' V113 71,! 1, 1 3;, * i!- A“. gs 4% 4—. *7." ' " h' 7':- “In...“ . m «oz-:53. m _ ‘. a I." Afr-u - "'7‘. , n - _ 4.- 66—2 "5’ “vsfia- ‘ ”- ”p...” ' w W . . N :- Tu ~J‘ ’- c, . 0 ---—~ -W '. -w I ,._,. ’“v—w—z - fi‘ *' - --...._. ‘ ."4— i 5’ _-_.. ’ ..~._' r» M4 -f- A Z—-‘— A ’L-.. "1 “— . o ' v}. ' - - ._. . .-. - g 1 .,_'_::- ‘ ‘7... .4 _ ‘ fi'.;- Ah; ‘ . a.’.¢(~ J‘— ‘11-..11 l‘ ”fat-VI |‘1”‘x ,{j‘t‘ ‘7‘.” . 2.1».-. ._ u r 1' ~' I‘.._‘- ‘ 13-..:7. 1 '15" 3 [(1 fl“ 1 1.1!“ 111. 1. “is???“ ul‘JIl‘lIfi’l‘I 121911" ‘ 1 11111 .. I :i‘j‘!‘ :ljd ‘114'11;"f_ ‘1‘? Eth'i'“ F1}, I ‘11 r , h: ‘ ‘ . 0 ”1‘. . .. . . . . .‘O . ' . W‘“ 5" . . -. '2er- . .-........ ..._ . ‘ . D . “11‘ {“11 N55: .1: .13. z ‘1“. I 1* =21; 1.111., .;‘ ' “11%.“; \‘M‘ I "1.13 {1116\1‘1111 1.1}; ‘31.; .. I 1: flag-11‘ U ,E‘ttz’n‘" ‘1‘ 1? .vp ‘1 "‘mfllg 11.." ".1 "A. '| '2“ .5 1 a. 1 "I “hq‘l’klw‘ 2;;1 3'1; 1.1I . 1:1“ 1151'” I‘ I‘ll 1 fl":'1 wk ‘13.“:- ‘3'“... :14 ‘1 I1 +19... '1'. \ ‘41.? l]. L 1 ‘ 111‘. vii" I I ‘ I‘M 1 1111 ”1'11 »» 1 111 21.111111! M 13111111111111.11fl‘111m ”WWW?“ 1111.11.31. " ' THESTS This is to certify that the dissertation entitled SELECTIVE TOXINS AND ANALOGS PRODUCED BY HELMINTHOSPORIUM SACCHARI: PRODUCTION, ISOLATION, CHARACTERIZATION, AND BIOLOGICAL ACTIVITY presented by ROBERT STANLEY LIVINGSTON has been accepted towards fulfillment of the requirements for H. D . degree in BbTANV IQNQI iaklflft’ iaifikc102)/ Major proTessor 7 {a MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 Datelkeo 2, [583 i L?” ”TRY .: r1: .: mm 1 U assay . ' 'bvifSIbl RETURNING MATERIALS: Place in book drop to LIBRARIES remove this Checkout from ”- your record. F__I______NES will be charged if book is returned after the date stamped below. ROON L532 Gui T M (“WW-113*“; “r“, '._,,'." ' 7" ‘QJJ .1 a u f'fiiJ-udfi ‘ la: 1 l" l i D SELECTIVE TOXINS AND ANALOGS PRODUCED BY HELMINTHOSPORIUM SACCHARI: PRODUCTION, ISOLATION, CHARACTERIZATION, AND BIOLOGICAL ACTIVITY By Robert Stanley Livingston 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 1983 ABSTRACT SELECTIVE TDXINS AND ANALOGS PRODUCED BY HELMINTHOSPDRIUM SACCHARI: PRODUCTION, ISOLATION, CHARACTERIZATION, AND BIOLOGICAL ACTIVITY 83! Robert Stanley Livingston A method was developed to isolate the host-selective toxins and toxin analogs from cultures of Helminthosporium sacchari. The procedure include the use of activated charcoal, ion exchange, gel, and flash chromatography, plus reverse phase HPLC. HS toxin was characterized in part by NMR, MS, derivatization and degradative chemical techniques. A structure proposed by other workers was shown to be wrong. Toxins contain two chains of 3-1.5 linked galactofuranose units attached to an unsymmetrical sesquiterpene. The three forms of toxin (A, B, and C) differed in relative abilities to induce electrolyte loss from susceptible sugarcane tissues. In cultures of the fungus, toxin concentration peaked at three weeks, followed by a rapid decline. 5h sacchari was found to produce a B-galactofuranosidase which removes galactose units from toxin, thus producing lower molecular weight analogs (toxoids). Twenty one different toxoids were produced by sequential removal of galactose from the three forms of toxin. Each of the three toxins and six of the toxoids with three galactose units were partially digested with enzyme; the resulting toxoids were separated by HPLC and the arrangement of galactose units was determined. Finally the biological activities of the toxoids were determined. One of the toxoids with three galactose units proved to be toxic to some Robert Stanley Livingston but not all fl. sacchari-susceptible clones of sugarcane. This toxoid was as toxic to certain sensitive sugarcane clones as was the lost active form of the toxin (four galactose units). The other toxoids were non-toxic on all other tested clones of sugarcane. All isomers of the toxoids with three galactose units gave protection against action of the toxin; the three galactose toxoids were more effective than were the toxoids with two galactose units. Toxoids with only one unit of galactose provided very little protection. Other experiments showed that the sesquiterpene isomer, the number of galactose units, and the arrangement of galactose units in the molecule all play a significant role in determining toxicity (as determined by induction of electrolyte loss) and relative ability to counteract the effects of toxin on sugarcane tissue. ACKNOWLEDGEMENTS I would like to thank Robert Scheffer, Keisuke Kohmoto, Mark Lesney and Steven Briggs for their support, guidance and friendship. I also thank Drs. Steven Tanis, Ray Hammerschmidt, Robert Bandurski and Harry Murakishi for being members of my graduate committee. 11 TABLE OF CONTENTS Page LIST OF TABLES...................................................... iv LIST OF FIGURES..................................................... vi GENERAL INTRODUCTION................................................ 1 LITERATURE CITED.................................................... 4 EXPERIMENTAL 1. Isolation and Characterization of Host-Selective Toxin from Helminthosporium sacchari....................... 5 II. Conversion of Helminthosporium sacchari Toxin to Toxoids by s-galactofuranosidase from HelminthosporiumOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.00.....0... 12 III. Toxic and Protective Effects of Analogs of Helminthosporium sacchari Toxin on Sugarcane Tissues...‘0.0000000000000000000000000.0.0.0000...0.0...0.. 18 IV. Selective Toxins and Analogs Produced by Helminthosporium sacchari: Production, Isolation, Characterization, and Biological Activity.................. 40 GENERAL DISCUSSIONOOOO0.0..0....OOOOOOOOOOOOOOOOOOOO0.00.00.00.00... 75 LITERATURE CITEDOCOOOOO0.0.0.0000...IO...O0.0.0....OOOOOOOOOOOOOOOOO 79 iii Table LIST OF TABLES Page EXPERIMENTAL I Proton NMR spectrum of toxin from Helminthosporium sacchari: shift position and number of protons for eaCh peakOOOOOOOOOOOOOOOO0.0.0....OOOOOOOOOOOOOOOOOOOOOOOO... 8 Solvents used in thin-layer chromatography of toxin from Helminthosporium saCChariO0.0000000000000.00000000000000 11 Chromatographic behavior of Helminthosporium sacchari toxin and related non-toxic substances (noxins). Preliminary processing included concentration of culture filtrates, followed by adsorption on Norit A, elution with 50% ethanol containing 1% ammonium hydroxide, and methanol precipitation........................ 11 EXPERIMENTAL II Galactose released by acid hydrolysis of purified toxin and toxoids (50 ug each). Galactose was quantified by a galactose oxidase/peroxidase assay........... 14 s-Galactofuranosidase activity in culture filtrates and fungal mats of Helminthosporium and other species. The fungi were grown in still culture (20 ml modified Fries solution in 125-ml flasks) for 23 d at 22°C ........... 16 EXPERIMENTAL III Effect of pretreatment with toxoid III on HS toxin- induced losses of electrolytes from sugarcane leaves (clone Co 453). Leaf discs were cut, preincubated in H20, and held in toxoid solutions (2 ml, 100 ug toxoid/ml) (B, C) or water (A). Tissues were then rinsed (C) or not rinsed (B), toxin solutions were added, incubated for 1 h, rinsed, transferred to 5 ml water, and leached for 3 h................................... 26 Effect of pretreatment with toxoids on HS toxin- induced losses of electrolytes from sugarcane tissue (clone NG 77-103). Toxin was used at 1.25 ug/ml............. 30 iv Table Page Protective effects of toxoids against activity of HS toxin, plus tox1c effects of toxoid III against certain clones. Protection was measured by effects of pretreatment on toxin-induced loss of electrolytes........... 33 EXPERIMENTAL IV Relative amounts of toxoids obtained from partial digestion of each of the six isomers of toxoid III, as determined by peak heights. Toxoid III was digested and the resulting toxoids were separated by HPLC and identified by their retention times. The galactose arrangement for the six isomers of toxoid III were determined by Macko (10)..................................... 57 Electrolyte leakage from sugarcane clone NG 77-234 induced by toxins (A2 2 and C22 ) and toxoid A1 2- Leaf disks weré exposed’ to test solutions for 0.5 hr, and rinsed thoroughly. Conductance values were taken after 3 hr incubation in leaching solution........ 63 Comparative abilities of three natural mixtures of toxoid isomers to protect sugarcane tissues (Clone NG 77-234) against toxin C. Toxoid I is a natural mixture of isomers with one galactose1 unit. Toxoid II' is a natural mixture of the isomers A1 and C11. Toxoid II is the natural mixture 816} isomers1 with galactose arranged 2,0 and 0,2. Leaf disks of clone N0 77- 234 were exposed to toxin and toxoids simultaneously for 0.5 hr, rinsed, and incubated in the leaching solution for 3 hr, when conductance was determined................................... 64 Comparative abilities of the six isomers of toxoid III to protect sugarcane tissues (clones NG 77-103 and NG 77-234) against toxin C. The treating solution contained toxin C at 0.75 pm without (none) or with toxoids at 50 pm. The standard protection bioassay was used........................ 66 Comparative abilities of toxoids containing sesquiterpene C in protecting sugarcane tissues (clones Co 453 and NG 77-234) against toxin C. The treating solution contained toxin C at 1.0 um without or with toxoid at the indicated concentrations. The standard protection bioassay was used........................ 67 Comparative abilities of four isomers of toxoid II to protect sugarcane tissues (clones NG 77-103, NG 77-234 and Co 453) against toxin C. The treating solution contained toxin C at 0.75 pm (for clones NG 77-103 and NG 77-234) and at 1.0 um (for cone Co 453), without (none) or with tox01ds at 50 pm. The standard protection bioassay was used............................................ 68 LIST OF FIGURES Figure Page EXPERIMENTAL I 1 Gas-liquid chromatogram of the chloroform-soluble hydrolytic products of toxin from Helminthosporium sacchari. The four major products had retention times of 2.53, 2.85, 5.95, and 6.47 min. Several other products were present in smaller amounts. All products had an apparent molecular ion in the mass spectrum at m/e 218; high resolution peak matching indicated an empirical formula of C15Hz10H................................ 7 2 Mass spectrum of toxin from Helminthosporium sacchari. Host-selective toxin (5 ug) was inserted into the source (temperature 200°C) by direct probe and ionized by electron ionization (35 ev). The probe was heated from 25 to 280°C (30°C/min); toxin was emitted at 280°C. Ions were monitored from m/e 29 to 500 on a Hewlett-Packard 5985 A mass spectrometer..................... 7 3 Proton NMR spectrum of toxin at 180 MHz. Host-selective toxin (10 mg) from Helminthosporium sacchari was dried under reduced pressure and rinsed several times with 010 to remove H20. The sample was dissolved in 0.5 ml 0 020 (99.7% deuterium) in a 5-mm NMR tube. The Fourier transformed spectrum was obtained from 25 transients on a Bruker WH-180 instrument. The chemical shifts in ppm for the major peaks are given in Table 1......................... 8 4 13C NMR spectrum of toxin at 45.2 MHz. Host-selective toxin (approximately 90 mg) from Helminthosporium sacchari was dissolved in 020 in a 10 mm NMR tube. The Fourier- transformed spectrum was obtained from 17,034 transients. The sodium salt of 2,2-dimethyl-2- silapentane-5-sulfonic acid was used as an external reference....................... 8 5 Mass spectrum of one of the chloroform-soluble hydrolytic products (C-15d) of toxin from Helminthosporium sacchari. Sample was introduced into the source (temp. 200°C) by gas chromatography and ionized by electron ionization (70 ev). The gas chromatography conditions are described in the text. Mass spectra of these compounds (C-15a, b, c, and d) obtained by chemical ionization with methane confirmed the molecular ions as m/e 218; high resolution peak matching gave the empirical formula at C15H210H................................ 11 vi Figure Page EXPERIMENTAL II Concentrations of toxin (0), toxoid 111 (n), toxoid II (0), and toxoid I (A) in culture filtrates of H. sacchari, as determined by quantitative GLC........TI........ 14 B-Galactofuranosidase activities in culture fluids (C)) and nwcelium (O) of H. sacchari. over a 6-week period. Cultures were grown in Roux bottles, each containing 200 ml of medium............................................. 15 GLC of products resulting from enzyme hydrolysis of HS toxin. Products are galactose (gal) and toxoids I, II, and III (1, 2, and 3, respectively). Determinations were made with trimethylsilyl derivatives, using a temperature regime of 130°C for 1.0 min followed by increases of 30°C/min up to 350°C, which was held for 5.0 min. A, Substrate control; B, products obtained when the enzyme preparation was diluted 1:4; C, products when enzyme preparation was diluted 1:2; 0, products when enzyme preparation was not diluted.................................. 15 EXPERIMENTAL III Effect of preincubation in water on sensitivity of sugarcane (clone NG 77-82) leaf tissues to HS toxin, as determined by electrolyte leakage. Leaf disks were cut and held on water for the indicated times. Disks were then placed in toxin solution or in water for 0.5 h, rinsed, and placed in 5 ml leaching solution. Toxin-induced leakage was determined by conductance of the leaching solution at 3 h. Toxin was used at 0.15 ug/ml (G) and 1.5 ug/ml (.). Standard deviation is shown by the vertical bars................................... 24 Effect of HS toxin concentration on loss of electrolytes from leaf tissues of sugarcane clone NG 77-103. Leaf disks were cut, incubated on water for 4 h, exposed to toxin for 0.5 h, rinsed, and leached in 5 ml of water. The values are conductances of leaching solutions after 3 h, minus the water control. Standard deviation is shown by the vertical bars................................... 27 Effect of toxoid pretreatment on HS toxin-induced leakage of electrolytes from leaf tissues of sugarcane clone NG 77-103. Leaf disks were cut, preincubated for 3 h, and exposed to toxoids I ( , II (563,) or III (I) for 1 h. Toxin was then added for a final concentration of 1.25 ug/ml; disks were then incubated for 30 min, rinsed, and leached for 3 h in 5 ml of water. Toxin-induced leakage is indicated by the conductance values. Each treatment was done in triplicate and the standard deviation is Similar to that Observed in Fig. 2.00....OOOOOOOOOOOOOOOOOOOO 28 vii Figure Page 4 Dosage-response relationship of HS toxin on leaf disks of sugarcane clone Co 453. Electrolyte leakage is indicated by conductance values of leaching solutions. Procedures are described in Fig. 2. Standard deviations are shown. Tissues used in this experiment were in an unusually sensitive condition, because of environmental conditions during growth............ 31 5 Effect of toxoid pretreatment on HS toxin-induced leakage of electrolytes from leaf tissues of sugarcane clone Co 453. Toxoids I (E ), II (fii), and III (II) were used. Toxin concentration was 1.25 ug/ml. Other procedures are described in Fig. 3........................... 32 EXPERIMENTAL IV 1 Elution pattern of HS toxin (0) and toxoids III (0), II' (I), II (A), and I (D) from a Sephadex LH-20 column (80x3 cm). The eluent was 50% methanol and 7 ml fractions were collected. An aliquot of each fraction was derivatized and the toxin and toxoids present were determined by GLC (5). The height of each peak indicated amounts......................... ..... .... 48 2 Elution of the three forms (A, B, and C) of HS toxin from reverse phase HPLC following application of 10 ug each to the column. The eluent was 20% acetonitrile in water and the flow rate was 2 ml/min. Absorption at 215 nm was monitored........................... 50 3 Electrolyte leakage induced by each of the three forms of HS toxin. Leaf disks of sugarcane clone NG 77-234 were exposed to toxin for 0.5 hr, rinsed, and incubated in 5 ml of water. Conductance of the leaching solution was measured after 3 hr.................... 52 4 Toxoids produced from toxin C by enzymatic removal of galactose units. The sesquiterpene in toxin A [ CIJY, ] and B [ CIDY. ] differs in arrangement of a double bond (11). Toxins A and 8 each produce a set of toxoids comparable in galactose arrangements to those shown above......................................... 53 5 HPLC separation of toxoids resulting from partial digestion of each of the 3 forms of HS toxin. The e—galactofuranosidase produced by E. charlesii was used for digestion. The letter in each panel indicates the sesquiterpene isomer present in the toxin and toxoids. Each peak is labeled to indicate the number and arrangement of galactose in each compound. HPLC conditions are described in material and methods.................................................. 55 viii Figure Page Bioassay of toxoids resulting from partial digestion of toxin A. The toxoids and undigested toxin were separated by HPLC and fractions (1 ml) were collected. Aliquots (10 ul) of the fractions were assayed with sugarcane clone Co 453 (O) and NG 77-234 (0). (a) Eluate from digested preparation. (b) A preparation identical to that chromatographed in (a) was partially digested and the resulting toxoids were separated by HPLC. The bars (i.e., A2 2) indicate the peaks that were present in each chromatogram. The solvent was 20% acetonitrile in water and the flow rate was 2 ml/min......... 60 ix GENERAL INTRODUCTION The disease of sugarcane known as "eyespot", caused by Helminthosporium sacchari, became severe in Hawaii, Puerto Rico, Java, and elsewhere when certain new cultivars were planted on a large scale (5). The major symptoms of the disease include "eyespot" lesions on leaves, with subsequent development of a necrotic lesion which can advance to the leaf tip. The fungus is easily isolated from the eyespot lesions, but usually is absent from the "runner" lesion above the eyespot. This situation was interpreted by Lee and other early workers to mean that a toxin, produced by the fungus in the eyespot, is transported up the leaf. Lee made some detailed early studies but came to no conclusions regarding the toxin (6). Lee's preparations appeared to be slightly more toxic to the susceptible than to the resistant clone of sugarcane, but his data were inconclusive. He suggested that the toxic component in culture filtrates of H. sacchari was a nitrite. Conclusive evidence for the presence of a toxin was presented in 1971 (8). The toxin was partially purified by a procedure involving solvent extraction and gel permeation chromatography. Many clones of sugarcane were tested for their reaction to the pathogen and to the toxin; those clones resistant to the pathogen were insensitive to the toxin and clones susceptible to the pathogen were sensitive to the toxin. Toxin was shown to be a small molecule which, when partially purified, could reproduce some of the symptoms of natural infection. Many papers on HS toxin have been published since it was first shown that filtrates of H. sacchari are selectively toxic to susceptible sugarcane. In fact, the disease and the toxin involved have become an important model case for studies on the molecular basis of plant disease development. In 1971 Steiner and Strobel published a paper containing a proposed structure for the toxin which was given the name helminthosporoside (9). The proposed structure was 2-hydroxycyclopropyl- a-D-galactopyranoside. A series of papers followed, from which was developed a theory which explained the high degree of Specificity and the mode of toxicity of the toxin (11). In 1975 a summary of this work was presented in "Scientific American" (12). The theory involved a receptor or toxin-binding protein in sensitive tissue; resistant tissue was said to contain a similar protein which did not bind toxin. Toxin binding caused a change in the conformation of the receptor and the surrounding lipids of the membrane. These changes induced the activation of a KT-Mg++ATPase. The result was an unbalanced flow of ions in the cell, a breakdown in normal cell functions, and eventual death of the cell. Several subsequent papers have been published which support and yet always change the details of the theory (2,3). However, a critical analysis of the work of Strobel et al. was published (1), and attempts in our laboratory to repeat the critical experiments were unsuccessful (7). My initial work was directed toward obtaining a purified preparation of the toxin produced by H. sacchari (HS toxin). The purified toxin was then characterized (in part) by spectral and chemical techniques; the original characterization by Steiner and Strobel (9) was shown to be wrong. These aspects of my work are presented as dissertation section I, which was published with the title "Isolation and characterization of host-selective toxin from Helminthosporium sacchari". During the course of this work, an enzyme was discovered which hydrolyzes the toxin molecule, freeing galactose units. This work is described in section II of the dissertation, which was published with the title "Conversion of Helminthosporium sacchari toxin to toxoids by B-galactofuranosidase from Helminthosporium“. Next, analogs of the toxic compound were discovered, these were characterized and shown to be inhibitors of the toxic effect of HS toxin. This part of the work was prepared as a manuscript for publication, and is given as section III of the dissertation; it was accepted for publication with the title "Toxic and protective effects of analogs of Helminthosporium sacchari toxin on sugarcane tissues". Finally, twenty-one non-toxic analogs of HS toxin were isolated and their protective effects against toxin were determined. One of the analogs proved to be as toxic as is HS toxin, but only to certain clones of sugarcane. This part of the research was described in a manuscript that was prepared for publication; it is presented as section IV of the dissertation. 