radioactive UDPG for glucan synthetase II assays). The reaction was initiated by the addition of 20 ul [14C1-UDPG in H20 (90,000 cpm, 240 mCi/mmol). After 1 hr at 25°C, 50 ul of 0.1 M MgCl was added for glucan synthetase I assays and 150 2 69 ul of 0.1 M MgCl was added for glucan synthetase II assays, and the 2 reaction was terminated by heating at 100°C for 1 min. Approximately 5 mg of carrier protein (using a pre-boiled 13,000 x g particulate frac- tion) was added and the radioactive reaction product was precipitated with 2 ml 70% ethanol and washed 3 times by centrifugation (Sorvall Cen- trifuge, SS-34 rotor, 5,000 x g) and resuspension in 2 ml 70% ethanol. After the final rinse, the pellet was resuspended in 10 ml scintillation fluid and radioactivity determined as above. Media and radiochemicals The grinding medium contained 0.25 M sucrose, 50 mM tris, 1 mM Naz-EDTA, 0.1 mM MgC12, pH 7.9; the binding assay medium contained 0.25 M sucrose, 10 mM sodium citrate, 5 mM MgSO 0.5 mM MgC12, pH 4. 6.0; the gradient medium (Figure 25, la and Ib) contained 10 mM tris, 1 mM Naz-EDTA, 0.1 mM MgC12, 1 mM KCl, pH 7.0. The pH of all media was adjusted at room temperature with glacial acetic acid. One milli- molar potassium metabisulfite was added to the grinding medium the same day, or one day before use. 8—[14CJ-BA (24 mCi/mmol) and 8-[14CJ-adenine (54 mCi/mmol) were obtained from Amersham Radiochemical Centre. [14CI-UDPG (uniformly labelled, 240 mCi/mmol) was obtained from New England Nuclear. p-[3HI-BA (10 Ci/mmol) was prepared as described (Section I, and Sussman and Firn, 1976). Results Comparison of the binding of 8- [VIC] -BA and p- [TH] -BA . . 4 . . In a typical experiment in which 8-[1 C]-BA binding was measured in vitro to an 80,000 x g particulate fraction of tobacco cells 70 . 14 (Table 4, third column), of a total of 2457 dpm [ Cl-BA bound, 517 dpm (21%) was saturable. Thus, 1,940 dpm (79%) [14C]-BA was non- saturably bound, i.e., was bound in the pellet even in the presence of 10.4 M non-radioactive BA. Saturability at higher concentrations of non-radioactive BA could not be tested since its solubility limit in H 0 lies at 1-2 x 10‘4 M. 2 In a separate experiment, 8—[14CJ-adenine was used to estimate the amount of radioactivity which was simply trapped in the pellet because of the volume of free water. Radioactive adenine was chosen . 2 . . for this purpose because [1 C]-adenine did not compete for 8-[14CJ-BA . . . 2 . binding and neither [1 C]-adenine nor [IZCI-benzyladenine competed for 14 . . . . l4 . . 8—[ C]-adenine binding. USing 8-[ Cl-adenine it was found that only 27% of the non-saturably bound 8-[14C]-BA could be accounted for by the trapped water of the pellet. This value is an upper limit since it assumes that there is no binding of 8—[14C1-adenine by the pellet beyond that trapped in the free water. . . . 3 . . The binding properties of p-[ Hl—BA at high (10 Ci/mmol) and low (24 mCi/mmol) specific activities were compared in the same experiment with that obtained with 8-[14CJ-BA (26 mCi/mmol). The results (Table 4) show that identical values were obtained for l4Cl-BA were used at the saturable binding when p-[3HJ-BA and 8—[ same specific activity and concentration. Results with the two iso- topes were also identical when the total level of saturable binding was changed by tissue pre-treatment such as heat or 1 M KCl. However, the non-saturable binding was consistently higher when p-[3H1-BA was 4 used as compared to 8-[1 Cl—BA at the same specific activity and concentration. 71 AHHoHV mm H mmm AwmmHv mam H Hmw AwMMHV ma H mom.m Awomav an H omo.m om H m~¢.m mom H omm.m Ammomv mm H mmo.H Awmmmv «om H va.H iwmmav m H moo.m AHmMHV on H mov.o hm H mva.v mNH H 5mm.n mm H ham «0H H was pH H ovo.a mma H 646.4 mm H smv.m he H omH.m Awooav mHH H mma.H Awquv «Ha H me.s vm H mva.m Awmnv Hma H mmm.a Awmvav hmH H vom.n up H mmH.m mHH H mon.H Hm H mnv.m mm H ov~.a «Hum steam 4mm HUM S H CHE ma .Uoooa Houucou Aaouucoo mo my .m.m +| uwaamm mom wagon Emu usmfiumwuuumum suHcHHHm-aoH z HIOH z 6-0H x m HOEE\HUE mm m g o a m H van m sHHchmmuaoH Z E v 0 -OH -OH x m Hoss\Hos em I IQ mm _mmH sHHcHummuach z snoH x m, z m-oa x v Han\Ho oH mm: m um _ mi umHAm>ummno mcfivcan manmusuMm mo mm>9 "onsu gm: ca cofiumuucmocoo mm “wasp gm: cH coHumHucwucoo mm .mnsu =¢= :H »HH>HHUH UHHHommm "umomuu m>wuomoflomm mHHmo ouounou scum cofiuuoum mamasofluumm m x ooo.om cm mcwm: mwmmmm mCHUCHQ ca mmnfimmgnm paw dmumuvaanm mo acmfiummfiou .v magma 72 High- and low-affinity saturable BA binding Low-affinity saturable binding was observed to increase in pro- portion to the increase in non-saturable binding after pre-treatment (Table 4). Thus, when the non-saturable binding rose to 158% and 133% of control, because of a pre-treatment with heat or 1 M KCl, respectively, the low-affinity saturable binding was also observed to increase, to 209% and 161% of controls, respectively. Similarly, when the non-saturable binding of 8-[l4cl-BA decreased, such as after a pronase or Triton X-100 pre-treatment, the low-affinity saturable binding also decreased to approximately the same extent. In contrast to these results, high-affinity saturable binding, measurable only at the low concentrations possible with p-[3H1-BA (Table 4), was unchanged by a l M KCl pre-treatment and decreased to 78% of controls following heating. In the same experiment, the non- saturable binding observed in the presence of 2 x 10.7 M non-radioactive BA and 4 x 10.9 M p-[BHJ-BA was increased to a similar extent (142% and 134% for pre-heating and l M KCl, respectively) as that observed for the low-affinity saturable and non-saturable binding. In order to examine in greater detail the effects of heat on high- and low-affinity, saturable binding and to obtain estimates for the Kd's and Rt's (total number of binding sites), binding was measured over a range of BA concentrations. The raw data obtained (Table 5) confirm the previous results, in that the heat lability is a property only of the saturable binding which is observed when the "B" tube con- tains low concentrations of non-radioactive BA. With the "A" tube at 6.2 nM p—[3HI-BA and the concentration of non-radioactive BA in the "B" tube increasing from 17 nM to 862 nM, there is a progressive 73 . 