130 571 THESlS This is to certify that the thesis entitled VIRULENCE MUTANTS 0F CEPHALOSPORIUM GRAMINEUM presented by SALLY LYNN VAN WERT has been accepted towards fulfillment \ of the requirements for Master .aLScianchegree in WPlant Pathology 113m; (4]. Ear/«:7; 111—— Major ofessor L" Dennis w. Fulbright ( F 5 Date AL?/»LJ—+ J/ {My} 0-7639 MS U is an Affirmative Action/Equal Opportunin Institution MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. ROOM US 03513 lilo NOT (I (MATE ~ VIRULENCE MUTANTS 0F CEPHALOSPORIUM GRAMINEUM BY Sally Lynn Van Wert A THESIS submitted to Michigan State University in partial fulfillment of the requirement for the degree of MASTER OF SCIENCE Department of Botany and Plant Pathology 1983 ABSTRACT VIRULENCE MUTANTS'OF CEPHALOSPORIUM GRAMINEUM BY Sally Lynn Van Wert To determine if graminin A (GRA) or polysaccharides, metabolites found in culture filtrates of Cephalospgrium gramineum, are important disease determinants of Cephalo- sporium leaf stripe (CLS), mutants were generated by UV light and chemical mutagenesis. Mutants varying in vir-' ulence were selected in a wheat seedling bioassay. GRA and polysaccharide porduction were determined by extraction from culture filtrates. No correlation between polysaccharide production and virulence was found. Gas chromatography/mass spectroscopy studies indicated that GRA was not a pathogeni- city factor and put its role as a virulence factor in doubt since GRA was not produced by moderately to highly virulent mutants. However, biological activity of GRA was observed in antimicrobial and phytotoxic assays. The seedling assay was also used to screen wheat lines for resistance to CLS. A positive relationship between tolerance and lack of symp- tom production in the seedling assay was found. To my mom and dad ACKNOWLEDGEMENTS I wish to thank Dr. Dennis Fulbright for his construc- tive advice, encouragement and financial support throughout' my graduate career at MSU. Thanks also goes to my committee members Drs. Bob Scheffermand Ray Hammerschmidt for their interest in my education and the editing of this manuscript. Special thanks goes to Al Ravenscroft for all his technical assistance and emotional support. Particular thanks is ex- tended to Brian Musselman, Facility Manager of the MSU Mass Spectral Facility. Without his assistance and expertise I would not have been able to complete this study. Thanks is extended to all my friends and the members of my lab for their patience and friendship. I would es- pecially like to thank Bob Livingston for his friendship and 'technical advice, and Mo Milligan and Steve Alderman for their guidance and instruction in the use of the MSU compu- ter system. Last, but not least, I extend my appreciation to my family for their love and support. Finally, love and appre- ciation are extended to Haggie McLeod, Jack Dingledine and 1 Dan Shelton for their invaluable friendship. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF APPENDICES LIST OF ABBREVIATIONS Literature Review PART I: iv Page vi viii 1 xi 1 Isolation and characterization of Cephalo- spgrium gramineum mutants that vary in Virulence 26 Introduction 27 Materials and Methods 28 Isolates of Cephalospgrium gramineum 28 Toxin ° 28 Mutagenesis of Cephalospgrium gramineum 28 Seedling assay 30 Colony morphology and relative spore production 32 Detection of revertants 36 Ability of isolates to overwinter and cause disease in the field 36 Culture harvest 37 Polysaccharide extraction 38 Toxin extraction and purification 39 Toxin detection by gas chromatography (GC) 40 Toxin detection by gas chromatography/mass spectroscopy (GC/MS) 4O Toxin detection by ultraviolet spectroscopy 41 Disc assay for antimicrobial activity 41 TLC plate assay for antimicrobial activity 43 Leaf-sheath assay for phytotoxicity 43 Seedling-flask assay for phytotoxicity 44 Inhibition of seed root growth by graminin A 45 Leaf-puncture assay for phytotoxicity 45 Electrolyte-leakage assay for phytotoxicity 46 Page Results -. 47 Mutagenesis and isolation of mutants with altered virulence 47 Colony morphology and relative spore production 50 Ability of isolates to overwinter and cause disease in the field 50 Polysaccharide and toxin production ' by isolates in culture 51 Characterization of culture filtrate preparations by gas chromatography/mass spectroscopy and ultraviolet spectroscopy 51 Disc assay 71 TLC plate assay , 72 Leaf-sheath assay__ _M_*mf .. ,73 Seedling-flask assay Inhibition of seed root growth 74 Leaf-puncture assay 74 Electrolyte leakage 79 Discussion 79 PART II: Screening wheat lines for resistance to Ceghalsospgrium gramineum with graminin A an wit iso ates varying in virulence 91 Introduction 92 Materials and Methods 93 Isolates of Cephalospgrium gramineum 93 Toxin 94 Plants 94 Seedling assay 96 Leaf-sheath assay 96 Results 97 Disease severity of wheat lines inoculated as seedlings with isolates of Cephalo- spgrium gramineum 97 The ef ect o graminin A on wheat lines as determined by the leaf-sheath assay 106 Discussion 109 Appendices 114 Bibliography 135 Table l. 2. 9. 10. ll. 12. 13. LIST OF TABLES Fungal isolates used Symptom rating system used in the seedling assay Disease severity rating as related to isolate virulence Comparative virulence of isolates and disease severity of cultivar Yortstar Production of polysaccharide in culture by isolates of Cephalospgrium ramineum which vary in viruIence Production of graminin A in culture by isolates.which vary in virulence Leaf-sheath assay symptoms for Yorkstar 3 days after treatment with graminin A Effect of graminin A on root growth by seeds of wheat and cress Isolates of Cephalosporium gramineum Wheat lines and their disease severity reaction to Cephalospgrium gramineum in the field Disease severity of winter wheat lines UT89099, Lenore, P1347738, MT77077, LRC 40, Lancer and PI094424 inoculated as seedlings with isolates of Cephalo- spgrium gramineum The effect of autoclaving and millipore filtering isolate shake suspensions on disease severity in winter wheat lines Yorkstar and Agrotritichum Leaf-sheath assay symptoms: The effect of graminin A on cut seedlings of l4-day-old winter wheat lines vi Page 29 33 35 49 52 54 75 76 94 95 99 104 107 Table A.1 B.1 -.c.1 E.1 E.2 E.3 F.l.1 F0301 Quantification and biological activity of graminin A samples after exposure to heat or pH change Quantification of graminin A in organic solvent extracts Migration of graminin A on silicic acidi thin layer chromatography plates using different mobile phases ' Mobile phases used and corresponding fractions eluted from column The presence and quantity of graminin A in samples as determined by thin layer chromatography and gas chromatography, respectively Biological activity of graminin A in combined fractions as determined by thin layer chromatography plate assay Tissue dry weight of M-13 and vis- ualization and migration of graminin a from culture filtrate extracts of M-13, with different harvest times, on thin layer chromatography plates Migration and visualization of graminin A from culture filtrate ex- tracts of M-l3, grown at different temperatures, on thin layer chrom- atography plates Migration and visualization of graminin A from culture filtrate ex- tracts of M-l3, grown in different media, on thin layer chromatography plates vii iPage 117 119. 120 125 125 126 128 129 130 LIST OF FIGURES Figure 1. 2. 3. 4. 5. 10. 11. 12. 13. 14. Structure of gragatins and graminin A Structure of 'butenolide' Structure of vitamin C Structure of tetronic acid Structure of the tautomer of tetronic acid Mass spectrum of authentic graminin A Mass spectrum of preparations from culture filtrates of M-l3 (A) and CG-18 (B) Selective ion monitoring spectra of authentic graminin A Selective ion monitoring spectra of preparations from isolate M-13 culture filtrates Selective ion monitoring spectra of preparations from mutant CG-l8 culture filtrates Ultraviolet spectra of isolate M-l3 culture filtrate preparations (A) and authentic graminin A (B) Ultraviolet s ctra of authentic graminin A (A and mutant CG-18 culture filtrate preparations (B) Electrolyte leakage of leaves (A) and sheaths (B) 'Disease severity of winter wheat lines Yorkstar, Marias, Agrotritichum and CIO7638 inoculated as seedlings with isolates of Cephalosporium gramineum viii Page l7 l7 l7 17 17 56 61 63 65 68 70 78 101 Figure 15. F.1 Page Disease severity of winter wheat lines CI11222, P1178383, F6-870 and P1278212 inoculated as seedlings with isolates of Cephalosporium gramineum 103 Growth kinetics of CG-18 and 7-54 in potato dextrose broth and potato, dextrose broth plus methionine (20 pl/ml) 134 ix LIST OF APPENDICES Appendix ' Page A. The effect of temperature and pH on graminin A ‘ 115 B. Efficiency of extraction of graminin A from water-by organic . solvents ' 118 C. Migration of graminin A on silica thin layer chromatography plates using different ~ mobile phases ’ ~9 120 D. High pressure liquid chromatography ' of graminin A 122 E. Flash chromatography of graminin A 123 F. The optimum number of days of growth, the optimum temperature for growth and the optimum growth medium for M-13 for maximum production of toxin 127 G. Growth kinetics of CG-l8 and 7-54 132 Aczo CA CLS BtOAc ' BtOH GC GC/MS GRA MeOH MTG em PDA + str ens TLC uv LIST OF ABBREVIATIONS acetone Davis' complete agar Cephalosporium leaf stripe ethyl methanesulfonate ethyl acetate ethanol gas chromatography gas chromatography/mass spectroscopy graminin A 1-1 ' - minimal agar methanol N-methyl-N'-nitro-N-nitrosoguanadine 'potato dextrose agar potato dextrose agar plus streptomycin potato dextrose broth thin layer chromatogaphy ultraviolet xi LITERATURE REVIEW General Stripe disease of wheat was first observed in Japan where it became an important disease in fields cropped con- secutively to wheat. Nisikado and Ikata named the causal fungus Cephalosporium gramineum Nis. and Ika. and the dis- ease Cephalosporium leaf stripe (CLS). Their