10. 11. 12. LITERATURE CITED Daly JM 1981 In "Toxins in plant disease", Durbin RD (ed.) Academic Press, NY pp 515 Kenfield 08, GA Strobel 1981 Selective modulation of helmintho- sporoside binding by alpha-galactoside-binding proteins from sugarcane clones susceptible and resistant to Helminthosporium sacchari. Physiol. Plant Path. 19:145-152 Kenfield 05, GA Strobel 1981 The status of a toxin receptor in sugarcane leaves in relationship to toxin sensitivity and leaf age. Biochemistry International 2:249-255 Kono Y, Hw Knoche, JM Daly 1981 In "Toxins in plant disease" Durbin RD (ed.) Academic Press, NY pp 515 Lee A 1926 The history and distribution of eye spot. Hawaiian Planter's Record 30:466-492 Lee A 1929 The toxic substance produced by the eye-spot fungus of sugarcane, Helminthosporium sacchari Butler. Plant Physiol 4:193-212 Lesney MS, RS Livingston, RP Scheffer 1981 Effects of toxin from Helminthosporium sacchari on nongreen tissues and a reexamination of toxin binding. Phytopath. 72:844-849 Steiner Gw, RS Byther 1971 Partial characterization and use of a host-specific toxin from Helminthosporium sacchari. Phytopath. 61:691-695 Steiner cw, GA Strobel 1971 Helminthosporoside, a host-specific toxin from Helminthosporium sacchari. The Journal of Biological Chemistry 246:4350-4357 Strobel GA 1973 The helminthosporoside-binding protein of sugarcane. The Journal of Biological Chemistry 248:1321-1328 Strobel GA 1976 Phytotoxins as tools in studying plant disease resistance. Trends in Biochemical Science 1:247-250. Strobel GA 1975 A mechanism of disease resistance in plants. Scientific American 232:80-88 EXPERIMENTAL I ISOLATION AND CHARACTERIZATION OF HOST-SELECTIVE TOXIN FROM HELMINTHOSPORIUM SACCHARI: '1er Jouamu. or 810mm Columns 256.1% 4 murmurs. pp 1705-1710 1961 War U..SA Isolation and Characterization of Host-selective Toxin from Helminthosporium sacchari * (Received («publication June 26. 1980. and in revised form. October 31. 19801 Robert s. Livingston and Robert P. Scheffer From the Department of Botany and Plant Pathology, Michigan State University, East Lansing, Michigan 48824 Helminthosporium sacchari infects certain clones of sugarcaneandproducesatoxinwiththesameplant selectivity as the fungus itself. The toxin was purified by use of activated charcoal plus thin layer, gel, and ion exchange chromatography. Gas chromatography (60) of a trimethylsilyl derivative of toxin gave a single peak. Toxin was characterised by GC, mass spectros- copy (MS). and NMR spectroscopy. The spectra of hy- drolytic products showed that toxin contains galactose plus a Cull" moiety which appears to be a sesquiter- pene. Spectral data and methylation procedures showed that toxin contains an oligosaccharide com- posed of B, 1 —. 5 galactofuranose units (probably 6 units). Several interconverdble forms of the Can moi- ety were evident after acid hydrolysis. Toxin was sep- arated from 8 closely related, nontoxic compounds C‘noxins”), which contained galactose plus the Cull” moiety. Comparative data show that the toxin exam- inedinthisstudyisthesameasthetoxindescribedby Steiner and Strobel (Steiner. G. W” and Strobel. G. A. (1971) J. Biol. Chem. 246, 4350-4357). The data also show that the previously proposed structure is incor. rect. At least 15 plant-infecting fungi are now known to produce substances with selective toxicity against msceptible hosts. Such toxins are not active against non-host species, and against host genotypes thatareredstanttothe fungus. Several of these “host-selective toxins" have been isolated and par tially characterised (6). However, only the toxin from Alter- naria mali affecting certain apple cultivars has been charac- terised completely (4), the structure confirmed (8). and the molecule synthesized (2). A. mali toxin is a cyclic depaipeptide with a 24,445. Helminthosporiwn sacchari (Van Breda de Haan) Butler selectively parasitizes some cultivars (clones) of sugar cane, causing a disease known as “eyespot" Several years ago, the fungus was shown to produce a toxin with selective eflects which matched those of the fungus Steiner and Strobel (7) isolated the toxic compound and characterized it as 2 hydrox- ycyclopropyl-a—nogalactopyranoside (trivial name. helmin- thosporoside) The proposed structure has not been con- firmedNeverthelmthisandotherworkonH. sacchari 'This work was supported by Grant PCM- 7611916 from the Na- tionalSciencePoundation. PartoftheworkwasdoneintheMichigan State University National Institutes of Health mass spectroscopy facility which is supported in part by United States Public Health Service Grant KIT-(D480. Published as Journal article No. 9684 ofthe Michigan Agricultural Experiment Station. The costs of publication of this article were defrayed in part by the payment of page charges. Th'n article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. toxin is often cited in discussions of the molecular basis of disease development and disease resistance in plants ( 1). We have re-examined the toxin from H. sacchari. Charac- terization is not yet complete. but we feel that the data should be published because of the importance of the work (7) and the controversies involved (10). An abstraCt describing some of our work was published (3). MATERIALS AND umons’ RESULTS Water-soluble Hydrolytic Products of Toxin —The aqueous phase of acid-hydrolyzed toxin was chromatographed on thin layer plates, using several different solvent systems, with diphenylaminezanilinezphosphoric acid as the indicator re- agent. The R, values and color reactions of the resulting spots matched those of galactose standards. The water-soluble frac- tion was then derivatized with 'I‘ri-Sil-Z and subjected to 60 MS’, using columns containing several different liquid phases. Retention times for peaks from gas chromatography matched those of derivatized galactose (a. If. and y forms); mass spectra confirmed the presence of galactose but there was no indica- tion of other sugars. Chloroform-soluble Hydrolyu'c Products of Toxin—The chloroform phase of hydrolyzed toxin was subjected to gas chromatography, using a column (1.8 m) packed with OV-l (3%) and a temperature of 170°C. Four major peaks and several minor peaks were observed (Fig. 1). Each major peak was later characterized by MS as a 15-carbon compound; for convenience, they are identified as 01511. C-15b. C-15c, and C-15d (Fig. 1). Enriched preparations of the four major C-15 products were made by TLC followed by chromatography with an LEI-20 column (see “Materials and Methods"). Possible interconversion of the 015 products was consid- ered. Aliquots of each of the 4 major C-15 products in aqueous trifluoroacetic acid (0.1. 0.05, and 0.01 is) were held at 95°C for 2.5 h. The solutions were then extracted with 3 equal volumes of chloroform and the combined extracts were sub- jected to 60. Results showed that each of the major C-15 poducts gave at least trace quantities of the others. For example, when C-15c was exposed to acid at 0.05 it. GC ' Portions of this paper (including “Materials and Methods.” some of the “Results," Fig. 5, and Tables II and III) are presented in miniprint at the end of this paper. Minith E easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry. 9650 Rockville Pike. Be- thesda. Md. 20014. Request Document No. SON-1305. cite authoris). and include a check or money order for $4.“) per set of photocopies M site photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press. ' The abbrevrations used are: GC, gas chromatography: MS. mass spectroscopy. Meso. dimethyl sulfoxide; MesSi. trimethylsilyl; TLC, thin-layer chromatography. El/D, electron ionization subsequent to thermal desorption from field emitters. 1705 1706 b C ° .1 2 4. 6 8 l0 minutes Pro.1.Gas-liquldchromatogramoftheoliloroform-sduble productsoftoxinfromfl ' sacchari. The four major products had retention times of 2.53, 2.85. 5.95, and 6.47 min. Several otherproductswerepresentinamslleramoimts. All psoductshadanapparentmolscularioninthemamspectrumatm/ e218; highrmolufioapsakmatchingindicaiedanempiricalformula 0‘ 01514310“. showed the presence of C-15d, plus two products with reten~ tion times very similar to those of 0.1511 and -b, plus two other products (retention times. 4.22 and 4.58 min). These data indicate that the C-15 products are imstable and are intercon- vertihle. Available data do not establish which form of the C- 15moietyisinthetoxinmolecule;indeed,toxinmightsa'nt as homers based on difi’erent forms of the C-15 moiety. Mass Spectroscopy—The following conditions were mod for intact toxin, using the Hewlett-Packard monument with the direct probe electron impact method: source temperature was ZXPC. probe was heated from ambient to 280°C. ioniaa~ tion voltage was 35 eV. and ions were monitored from rule 35 to 500. The spectrum (Fig. 2) showed the highest visible mam ion at 380. The Varian CH-5 mass spectrometer was used for high resolution peak matching. ionisation was by electron irnpsct with accelerating voltage at 70 eV. Peak matching with title 380 indicated that the most probable empirical formula was Criaasos- The low resolution spectrum had a peak at m/e 201, (Can) which appears to be Cad-1.0. minus galactose. High resolution peak matching on the rule 201 confirmedthisempiricalformula Therewasathirdpeakat iii/e 217; peak matching indicated that this was CanO, a carbon-hydrogen compound plus oxygen from the galactoaide linkage.Apeakatm/e259waspredicted;thisshouldbe 0113902 (Crane: 4’ CsHeOr -’ Cirflasos. 01 I 6"“th unit plin a portion of galactose, a known break for the galac- toside linkage).T'helowresolutionspectrurnhadtheexpectod peakatm/e259,andpeakmatchingconfirmedthepredicted formula. IntacttoxinwasdholvedinDro(99.7%)tosxchange deuterhrmforthehydroxylprotomThemamspecmimof deuterium-labeled toxin should show an increase of one atomic mass unit/exchangeable proton. Results showed that m/e 330 was shifted to m/e 384, indicating four exchangeable protons. Related m/e values were shifted comparably. Again, Isolation and Characterization of Toxin from H. sacchari In- -1 ., 904? ‘ l ! I - 1:.- l 1 vvvvvvvvvvvv 50 100 ISO 5 200 250 300 350 Pro. 2. Massspectrumoftoxinfiomflsbnwhosportannsnb MHost-selectivetoxin(5ug)wasinsertsdintothesoiu~cs(tem- psrauue2m°Clbydirectprobeandionissdbyelchonionimtion (35 so). The probe was heated fixln 25 to 2N°C (N'C/min); toxin wasemittedatM‘CJonsweremonitoredfmmm/efitomona HewlettaPackardMAmauspectromstsr. thesedatsarecona’ntentforastructurocontaining galactose, plusaunitcontaining 15carbonswithhydrogen (Cd-11.0.4» Gui-1,, —. Gui-1,0,), with the 15-carbon unit attached to the galactoseatssinglepoaitionThiswouldlesvefourfree hydroxyl groups which would exchange protons for deuterium The mild conditions required for hydrolysis (with release of galactose and a 1&carbon unit from intact toxin). plus the MS data. Most a galactosidic linkage. Mamspectraformethylatedtoxinwereobtainedwiththe Varian CPI-5 spectrometer. using the EI/D method. Acceler- ation voltage was 1.0 kV, which gives increased sensitivity and lower accuracy ($1.0 rule) with mamas >1w0. Filament cur- rent was 18 mA, and ions were monitored up to m/e 1150, which is maximum for the spectrometer. The spectrum showed ion at m/e 1060 and 1093. indicating that toxin contains at least 5 galactose units plus a Catlin moiety. The four chloroform-soluble. 15-carbon hydrolytic products of toxin (C-l5, a-d) were characterised by GC-low resolution MS and by high resolution peak matching. Peak matching of the rule 201 fragment indicated that the empirical formula wasthesameasthstdeterminedfortheionatm/eZOI (Can) in the spectrum of intact toxin Spectra were collected in both electron impact and chemical ionization (methane) modes; these data indicated that 218 was the probable molec- ular mass for all 4 products. Peak matching of the rule 218 produced the empirical formula CulinOH. The C-15 products may be suquiterpeneotype compounds. with a single hydroxyl formed during hydrolysis, and with 5 points of unsaturation (double bonds or ring structures). All four C-15 hydrolytic poducta produced a similar fragmentation pattern (see Fig. 5 inminiprint),withvariationintherelativeab\mdanceofthe uvu-alfi'agmenuThisiridicatesthattheC-wcompoimds areveryaimilarinstructure,podblydifleringonlyinthe position of double bonds. The conch-ion '- supported by protonNhflldataontheC-wbreakdownpoductsuiven below). Mostionsinthemammectrinnoftoainwereahopressnt intheC-l5moiety.0nlysfewionswerefiomgalactose,which '3 not surprising because sugar moieties are known to give weak mam spectra. Puke for intact toxin at rule 43, 73, and 91canbeattributedtoboththegalactoseandtheC-15 fragment (overlapping ions were confirmed by high resolution ma-specuoscopy).Theionsatm/emand61aretypicalof galactoseandwsrenotfoundintbemecu'aoftheC-m moiety.TTiesesetsofiornwer-eemittedfromtheprobeatthe same high temperature. indicating that the toxin preparation is a single. large. relatively nonvolatile compound. ProtonNMRStudies—T‘heNMRspsctrumoftoxininDzo contains 14majorpeaksordistinctregions0'ig.3.TableI). Isolation and Characterization of Toxin from H. sacchari D” Freak-moral“!!! oftoxinatIIOmRLl-lost- selective toxin (10 mg) from Hebmnthosporimn sacchariwas dried under reduced pressureand rinaedseveral timeswith Mwnmove 14.10 TbesamplewasdissolvedinO.5mlofDr0(99.7%deuterium)in a5-mm NMR tube. The Fourier transformdapectrumwasobtained hom25u'ansientsonaBrukerWl-1180instrument.'1’nechemical shiftsinppmforthemajorpeahsaregiveninTablel. Tassel ProtonNMRspectrwno/toxinfiomflebninthomoriumsacchari: shiftpou’tionandmanberofprotomforeachpcak . Shift Number of . 4 Number of m- m A 5.41 1 n 3.74 —" B 5.77 ' 1 1 3.66 —‘ c 5.17 3 J 257 1 D 5.06 1 K 2. 0.1. . s E 4.95 1 L 1.64 3 F 4.10 —‘ M 1. 5- 12 7 C 4.03 —' N 0.85 3 ' See Fig. 5. ‘CalculatedfromtheHDOpeakat 4.74. ‘Determinedfromthesreaundereschpeak. ‘PesksF.G.l-l.andlmresentatotalof38protona The spread of the seven peaks in the 4.91 to 5.41 ppm region is too great to result from splitting of a single group of protons. Therefore. these peaks represent five groups of protons. One group at 5.17 ppm contains three identical protons; the other peaks in this region represent one proton each These peaks probably are from olefinic protons. Strong absorbsnce at 3.6 to 4.2 ppm represents approximately 35 to 40 protons on carbon with s bydroxyl group; most of this sbsorbance is from protons in galactose. This number of protons suggests as many as 5 or 6 galactose units/molecule of toxin. 3 anomeric protonsalsoabsorbinthelowfieldendofthe 3.6to4.2ppm region.Thepeaksat4.03ppm maybefrom galactoaeprotons. but protons on the C-15 moiety that are involved in the galactosidelinkagecouldalsoabsorbinthisregionhaingle C-15 unit/molecule of toxin probably has no more than two protonsabeorbinginthe4.03region;thisisnotenoughto account forthe large peak at 4.03 ppm. Anotherpo-ibility could be more than one 015 moiety/toxin molecule. Proton NMR studies of the hoisted C-15 moieties showed that they varied in the number of olefinic protons (2 to 4). There was a sharp singlet upfield. which represents an isolated methyl group. Theaharpsingletat4ppm wassmignedtothe two protons on the carbon with the hydroxyl. which probably '3 adjacent to a carbon with a double bond. The NMR spectrum of the Me,Si derivative of toxin. in deuterated chloroform, was dominated by absorption at 0.25 t00.05ppm.Thearesofth'nabaorptionwasproportionalto thenumberofprotonsonthoeeMaSigroupsthathsdre- 1707 placed each hydroxyl group on the original toxin. The two peaks at 1.64 and 0.85 ppm, each representing 3 protons. were used to determine the units of area/proton. To determine the number of Me;Si groups/toxin molecule. the number of pro- torn in the 0.25 to 0.05 region (125 to 150 protons) was divided by the number of protons/MeaSi group (9 protons). There were 14 to 17 Me.Si groups/molecule of toxin. indicating 5 galactose units. ”C NMR Studies—The dominant characteristic of the ”C NMR spectrum of intact toxin was the strong intensity of peaks from 63.9 to 85.5 ppm (Fig. 4). These peaks are from carbons in galactose; they are much more intense than the peaks given by the C-15 moiety. The difference in intensity of thegalactoseandtheC-lspeaksisgreatenoughtomggest that there are several galactose units per 015 moiety in the toxin molecule. The six to eight peaks from 116.7 to 151.2 ppm indicate six to eight olefinic carbons. These data indicate the presence of three double bonds and two ring structures per C- 15 fragment. suggesting a sesquiterpene. This is consistent with the empirical formula predicted by mass spectral peak matching data. The region of the spectrum in which aliphatic carbons absorb (18 to 48.6 ppm) contains 9 to 13 peaks. The carbon atom of the C-15 moiety which shares an oxygen with the oligosaccharide will absorb in the same region as does galactose (63.9 to 85.5 ppm). Thepeakst63.9ppmisverycloeetotheassignedshift position for the C-6 of a galactofuranoside (63.6 ppm). The shift position ofa 01 is characteristic for a and B anomu'ic forms of pyranosides and furanoaides. The lack of peaks at 101-104 ppm rules out a- and Bogalactopyranosides and so galsctofuranoaide (9). The peak at 109.8 ppm indicates a p. linked galactofuranoside (5). Mass Spectroscopy of Hebninthoaporoside—MS data on helminthosporoside from G. Strobel ( Montana State Univer- sity) weretakenforcompar'non withourpreparationoftoxin 14040120 ”9.5.33— Pro. 4. “cm oftoxinstltljmflsflost-eelective toxin(approximstely90mg)fiomfle ' ' sacchariwas dbolved in D40 in a 10 mm NMR tube. The Fourier-transformed mmobtsmadfromflmetramienu'fliesodhnnsaltofu- dimethyl-Z-eilapentane-b—ulfonicsddwnuadasanexternslreter- ence. 1708 from H. sacchari. First. the Hewlett-Packard 5985-A spec- trometer was used with a series of ion source temperatures and ionization voltages in the electron ionization mode. With the ion source temperature at 250°C and the ionization voltage at 70 eV. the man spectra for the two preparations were essentially the same. except for minor peaks; these spectra were similar to the published spectrum for helminthosporo- side ( 7). Next. the fragmentation conditions were altered to favor survival of higher mass ions: this was accomplished by using a lower ion source temperature (@0°C) and a lower ionisation voltage (35 eV). Under these conditions, the two preparations gave similar spectra. including the m/e 380 ion. However, Strobel’s preparation had an ion at m/ e 236 which was missing from all spectra of our preparations. This difier— ence is important because m/e 236 was considered to be the molecular ion of the toxin (7). Further compar'nons of Strobel's preparation with ours were by high resolution peak matching. using the Varian CH- 5 mam spectrometer. Ions resulting from fragments of the aglycone moiety of toxin were selected for examination. be. cause that unit (the C-15 compound) was not included in the previously proposed stucture ( 7). Thus. the empirical formulae were determined for the ions at m/e 145. 