“- ,.' I ax .,_ 2’ : .1 v ’2. E. ‘l h " - Q’JDI 0". 1'. o ' '. . J '.' f -( \|.- - .- nfi, v y 5‘14 .\ “4‘31, L1 LIBRARY Michigan State University This is to certify that the dissertation entitled Host-selective toxins from Helminthosporium carbonum: purification, chemistry, biological —-——b——— activities, and effect on chlorophyll synthesis in maize. presented by Jack Bryan Rasmussen has been accepted towards fulfillment of the requirements for $.11.— degree in W ogy W Major professor Date 4 I 9 "(Iii-n- tin—d A ' “in IA - . t . 042771 MSU LlBRARlES “ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. HOST-SELECTIVE TOXINS FROM HELMINTHOSPORIUM CARBONUM: PURIFICATION, CHEMISTRY, BIOLOGICAL ACTIVITIES, AND EFFECT ON CHLOROPHYLL SYNTHESIS IN MAIZE BY Jack Bryan Rasmussen 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 1987 ABSTRACT HOST-SELECTIVE TOXINS FROM HELMINTHOSPORIUM CARBONUM: PURIFICATION, CHEMISTRY, BIOLOGICAL ACTIVITIES, AND EFFECT ON CHLOROPHYLL SYNTHESIS IN MAIZE. BY Jack Bryan Rasmussen Helminthosporium carbonum race 1 is the causal agent of a leafspot disease that affects only certain inbred lines and hybrids of maize. The fungus produces a host- selective toxin (HC toxin) that is required for disease development. A new and simpler purification scheme was developed for the major form of toxin (HC toxin I). The procedure eliminated the need for TLC and HPLC as preparative steps and resulted in rapid accumulation of crystalline toxin I. Yields were over 80 mg toxin per liter of culture fluid. Three analogs of toxin I were isolated from culture fluids of the fungus using the new purification scheme. The analogs had the same specificity as does the pathogen, and were designated HC toxins II, III, and IV. HPLC was required for final purification of the analogs. Spectral and other data indicated that the analogs, like the previously characterized toxin I, are cyclic tetrapeptides with one unusual epoxide-containing amino acid. HC toxin IV differs chemically from toxin I only at the carbon adjacent to the epoxide. Toxin IV has a hydroxylated carbon and toxin I has a carbonyl carbon. This conclusion was based on amino acid analyses, fast atom bombardment mass spectroscopy, and 130 and 1H NMR. NaBHu reduction of the ketone in toxin I produced a compound with the same chromatographic and spectral properties as toxin IV, confirming the structure. EDSO values based on inhibition of susceptible seedling root growth for toxins I, II, III, and IV were 0.2, 0.4, 2.0, and 20 ug/ml respectively. The preparation of toxin II was found to be more active than was reported previously. Resistant seedlings tolerated 100-fold higher concentrations of pure toxin I than did susceptible seedlings. Hydrolysis of the toxin I epoxide to a diol destroyed toxicity to susceptible and resistant seedlings, suggesting that the same mechanisms are affected in resistant and susceptible plants. HC toxin I was found to have a rapid inhibitory effect on the synthesis of chlorophyll in etiolated maize leaves. Approximately 50% inhibition of chlorophyll synthesis was observed in susceptible leaves with toxin at 20 ug/ml 6 hr after the initiation of greening. There was inhibition with < 1.0 ug toxin per ml and linear increases over at least five orders of magnitude. Resistant maize tissues were similarily affected when toxin concentrations were loo-fold higher than were required for susceptible tissues. The application of 6-aminolevulinic acid (ALA), the first and limiting step of the chlorophyll synthesis pathway, prevented the toxin-induced inhibition. The data indicate that toxin somehow causes a block in the synthesis of ALA in etiolated maize leaves exposed to light. This is the most rapid inhibitory effect observed to date for RC toxin. AW I thank Dr. R.P. Scheffer for directing this research, and for sharing so many scientific insights. The rest of my graduate committee, Drs. R. llarrmerscl'midt, R.S. Bandurski, and N.E. Good, provided valuable discussion and asked fascinating exam questions. I am deeply indebted to Linda Matthews-Rasmussen for encouraganent, support, and patience, particularly at the end of the dissertation. ii TABLE OF CONTENTS Page LIST OF TABLES ........................................ LIST OF FIGURES ....................................... INTRODUCTION .......................................... 1 LITERATURE REVIEW ..................................... 3 METHODS AND MATERIALS 0.0...00....OOIOOOOOOOOOOOOOOOOOO1” Toxin Production and Initial Purification .................................. 14 Thin Layer Chromatography ........................ 15 Flash Chromatography ............................. 15 HPLC ............................................. 16 Hydrolysis of Epoxide ............................ 17 NaBH Reduction of Toxin I ....................... 17 Specflroscopic Analyses ........................... 17 Root Growth Inhibition Bioassay .................. 18 Chlorophyll Synthesis in Etiolated Maize Leaves .................................. 19 Determination of Chlorophyll ..................... 20 Formation of Heat Shock Proteins in Maize COO0....OOOOOIOOOOOOOOOOOOOOOO0.0.0.0.... 20 RESULTS OOOOOOOOOOOOOOOOIOOOOOOCOIOOOOOOOCOOOOOOOOOOOOO 23 Isolation and Identification of Major Toxin ................................... 23 Isolation and Identification of Toxin Analogs ................................. 32 Structure of HC Toxin IV ......................... 41 Effect of Toxin on Chlorophyll Synthesis in Etiolated Leaves ................. 51 Recovery of Toxin from Homogenized Leaves ........................................ 60 iii DISCUSSION O0.0.0.000....0...OOOOOOOOOOOOOOOOIOOOOOOOOO 63 APPENDIX OOOOOOOIIO...00.000.00.00...OOOOOOOOOOOCCOOOOO 7” LITERATURE CITED 00.0.0000...0.000000000000000000000000 79 iv LIST OF TABLES Table Page 1 Inhibition of susceptible seedling root growth by HG toxin II as determined by two different assays 0.00.00.00.00...OOOOOOOOOOOOO... no 2 Assignments of 13C resonances for RC toxins I and Iv 00......0.000.0.0.000000000000000000000000 “8 3 Effect of HC toxin I (20 ug/ml) on electrolyte leakage from susceptible maize leaves 0....O0.0....O.I0..OOOOOOOOOOOOOOOOOOOOOOO 57 A Effect of ALA on the lag phase of chlorophyll synthesis in eiolated maize leaves .............. 59 5 Uptake and incorporation of 35S-methionine by root tips of susceptible of resistant maize at 25 and “5C .0.000000000000000000000.0.0.0.... 76 LIST OF FIGURES Figure Page 1 HPLC of HC toxin I crystals. Chromatography was on a column (0.4 x 25 cm) of Partisil 10. Elution was with hexane/ethanol (95/5, vzv) at 4.0 ml/min ................................... 