studies re- sulted in a comprehensive report, which included discussion of the taxonomic position of the fungus and observations on factors influencing disease development (Nisikado et a1, 1934). Cephalosporium is a member of the Deuteromycetes (Fungi Imperfecti). The genus is characterized by its well dev- eloped hyaline mycelium and its slender unbranched, singly arising phialides (conidiophores) (Buchanan, 1911; Barron, 1968; Hawksworth and Waller, 1976). Phialospores (conidia) are abstricted successively in basipetal succession from an open growing point at the apex of the sporogenous cell, which lacks a distinct apical collarette. Phialospores remain in a ball or rarely in fragile chains at the apex. The spore balls on top of solitary, tapering phialides, are the diagnostic feature of the genus (Barron, 1968). In 1971, Gams (in: McGinnis, 1980) concluded the genera Acremonium and Cephalosporium were cogeneric and that l .Acremonium. should be considered to have priority. However, this suggestion has not been followed; Cephalosporium will be retained as the genus name in this paper. Grasses are the only known hosts; as is implied by the species ephitet gramineum. Several major cereslfcrops are susceptible (Wiese, 1977) and CLS has been found on many grasses, including the species of Dactylis, gggggg, 511523, Agropyron, Egg and Arrhenatherum (Bruehl, 1957; Howell and Burgess, 1967). Spring cultivars of wheat, oats, barley, rye and triticale are susceptible, but they do not suffer serious damage from the disease in the field (Wiese, 1977). The disease is of major importance only on winter wheat (Triticum aestivum L.). Cephalosporium leaf stripe is found in Japan (Nisikado et al., 1934), Europe (Gray, 1960; Slope, 1965: Hawksworth and Waller, 1976) and North America (Wiese, 1977). It was first reported in the United States by Bruehl (1956a, 1956b). In North America CLS occurs frequently in the Great Lake States, the Pacific Northwest, Kansas and Montana. Disease incidence in these areas has ranged to 80% (Wiese, 1977), and yield has been reduced by 80% under severe dis- ease conditions (Richardson and Rennie, 1970; Johnston and Mathre, 1972). In Michigan, CLS has been observed every year since 1961. Smith et al. (1966) found symptoms in every wheat field observed in 1965. Recently, Kansas estimated the loss of 500,000 bushels of wheat, or 1.5% of the crop to CLS. This is greater than the losses from pow- dery mildew and take-all (NCA-Meetings, 1979). Bruehl (1963) identified a fungus that formed a spor- odochial stage on winter wheat stubble and straw as Hymenula cerealis Ellis. and Everth.. Cultural and inoculation ex- periments showed that E; cerealis is the saprophytic-spor- ulating stage of g; gramineum. The causal organism of CLS, therefore, has two imperfect names: Cephalospgrium gramin- ggm, the parasitic stage, and Hymenula cerealis, the sapro- phytic stage. Bruehl stated that g; cerealis should take precidence over 9; gramineum but recommended that Cephalo- sporium leaf stripe remain the name of the disease. To date no perfect stage has been described for the fungus. Saprophytic stage The saprophytic life cycle of g; gramineum begins after harvest and tillage operations return wheat debris to the soil. Sporodochia appear in cool, wet weather during the autumn months. According to Wiese and Ravenscroft (1978b), growth of the fungus within the straw is non-directional, inter- and intracellular. Stomata and severed ends of straw segments are the avenues to the exterior. Spore production and sporodochial development are initiated only after hyphae emerge from straw. Rapid conidiogenesis along with mucopolysaccharide production results in a mass of tightly adhering.phialospores (Bruehl, 1963; Hawksworth and Waller, 1976; Wiese and Ravenscroft, 1978b). Conidiogenesis is highly efficient even without sporodochia formation (Wiese,‘ 1977). Wiese and Ravenscroft (1975) found the half-life of conidia to be 0.5 to 2.5 weeks at 25 C, whether the soil was moist or dry, and 26 weeks at 7 C in moist soils. As long as the integrity of the colonized host tissues was main- tained, the population renewed itself. When the host residues remain undisturbed near or at the soil surface, the fungus can survive saprophytically and produce infectious propagules (conida) for at least three growing seasons. Cephalospgrium gramineum is a poor competiter in colon- izing refuse, grows slowly (Wiese, 1977). Lai (1967) and Lai and-Bruehl (1968) showed that the fungus dominated buried straw for the first six months and then declined ate: constant rate. After the g; gramineum titer decreased, 2;;- choderma spp. and Fusarium spp. became the predominating ‘colonizers, in moist and dry soil, respectively. Thus, 9; gramineum, when established in straw, exhibits some ability to dominate the substrate. There is evidence that anti- microbial production may play a role in excluding other fungi from previously infested wheat straw (Lai, 1967; Lai and‘Bruehl, 1968; Bruehl et al., 1969). An antibiotic pro- duced by g; cerealis (Lai, 1967; Lai and Bruehl, 1968; Bruehl and Lai, 1969) was most active in a medium at pH 5 or lower, and might aid in competition. Bruehl et a1. (1972) found antibiotic production to be related to growth rate: antibiotic production was reduced when the fungus grew rapidly and increased when the fungus was under moderate water stress and growing slowly. Among fungi, effective substrate possession is not unusual, and.may be of more im- portance to pathogens of weak saprophytic ability (Bruehl and Lai, 1966). In summary, untiI“infection occurs, survi- val during the fall and winter is a function of spore pro- duction, longevity and substrate possession. The development of a green wheat agar as a selective medium (Wiese and Ravenscroft, 1973), allowed the quantita- tive determination of g;_gramineum in soil . Wiese and Ravenscroft (1975) monitored propagule numbers in Michigan wheat fields and showed that propagules began to increase dramatically in September and peaked in midwinter. Little or no detection of propagules occured in June and July. Spores were the primary source of inoculum and declined rapidly in number with the advent of warm temperatures. Once a field became infested with g; gramineum, continued wheat monoculture normally resulted in a build-up of g; gramineum populations. Wiese and Ravenscroft (1978a) demon- strated a longterm decrease in the pathogen population and disease incidence with continued wheat monoculture over an eight-year period. Other than in Michigan, the decline of CLS has not been reported in any major wheat growing area where wheat monoculture is commonly practiced. Bailey (1980), studying the decline phenomenon in the same fields observed by Wiese and Ravenscroft (1978a), did not demon- strate the transferability of the 'decline factor' from a field in decline to fields which had been in monoculture but not showing decline. Bailey suggested that pathogen popula- tions and disease fluctuations over time may be a character- istic of individual fields regardless of seemingly identical cultural practices. Parasitic stage It is generally agreed that the soil-borne pathogen, g; gramineum, invades wheat through the roots at wounds ass- ociated with spring heaving of soil (Bruehl, 1957; Pool and Sharp, 1969a: Johnston and Mathre, 1972; Mathre et al., 1977; Wiese, 1977) or through wounds associated with insect injury (Bruehl, 1968; Wiese, 1977). Spring grains escape infection, or avoid injury, as spring infection rarely builds to damaging proportions (Wiese, 1977). In Britain, leaf stripe has been associated with wireworm damage (Slope and Bardner, 1965.). Otiano (1962) presented evidence for the entrance of the pathogen through openings resulting from rupture of the pericarp and coleorhiza by emergence of roots during germination. Experiments by Mathre and Johnston (1975b) indicated that root exudates increased conidia production by hyphae from infested straw. They suggested that passive entry of conidia into the root system took place in the spring after heaving had severed the roots. At this time, conidia would be drawn in with transpirational water, enter the vascular system, reproduce and colonize the plant. This type of mechanism would separate 9; gramineum from other vascular pathogens such as Fusarium spp. and Verticillum spp., which can actively penetrate and grow into root tissue, as well as entering through wounds. Bailey (1980; Bailey et al., 1982) showed by electron microscopy, that conidia will germinate and produce runner hyphae on the root surface in response to increased root ex- udation following freeze stress. Once penetration had occured the epidermis and cortex did not prevent coloniza- tion by g; gramineum. Bailey suggested that freeze stress and not wounds per se may be an important factor affecting the predisposition of wheat plants to active penetration and infection by g; gramineum. When the fungus was introduced directly into the root xylem, at room temperature, visible leaf stripe was seen in about ten days (Wiese and Ravenscroft, 1978a). In cereals, the fungus became established as a vascular pathogen and caused chlorotic and necrotic xylary stripes. Formation of stripes started near the crown and moved acropetally, later accompanied by chlorosis and necrosis of leaf tissue (Nisi- kado et al., 1934; Bruehl, 1957; Hawskworth and Waller, 1976). One to 4 linear stripes develop on on the culms, leaf blades and leaf sheaths. Acropetal symptom development leads to early senescence of lower leaves resulting in stunting, the bleaching of spikes prior to normal ripening, poorly filled or unfilled seed heads, smaller seed size and decreased flour quality (Mathre and Johnston, 1975a; Mathre, et al., 1977). Overall quality decrease was measured by kernel weight, flour yield and physical properties of dough. Microscopic examination of leaf segments showed that the fungus was present