157, 201, 217, and 380. which were evident in the spectra of both preparations. The predicted empirical formulas for these ions were identical for both preparations. The data indicate that our toxic mole- cule is the same as the one reported elsewhere. and that the promd structure (7) must be revised. 0180083108 Our highly purified toxin contained no detectable contami- nants. as shown by thin layer chromatography and by gas chromatography of derivatives. Toxin purified by gas chro- matography was hydrolyzed and the hydrolytic products were analysed. The experiment confirmed that the products of hydrolysis (galactose and a 15-carbon compotmd) were de- rived from a single toxic molecule. A preparation of helminthosporoside. kindly supplied by G. Strobel, was compared with our preparation of host-selective toxin from H. sacchari The preparations had identical gen- otype specificity to sugar cane. identical behavior in all thin layer chromatography systems that we used, and very similar IR spectra. However. preparations described elsewhere were yellow ( 7); the sample we obtained from Strobel was yellow. Our highly purified toxin was colorless and our impure prep- arations were yellow. The mass spectra which we obtained from Strobel’s and our preparations were very similar. with high mass ions at m/e 380. However. an ion at m/e 236 was presentinStrobel'spreparation butwasmfinginoumthe rn/e 236 ion appears to be a contaminant. We conclude from themamapectraldatathatthesametoxicmoleculewas present in both preparations ManydetailsseeninourNMRspectnnnofthehost-selec- tivetoxinarenotevidentinthepublishedspectnnnofhelo minthosporoside (7). The published spectrum (7) has a low signal to noise ratio. possibly resulting from low contamination with water, a low concentration of toxin, or other factors. The four distinct one-proton peaks in our spectrum at 4.95 to 5.41 ppm may have been lost in the signal noise in the publ‘mhed spectrum of helminthosporoside. The series of peaks at 1.35 to 1.2 ppm were interpreted to be from a proton in a cyclopro- pyl ring (7). In our experience, such peaks were evident at 1.19 to 1.04 ppm in spectra of impure toxin preparations (data not given); cleaner preparations retained selective toxicity but lacked these peaks. Limited data (not given) indicated that the 1.19-1.04 peaks were from peptide contaminants contain- ing valine. A trace amount of the valine-containing peptide is Isolation and Characterization of Toxin from H. sacchari suggested byasrnallpeakat 1.08ppmintheNMRspectrum of our preparation (fig. 3). Several kinds of data indicate that the true molecular weightofthetoxinis>1000.Theionatnt/e380(l-‘ig. 2) probably does not represent the true molecular ion. because of limitations in the instrument and the method used. The rule 380 ion. with the empin'cal formula 02:11:40.. is consistent withastmcturecontainingonegalactoseandonel5-carbon unit. However. the man spectrum for methylated toxin. ob- tained with the Varian CH-5 spectrometer with the EI/D method. contained an ion at m/e 1060; this indicates that toxin contaim at least 5 galactose units plus a Cian moiety. The proton NMR data indicate 4-6 galactose units/toxin molecule. The 1“C NMR data. although not quantitative. also are consistent with 4-6 galactose units. The proton NMR of the Me,Si derivative of toxin indicates 14-17 hydroxyIs/toxin molecule, which is consistent with 5 galactose units. Results of the methylation analysis of galactose from hydrolyzed toxin mggest5galactoaeunitsinanunbranchedchain.”Cth spectra indicate that the oligossccharide contains B. 1 -t 5 linked galactose in the furanose form (5. 9). An attempt was made to hydrolyze toxin with a and B- galactosidases (Sigma Chemical 00.). There was no liberation of galactose or C-15 compounds by these enzymes, used singly or in combinations (data not given). Four different 15-carbon compoimds were isolated aim hydrolysis of the toxin by dilute acid. All the 15-carbon compounds had a molecular weight of 218, as determined by man spectroscopy. These compounds appear to be convertible from one to the other. NMR data indicated that the com- pounds differ from each other in the positions of their double bonds and rings. The compounds may be sesquiterpenes. as indicated by the 15-carbon akeletom with double bonds. Fur- thermore. the molecules of some sasquiterpenes are known to be rearranged in dilute acid. The MS and hydrolysis data indicated that the 15-carbon imit is attached to the galactose chain by a galactoaidic linkage. Acid hydrolysis of this linkage shouldgivesl5-carbonunitbearingahydroxylgrouponthe carbon that was involved in the galactosidic linkage. Further characterization of the 15-carbon moiety is underway. In summary, the data discussed above show that toxin contains galactofuranose units linked by 8.1- 5 bonds, plus a Cadiz; moiety attached to the reducing and of the oligossc- charide. Several lines of evidence indicate five galactose units. The empirical formula of the aglycone unit indicates 5 points of unsaturation; spectral data indicate 3 double bonds and 2 rings (a sesquiterpene). Molecular weight of the toxin was calculated. tentatively. to be 1028. AWb—WOU‘MWWMWOy andW.H.Remchformggestionsmdhelpininterpretationofdata. andtoDr.H.Nunasforintarpretstiond“CNMRdata.Wealso thankProfucssN.EGood.C.J.Pollsrd.andK.Kohmotofor helpfulcommsntsandd'nctnion. NoteAddedinProol—Werecuitlybecsmaawareofworkonlf. saccharitosinbyRC.Beier((1980)Ph.D.theaia.Dsparnnentof Chain-y.MontansStateUniversity).Bsisrmggeststhatthetoxin maycontain2galactoseuniusndanaglycone(CanOa). REFERENCES 1. Dickinson.C.l-1..sndla1cas.J.A. (1977)PlantP¢thologyand Plant Pathogens John Wiley and Sons. New Yak 2. Lee. S.. Aoyagi. l-L. Shimohigsshi. Y.. lsumiya. N.. Ueno. T.. and Fukami. H. (1976) Tetrahedron Lett. 843-846 3. Livingston. R. 8.. and Schafl’er. R. P. (1550) Phytopatholoo 70. Abstr. 449 4. Okuno. T.. lshita. Y.. Sawai. 1L. and Matsumoto. T. (1974)'Chent 10 Isolation and Characterization of Toxin from H. sacchari 1709 Lett. (Chem. Soc. Jpn) 1974.635-638 4350—4357 5. Ritchie. RC .8. Cyr. N.. Korech. 13.. Koch. H. 1.41111 Perlin. a 8. Dana. T.. Nakaahima. T.. Hayashi. Y.. and 1mm 11. (1975) s. (1975 Can. J. Chem. 43. 1424-1433 Agric. Biol. cm 49. 1115-1122 6. Schefler. R. P. (1976) in Physiokrgical Plant P41115169 (Ency- 9. was“, 'r. a. Won. 11. a. Whaley. 'r. w.. Barker. FL. and in?) of Plant tPhysiology. New Series. Vol. IV) (Heitefuss. Matwiyoff. N. A. (1976) J. Am. Chem. Soc. 98. 5807-5813 FL, and Williams. P. 11.. eds) pp. 247-w9. Springer Vex-lag, 10. Wheeler. 11. (1976) Specificity in Plant Diseases. (Wood. R K. ' S.. and Graniti. A.. eds) pp. 217-235. Planttm Press. New York 7. Steiner. G.W ..and Strobel. G. A. (1971) J. Biol Chem 246. Additional referencesarefoundonp. 1710. E amnesia” Irma lstss 14 treat 44 4 flee-em 11's .slstely stter 4.441 h- the eel—t teas. mmm 44.4"". 4' the 441m reenter ta 4 mounts. 4t tm may-444 tells he res-s mull-44 441m; lessees it sees 41444444444 seasan tas ascent, 4' to tan preteens 4 n as strve e' the 11.: slats see .4444- some! attn lestcster. the statues tens 4144.74 mum to tan mnllsstea 4444. l' .14 usefully. ten 11: Mr! 4ll-es 41-141 c-1414 snaretlu 4" team '4- 4 statlsr see-teal: m Iselsttea .14 tasrectmaatlee 4' Inst-talent” 7441' (mats-l). the 14414401141414er (Ir-0.51.4: ace-e4 tr- uss slate as tests .1 n. hlmmr-sscc 41M 94- sh gs) atta sever-4 411m h selattse (0.5 41) 4' test- ". 414 mtsolsyer slates .4 We. a, alts ester 44 tee 4514441. on 4 to)- (0.9 a) cannula. layen (tr- t. to son.) 4' Insea- s-rs. 1 a. 14441444- 5' c4114: earns—44 r4414. 1o :4. tosses- 5-26. I a. an: I“: 041 est-1 sateen.- min. 10 ca. 14414 .2 14 the 10 1414 a: serum 4' t- slasts. Omen s. unseat. sea Inert '. nastier tame trestle-s an ..1-14¢. seine-tram. ans 414441444 14 0.1 41 4' new. mat. the mutt-1 ass assess tar-saw) 4 .1491 3.: col. (iso-too 4441.111 I 1.31 n). 0441444: wtta user. 14414 mi 14 the elaste at 120-18 41. trestle-14 '4. 110-114 s1 are -1114: all eer-seatretea. 1's- ereeentw teases lease. 4' electrolytes tr. reasse- clear (4 411 lamuu D ms at 10 44141. 14414 .4 414414 n44 ate-es 14 nae-41 4t -15‘c. 144m. 4' 14141 n 4444 .4414- rns Jen t. (.4441: 4' 1 444 4' 14m em 1 > a tn-thymlr‘ (114431) “441144 41 14-14 444 fihfivvl eaten 4444414114: 14' mate-4 54444 441mm. 41a 44:1 scene-nu 4 41.4mm net-em. av .44 14 441 car—turn». 14414 14 44444441 4.. 5404' the 0.3. o. L snares-4 seesaw-ts 51.41141 4t (4441 94141.31. Sun talus. 4' tee «144 14 4 rescues 414 at am. mag 4 .141 4' 4111-4444. ins-tayuuyltsteenlo 14 9, isolates. sates an "4114' 1a mutt; to reunite: .m mute 1.4 a» can I44' 444' fine!- (Yfl-Sl‘ 1.91"“ (ta-1:41 to.)— sea are!) its 441411- uaas 411mm to 5 mag (01 Cultures tar 1.4414 mien save 4": ter 11 can 41 41.21%." am sultan 14 444 tens/s1. convenience m tie-eluent «1 41441.44 41 arcs-411.4." ltw saux settles 44:1- seetsmaq too 41 4' "144 .414- 4.01th 411- 0.11 4.41 avert car-stagnant s- 4 1 n 450 I :41. 4' a 24411 no 4' ll 07-1. 441114 4 mltttwecure 403 (11. 1441s as: present he sultan '141. ‘10 save harvester 44 "Ivan. '41 areas 4.1144441 ans 4 new seat tateenter. Yes t-erstan '1 4y meaeecleta an ants-1 Is. lesser. or: 44' 414.414.114.171” to ram. eel tas tall- fle- rats 444 wallet-n4. leather-41 owl's-t .4 a Wt. else alts 4 .11. has rate 4' to elm-rte 14-14 .1u1-mmuomnmll1w-owmma electre net 1‘ ”44411414 eat an '1'. resist“ 1144.. all assays a. all used-lat stmulstlyl (1B1 Crust!" a! latent mm 444 - 44444-414441. .44 nsestee tee er sen 11444. a 44414-44441". ((4 48), 444 4 testsdueafltue class .444! annals ut- teua -lssals.lr1-SIl-Z ass 4.- ts 4441441144 tam (111‘n 441 use-mt 4r lat-4410) an aael 14 secs 4444;. "am are or. . asserts- ale-nave (4). harness .444. tats-14,44 mm .4 .44 4.4 scents 15441114 44444 is. 14m: 44- 0444 muses-14. 7.4 441411. .4 strand 0 4 20- to ca slats s- eenlessa MW (altars filtrate (s l1t4n1444 cue-antes q mic-ll to 4 17 a "an ant- ruler-414's: «stay! star (1.1). ha fro-ten- .4 Issues 41th a m pressure to to 4 actustaa m1 (lam A tn- SIps Clea £4.14. asset to manner (11 44411114 14 4,10.) 4441144 14 4 strip 444444 h- is. n: 41414.4:4144 .1 mace-11'4“. lance n4 sealer to PC. 4114444 rev it seen. all 44444114444 It 10 000 I ssnlesed at 140° t '4' s 414.541 4'me men-enact: 4' tea rescues ass l4 441444 44144 '4' I limes Yb stereos) pellet castsintn swam taste ass re-suseeaas 1- 900 ll 4' 104' 0.1. Mneattses 1.4414 (s, 0.741444 4.444 .4 41.4.4 alts n... n.“ u 41 “a, .4 steel-s1. sun-44 rev 5 41114144. .14 cents" 44414. 744 must .4 «seems. the '7 414m! stun 7h aelatlu ear filter-m the” 4 41421 1144' filter ear M to 441141 4144 M 14 t“ nae-s) 1400 s 1. 4111-1144. wrflasst. e- u- aeorfiauat m4 “as none. ”sun. the reside 1.4 41444144414 0.! sl s' mes talent" 0.0111 «trams. 14414 .1 slates 'v- aertt 4 s, museum-4 the .1141 14 am 41 4' 501 4041141 tutu-144 11 -1. males. 14 a this. t. slat-'7 n4 senate! 'e' 10 slates W. 111,1, .41". 44414 ‘ 4144414. 14 aunt o. as a sat tantrum. 1'01: 01-414mm sue n4 neestee an tea me 1m 41 must Ian an" as a 41.541 4141. la» sol-s 4' mint-I an teen 44444.1a4 4141 ”less. 911141-44 three. Insta- 41444 "err "144' sear ts v.44 were: 4' but A. 444 — 444144 '14.), arts 4 74414441444 see. all tn sressretlu a seam to 95’: 14 4 mates to 5 s1. 114144451 (to sl) ass “In-es slealy me a mater-tsetse ”are“. .4414. 414:4. At 40 stsate intervals the 4141 .4 44.144 nicely. tas alanine- .444 .4 see u- aslatlu a- aala st 45‘: 44' :4 er. 4 anusltste ass 4144144444. .4444. -411“ 4' ruler-overseas 4044.01 tn vtsl ass rater-u to tea seeing slate 14t4l use 4' wanna" st 95': n4 tao seen. after 401th 04 4444441 ~14 tot the .030“ “mu" ‘5 consent '90” OHM. '“ ”4 1° 3 " fl“ 5“ In." extreme! the. "at. each tlu 41th an 4411- e' calsre'sve. all celeroten estretu son 444 places an 4 1444444. 111-20 col-1 (so 4 Ira-1.174 He- r414 14' um 44a 41- 4th" e-tma .4 mum. lee season or me also-4'44. .4444 .44 4441,44. 44' ymmu tell-u ear 0. 0': - 0.1 all-14 lee cal-1 n4 “lasso 41th 501 means. as l s‘ traction: mu. “ collected. tom .41 recover-ea la Haulers Ores 417-669 al. 77411 at. sessrstc tam 'na 1 values an see-um: 44441.net ('nn1st'14e1cs 4111 44 firearm in maltt- [4:5 ntlee 4' telenfon-teloel 11 c' tam vol 111. he mlere'ere use 4' 'rectlu has the Lit-20 (see other) eel-I ass escapes 'er aIelestce‘ scttvlty a: easiest“ “on to: 4 i immune 41 resets-1 worsen at es en 4 nt slate. at- 44441944: to TLC to Our: "setter-4 cesium-g tam ms .IIOI. The 4124 as late-11.4 4' 4441.124 417: treat 411.4 chlorofor- 4141a,! usev (1 11.144 44444 are must. .4 at l 44441444: 44 Tu. 414444 was 410- stsaelvlsalaezsentsszsseaeaenc 4:14. 1441:4144 nlsttn 0.15 444 tea “410.44. theatre-444414 runs '1. tea slne. sent-es. era slam .4!“ 4' tam as males. Only the tram-s "- ue a)- mat-re all-tea u with 5 fine! 4' (Neuter-e. bless flares I" "Item as Muted to 0.1 al. In estate )0! tall» are mad tart-r. 4.14 an w 4144144 I) 4 111-211 to)- (1 o 4105 43141- .44)“ alts «calm Duane): leases (2; 1: 2) last trestle- (175 all '1‘ he a)- -.s more: acted 1.4 m4 lute-mm" trestle-4 'r- the tit-D talus an esmlme. asses-tastes 1_4 :5! e- ear—steam. he 'rectten In- 45 ts 77 41 «4141-44 t lb 444 t (C- its seem at 70 41 4144414411 14 as 41 4' aster 111.1411 (0.5 all are shoes 0 4 14414444- M 4111; OICIOI' us t-ltt at 72 all. fines "4:114” 4'. .7 to 105 41 eeat41444 can .4 4 “.15; .4444 «144114: 11.5 C144! "laser alts .ta'. ta) "ecu-14 an collector 1. OK at 94 s1 444 (~18 assess 41 Mel. "aeti- cellist-m ”a far 44¢ 4' ta- ta: eve-lets WlflHHDHO-llul- H-asattsmtrsflwtsssaflssmtel-I. Yea "Wares-saws)ttmubethfltel-.tswcataacmcestret 414' tests ur 14 "ecu-s tr- fl-lm s1.leat:-east41414! from-Ia an “1114:. 1444» to ease. mp1; unease resent"- 4' but 444 Collie-re Islam ta this my. s: t- 154 417-44 41144144414441.1041“- 44.1412 seen-t1eerstnsserenescr4' sashllauanntamemassanasssozeontu. several tats-1444' slates (W 4111“ plan-211.c- 0.15 a tutti) set nuclease u 4 17 a treat me seats-4: inter 14:11 the tale-«41414144 1444 444 Ieutes 4, 4416144 tan Isolation and Characterization of Toxin from H . sacchari W. 1.14 14.4 mm syn-e11 estns (ll. 44 n sessee I 4 ..1-tsy1 eel'sslss 1:450) 15.1 ell 444 ..ste tat; I." (11.4 .1 see tn salstl. e44 4.1.4.144 5-10 414. 4 441414141 41. 144,1 4441. (11.) e1 .s n nee 41".... ts .etln sue-else 144 441.1: .1 .ales . lee. sl-|.y. a. tn reattl. e'staee eas leseestes alts ants-.1 a't. celare'see 11 411444 44444. 4. en 44141144 .4 .teettas 44444 "assent-stat (total 4444".) 1114 111114.04. evenness sense-testes 4444' 41444.4. tn 4441.4 .4 414441444 14 nets-410.4 411444 centeresns e1 '1 111-10 441.1 (OJ 4 It tel. 4144 seetn es en sales-t. n 41 trestle. .ee sellestas. 'eastlns ntalate. .t 141.44 44414 .44 10.14144 4, nettle. t .1 41144.4 . tale-1414' slates .1. .es .es .es an: al.... .441" slates .es sever. alts 41.4-41.1. 44111.: sneseaele .14 .4 .44.. ea 1ft es .441. sens. 'eastlees 10 ea 14. .10 :uns .tnlates .414. .44 nlns n .1. n41- alt . 1n .tlelatl. In.“ assess-lets .0414". 4' 44414. .tnlstes tests .4 as I .aaeselee :1. 4444 U' .ulnuatles . '4. mu. “lat. .414 .4 4444414444 14 0.0! 1”. at II' '4e 1. sears n n 4.41. tn 441411. m4 4.41 1.1 (1 e1 441-) 4' his. 114 (”400 neltntnsnee-seetes teteentate'ere. 1nnenlneta.seelleetes.ntn .1. 4. else. alts t 41 4' .tnnl. 7. slate a. nee. aen 4.1.4. an. ..1 i 1.. .. . o O .— v at a. ate-1144. 44 4.144 tel 4'.teen1 ees te-e. 4' 44411: 4414.1. 4141 444 444144 =4 .14 st 145-: tar s elsetes. Dee sens-14144 net 414 4.14444 1n .tanntatles seeseeen ”1.44s. ".4 1441 14411-4414..serles essnasatylateslelMels'esetlt naruezsyvlsln (1: 1 441/141) at '1’! 'ae 1. 4 44444 141.114 .4 asses-14444404144 .44! 414.444. ten-444 8m 11141441444 .4 414441444 14 est-41444 41119144 (1 e11 444 sets-est. J et- alts .tae 111.4 41 .es tlnz. 1n .es-1t n. .4 unseat. tee-44 411 tees. d .eae n 4144414. 14 44444.. .tnlee selerl. its me nests... .4 .es es least", ease ssetlelly .tsylatee 4141141443444 nee-41. teen... 4' 444 seesests . 4 tel— 41 n anus 1140 as e 1.1, “14444 lest-44.11, at 1111‘. .es-sees 41t- nst 4's. tnsenses. 1.4-41-4-444141- t .8. CJ-tlt'e-Oftnlgalatt'ul (.es-t1. t1.. 3.11 41.1441. 41.1 eaa'leests. 4'44:- stseeattaee see ends. eseetts stellsns .44 .estn (t. .set1tet144 .taeetntlen 'ee .es seetlell, .tnlst. 4141441 ntn sent! n esteem-s h 4 Delete-eaten!” 1m “Y! 4 . Danes. 44414 .4 We! tl14- as. a ates. .4411 .4444 sees-1t tset— (14414 t) 444 ..1-41 anneal-t th4- 1| mm.“ :1! 4 414414 41411441 asst 0.1.1.. m4 as. ea .test tesls as 411144 .1 slates 44.1.4441. nettl. ntn 4414' 1.4th - enelteeess 44194414 44.4."... at 1” '4' 14 e10 (ales-1. 144'. Inst 1144). 4141.44 trlselsel..|e en's. .latlee. nstesl (1.1. ass-4144m 441444.441"th eels (ls/t 41/111 41. la 14441 state-41 (414441. 4. 44411114 4 ‘C n 4 .14 :- ulna .es 14 014-1.. W 4' ~14 '4. m m. M m M hat-41.1.44 0.1 0.! ntnznev l, l 0.15 "hunt .lsx.tae 4:1:1 t.) 1:4tn1 .ltatazne' 7:1:t Ill ..1 :.444 than 0.14 Calm-ens“ :1 e M an 311144 44141:. M! . nee-assesses slates '4. t. .444. IesWinn‘tlnnuns'nnunaslssle... 7n “nan-elevenenllestes eneslm ”..1 W 44 41. 11 4144 444 Is an I (.15) e1 eels-laser slates (.eelns elte .141... sleenl atssr. 3:13. . 414 tn eel 441 see144t1ess nestles. 441.444 . ..ter meet-eel 41.44414. .4144 .4 cent... anlae'nnfil-tasle .eltlesnnmn nlna' ten-4‘ eel-4' .es-41. Inlseletee sense-stasetl'telteesnslaleape 444441- nlnnannttelntnnstlus'nm .teenl ”SI-fr!” '1. .4 .414. an.“ .4 mrnl “.314. ..- ."ltln leans-'1. 411 1t as sales. tn eel- leseeaeas ta I.“ e1 n1..te1'la-.aet1t .14 414144 4 11.. I 41.1 nestles 4' tel'l.e..tlt nls. ealatlelnleatn'ee l net .OJeldtaleevs'see- I. at... seeeeeaes '4' 4.14 "41,414 .s '411 . l'tae noel ls. sslastns .4 'ns tstnnsesesntnetsae n s'taslsaeee'ns tel 1.14). Den tntrdamagnunnlatasn-nlt nstane 1.4144. 4441141440 sens reel. 4' .ltaee "lens. '41 1.. n eases". .Iant 541.1. ates It stenlnts natal. line .n‘. . “*1 "I’M!“ flan vessel 4.44141. 4.14 4e ..ul‘. . ealn M 1:12 ts-e 441.- 1 41 144441. 41-41 cm M .414 g 444441334 1. 4, 4.4 4.44 an 4.1 m: uses-4.344144 35.111 4.14 .' 4.1 ~ 4.1 4.: 4.4 m 441-. l 41 tesaln M It hit - N .ln. 2 e1 ”.41- .J . 11 07144. leaves '4. . sle- n .444 tests attests lease (lee 4 4114' sane-4.4.4141. 4414114- aaee 414444 .nll .nsnavnanealnss' .slsaslaa'eastln..stnleases.ee .14 n t as. eeealsan ette at. 44144144 nslty 151. “41.1, ..ltlea 1.4. ens ..1 seen. Hen-41.1.4 (tleyts 4.44.414 Ill-Masts.“ .es a'tse 4.444. . tule. ma. '1. 41.1, restate-s tlee. .es s. 414141, sweet... as 4 .es-41.1144 sees: (41. Ynnnaesnsas (l)-..n.teee1nt.elenn 1.4144414444144111 1. 4' elects-else. 'e. ..1t144. manta. n 41 1, 1n1t144 41.44 d 41'!) 441-14144 ten-444.414.4414..“ 1441 mesatleete 0w nnee'nteatneelslelaeesttlntseeslsesnetnns is. 141.1, .41". taste ”as east '40; asset set.- .4 ta 4! 444 a. a It ta metre 1e tn 41'. fl seessratlse teen tt 4 .leeless UM. Kilauea testtes 4.17 mu. ne tn aleeless n'. .4 tackles. tn .seaeene. In alelaslssl astle'ty .es h 4.. .'er tn aflslesl nestles. 44414144 tesls .1 easy 441.14 14 met .4 menu 1t .4 sllptly less 441.14 14 steetnl salts-144. ”414144. 4444441. 4. eats-41. 14414 ass seen-sly 4414414 14 .44... n .4 lssaleale 14 al.... .441 natata. n ..441 44444 1.44144 nlaans. 1441' «4 a. sense let‘s tease. lee see-s1. 4.444. a. .14 4 .snlea 41.4414 e.ttlee. n.ee .1 44.444141 44 4144144 ans-4414.14. 44414 tree ,1 .es-sew 441.14. 1.1-.e14atla4 4' "statt 444111 ..1". atmta'e 'ea' :40 ( ta n 41.44. 1.1144 test tesle 44.4414 14 4 441441441, 141-...4-44141114 an... 1.14 aalattn 414 .t .4” “'4 Is .4 In T. n. 210 I. n 14 444 4141414 eases. hell seem-..ees mlunthr nnsssesteeeneps'eeneelse .44 ans Ml. at lull a". 