24 2 1H NMR spectrum of crystalline HC toxin I. Spectrum was collected on a Bruder WM-250 spectrometer at 250 MHz. Chemical shifts are relative to internal CHCl3 ...................... 25 3 13C NMR spectrum of crystalline HC toxin I. Spectrum was collected on a Bruder WM-250 spectrometer at 68.9 MHz. Chemical shifts are relative to internal CHCl3 .................. 27 4 Structures of HC toxins I, II, and III .......... 29 5 Effect of crystalline HC toxin I on seedling root growth. Susceptible (O) and resistant (A ) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions for 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water ........................................ 30 Figure Page 6 Effect of HC toxin I (O) and its diol (O) on root growth of resistant seedlings. Five seedlings per duplicate 9-cm Petri dish were incubated in water or in dilutions of each compound for 72 hr; 10 ml solution was used in each dish. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water .............................. 31 7 1H NMR spectrum of HC toxin I diol. Spectrum was collected on a Bruder WM-250 spectrometer at 250 MHz. Chemical shifts are relative to internal CHCl3 .................................. 33 8 HPLC of a preparation containing minor HC toxins. Chromatography was on a reverse phase column (Cj ) (2.2 x 30 cm) of Waters Bondupak. Elutign was with a gradient; the initial solvent of 7% ethanol in water was changed linearly to 20% ethanol in water over 30 min. The solvent composition was then held constant at 20% ethanol in water for the next 15 min. The flow rate was constant at 2.0 ml/min ...................................... 35 9 Partisil 10 HPLC of peak RP1 (from Figure 8), separating HC toxins II and III. Column dimensions were 0.4 x 25 cm. Elution was with hexane/ethanol (95/5, vzv) at 4.0 ml/min .......................................... 36 10 Effect of HC toxin II on seedling root growth. Susceptible (O) and resistant (A) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions for 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water ........................................ 38 Figure Page 11 Effect of HC toxin III on seeling root growth. Susceptible (O) and resistant (‘) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water ........................................ 39 12 Final purification of HC toxin IV by HPLC. Chromatography was on a column (0.4 x 25 cm) of Partisil 5. Elution was with hexane/ethanol (93/7, v:v) at 3.0 ml/min ........ 42 13 Effect of HC toxin IV on growth of seedling roots. Susceptible (O) and resistant (A) seedlings (5 per duplicate 9—cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water ........................................ 43 14 Proposed structure for RC toxin IV .............. 45 15 13C NMR spectrum of HC toxin IV. Spectrum was collected on a Bruder WM-250 spectrometer at 68.9 MHz. Chemical shifts are relative to internal CHCl3 .................................. 46 1H NMR spectrum of HC toxin IV. Spectrum was collected on a Bruder WM-250 spectrometer at 250 MHz. Chemical shifts are relative to internal CHC13 00.0.0.0...OOOOOOOOOOOOOOOOOOOOOOI “9 16 17 Effect of HC toxin (20 ug/ml) on chlorophyll synthesis in susceptible maize. Leaves accumulated solutions of buffer (C) or toxin (1‘) for 2 hr in the dark (through the transpiration gtream) prior to exposure to light (33 uE/m /sec) ............................ 52 18 Effect of toxin (20 ug/ml) on susceptible and resistant leaves, expressed as 1 inhibition of chlorophyll synthesis. These include data given in Figure 17, plus data with resistant tissues ......................................... 53 viii Figure Page 19 20 Toxin concentration effects on chlorophyll synthesis in susceptible and resistant leaves. Toxin was administered for 2 hr in the dark; leaves were then placed in water and in the light (6 hr) to induce greening ...... 55 Effect of ALA (1.0 mg/ml) on toxin-induced inhibition of chlorophyll synthesis. Toxin concentration was 175 ug/ml. Leaves took up solutions in the transpiration stream in the dark for 2 hr prior to exposure to light (1.7 uE/m /sec) ...................................... 61 ALA 808 FAB MS HPLC NBP NMR TLC LIST OF ABBREVIATIONS é-aminolevulinic acid 2-amino-8-oxo-9,10-epoxydecanoic acid fast atom bombardment mass spectroscopy high pressure liquid chromatography p-(nitrobenzy1)-pyridine nuclear magnetic resonance thin layer chromatography Introduction HelminthOSporium carbonum (Ullstrup) race 1 is a pathogen of certain varieties of maize. The fungus, which causes a severe leaf blight, produces a host-selective toxin (HC toxin) as its major disease determinent (36). When this research was initiated, only one form of toxin (HC toxin I) had been purified and characterized chemically. The existing purification scheme was slow and tedious; thin layer‘chromatography (TLC) and analytical high pressure liquid chromatography (HPLC) were limiting preparative steps. In the first part of this research a new purification scheme was developed for toxin based on flash chromatography (46) on silica gel. The scheme eliminated the need for TLC and HPLC, thus greatly simplifying and expediating toxin purification. At least three analogs of toxin I were isolated from culture fluids. All analogs required HPLC for final purification. Two analogs (HC toxins II and III) were characterized chemically as part of another thesis (47); the structure of the third analog (HC toxin IV) was elucidated in my research using amino acid analyses and spectroscopic 1H NMR). Shortly after the techniques (FAB MS, 13C and research was initiated, another group reported the structure of toxin II (19). However, my preparation was much more active, suggesting higher degree of purity. The second portion of my research was concerned with the mode-of—action of HC toxin. This has been an elusive problem for toxin researchers since all early effects of HC toxin on susceptible cells to date are stimulations or increases rather than inhibition of activities (22,57,58). I found that toxin has a rapid inhibitory effect on chlorophyll synthesis in etiolated leaves of susceptible maize. HC toxin inhibitied chlorophyll synthesis within 4 to 6 hr after exposure to light. The data are the most rapid inhibitory response to H0 toxin observed to date, and may probide the basis for an improved and more rapid bioassay for toxin. The inhibition of chlorophyll synthesis could be prevented by