within 1-2 vascular bundles per leaf. Only proto- and metaxylem vessels were inhabited (Wiese and Ravenscroft, 1978a). Morton (1980) reported that xylem restriction was associated.with xylem maturation gradients between internodes, within nodes, and within leaves. With increased colonization, the disruption of protoplasts in phloem and mesophyll cells bordering vascular bundles was detected. A marked decrease in chloroplast number within affected mesophyll cells was associated with external manifestations of chlorotic stripes. Zones of mesophyll disruption coalesced when vascular bundles were colonized. This evidence inducated that g; gramineum is incapable of penetrating living cells at any time during the disease cycle. Control ‘ Cephalosporium leaf stripe is favored by cool spring temperatures,.soils having high moisture levels, early planting dates and fields planted consecutively to wheat (Nisikado et al., 1934; Pool, 1967; Bruehl, 1968: Bruehl and Lai, 1968a; Pool and Sharp, 1969a; Wiese and Ravenscroft, 1976, 1978a: Bailey, 1980: Bailey et al., 1982). Bruehl and Lai (1968a) indicated CLS was associated with soil types having low pH and poor internal drainage. Pool and Sharp (1967, 1969b) showed that disease was more prevalent in lower and more poorly drained soils. Poor drainage leads to poor aeriation, compact soils and slow decomposition of straw.so that survival of the pathogen in refuse is pro- longed. Pool and Sharp (1967, 1969b) also indicated that seedlings from early plantings with fertilization were sub- ject to increased root breakage as compared to those from later plantings or early plantings without fertilization. The larger root system in the fertilized seedling could yield more entry points for the pathogen. Latin et al.(1982) recently reported that the amount of applied nitrogen appeared to have no measureable effect on CLS incidence. The major means of control of CLS are through the cult- ural practices of crop rotation, sanitation and late sowing. As these three types of control are not always practical, the most desirable control measure would be the use of resistant varieties. Seed treatment is not necessary because 10 the pathogen is not disseminated through seed (Nisikado et al., 1934). No chemical means of controlling this disease are currently available. Wiese and Ravenscroft (1975) and Latin et a1. (1982) have suggested three year rotations of wheat with nonhost crops. CLS was shown to be self-sustaining in fields with a history of grassy weeds in the 2nd, 3rd or 4th year of wheat production (Wiese and Ravenscroft, 1978a). Late sowing (Wiese and Ravenscroft, 1976; Wiese, 1977) limits autumn root growth, thus minimizing sites for injury and infection. Wiese and Ravenscroft (1976) advocated planting 10 days af- ter the local Hessian flyfree dates in Michigan. Sanitation is of extreme importance as a cultural means of control, and can be accomplished by burning infested straw and stubble, mechanical removal of residue prior to planting, and disking or deep plowing (Nisikado et al., 1934: Wiese and Ravenscroft, 1975; Latin et al., 1982). Removal of residue eliminates the pathogen. Wiese and Rav- enscroft (1975) showed that levels of fungus and disease are related to the amount of residue in and on the uppermost 7.6 'cm of soil. Disking was shown to be inferior to deep plowing for removing residue. Mathre and Jphnston (1975b) stated that deep plowing to 30 cm is an excellent means of control. Although Latin et a1. (1982) have confirmed Wiese's work in Michigan, they indicated that minimal tillage has promise when combined with longer rotations and 11 resistant wheat cultivars. This combination was suggested because of significant soil erosion problems in the Pacific Northwest and the need for maintainance of soil conservation practices. To date there are few winter wheat cultivars showing tolerance to CLS. Severing the roots of wheat seedlings, applying conida in liquid suspension cultures and rating symptoms after a given period, has been used to differen- tiate between resistant and susceptible varieties (Mathre and Johnston, 1975a: Mathre et al., 1977; Bailey 1980). Mathre (Mathre et al., 1977) contends that a lower inoculum potential of g; gramineum differentiates best and that these levels approximate those encountered in the field. Morton .et a1. (1980) also discouraged the use of nonvernalized win- ter wheat plants because itwas_assumed they a .- provide an atypical environment for the pathogen. Bailey (1980) showed that a freeze screening technique correlated better with field determined resistance than the test utilizing root severing as the predisposition factor. Mathre et al. (1977) observed that yield components in infected plants (seed size or kernel weight and seed number per head) vary in different wheat lines suggesting that lines may react to the pathogen in different ways. As seed size is usually reduced more than seed number, Morton and Mathre (1980a) suggested that selection on the basis of seed size could be an effective and simple means of identifying 12 and evaluating resistance in infected plants derived from winter wheat germplasm. Morton and Mathre (1980b) have identified three types of resistance to CLS: l) a reduction in the number of dis- eased plants in a population, 2)‘a reduction in number of diseased tillers within a plant, and 3) a reduction of the rate and severity of disease development within a plant. The latter two responses restrict the pathogen after successful ingress and have been observed only in one wheat cultivar, Crest LRC 40. Pathogenesis ' There has been much speculation regarding which fungal ‘ metabolite, if any, might be a primary agent in symptom pro- duction in CLS. Most studies have focused on polysacchar- ides or toxins. In 1938, Ikata and Kawai (1938), showed that filtrate from nutrient solutions in which the organism had grown contained a toxin which inhibited seedling growth. Bruehl (1963) suggested that symptom development may be the result of toxic metabolites in addition to the plugging of vessels by mycelium, because brown discoloration of vascular elements occured when little mycelium was present. Spalding et al., (1961) found that g&' gramineum produced polygalac- turonase which was isolated only from infected plants. Diseased tissues were also found to have a lower moisture content, which they attributed to increased viscosity of 13 solutes in the xylem and xylary plugging by a polysaccharide produced by the fungus. As lateral and acropetal dye move- ment could not take place in stripe tissue, it was felt that hyphae and pectin plugs derived from the degradation of host tissue by pectinolytic enzyme action further inhibited water . movement. Such 'vascular distress' was concluded to contri- bute to cellular dysfunction and death of wheat plants. This work was supported by Pool and Sharp (1969a) who ex- tracted a polysaccharide produced by g; ggamineum, only from infected wheat tissue, and showed that it would restrict fluid movement in healthy wheat leaves. Wiese (1972) suggested that a diffusable fungal by- product(s) and not pectin or hyphal plugging in the xylem was responsible for stripe development. Electron micro- graphs of infected vascular bundles showed an accumulation of an electron-dense material surrounding conidia. This electron-dense material was found to line the walls of in- fected vessels after liberation from conidia. In leaves bearing prominent symptoms, xylem vessels were densely packed with conidia but no plugs of polysaccharide or pectin were seen. Wiese believed that occlusion of main vascular bundles as a result of fungal proliferation is of secondary importance in pathogenesis as occlusions were found to follow rather than preceed lateral extension of leaf striping. 14 In 1977 Kobayashi and Di (1977b, 1979) isolated and characterized a toxic metabolite from culture filtrates of g; gramineum which reportedly discolored the vascular tissue .' and induced chlorosis and vascular browning at low concen- trations in wheat leaf cuttings. This compound, graminin A, was also found to have antibiotic properties effective ag- ainst some bacteria and fungi. Gregatin A, isolated earlier by Robayashi and Ui (1977a) from culture filtrates of g; gregatum, had nearly the same structure and antimicrobial activity as graminin A (Figure l). Gregatin A is a toxin that apparently mimics the symptoms of brown stem rot of adzuki bean, soybeans and mung beans. Gray and Chamberlain (1975), four years earlier, found that soybeans bred for resistance to brown stem rot did not wilt when placed in a crude culture extract of g; gregatum but susceptible culti- vars wilted in 3 days. They suggested that a toxin was in- volved in pathogenicity and that resistance was partially a result of toxin resistance. More recently, Kobayashi (1980) attempted to extract gregatin A and graminin A from diseased adzuki bean and wheat tissue, respectively, and from plants injected hypodermically in the stem base with their respec- tive toxins. As no toxins were isolated, he postulated that the compounds may have been converted into other compounds 1 vivo. In both circumstances inoculation with toxin produced symptoms similar to natural infection. 