1.. .t4 “teats m M 4' ”antes 4.44.4 t "1 la 14 l 1 14411110 1 .1441 .0" 4.4.4 as at l‘t alts 4 set 4' etteepe. tn lasts. .4 ".es "'4‘ 1.1.4 In“ .teasnl .4 «teller-then tale-4 test. 4' teltlseeeaestlt asle. 1n .414. .4 414441444 14 asset see a 411.. asetalsseg teen-s nH'mttue'tasls.etlasleaeesttla-1n41aesneeeselt «sales-tassel. 44.44 tree. 4' aster. 141-5114 110 411.4 44444 a. tn 4141 .s .14 at 411': he 10 41.444 44 14444.14“ emvetuanas. 4 eta-vases .s sees-1'"! eaelsstluee 441441444 (11.! 4414114144 141.311 1. .10 tee. lee sass». 514441444144 411.4t4(s.114'tntes 14:14.:ts Menstsese 44441444644111. natal-18 01.1.1 1.4 esters 1.41 444.441.. leans-11 y at 110': test 41.4 .44 later... 1. eeaalts 1.144s. test .aeaInls 4' 1'1 44 4' teen 4414.44 11.1 44 4' 441444444. l 1 1 1 14. n 444 wlnls .es 1.1m. .es teas ee plates. .1.) se' . ese 4 as 4. nlatl. nlnls n .. : reel. 4. 444441141141 asttaee e' tease lastssa .1ts 14 44414 1n see... e114,...“ alts eats-else 11.4. saw «Lots-44.0541... lease-ltltsese'tneeeeistsnetne .tseelees e; 44.441.44.411." .44 4.4444 4144 pals-44 .estee (11. 1’44 seeseese 4' 4444441 .1444 la. 14 assessast 4'tn he ssnlelly .tnlates 4141141 netates [(‘.s-41-0-444t--2. ). 5.6-tetea-o-nt11rlastltel 1ele SO. ”.1051“ 1.4.t-tn-o-4s4tyl-t. 1.0-tH-Ont‘yleslaetltel (en It Meet 4441. 4' tease“ t. tnens .Masalte Inns-st 1.4'atls.es14neeelysls 1414-4444-4444. entnlllu 4' tn 44146.4 .1. ls 14 tn '4... use. 1. 1.4.1 meatl- 4' 114214 test nlataly neeelnas tee .tnlatss .4111 n 444444 4141- menu“. 4' enema .4 4.4 s. 1414 .euat noun. 'er en a. .etlally .tnlatas 4141t41 .atstss .s nee-1.4 n1 tleuottnsses-esslstss rs .Ynaseearnleeeetlee' at: 14.1.5. 6-tetr4-oneylsa sstltalta .l totrl-O-aesty -t.).t-tel-O- .tnlgsleeutol ass 1 4.4 altna'est reel. 4' 1;) sea blasts .ta .4. 14444.net. el 1414 4144 «test 4' he 411144444 alrllnassea . (4 4 tee-1.1 . latte'aeense nth 14444.1 10111.44 .lseteweenss 44m 144144 441 u 3t-ee 4.41.0141 411 platelets-444 44114 41141414444444. ”1 41 lasts sheet. '4. .411 A .14 It. teele pet'leatl. m a. t .1 I. seen u .14 .es? el elaate 'eettl. n slat. . 4 ten-lapse slats a. .1441... essteee: .te' (14:11 'n 4141.44 .44 tn- .444. 41th 414.441.1441 44111.: nae-4:“ 4:14 sesnatas hlsnnnst'ern(altseltteeeettqeeleesntnnt slates) .ee 41.. '4. tn L11- s-tll 441- 1441".- "ecu-14.41.444.44 sees. tn 4... slates 'leet (teestln 01-67. 1.14 31 .s .4 humane-441.44 14 latae 'nstlan teen. teslt aunts 44441.. 1.444. n ta... 441.44. 4' slatteelstes 91.14“ 41444 lee seen-1.4.4.44.“ 44.44444 4.414114444411141 11. .4111. In 41.144 441.. .4 4' 441444 tee en 1.14414 4444441 "14444444 use. see 414. '4 1.14 J. .41. 1.411 as. 444144. '4. asleeee '11"... n has. 144444444 Isle ll! .4 41.1; 441.14 14 .tnnl n nemesis 441.14 14 .es? .4 44an see. .414 111 e. slates '4. tn lee-e .lna at“ m .tnnl a. 'e. tn M tel-1 4114 nets-41.4. .ts' (19 11.4 .44 44 .4 .14. 14 tens-laser m e' .sles 11 see 11;. ~ 144141.. .4144 .es 144.114 as 414 444 ”417114 seesatts ..1444 ey tn a...“ sauna. tr tn taut. islestn .san tn 441, sense ten 14 tn 44.444 nea- 1. 4414441444 .4441 '4. neeelnes .4144 I ass 1| 1.. .4114. 4,4 «41.1.144- .ue seats .44 .talne 1n eats-t1. 11.4 4. .41 4444144 4' tsasa nseelyt stee.ect 4' nan-re .eelnsltltsl "tenet. ne'eaelt-ta eta-Inseaytlt sees-use! 4441410144. 4. t 4. 41 ('14. l;- 114 tests a. .4144 1 .4 ll “"44. 14 en eelatus nens 414444 e' .4 .eslt'eeeet colt sens 4' nan-41,414. 1n aleee'eee nee 4' noel”. .414 111 4.41.4 t- 141: e. s alts t-lts a. 4 14 teen nans.4 .4 14-444.- aeeesst. 441.4 314.4." 111. .4 n 1441 at. nut-4.11 n4 4 manl- t1. ntne use i n a. fines“trainee-14.404-144.01...”teselstlsenss's ' s-eesu "and". ”tn-lien-‘ -e'entennsleslnllnlntnnnees. 144.4 sens'nnstteleetnnnsntn..last.s...est1. .mflnhnnmlynnndplastn ..tmn‘fieln'eenls 111. um-‘t~ 314. t. .44 .es". 4' n 4' tea .1....ea1.le noelstlt vents 111-1441 4' n1- '4. 41. Inla— llnn ntn- "(1 1. 441“ .ent 4144444- 14414411. (1! 441‘s: 4444111 anaestelsesletntut .ssnstee s'tn. (t-l a.t.nseat41.se4 nul eeelaatl. Intense. cum-s “01......" In. Us 414. I1. 1.41.1. .41 not. .44 h nlettal “la . ‘1!“11 1. ..1-41. I- new J. It“ I” t. 44.4..v..lseeet..tl.ane.l..tlnsea.|-..stn..e. 1141415444. (as I“. .- I 1441. 1|). ~- 0' Mt ”In. ”mt. fl 44444441. I. Me. I- L. n 44441144. I. D. II”) M nan-141. 4. 4.44144. 4. h. 444 1.141”. t. t. (1.) Ml. 10.“. t. ”.41. C. L. .4 net-r. ‘- I- "’72) ~01“. H. 444441. 4.14412. 6. I... c. L. n W. I. (II’) N..- u. b. *- 1444441. quU-llu. EXPERIMENTAL II CONVERSION OF HELMINTHOSPORIUM SACCHARI TOXIN TO TOXOIDS BY B-GALACTOFURANOSIDASE FROM HELMINTHOSPORIUM 12 Plant Physiol. (1983) 72. 530—534 0032-0889/83/72/0530/05/SOO.50/0 Conversion of Helminthosporium sacchari Toxin to Toxoids by B-Galactofuranosidase from Helminthosporium‘ Received for publication November 8. 1982 and in revised form January 25. I983 Roar-In S. LIVINGSTON AND ROBERT P. Scmarm Department of Botany and Plant Pathology. Michigan State University, East Lansing, Michigan 48824-1312 ABSTRACT HWWMnhost-aeleeflvetoxinnndmo alyrelated nontoxic WJerereferredtoas‘toxolds'Toxlnand thethreetoxoidswereeachbolatedtoahighlevelolpuflryandwere Wderuflicmwmmnleasedgalnctosewnsm hyagalaetoae oxidase/peroxidase assay. Toxin was formdtocontahifonr nits of galactose per molecule. as previously reported. Toxoids I. IL and lflcontainedonetwo.andthree-ltsofgalactoae.respecdvely.ln dtuesoftheiugmtoxheoncentrationpeahedatiiweekalollowedby nnpiddecline;astoxin|evelsfell.thetotalamormtoftoxoldslncrenaad. Aneurymewlthfl,‘ ‘ “ activitywasfoundlnnnallamo-ts hthecdtuesoffl.sneehari:theenzymeconvertedtoxlntothetoxoids arrirru.,‘.u“ " " waspreviouslyhnownfroniveryfew-icro- organisniszthereforeaeveralpathogenieflehnrhosporiaandotherl-gl weretestediorprodnctionwfl-I‘ ‘ " activitvinedt-efl- untesandmyceliaofflnieroriaeJLnaydinflccba-niflandflm wasniuehgreaierthanlnflttratesandmyceliuolflJoeeha-i.Moreworls isneededtodeterininethesignlficanceofenzymeprodoctioobythese lI-gi.No,‘3_“' ” wasevidentfromfnsciimaxm flashed-um Helminthosporium sacchari (Van Breda deHaan) Butler pro- duces in culture a host-selective toxin (1 l). Clones of sugarcane which are highly sensitive to the mm are also susceptible to the pathogen. whereas clones that are insensitive to toxin are resistant to the pathogen (10). The toxin was first characterized as 2- hydroxycyclopropyloa-n-galactopyranoside. or ‘helminthosporo- side‘ (12). Our data indicate that this structure is incorrect. and that the toxin molecule contains four to six B—galactofuranose units plus a sesquiterpene (3. 4). Macho er a1. (7) have proposed a structure which contains four galactose units attached to an aglycone (Cioflu02). Several nontoxic compounds similar to toxin have been found in culture filtrates. along with toxin (3, 4). Pretreatment of sensitive sugarcane tissues with the related com- pounds (toxoids) reduced toxin-induced electrolyte leakage (5). The toxoids were purified and were all found to contain an aglycone with the same mol wt as the sesquiterpene that is part of the toxin molecule. Thus. the toxoids differ from toxin only in the number of galactose units in the molecule (5). Early observations indicated that toxin in cultures of H. sacchari" reaches a peak in 3 weeks. followed by a rapid decline. The decline in toxin could be caused by a fl-galactofuranosidase such as that described by Rietschel-Berst er al. from Penicillium charlesii (9). 'Supported in part by the National Sdence Foundation (grant no. PCM.8100‘7| I). Michigan Agricultural Experiment Station Journal Article No. 10784. We have found that the Penicilli‘ian enzyme will release galactose from HS toxin2 and from toxoids. thus confirming the fi-galacto- furanose conformation of these compounds. We report here the isolation of an enzyme having). 0 3““: u-fn mu " activity from cultures of several species of Helminthosporium. The enzyme may be responsible for the rapid drop in toxin concentration in culture fluids of H. sacchari and the concurrent increase in toxoids. An abstract of some of this work was published (6). The term ‘toxoid' is rational and convenient for the nontoxic compounds related to HS toxin. This follows previous use of toxord for inactive forms of toxins involved in animal diseases. However. the toxin and toxoids from H sacchari are not known to be antigenic. in contrast to toxins and toxoids from animal pathogens. Three toxoids from H. sacchari are designated 1. II. and 111. in reference to the numbers of galactose units in the molecules. but with no reference to isomers. These same three toxoids were called ‘noxins‘ Ill. 11. and 1. respectively. in an earlier report (4). MATERIALS AND METHODS HeWthospan’ian sacchari was for toxin. toxoid. and enzyme production in still culture at 21 to 23°C, in H. Roux bottles each containing 200 ml of Fries medium supplemented with 0.1% yeast extract (8). In time course studies, the ailture fluids from three bottles were harvested each week and filtered through Miracloth. 1n toxin studies. the bottles and fungal mats were rinsed with 25 ml of 50% methanol; the rinse solution was added to the culture filtrate which was then concentrated under reduwd pressure (at 37°C) to one-tenth the original filtrate vol- ume. Norit A (2.5 gm) was added. the concentrate was stirred at 4°C for 15 h. and the toxin and toxoids were extracted as previ- ously described (4). To recover all of the toxoid with lowest mol wt, a final dichloromethanezmethanol (1:1) extraction of the Norit A was required. The solutions were combined. filtered. and con- centrated to a syrup as described above. Methanol (50 ml) was added slowly. the solution was held at -15°C overnight. and the resultant precipitate discarded. The supernatant solution was con- centratedandmadeto3.0mlwithmethanol. Thispreparation, designated ‘A,’ was used for GLC determinations of toxin and toxoids Highly purified preparations (designated '3’) of toxin and toxoids were obtained by the procedures described previously (4). A trimethylsilyl derivative of the preparations was subjected to GC to verify the absence of free galactose or other compounds. Toxin and toxoids were stable when stored in methanol at - 15°C. Dryweightsofthepreparationsweredeterminedafterdryingat 110°C. An aliquot (IO—20 p1) of preparation A was dried and MesSi derivatives (50—100 pl. total volume) of the toxin and the toxoids ”Abbreviations: HS toxin. the host-selective resin from H. art-claim; MesSi. trimethylsilyl. 530 13 14 ENZYMIC CONVERSION OF H. SACCHAR] TOXIN prepared as previously described (4). The MesSi derivatives (2-6 til/injection) were chromatographed on a 2 x 450 mm column of 2% OV-l in a Varian 3700 gas chromatograph with a flame ionization detector. Temperature was increased by 30°C/min from 130°C to 350°C. followed by an isothermal run at 350°C for 5 min. The recorded peaks were cut out. weighed. and quantified by comparison with standard curves for purified toxin and toxoids. Toxin was also measured by an electrolyte leakage bioassay (10). To determine the amount of galactose present in the toxin and toxoid molecules. aliquots of purified preparations (B) were by- drolyzed in 0.5 M TFA at 95°C for 2 h in a sealed vial. The acid was removed from the opened vial at 40°C. using a jet of N2. The residue was dissolved in 0.2 M phosphate buffer (0.5 ml. pH 7.0). Galactose released from toxin and toxoids was measured by a modified version of the procedure of Fischer and Zapf (2). The assay solution was adjusted to 0.5 ml with 100 mM phosphate buffer (pl-1 7.0). A solution (200 p1) which contained peroxidase (5 units) and o—cresol (3.8 mmol) was added. followed by 50 pl of galactose oxidase solution (1.25 units). The reaction was allowed to proceed at 37°C for 30 min. when the absorption at 410 nm was measured. Peroxidase (type 11) and galactose oxidase (type V) were obtained from Sigma Chemical Co. Fungal cultures for enzyme production were grown as described above. Cultures were harvested and the fungal mat was squeezed to remove excess liquid. The solution was filtered and precipitated with ammonium sulfate (476 mg/ml ) at 4°C. A pellet was collected by centrifugation at 20.00ng for 20 min; the pellet was resuspended in 2.0 mu phosphate buffer at pH 7.0. Enzyme in the mycelium was extracted in a grinding medium containing phosphate buffer (50 mar. pH 7.0) and 2-mercaptoethanol (10 mM). The mycelium was disrupted with a Sorvall Omni-mixer run at high speed at 4°C for 4 min. The resulting slurry was cenuifuged at 20.000g for 15 min. and the supernatant solution was precipitated with ammo- nium sulfate (476 mg/ml) at 4°C. A pellet was collected and resuspended as described above. This preparation was used when comparing enzyme activities of various fungal species and for weekly determinations of enzyme activities. Another procedure was used to obtain a more purified enzyme for GLC studies of enzyme hydrolysis products of toxin. The pellet from the ammonium sulfate precipitation was suspended in water. dialyzed against phosphate buffer (10 ms. pl-i 7.0). and centrifuged (20.000g for 15 min) to remove insoluble materials. An aliquot of the dialyzed solution. containing 10 mg of protein. was applied to an upward-flowing Bio-Gel P-lSO column (2.6 x 31 cm). which was developed with phosphate buffer (10 mar. pH 7.0). Enzyme activity was eluted as a peak in fractions from 58 to 74 ml: enzyme elution came immediately after void volume. indicating a mol wt >150.000. The active fractions were combined and dialyzed against phosphate buffer (1.0 mu. pH 7.0). The dialyzed solution was then concentrated under reduced pressure at 25°C to one-fourth of the original volume and the pH was adjusted to 7.0. Aliquots were frozen for storage. The enzyme was assayed using HS toxin as the substrate. Toxin was purified as previously described (4). omitting the TLC and final gel column steps. The preparation was free of galactose but contained a trace of toxoids. Enzyme activity was determined at 37°C. using 50 or 100 pl of a solution containing acetate buffer (20 mm. pH 4.5); the preparation was incubated for various times. up to 20 h. The amount of free galactose was determined by the galactose oxidase assay (2). as described above. One unit of enzyme activity was defined as the amount of enzyme required to release 1.0 pg galacmse/h from 1.0 mg of toxin in a volume of 50 pl at 37 °C. When possible. all assays were run at enzyme concen- trations able to release 20.0 pg galactose in 6 h. Specific enzyme activity is presented as activity units per pg of protein. The galactose in toxin and toxoids is in the furanose form (3. 4) and therefore is not a substrate for galactose oxidase. Pretein was 531 determined by the procedure of Bradford ( l ). All experiments were repeated one or more times. RESULTS Desalination ofGalnctoae hi Tank and Toxoid Molecules. The mol wt for the toxin and the three toxoids (I. II. and 111) were calculated to be 884, 398. 560. and 722. respectively. These weights are calculated on the assumption that the molecules contain the aglycone (C...H..0,) (7) plus I or more units of galactose. Purified samples (50 pg each. preparation 8) of toxin and the three toxoids were individually hydrolyzed and the amounts of galactose released were determined by the galactose oxidase/ peroxidase assay. Hydrolysis of 50 pg toxin should release 40.7 pg galactose; the 40.9 pg obtained experimentally (Table 1) con- firms the assignment of 4 units of galactose/toxin molecule. Toxoid 11] released 36.4 pg galactose (theoretical. 37.4). indicating that the compound contained 3 units of galactose. Toxoids 11 and 1 released 33.4 and 21.9 pg galactose (theoretical. 32.1 and 22.6): thisconfirmstheassignmentonand l unitsofgalactoserespec- tively. AecnmnlationofToxinandToxoidslnCnltnrel-‘hldsofll. sacchari. Toxin production in culture was shown by electrolyte leakage assays (10) to decline after reaching a peak at 3 weeks (data not shown). We then reexamined the time course of toxin production in culture. using quantitative GLC of MesSi derivatives of the toxin and toxoids. The GLC data showed clearly that toxin concentration peaked in 3-week-old cultures. and that toxin titers declined forthenext3weeks(Fig. l). From3to4weeks. the total Table l. Galactose Released by Acid Hydrolvsr’s of Purified Toxin and Toxoids (50 pg each) Galactose was quantified by a galactose oxidase/ peroxidase assay. Galactose Recovered Toxin Toxoid Ill Toxoid ll Toxoid 1 F8 Experimen' tal value' 40.9 1 0.9 36.4 :i: 1.0 33.4 :t 0.7 21.9 : 0.7 Theoretical value’ 40.7 37 .4 32.1 22.6 ' Mean value and so of four replicates. ° These values apply if the molecules of toxin. toxoid Ill. toxoid ll. and toxoid 1 contain 4. 3. 2. and I units of galactose. respectively. FIG. 1. Concentrations of toxin (0). toxoid 111 (I). toxoid 11 (O). and toxoid l (A) in culture filtrates of H. median; as determined by quantitative GLC. 532 toxoid levels increased as the toxin titer fell. The results suggest that the cultures may contain an enzyme capable of cleaving the B-lé-lined galactofuranose units from HS toxin. thus producing toxoids. Detectionoffi-GalactofuranosldaseActivitylnCultm'esofH. seed-1°. Therapiddropintoxintiterfromthe3rdtothe 5thweek in culture suggested that maximum activity of a hypothetical enzyme should be found in the culture filtrate during this period. Filtrates from cultures that were 2 to 6 weeks old were dialyzed against water. then against 25 mm acetate buffer (pl-l 4.5). The volume was maintained at that of the original filtrate. Toxin ( 1 mg) was added to a lOO-pl sample ofthe dialyud enzyme prepa- ration. no galactose was released during incubation for 15 h at 24°C. Therefore. a more concentrated solution of the enzyme was from culture filtrates. The proteins from 3-week-old cultures were obtained as described in “Materials and Methods." the ammonium sulfate precipitated pellet was collected by cen- trifugation. resuspended in water. and dialyzed. Half the prepa- ration was brought to pH 4.5 with acetate buffer and the other half to pH 7.0 with phosphate buffer. The final volume was one- fortieth of the original filtrate volume. Toxin (0.5 mg/SO pl of enzyme solution) was added as substrate and the mixture was incubated at 24°C for 20 h. The reaction was stopped by adding eight volumes of methanol. A precipitate was removed by centrif- ugation and a portion of the solution chromatographed on thin- layer silica plates with acetonezwater (9:1). Toxin. toxoids. and galactose were made visible by spraying the plates with an indi- cator (diphenylamine. 2 g; aniline. 2 ml; phosphoric acid. 10 mt acetone, 90 ml) and heating to 100°C. The enzyme preparation at pH 4. 5 cleaved 1 unit of galactose from approximately 25% of the toxin molecules. producing toxoid 111. A trace of toxoid II was detected (produwd by the removal of l galactose unit from toxoid 111): no toxoid 1 was detected. When the reaction mixture was incubated at 24°C (pH 7.0) or at 0°C (pH 4.5 and 7.0). only a trace of galactose was released and a trace of toxoid 111 was detected after 20 h. The enzyme preparation was held at 100°C for 20 min. then was incubated with toxin at 24°C for 20 h; again. only traces of galactose and toxoid were detected. The data show that the culture filtrate contains small amounts of an enzyme capable of hydrolyzing the fi-l.5-galactofuranoside linkage in HS toxin and in the toxoids. The enzyme had maximum activity below pH 5.0; activity dropped rapidly at pH levels above 5.0. The pH of culture fluids was 3.5 at 21 d and increased to pH 5.0 at 32 d. Thus. conditions in culture favored enzyme activity. yet there was insufficient activity in the culture fluids to account for the rapid decline in toxin cones-tration which occurred from Instr-n Wits/Mature") 3 Weeks FIG. 2. fi-Ga‘lactofuranosidaae activities in culture fluids (0) and my- celium(.)ofH.sacehariovera6-weekperiod Culturesweregrownin Roux bottles. each containing 200 ml ofmedium. LIVINGSTON AND SCHEFFER Plant Physiol. Vol. 72. 1983 A w); .. .. 1qu ninja); Detector Response U 14,1 5 Minutes FIG. 3. GLC of products resulting from enzyme hydrolysis of HS toxin. Products are galactose (gal) and toxoids 1. 