supplying ALA, the first limiting step of the chlorophyll synthesis pathway, to toxin-treated leaves. The data provide clues to the mode- of-action of HC toxin. Literature Review Remarkable selectivity is involved in many plant diseases; resistance or susceptibility often is controlled by one gene pair. Also, pathogenicity or the ability to induce disease is known or hypothesized to be under single gene control in many fungal pathogens. Most genetic studies of diseases involving fungi have progressed little beyond this observation of genotype or race specificity. Little is known about gene products for susceptibility or resistance in plants, and few pathogenicity factors are known from fungi. The only known exceptions are in the diseases involving host-selective toxins (36). In these systems, a single metabolite from the fungus, the host- selective toxin, is known with confidence to be the major disease determinant. Host-selective toxins are low molecular weight compounds produced by the pathogen that have the same host specificity as does the pathogen; resistant genotypes and species are not affected or are highly tolerant (36). To date, 15 host-selective toxins are recognized (37). The host-selective toxin from H; carbonum race 1 (HC toxin) is the subject of this study. The fungus causes a leaf spot and ear rot that affects only certain genotypes of maize (50). Scheffer and Ullstrup (42) first demonstrated that the fungus produced a host-selective toxin in culture. That report came shortly after the discovery of the host-selective toxin from Periconia circinata (PC toxin), a pathogen of grain sorghum (39). The discovery of HC toxin was also preceded by the discovery of host-selective toxins from Alternaria alternata f. kikuchiana (AK toxin) and Helminthosprium victoriae (HV toxin), pathogens of Japanese pear cv. Nijissicki and certain cultivars of oats, respectively (29,37). The physiological effects of HC toxin have often been compared to those of PC and HV toxins. These comparisons are described below. HC toxin is well-characterized chemically. Pringle (33) offered a partial characterization of toxin based on amino acid analyses and paper chromatography. Leisch £3 31; (25) provided strong spectral evidence that toxin is a cyclic peptide containing alanine, proline, and an unusual epoxide-containing amino acid, 2-amino-8-oxo-9,10-epoxy— decanoic acid (ace), in a 2:1:1 ratio, respectively. This amino acid composition was confirmed by two other groups of researchers, (32,52), who reported an amino acid sequence different from that proposed by Leisch gt El; (25). Walton 33 El; (52) first identified the toxin as cyclo-(2-amino-8- oxo-9,10-epoxydecanoyl-prolyl-alanyl-alanyl). Synthesis of the toxin has confirmed the structure (18). The epoxide was shown to be necessary for toxicity (6,53). Hydrolysis of the epoxide to a diol resulted in a nontoxic compound that did not protect susceptible seedling from toxin in molar ratios of 7:1 (diol:toxin) in root growth inhibition bioassays (6). An analog of toxin that differed structurally from the major form of toxin by the substitution of glycine for alanine adjacent to ace was found in culture fluids of the fungus (19). The analog had the same specificity, but was reported to be 35-fold less potent, on a molar basis (19). HC toxin was used in several classic experiments that established host-selective toxins as pathogenicity factors. The teleomorph of the fungus, Cochliobolus carbonum, was crossed with g; victoriae, the producer of RV toxin, and the ascospore progeny were analyzed (38). The ascospore progeny produced either the maize toxin, the oat toxin, both toxins, or neither toxin in a 1:1:1:1 ratio. All ascospores which produced the maize toxin were pathogenic to susceptible maize, and all ascospores which did not produce the toxin were not pathogenic to maize. These findings were confirmed and extended by use of laboratory mutans and wild type isolates of the fungus from around the world (55). These data are convincing evidence for toxins as pathogenicity factors, and indicate that HC and RV toxins are produced under control of single, unlinked genes (38). HC toxin is required for colonization of susceptible maize. Conidia of g; victoriae and of a nonpathogenic isolate of g; carbonum germinated and penetrated cell walls of susceptible maize but did not develop further; at most, a few cells were penetrated, and a hypersensitive death of the cells occurred. Resistance was evident by 16 hr post- inoculation (7). When toxin was administered exogenously to the infection court, these non-pathogenic isolates colonized susceptible tissue in the same manner as did the pathogenic isolate. Resistance of maize to H; carbonum is inherited in a simple dominant fashion (51). The major locus controlling the disease reaction, Hm, was mapped on chromosome 1. Further work demonstrated that the Hm locus has 3 alleles. The degree of resistance in plants not carrying the dominant allele (Hm) was conditioned by a second locus, Hm located on chromosome 9 (31). Resistance to the 2, fungus was then compared with resistance to the toxin (23). These comparisons were only made with the Hm genotypes; Hm2 was not available for tests. Genotypes that were highly susceptible to the fungus were most sensitive to toxin, whereas the genotypes that were resistant to the fungus were insensitive to toxin; genotypes intermediate in susceptibility were intermediate in sensitivity to toxin (23). Thus, resistance to the toxin equals resistance to the fungus. Plants treated with toxin exhibit many of the same responses as do plants infected with the fungus. Infection by H; carbonum was reported to stimulate maize leaves to fix more C0 in the dark than do control plants (27). The 2 level of C0 fixation by infected tissue eventually 2 decreased relative to control plants, apparently because of the accumulation of inhibitors (27). RC toxin was shown to cause the same effect in susceptible leaves 4 hr after exposure to toxin, but the levels always remained above those of control plants (22). Apparently, no inhibitors accumulate after toxin treatment (22). Toxin-treated resistant leaves also fixed more CO2 in the dark than did their water-treated controls, but the level of increase was always smaller than in toxin-treated susceptible leaves at a given toxin concentration (22). Susceptible leaves consumed about 30% more oxygen than did control leaves in response to an eight hr exposure to toxin, and respiration remained above control levels for at least 