15 Creatura et a1. (1981) found that stomates opened wider and responded slower to water potential changes when leaves were treated with graminin A than leaves not treated with. the compound. This response also occured in g; gramineum infeted plants where it preceeded stripe development and was more pronounced under conditions of water stress. They showed that differences in stomatal activity were not a function of differences.in the leaf water status. Results indicated that the toxin graminin A, and not blockage of the lxylem, was involved in early stages of pathogenesis. Morton and Mathre (1980a) reported that the pattern of stripe formation in g; gramineum infected flag leaves of a sus- ceptible winter wheat cultivar at heading was closely ° correlated with depression of realitve water content, net photosynthesis, chlorophyll content and stomatal conduc- tance. All four physiological parameters were interrelated as indicated by regression analysis. They concluded that chlorosis around colonized vascular bundles could be attributed to effects of localized restriction of lateral water movement rather than a diffusable toxin, like graminin A. It should be noted that those data were recorded late in pathogenesis, in contrast to the work by Creatura et al., whose results apply to early pathogenesis. Graminin A (Figure l) and gregatin A (Figure l) are FIGURE 1. FIGURE 2. FIGURE 3. FIGURE 4. FIGURE 5. Structure Structure Structure Structure Structure of of. of of of 16 the gregatins and graminin A. 'butenolide'. vitamin C. tetronic acid. the tautomer of tetronic acide cnao 17 2 J11. GREGATIN A H CH'CH’CH: a w cua _ s on c143 onmmm A H omen-can, HO-CH O "0 J5" 0 FIGURE 3 0\ 1“: ~ H2 0 O 18 secondary metabolites that are derivatives of tetronic acid. The proper name for tetronic acid is 3-hydroxybut-2-enolide (Pattenden, 1978). Tetronic acid and its simple acyl der- ivatives are polar substances with high melting point (Haynes and Plimmer, 1960). These compounds are found in a group of natural products which includes the butenolides. The term 'butenolide' describes the unsaturated 7-lactone system (Figure 2) (Pattenden, 1978). The nucleus of both butenolides and tetronic acids is the 7-lactone, 5-membered ring; a cyclic ester (Morrison and Boyd, 1975). Tetronic acid derivatives are a and/or 7-substituted derivatives of the parent acid (Figure 4). This form normally predominates in the tautomeric system (Figures 4 and 5) (Haynes and Plimmer, 1960). Those dissubstituted in the 7-position are not regarded as tetronic acids. The tetronic acid nucleus occurs in the important nat- ural product vitamin C (Figure 3), which is a 7-sub- stituted-a-hydroxytetronic acid (Haynes and Plimmer, 1960). A group of complex tetronic acid derivatives were found re- sponsible for coloration in the lichens (Haynes and Plimmer, 1960; Pattenden, 1978). Tetronic acid derivatives are also found as fungal metabolic products from several Penicillium spp.. The bulk of these have been characterized from g; charlesii, and many other fungi (Haynes and Plimmer, 1960; Pattenden, 1978; Gudgeon et al., 1979). 1.9.; Apart from'vitamin C, most tetronic acid derivatives appear to show little physiological activity. However, penicillic acid is more active against gram-negative bac- teria than is penicillin (Haynes and Plimmer , 1960): vul- pinic acid, produced by the fungal constituent of lichen, is toxic to animals (Pattenden, 1978); and the tetronic acid derivatives found in marine sponges possess strong anti- biotic activity (Pattenden, 1978). Patulin, biosynthesized by g; pgtulin, has been used as a fungal toxin in food and as a general plant toxin (Ellis and McCalla, 1973). The gregatins and graminin A have been shown to possess both antimicrobial activity and phytotoxicity (Kobayashi and Ui, 1977a, 1979, Kobayashi, 1980; Creatura et al., 1980). Gre- gatin production has also been reported for Aspgrgillus panamensis (Anke et al., 1980). Interestingly, aspertonin A and B, produced by A; rugulosus (Ballatine et al., 1969), are enantiomers of gregatin A and D, respectively (Kobayashi and Ui, 1977a; Anke et al., 1980). Many lactone-ring con- taining compounds are carcinogenic to animals (Ceigler et al., 1971; Uraguchi and Yamazaki, 1978). Anke et al. (1980) consider the double bond of the a-side chain of gregatin A (Figure l) as the structural fea- ture most important for biological activity. The reduced activity of gregatin D (Figure l) is a result of the reduc- tion of this double bond. Gregatin D can theoretically be 20 derived from gregatin A through addition of various alcohols or water to the double bond of thecz-side chain. The longer acyl chain of graminin A leads to a change in selective toxicity, as the data of Kobayashi and Di (1979) indicated. Two separate biosynthetic pathways to tetronic acids have been established: a) oxidative cleavages of polyketide derived aromatic intermediates, and b) condensation of an acetate-derived chain(s), usually fatty acid, with tri- carboxcylic acid (TCA) cycle intermediates (Turner, 1971: Pattenden, 1978: Gudgeon et al.. 1979). The relatively high incorporation of tracer molecules, primarily 14C-labeled precursors, into secondary metabolites in microorganisms has enabled detailed studies to be made of the biosynthesis of several fungal metabolites. Based on these works, Stoessl (1981) suggested that graminin A and gregatins A and D are compounds of mixed fatty acid~TCA cycle origin, with C-4-6 likely from C-l to C-3 of succinic acid or equivalent and C-7 - C-12 and C-2 - C-16 (or C-18) from fatty acid residues. Determining the role of a toxin in pgthogenesis In plant pathology, the term toxin is generally defined as a 'non-enzymatic product of a microorganism or a micro- organism-host interaction which is harmful to plants in low concentration' (Rudolf, 1976). Toxins in plant pathology have three essential features: a) toxins are products of 21 microbial pathogens of plants; b) toxins cause obvious dam- age to plant tissues: and c) toxins are known with confi- dence to be involved in disease development (Scheffer, 1983). VSome authors (Luke and Biggs, 1976; Rudolf, 1976) distinguish between phytotoxin and pathotoxin, the former referring to those which are toxic to plants and the latter defined as a toxin which induces all typical disease symp- toms in reasonable concentration and whose production is correlated with pathogenicity. The term 'phytotoxin' can be misunderstood as many mycotoxins may be toxic to plants but may never be involved in the development of plant diseases (Reiss, 1978). The term is also used to indicate a toxic substance from higher plants (Harper and Balke, 1981). Rather than define toxins in terms of their toxic pro- perties, Yoder (1980, 1983) distinguishes these molecules in accordance with their requirement for successful infection of the host; 'pathogenicity factors' for those toxins re- quired for pathogenicity, and 'virulence factors' for those toxins involved in the degree of disease expression. Phyto- toxic metabolites can further be classified as non-specific or host-specific (or host-selective). Non-specific toxins affect plant species other than those hosts of the toxin-producing pathogen. Host-specific toxins affect only hosts of the pathogen. The toxic metabolite graminin A may be a non-specific toxin as it has limited host selectivity (Kobayashi and Ui, 1979: Scheffer, 1983). 22 There,are a number of guidelines proposed for the . evaluation of the significance of toxins in disease (Rudolf, 1976, Yoder, 1980). Criteria commonly used include the iso‘ lation of toxin from diseased plants, reproduction of typical disease symptoms when toxin is applied to healthy plants, correlation of pathogen and toxin specificities to- ward plants and correlation of virulence with ability to produce toxin. None of these criteria alone provide ade- quate evidence for the involvement of a toxin in disease. Assay systems to evaluate the phytotoxic action of sus- pected toxins when applied to healthy plants include the use of intact plants, seeds or seedlings, plant parts, tissue culture cells or the use of other target organisms. The placement of cuttings into test solutions has been fre- quently used to assay for wilt toxins and other toxins (Rudolf, 1976). Van Alfen (van Alfen and MbMillan, 1982) recently discussed the misleading nature of the wilt bio- assay method when a suspected wilt toxin is a macromolecule and causes wilt by physically interfering with water move- ment through cuttings regardless of other activities it might possess. He suggested alternatives to the wilt bio- assay which considered the potential of these macromolecules to physically disrupt water transport. An inherent diff- iculty with the use of cutting assays is that the concentra- tion of the toxin to which living cells are exposed can 23 never be determined. The toxin dose taken up per gram fresh weight of a cutting must be determined (Rudolf, 1976). The site of primary or secondary site for action of some toxins is the plasmalemma of the host (Rudolf, 1976; Scheffer, 1976; Yoder, 1980). The measurement of electrolyte leakage is used to assess this action when plant parts are exposed 'to various concentrations of toxin. However, only gross - changes in permeability are detected and it is assumed that electolytes leak atthe same rate as non-electrolytes (Ayers, 1978). Toxins and allelopathic compounds have been assayed by the inhibition of seedling root growth (Pringle and Braun, 1957; Pringle and Scheffer, 1963; Scheffer and Pringle, 1961:.Tang and Young, 1982). The selectivity of the toxin produced by Periconia circinata for susceptible sorghum seedlings but not resistant sorghum seedlings, was demonstrated with this assay (Scheffer and Pringle, 1961). A more rapid and sensitive microbiological assay for toxin, involving inhibition of other organisms by the toxin, may reveal toxic activity at picogram levels (Staskawicz and Panopoulos, 1979). Screening plant lines for resistance to disease with toxin has been used primarily for host-specific toxins (Wheeler and Luke, 1955: Schertz and Tai, 1969; Steiner and Byther, 1971). Disease-resistant plants or cells can be efficiently selected both 1g yiggg and 12,3139. A high level of resistance is expected if a pathogenicity factor is 24 used; an intermediate level of resistance is expected if a virulence factor is used (Yoder, 1983). The best test of pathological significance is the specificielimination of the toxin from the biological system. and the observation of changes ( if any ) which may occur in disease development or initiation. The most powerful technique to attempt specific elimination of a molecule from a complex system is genetic manipulation, either by muta- tional analysis or by the use of naturally occuring varia- tion. An example of elimination of a toxin was afforded by Patil et al. (1974). By mutational analysis, they were able to establish the role of phaseolotoxin, a non-specific toxin, in the bean halo blight disease. A toxin-less mutant of Pseudomonas syringae pv. phaseolicola, though able to. multiply as well 33,3112 as the wild type, could neither cause systemic chlorosis nor invade inoculated host tissue systemically. Phaseolotoxin was therefore determined to be a virulence factor. Naturally occuring variations of toxin production is present in the pathogen £5; syringae pv. tabaci . Loss of tabtoxin-producing ability results in loss of chlorotic halos, as is characteristic of fig; syringae pv. angulata (Braun, 1937). These results indicated that the tabtoxin, a non-specific toxin, contributed and was re- sponsible for certain symptoms, but was not required for pathogenicity. In contrast, genetic and other data indicate that the host-specific toxins from Helminthospgrium .25 victories and g; carbonum are required for pathogenicity of the producing fungi (Sdheffer, 1976). PART I: ISOLATION AND CHARACTERIZATION OF * -.