2258 o: «9.5me gm m8 353:8 3333 Sam mo Nva + mun Hm + mmva mma + Oman mm .H mmmm Ed Now on H: H mom 3. H ms: 82 H R8 3 H mmmm 2: HQ mm 3 H m3 mm H «.8 mm H 23 mm H 3% 2: SN 8 S H NS .8 H mom 3 H mmmm 3 H 636 2a 02 am am H 63 i H Rm S H mmmm mm H «88 z: 3 03 mm H 87 «3 H mmm ma H 38 m: H ammo 2c 2 u n .. mm H 38 S H :2. . mHHanuumwn pmumwnuwum Houucoo omumwcumum Houucou cocoa mmnma mH nownz mcHUcHQ umHHmm uwmlsmo umaamm mom Emu coHumuucmucoo mandudumm mo w muH>Huum0Hpmm madam wanmusumm >HH>HuumOHomm pcdom Hmuoa mHHmu ouomnou Scum cowuumuu mumHsoHunmm m x ooo.om cm 0» mcprHn CHconuwu wuHcmemlamw£ .mHHanuummn mo mocmpcmmmp coHumuucwocou .m magma 74 decrease in the percent of the saturable binding which is heat-labile, from 100% to 66%. When the data for binding in the control samples are plotted according to Scatchard (1949), a non-linear relation is obtained (Figure 22) which can be resolved into two straight lines, repre- senting the equilibrium binding parameters for high- and low-affinity binding sites. In contrast, the Scatchard plot for binding in a pre-heated sample clearly demonstrates the absence of the high- affinity binding component. These data for the pre—heated sample were fitted to a single straight line by least squares analysis with a high correlation coefficient (r = -0.964). Another effect of the heat pre-treatment, which is evident from this plot, is a 10% increase in the number of low-affinity binding sites, with no effect on the K8. A graphical technique (Rosenthal, 1967) to obtain pre- cise values for the high—affinity binding parameters was found unsuitable, due to the fact that the high-affinity binding accounts for at most only about 10-15% of the total bound/free. Thus, the values shown in Figure 22 (Kd = 1.4 x 10.7 M, Rt = 8 x 10"9 mol/Kg fresh weight) were calculated by visually estimating a straight line parallel to the steepest part of the high-affinity portion of the control binding curve. Because of interference by the large excess of low-affinity binding sites, the Kd and Rt values for the high- affinity site are probably upper limits and should only be considered as estimates of the true values. Structure-activity requirements of in vitrgJ3H-BA binding (i) 80.000 xeg,particulate fraction from tobacco cells. The . 3 ability of the cytokinin analogues to compete with p-[ Hl-BA at the 75 .ACHE ma .Uoooav ucwEummuunwum umwn m ou huH>HuHmcmm mCHummuucoo ufiwau can maamu ooomnou Bonn coHuomum mamasoHuumm m x ooo.om cm CH mouHm mchcHn CHConuho huHcfimuMJkoH paw huHcHumanmwa 305m ou mcoHumuucwocoo mm ucmumMMHv um mcprHn mmnnmmalm no uon phonoumom .Nm mudem 76 mm mHson .2. _Hozaomg ab. ow on ON 0_ .3mox\»o_oe.-o.xo¢¢..m \. .-o. x t. . 3. £55. -33 .._Omkzoo .3... 3. 228. .636 2 .m womr 1 .3... fuses .6.ch .m .2 ob. :N «ax om._. o-F~BA = p-F-BA = kinetin > m—F-BA > o~Cl-BA > o-Br-BA > trans-zeatin > cis-zeatin > m-Cl-BA > p-Cl-BA > p-Br-BA. This relative order of binding activity closely parallels that observed for the biological activity of these analogues in the same tissue (Figure 17). When the ability of the cytokinin analogues to compete with p-[3Hl-BA at the low-affinity binding site was measured, all of the analogues showed approximately equal activity as BA, except for o-F-BA and p-F—BA, which exhibited slightly higher activity (Table 7). (ii) Tale, p-[3HJ-BA was found to bind saturably to talc powder (Mallinkrodt, U.S.P.) in a manner superficially resembling that which we observed with a particulate fraction from tobacco cells. In a Scatchard plot (Figure 23), p-IBHl-BA binding to talc gave a non-linear curve which was visually resolved into two straight lines 4.5 x 10'8 M, R 1.3 x 10'8 mol/Kg representing high-affinity (Kd t 2.3 x 10"6 M, R 2.4 x 10"7 mol/Kg t dry talc) and low-affinity (Kd dry talc) binding sites. The low-affinity saturable binding was found to be non—specific in its structural requirements since all of the cytokinin analogues tested competed approximately equally well (Table 8). In contrast, the analogues showed differing abilities to compete for the high-affinity saturable binding and their observed order was: p-Br-BA = o-Br—BA = o-Cl-BA > m-Cl-BA > p-Cl-BA > p-F-BA = BA > o-F-BA = m-F-BA. This is totally unlike the specificity obtained in any bioassay and biological binding test. 78 Table 6. Analogue specificity of high-affinity saturable cytokinin binding in an 80,000 x g particulate fraction from tobacco cells a Saturably bound radioactivity Compound tested cpm per pellet :_S.E. BA 318 i_27 o—F—BA 279 :_20 m-F-BA 249 :_26 p-F-BA 271 :_27 o-Cl-BA 227 :_25 m-Cl-BA 118 :_34 p-Cl-BA 72 :_28 o-Br-BA 200 :_32 p—Br-BA 34 :_24 Kinetin 284 :_27 cis-Zeatin 134 :_22 trans-Zeatin 166 i_l6 aAll tubes contained 5 nM p-[3H]-BA (10 Ci/mmol in "A" tubes). The non-radioactive compounds tested above were added to "B" tubes at 500 nM. 79 Table 7. Analogue specificity of low-affinity saturable cytokinin binding in an 80,000 x g particulate fraction from tobacco cells a Saturably bound radioactivity Compound tested cpm per pellet :_S.E. BA 1053 :_125 o-F-BA 1401 :_104 m-F-BA 964 :_ 21 p-F-BA 1411 :_ 72 o-Cl-BA 942 i. 65 m-Cl-BA 1187 :_187 p-Cl-BA 1047 :_130 o-Br—BA 1248 :_159 p-Br-BA 969 + 40 aAll tubes contained 5 nM p-[3H]-BA plus 1500 nM 1H-BA (final specific activity of 400 mCi/mmol). The non-radioactive compounds tested above were added to "B" tubes at 10 nM. 80 Figure 23. Scatchard plot of p-[3H]-BA binding to talc. 81 mm mstHm :2. 3230.... m.o. on ON 0. In, (a) 22 to 3.9.. to. x 3 u .m o 20.0. x 3 «av. b.c.t<-33 o 0.2 .5 8.8. ob. x n.» «.1 2 ob. x m... «ox 5.5.3-6.... mum“. 02300 adj .Hootxch—zv 044% 0... 02525 HuomoHUMHIcoc 2: com .mHHmo ooomnou mo uzmwmz nmmum m N.N ou popcommmuuoo uwaamm 36mm .mmnsu ..mmnsu =<= 0H Hoea\Ho 0H. «mI_mm.Im :0 m.m cmchucoo mmamemm Hflmm Q =m: 0» compo n00.H 00H.H 00 H 0MH I I NH.0 H09 2 H I I I 0 x 000.0: + I I . CHE . mmv + 0am 000.” 000 am .I 00 0H 0 . ma 0.00H I 0 H 000.00 00m + 000 me + 00H 00H.0 0H0.0 0H.0 HoHHcoo 0nm.H v~0.~ mmm.H mnm.a I I «0.0 Hum 2 H 50H.H 000 00 .H mHs 00 .H 0N0 I I 00.0 :Hs 0H .0.00H 0 x 000.00 .I .I . ,I 0 x 000.MH 00H + 0»0.H 00H + mmm.H 000.0 000.0 «0.0 HoHucoo 00H.H 000.H 00H.H mmm.H I I NH.H H09 2 H 00H.H N00 00H.H 0mm 0HH.H mom I I 0H.H 0H5 0H .0.00H 0 x 000.mHI0 H0H.H 000.H hHm.H NH0.~ H00.0 mma.0 mm.H HoHHcou chuoum mm. chuomm.mE umaamm cHwHOHQImE umaawm umaamm ucwfiummquwum COHuomum \500 \500 n \500 \05 n \05 n \05 mHHanIuwmn 42m chuoum mmcHuaHm manmusumm suHcHHHHIamHm coHummDMHuucmo waucmummmwp ha pmchuno mcoHuomum mudeoHunmm Hamo ooomnou msoHum> CH ucmucoo Ckuoum can 42m .muH>Huom mcHocHQICHconu>u mo cemHHmmEou .m wanna 85 determine whether specific binding was associated with a particular membrane or organelle (Figure 24). Membrane marker enzymes were used to locate the Golgi apparatus (peak of glucan synthetase I activity at 35% (w/w) sucrose), a plasma membrane fraction (peak of glucan synthetase II activity at approxi- mately 39-40% (w/w) sucrose), and mitochondria (peak of cytochrome c oxidase at 41-42% (w/w) sucrose). Visual examination of the centrifuge tubes following sucrose gradient centrifugation (Figure 24) consistently showed 4 distinct bands. Since bands 2 and 4 were the only green bands and band 4 was more prominent when less harsh grinding conditions were used (i.e., grinding by hand with a porcelain mortar and pestle instead of the Super Dispax Tissumizer), these fractions were identified as con- taining stripped (Band 2) and intact (Band 4) chloroplasts (Miflin and Beevers, 1974). A marker enzyme for smooth endoplasmic reticulum was not measured, but this membrane was tentatively assigned to band 1 since it is the only prominent lower density visual band and a similar highly turbid zone which is considered to be smooth endoplasmic reticulum is found in gradients of corn homogenates (R. Hertel, personal communication). Band 3 was the most highly concentrated, intense visual band and was identified as containing mitochondria because it coincided with the peak of cytochrome c oxidase activity. Because the enzyme assays required only low quantities of sample, many fractions could be collected and tested for activity, allowing finer resolution of overlapping peaks (Figure 24, II). In contrast, resolution was greatly sacrificed in the fractions obtained for the cytokinin binding assay (Figure 24, Ia and Ib). In order to 86 Figure 24. Visually observed bands in linear sucrose density gradients after centrifugation of an homogenate from tobacco cells. BAND 1: BAND 2: BAND 3: BAND 4: 87 ............................ ............................... ................................ ................................ ................................ ........................... DI FFUSEJWHI TE ___ :::::.:::::::::::::::::::::::::: ................................ -------------------------------- DENSEJGREEN DENSEJWHITE DIFFUSEJGREEN Figure 24 SAMPLE LAYER (IS-AS Z (w/w) SUCROSE GRADIENT 88 obtain sufficient cell material for the 12 binding assay tubes which were needed for each fraction (3 "A" tubes and 3 "B" tubes for a control sample and 3 "A" tubes and 3 "B" tubes for a pre-heated sample), it was necessary to (l) collect a smaller number (six) of fractions with larger volumes from each gradient and (2) pool the fractions collected from six gradients. The problems of processing sufficient biological material is further aggravated by the background of heat-stable, low-affinity cytokinin binding. These experiments were originally attempted to improve the signal-to-noise ratio based on the possibility that high-affinity binding might reside exclusively in a membrane fraction readily separable from the bulk of all others. Unfortunately, as detailed below, this did not prove to be the case. Despite the difficulties, the following limited conclusions may be derived from the data. The level of non-saturable binding closely followed the protein content of each fraction (Figure 25, Ib) and was observed, as in all other experiments, to increase following a heat pre-treatment. High-affinity, saturable binding was also most' prominent in the fractions containing the highest protein content. In this particular experiment (Figure 24, Ia), fractions #4 and #5 were the only ones containing heat-labile, high-affinity saturable binding sites. 0f two separate repeats of this experiment, one yielded similar results and the other showed heat-labile, high-affinity saturable binding only in fractions #5 and #6. Thus, fractions #4 and #6 showed variable amounts of heat-labile, high-affinity saturable binding, while fraction #5 most consistently exhibited this activity. 89 Figure 25. Fractionation of membrane marker enzymes and BA binding activity after centrifugation of a tobacco cell homogenate on linear sucrose density gradients. The results in Ia and Ib were obtained with aliquots from the same pooled gradient fractions in one experiment. The results in II were obtained with a gradient in a separate experiment. All conditions in II were the same as in Ia and lb (see Materials and Methods for media) except that the grinding medium contained 50 mM MES (2(N-morpholino)ethane sulfonic acid)-NaOH, pH 6.0 instead of 50 mM tris-NaOH, pH 7.9, and 0.1 mM MgC12 was deleted. The gradient medium contained 10 mM MES-NaOH, pH 6.0 instead of 10 mM tris-NaOH, pH 7.0, and the 0.1 mM MgC12 and 1 mM KCl were deleted. In all experiments, 17 ml of a Miracloth filtrate obtained from a tobacco cell homogenate was layered onto 21.5 ml of a linear 15-45% (w/w) sucrose gradient and centrifuged in a Beckman Preparative Ultracentrifuge (SW-27 rotor, 27,000 rpm) for 3 hr at 95,000 x g. Fractions were collected by tube puncture and gravity-feed, from the bottom. In Ia, o———o is sucrose concentration [% w/w]; A—-—A and A—-——i represent high-affinity saturable BA binding in a control and pre-heated (100°C, 15 min) aliquot of the fraction, respectively. In Ib, o-oooo is protein; A-——A and A---A represent non-saturable BA binding (5 nM p-[3HI-BA in "A" and "B" tubes, 100 nM non— radioactive BA in "B" tubes) in control and pre-heated samples, respectively. In II, o--o is % sucrose concentration [% w/w], o---o is glucan synthetase I activity, x——fix is glucan synthetase II activity and A---A is cytochrome c oxidase activity. 3H-BA BOUND 0PM / pellet x l0" 3: S.E. RELATIVE ENZYME ACTIVITY 90 3.0 I5.0 I0.0 5.0 0.0 I.O 0.5 Figure 25 I SATURABLE BINDING ° (HI-Affinity) ' ‘ 45 Control )- d 3 O f ‘ I 5 I ‘\ ,"KPro-Matod .. \\i’ .- o Ib NONSATURABLE BINDING (LOW- Affinity) ‘ «0.3 " ‘ 0.2 " ‘ O.I EM?) GOLGI. PM. MITO. II .LCIlLORJ, ,l, t A 45 . I 0” °. . I I x. 30 I ' .' 5 b l 5’ .‘ : I I" 1 j ,4' 2* I5 9. I 4' ‘ = -0- ' ‘0' ‘ 'd’ h‘ ”was! 4; ---- ". 4°" 4 I 2 3 4 5 6 FRACTION PROTEIN 96 SUCROSE (VI/w) (mg/pellet/froction) °/o SUCROSE (IN/w) 91 Discussion Low-affinity binding as measured with 8-[14c1-BA and -[5H]—BA Binding assays with benzyladenine labelled with 3H or 14C at the same low specific activity gave identical results for saturable binding, but not for non-saturable binding. In all treatments, there was approximately twice as much p-[BHJ-BA non-saturably bound than 8-[14C1-BA. There are two possible explanations for this result: (1) the higher level of p—[3B1-BA binding was due to a radioactive impurity in the sample or (2) the difference was caused by an isotope effect. The p—[3H]-BA sample used had been purified by chromatography on both cellulose (Figure 4) and Sephadex LH-20 (Figure 5), and re-chromatography in these and other systems (see Section I) showed a radiochemical purity of 99%. If there was a radiochemical impurity in the p—[BHJ-BA sample, it must either be chromatographically very similar to BA (and have a similar UV absorption spectrum) or else it must be present at very low radioactive concentrations. Since this extra binding of p-[BHJ-BA was totally non-saturable (i.e., it.occurred even in the presence of 10-4 M BA), it did not interfere with the measurement or interpretation of saturable cyto- kinin binding. In other words, its only effect was to increase (double) the background ”noise" without affecting the signal. Specific versus non-specificBA binding Only the high-affinity saturable p-[3fil-BA binding which I observed in a particulate fraction from tobacco cells fulfills the most important criterion which characterizes "specific" binding; 92 that is, the structural requirements for binding were identical to those observed in the biological response. The conclusion that this binding was specific is supported by the contrasting lack of speci- ficity for low-affinity binding with this fraction from tobacco cells and also by the observation that genuinely artifactual binding to a non-biological material (talc) had a structural specificity totally unlike the specificity of the biological binding and bioassay results. The observation that only the high-affinity, saturable binding was heat-denaturable is additional evidence that only this binding site resides in a specific protein. In contrast, a heat pre-treatment caused a 10% increase in the number of low-affinity saturable binding sites. A similar observation was reported by Puca et al. (1971) for the non-specific binding of estrogen by bovine plasma albumin. These workers interpreted this increase in non-specific binding as due to a loss of protein conforma- tion, with a resultant increase in protein surface and, thus, in non- specific binding sites. This explanation may also hold for the heat- induced increase