30 hr (22). Increased respiration is the usual response by plants to infection. The increased respiration of maize leaves in response to BC toxin was less in magnitude and slower to develop than that which occured in susceptible oat leaves treated with HV toxin (35,40). Many of the physiological effects of HC toxin differ from the effects of other host-selective toxins. A toxin preparation that had an ED of 1.0 ug/ml against 50 susceptible seedlings stimulated root growth of the same seedlings at 0.125 ug/ml over 48 hr (23). Higher toxin concentrations stimulated root growth of resistant seedlings (23). The same preparation was tested for inhibition of seedling root growth of various non-host plant species. All the non-host plants were much more resistant to toxin than was susceptible maize, but some species, for example tomato and radish, were not as tolerant as was resistant maize (23). Many of the non-host species were stimulated in their root growth by low toxin concentrations, suggesting that some processes were affected by toxin in all plants tested (23). RC toxin at 5.0 ug/ml caused a rapid but transient increase in the negative electropotential across the plasmalemma of susceptible maize cells (13). The increases, 10 to 40 mv in magnitude, were evident within the first 3 minutes of toxin exposure, but the electropotential returned to initial levels within 30 min (13). Such increases were in sharp contrast to the effects of RV and PC toxins on their susceptible hosts. Those toxins gradually decreased the electropotential of susceptible cells at a rate of approximately 50 mv/hr (13). The data indicate that the effect of HC toxin on the plasmalemma differs from that of the other toxins (13). There have been attempts to demonstrate uptake of toxin by plant tissues. Roots of seedlings were placed in toxin solutions for 12 hr, after which seedlings were removed and another set of seedlings were exposed for 12 hr. A third set of seedlings was handled the same way. The residual toxin solution was then bioassayed and found to contain as much toxic activity as did control toxin solutions which had never been in contact with seedlings (21). The same negative results were obtained in experiments with resistant and susceptible seedlings (21). If toxin was removed from solution by either genotype, the amounts were not detectable by the bioassay (21). In another experiment, leaves took up toxin in the transpiration stream for 20 hr. The leaves were then ground, and water extracts were bioassayed. No host- selective toxicity was recovered from leaves of susceptible or resistant seedlings. These negative results are similar to those for HV toxin (41). Host-selective toxin activity, as determined by the root growth inhibition assay, was recovered from leaves of sorghum plants exposed to PC toxin (12). Equal amounts of PC toxin activity were recovered from susceptible and resistant plants in those experiments (12). Experiments indicate that uptake or activity of HC toxin is governed by rate-limiting mechanisms, and requires metabolic energy (21). Small seedlings were placed in toxin solution for various times (4, 6, or 10 hr) under a variety of conditions. The seedlings were then rinsed and placed in water for 96 hr. The inhibition of root growth after the 96 hr incubation was used as a measure of toxin activity during the brief exposure to toxin. Activity increased with toxin concentration, temperature, and exposure time (21). The metabolic inhibitors azide, cyanide, and 2,4-dinitrophenol decreased toxin activity when they were administered during the brief exposure to toxin (21). Similar reductions in activity were observed when anaerobic conditions were imposed during the toxin exposure time (21). The data could be interpreted in several ways, but the most simple explanation is that toxin uptake was affected by the manipulations (21). The effect of toxin on nitrate reductase, a substrate inducible enzyme system, was determined. Toxin at 20 ug/ ml increased in 1113 nitrate reductase activity by 20 to 30% within 4 hr after exposure to the nitrate substrate (57). Toxin treatment also doubled the uptake of nitrate by susceptible maize by one hr after exposure. Nitrate reductase activity from plant extracts was not affected in 11339, indicating that toxin had no direct or primary effect on the enzyme (57). Other experiments indicated that toxin had no influence on the ability of cells to retain nitrate; overall, the data indicate that the increased level of nitrate reductase activity probably was the result of increased availability of substrate caused by the stimulated uptake of nitrate (57). RC toxin similarly increased the uptake of Na, Cl, leucine, and 3-o-methylglucose in susceptible maize roots, 10 but had no effect on uptake of N0 K, Ca, phosphate ions, 2: SO“, and glutamic acid (58). These data indicate that HC toxin does not cause the general disruption of plasma membranes that is caused by RV and PC toxins. Rather, the increased uptake of certain solutes appears to be the result of other toxin-induced changes in the plasmalemma. Electrolyte leakage from susceptible maize tissues is increased by toxin, but this does not occur until 10 to 16 hr after exposure to toxin (32). The slow response indicates that the plasmalemma is not disrupted in the early events of toxin action. The data indicate that HC toxin is subtle in its action; the known physiological processes are stimulated or increased rather than disrupted. Additionally, resistant tissues respond as susceptible tissues, provided toxin concentration is approximately 100-fold higher (23.57.58). This has led to the idea that similar processes in many plants may be affected by toxin. However, the available data do not allow us to propose an initial lesion site for toxin. Because disruptive effects of toxin are slow to develop, the only Suitable bioassay for toxin to date is based on inhibition of susceptible root growth. This assay is quantitatively reliable, but requires at least 48 hr exposure to toxin (6,23,38). This has limited the scope of experiments concerning the mode-of—action. An objective of my research was to investigate physiological systems that might be used to gain clues to 11 the mode-of—action, and perhaps provide the basis for a more rapid bioassay. I found that the toxin has a rapid inhibitory effect on chlorophyll synthesis in etiolated corn leaves, so it is appropriate to briefly review the pertinent literature. A detailed review on chlorophyll synthesis has been published recently (5). Chlorophyll is a porphyrin derived from a branch of the tetrapyrrole pathway. o-aminolevulinic acid (ALA) is the first identified precursor of the pathway, which gives rise to heme synthesis in all organisms (including bacteria and animals), as well as to chlorophyll in plants (1). In plants, the pathway branches after protoporphyrin IX; insertion of Fe2+ into this intermediate leads to heme synthesis whereas chelation of Mg2+ results in chlorophyll synthesis (1). Etiolated leaves, when placed in light, have a lag phase of about 2-4 hr before chlorophyll is synthesized (15,30). When ALA is supplied in the dark, this lag phase is eliminated; the leaves begin making chlorophyll immediately upon exposure to light (30). Leaves supplied with ALA and held in the dark accumulate protochlorophyllide, an intermediate of the pathway which requires light for the enzymatic reduction to chlorophyllide. These experiments have led to two important conclusions: 1) all of the enzymes required for chlorophyll synthesis are non-limiting in etiolated leaves, 12 except for the enzyme complex which makes ALA; and 2) ALA is the major controlling point of the pathway. The lag phase of chlorophyll synthesis is thought to be associated with the 33 £319 synthesis of the unnamed enzyme complex which makes ALA (30). Inhibitors of protein synthesis prevent the formation of chlorophyll in etiolated leaves placed in the light, but the application of ALA reverses this effect (30). Other studies directly show that ALA formation in maize leaves is light-dependent (15). Isolation and characterization of the enzyme complex which makes ALA in plants has been accomplished only recently. The system is remarkably different from that in bacteria and animals, where ALA is formed from succinyl CoA and glycine (17,43). In plants, ALA is formed from glutamic acid (2) by an enzymatic system that consists of two protein fractions and an RNA species (17). The RNA is a chloroplast glutamate tRNA that functions as a cofactor in the synthesis of ALA (43). 13 Methods and Materials Toxin Production and Initial Purification. Procedures for production and initial purification of toxins were modified from those used previoulsy (34). Single spore isolates of H; carbonum race 1 were maintained on potato dextrose agar slants at 4 C. For toxin production, the fungus was grown in a modified Fries solution containing yeast extract (34). Fluids (10-20 liters) from 21 day old cultures were filtered through cheesecloth, and through paper (Whatman #1), and then concentrated to approximately 0.1 original volume under reduced pressure at 40 C. An equal volume of methanol was added to the concentrated filtrate and the solution was stored overnight at -20 C; a precipitate was removed by filtration through paper (Whatman #42). Methanol was then removed under reduced pressure and the aqueous solution was extracted three times, each time with an equal volume of methylene chloride. Steps to this point were completed as quickly as possible to avoid hydrolysis of the epoxide. All subsequent preparations were dried under reduced pressure and stored under argon or nitrogen at -20 C. The combined methylene chloride extracts were dried under reduced pressure. The remaining reddish oil (several ml) was dissolved in methanol and placed on a column (4.1 x 14 35 cm) of Sephadex LH-20 (Sigma Chemical Co.) previously equilibrated with methanol. The column was eluted with methanol and five ml fractions were collected. Fractions containing toxin were identified by bioassay (38) and by TLC. Thin Layer Chromatography (TLC) Thin layer chromatography (TLC) was on .25 mm silica gel plates (Merck) developed with acetone/methylene chloride (1:1, v/v). Toxins were detected by spraying the chromatographs with the epoxide indicator 4-(p- nitrobenzyl)-pyridine (NBP) (14). Flash Chromatography The toxin containing fractions from the LH-20 column were combined and concentrated under reduced pressure, giving a viscous orange oil that was fractionated by flash chromatography (46) using 230 - 400 mesh flash silica. The column diameter was dictated by sample dry wieght (46), but the length of the packed silica bed was approximately 15 cm regardless of column diameter. The column was eluted with hexane:methylene chloride:acetone (1:1:1, v/v/v; solvent A) and fifty ml fractions were collected. The fractions were periodically examined by TLC, using the epoxide indicator for the presence of a reactive spot at Rf .55. This spot was shown to be toxin I. After toxin I was eluted in solvent A, the mobile phase was changed to methylene chloride:acetone (1:1, v/v; solvent B). The column was 15 then eluted with 1000 ml of solvent B and the eluate was collected as a single fraction. TLC of this eluate indicated epoxide-containg spots between Rf .25 and .35. These spots were shown to have host-selective toxicity, indicating that they are possible toxin analogs. The entire preparation that eluted in solvent B was refractionated by flash chromatography on silica gel using a smaller column (20-30 mm diameter) with both solvents A (300 ml) and B (500 ml) as described above. Twenty ml fractions were collected. The solvent A eluate from this step was pooled with that from the previous step, and toxin I was crystalized with diethyl ether (25). Toxin analogs were located in the solvent B fractions by TLC; the fractions were pooled to form the minor toxin preparation. HPLC Solvent B eluate from the flash chromatography was dissolved in water (25 ml) and placed on a column (2.2 x 15 cm) of C (40 um, J.T. Baker Chemical Co.). Loosely 18 absorbed compounds were removed by passing 100 ml of water through the column. Analogs of toxin I were then eluted with 30% ethanol in water (100 ml). This eluate was loaded onto a Waters uBondapac C18 HPLC column (.78 X 30 cm) and chromatographed with a Varian 5000 instrument. The initial solvent of 7% ethanol in water was linearly changed over 30 min to 20% ethanol in water , which was then held constant for the next 15 min. The flow rate was constant at 2.0 ml/min and absorbance was monitored at 215 nm. Final 16 purification of the toxic analogs was with a 0.4 X 25 cm column of Whatman Partisil 5. Elution was with an isocratic mixture of hexane/ethanol (95/5 or 93/7, v/v) at 3.0 ml/min and absorbance was monitored at 215 nm. Hydrolysis of Epoxide The epoxide of toxin I was hydrolyzed to a diol with 0.1% (v/v) triflouroacetic acid (TFA) in water (6). The diol was purified on the 2.2 x 15 cm column of C18 as described above for the minor toxins, except that elution was with 100 ml of 10% (v:v) ethanol in water. The eluate, rich in toxin I diol, was subjected to HPLC, using the Partisil 5 column and hexane/ethanol (85:15, v/v) as described above. NaBH” Reduction 3: Toxin I The carbonyl adjacent to the epoxide of toxin I was reduced with NaBH in anhydrous methanol. NaBH (1 to 2 M) 4 4 was added to 0.2 M toxin. The reaction mixture, capped under dry N2, was placed on ice. The reaction was stopped after one hr incubation by the addition of several ml water. Reduced toxin was immediately extracted with several volumes of methylene chloride. This extract was subjected to HPLC, using the Partisil 5 column as described above for toxin analogs. 