-\ CEPHALOSPORIUM GRAMINEUM MUTANTS THAT VARY IN VIRULENCE 26 27 INTRODUCTION Cephalosporium leaf stripe (CLS), caused by the soil-borne fungus Cephalospgrium gramineum, produces vascu- lar chlorosis and necrosis of leaf tissue resulting in yield reduction in winter wheat. Polysaccharides and toxins pro- duced by the fungus have been suggested as the cause of such symptoms. Polysaccharides produced by the fungus were im- plicated in plugging of the xylem (Spalding et al., 1961; Pool and Sharp, 1969a). However, Wiese (1972) could find no evidence of plugging by polysaccharides. Instead, Wiese found xylary occlusions resulting from fungal proliferation. These developed after leaf striping was evident. Kobayashi and Ui (1977b, 1979) isolated and characterized a toxic sub- stance, graminin A (GRA), from culture filtrates of g; gramineum . Graminin A, which produced vascular browning and chlorosis at low concentrations (25 pg/ml) in excised leaves, is structurally similar to gregatin A, a toxic sub- stance isolated from culture filtrates of g; gregatum . Gregatin A mimics the symptoms of brown stem rot, caused by g; gregatum , on adzuki beans, soybeans and mung beans (Kobayashi and Ui, 1977a). The purpose of my study was to determine whether or not GRA and polysaccharides are impor- tant disease determinants of CLS. This was approached by isolating virulence mutants, and screening them for toxin and polysaccharide production in culture. 28 MATERIALS AND METHODS Isolates of Cephalospgrium gramineum These are described in Table l. Cultures of g; . gramineum were grown onpotato dextrose agar (1.5% agar) plus streptomycin (100 pg/ml) (PDA + str), or grown in potato dextrose broth (PDB) (Tuite, 1969). Cultures to be assayed for GRA and polysaccahride production were grown in a broth described by Kobayashi and Di (1979). reuse Authentic GRA was kindly supplied by K. Kobayashi (Hokkaido University, Sapporo, Japan). The preparations had a minor contaminant (Kobayashi, personal communication) which could be detected by gas-liquid chromatography (GC), high pressure liquid chromatography, and GC/mass spectros- copy (GC/MS). This toxin preparation was used for all bio- assays unless otherwise indicated. Mggggenesis of Cephalosporiumgramineum Ultraviolet light (UV) induced mutants of g; gramineum were obtained by plating 0.1 ml (108 cells/m1) of a 3-day-old shake culture of CG-18 grown on a Lab-line orbit environ shaker (1800 rpm, 27 C) onto PDA + str and exposing the plate to shortwave (zoo uW/cmz) av light for 2.5 minutes from a height of 3.6 inches with a UVSL-25 Mineralight lamp 29 TABLE 1. Fungal Isolates Used Morphology on Isolates Various Media Virlulence pm CA gr; WA CG-82 myc myc myc myc very high M-l3 myc myc myc myc very high CG-l8 yst yst yst myc moderate N20 yst yst yst myc high E53 yst yst yst myc low 7-54 yst yst - - very low E67 yst yst yst yst very low Source Michigan field isolate Michigan field isolate UV mutant of M-13 * NTG mutant of CG-l8 EMS mutant of CG-18 UV mutant of CG-l8 EMS mutant of CG-l8 PDA - potato dextrose agar CA - Davis' complete agar (Lederberg, 1950) MA - Davis' minimal agar (Lederberg, 1950) WA - water agar myc - mycelial yst - yeast-like - - no growth * Fulbright and Ravenscroft, 1981 3O (Ultra-Violet Products Inc., San Gabriel, CA 91778). The plates were incubated in the dark at room temperature (24-26 C) for 6 days before selecting colonies for virulence test. Mutagenesis with N-methyl-N'-nitro-N-nitrosoguanadine (NTG) and-ethyl methaneaulfonate (EMS) was performed as described by Miller (1972) for NTG mutagenesis. The concen- tration of NTG was approximately 40 ug/ml. Approximately 150 pl of concentrated EMS was added to 12 ml of a shake culture of mutant CG-l8. Aliquots were taken from the shake culture at 0, 1, 2, 4, 10 and 24 hours and plated on PDA + str. After 4 days growth,.colonies were selected ar- bitrarily and tested for virulence using the seedling assay (described below). Auxotrophic requirements were found by supplementing minimal agar (MA) with various combinations of amino acids, vitamins, purines and pyrimidines (Holliday, 1956L Seedling assay ("WHI'LThe soft, winter white wheat cultivar Yorkstar was used for all assays requiring plant material. This variety is very susceptible to CLS. Plants for the seedling assay were grown for 10 days in autoclaved sand in a growth chamber with a l4-hr-day-photoperiod (9 x lO‘ergs/cmz- sec) and 21 C. The plants were fertilized at 3 days and 9 days with a solution of Rapid Gro (25, 19, 17; 1 tbsp/gal water). After 10 days growth seedlings were removed, the roots severed to 1 inch in length and washed free of and particles, and plawed im conidial suspensions. Conidial suspensions were 3] prepared by growing 9; gramineum in 50ml PDB for 6 days on-a reciprocal shaker (96 strokes/min) at 24-26 C. Six seed- lings were placed in each conidial suspension (107- 109 conidia/ml) for at least 15 minutes before transfer to a pot containing sterilized soil. The remaining conidial sus- ' pension was poured equally into the holes in which the seed- lings were planted. After inoculation, seedlings were grown in a chamber at 17 C, with a lS-hr-day-photoperiod (9 x 104 ergs/cmz- sec). The temperature favored growth of the pathogen. Plants were fertilized 7 days after inoculation and symptoms were rated 14 days after inoculation. If no symptoms appeared the 14 days, the plants were kept in the chamber for another week for observation of possible symptoms. The initial screening procedure was repeated 3 times, to select mutants with various degrees of virulence. Four mutants, 3 with decreased virulence and 1 with increased virulence, were selected and tested 3 more times. Each seedling assay included the following treatments : uninocu- lated PDB, isolates CG-18, M-13, CG-82 and autoclaved (20 min, 15 psi) shake cultures of CG-18 and M-l3. A symptom rating system (1-15) was based on observa- tions of symptom development in the growth chamber (Table 2). Each plant in a pot was given a rating and the ratings were totaled per pot. Each pot represented one replication of one treatment. The experiment was set up and analyzed as 32 a randomized complete block design, blocking over time (Steele and Torrie, 1980). A disease severity rating as re- lated to isolate virulence was obtained by transforming the symptom rating system to a scale of 6-100 (Table 3). -This was accomplished by multiplying total pot symptom rating values (symptom rating total for 6 plants) by a factor of 1.11. A seedling was randomly selected from pots inoculated with each isolates for reisolation of the organism. Sections from the leaf-sheath and the leaves were surface sterilized in a 1.5% NaOCl.solution, and placed on PDA + str. Plates were observed for one week for fungal growth from the vascular tissue. Colony morphology and relative spggg_production Growth of mutants was obseved after 4 days on PDA, Davis' complete agar (CA)(Lederberg, 1950), Davis' minimal agar (MA) (Lederberg, 1950) and water agar (WA) for compari- son with colonies of mutant CG-l8 and isolate M-l3. Morphology of conidia and sporogenous cells were examined with the light microscope to confirm identity of the mutants as being members of the genus Cephalospgrium. To determine relative spore production of mutants and isolates, PDB (50 ml in 125 ml Erlenmyer flasks) was inocu- ated with mutants or wild type isolates grown for 7 days in 33 TABLE 2 Symptom rating system used in the seedling assay m WGOQO‘GUIMO‘IfiwUNI-J 10 11 12 13 ** Sygptom no symptoms faint, general chlorosis lst 2nd lst lst 2nd 3rd lst 2nd lst 1st lst lst 2nd lst 2nd 1st leaf chlorotic leaf chlorotic and 2nd leaves chlorotic leaf striping leaf striping leaf striping leaf chlorotic and striping leaf chlorotic and striping and 2nd leaves striping leaf chlorotic, 2nd leaf striping leaf striping, 2nd leaf chlorotic leaf chlorotic and striping, leaf chlorotic or striping leaf chlorotic or striping, leaf chlorotic and striping and 2nd leaf chlorotic and striping lst lst lst, leaf dead leaf dead, 2nd leaf chlorotic leaf dead, 2nd leaf striping 34- TABLE 2 can't. 14 entire plant nearly dead, stem is green 15 . entire plant dead, no green tissue * Astericks indicate the same rating was used for variations of the same type of symptom. 