in low-affinity saturable binding which we observed in fractions from tobacco cell. Localization of specific cytokinin binding site In the assay for specific cytokinin binding, the signal was low, as was the signal—to-noise ratio. For.this reason, it was obviously difficult to determine precisely with which cell structure this binding site was associated. In spite of these difficulties, I conclude from results which were obtained consistently that it is associated with a structure which is present in all fractions obtained by differential centrifugation, but is enriched in a 13,000—80,000 x g 93 particulate fraction. This structure also bands at a density cor- responding to 35-42% (w/w) sucrose under conditions at which equilibrium sedimentation of cell membranes and organelles is obtained (Hodges and Leonard, 1973). The results on insensitivity of specific binding to 1 M KCl indicate the high-affinity binding site which I have detected in a particulate fraction from tobacco cells is unlike that obtained by Fox and Brion (1975) from ribosomes of wheat germ. My differential centrifugation studies further suggest that this binding site is not associated with ribosomes; however, since the differences in specific binding activity between the particulate fractions were barely significant, these conclusions remain tentative. Finally, because the high-affinity binding site was found to be associated with a structure which banded in equilibrium sucrose gradients in the vicinity of the bulk of the cell organelles and membranes, it was not possible to determine which of these structures were responsible for the binding activity. Where do we go from here? At this point, it is clear that a more conclusive localization of the high-affinity, heat-labile cytokinin binding site will require an increase in the signal-to-noise ratio generated in the binding assay. This may be obtained by including 10-5 M p-Br-BA in all of the binding assay mixtures since this compound was observed to compete well for the low-affinity, non-specific binding site but not at all for the high-affinity binding site in particulate fractions from tobacco cells. A further increase in the signal-to-noise ratio may be obtained by increasing at least lO-fold the amount of biological 94 material over that used in this study. The tobacco cells which I have used are easily grown to high densities, and there would be no problem in obtaining sufficient cells. However, for sucrose density gradient studies with larger amounts of material, it would be neces- sary to use a zonal rotor (e.g., the Sorvall sZ-14). Another possible approach is to attempt to increase the signal and improve the signal-to-noise ratio by optimizing the extraction and assay conditions. This was not attempted in the present study beyond the initial, preliminary experiments. This approach was used successfully by Dohrman (1975) to extend the initial studies on an auxin binding site from corn coleoptiles (Hertel et al., 1972) and to more definitively characterize its localization, pH sensitivity, etc. As an alternative approach, I synthesized a cytokinin derivative which has the capacity to covalently combine with its receptor. My work using this technique, known commonly as photoaffinity labelling, is described in the next section. SECTION IV CHEMICAL SYNTHESIS AND BIOLOGICAL ACTIVITY OF B-AZIDO‘BENZYLADENINE, A POTENTIAL CYTOKININ PHOTOAFFINITY REAGENT Introduction Carbene and nitrene generating, light-sensitive derivatives of hormones and substrates have proven useful to analyze in animal tissues hormone receptors and active sites of enzymes (see review by Knowles, 1972). Recently, Haley and co-workers described the use of radioactive 8-azido-substituted purine derivatives to covalently label ATP and cAMP binding proteins, including ATPases and a protein kinase (Haley and Hoffman, 1974; Haley, 1975; Pomerantz et al., 1975; Malkinson et al., 1975). A striking feature of these and other photo- affinity studies (Guthrow et al., 1973; Hanstein and Hotefi, 1974; Maassen and Moller, 1974) has been the ability to use the radioactive, light-sensitive derivatives with an impure, unfractionated cell-free extract to radioactively label only those proteins which have a spe- cific binding site for the particular ligand. I have undertaken the synthesis of a radioactive cytokinin photo- affinity derivative (8-N3-BA) in the hope that, in combination with SDS gel electrophoresis, it will permit identification of a specific cytokinin-binding protein(s), as has been achieved in analogous studies with 8-azido-ATP and 8—azido—cAMP using extracts of animal tissues. In addition to the obvious advantages that a light-generated covalent bond between a hormone and its receptor has for analysis of 95 96 hormone binding in vitro, it may also be possible to perform the photolysis and covalent-bond formation in the intact tissue prior to cell rupture, as, for example, has been suggested to occur during the in vitro analysis of auxin-stimulated enzyme activity in the plasma membrane of coleoptiles (Cleland, 1975). In this section, I describe the chemical synthesis and photo- lytic properties of non-radioactive 8-azido-benzyladenine, and also results of biological experiments which were conducted prefatory to its use as a radioactive compound. Materials and Methods Synthesis of non-radioactive 8-N,-BA Attempts to brominate BA directly in the 8-position, according to procedures used to obtain 8-bromo-cAMP (Muneyama et al., 1971) and 8-bromo-adenosine (Holmes and Robins, 1964) yielded only 2-bromo- benzyladenine, as identified by UV and NMR spectroscopy and mass spectrometry. 2-Bromo-BA did not undergo displacement reaction to form the azido derivative when heated with NaN3 in dimethylformamide under conditions used to obtain 8-azido-cAMP (Muneyama et al., 1971), 8-azido-adenosine (Holmes and Robins, 1965) and 8-azido—AMP (Haley and Hoffman, 1974). We thus chose an alternative route for the prepa- ration of 8-N3-BA, based on the synthesis of N6-substituted adenine derivatives from adenosine, according to the alkylation and rearrange- ment procedure described by Leonard and Fuji (1964): 97 8-BrAdenosine (I) NaN3 V 8-N3Adenosine (LE) benzyl bromide V l-benz 1-8-N Adenosine (III) 3 0.2N NaOH N6Benz 1-8-N3Adenosine (I!) 0.5N HCl N6Benzy1-8-N Adenine (Z) 3 8-Azido-adenosine (15) was synthesized from 8-bromo-adenosine (I) (Aldrich Chemical Co., Milwaukee, WI, USA) according to Holmes and Robins (1965). To 165 mg NaN (3 mmol) in 50 ml dimethylformamide 3 692 mg of I_(2 mmol) was added and the solution heated at 75°C. The progress of the reaction was monitored by measuring the UV absorption spectrum of the product (Table 10). After 15 hr, when the reaction was completed, the solvent was evaporated in vacuo at 80°C and the product, I}, was precipitated from the viscous residue by addition of 50 m1 methylene chloride. The precipitate was washed with 2 ml distilled H 0, dried in vacuo at 2 50°C, dissolved in 6 ml dimethylformamide, and 0.43 g benzyl bromide (2.5 mmol) were added to it. The reaction mixture was stirred and heated at 37°C, and the formation of III_was followed by TLC (Table 11). An additional 0.29 g benzyl bromide (1.7 mmol) was added after 24 hr. By 48 hr, the reaction was complete and the solution was added to 40 ml 0.2 N NaOH. A brown, immiscible oily residue (pre- sumably unreacted benzyl bromide) which formed immediately was removed by centrifugation. The supernatant was then heated at 85°C 98 Table 10. UV spectra of intermediates and related compounds used in the synthesis of 8-N3-BAa 0.1 N HCl Neutralng 0.1 N NaOH Compound 1 A . A A . l l . max min max min max min 8-Br-Adenosine (I) 262 231 264 231 265 237 213 8-N3-Adenosine (II) 281 b 244 281 248 280 249 206sh 221 l—Benzy1-8-N3— 282 248 282 255 282 255 Adenosine (III) 2125b 3255b 3255b N6-Benzyl-8-N3- 290 250 289 255 289 254 Adenosine (I!) N6-Benzyl—8-N3— 298 250 290 254 293 256 Adenine (y) 2308b 221sh Adenine 262 227 260 224 267 238 207 277sh Adenosine 257 230 259 226 259 234 1-Benzyladenosine 258 235 259 232 259 237 267sh 267sh N6-Benzyladenosine 264 234 269 231 268 236 (BA-riboside) 208 N6-Benzyladenine (BA) 276 236 270 231 276 242 210 2825b aA11 spectra were obtained in 50% ethanol. bsh denotes a shoulder. 