17 Spectroscopic Analyses Amino acid analysis was performed by the Macromolecular Structure Facility at Michigan State University using a Waters Associates Pico-Tag analyzer. l H- and 13C-NMR spectra were collected in CDCl at 250 3 and 68.9 MHz, respectively, on a Bruder WM-250 spectrometer. Chemical shifts are relative to internal CHC13. FAB MS was performed by Michigan State University Regional Mass Spectroscopy Facility, Department of Biochemistry, on a JEOL RX 110 HF instrument. High resolution masses were relative to internal glycerol or RC toxin I. Root Growth Inhibition Bioassay A root growth inhibition bioassay was used to determine activity of toxins (38). Seeds were near isogenic Pr X K61 (susceptible) and Pr1 X K61 (resistant) maize. Seeds (5 per duplicate dish) were placed in 9-cm Petri dishes containing 10 ml water or dilutions of toxin in water. Root lengths were measured after incubation for 72 hr. Percent inhibition and ED50 values were determined as described elsewhere (38). The assay data were from the results of a single representative experiment and all assays were performed three times with similar results. Toxins used in the assays were freshly purified by HPLC to minimize the possibility that inactive toxin (6) contaminated the preparations. 18 Electrolyte Leakage Experiments Electrolyte leakage from maize leaves was determined by a method modified from a procedure developed for use with other toxins (4). The terminal 2 cm of the second true leaf of plants (10 to 14 days old) grown under Sylvania Gro-Lux lights was excised and out once to give two 1-cm pieces. Leaf pieces (200 mg per duplicate sample) were enclosed in cheesecloth bags and submersed in 10 ml of distilled water or RC toxin I solution (20 ug/ml) contained in scintilation vials. The samples were vacuum-infiltrated for 15 min, and incubated in experimental solutions was for a total of 2 hr (including the time under vacuum). After incubation, the leaves were rinsed thoroughly in distilled water, and 10 ml of fresh distilled water was added to each vial. The conductivity of the ambient solution was measured periodically with a conductivity meter (Markson Science, Inc). Duplicate samples were used for each treatment. Chlorophyll Synthesis £3 Etiolated Maize Leaves Crystalline toxin I were used in all experiments on chlorophyll synthesis. Toxin and other experimental compounds were used in 6 m1 of 0.01 M potasium phosphate buffer in 10 ml vials. ALA was purchased from Sigma Chemical Co. Protocol for experiments on chlorophyll synthesis was modified from that used in work on maize (15) and barley (30). Near isogenic susceptible (Pr x K61) and 19 resistant (Pr1 x K61) maize seedlings were grown in the dark for 9 to 11 days. The etiolated leaves were excised above the coleoptile while working under a green safe-light (30). The basal ends of excised leaves (2 to 4 gm fresh weight per duplicate sample) were placed in the experimental solutions and held for two hours in the dark; a small fan was used to increase transpiration and the uptake of solutions (30). In most experiments, leaves were then removed from toxin solutions and the basal ends were placed in water. The leaves were held under Sylvania Gro- Lux lights at 33 uE/mZ/sec to promote chlorophyll synthesis. In experiments involving the exogenous application of ALA, leaf bases were left in the experimental solutions while greening. Light intensity for those experiments was 1.7 uE/mZ/sec. Determination of Chlorophyll After exposure to light, leaves were blotted dry and weighed. Chlorophyll was extracted with acetone/0.1 M CaCO3 in water (9:1, v:v) as described elsewhere (30). The leaves were ground in 30 ml of the solution; an additional 20 ml were used to rinse the blender. The extracts were filtered through paper, placed in capped vials, and allowed to clarify overnight at 4 C in the dark. Chlorophyll content was measured with a Gilford 240 spectrophotometer, using the equations of MacKinney (26). Duplicate samples were used in each treatment. Standard deviations generally were less than 10% of the mean. Each experiment was 20 performed at least three times with similar results. Data given are the results of a single experiment. Formation of Heat Shock Proteins 13 Maize A procedure modified after that of Cooper and Ho (8) was used to induce the synthesis of and to analyze heat shock proteins in maize. Seeds of susceptible and resistant maize were germinated between moist filter paper for three days at room temperature. The terminal 5 to 10 mm of each root tip was excised and placed in Eppindorf tubes containing 100 ul water or HC toxin I at 20 ug/ml four root tips were used per tratment. The tissues were incubated for 1 hr at room temperature, after which 50 uCi 3SS-methionine (specific activity 1330 Ci/mMol) was added to each treatment. The samples were capped and held at room temperature (controls) or were heat shocked at 45 C for an additional 3 hr. At the end of this incubation the root tips were rinsed thoroughly, blotted dry, placed in new Eppindorf tubes, and frozen in liquid nitrogen. Proteins were extracted by homogenizing the frozen root tips in 100 ul of an extraction buffer with the aid of a ground glass tissue homogenizer. The extraction buffer, modified from (24), contained 140 mM tris (pH 7.5), 2mM phenylmethylsulfonyl flouride, 2% (w:v) sodium docylsulfate, and 1% (v:v) dimethylsulfoxide. The homogenate was microfuged for 15 min, and the supernatant was transferred to a new Eppindorf tube; the pellet was 21 35S discarded. Incorporation of -methionine into protein was determined by precipitation with 5% and 10% trichloroacetic acid in water (28). Equal cpm of incorporated 35S-methionine per treatment was subjected to polyacrylamide gel electrophoresis. Samples were subjected to electrophoresis overnight in buffer, modified from (24) of 10% glycerol (v:v), 2% beta-mercaptoethanol (v:v), 2.1% SDS (w:v), 625 mM Tris (pH 6.8). The gels were stained with a solution of 0.1% coomassie brilliant blue (w:v), 7% glacial acetic acid (v:v), 50% methanol (v:v) for 1 hr, and destained with a solution of 5.4% (v:v) glacial acetic acid, 15.4% (v:v) methanol, and 2.3% (v:v) glycerol (8). 