35 TABLE 3 Disease severity rating as related to isolate virulence Total Symptom Disease Rating for 6 Severity Disease Isolate Seedlings Rating Severity Virulence 6 6.66 0-25: no disease ‘absent or or 12 13.32 very mild very low 18 19.98 24 26.64 25-45: mild low 30 33.30 36 39.96 42 46.62 45-60: moderate moderate 48 53.28 54 59.94 60 66.60 60-85: severe high 66 73.26 72 79.92 78 86.58 84 86.58 85-100: very very high severe 90 99.90 36 PDA + str. Inoculated flasks were incubated on a reciprocal shaker. After 6 days the number of conidia/ml for each cul- ture was determined using a hemacytometer. Detection of revertants " - An auxotrophic, methionine requiring mutant isolated after UV light mutageneis of mutant CG-18 was assayed for reversion by the following method. A culture of approx- imateiy‘io8 conidia/ml ens was centrifuged at 8500 g to: 10 minutes. The pellet was resuspended in sterile physiologi- cal saline (0.85% NaCl), centrifuged and then resuspended in 10 ml saline. A 0.2 ml sample of the suspension was plated on MA. After one week plates were checked for growth. Any colonies appearing on MA would be considered revertants and tested for virulence in the seedling assay. Agility of isolates to overwinter and cause diseaSe in the field Spore suspensions were obtained by growing the wild type isolates CG-82 and M-l3 and the mutants CG-18, 7-54 and E-67 in modified Eckert's medium (Johnston and Mathre, 1972) in shake culture (1800 rpm, 27 C). Oat inoculum was pre- pared by adding 10 ml aliquots of a heavy spore suspension of each isolate to 150 g of oat seeds which had been moist- ened with 100 ml distilled water and autoclaved for 90 minutes in glass quart jars. The jars were capped with a screw band placed over 3 layers of Whatman #3 filter paper. The bottles were shaken to distribute inoculum and incubated 37 at room temperature (24-26 C) for 8 weeks. The jars were shaken at 2-3 week intervals and at 4 weeks 50 ml of Eckert's medium was added (Mathre and Johnston, 1975a). Field plotsiwere inoculated with oat kernel inoculum at Michigan State University in the fall of 1982. Inoculation was accomplished by placing the inoculum in a 7-row, Auger-feed seed drill along with vitavax treated wheat seed. 'The seed and inoculum were added to the row simultaneously in a field which had been grown in dicots for several years. A completely randomized design was used. Treatments in- cluded wild type isolates CG-82 and M-l3 and mutant isolates CG-18, 7-54 and E67 grown on oat kernels. Controls included vitavax treated seed and oat seeds treated with PDB. Isolation of the fungus from infected tillers (leaves showing stripes) was attempted in the spring of 1983. Iso- lations were attempted from plants at growth stage 5 (pseudo-stem, formed by sheaths and leaves, strongly erect) and 10.1 (first ears just visible, ear escaping through split of sheath) on the Feekes' scale (Large, 1954). Culture harvest Isolates (3 replications/isolate) were grown in 4 l diptheria bottles containing 1 liter of medium (Kobayashi and Ui, 1979) for 28 or 29 days (24-26 C). Three 1 liter replications of each isolate were grown. Cultures were har- vested by filtering through 11 cm Whatman #4 filter paper 38 discs which had been dried to a constant weight in a 45 C drying oven. Filters with mycelium were dried again in the oven and weighed. For all isolates 200 ml of filtrate was centrifuged at 4100 g for 40 minutes. The resuspended coni- dial pellet placed in a predried, preweighed aluminum pan, and then dried and weighed. The weights of culture medium which collected on filters, along with the mycelium and that which was pelleted by centrifugation, were subtracted from dried mycelial and conidial pellet weights, respectively. Polysaccharide extraction Polysaccharide was precipitated by adding an equal volume of 95% EtOH to 200 ml of the culture filtrate. The solution was refigerated (4 C) overnight and then centri- fuged at 4100 g for 30 minutes to pellet the precipitate. The pellet was redissolved in a small volume of distilled water and dialyzed in cellulose dialysis tubing (molecular weight cut off 12,000 - 14,000, Spectrapor, VWR Scientific Inc.) against distilled water for 6 hours in the cold (4 C). The polysaccharide content was determined by the Anthrone method (625 nm) (Hodge and Hofreiter, 1962) using galactose was used as the standard. An analysis of variance and the LSD multiple comparison test (Steele and Torrie, 1980) were performed to determine significant differences in polysaccharide production between wild type isolates and mutants in culture. 39 Toxin extraction and purification The remaining 800 ml of culture filtrate was evaporated under reduced pressure at 50 C to 100 ml. Each 100 ml sample was then extracted four times with equal volumes of methylenechloride (CH2C12). The efficiency of extraction was 82% based on recovery of known amounts of GRAX The or- ganic solvent extracts were evaporated to dryness under re- duced pressure at 38 C and the residue redissolved in a small volume of chloroform (CHC13). The CHCl3solutions were transferred to l dram vials and dried under a stream of nitrogen gas. Vials were sealed with a teflon lined cap and stored at -20 C until futher purification. The contents of each dram vial were redissolved in 200 pl CHCL3 and applied to Supelco silica gel G , preparative thin layer chromatography (TLC) plates containing a phosphor (254 nm). The plates were developed with CHC13:methanol (MeOH) (98:2, v/v) and the toxin detected as a dark blue band under shortwave (300 pW/cmz) uv light, but not longwave (640 pW/sz) UV light. The toxin bands were removed form the TLC plate and eluted with CHC13. The CHCl3 eluate was filtered through Whatman CG/F, 3.1 cm glass-fiber filters to remove silica gel paricles. The filtrate evaporated to dryness un- der reduced pressure at 38 C. The residue was redissolved in 50 ul CHCL3 and a small volume spotted on a Fischer silica gel GF, redi-plate to check for homogeneity of the sample. The efficiency of recovery toxin from TLC prepara- 40 tive plates was 79%. .At all times a 1 or 2 pg samples of authentic GRA/pl CHC13 was used as a standard. Dried cul- ture filtrate preparations were stored at -20 C in 1 dram vials with teflon caps until analysis. Toxin detection by gas chromatography (GC) 50 fl CHCl3 was added to each vial containing prepara- tions from each isolate culture filtrate. One pl samples were analyzed by injection into a Varian 3700 GC equipped with a Supelco glass column (6 ft, inner diameter 2 mm) packed with 3% OV-l7 on 100/120 gas chromosorb Q. Para- meters of each run included: 300 C detector, 270 C in- jector, 190 C ( l min ) - 290 C ( 1 min ) column, with a 0-10, attenuation 4 or program of 10 C increase/min, range 1 8 and flame ionization detection. The carrier gas (N2) had a flow rate of 30 cc/min. To insure complete delivery, samples were injected by the sandwich technique where the sample being assayed was held in the syringe between two equal volumes of solvent. Along with each set of culture filtrate preparations, a sample of 1 pg authentic GRA/pl CHCl3 was assayed. Samples were also assayed by coinjection with 1:11 autnentic GRA/p1 CHCl3 to verify cochromatography of sample and standard. . Toxin detection by gas chromatographylmass spectroscopy(GC/MS) Culture filtrate preparations and authentic toxin were analyzed by GC/MS using a Hewlett-Packard, 5985 GC/MS 4] (Iquadropole system equipped with an 18 inch glass column and packed with 3% OV-l7. The GC column temperature program for each run was 190 C (1 min) - 285 C, with a 20 C increase per minute. The carrier gas (He) had a flow.rate of 30 cc/min. Mass spectra were obtained by electron impact of compounds. In addition to total mass spectra, selective ion intensity and mass spectra were produced by selective monitoring of ions of particular interest. Toxin detection by ultraviolet spectroscopy The spectra of culture filtrate preparations were ob- tained by scanning samples (in MeOH) from 210 nm - 340 nm in a Gilford 2600 spectophotometer. The spectra of prepara- tions from wild type and mutant isolate culture filtrates were compared with the spectrum of the authentic toxin and with published spectral values for GRA (Kobayashi and Ui, 1979). Qigcassay for antimicrobial activity Toxin in 2% EtOH was added to individual Whatman 540 (2.1 cm) hardened ashless filter discs to final quantities of 0, 30, 50 and 100 pg GRA. The discs were allowed to dry and placed on plates of PDA + str. These plates were then sprayed with a log-phase culture of Escherichia ggli, Rhodotorula spp. or Bacillus megaterium ,grown in complete medium, using a DeVilbiss 15 spray atomizer. Plates were incubated at 27 C and observed for zones of inhibition after we: af1 re: My x Slll Vii Pl: b0l 0f 42* 24 hours. This assay was repeated using Cladosporium ggggg- erinum. Plates were incubated at 17 C and observed for zones of inhibition after 30-36 hours. i The disc assay developed by Staskawicz and Ponopoulus (1979) was also used._ A 2 m1 log-phase g; ggli culture grown in minimal broth or Rhodotorula spp. grown in complete broth was mixed with 2 ml of molten WA (2%) kept at 65 C. After overlaying on MA, toxin-impregnated filters were placed on the overlay. Alternatively wells were cut into the overlayed agar plates with a No. 3 cork borer (0.6 cm. diameter) and filled with 20-25 pl toxin solution. Plates were incubated as above and observed for zones of inhibition after 5 hours and 24 hours for g; ggli and Rhodotorula spp., respectively . This assay was also used with 9; cucumeri- ggg. Complete broth (25 ml) was added directly to lO-day-old cultures of g; cucumerinum grown on PDA and the surfaces of the cultures were rubbed lightly with a glass rod. Two ml of the resulting spore suspensions were mixed with 2 ml of molten WA. Toxin-impregnated filters were placed on the overlay or wells were cut in the agar, as a- ‘bove. Plates were incubated at 17 C and observed for zones of inhibition after 30-36 hours. Toxin concentrations from 0 to 100 g GRA in 2% EtOH were used in all assays. the culture filtrate preparations of wild type isolates and mutants were also tested by these methods. 