11. and Ill (1. 2. and 3. respectively). Determinations were made with trimethylsilyl derivatives. using a temperature regime of 130°C for 1.0 min followed by increases of 30°C/min up to 350°C. which was held for 5.0 min. A. Substrate control; 3. products obtained when the enzyme preparation was diluted 1:4; C. products when enzyme preparation was diluted 1:2; D. products when aizyme preparation was not diluted. 21 to 32 (1 (Fig.1). The amount of fl-galactofuranosidase in the fungal mat and in culture filtrate was then determined at weekly intervals for 6 weeks (Fig. 2). Enzyme activity in the culture filtrate was barely detectable during the first 4 weeks: values ranged from 10 to 100 units/bottle. By week 5. the pH of the filtrate was above 6.0 and most of the toxin was gone. The amount of enzyme in the culture solution remained below the level that would be required to convert toxin to toxoids at the observed rate. There was substan- tiallymoreenzymeactivityinthemyceliumthanintheculture fluids; activity was detected in l-week-old cultures. and reached apeak at4weeks(Fig. 2). However.wewerenotabletodetermine the amount of B-galactofuranosidaae activity in the intact myce- lium; therefore. we could not determine losses during preparation of enzyme. ConverslonofToxintoToxolds Ia thyfl-Galaetofurano- ddaae. HS toxin was hydrolyzed with a highly purified prepara- tion of B-galactofuranosidaae which was kindly provided by Dr. .1. E. Gander. At pH 4.6. the enzyme cleaved galactose from toxin. producing all the toxoids. The enzyme was prepared from the mycelium of 4-week-old cultures of H. sacchari. using the method which included chro- matography on a Bio-Gel P-150 column (see “Materials and Methods"). Several dilutions of the tion were allowedtoreactwith toxin(1.0mg)ia tooplacetate buffer(200 mu, pH 4.6) at 37°C for 18 h. The reaction was stopped by adding ENZYMIC CONVERSION OF H. SACCHAR! TOXIN Table 11. B-Gulacrofiirmoddase Activity in Culture Mir-ares and Fungal Mars of Hebuhuhoqioruau and Other Species Thefungiweregrowninstillculture (20mlmodified Friessolutionin 125-ml flasks) for 23 d at 22°C. Culture Filtrates Mycelial Mat F ‘ isolate r 8 No. Units/ Specific Units/ Specific flask activity flask activity’ H. neydr‘s race T l 9.21!) 3.1 3.200 3.2 H. mam race T 2 13.600 3.2 5.300 4.5 H. maydtt race '1' 3 34!!) 1.0 2.500 2.0 H. whom race 3 5.81!) 5.0 18.0“) 11.8 H. whom race 1 1.920 5.0 12.750 5.8 H. micron race 2 12.11!) 5.5 11,750 6.4 H. Victoria: 244.1113 1.8 12.250 4.8 H. sacchari‘ l 80 0.08 1.800 0.45 H. sacchari 2 180 0.05 775 0.65 H. sacchari 3 40 0.05 775 0.24 H. sot-chat 4 160 0.05 425 0.63 H. sacchari 5 20 0.05 1.550 0.35 H. sacchari 6 60 0.07 1.250 0.39 H. sacchari 7 60 0.09 1.425 0.27 H. sacchari 8 60 0.05 1.550 0.71 H. sacchari 9 40 0.05 1.400 0.30 H. aacdlari 10 20 0.05 1.6“) 0.55 I". estimation ND‘ ND C. cum ND ND ' Enzyme units/pl divided by the pg protein/p1 of the enzyme prepara— tion. ° isolates of H. sacchari are from Florida (1-6). Hawaii (7-8). and Australia (9-10). ‘ None detected. 0.9 ml of methanol and the precipitate was removed by centrifu- gation. An aliquot of the solution was dried. MciSi derivatives were prepared. and the derivatized products were separated by GLC. The highest concentration of enzyme converted all the toxin to toxoids and galactose within 18 h (Fig. 3); most of toxoid 111 also was hydrolyzed. Less toxin was hydrolyzed by lower concen- trations of the enzyme. There was no evidence of galactose dimers in the reaction solutions. indicating that the enzyme cleaves only the terminal unit of galactose. as does the enzyme described by Rietschel-Berst er al. (9 ). Toxoid 1 did not increase to high concentrations under these reaction conditions. although it does in the culture filtrates. The small peaks adjacent to the large galactose peaks were identified, by comparison with standards. as two sesquiterpenes which were found previously as acid hydrolysis products of toxin and toxoids (4). This indicates that some of the toxin was completely hydrolyed, liberating galactose and the aesquiterpenoid core of the toxin. Production of fl-Galactoflranosidaae by Several Hm riauandOtherSpecles. Large amounts ofB- furanosidase have been isolated from culture filtrates of Paucilli'um charlesii (9). in contrast to the low activity found in filtrates of H. sacchari Accordingly. several Helminthosporia and other fungi were tested for B—galactofuranosidase activity (T able II). The fungi were grown in stationary culture for 23 d in 125-ml Erlenmeyer flasks. each containing 20 m1 of modified Fries medium. Enzyme solu- tions were prepared from filtrates and from fungal mats. Assays showed that pH 4 to 5 was optimum for activity of the enzymes from each species. The levels of B-galactofuranosidase activity in the culture fil- trates of each of four species of Helminthosporium (H. mavdi's. H. carboman. H. victoriae. and H. Mcicum) were much greater (11- 533 680 times) than the activity in the culture filtrate of H. sacchari (Table II). Fungal growth and protein levels in cultures of the several species were comparable: each isolate gave reproducible results. The amount of B-galactofuranosidase activity in the my. celium of H. sacchari was in all cases less than the amounts in the other Helminthosporium species. There was no detectable B-gal- actofuranosidase activity in the mycelium or culture filtrates of F. oxysponan or C. W. ’ DISCUSSION We reported previously that a host-selective toxin and three different toxoids. all containing galactose and a sesquiterpene. are produced by H. sacchari (3, 4). However. we did not determine the exact number of galactose units in the molecules. Mass spec- tral. NMR. and other data indicated that the toxin and toxoid molecules contained an aglycone. and that the aglycone from toxin and toxoids had the same mol wt. We also found that galactose in the toxin molecule is in the furanose form and is linked by B—1.5 bonds. We now report that there are 4. 3. 2. and 1 units of galactose in HS toxin. toxoid Ill. toxoid 11. and toxoid I. respectively. These values were determined by accurate meas- urement of the galactose released by acid hydrolysis of toxin and toxoids. The results for toxin confirm the report of Macko er al. (7). Each of the toxoids may exist in three different isomers. as indicated by Macko er al. for toxin (7). A consideration of the isomers is outside the scope of this report. GLC data showed a rapid drop in toxin concentration in H. sacchari“ cultures from week 3 to week 5. Electrolyte leakage assay also showed that toxin levels peaked at 21 d. and declined to nondetectable levels at 42 d. Sucrose in the medium was depleted by 15 d. and fungal growth stopped by 18 (1 (data nor given). As toxin concentrations dropped. there was an increase in the amount of toxoids in the filtrates. The toxoids could result from enzyme or other breakdown of toxin. but there are other possibilities. Toxin and each of the toxoids could be synthetic end-products. or toxoid synthesis and enzymic breakdown of the toxin could occur concurrently. Cultures always contained toxoids when toxin was present. suggesting that toxin and toxoids could be synthetic end- products. Toxin was shown to be converted to toxoids by a B-galactofur- anosidase that was detected in cultures of H. sacchari. However. no enzyme activity was detected when dialyzed preparations of the culture fluids were added to purified toxin at concentrations up to 100-fold greater than that originally present in culture fluids. Thus. there was not enough enzyme activity in culture fluids to convert toxin to toxoids at the observed rates. B-Galactofuranos- idase activity was detected when proteins in culture filtrates were precipitated with ammonium sulfate and dissolved in a small volume of water. Conditions for detection included concentrations of enzyme and toxin that were 4.000-fold greater than those found in filtrates of 3- to 5-week-old cultures. Under these conditions. only one-fourth of the toxin was cleaved; thus. the enzyme prep- aration from the culture fluids had <0.1% of the B-galactofura- nosidase activity required to hydrolyze toxin at the rate that it disa from cultures. If B-galactofuranosidase activity leads to loss of toxin from cultures. then the enzyme must be associated with the mycelium. When toxin loss from culture fluids was most rapid. 98% of the B—galactofuranosidase activity was in the mycelium and only 2‘7: was in the fluids. Even the activity in the mycelium may not be sufficient to account for a conversion of toxin to toxoids at the observed rate: thus. other mechanisms may be involved. Perhaps toxin or toxoid are brought together with the enzyme at particular locations in the cell. Another possibility is that the enzyme is attached to the cell wall and can convert toxin to toxoids without movement through the plasma membrane; this has not been examined. Finally. we were not able to determine the amount of 17 534 B-galactofuranosidase activity lost during the isolation of the enzyme from the mycelium; this could account for the shortage of enzyme needed to convert toxin to toxoid. B-Galactofuranosidase has been reported from very few micro- organisms; therefore. we examined several plant pathogens for ability to produce the enzyme. isolates of H. maydr’s. H. earbornon. H. trimester. and H. vieton'ae accumulated 11- to 680-fold more B-galactofuranosidase activity than did any one of the 10 isolates of H. sacchari. The enzyme activities in both the mycelium and the culture fluids of all tested isolates of H. sacchari were far less than the activities of the other Hebnr'nthospormm species. In contrast. cultures of Cladomorr‘um cucwnerr'mmr and PW oxvsponan contained no detectable &g-l=~nftxnm "‘ activity, indicating that the enzyme is not ubiquitous among fungi. Further work is necessary to determine the significance of high enzyme production by some species. low production by others. and no production by still others. The data do not prove that HS toxin is absent from cultures of some Helminthosporia because enzyme levels are high. It seems likely that the aetivity of B-galactofuranosidase in young cultures of H. sacchari is low enough to allow HS toxin to accumulate. The enzyme probably contributes to the disappear- ance of toxin in mature cultures. The possible production and significance of,‘3 ;-'“‘ J" "‘ ‘ indiseaaedtissueremainsto be determined. ' AWN—Wewishwthankthefollowmgpcoplefmtheirhelpzm Hancoekforexperttechntalassistance; J E Gander. Depanmentofhtochemsstry UniversityofMinnesotaforsamplesofB- «JackCCcmstoek (HawauSugarPlanters). JackDeaanDA. CanalPomt. FLLandOwenSturgem LIVINGSTON AND SCI-IEFFER mmudmwmmcopilly.mmlfor mauw Plant Physiol. Vol. 72. I983 LITERATURE CITED LinemanuulfltiArapidandsenuvemsthodfor theqnantitauonof mierogramquentiti-ofproseinnsihnagtheprmdpleofplm-dyebmding Anall'tochem72;248—254 2. Ftscrrsz. JZanlWQuanutauveBesummungderGalaktosemittelsGalak- moxydueausboetylamm. LZPhysiolCham337: 136-195 3. Ltvmosronlts. ”Saturn 198! lsolationandeharacta'uauonofhost dearvetoainfromHMawu-uucemnytopathologyfl. 237(Ahstr no.449. l980meetingsofAmPhytopatholSoc) WomnRS.RPSaml9811solauona-dcharacteriaaticnofhost- selective toxin from HWmccha-r. lhtolChem256; 1705-1710 .lJerosronllS "Scan-ml”! Fungal produasch-nnllyrelatedto Han-W toxinprctefl sugarcane nastier from the toxin Phytopathology 71 891(Abstr) 6. Lrvnwsron RS. RP 5cm 1982 W aenvny of Hel- mho ' apeetescnnvertsH.meeh-r'tcamtotoaoids.l’hytopathology 72:933(Abstr) 7. MacxoV. lthooonm. TWmJAAmeranAchm. DAaroom 1981 Charlaeruauonofthehoa-nlectrvemproducedvadn-arhospo MMIhecausalcrgan-nofeympotdheaseofmgamne. Eaperienua 37: 923-924 8. Patricia”. ”Scum 1963Purificatioaofthesalecuvetoxinofhrm ml’hytopathologyfl: 785-787 9. karma-Dunn NHeror-r. PDMChncmf-‘Fano. IEGsnnn 1977Enracllularaeflr‘ " fromfaxulamchataulbiol Qan252: 3219-3226 10. mn.RSUVmosro~l9w5msiuvnyofugarcanedonestotoxinfm umwuwsymw.w. 11.8rumOWMRSBml971Putnlehamuerinumndueofahon-selecuve tcamfrom HWWonnmrane.Phytopthology61:691- up 695 12 SNOW. GASraonnl'fllfldminthoepuuaahoat-qedfletoamfrom I‘m-ochre WJMMZ“: 4350-4357 EXPERIMENTAL III TOXIC AND PROTECTIVE EFFECTS OF ANALOGS OF HELMINTHOSPORIUM SACCHARI TOXIN 0N SUGARCANE TISSUES 18 TOXIC AND PROTECTIVE EFFECTS OF ANALOGS OF HELMINTHOSPORIUM SACCHARI TOXIN ON SUGARCANE TISSUES Abstract Host-selective toxin and three groups of smaller, structurally related compounds were isolated from culture fluids of Helminthosporium sacchari. The lower mol wt compounds (toxoids) differ from HS toxin in the number of galactose units per molecule. Toxoids protect sugarcane tissues from toxin, as shown by use of the assay based on toxin-induced loss of electrolytes. Toxoid III (3 units of galactose/molecule) was most effective at preventing toxicity; it gave 90% protection at a 24:1 ratio, on a molar basis. Toxoid II (2 units of galactose) was intermediate, and toxoid I (1 unit of galactose) provided the least protection against HS toxin. Protection was apparent when tissues were exposed to toxoid III for one h, rinsed to remove free toxoid. then exposed to toxin. The data suggest that toxoids act as competitive inhibitors of HS toxin. This is consistent with the hypothesis that toxin binds to a receptor molecule in susceptible cells. Toxoids I and II induced no loss of electrolytes from any of the tested clones of sugarcane. Toxoid III caused losses and runner lesions on clones NG 77-234 and NG 77-82, but not on clones Co 453 and NG 77-103; all four clones are equally susceptible to E. sacchari. 19 20 INTRODUCTION Helminthosporium sacchari (Van Breda deHaan) Butler, the cause of eyespot disease of sugarcane, produces in culture a host-selective toxin (HS toxin) (13) and several structurally related compounds (3,4,7). Clones of sugarcane that are highly susceptible to the pathogen are sensitive to H5 toxin and clones that are highly resistant to the pathogen are insensitive to the toxin. HS toxin, which was initially characterized as Z-hydroxycyclopropyl-a-D-galactopyranoside (14), has now been shown to contain a sesquiterpene and 3 1->5 linked galactofuranose units (4). Macko et al. (9) proposed a structure for HS toxin which contains two galactose units on each side of an aglycone residue (015H2402), which can exist as three isomers. The three forms of the selective toxin were reported to cause runner lesions on susceptible sugarcane leaves. The non-toxic compounds related to HS toxin can be produced by removal of galactose units from toxin. There are several isomeric forms of HS toxin and therefore of each of the lower molecular weight compounds. The lower mol wt compounds were called toxoids because they reduce the effects of HS toxin on susceptible tissue (5), and because they are structurally related to the toxin. Use of the tenn "toxoid" for these analogs is explained and justified in an earlier paper (7). An enzyme with B-galactofuranosidase activity has been isolated from cultures of H. sacchari (6,7). This enzyme may be, in part, responsible for removal of galactose from HS toxin and for accumulation of toxoids in culture fluids of the fungus. Toxoids I, II and III represent groups of isomeric forms that contain 1, 2 and 3 units of galactose, respectively (7). The primary objective of this study was to determine the 21 comparative effects of pretreatment with the three toxoids on toxin-induced losses of electrolytes. The toxoid most similar to H5 toxin in structure gave maximum protection. To date, there are no data on production of toxoids in infected tissue. MATERIALS AND METHODS Isolates of H. sacchari were obtained from Jack L. Dean of the USDA Sugarcane Research Station at Canal Point, FL, Jack C. Comstock of the Hawaii Sugar Planters' Association, and Owen Sturgess of the Bureau of Sugar Experiment Station, Indooroopilly, Queensland, Australia. Stock cultures of the fungi were maintained on cane leaf agar (16). Cultures for toxin and toxoid production were grown in a modified Fries solution (10). Sugarcane clones NO 77-82, NG 77-234, NO 77-103, CP 73-1000 and Co 453 were obtained from J.L. Dean. Clones H52-4610 and H50-7209 were obtained from G.A. Strobel of Montana State University. Plants were grown in 5 gallon plastic pots in the greenhouse at temperatures between 18 and 24°C. The youngest fully-expanded leaves from plants of uniform size (1.5-2.5 meters high) were used in bioassays. Purified toxin and toxoids were prepared by the use of activated charcoal plus thin layer, gel, and ion exchange chromatography, as previously described (4,7). Relative purity and cross contamination were monitored by GLC (7), which detects unwanted toxins and toxoids down to 1% (by wt) of the preparation. However, the procedure did not separate isomeric forms of each toxoid; thus, each preparation was a mixture of compounds with identical molecular weights. Concentrations of purified HS toxin and toxoids in solution were determined by dry weight (after 22 thorough drying at 110°C) and by measurement of the galactose released by acid hydrolysis of the toxin or toxoid, as previously described (7). The assay to measure protective effects of toxoids was based on toxin-induced loss of electrolytes; this assay is more reliable than the assay based on development of runner lesions (12). Leaf disks (1.0 cm in diameter) were cut, immediately placed in water, and held for 3 h. Eight disks were then placed in each assay vial. The standard toxoid protection bioassay involved exposure of the disks to water (2 ml) or to a toxoid solution (2 ml) for one h; an aliquot (100 pl) of HS toxin solution was then added to bring the solution to the necessary toxin concentration. After exposure to toxin for 0.5 h, the disks were rinsed several times with water and placed in 5 ml of water (leaching solution). The disks were incubated on a shaker and the conductance of the solution was determined at intervals with a conductivity meter. Toxin or toxoid-induced loss of electrolytes was based on the conductance values of leaching solutions taken after 3 h incubation, less the values for the water controls. The level of leakage varied from assay to assay because of environmental factors prior to leaf harvest (1). Therefore, all assays included a series of known toxin concentrations for standards. Any variation in this procedure is specified in the results. Assays, which were run at 22° under laboratory lighting, were replicated at least three times; conductance values for replicates varied less than 15% of the reported averages, unless stated otherwise. Abilities of toxin and toxoids to produce runner lesions on leaves was determined as described elsewhere (12,13). All experiments were repeated three or more times over a two year period. 23 RESULTS Effect of excision and preincubation on sensitivity of leaf tissues to HS .EQlifl- Variations between assays in rates of toxin-induced leakage of electrolytes were observed; accordingly, an attempt was made to determine the factors involved in variability. The rate of leakage was low when leaf disks were cut and immediately exposed to toxin. Rate of leakage was much higher when leaf disks were allowed to stand in distilled water at 23°C for 6 to 10 h prior to exposure to toxin (Fig. 1). Incubation in water for more than 10 h, prior to toxin treatment, resulted in decreases in toxin-induced loss of electrolytes. When a low concentration of toxin was used, a slightly longer incubation time may have been required to reach maximum sensitivity (Fig. 1). Greater sensitivity to toxin, expressed as higher rates of electrolyte loss and total losses of electrolytes, also was observed when leaf disks were washed for 0.5 h and held on wet filter paper (rather than on water) at 23°C fbr 6 h. Effect of toxoidgpretreatment on sensitivity of tissues to H5 toxin. Leaf disks were cut and incubated in water for 6 h, then were exposed to toxoid III (100 ug/ml) for 0, 1, 2, or.3 h prior to toxin treatment. Toxoid pretreatment consistently gave protection against toxin; prbtection was maximum when toxoid III was applied one h before exposure to toxin. However, toxoid III gave protection even when tissues were exposed simultaneously to toxin and toxoid. Prior exposure to toxoids for more than one h gave erratic results. Experiments were designed to determine whether or not maximum protection requires the presence of excess toxoid during toxin exposure. In a representative experiment, toxin at 5.0 and 1.0 ug/ml caused losses that gave conductance readings of 144 and 43 umhos above the water 24 loo ‘~.