22 Results Isolation and Identification of Major Toxin Flash chromatography with solvent A gave a relatively pure toxin preparation. Toxin I was crystalized from the eluate in diethyl ether; one liter of culture filtrate yielded well over 80 mg of crystaline toxin I. HPLC on the Partisil 10 column indicated that crystaline toxin I had a retention time of 11 min; no impurities were detected (Fig 1). High resolution FAB MS indicated a molecular formula of C21H32N4O6’ which is identical to that reported for HC 1H and 13C NMR spectra (Fig 2 and 3) were toxin I (32). also identical to those published for toxin I (32); this confirms the chemical identity of the compound (Fig 4). ED50 for the crystalline toxin I was 180 to 230 ng/ml (ca. 0.4 umolar) in three assays against susceptible seedlings; resistant seedlings were not affected at concentrations up to 5.0 ug/ml (Fig 5). This specific activity matches that of the most active preparation reported for HC toxin (32). The ED50 value of toxin I against resistant seedlings was about 20 ug/ml, or about 100-fold higher than that for susceptbile seedlings (Fig 6). Crystals were very stable when stored for several months; no reduction in specific activity was ever observed for any preparation of crystals. Crystals were usually 23 Detector Response 1 J l L J 0 10 Minutes Figure 1. HPLC of HC toxin I crystals. on a column (0.4 x 25 cm) of Partisil 10. hexane/ethanol (95/5, v:v) at 4.0 ml/min. 24 20 Chromatography was Elution was with Figure 2. 1H NMR spectrum of crystalline HC toxin I. Spectrum was collected on a Bruder WM-250 spectrometer at 250 MHz. Chemical shifts are relative to internal CHCl3. 25 J 2; ;: 1< ,: (a {W :11 7 Figure 3. 13C NMR spectrum of crystalline HC toxin I. Spectrum was collected on a Bruder WM-250 spectrometer at 68.9 MHz. Chemical shifts are relative to internal CHC13. 27 fuzv >. 1.5).? 7' hr; .r.l.....:a.u1r.r.u:st $734731... 14.41.11]. _ 28 Figure 4. H o\\C/N\ IR' 013/ H \ -o c C’ H / \ HN NH \ c H o H - - _ - 0 \CH /C (CHZ)5 fi C\ICH2 Rz’c” I *0 \ ,CH2 CH2 TOXIN l R'2cw3 RzzH TOXIN H 3,21: Rzzfl TOXINIII n,:CH3 R2:OH Structures of HC toxins I, II, and III. 29 90- Susceptible 70 4 c 50 " 2 3.3; i 30- 89 ‘0 ' Resistant I — A q l H I I I I I I I I O 0.05 0.1 0.2 0.5 1.0 2.5 5.0 Toxin concentration (99 -ml '1) Figure 5. Effect of crystalline HC toxin I on seedling root growth. Susceptible (O) and resistant (A) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions for 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water. 30 80 O O Toxinl ° Toxinl Diol 60 *- c O o: 40 '- f .c E 5‘9 20 *- 1 O r- 0 _fi/,/ 1 l l l J l O 5 10 20 50 100 250 Concentration (pg - mid) Figure 6. Effect of HC toxin I (O) and its diol (O) on root growth of resistant seedlings. Five seedlings per duplicate 9-cm Petri dish were incubated in water or in dilutions of each compound for 72 hr; 10 ml solution was used in each dish. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water. 31 stored at -20 C, although they appeared to be equally stable at room temperature. The epoxide of toxin I was hydrolyzed to a diol by use of 0.5% (v/v) triflouroacetic acid in water (6). The diol was purified by column chromatography and HPLC. 1H NMR (Fig 7) and FAB MS data matched those previously published for toxin I diol (6), confirming the chemical identity of the compound. Toxicity against susceptible and resistant seedlings was abolished by hydrolysis of the epoxide to a diol (Fig 6); root growth of resistant seedlings was not inhibited by diol concentrations up to 250 ug/ml (Fig 6). Isolation and Identification of Toxin Analogs HPLC was required for purification of toxin analogs in the solvent B eluate from flash chromatagraphy. Co- chromatography of compounds was a serious problem on HPLC with C18 sorbent; satisfactory resolution of toxins was never achieved with that column. Three major peaks, designated RP1, RP2, and RP3 in order of elution, were obtained in a gradient of water and ethanol on the C18 column (Fig 8). Peaks RP1 and RP2 contained epoxides as indicated by reactivity with MB? and were selectively toxic in root growth inhibition bioassays; peak RP3 lacked an epoxide and did not inhibit the growth of maize roots. RP1 was difficult to purify further on C18’ but was easily separated into two major components by use of the Partisil 5 column (Fig 9). 32 Figure 7. 1H NMR spectrum of HC toxin I diol. Spectrum was collected on a Bruder WM-250 spectrometer at 250 MHz. Chemical shifts are relative to internal CHC13. 33 'I CLL m : infill - razv 34 Detector Response . l l J l 1 1 0 IO 20 30 4O 50 Minutes Figure 8. HPLC of a preparation containing minor HC toxins. Chromatography was on a reverse phase column (C18) (2.2 x 30 cm) of Waters Bondupak. Elution was with a gradient; the initial solvent of 7% ethanol in water was changed linearly to 20% ethanol in water over 30 min. The solvent composition was then held constant at 20% ethanol in water for the next 15 min. The flow rate was constant at 2.0 ml/min. 35 TCDXJIQ II Detector Response )— - l 1 l 0 10 2O 3O 40 AAinutes Figure 9. Partisil 10 HPLC of peak RP1 (from Figure 8), separating HC toxins II and III. Column dimensions were 0.4 x 25 cm. Elution was with hexane/ethanol (95/5, v:v) at 4.0 ml/min. 36 Both peaks from the Partisil 5 column had host selective toxicity; they were designated HC toxins II and III in order of elution (Fig 9). Toxin II had an ED50 value of 370 to 400 ng/ml (ca. 0.8 umolar) in three assays against susceptible seedlings (Fig 10). The ED50 for toxin III was 1800 to 2100 ng/ml (ca. 4.4 umolar) in three assays (Fig 11). Growth of resistant seedling roots was not inhibited by toxins II and III at concentrations up to 10.0 and 20.0 ug/ml, respectively (Fig 10 and 11). Spectral data for toxins II and III were collected and published as part of another thesis (47). HO toxin II was shown to be the glycine-containing analog first described from culture fluids of Ht carbonum by Kim gt El; (19) (Fig 4). Toxin III was shown to contain trans-3-hydroxyproline rather than proline as its only difference from toxin I (Fig 4). Toxin III had not been reported previously. Kim ‘Et al. (19) reported an ED of 7.0 ug/ml for toxin II in a 50 root growth inhibition bioassay that involves 96 hr incubation in 15 ml solutions. My preparation of the same compound was considerably more active in a root growth inhibition assay that used 10 ml solutions and 72 hr incubations (Fig 10). My preparation of toxin II was tested simultaneously under both assay conditions, and gave nearly the same specific activity in both assays (Table 1). Toxin II gave an ED of 0.4 ug/ml in the 72 hr assay using 50 10 ml solutions; an ED50 of about 0.3 ug/ml was observed with the assay conditions of Kim gt al. (6,19) (Table 1). 