43 flC plate assay for antimicrobial activity A nutrient medium (Allen and Kuc, 1968) was added directly to lO-day-old cultures of g; cucumerinum grown on PDA. The surfaces of the cultures was rubbed lightly with a glass rod and the‘resulting-spore suspension placed in a glass DeVillbiss 15 atomizer. The suspension was sprayed on TLC plates on which authentic toxin samples and preparations from wild type isolate and mutant culture filtrates has been run or spotted. The sprayed plates were placed in humidity chambers (Allen and Ruc, 1968). Plates were supported by 10 ml glass beakers and the chamber was sealed with masking tape. The plates were incubated for 48 hours and inspected for inhibition of g; cucumerinum at locations corresponding to migration of the toxin. Zones of inhibition appeared as white spots on a dark green background of spores and mycel- ium. , Leaf-sheath assay forphytotoxicity Plants for the leaf-sheath assays were grown in sterilized soil in a growth chamber with a lZ-hr-day-photo- period (2 x losergs/cmz- sec) and a 21 C temperature regime for 15 days. These plants were fertilized 7 days after planting. Following the procedure of Kobayashi and Ui (1977b, 1979), lO-day-old wheat seedlings were cut at the soil line, cut again under water and placed through slits in parafilm sealed 1 dram vials containing 1 ml of toxin sol- ution (0, 25, 50 and 100 pg GRA/ml 2% EtOH). Each treatment '44 was performed in duplicate. Vials were placed under flour- escent light banks (6 x 103ergs /cm2- sec) with a lO-hr-day-photoperiod: the temperature was 28 C. Leaves were evaluated for chlorosis and wilt, as compared to the 2% EtOH control, after 3-5 days. The leaf-sheath assay was also performed by placing cut seedlings in 100 pl of toxin solution (0, 25, 50, and 100 pg GRA/ml 2% EtOH), allowing the toxin solution to be taken up and then filling the vial with distilled water whenever necessary. The leaves were evaluated as above. This procedure allowed the amount of toxin applied per gram fresh weight of wheat tissue to be determined. Fresh weights of five cuttings were averaged to yield a mean fresh weight. Seedling-flask assay for phytotoxicity Seeds were surface sterilized by placing in 70% EtOH for 5 minutes, under vacuum, followed by 5 minutes under vacuum in 0.5% NaOCl containing 1 drop of 5% Triton-x 100 per 500 ml 0.5% NaOCl. Seeds were then rinsed twice with sterile distilled water and single seeds were placed in scintillation vials containing 3 ml 0.1% nutrient broth and allowed to germinate (Bailey, 1980). After 4-7 days, noncontaminated, germinated seeds were asepticaly trans- ferred to 125 ml Erlenmyer flasks containing 20 ml PDA + str. One week later a 1 ml solution of GRA in 2% EtOH or 1 ml of 10, 20 or 40% EtOH was asepticaly pipetted onto the agar surface and the flask swirled so the liquid covered the CUE a '8 45 surface. After 2 days roots were inspected for browning. Inhibition of seed root growth by graminin A . Inhibition of seed root growth was assayed by the methods of both Pringle and Braun (1957) and Tang and Young (1982). Following the procedure of Pringle and Braun, five wheat seeds, with the radicle just beginning to show, were placed in 60 x 15 mm petri dishes with 5 ml of a toxin solu- tion (0, 25, 50, 100 and 200 pg GRA/ml 2% EtOH) or distilled water. After 48 hours the length of roots was measured. Following the methods of Tang and Young, dormant cress curled seeds (Herpt Brothers Seedmen Inc., Brewster, NY 10509) or pregerminated wheat seeds were placed in petri dishes on filter discs impregnated with toxin and moistened with 200 pl of water. The petri dishes were sealed with parafilm. Cress seed root length was measured after 3 days and wheat seed root length after 8 hours. For both seed types the following toxin solutions and controls were used: 0, 5, 10, 30, 50, and 100 pg GRA in 2% EtOH and distilled water. An analysis of variance was performed on each set of data. If a significant difference was found with the F-test, Dunnett's test (Steele and Torrie, 1980) was per- formed to determine which toxin treatment varied signifi- cantly from the 2% EtOH control. Leaf:puncture assay for phytotoxicity Plants for the leaf-puncture assay were grown the same as those grown for the leaf-sheath assay. Cut lS-day-old It") (1 46 wheat leaves were placed in pyrex petri dishes (15 x 100 mm) on glass supports over wet filter paper. A small hole was made near the leaf base and a 5 pl solution of GRA (25, 50, 100, 200 pg GRA/ml 2% EtOH) was placed on the wound (Scheffer and Livingston, 1980). The plates were sealed with‘parafilm and the wheat leaves observed for symptom production in 24 hours. Electrolyte leakage assay for phytotoxicity Ten to twelve day-old greenhouse grown wheat seedlings used for the electrolyte leakage assay. The first and sec- ond leaves of seedlings were cut into 1 cm lengths and rinsed. Duplicate 0.2 9 samples were tied in 6 x 6 inch pieces of cheesecloth, the cheesecloth bags were placed in distilled water and rinsed 3 times during the following 10 minutes. The bags were then placed in scintillation vials containing 5 ml of toxin solution (0, 25, 50, 100 and 200 pg GRA/ml 2% EtOH) for 1 hour. The leaf pieces were vacuum in- filtrated for 10 minutes and then incubated on a shaker (122 strokes/min, 24 C) for the remaining 50 minutes. The bags were rinsed thoroughly with distilled water and 5 ml distilled water were added as a leaching solution. The initial zero reading was taken at this point. Vials were incubated on a shaker and conductivity (pmhos) readings were made at 0.5, l, 2, up to 10 hours with a pipette-type elec- trode (k - 1.0) coupled with a conductivity meter (Scheffer and Livingston, 1980). A comparison of electrolyte leakage 47 between leaves and sheaths was also made using 12-day-old seedlings. Readings were taken hourly up to 12 hours and then sporadically up to 46.5 hours. Duplicate sample readings were averaged and plotted. Data were also analyzed by linear regression (Steele and Torrie, 1980). The slopes (rate of electrolyte leakage) of individual regression lines were compared for homogeneity. RESULTS Mutagenesis and isolation of mutants with altered virulence Propagules of mutant CG-18 were treated with NTG, EMS and UV light. Survival of 10% of the propagules was achieved after 24 hours, 10 hours, and 2.5 minutes, respec- tively. One auxotrophic mutant requiring methionine was ob- tained by UV mutagenesis. Revertants of the auxotroph were not obtained after several attempts to isolate prototrophs. An identifying number was given to each mutant. Mutants generated by NTG were given a number preceeded by the letter N. Those generated by EMS were given a number preceeded by the letter E. Several hundred mutagenized isolates were screened for virulence in The seedling assay. Four mutant isolates, N20, E53, 7-54 and E67, were selected for further study. Seedling plants were inoculated and severity of disease in- duced by each isolate was determined. Virulence of wild .48” type isolates and mutants of cultivar Yorkstar are summar- ized in Table 4. The wild type isolates CG-82 and M-l3 were highly virulent, and the mutant CG-18 was moderately viru- lent. Disease induced by isolate N20 was rated as more severe than that induced by its parent CG-18. Disease sev- erity levels rated for isolates E53, 7-54 and E67 were less severe than CG-l8. There was no significant difference among the disease severities produced by the autoclaved cul- tures, uninoculated PDB, or mutants E67 and 7-54. Disease severity produced by CG-82 and M-13 were not significantly different. The fungus was routinely isolated from the sheath but rarely from the third leaf of inoculated seedlings, depend- ing on the virulence of the isolate. CG-82 and M-l3, highly virulent wild type isolates, were isolated from all parts of seedlings in most cases. E67, a mutant of very low vir- ulence, was sometimes isolated from the sheath and rarely from the first leaf. The same was true for mutant 7-54. E53 was usually isolated from the sheath and first leaf. Mutants CG-18 and N20 were found in the sheath and in the first and second leaf, but rarely in the third leaf. 4’) TABLE 4 Comparative virulence of isolates and disease severity of cultivar Yorkstar _ Disease Sevfrity Disease . 2 Isolate Rating Severity Virulence CG-82 93.24 e very severe very high M-l3 (85.47 e very severe very high CG-18 ‘ 44.76 c moderate moderate N20 61.43 d severe high E53 27.93 b low low 7-54 13.32 a very low very low control(PDB) 6.66 a very low very low M-13 6.66 a very low very low auto CG-18 6.66 a . very low very low ' auto 1. Isolates with the same letter do not differ sig- nigicantly (p - 0.05) according to the LSD test (Steele and Torrie, 1980). 2. The relationship of disease severity to virulence is summarized in Table 3. 