99 Table 11. Thin layer chromatography on silica gel of intermediates and related compounds used in the synthesis of 8—N3-BA Compound Rfa 8-Br-Adenosine (I) 0.44 8-N3-Adenosine (II) 0.44 l-Benzyl-B-NB-Adenosine (g3) ' o . 31 N6-Benzy1-8-N3-Adenosine (IV) 0.71 (0.38) N6-Benzyl-8—N3-Adenine (y) - (0.69) Adenosine 0.30 Adenine 0.32 8-Br-Adenine 0.44 N6-Benzyladenine (BA) 0.61 N6-Benzyladenosine (BA-riboside) 0.59 1-Benzyladenosine 0.10 aDeveloping solvent was 16% methanol in methylene chloride, except for the values in parentheses which represent the Rf in 8% methanol in methylene chloride. 100 to effect rearrangement to £2, The reaction was monitored by TLC and found to be complete within 45 min. Estimated yield (by UV spectroscopy) was 350 mg (0.88 mmol) of 1!, Cleavage of the ribose to give the 8-azido-substituted free base (y) was achieved by the addition of 6 N HCl to the above, cooled reaction mixture, to give a final concentration of 0.5 N HCl. After 45 min at 80°C, the reac- tion was complete, as judged by TLC. The pH was adjusted to 4.5 with 1 N NaOH and the product, 2, was separated from the reaction mixture by extraction with ethyl acetate. Forty-two milligrams (0.12 mmol) of chromatographically pure V_was recovered following recrystalliza- tion from ethanol. The major loss in overall yield (6% of theoretical final yield from I) was thus in the last, acid hydrolysis, step. The identity of the final compound (2) as 8-N3-BA was confirmed by its mass spectrum (Figure 26) and by its photo~labile properties (see below). Cytokinin bioassays For measuring cytokinin activity, two bioassays were used: growth of a cytokinin-dependent tobacco cell suspension culture and bud formation in 2-week-old moss protonemata. Both assays have been described in Section II. The reader will notice that the maximum fresh weight per flask attainable with BA (10 g) in the more recent dose-response experiment with the tobacco cells (Figure 29) is approximately two-thirds that shown (15 g) in Figure 17. The experiments of Figure 29 were per- formed in the summer and fall of 1976, and those of Figure 17 during spring and summer of 1975. In the period between, it was necessary to discard the liquid cell cultures due to contamination and to start 101 .dml m zIw mo 55.30QO mum: .0m mHsmHm 102 0N stmHm ®\E Com OON P L . . - b _ p p b . _ bu _ . _ pr. ..r._._+III.—1_.—+_41..-d|u|+c1II-Lfi O \ x. z I .— _— _4 ./ I ZH—MHZIA/ _ Z I-Im._l IIIIIII II [I' on ......... IIIZI: NZI . _mw N 0 I0 ®®N mmN I +2 NMN 00_ 00:20:34 00. oo. 00 02.0.01 41.1.1. I -._.. . .. .s 0 . Q0 I On. _®_ _m OO— 103 a new suspension culture from a stock of agar-grown callus of the same strain #21 which had been continuously subcultured for just such an evantuality. No other aspect (growth curve, analogue speci- ficity) of the bioassays had changed. It is interesting that with the "newer" liquid cells the higher yield of cells (15 9) could be obtained at high 8-N3-BA concentrations, while the maximum obtainable yield on BA had been lowered. Photolysis of 8-Na-BA The UV light sources employed in this study were common, hand- held lamps used to detect UV absorbing or fluorescing zones on paper and thin-layer chromatograms: UVS-ll Mineralight for short wave- lengths and UVL-21 Blak-Ray for long wavelengths (lamps were obtained from Ultra-Violet Products, San Gabriel, CA, USA). Vessels containing light-sensitive azido derivatives were wrapped in aluminum foil to protect the photoaffinity reagents from exposure to room light. Ultra- violet light intensities were measured with a Kettering Radiant Power Meter (Laboratory Data Control Division, Milton Roy Company, P.O. Box 10235, Riviera Beach, FL 33404). Spectroscopy, Ultraviolet spectra were recorded on a Carey Model 15 Spectro- photometer and the low resolution mass spectrum was obtained by direct probe using a Varian (Bremen, Germany) Mass Spectrometer Model 104 Results Chemical properties and photolysis of 8-Na-BA J— As judged by increased solubility in ethanol and decreased solu- bility in H O, 8-N -BA was found to be more lipophilic than BA. 2 3 . . -5 Saturation in water was reached at 5 x 10 M 8-N3-BA, as compared to 1-2 x 10-4 M for BA. The photolysis reaction of 8-N3-BA could be conveniently moni- tored by a change in the UV spectrum (Figure 26). This spectral change (in 50% ethanol) was accompanied by a disappearance of the 8-N3-BA zone at R 0.69 in silica gel TLC (Table 10) and the appearance f of UV-absorbing compounds at the origin and at Rf 0.08 and also smaller amounts at Rf's 0.04 and 0.89. The photolysis products were not further characterized. The relative effectiveness of various lamps was determined by measuring the absorbance of unreacted 8-N3-BA at 305 nm. The photolysis of 8-N3—BA followed first order kinetics (Figure 28). The short- wavelength UV lamp gave a photolysis rate (tl/2 = 1.1 min) twice that of the long-wavelength UV lamp (tl/2 = 2.2 min). However, for experi- ments involving photolysis of 8-N -BA in intact tissues, the long- 3 wavelength UV lamp may be more suited. It can also be seen in Figure 28 that a polystyrene petri dish cover reduces the effectiveness of the long-wavelength UV by only 1/3. If it is necessary to maintain sterile conditions during photolysis, the petri dish cover can there- fore be kept on and the photolysis time can be increased accordingly. Biological activity of 8-N,-BA In the absence of actinic light, 8—N3-BA is as active as BA in eliciting bud formation in moss protonemata (Figure 30) and 105 Figure 27. UV spectrum of 8—N3-BA before and after various times of photolysis with a long—wavelength UV lamp (900 uW/cmz). 2 0* Optical Density 200° 106 OIMn. Oufin. ‘I I I“. \\ O. o ‘ 1 ID 0 O N ‘0 O N N '0 Wave Length (nm) Figure 27 107 Figure 28. Time course of 8-N3-BA photolysis comparing the effectiveness of long (0) and short (A) wavelength UV light, with (---) and without (-——) a polystyrene petri dish cover as short wavelength filter. The light intensities 0.5 cm from the lamp surface were: short wavelength, 280 uW/cm2 (90 pW/cm2 with petri dish cover), long wavelength, 900 uW/cm2 (630 uW/cm2 with petri dish cover). 108 I00 ‘ "95‘ ‘H:1.