37 90- 70 " Susceptible c 50 " .2 I-E ii 30 4 o k A 10 . Resistant I 0 OJ 02 (L4 L0 2L) .50 100 Toxin concentration (pg .ml -1) Figure 10. Effect of HC toxin II on seedling root growth. Susceptible (O) and resistant (I) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions for 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water. 38 90-1 Susceptible 7o- : 50‘ 3 1 i 30 be 10-1 1 0 Figure 11. of each seedling after the incubation time. A Resistant A A 0.5 10 2.0 5.0 10.0 20.0 Toxin concentration (pg Jul '1) Effect of HC toxin III on seeling root growth. Susceptible (O) and resistant (A) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions 72 hr. was calculated for each treatment based on the longest root An average Percent inhibition is relative to control seedlings in water. 39 Table 1. Inhibition of susceptible seedling root growth by HG toxin II as determined by two different assays. s Inhibitiona HC Toxin II b c (ug/ml) 72 hr assay 96 hr assay 10.0 84 86 5.0 82 84 2.0 76 86 1.0 71 80 0.4 50 59 0.2 39 46 0.1 8 14 aTen seedlings (5 per duplicate plate) were incubated in each treatment. Percent inhibition is relative to appropriate water control. bIncubation was in water or in toxin (10 ml per duplicate plate) for 72 hr. This is my standard assay. cIncubation was in water or in toxin (15 ml per duplicate plate) for 96 hr. This is the assay of Kim gt al. (15). no Yields of toxin II were generally 1.6 to 1.7 times greater than the yields for toxin III; yields of each were less than 5% of those for crystalline toxin I. Another isolate of 5. carbonum race 1 was examined and was found to produce the three forms of toxin in the same ratios as the original isolate. Peak RP2 (Fig 8) was resolved into several components when chromatographed on Partisil 5; many of these components gave a positive reaction with the epoxide indicator on TLC. The retention time of the major component was 17 min; TLC indicated the compound, which had an Rf of .33, possessed an epoxide. Root growth inhibition bioassays indicated host-selective toxicity, and the compound was designated HC toxin IV. To use the HPLC more efficiently, peak RP2 was subjected to flash chromatography on a column (1.0 X 15 cm) of silica prior to further purification of toxin IV by HPLC. Toxin IV-containing fractions were identified by TLC. Final purification was by HPLC (Fig 12). The compound gave an ED of 20 ug/ml 50 against susceptible seedlings in the root growth inhbition assay; resistant seedlings were unaffected by toxin concentrations up to 100 ug/ml, the higherst concentration assayed (Fig 13). Structure gt Ht Toxin t! Spectral data were collected for toxin IV. High resolution FAB MS established the empirical formula as 021H3406N4’ or tox1n I plus two hydrogen atoms (calculated 41 Detector Response AJ 1 l O 10 20 Minutes Figure 12. Final purification of HC toxin IV by HPLC. Chromatography was on a column (0.4 x 25 cm) of Partisil 5. Elution was with hexane/ethanol (93/7, v:v) at 3.0 ml/min. 42 9O 5 7o '3 LE 50 E O k 30 1o 0 2 5 1O 20 50100 Toxin concentration (vs - ml '1) Figure 13. Effect of HC toxin IV on growth of seedling roots. Susceptible (O) and resistant (I) seedlings (5 per duplicate 9-cm Petri dish containing 10 ml solution) were incubated in water or in toxin solutions 72 hr. An average was calculated for each treatment based on the longest root of each seedling after the incubation time. Percent inhibition is relative to control seedlings in water. 43 C21H3506N4 439.2556, found 439.2561, +1.0 ppm). Amino acid analysis indicated alanine and proline in a molar ratio of 2:1 (identical with toxin I) suggesting that the two additional hydrogens were on the epoxide-containing amino acid. Allowing for an intact epoxide, as indicated by reactivity with NBP, the data were best explained by a hydroxylated carbon rather than a carbonyl adjacent to the epoxide (Fig 14). 13C NMR spectra for toxins I (Fig 3) and IV (Fig 15) were very similar. Assignments of carbon resonances were made with assistance of published spectra for toxins I, II, and III (32,47), and are presented in Table 2. The spectrum for toxin I indicated 21 total carbons with carbonyl carbons at 171.2, 173.1, 173.5, and 173.6 ppm, assigned to each of the four amino acids, and at 207.3 ppm, in accordance with published data assigned to carbonri of aoe (32,47). Toxin IV contained 21 carbons, agreeing with the calculated empirical formula, and showed only the four amino acid carbonyls; the resonance at 207.3 ppm was absent and a new signal was observed at 71.5 ppm. This is consistent with hydroxylation of the carbon adjacent to the epoxide, therefore, the resonance at 71.5 ppm was assigned to that carbon (carbon n). No other 13 differences were observed between the two C-NMR spectra (Table 2). 1H NMR data for toxins I and IV were also very similar (Fig 2 and 16). The only difference was in protons assigned to the epoxide. Toxin I appears as three double 44 o H CH C HC 3/ \ \ cfio CH / \ HN NH \ ’xc “c/ (CH ) 2-21- CH C) \\~CJ* <:/// I \ I ‘2 N C52 1 \o H C , 2 CH2 Figure 14. Proposed structure for HC toxin IV. 45 Figure 15. 13C NMR spectrum of HC toxin IV. Spectrum was collected on a Bruder WM-250 spectrometer at 68.9 MHz. Chemical shifts are relative to internal CHCl . 3 46 47 13 Assignments of C resonances for HC toxins I and IV. Amino Acid Carbon Toxin I Toxin IV Alal o 45.9 45.1 6 14.0 14.1 Ala2 O 47.3 47.4 B 14.6 14.6 Pro a 57.7 57.8 3 28.6 29.2 Y 22.6 24.9 6 48.0 48.1 Epoxide- o 51.8 52.0 containing 8 29.0 29.2 Amino Acid Y 24.8 25.0 6 24.9 25.1 E 25.4 25.6 C 36.1 34.2 n 207.3 71.5 8 46.9 47.1 1 53.2 55.3 Amide C=O 173.6 173.6 173.5 173.5 173.1 173.2 173.3 171.4 us Figure 16. 1H NMR spectrum of HC toxin IV. Spectrum was collected on a Bruder WM-250 spectrometer at 250 MHz. Chemical shifts are relative to internal CHC13. 49 CLL m a. m. a~ m~ an mm a: m: an mm am new 9 leJJljiljjjjjjjlel 141 (ll .1 . J 424 ‘47:.) a