~..‘§.| 75 m .2 50 n E a I 25 - '\ I (15 2 6 IO Preincubation Time (hours) Fig. 1. Effect of preincubation in water on sensitivity of sugarcane (clone NO 77-82) leaf tissues to H5 toxin, as determined by electrolyte leakage. Leaf disks were cut and held on water for the indicated times. Disks were then placed in toxin solution or in water for 0.5 h, rinsed, and placed in 5 ml leaching solution. Toxin-induced leakage was determined by conductance of the leaching solution at 3 h. Toxin was used at 0.15 ug/ml (a) and 1.5 ug/ml (.). Standard deviation is shown by the vertical bars. controls, respectively. Pretreatment of tissues with toxoid III (100 ug/ml) for one h, followed by exposure to the toxin (5 or 1 ug/ml) plus the toxoid for 1 h, gave conductance values in the leaching solutions of 5 and 0 umhos, respectively. Thus, the toxoid gave essentially complete protection against toxin at both concentrations. This same procedure was repeated, except that the tissues were rinsed thoroughly after pretreatment with toxoid III, to remove free toxoid; no toxoid was added along with toxin. Electrolyte leakage was reduced well below that induced by toxin alone (no pretreatment with toxoid), although the protection was less than that obtained when toxoid was present in the toxin solution (Table 1). Therefore, much of the protective effect of a toxoid pretreatment appears to be retained even when excess toxoid solution is removed prior to toxin exposure. The dosage-response curve for H5 toxin-induced loss of electrolytes from clone NO 77-103 showed nearly linear increases in electrolyte losses with increases in toxin concentrations from 0.156 to 1.25 ug/ml (Fig. 2). None of the three toxoids caused loss of electrolytes from this clone of sugarcane. When leaf discs were treated with a toxoid prior to and during exposure to toxin, the amount of leakage was less than that induced by toxin alone. Toxoid III provided more protection than did toxoid II; toxoid I was the least effective (Fig. 3). Figures 2 and 3 show data from the same experiment and therefore can be compared. The standard curve of toxin-induced losses of electrolytes (Fig. 2) was used to calculate the apparent reduction in toxicity imposed by toxoids. For example, toxoid III at 100 ug/ml reduced the leakage induced by toxin at 1.25 ug/ml to the level that would be induced by toxin (without toxoid pre-treatment) at 0.08 ug/ml. This was calculated to be a 94% reduction 26 Table 1. Effect of pretreatment with toxoid III on HS toxin-induced losses of electrolytes from sugarcane leaves (clone Co 453). Leaf discs were cut, preincubated in H20, and held in toxoid solutions (2 ml, 100 ug toxoid/ml) (B, C) or water (A). Tissues were then rinsed (C) or not rinsed (B), toxin solutions were added, incubated for 1 h, rinsed, transferred to 5 ml water, and leached for 3 h. Conductance Treatment 5 ug toxin/ml 1 ug toxin/ml (uthS) (umhos) A. Water -> Toxin 144 43 B. Toxoid -> No Rinse -> Toxin 5 0 C. Toxoid -> Rinse -> Toxin 53 4 27 I 50 I ! 40 -§ 30 i E a 20 E 10 .1 Al— .31 2 .625 .95 1.25 Toxin Concentration (pg/ml) Fig. 2. Effect of HS toxin concentration on loss of electrolytes from leaf tissues of sugarcane clone NO 77-103. Leaf disks were cut, incubated on water for 4 h, exposed to toxin for 0.5 h, rinsed, and leached in 5 ml of water. The values are conductances of leaching solutions after 3 h, minus the water control. Standard deviation is shown by the vertical bars. 28 o 6.25 25 100 Toxoid (DQImI) Fig. 3. Effect of toxoid pretreatment on HS toxin-induced leakage of electrolytes from leaf tissues of sugarcane clone NO 77-103. Leaf disks were cut, preincubated for 3 h, and exposed to toxoids I G ), II (58) or III (II) for 1 h. Toxin was then added for a final concentration of 1.25 ug/ml; disks were then incubated for 30 min, rinsed, and leached for 3 h in 5 ml of water. Toxin-induced leakage is indicated by the conductance values. Each treatment was done in triplicate and the standard deviation is similar to that observed in Fig. 2. 29 in toxicity (Table 2). Toxoid III at 6.25 ug/ml gave 78% protection against toxin at 1.25 ug/ml. These percentages of protection provided by toxoids were calculated by the formula indicated in table 2. The experiment was repeated with sugarcane clone Co 453. The leaves used in this experiment were in an unusually sensitive condition; as a result, HS toxin at 0.625 ug/ml caused maximum losses of electrolytes (Fig. 4). Toxin in this experiment was used at 1.25 ug/ml. Again, toxoid III provided more protection than did toxoid II, and toxoid I gave no detectible protection against toxin (Fig. 5). However, toxoid I gave measurable protection against toxin in other experiments with Co 453, using non-saturating levels of toxin (Table 3). The data clearly show that prior exposure to toxoids can reduce HS toxin-induced loss of electrolytes. The degree of protection was directly proportional to the concentration of the toxoid used. Protection was greatest with the toxoid that was most similar to toxin. Clonal differences in protective and toxic effects of toxoids. The three toxoid groups were tested for toxic and for protective effects against HS toxin, using five sensitive clones of sugarcane (Co 453, N6 77-103, NS 77-82, NO 77-234, and CP 73-1000). Tissues were pre-treated with toxoids in the usual way and then were exposed to non-saturating concentrations of toxin. Effects on electrolyte losses were determined 3 h after tissues were exposed to toxin and rinsed. Several control treatments were included: tissues exposed to water alone; tissues exposed to toxin without toxoids; and tissues exposed to each toxoid without toxin. The three toxoids, used alone at 100 ug/ml, caused no losses of electrolytes from tissues of clones Co 453 and NG 77-103. However, toxoid III used alone at 25 or 100 ug/ml caused leakage of 30 Table 2. Effect of pretreatment with toxoids on HS toxin-induced losses of electrolytes from sugarcane tissue (clone NO 77-103). Toxin was used at 1.25 ug/ml. Protective effects of toxoids III II I Toxoid Appar. Protec- Appar. Protec- Appar. Protec- concentration toxina tionb toxin tion toxin tion (uglmI) uglml % ug/ml % ug/ml % 100 0.08 94 0.14 89 0.68 45 25 0.12 90 0.19 85 0.87 30 12.5 0.23 82 NDc ND ND ND 6.25 0.27 78 0.58 53 1.03 18 o ' 1.25 1.25 1.25 aApparent toxin = Toxin-induced leakage equivalent to that for toxin at indicated levels. See Figs. 2 and 3. 5 [Actual ToxinJ,-,[Appar. Toxinln % PT‘OtECtIO" = fACtuajl TOXTHT A 100 cNot determined. 31 lOO g...-.---%""—-———Il 75 m E ;_50 . 25 fij J .4 .312 .625 .75 L25 Toxin Concentration (pg/ml) Fig. 4. Dosage-response relationship of HS toxin on leaf disks of sugarcane clone Co 453. Electrolyte leakage is indicated by conductance values of leaching solutions. Procedures are described in Fig. 2. Standard deviations are shown. Tissues used in this experiment were in an unusually sensitive condition, because of environmental conditions during growth. 32 Toxoid (pglml) Fig. 5. Effect of toxoid pretreatment on HS toxin-induced leakage of electrolytes from leaf tissues of sugarcane clone Co 453. Toxoids I ug/ml. Other procedures are described in Fig. 3. , II (a), and III (I) were used. Toxin concentration was 1.25 Table 3. Protective effects of toxoids against activity of HS toxin, plus toxic effects of toxoid III against certain clones. Protection was measured by effects of pretreatment on toxin-induced loss of electrolytes. Conductance,(pmhos) Toxoidgpretreatment and concentrations (Hg/ml) Nonea III II I Sugarcane Toxin con- clone centration 0 25 100 25 100 25 100 nQ/ml NG 77-103 1.25 51 4.5 1 8.5 6 39 32 0 0 0 0 0 0 0 Co 453 0.625 62 11 1 21 8 55 40 0 0 0 o 0 0 0 NO 77-82 0.625 68 65 50 24 8 75 44 0 45b 51 0 0 0 0 N6 77-234 1.25 99 103 108 50 32 100 76 0 NDC 106 ND 6 ND 0 cp 73-1000 2.5 65 58 52 57 29 72 64 0 20 25 0 2 0 2 H52-4610d 100. 0 ND 0 ND 0 ND 0 0 0 ND 0 ND 0 ND 0 H50-7209d 100. 0 ND 0 ND 0 ND 0 0 0 ND 0 ND 0 ND 0 aToxin-induced electrolyte loss with no pretreatment. bToxoid-induced electrolyte loss with no exposure to toxin. cND = not determined. dClones H52-4610 and H50-7209 are resistant to H. sacchari; all others are susceptible. 34 electrolytes from tissues of clones NG 77-82, NO 77-234, and CP 73-1000 (Table 3). Toxoid III-induced losses were smaller than the HS toxin- induced losses; for example, the leaching solution for clone NG 77-82 exposed to toxoid III at 100 ug/ml had a conductance value of 51 umhos whereas the value for toxin at 0.625 ug/ml was 68 umhos. The conductance value for toxin-treated tissue (65 umhos) was less than half the value induced by a concentration of toxin which induced the maximum rate of electrolyte loss. Leaching solutions for tissues that were pretreated with toxoid III (100 ug/ml) and exposed to H5 toxin (0.625 ug/ml) had a conductance value of 50 umhos (Table 3). Therefore, the tissues of clone NG 77-82 that were exposed to toxin did not leak more than that induced by the toxoid III alone. This showed that toxoid III induced some losses of electro- lytes from tissues of clone NG 77-82, but the toxoid also protected against further toxin-induced losses. A lower level of toxoid III (25 ug/ml) induced even less electrolyte loss and provided less protection against toxin-induced loss of electrolytes. In contrast to toxoid III, toxoids II and I induced little or no loss of electrolytes from the sugarcane clones tested. Resistant clones (H52-4610 and H50-7209) were insensitive to HS toxin (100 ‘ug/ml) and to the toxoids (100 ug/ml). Tests showed that both HS toxin and toxoid III (100 ng each) caused runner lesions on leaves of sugarcane clones NG 77-82 and NG 77-234. Leaves of clones Co 453 and NG 77-103 developed runner lesions in response to toxin but not to toxoid III. Experiments are underway to isolate the 6 isomeric forms of toxoid III and to determine which forms give maximum protection and which will induce losses of electrolytes from selected clones of sugarcane. 35 DISCUSSION Greenhouse-grown sugarcane can vary in sensitivity to HS toxin, as indicated by toxin-induced loss of electrolytes. Greenhouse temperature is a major factor in this variability (1); the temperature of the greenhouse must be maintained below 24°C for tissues to have maximum sensitivity to toxin. Another source of variability was traced to the treatment of leaf disks used in bioassays. Leaf disks that were cut and exposed immediately to toxin leaked much less than did disks that were held for 6 h in water prior to exposure to toxin. Also, the minimum concentration of toxin that induced leakage in the preincubated disks was much less than the minimum required by the freshly-cut disks. All these sources of variability must be controlled for accurate data in toxin bioassays and in toxoid protection experiments. The ability of toxoids to protect leaf tissues against subsequent losses of electrolytes induced by HS toxin was examined. Protective effects of toxoids were evident when tissues were pre-treated with toxoids or treated concomitantly with toxoids plus toxins. However, pre-treatment may leave significant amounts of toxoids in intercellular spaces. Therefore, an attempt was made to remove as much toxoid as possible, by thorough washing prior to toxin exposure. Toxoids gave somewhat more protection by pre-treatment without washing, but most of the protective effect was evident in the thoroughly washed tissues. These results suggest binding between toxoid and a putative receptor; such binding could reduce the subsequent interaction of toxin with a receptor. Binding of toxin to a protein in susceptible sugarcane was claimed in earlier work (15), but attempts to repeat the crucial experiments were not successful (2). The toxoid data indicate that 36 binding sites should still be considered. However, mechanisms other than the simple competition of HS toxin and toxoid for a common receptor site could be involved. Toxoid III, which has 3 units of galactose and is most similar to HS toxin, gave better protection against toxin than did the other toxoids, as determined by toxin-induced losses of electrolytes. Toxoid II (2 units of galactose) gave protection, but was less effective than was toxoid III. Toxoid I (1 unit of galactose) which is least like toxin in structure, gave little protection against toxin. The correlation of structural similarity to toxin and ability to protect tissues against toxin-induced losses of electrolytes supports the hypothesis that the toxoids are competitive inhibitors of HS toxin. The toxin and toxoid preparations used in these studies can be separated by HPLC into several isomeric forms (9). Preliminary data indicate that the isomers may not behave identically (8). A more detailed kinetic analysis involving the HPLC-purified isomeric forms of toxin and toxoids will be necessary to determine whether or not the toxoids are competitive inhibitors of toxin. The molecular weight of HS toxin and toxoids were not known when these experiments were completed, but the values are now available. On a molar wt basis, the ratio of toxoid to toxin required for 90% protection was 24:1 for toxoid III and 140:1 for toxoid II. A ratio of 180:1 for toxoid I gave only 45% protection. These values are calculated from the data given in Fig. 3. Earlier studies (5) indicated that the toxoids did not induce loss of electrolytes from certain clones of sugarcane (CP 52-68, NG 77-103, Co 453) that are sensitive to HS toxin. When the toxoids were tested for 37 ability to induce loss of electrolytes from other toxin-sensitive clones, three (NG 77-234, NG 77-82, CP 73-1000) were found that were sensitive to toxoid III. This toxoid also caused runner lesions to develOp on these same toxin-sensitive clones, but not on the other toxin-sensitive clones, and not on H. sacchari-resistant clones. However, the natural mixture of toxoid III isomers caused less damage on a weight basis (as measured by electrolyte loss and production of runner lesions) than did HS toxin. Although toxoid III induced electrolyte losses, it also protected tissues against further toxin-induced losses of electrolytes. The various toxin analogs may be used to determine the structural requirements for maximum affinity to sensitive sites, and for toxic action. These findings could have relevance to an understanding of disease development in certain clones. 1. 3. .5. 6. 7. 8. 9. 38 REFERENCES Byther RS, GW Steiner 1975 Heateinduced resistance of sugarcane to Helminthosporium sacchari and helminthosporoside. Plant Physiology 56:415-419 Lesney MS, RS Livingston, RP Scheffer 1981 Effects of toxin from Helminthosporium sacchari on nongreen tissues and a reexamination of toxin binding. Phytopathology 72:844-849 Livingston RS, RP Scheffer 1981 Isolation and characterization of host-selective toxin from Helminthosporium sacchari. Phytopathology 71:237 (Abstract) Livingston RS, RP Scheffer 1981 Isolation and characterization of the host-selective toxin from Helminthosporium sacchari. The Journal of Biological Chemistry 256:1705-1710 Livingston RS, RP Scheffer 1981 Fungal products chemically related to Helminthosporium sacchari toxin protect sugarcane tissues from the toxin. Phytopathology 71:891 (Abstract) Livingston RS, RP Scheffer 1982 B-galactofuranosidase activity of Helminthosporium species converts H. sacchari toxin to toxoids. Phytopathology 72: 933 (Abstract) Livingston RS, RP Scheffer 1983 Conversion of Helminthosporium sacchari toxin to toxoids by 8-galactofuranosidase from Helminthosporium. Plant Physiology 72:530-534 Livingston RS, RP Scheffer 1983 Biological activities of Helminthosporium sacchari toxins and related compounds. Phytopathology 73:830 (Abstract) Macko V, K Goodfriend, T Wachs, JAA Renwick, W Acklin, D Arigoni 1981 Characterization of the host-specific toxins produced by 10. 11. 12. 13. 14. 15. 16. 39 Helminthosporium sacchari, the causal organism of eyespot disease of sugarcane. Experientia 37:923-924 Pringle RB, RP Scheffer 1963 Purification of the selective toxin of Periconia circinata. Phytopathology 53:785-787 Pringle RB, RP Scheffer 1964 Host-specific plant toxins. Annual Review of Phytopathology 2:133-156 Scheffer RP, RS Livingston 1980 Sensitivity of sugarcane clones to toxin from Helminthosporium sacchari as determined by electrolyte leakage. Phyt0pathology 70:400-404 Steiner GW, RS Byther 1971 Partial characterization and use of a host-specific toxin from Helminthosporium sacchari on sugarcane. Phytopathology 61:691-695 Steiner GW, GA Strobel 1971 Helminthosporoside, a host-specific toxin from Helminthosporium sacchari. The Journal of Biological Chemistry 246:4350-4357 Strobel GA 1973 The helminthosporoside-binding protein of sugarcane. The Journal of Biological Chemistry 248:1321-1328 Wismer CA, H Koike 1967 Testing sugarcane varieties against eyespot, brown spot, red rot and leaf scald diseases in Hawaii. Proc. Int. Soc. Sugarcane Technol. 