~ \~\ \\ “ ‘ \ ‘k s “ \ \ Ox‘ ‘~~ \ \ \ \ \ \ 50- ‘~ “~ \\ \~‘ \ \\O ‘s‘ % ‘ \ \ ‘4 00305 \ \ \ Remaining ‘. 25 ‘ '0 l l l 2 4 6 Time(min) Figure 28 109 significantly more active than BA in the tobacco cell suspension bioassay (Figure 29). In addition, at concentrations greater than 10-6 M, where BA consistently and strongly inhibits growth of tobacco cells, 8-N3-BA does not. Thus, it appears that 8-N3-BA has all of the growth-promoting but none of the growth-inhibitory properties of BA in this liquid tobacco cell suspension bioassay. When a high concentration of 8-N3-BA (2.5 x 10.6 M) was mixed with a supra-optimal concentration of BA (5 x 10-6 M), the yield per flask was 1.3 :_0.3 g, as compared to 16.5 i 0.9 g and 16.8 :_0.9 9 when the 8-N -BA (2.5 3 x 10.6 M) was used alone or with S x 10-7 M BA, respectively. Thus, cells grown in the presence of high 8-N3-BA concentrations have not lost their sensitivity to supra-optimal BA concentrations. The inhibitory response at high BA concentrations is less striking and more variable in the moss bioassay. When a sample of 8—N3-BA is exposed to short-wavelength UV irradiation (sufficient to obtain 299% photolysis, as checked by TLC) prior to bioassay, the cytokinin activity is reduced to ca. 1/100 in the moss bioassay (Figure 30) and ca. 1/50 in the tobacco callus bioassay (Figure 29). Discussion By a four-step reaction sequence from 8-bromo-adenosine, I was able to synthesize a highly active, light-sensitive cytokinin analogue. Our finding that an 8-azido substitution does not reduce cytokinin activity of BA is in accordance with the known lack of effect of halogen substitution at the 8-position of the purine ring (Dammann and Leonard, 1974) and the chemical similarity of the azido group and the halogens (Boyer and Cantor, 1954). Similarly, the increased 110 .mxmnam wumaHHmsuonnw ummma um Baum mHHma an» no .m.m.H muanmz zmauw cme an» mm ommmmumxa mum muasmmm .SMmmnoHn on» :H mm: on uoHum AwnmmumoumEOHna Mahna :Hnu an camps“ may 4mIszw mchHmEmu Haw ou omHMHpmnHH some can meEnm mmIszm panhaouonmnmum one .uanH aHaHuaa mo oucmmnm we» aH .Amusuaau concwmmamv anmmMOHa :onH>Hp HHmu oaawnou may cH AOIIIOV «mIszm panaaouonm Iwum can A1IIIHH>Huam CHconuha mo camHummSoo .om musmHm 113 0. on stmHm A Evcozoxcoocoo 33:32.an3 vxxx «mnzolwss \ \ fl\ (011v «mnzm 000.. l 1 000.0 +0005 IuaId Jed spng Io quumN 114 activity of 8-N3-BA over BA in the liquid tobacco-cell suspension bioassay is consistent with the higher activity of 8-methy1-2iP as compared to 21F in the tobacco callus bioassay (Dammann and Leonard, 1974). Like a methyl group, the azido group is considered a strong electron donor in an aromatic system (Smith et al., 1962). One interpretation of the lack of inhibitory activity at high concentrations of 8-N3—BA as compared to BA (Figure 29) is that there are two receptor sites for BA, one with a low Kd the occupancy of which regulates growth stimulation, and one with a higher Kd which regulates growth inhibition. The lack of inhibition observed with 8-N3-BA could then be explained by a lack of affinity of 8-N3-BA for the receptor with the high Kd. The inhibition of growth at high, probably non-physiological concentrations (5 x 10"6 M) of BA may be caused by competition of the cytokinin with substrates of purine- utilizing enzymes. If this is true, there may be more than one "receptor—site" with higher K . Alternatively, 8-N d -BA may be more 3 readily metabolized, and its actual concentration may therefore never reach toxic levels. The low biological activity we observed with pre-photolyzed 8-N3-BA in the moss bioassay can probably be accounted for by the presence of remaining, non-photolyzed 8-N ~BA. Thus, in this system, 3 the products of 8-N3-BA photolysis are very likely inactive. The results with tobacco, however, are not as clear cut and indicate that the product(s) of photolysis may retain some biological activity in this system. Since 8—N3-BA is a fully active cytokinin and has photolytic properties similar to those observed by others with 8-N3-cAMP and 8-N3-ATP, its use as a radioactive reagent for a covalent labelling 115 of cytokinin binding sites is justified. A suitable route to syn- thesizing tritiated 8-N3-BA at a high specific activity is available, starting with commercially obtainable 2-3H-adenosine (Amersham Radiochemical Centre). In preliminary experiments, this compound has been successfully brominated to give 2-[3H]-8-bromoadenosine in high yield. GENERAL DISCUSSION Contrary to my earliest expectations, not all saturable cyto- kinin binding is specific binding. This proved to be a much more difficult problem with cytokinins than was reported for analogous auxin binding studies (Hertel et al., 1973). However, in binding studies using radioactive cytokinin of high specific activity (p-[3H]-BA, 10 Ci/mmol) and at low concentrations (10'.9 to 10.8 M) a high-affinity binding site could be detected in a particulate fraction of tobacco cells. On the basis of its heat lability and its structure-activity properties, which are very similar to those observed in a bioassay using the same cell line, the observed binding site is a likely candidate for the cytokinin receptor. In a typical experiment, the specifically (high-affinity site) bound radioactivity represents only 0.2 to 0.5% of the total radio- activity, similar to the levels reported in the first study on specific auxin binding sites (Hertel et al., 1973). Ordinarily, this might be sufficient for localization of the binding site, but due to the large amounts of non-specific, low-affinity binding found with cytokinins, the results are not yet sufficient to resolve a location for specific binding site in the cell. I observed a distinct difference in the structure-activity requirements in two cytokinin bioassays (moss bud formation and tobacco cell division) which was interpreted on the basis of differ- ences in binding sites of the respective receptors. Though other 116 117 interpretations are possible, this one can be directly tested by binding experiments using moss homogenates in a fashion similar to those which have been carried out with tobacco. However, it should be noted that a 75-90% reduction in the number of high-affinity binding sites observed in the particulate fraction of tobacco cells (see Figure 22) would probably have made their detection impossible because of the high background of low-affinity, non-specific binding. This problem may be insurmountable in the moss system since only a fraction of the cells in any one protonema are responsive to cytokinin. For a more detailed characterization of the specific, high- affinity cytokinin binding site in tobacco cells, I synthesized a biologically active cytokinin photoaffinity reagent, 8-N3-BA. This approach is expected to reduce the problems arising from non-specific cytokinin binding because light-generated covalent binding via nitrene insertion to a specific, higher-affinity binding sites has been reported to have a higher efficiency than that to lower affinity, non-specific binding sites (see Hixson and Hixson, 1973, for discussion of this puint). In addition, this technique is amenable to enrichment of the specific binding signal by the use of gel electrophoresis under denaturing conditions (SDS), once the covalent bond is generated. Thus, if a gel is divided into 100 slices, each containing one theoretical non-specific binding site, the signal-to-noise ratio for detecting specific binding to a protein in one slice is effectively increased by a factor of 100. Ultimately, proof must be obtained that a specific binding site is or is not involved in the physiological response. As was pointed out by Kende and Gardner (1976), this can only be achieved by two 118 approaches: (1) examining tissues with a single point genetic lesion affecting the receptor, and (2) characterization of in vitro bio- chemical reactions which are also observed in the early in vivo response of the tissue to the hormone. REFERENCES REFERENCES Armstrong, D. J., Burrows, W. J., Evans, P. K., Skoog, F. S.: Iso- lation of cytokinins from tRNA. Biochem. Biophys. Res. Commun. 31, 451-456 (1969). Batt, S., Wilkins, M., Venis, M. A.: Auxin binding to corn coleoptile membranes: kinetics and specificity. Planta (Berl.) 130, 7-13 (1976). Batt, S., Venis, M. A.: Separation and localization of two classes of auxin binding sites in corn coleoptile membranes. Planta (Berl.) 130, 15-21 (1976). Berridge, M. V., Ralph, R. K., Letham, D. S.: The binding of kinetin to plant ribosomes. Biochem. J. 119, 75-84 (1970). Berridge, M. V., Ralph, R. K., Letham, D. 5.: On the significance of cytokinin binding to plant ribosomes. In: Plant Growth Substances, 1970, pp. 248-255. Ed.: Carr, D. J., Berlin: Springer-Verlag 1972. Bezemer-Sybrandy, S. M., Veldstra, H.: Investigations on cytokinins. IV. The metabolism of BA in Lemna minor. Physiol. Plant 25, 1-7 (1971). Boyer, J. H., Cantor, F. C.: Alkyl and aryl azides. Chem. Rev. 56, 1—57 (1954). Brandes, H., Kende, H.: Studies on cytokinin-controlled bud forma- tion in moss protonemata. Plant Physiol. 43, 827-837 (1968). Cleland, R. E.: Auxin-induced hydrogen ion excretion: correlation with growth, and control by external pH and water stress. Planta (Berl.) 127, 233-242 (1975). Cuatrecasas, P., Tell, C. P. E., Sica, V., Parikh, 1., Chang, K.: Noradrenalin binding and the search for catecholamine receptors. Nature 247, 92—97 (1974). Cuatrecasas, P., Hollenberg, M. 0.: Binding of insulin and other hormones to non-receptor materials: saturability, specificity and apparent "negative cooperativity." Biochem. Biophys. Res. Commun. 62, 31-41 (1975a). 119 120 Cuatrecasas, P., Hollenberg, M. 0., Chang, K., Bennett, V.: Hormone receptor complexes and their modulation of membrane function. In: Recent Progress in Hormone Research, Vol. 31, pp. 37-94. New York: Academic Press 1975b. Cuatrecasas, P., Hollenberg, M. D.: Membrane receptors and hormone action. Adv. Protein Chem. 59, 251-451 (1976). Dammann, L. 6., Leonard, N. E.: Cytokinins: synthesis of 2-, 8-, and 2,8-substituted 6-(3-methyl-2-butenyl)aminopurines and their relative activities in promoting cell growth. Phytochem. 13, 329-336 (1974) . — Davies, E., Larkins, B. A.: Polyribosomes from peas. V. An attempt to characterize the total free and membrane-bound polysomal popu— lation. Plant Physiol. 55, 749-756 (1975). Dohrmann, U.: Charackterisierung der in-vitro-bindung des pflanzen— hormons auxin bei koleoptilen von Zea mays L. Inaugural- Dissertation Albert-Ludwigs-Universitat, Freiburg i. Br. 85pp (1975). Dollstadt, R., Hirschberg, K., Winkler, E., Hubner, G.: Bindung von indolylessigsaure und phenoxyessigsaure an fraktionen aus epi- kotylen und wurzeln von Pisum sativum L. Planta (Berl.) 559, 105-111 (1976). Dyson, W. H., Hall, R. H., Hang, C. I., Dutta, S. P., Cheda, G. B.: Cytokinin activity of a group of ureidopurine derivatives related to the tRNA component, N-(purin-6-ylcarbamoyl)threonine. Can. J. Bot. 59, 237-243 (1972). Elion, G. B., Burgi, E., Hitchings, G. H.: Studies on condensed pyrimidine systems. IX. The synthesis of some 6-substituted purines. J. Am. Chem. Soc. 12, 411-414 (1952). Elliott, D. C., Murray, A. W.: A quantitative limit for cytokinin incorporation into tRNA by soya-bean callus tissue. Biochem. J. 130, 1157-1160 (1972). Evans, E. A., Sheppard, H. C., Turner, J. C.: Validity of tritium tracers. Stability of tritium atoms in purines, pyrimidines, nucleosides and nucleotides. J. Labelled Cmpds. 5, 76-87 (1970). Farrow, J. T., Vunakis, H. V.: Binding of d-lysergic acid diethyl- amide to subcellular fractions from rat brain. Nature 237, 164- 166 (1972). Fleck, A., Munro, H. N.: The precision of UV absorption measurements in the Schmidt-Thannhauser procedure for nucleic acid estimation. Biochim. Biophys. Acta 55, 571-583 (1962). Fox, J. E., Brian, J. L.: A cytokinin binding protein from higher plant ribosomes. Biochem. Biophys. Res. Commun. 53, 694-700 (1975). 121 Fox, J. E., Sood, C. K., McChesney, J. 0.: Effect of substituents at the 9-position on cytokinin activity. Phytochem. l2: 1531-1533 (1973). Goldstein, A.: Opiate receptors, a minireview. Life Sci. 52, 615- 623 (1974). Guthrow, C. E., Rasmussen, H., Brunswick, D. J., Cooperman, B. 8.: Specific photo-affinity labeling of the cAMP receptor in intact ghosts from human erythrocytes. Proc. Nat. Acad. Sci. (Wash.) 19, 3344-3346 (1973). Haley, B. E.: Photo-affinity labeling of cAMP binding sites of human red cell membranes. Biochemistry l2! 3852-3857 (1975). Haley, B. E., Hoffman, J. F.: Interactions of a photo-affinity ATP analogue with cation stimulated ATPases of human red cell membranes. Proc. Nat. Acad. Sci. (Wash.) 25, 3367-3371 (1974). Hall, Z. W.: Release of neurotransmitters and their interactions with receptors. Ann. Rev. Biochem. 35, 925-952 (1972). Hanstein, W., Hatefi, Y.: Characterization and localization of mito- chondrial uncoupler binding sites with an uncoupler capable of photo—affinity labeling. J. Biol. Chem. 249, 1356-1362 (1974). Hecht, S. M., Frye, R. B., Werner, D., Hawrelack, S. D., Skoog, F., Schmitz, R. Y.: On the "activation" of cytokinins. J. Biol. Chem. 250, 7343-7351 (1975). Hertel, R., St. Thomson, R., Russo, V. E. A.: In-vitro auxin binding to particulate cell fractions from corn coleoptiles. Planta (Berl.) 2, 325-340 (1972). Hertel, R.: Auxin transport and in vitro auxin binding. In: Membrane Transport in Plants, pp. 457-461. Ed.: Zimmerman, M., Dainty, J., Heidelberg: Springer 1974. Hixson, S. S., Hixson, S. H.: Photochemical labeling of yeast alcohol dehydrogenase with an azide analog of NAD+. Photochem. Photobiol. 55, 135-138 (1973). Hodges, T. R., Leonard, R. T.: Purification of an ion stimulated ATPase from plant roots. In: Methods in Enzymology, Vol. 32, pp. 392-406. Ed.: Colowick, S. P., Kaplan, N. 0., New York: Academic Press 1973. Holmes, R. E., Robins, R. K.: Purine nucleosides. VII. Direct bromination of adenosine, deoxyadenosine, guanosine, and related purine nucleosides. J. Amer. Chem. Soc. 55, 1242-1245 (1964). Holmes, R. E., Robins, R. K.: Purine nucleosides. IX. The synthesis of 9-8-D—ribofuranosyl uric acid and other related 8-substituted purine ribonucleosides. J. Amer. Chem. Soc. 52, 1772-1776 (1965). 122 Jensen, E. V., DeSombre, E. R.: Mechanism of action of the female sex hormones. Ann. Rev. Biochem. 25, 203-230 (1972). Jouanneau, J. P., Réaud-Lenoél, C.: Croissance et synthese des proteines de suspensions cellulaire de Tabac sensible a la kinetine. Physiol. Plant 55, 834-850 (1967). Keates, R. A. B., Trewavas, A. J.: Protein kinase activity associated with isolated ribosomes from peas and Lemna. Plant Physiol. 