12:1144-1153 EXPERIMENTAL IV SELECTIVE TOXINS AND ANALOGS PRODUCED BY HELMINTHOSPORIUM SACCHARI: PRODUCTION, CHARACTERIZATION AND BIOLOGICAL ACTIVITY 40 SELECTIVE TOXINS AND ANALOGS PRODUCED BY HELMINTHOSPORIUM SACCHARI: PRODUCTION, CHARACTERIZATION AND BIOLOGICAL ACTIVITY Abstract Helminthosporium sacchari (HS) toxin and several lower mol wt, non-toxic analogs (toxoids) were isolated from culture filtrates. Three isomers of the toxin (A, B, and C) were separated by HPLC. Each differed from the others in relative toxicity to susceptible sugarcane; resistant sugarcane was not affected. Next, each toxin isomer was partially digested with a B-galactofuranosidase and the resulting toxoids (seven from each toxin isomer) were separated by reverse phase HPLC and identified. Each isomer of toxoid III (three galactose units/mol) also was partially digested and the arrangement of gal units was determined. One of the toxoids (A1,2, which has one gal on left and 2 on right of sesquiterpene A) was found to be toxic to certain clones of H. sacchari- susceptible sugarcane, but not to other susceptible clones. This toxoid was derived from the least active form of toxin (form A); nevertheless, toxoid A1,2 was as toxic to certain clones as was the most active form of toxin (form C). The degree of protection obtained with each of the six toxoid III isomers was determined by use of the electrolyte leakage assay. Toxoid C2,1 was more effective at preventing toxin C-induced electrolyte losses than was any other toxoid. Each isomer of toxoid III protected better than did any isomer of toxoid II (two gal units/mol). Toxoids with the 1,1 gal arrangement did not protect as well as did 41 42 toxoids with the 2,0 or 0,2 gal arrangement. Thus, the sesquiterpene isomer, the number of gal units, and the gal arrangement pattern determine the effectiveness of the compound in induction of electrolyte loss and in prevention of toxin C-induced electrolyte loss from sugarcane tissues. INTRODUCTION Four plant pathogenic fungi from the genus Helminthosporium are known to produce selective toxins (16) which are active only against host genotypes. The genotypes that are sensitive to toxin are always susceptible to the pathogen which produced that toxin. Several of these toxins have been isolated and characterized (13). It is becoming apparent that the toxins, as produced in culture, exist in several different but closely related forms (1,2,4,7). There are also toxin analogs which can be either non-toxic or selectively toxic; certain analogs have a more narrow selectivity than do the major toxins (8). H. sacchari produces toxin (4 gal/mol) in three isomeric forms. Each form contains a different isomer of a sesquiterpene (C15H2402) (11). Attached to this unsymmetrical sesquiterpene are 2 groups, each composed of two 8 1,5-linked galactofuranose units. Non-toxic, lower mol wt analogs, termed toxoids (7), have been found in the culture fluids of H. sacchari. The toxoids can be grouped by the number of galactose units in the molecule. Toxoids III, II and I were shown to contain 3, 2 and 1 units of galactose, respectively (7). A 8-galactofuranosidase produced by H. sacchari was shown to convert HS toxin to toxoids in XiEEQ- The studies also suggested that this enzyme probably converts toxin to toxoids in cultures (7). Sequential 43 removal of one, two, or three galactose units from the three sesquiterpene isomers of toxin should give 21 different, lower mol wt toxoids. We have examined this situation by enzymatic removal of galactose units from each form of toxin, and have identified the resulting compounds by chromatography. A previously reported method of nomenclature for the toxoids was followed (10). The toxin identification letter (A, B, or C) refers to the sesquiterpene core isomer present in the molecule. The subscript numbers refer to the number of galactose units present and their position in relation to the core. The first number refers to galactose attached to the left side of the molecule and the second number to galactose on the right side of the sesquiterpene. For example, toxoids C2,1 and C1,2 were produced by removal of a single galactose unit from toxin C2,2. Previous work showed that toxoids with 3 galactose units protect tissues from toxin more effectively than do the 2 galactose toxoids, as shown by the electrolyte leakage assay. A small but detectable level of protection was provided by the toxoids with 1 galactose unit (9). However, in each case the three, two, and one galactose toxoids were a mixture of isomers. These mixtures of toxoids have now been separated by reverse phase HPLC. The protective and toxic ability of several of these purified toxoids will be reported here. Some of this work was reported in an abstract (8). MATERIALS AND METHODS Isolates of Helminthosporium sacchari and the sugarcane clones used were obtained from previously reported sources (7). Toxins and toxoids were isolated from cultures of H. sacchari by a modified version of the 44 procedure described previously (5). The initial separation between toxin and the toxoids was achieved on a Sephadex LH-20 column as previously described. The purification procedure was then modified by using flash chromatography instead of TLC and by including reverse phase HPLC as the final step. The toxin and toxoid content of each fraction from the LH-20 column was identified and measured by GLC (7). For flash chromatography, 30 gm of Whatman LPS-2 was poured as a slurry in acetonitrile into a column giving a bed with dimensions of 2.2 X 19 cm. A methanol solution of the preparation to be chromatographed was mixed with 1.5 9m of LPS-2 and gently stirred while the solvent evaporated. The final powder was applied to the top of the column bed and the sample was eluted, under pressure, with stepwise gradient solutions consisting of 100 ml volumes of acetonitrile-water in the following ratios: 100:0, 97.5:2.5, 95:5, 90:10, 85:15 and finally 80:20. Fractions (9 ml) were collected and monitored for toxin and toxoid content by TLC as previously described (5). A Varian 5000 Liquid chromatography instrument equipped with a Waters pBondapac C13 column (0.78 X 30 cm) was used for HPLC. Toxins and toxoids isolated from the culture fluids were chromatographed with acetonitrile-water mixes, as follows: for toxins, 20:80, for toxoids III, 19:81, for toxoids II', 20:80, and for toxoids II, 26:74. The flow rate was 2 ml/min. The change in absorbance of the eluate at 215 nm was monitored and fractions were collected manually. A second method was used for obtaining purified toxoids for determination of biological activities. HPLC-purified toxins A, B, and C (60 mg of toxin C, 20 mg each of toxins A and B) were partially hydrolyzed at 37° for 1.5 hr in a 1 ml solution of 8-galactofuranosidase from Penicillium charlesii. The 45 resulting solution of toxoids was evaporated to dryness, the residue was dissolved in 26% acetonitrile, and the solution chromatographed by HPLC as described above. The eluate containing each toxoid isomer was collected, concentrated, and dissolved in methanol for storage. To determine the concentrations of purified toxin and toxoids in solutions, the amounts of galactose released by acid hydrolysis were measured by a modified version of the procedure by Fischer and Zapf, as previously described (7). The calculations were based on molecular weights of toxin and toxoids as follows: 884 for toxin (4 gal/mol); 722 for toxoids III (3 gal/mol); 560 for toxoids II (2 gal/mol); 398 for toxoids I (1 gal/mol). Purified toxins A, B and C and each isomer of toxoid III were partially digested using a-galactofuranosidase isolated from Helminthosporium sacchari, H, maygis (7), or Penicillium charlesii (14). This enzyme was partially purified from the culture fluids of g. charlesii and H. maygis, and from the mycelium of H. sacchari (7). The procedure involved concentration of the crude enzyme solution (160 ml) to a volume of 32 ml, precipitation of the proteins with ammonium sulfate at 75% of saturation, and extensive dialysis of the resuspended enzyme pellet against 2.5 mM acetate buffer (pH 4.6). The volume of the final preparation was adjusted to 20 ml. Toxins A, B, and C (300 ug of each) were dissolved in 270 pl of 2.5 mM acetate buffer (pH 4.6) and 30 ul of enzyme solution was added. The preparation was partially digested at 37° for 40 minutes for toxin C, and 30 minutes each for toxins A and B. The reaction was stopped by adding 0.5 ml of methanol and the preparation was evaporated at 45° under a jet of nitrogen. The residue was then dissolved in 300 pl of 20% acetonitrile in water. Aliquots (50 ul) of 46 the hydrolyzed solution of each isomer of toxin were then subjected to HPLC, using a program of 0-5 minutes acetonitrile-water (20:80), followed by a 10 min linear gradient to 27% acetonitrile and was then maintained at 27% for 20 min. Other conditions for HPLC were as stated above. The six toxoid III isomers were also purified from culture fluids. A sample of each isomer (50 pg) was then dissolved in 90 pl of acetate buffer (2.5 mM, pH 4.6) and 10 pl of the Penicillium enzyme preparation was added. After 25 minutes at 37°, the sample was prepared for HPLC as was done for the partial digests of toxin. The residue was dissolved in 80 ul of 20% CH3CN. Aliquots (40 pl) were chromatographed by HPLC using the same gradient used for analysis of the partial digests of the toxins. The biological activities of the toxins and toxoids were measured with the electrolyte leakage assay (15). The assay is based on induction of leakage from susceptible leaf tissue by toxin, and prevention of toxin-induced losses by the toxoids. Leaf disks that were cut from young but fully expanded sugarcane leaves were preconditioned by incubation in water for 6 hr (9). In protection assays the toxin and toxoid solution was pre-mixed to provide simultaneous exposure of the disks. After 0.5 hr of exposure to the test solution, the disks were rinsed and placed in 5 ml of water (leaching solution). Toxin or toxoid-induced loss of electrolytes was based on the conductance values of leaching solutions taken after 3 hr incubation, minus the values for the water control. The ability of toxin and toxoids to produce a runner lesion was tested as previously described (15). Resistant tissue controls were used in all assays. 47 All solvents were redistilled, acetonitrile was HPLC grade, and water was double distilled. Each treatment was done in triplicate and each assay was repeated. All numerical values are averages from one experiment. The variation for each treatment never exceeded 25% of the average. RESULTS Isolation and identification of toxins and toxoids. Toxoids obtained by removal of galactose (gal) are more hydrophobic than is the parent toxin. Sephadex LH-20, being lipophilic, gave a more dramatic separation of the toxin and toxoids than was achieved on a gel permeation column. Therefore, chromatography on an LH-20 column was used as an early step in the purification procedure. GLC of an aliquot of each fraction was used to measure the toxin and toxoid elution profile from the LH-20 column (Fig. 1). Aliquots of fractions from each peak were also subjected to HPLC. The results indicated that the LH-20 column separated these compounds primarily on the basis of number of gal units/molecule (data not shown). All isomers of each toxoid group eluted as a single peak. One exception was the peak labeled toxoid II'. This peak contained the 3 sesquiterpene isomers of toxoid II that had each gal unit attached directly to the sesquiterpene (A1’1,81,1,C1,1). Sesquiterpene and gal conformational isomers of each toxoid group were separated successfully by use of reverse phase HPLC. HPLC was used also for separating the 3 isomeric forms of toxin (Fig. 2) which differ from each other in their absorbance at 215 nm. The activities of the three forms of toxin were compared using the electrolyte leakage bioassay. All three forms were toxic to susceptible 48 Fig. 1. Elution pattern of HS toxin (0) and toxoids III (0), II' (I), II (A), and I (CD from a Sephadex LH-20 column (80x3 cm). The eluent was 50% methanol and 7 ml fractions were collected. An aliquot of each fraction was derivatized and the toxin and toxoids present were determined by GLC (5). The height of each peak indicated amounts. 15 49 ‘0 2 1n asuodsea 10139180 77 85 93 I 101 109 69‘ Fraction Detector Response 50 Au n C23 32.2 5 IO 15 2 Minutes 0 Fig. 2. Elution of the three forms (A, B, and C) of HS toxin from reverse phase HPLC following application of 10 pg each to the column. The eluent was 20% acetonitrile in water and the flow rate was 2 ml/min. Absorption at 215 nm was monitored. 51 sugarcane but had no measurable effect at 100 pg/ml on resistant sugarcane. Toxin C was more active than were the two other forms and toxin 8 was slightly more active than was toxin A (Fig. 3). Saturating concentrations of each toxin induced the same rate of electrolyte loss for the first 4 hr following exposure to toxin (data not shown). The large number of toxoids found in the culture fluids of H. sacchari made separation and identification difficult. To simplify the procedure, toxins A, B, and C were separated and then partially digested with the 8-galactofuranosidase from Penicillium charlesii. The chromatographic profile of each partial digest of each form of toxin (A, B, and C) should contain seven toxoid peaks (Figs. 4, 5), each with the same sesquiterpene isomer. However, the two toxoid I isomers with the A form of the sesquiterpene had nearly the same retention times and eluted as a single peak; as a result only six toxoid peaks were seen. Removal of either of the two terminal galactose units from each form of toxin produced two isomers of toxoid III (Fig. 5). The following question was then posed: how is the galactose arranged in each of the toxoids? This was answered in part by use of partial digests of each of the six toxoid III isomers and identification of the resulting toxoids (Table 1). These data do not tell whether the galactose is attached to the left or the right side of the sesquiterpene core; i.e., the data do not tell whether the galactose is arranged 2,1 or 1,2. The data only show that two galactose units are on one side and one galactose unit is on the other side of the sesquiterpene. Because of this, we have adopted the galactose arrangement for the toxoid III isomers as indicated by Macko (10, and personal communication). 52 /. / ///-/ l/I :/ . 1 . 1.5 2.5 3.5 4.5 Concentration log nM Fig. 3. Electrolyte leakage induced by each of the three forms of HS toxin. Leaf disks of sugarcane clone NG 77-234 were exposed to téxin for 0.5 hr, rinsed, and incubated in 5 ml of water. Conductance of the leaching solution was measured after 3 hr. 53 Fig. 4. Toxoids produced from toxin C by enzymatic removal of galactose units. The sesquiterpene in toxin A [ QXCH/ ] and B [ EjCRK ] differs in arrangement of a double bond (11). Toxins A and 8 each produce a set of toxoids comparable in galactose arrangements to those shown above. 54 -c-Pem too o .u 5’ \ . J \ pam-oi—mm-oi ioipmmioipam N6 . o o Nu I... I \ I pmmioipmmioi pmm-oi -38 8-8-9.8 N J _ Nu L‘n _ ..r _em-o-_eu-oi -o-Fee-o-Fem N No 55 Fig. 5. HPLC separat1on of toxoids resulting from partial digestion of each of the 3 forms of HS toxin. The B-galactofuranosidase produced by E. charlesii was used for digestion. The letter in each panel indicates the sesquiterpene isomer present in the toxin and toxoids. Each peak is labeled to indicate the number and arrangement of galactose in each compound. HPLC conditions are described in material and methods. 56 esuodeen 1o1aeaea 25 IS 20 Minutes IO Table 1. each of the six isomers of toxoid III, as determined by peak heights. 57 Relative amounts of toxoids obtained from partial digestion of Toxoid III was digested and the resulting toxoids were separated by HPLC and identified by their retention times. the six isomers of toxoid III were determined by Macko (10). The galactose arrangement for Substrate A2,1 A1,2 B2,1 B1,2 C2,1 c1,2 Retention times (min) Toxoid Assignments Retention times (min) Toxoid Assignments Retention times (min) Toxoid Assignments Peak Heights (cm) and Toxoid 15.23.16; 10 O 17.3 22.1 22.8 1 30 1 O A1,1 A2,0 19.1 22.9 0.25 13 1 0 B1,1 B2,0 19.8 24.4 0 9.5 I10,2 25. 0 11 '30,2 26.1 Assignments1 25.1 8.5 27.9 25.2 0 410,1 gee 0 B0,1 29.4 4 20 7 0 c1,1 c2,0 12. C0,2 10.5 4.5 C1,0 1.0 6.0 Co,1 1See text. 58 The arrangements of galactose units in toxoids I and II were ascertained from the arrangement in the toxoid III isomers. For example, removal of one of the terminal galactose units from toxoid C2,1 will produce C2,O or C1,1 but not C0,2 (see Fig. 4). Therefore, the toxoid III isomer not present in the partial digest of toxoid C2,1 must be toxoid CO,2° The same logic was used to identify toxoid C2,0 after partial digestion of toxoid C1,2. The toxoid III isomer that was present in the partial digests of both 02,1 and C1,2 must be toxoid C1,1. Next, the galactose arrangement of the two toxoid I isomers containing the C sesquiterpene was determined. The predominate isomer of toxoid I in the partial digest of 02,1 must be C1,0. This is because it is produced by further hydrolysis of both C2,O and some of the C1,1. The small amount of C0,1 is produced only from C1,1 because no CO,2 was found in the partial digest of toxoid C2,1' Again, the same logic was used to identify tOXOId Co,1 in the partial digest of toxoid €1,2- The galactose arrangement for the toxoids resulting from digestion of the other forms of toxoid III, which contain the A and B sesquiterpene isomers, were determined by the same logic that was used for the C isomers. Each isomer of toxin (A, B, and C) was also partially digested with the B-galactofuranosidase isolated from E. 091915 and fl. sacchari. The same toxoids, although in different relative amounts, were produced using these enzymes as the toxoids produced with g. charlesii enzyme. Biological activitygof the toxoids. We previously reported that the natural mixture of toxoid III isomers induced loss of electrolytes and produced runner lesions on some but not all clones of sugarcane that are sensitive to the 4-gal toxin molecule (9). The 5 clones of sugarcane used in this experiment were highly susceptible to the pathogen and 59 sensitive to the 3 forms of toxin. Each form of purified toxin was partially digested, to identify the toxic moiety in the toxoid III preparation. The resulting mixtures of toxin and toxoids were separated with reverse phase HPLC (20% CH3CH, 2 ml/min) and fractions (1 ml) were collected. Aliquots taken from each fraction were bioassayed on sugarcane clones Co 453 and NG 77-234. The digests of toxins B and C contained no toxoids that caused damage to any of the five clones; only those fractions containing undigested toxin (4 gal units/mol) induced loss of electrolytes. In the partial digest of toxin A, aliquots of fractions containing the toxin (4 gal/mol) caused loss of electrolytes from clones NG 77-234 and Co 453 (Fig. 6). No other fractions induced electrolyte loss from Co 453. Those aliquots taken from fractions containing toxoid A1,2 (but not from any other toxoid-containing fractions) induced electrolyte loss from clone NG 77-234 but not from Co 453. As a control, undigested preparations of each toxin were chromatographed as above, and as expected only fractions containing the toxins (4 gal/mol) induced electrolyte loss. This confirms that toxoid A1,2, produced by the action of 8-galactofuranosidase on toxin A, is the toxic moiety in the toxoid III mixture of isomers. Toxoid A1,2, purified from culture fluids of H. sacchari, was tested for ability to induce electrolyte loss from several clones of sugarcane. Toxoid A1,2 was toxic to clones NO 77-82, NG 77-234, and CP 73-1000 but was not toxic to Co 453 and NO 77-103. The 5 non-toxic isomers of toxoid 111 can act as inhibitors of A1,2 (data not shown). The electrolyte loss- inducing ability of toxoid A1,2 was compared to its parent molecule (toxin A) and to toxin C, the most active form of toxin. Although toxin A was less active than was toxin C, the removal of the left terminal 60 Fig. 6. Bioassay of toxoids resulting from partial digestion of toxin A. The toxoids and undigested toxin were separated by HPLC and fractions (1 ml) were collected. Aliquots (10 pl) of the fractions were assayed with sugarcane clone Co 453 (O) and NG 77-234 (0). (a) Eluate from digested preparation. (b) A preparation identical to that chromatographed in (a) was partially digested and the resulting toxoids were separated by HPLC. The bars (i.e., A2,2) indicate the peaks that were present in each chromatogram. The solvent was 20% acetonitrile in water and the flow rate was 2 ml/min. cotuoi cc an on em an on an em .I.I.|.Io Iolo|.|.l.la 0|. / 2 fl/ 4 .l. .x._ .mN . On 3 O A7U\O 1 m“ ”I n D o n a 0 D J v w < N w < r N < N N < w OLIOI ”I'MIJMIOIIM '0 ”am, a T. /e / 1 on 0' OO— 31 Io «.u < 62 galactose unit resulted in a compound (A1,2) which was about as active on the three sensitive clones as is toxin C (Table 2). Protective effects of toxoids. It is apparent that the positions of galactose and the arrangement of double bonds in the sesquiterpene core are very important in determining toxicity. 00 these structural charac- teristics play a significant role in determining how effective a toxoid will be at protecting tissues against toxin? In our previous examination of the protective abilities of the natural mixtures of toxoids we omitted one group of toxoid II isomers (9). This natural mixture of toxoids, designated 11', contains the isomers A1,1, 81,1, and C1,1. They elute from the LH-ZO column as a single peak that is separate from the peak containing the other 6 isomers of the toxoid 11 group (Fig. 1). The protective abilities of the natural mixture of toxoids II', II and I were compared using the standard protection bioassay (Table 3). Toxoid 11' required four-fold higher concentrations than did toxoid II to give equal protection. Although all toxoids II and II' contain 2 galactose units, their arrangement plays a significant role in their effectiveness as inhibitors of toxin. Toxoid I, with only one galactose, provided significantly less protection than did toxoid II'. The six isomers of toxoid III were tested for their protective ability on 2 clones of sugarcane (Table 4). As a control, each toxoid was again tested for ability to induce electrolyte loss on both sugarcane clones. Only toxoid A1,2 induced loss of electrolytes from clone NG 77-234; no toxoid isomer induced losses from NG 77-103. In general, the toxoids protected against toxin C more effectively on clone NG 77-103 than on any other clone tested. Toxoid C2,1 was consistently more effective than was C1,2 Or any other toxoid at protecting against 63 Table 2. Electrolyte leakage from sugarcane clone NG77-234 induced by toxins (A2,2 and 02,2) and toxoid A1,2. Leaf disks were exposed to test solutions for 0.5 hr, and rinsed thoroughly. Conductance values were taken after 3 hr incubation in leaching solution. Concentration (uM) Losses induced by A1,2 A2,2 C2,2 (pmhos) (pmhos) (pmhos) 0.1 30. 4. 35. 1.0 109. 82. 116. 10.0 142. 136. 134. 64 Table 3. Comparative abilities of three natural mixtures of toxoid isomers to protect sugarcane tissues (clone NO 77-234) against toxin C. Toxoid I is a natural mixture of isomers with one galactose unit. Toxoid II' is a natural mixture of the isomers A1,1’ 81,1 and C1,1. Toxoid II is the natural mixture of isomers with galactose arranged 2,0 and 0,2. Leaf disks of clone NG 77-234 were exposed to toxin and toxoids simultaneously for 0.5 hr, rinsed, and incubated in the leaching solution for 3 hr, when conductance was determined. Toxoid Concentration Losses induced by toxin-C (0.75 pm) _ guM2____. in the presence of toxoid 1 II'_ II (pmhos) (pmhos) (pmhos) 0 127 127 127 50 -a - 52 100 120 85 - 200 103 53 - 400 95 30 - aNot determined. 65 toxin C. In general the toxoids with the A sesquiterpene were slightly less effective than were those with the B sesquiterpene (Table 4). The arrangement of the gal units made little or no difference with toxoids containing the A or B sesquiterpene, but made a significant difference for toxoids containing the C sesquiterpene. The effectiveness of toxoids C2,1 and C1,2 at protecting against toxin C were tested at several toxoid concentrations. Toxoid C2,1 was more effective than was C1,2 at protecting leaf tissue against toxin C—induced loss of electrolytes at each of several toxoid concentration (Table 5). It was difficult to obtain purified preparations of the toxoid II isomers. Those toxoids with the A or B sesquiterpene were purified from the culture fluids, which had a very low amount of A0,2 and 30,2 isomers. Thus, only A2,0 and 82,0 were obtained in amounts required for protection assays. Toxoids containing the C sesquiterpene were purified from a partial digest of toxin C. The protective effectiveness of these four toxoid II isomers were tested, using the standard protection bioassay with three clones of sugarcane (Table 6). Toxoid C0,2 was consistently more effective than was C2,0 at protecting tissues against toxin C. 82,0 and C2,O were equal in effectiveness, but toxoid A2,0 was nearly as effective as was C0,2. DISCUSSION Three different isomers of HS toxin were found in preparations purified by a previously reported procedure (5). These forms of toxin were reported to differ in the sesquiterpene core (12). Each form was toxic only to sugarcane susceptible to the pathogen, with no effect on 66 Table 4. Comparative abilities of the six isomers of toxoid III to protect sugarcane tissues (clones NG 77-103 and NG 77-234) against toxin C. The treating solution contained toxin C at 0.75 pm without (none) or with toxoids at 50 pm. The standard protection bioassay was used. Toxin-induced losses from clones Toxoid preparation NG77-234 NG 77-103 (uthS) (pmhos) III mixa 39 3 A2,1 36 12 A1,2 - 13 82,1 21 4 81,2 32 6 C2,1 8 0 C1,2 38 6 None 116 87 aThe natural mixture of all toxoid III isomers. 67 Table 5. Comparative abilities of toxoids containing sesquiterpene C in protecting sugarcane tissues (clones Co 453 and NG 77-234) against toxin C. The treating solution contained toxin C at 1.0 pm without or with toxoid at the indicated concentrations. The standard protection bioassay was used. Toxin-induced losses in the presence of toxoid Toxoid concentration Clone Co 453 Clone NG 77-234 ("5) c2,1 c1,2 c2,1 C1,2 (pmhos) (pmhos) (pmhos) (pmhos) 50 3 7 16 66 25 13 27 58 102 10 44 67 106 112 O 81 81 '110 110 68 Table 6. Comparative abilities of four isomers of toxoid II to protect sugarcane tissues (clones NG 77-103, NG 77-234 and Co 453) against toxin C. The treating solution contained toxin C at 0.75 pm (for clones NG 77-103 and NG 77-234) and at 1.0 pm (for cone Co 453), without (none) or with toxoids at 50 pm. The standard protection bioassay was used. Toxin C-induced electrolyte losses from clones Toxoidgpreparations NG77-103 NG77-234 Co 453 (pmhos) (pmhos) (pmhos) II mixa 14 34 35 A2 ’0 20 49 41 82,0 34 61 58 C2,0 3O 59 56 C0,2 18 44 34 None 87 72 76 3The natural mixture of all toxoid II isomers with galactose arranged 2,0 or 0,2. 69 sugarcane resistant to the pathogen. The relative toxicity of each form of toxin was tested and compared. Each toxin isomer produced runner lesions on susceptible sugarcane and each caused electrolyte leakage from all susceptible clones that were tested. However, the three forms differed in relative severity of induced losses. Apparently, differences in the position of a double bond in the sesquiterpene core of toxin, which results in differences in shape of the toxin mol, has a major role in determining how active the toxins are at inducing electrolyte loss. The HS toxin molecule apparently has two chains of two galactose units each, attached to an unsymmetrical sesquiterpene core (12). This arrangement of gal has been designated 2,2 (12). Previous work has shown that the toxoids can be produced by enzymatically removing 1, 2, or 3 units of galactose from toxin. The earlier study was done with a mixture of the three toxin isomers (7). We have now partially hydrolyzed each isomer of toxin and identified the seven resulting toxoids. Two isomers of toxoid III were formed from each isomer of toxin. This is consistent with a galactose arrangement pattern in the toxins of 1,3; 3,1; or 2,2. It is not consistent with an 0,4 or 4,0 arrangement, which would produce only 1 form of toxoid III. Further, three isomers of toxoid II were found to be produced from each isomer of toxin. This rules out a linkage of 1,3 or 3,1, which would result in only 2 forms of toxoid II. The toxin must therefore have a linkage of 2,2. The existence of three isomers of the sesquiterpene results in toxins A2,2, 82,2, 02,2. We reported previously that a mixture of toxoid III isomers was toxic to some sugarcane clones, but not to others. Data reported here show that of these isomers, only A1,2 induced electrolyte losses and produced a runner lesion. It is interesting to note that this toxoid was 70 produced by removal of the left terminal galactose unit from toxin A, the least active form of toxin. Toxoid A1,2 is just as active at inducing electrolyte losses in certain sugarcane clones as is toxin C, the most active form of toxin. Furthermore, the fact that only some susceptible clones of sugarcane are sensitive to toxoid A1,2, whereas many other clones are sensitive to the toxins (4 gal/mol), suggests that the mode of toxicity for toxin and toxoids may be complex. Possibly the putative receptor proteins of sugarcane differ slightly, or the mechanism of leakage differs from clone to clone. The data indicate that the sesquiterpene isomer and the number and arrangement of galactose units is very important in determining toxicity of a compound. Are these factors significant in determining whether or not a toxoid will effectively prevent action of toxin? We previously reported that the natural mixture of toxoids with 3 gal units protected better than those with two gal units (9). Each of the six toxoid III isomers was tested for effectiveness at protecting against toxin C. Toxoid C2,1 protected much better than did toxoid C1,2. Apparently the location of galactose units is important in protective ability. The arrangement of galactose units in the toxoids containing sesquiterpenes A and B did not make as much different as did the arrangement in toxoids containing the C sesquiterpene. In general, toxoids wfith the B sesquiterpene protected better than did those with the A sesquiterpene. It will be interesting to use toxin 8 to determine whether or not the B sesquiterpene toxoids will protect better than will those with the C sesquiterpene. Only seven isomers of toxoid II were obtained in amounts necessary for protection assays. Toxoids with the 1,1 arrangement did not protect 71 as well as did toxoids with the 2,0 or 0,2 arrangements, even though all three contain 2 gal units/molecule. Apparently a two galactose chain is an important factor in determining how effective a toxoid will be at inhibiting the action of the toxin. Toxoid C0,2 was more effective than was C2,O at protecting against toxin C. This is in contrast to the results with toxoid III showing that C2,1 was more effective than was C1,2. The other toxoid II isomers differed in effectiveness, but none was as effective as was any of the isomers of toxoid III at preventing toxin C induced electrolyte losses from sugarcane tissue. Are our data compatible with the hypothesis that toxin binds to and changes a specific protein found in sensitive sugarcane cloneS? The data on protective effects of toxoids appear to be compatible with such a hypothesis. It is unlikely that toxin and toxoids are interacting with each other. The simplest explanation of protective effects of toxoids is that they interact with a toxin receptor, thus reducing toxin interaction with the receptor. The toxoids are not toxic; thus their interaction with a receptor apparently is not sufficient to result in electrolyte loss. The interaction of a receptor with toxin may be qualitatively different from the interaction of a receptor with a toxoid. Perhaps only the toxin, because of its specific shape, can cause a conformational change in the receptor which can result in electrolyte loss. These are tentative suggestions, because no kinetic data on toxin uptake or action are available. Whole tissue does not provide sufficiently precise kinetic data to determine whether or not the protective effects of toxoids are competitive. Better assays will be required; among the possibilities are use of cell cultures or protoplasts, which should be uniform host-cell populations. Eventually, cell-free preparations of 72 receptor proteins and radioactively-labeled toxin and toxoids will be required for a firm conclusion regarding toxic action. 73 REFERENCES Ciufetti LM, MR Pope, LD Dunkle, JM Daly, HW Knoche 1983 Isolation and structure of an inactive product derived from the host-specific toxin produced by Helminthosporium sacchari. Biochemistry 22:3507-3510 Danko SJ, Y Kono, S Kim, JM Daly, HW Knoche 1983 Purification of Phyllosticta maydis toxin. Phytopath 73:828 (Abstr) Gander JE, NH Jentoft, LR Drewes, PD Rick 1974 The 5-0-8-D-galactofuranosyl-containing exocellular glyc0peptide of Penicillium charlessii. J Biol Chem 249:2063-2072 Keen NT 1983 Purification of victorin. Phytopath 73:830 (Abstr) Livingston RS, RP Scheffer 1982 Isolation and characterization of the host-selective toxin from Helminthosporium sacchari. J Biol Chem 256:1705-1710 Livingston RS, RP Scheffer 1981 Fungal products chemically related to Helminthosporium sacchari toxin protect sugarcane tissues from the toxin. Phytopath 71:891 (Abstr) Livingston RS, RP Scheffer 1983 Conversion of Helminthosporium sacchari toxin to toxoids by 8-galactofuranosidase from Helminthosporium. Plant Physiol 72:530-534 Livingston RS, RP Scheffer 1983 Biological activities of Helminthosporium sacchari toxins, and related compounds. Phytopath 73:830 (Abstr) Livingston RS, RP Scheffer 1983 Toxic and protective effects of analogs of Helminthosporium sacchari toxin on sugarcane tissues. Physiol Plant Path 10 11 12 13 14 15 16 17 18 74 Macko V, G Grinnalds, J Golay, D Arigoni, W Acklin, F Weibel, C Hildebrand 1982 Characterization of lower homologues of host-specific toxins from Helminthosporium sacchari. Phytopath 72:942 (Abstr) Macko V, D Arigoni, 1982 Structure of host-specific toxins produced by Helminthosporium sacchari. Abstract No. 252, Am Chem Soc 183rd National Meetings Macko V, K Goodfried, T Wache, JAA Renwick, W Acklin, D Arigoni 1981 Characterization of the host-specific toxins produced Helminthosporium sacchari, the causal organism of eyespot disease of sugarcane. Experientia 37:923-924 Macko V 1983 Structural Aspects of T0xins. In Daly JM, BJ Deverall, eds, Toxins and Plant Pathogenesis. Academic Press, Sydney pp. 41-80 Rietschel-Berst M, NH Jentoft, P0 Rick, C Pletcher, F Fang, JE Gander 1977 Extracellular exo-B-galactofuranosidase from Penicillium charlessii. J Biol Chem 252:3219-3226 Scheffer RP, RS Livingston 1980 Sensitivity of sugarcane clones to toxin from Helminthosporium sacchari as determined by electrolyte leakage. PhytOpath 70:400-404 Scheffer RP 1983 Toxins as chemical determinants of plant disease. In Daly JM, BJ Deverall, eds, Toxins and Plant Pathogenesis. Academic Press, Sydney pp 1-40 Steiner, GW, GA Strobel 1971 Helminthosporosides, a host-specific toxin from Helminthosporium sacchari. The Journal of Biological Chemistry 246:4350-4357. Strobel GA 1973 The helminthosporoside-binding protein of sugarcane. J Biol Chem 248:1321-1328 GENERAL DISCUSSION Helminthosporium sacchari was shown in 1971 (4) to produce a highly active toxin that is selectively toxic to sugarcane clones (cultivars) that are susceptible to the fungus. Soon after this discovery was published, a structure of the toxin was presented (5). This work initiated Dr. Strobel's very active research program which included studies on mode of toxic action and resistance to toxin (7). The work attracted much attention, and became the most-cited model for the molecular basis of plant disease development. My dissertation research began as a reexamination of the chemical characteristics of the selective toxin from H. sacchari. Improved procedures were developed to isolate and purify the toxin (section I and IV). Characterization by NMR, MS, derivatization, and other techniques showed clearly that the earlier description of toxin (5) was incorrect. The toxin contains four units of galactose (not one as originally described); the galactose is in the relatively rare furanose form, with B linkages. The core of toxin is a sesquiterpene, not cyclopropanol as first described. This part of the dissertation has been published (section I). Several analogs of toxin which contain three, two, or one units of galactose were found in cultures of H. sacchari. These compounds might be precursors or degradation products of toxin. They were called toxoids, for convenience, because they are analogs of toxin that protect 75 76 susceptible tissues against toxin. Another significant finding was that 'H. sacchari produces an enzyme (8-galactofuranosidase) which removes galactose units from the toxin molecule, forming toxoids. The kinetics of the production of 8-galactofuranosidase, toxin, and toxoids indicated that the toxoids are degradation products of toxin. It is interesting that the 8-galactofuranosidase activity in H. sacchari cultures is very low when compared to the activities of four other species of Helminthosporium. If the B—galactofuranosidase activity was as high in liquid cultures of H. sacchari as it is in cultures of other species, then very little toxin would accumulate. It is only because HS toxin can be obtained at up to 200 mg/liter of culture fluids that much of this work could be done. When HPLC became available to the laboratory, 1 found that my preparations of toxin and toxoids were a mixture of isomers (section IV). Three forms of HS toxin were separated with reverse phase HPLC. Each isomer of HS toxin was toxic to susceptible but not to resistance clones of sugarcane, although each differed from the others in relative ability to induce loss of electrolytes from susceptible sugarcane tissues. Using a B-galactofuranosidase, each isomer of HS toxin and of toxoid III was partially hydrolyzed and the resultant toxoids were separated and identified with HPLC. This information allowed me to identify the sesquiterpene isomer and the number and arrangement of galactose units in the toxins and toxoids. The work also confirms other reports (2) that toxin contains two chains of B-l.5 linked galactofuranose units. The toxoids were isolated by HPLC and tested for biological activity (section IV). An assay based on measurement of the galactose released by acid hydrolysis of these compounds was developed to measure small amounts 77 of the toxin and toxoids. Only the three isomers of toxin and one isomer of toxoid III are toxic to sugarcane; the toxoid III isomer was toxic to three but not to two other clones of sugarcane that were all susceptible to the pathogen and sensitive to the toxin (four galactose units). It is interesting to note that this isomer of toxoid III was more toxic than was the toxin from which it was derived. The other five isomers of toxoid III and the toxoids with two or one units of galactose were not toxic to any of the sugarcane clones tested. Each of the six isomers of toxoid III and seven of the toxoid_II isomers were tested for ability to protect sugarcane tissue from toxin-induced damage. The number of galactose units in each toxoid was the most important factor in determining how effective a toxoid was at protection. Toxoids that contained three units of galactose protected better than did any of the toxoid II isomers. Toxoid I provided much less protection than did the toxoids that contained two units of galactose. These and other results indicated that the sesquiterpene isomer and the number and arrangement of galactose units are all important in determining how effective a compound will be at inducing or preventing toxin-induced loss of electrolytes from sugarcane tissue (section III and IV). The current hypothesis for explaining the high degree of specificity which is characteristic of HS toxin involves a receptor protein (7). Many data have been published to establish this hypothesis but the work has been severely criticized (1). The hypothesis states that a protein in sensitive tissues binds toxin whereas a similar but different protein in insensitive tissue is unable to bind toxin. Binding was said to cause a change in the shape of the receptor protein. The change in some way 78 allows the pathogen to colonize the host tissue. Are the data presented in this dissertation compatible with the receptor protein hypothesis. If toxin is interacting with a receptor protein, then the toxoids may also interact with this receptor. However, a simple interaction may not be sufficient to cause tissue damage. Toxoids are not toxic even at high concentrations (100 pM). Toxin may be unique because it cannot only interact with the receptor but its binding may change the shape of the protein enough to alter its function. Toxoids could bind to a limited degree but may not induce sufficient change in shape of the protein to exert a toxic effect. These considerations indicate that my data could be compatible with a protein receptor model. My aim was to contribute to an understanding of plant disease development at the molecular level. However, the full significance of the work will require as a first step a quantitative determination of toxin and toxoid production in infected tissue. Toxin production at the site of initial colonization has not been shown, although toxin is clearly being produced in older lesions as evidenced by the development of runner lesions and by isolation of toxin from these lesions (6). Preliminary studies of the toxin produced by H. sacchari fit the pattern of a host selective toxin yet only suggest that the toxin is necessary for disease development. Studies such as those done with H. carbonum and H. yjctggiag are needed for H. sacchari (3). The production of radio-labeled toxin and attempts to isolate a receptor protein also should have a high priority. 3. 4. 5. 6. 7. LITERATURE CITED Daly JM 1981 In "Toxins in Plant Disease", Durbin RD (ed.) Academic Press, NY pp 515 Macko V, T Goodfriend, T Wachs, JAA Renwick, W Acklin, D Arigoni 1981 Characterization of the host-specific toxins produced by Helminthosporigm sacchari, the causal organism of eyespot disease of sugarcane. Experientia 37:923-924 Scheffer RP 1983 In "Toxins and Plant Pathogenesis", Daly JM, BJ Deveral (eds.) Academic Press, Sydney pp 181 Steiner GW, RS Byther 1971 Partial characterization and use of a host-specific toxin from Helminthosporium sacchari. Phytopath 61:691-695 Steiner GW, GA Strobel 1971 Helminthosporoside, a host-specific toxin from Helminthosporium sacchari. The Journal of Biological Chemistry 246:4350-4357 Strobel GA, GW Steiner 1972 Runner lesion formation in relation to helminthosporoside in sugarcane leaves infected by Helminthosporium sacchari. Physiol Plant Path 2:129-132 Strobel GA 1982 Phytotoxins. Ann Rev Bioch 51:309-333 79