55, 95-99 (1974). Kasamo, R., Yamaki, T.: In Vitro binding of IAA to plasma membrane- rich fractions containing Mg+2-activated ATPase from mung bean hypocotyls. Plant Cell Phys. 52, 149-164 (1976). Kende, H.: The cytokinins. Int. Rev. Cytology 55, 301-337 (1971). Kende, H., Tavares, J.: On the significance of cytokinin incorpora- tion into RNA. Plant Physiol. 55, 1244-1248 (1968). Kende, H., Hahn, H., Kays, S. E.: Enhancement of nitrate reductase by benzyladenine in Agrostemma githago. Plant Physiol. 55, 702- 706 (1971). Kende, H., Gardner, 0.: Hormone binding in plants. Ann. Rev. Plant Physiol. 52, 267-290 (1976). Knowles, J. R.: Photogenerated reagents for biological receptor- site labeling. Accounts Chem. Res. 5, 155-160 (1972). Kuraishi, 8.: Effect of kinetin analogs on leaf growth. Sci. Papers College Gen. Educ., Univ. Tokyo 5, 67-104 (1959). Leonard, N. J., Fuji, T.: The synthesis of compounds possessing kinetin activity. The use of a blocking group at the 9-position of adenine for the synthesis of l-substituted adenines. Proc. Nat. Acad. Sci. (Wash.) 5;, 73-75 (1964). Leonard, N. J.: Chemistry of the cytokinins. In: The Chemistry and Biochemistry of Plant Hormones, Vol. 7, pp. 21-56. Ed.: Runeckles, V. C., Sondheimer, E., Walton, D. C., New York: Academic Press 1974. Letham, D. 8.: Regulators of cell division in plant tissues. V. A comparison of the activities of zeatin and other cytokinins in five bioassays. Planta (Berl.) 22, 228-242 (1967). Letham, D. 8., Young, H.: The synthesis of radioisotopically labelled zeatin. Phytochem. 15, 2077-2081 (1971). Linsmaier, E. M., Skoog, F.: Organic growth factor requirements of tobacco tissue cultures. Physiol. Plant 55, 100-127 (1965). Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R. J.: Pro- tein measurement with the Folin-Phenol reagent. J. Biol. Chem. 193, 265-275 (1951). 123 Maassen, J. A., Moller, W.: Identification by photo-affinity labeling of the proteins in Escherichia coli ribosomes involved in elonga- tion factor G-dependent GDP binding. Proc. Nat. Acad. Sci. (Wash.) 25, 1277-1280 (1974). Malkinson, A. M., Krueger, B. K., Rudolph, S. A., Casnellie, J. E., Haley, B. E., Greengard, P.: Widespread occurrence of a specific protein in vertebrate tissues and regulation by cAMP of its endogenous phosphorylation and dephosphorylation. Metabolism 52, 331-341 (1975). Miflin, B. J., Beevers, H.: Isolation of intact plastids from a range of plant tissues. Plant Physiol. 55, 870-874 (1974). Miller, C. O., Skoog, F., Van Saltza, M. H., Strong, F. M.: Kinetin, a cell division factor from deoxyribonucleic acid. J. Amer. Chem. Soc. 11, 1392 (1955). Muneyama, K., Bauer, R. J., Shuman, D. A., Robins, R. R., Simon, L. N.: Chemical synthesis and biological activity of 8-substituted cAMP derivatives. Biochemistry 55, 2390-2395 (1971). Nathanson, J. A., Greengard, P.: Serotonin-sensitive adenylate cyclase in neural tissue and its similarity to the serotonin receptor: a possible site of action of LSD. Proc. Nat. Acad. Sci. (Wash.) 25, 797-801 (1974). Okumura, F. 5., Kotani, Y., Ariga, T., Masumura, M., Kuraishi, S.: Syntheses of kinetin-analogues. Bull. Chem. Soc. Japan 55, 883-887 (1959). ’ Oostrom, H., Van Loopik-Detmers, M. A., Libbenga, K. R.: A high affinity receptor for indoleacetic acid in cultured tobacco pith explants. FEBS Letters 55, 194-197 (1975). Pomerantz, A. H., Rudolph, S. A., Haley, B. S., Greengard, P.: Photo-affinity labeling of protein kinase from bovine brain with 8-azido-cAMP. Biochemistry 55, 3858-3862 (1975). Puca, G. A., Nola, E., Sica, V., Bresciani, F.: Estrogen-binding proteins of calf uterus. Partial purification and preliminary characterization of two cytoplasmic proteins. Biochemistry 55, 3769-3779 (1971). Ralph, R. K., McCombs, P. J. A., Tener, G., Wojcik, S. J.: Evidence for modification of protein phosphorylation by cytokinins. Biochem. J. 130, 901-911 (1972). Ralph, R. K., Bullivant, S., WOjcik, S. J.: Effects of kinetin on phosphorylation of leaf membrane proteins. Biochim. Biophys. Acta 421, 319-327 (1976). Riggs, A. D., Bourgeois, 8.: On the assay, isolation and characteri- zation of the lac repressor. J. Mol. Biol. 55, 361-364 (1968). 124 Rosenthal, H. E.: A graphic method for the determination and presen- tation of binding parameters in a complex system. Anal. Biochem. 55, 525-532 (1967). Scatchard, G.: The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 55, 660-672 (1949). Sharma, S. K., Nirenberg, M., Klee, W. A.: Morphine receptors as regulators of adenylate cyclase activity. Proc. Nat. Acad. Sci. (Wash.) 25, 590-594 (1975). Shelton, K. R., Clark, Jr., J. M.: A proton exchange between purines and water and its application to biochemistry. Biochemistry 5, 2735-2739 (1967). Smith, P. A. S., Hall, J. H., Kan, R. 0.: The electronic character of the azido group attached to benzene rings. J. Amer. Chem. Soriano-Garcia, M., Parthasarthy, R.: Structure-activity relation- ship of cytokinins: crystal structure and conformation of kinetin. Biochem. Biophys. Res. Commun. 55, 1062-1068 (1975). Spudich, J. L., Koshland, D. E., Jr.: Quantitation of the sensory response in bacterial chemotaxis. Proc. Nat. Acad. Sci. (Wash.) 15, 710-713 (1975). Sussman, M. R., Firn, R. D.: The synthesis of a radioactive cyto- kinin with high specific activity. Phytochem. 55, 153-155 (1976). Takegami, T., Yoshida, K.: Isolation and purification of cytokinin binding protein from tobacco leaves by affinity chromatography. Biochem. Biophys. Res. Commun. 51, 782-789 (1975). Tandeau de Marsac, N., Jouanneau, J.: Variation de l’exigence en cytokinine de lignees clonales ce cellules de Tabac. Physiol. Veg. 55, 369-380 (1972). Thomson, K. S.: The binding of l-N-naphthylphthalamic (NPA), an inhibitor of auxin transport, to particulate fractions of corn coleoptiles. In: Hormonal Regulation in Plant Growth and Development. Proc. Adv. Study Inst. Izmir 1971, pp. 83-88. Ed.: Kaldeway, H., Vardar, Y., Weinheim: Verlag-Chemie 1972. Thomson, K., Hertel, R., Muller, S. Tavares, J. E.: NPA and TIBA in vitro binding to particulate cell fractions and action on auxin transport in corn coleoptiles. Planta (Berl.) 555, 337- 352 (1973). Thomson, K. S., Leopold, A. C.: In vitro binding of morphactins and l-N-naphthylphthalamic acid in corn coleoptiles and their effect on auxin transport. Planta (Berl.) 115, 259-270 (1974). walker, G. C., Leonard, N. J., Armstrong, D. J., Murai, N., Skoog, F.: The mode of incorporation of BA into tobacco callus tRNA; a double labeling determination. Plant Physiol. 55, 737-743 (1974). 125 Whitaker, B. D., Kende, H.: Bud formation in Funaria hygrometic: A comparison of the activities of three cytokinins with their ribosides. Planta (Berl.) 121, 93-96 (1974). Whitfield, P.: Detection of purines, pyrimidines, nucleosides and nucleotides. In: Data for Biochemical Research, p. 565. Ed.: Dawson, R. M. C., Elliott, W. H., Jones, K. M., New York: Oxford University Press 1969. Zubay, G., Schwartz, D., Beckwith, J.: Mechanism of activation of catabolite-sensitive genes: a positive control system. Proc. Nat. Acad. Sci. (Wash.) 55, 104-110 (1970). "IIIIIIIIIIIIIIIIIIIIIIs