50 Colony morphology and relative spore production of isolates Colony morphology of wild type isolates and mutants on four agar types is summarized in Table 1. All fungi pro- duced yellow-beige colored colonies. Mutant CG-18 was yeast-like when examined with the light microscope. The most common form of CG-18 was as the shape of the letter Y, where each of the three sections making up the letter were elongated, oval sporogenous cells. Sometimes short, randomly branched chains and clumps of sporogenous cells were found. Single, round and elongated, oval conidia were also seen. All sporogenous cells con- tained two round inclusions, one at either end of each cell. The conidia of M-l3 and CG-82 had the same morphology as conidia of CG-18. Examination of spores and sporulating structures of mutants with the light microscope confirmed they were 9; gramineum and derived from CG-18. The relative spore production of wild type isolates and mutant shake cultures used for inoculation of seedlings in the seedling assay was 0.80 - 5.20 x 108 conidia/ml PDB. CG-l8 produced the most spores and E67 produced the least number of spores in PDB. There was no relationship between virulence and spore production of mutants and isolates. Ability of isolates to overwinter and cause disease in the field Reisolation of g; gramineum in the spring of 1983 from the early season (growth stage 5) and the late season 5} (growth stage 10.1) plants showing striping symptoms was successful for CG-82, M-13 and CG-18. The results were identical for both plant ages. The Q; gramineum reisolated from CG-82 and M-13 inoculated plants always had mycelial growth on PDA. The control plants were infected with a g; gramineum which had mycelial growth on PDA. The pathogenic yeast-like CG-18 was always reisolated from CG-18 inoculated plants. A g; gramineum isolate with mycelial growth on PDA was isolated from E67 inoculated plants indicating a natural contaminate from the soil. No Q; gramineum was isolated from 7-54 inoculated plants. Polysaccharide and toxin production by isolates in culture Polysaccharide production in culture is summarized in Table 5. There were, however, significant differences in the amount of polysaccharide produced in culture between wild type isolates and mutants. However, there was no rela- tionship between virulence in the seedling assay and poly- saccharide production in culture. Characterization of culture filtratepggparations by ggg chromatogggphy/mass spectroscopy and ultraviolet gpectroscopy “' Preparations from culture filtrates coinjected with autnentic GRA resulted in one peak when analyzed by conven- tional GC. Subsequent analysis of toxin preparations by 52 TABLE 5 Prodution of polysaccharide in culture by isolates of Cephalospgrium gramineum which vary in virulence Isolate Virulence1 m ol saccharidez'3 co-sz very high 0.33 a M-13 very high 0.95 c CG-18 moderate 0.81 bc N20 high 0.47 ab 353 ' low ' 0.77 bc E67 very low 0.94 bc 1. Virulence was determined by the seedling assay, using cultivar Yorkstar. 2. Polysaccharide content was determined by the anthrone assay. Fungal tissue was determined by dry weight. 3. Each value is the mean of 3 replications. Values followed by the same letter do not differ sig- nificantly (p - 0.05) according to the LSD test (Steele and Torrie, 1980). 53 GC/MS revealed no production of GRA by all mutants (Table 6). CG-82 and M-13 did produce GRA as determined by GC/MS. The total ion mass spectrum of authentic GRA from Rob- ayashi is shown in Figure 6. The molecular ion (M) m/e 304 was detected as were (M-l) m/e 303 and (M+l) m/e 305. Two mass spectral fragments of GRA were observed at m/e 207 (M-CGHQO) and m/e 97 (-C6H90) (Robayashi and Ui, 1977). The base peak was m/e 93. The heights of the ion peaks in the spectra indicate the relative abundance of one peak to that of the base peak. The total ion mass spectra of culture filtrate prepara- tions from M-l3 and CG-18 are shown in Figure 7. The spec- trum of culture filtrate preparations from M-l3 culture fil- trates (Figure 7A) is representative of both CG-82 and M-13. The moleclar ion (M) m/e 304 was observed. The apparent base peak, however, was a spectral fragment at m/e 129. The spectrum of the culture filtrate preparation from CG-18 cul- ture filtrate (Figure 7B) is representative of CG-18, N20, E53 and E67. No molecular ion (M) m/e 304 was detected.’ The base peak was m/e 129. The spectral fragment, m/e 129, was also seen in the spectrum of authentic GRA (Figure 6). The authentic GRA sample and preparations from culture’ filtrates of the tested isolates were examined more closely for GRA by selective ion monitoring. The ions m/e 129, 224, 248 and 304 were chosen. The ion m/e 129 was .54 TABLE 6 Production of graminin A in culture by isolates which vary in virulence Isolate Virulence1 gtggagtionz CG-82 very high + M-l3 very high + CG-18 moderate - N20 high - :53 low ' ' - E67 very low - 1. Virulence was determined by the seedling assay, using the cultivar Yorkstar. 2. The presence of graminin A was determined by gas chromatography/mass spectroscopy. + - graminin A produced, - - no graminin A detected. FIGURE 6. Mass spectrum of authentic graminin A. 56 a}: .0 masons...“ ogql OWN Ova — .ww 0.2. _ va ._i____ 1 . .ai» our pi j. 00—. :_ i. a: 4.4— a fin—«q!- b b v a: a. . :1-« q P41: __ _:Efi _ b 1:141: : _ _fi._....._.___ 0 =3:— _ aMlelafl souspunqv 5 -7 FIGURE 7. Mass spectra of preparations from culture filtrates of M-l3 (A) and CG-18 (B). Relative Abundance Relative Abundance loo. sol, 604 404 20. L00. 20 4b a so 10 120 '4'.- 140 12"!) 135 '29} “22h rzji' Ill/O FIGURE 7. 59 monitored as it was the base peak of the isolate and mutant culture filtrate preparations. The remaining ions were not found in the spectrum of preparations from CG-18 culture filtrates (Figure 7B), but were common to the authentic GRA mass spectrum (Figure 6). The ion intensities of the selected ions in the authen- tic GRA sample are shown in Figure 8A. The number of ions detected for a peak reaching 100% on the relative abundance scale is given for each selected ion. The peaks at approx- imately 180 seconds (3 min) represented GRA. The ratio of the number of ions of m/e 304 detected to the number of ions of m/e 224 detected was 1:2. That the ion peaks for m/e - 224, 248 and 304 had the same retention time indicated they were derived from the same molecule. Figure 8B is the mass spectrum of the selected ions in authentic GRA at the reten- tion time 3:02 minutes. The spectral fragment m/e 129 was present in the authentic GRA. The ion intensities of the selected ions in the prep- arations from M-l3 culture filtrate are shown in Figure 9A. The peaks at approximately 200 seconds (3.5 min) represented GRA. ALthough GC alone did not allow the detection of a contaminating compound, the ion intensity spectrum resolved this compound from GRA. This was seen by the different re- tention time for m/e 129. The ion m/e 129 had a shorter re- tention time than the spectral fragments of GRA. There was, however, incomplete resolution of the spectral fragments of FIGURE 8. 60: Selective ion monitoring spectra of authentic graminin A. A. Selective ion intensity spectrum. The number of ions detected is given for each peak reaching 100% on the relative abundance scale. . B. Selective ion mass spectrum. The height of the ion peaks in the spectra indicate the relative abundance of one peak to that of the base peak. Relative Abundance Abundance Relative sstsc_r§g i0! ‘3 100%“ 103.. A 304 ‘ I I I I T I I I T I I I 1 100%: 9 ' 24s 3 W -I 1009's- 230 k 224 ‘ I I I r I I r T T I I I 1 100%- 54 A 129 J¢ffi I r r T' r rr ' I t r € 10036- ass o is 7’s 113 152 159 227 ass ans 341 379 417 455 sec. TIME-3:02min. 22‘ 200 pg GRA/ml). Perhaps the inhibition of growth of §g_ggli by GRA was not detected in the disc assay as not enough GRA was used. In contrast to reports by Kobayashi and Di (1979), GRA did'not produce browning of leaves and vascular tissues of wheat cuttings at high or low concentrations. However, chlorosis of leaf-sheath tissue exposed to GRA was seen. Leaf-sheath tissue exposed to 2% EtOH and water also became chlorotic after 3 days. The major distinction between the toxin treated tissues and the control tissues was wilting and leaf drying often associated with toxin treated tissue. Chlorosis was not observed when leaves without intact sheaths were assayed in this manner. Apparently, intact 87 sheaths were necessary for the observation of the toxic ac- tion of GA in the leaf-sheath assay. Toxin application did not reproduce the typical chlorotic striping symptoms of CLS in excised leaves, possibly because the entire vascular sys- tem was exposed to GRA. The remaining bioassays were performed in an attempt to find an assay, more sensitive and less subjective than the leaf-sheath assay, for phytotoxicity. Most of the bioassays were selected because they are commonly used to indicate the phytotoxic action of a suspected toxin, and can be used to screen plant lines for resistance to disease. However, the leaf-sheath assay was the only assay in which GRA consis- tently exhibited phytotoxicity. The seedling-flask assay was developed to examine the effect of GRA on root tissue in young seedlings. Although browning of roots and root tips was observed, the assay re- quired too much preparation to be practicle. Contammination of seedling cultures was very high as the large amount of seed pubesence prevented complete surface sterilization. An assay using roots and observing the time required for root hair cell death after toxin treatment might differentiate resistant and susceptible wheat lines. Root hair cell death could be determined by microscopic observation after roots were stained with a vital stain. Inhibition of seed germination provided an assay for phytotoxicity, but inhibition of root elongation was 88 inconsistent. Immersing the germinated wheat seeds in toxin solutions (Pringle and Braun, 1937) provided inhibition of root elongation, but results differed in two separate ex- periments using this method. Nonhomogeneity of wheat seed could be partially responsible for this inconsistency. The lack of phytotoxicity of GRA impregnated filter discs (Tang and Young, 1982) may have been due to the insolubility of GRA in water or binding of GRA to the cellulose ot the paper. The leaf-puncture assay used to distinguish sugarcane clone sensitivity to Helminthospgrium sacchari toxin (Steiner and Byther, 1971) was adapted to wheat. As little as 58 ng of g; sacchari toxin produced runner lesions on susceptible sugarcane leaves (Steiner and Strobel, 1971). The highest amount of GRA applied to wheat leaf punctures was 1 pg GRA. Perhaps greater quantiites were needed to ob- serve toxic action. Additional problems with this assay were the incomplete uptake of the toxin droplet into the puncture and the inability to quantify actual toxin uptake. The electrolyte-leakage assay would have provided the best assay for quantitative work with authentic GRA prepara- tions. It does not require visual assessment and is the most rapid of the assays discussed. There was no difference in the rate of toxin-induced leakage among all the various concentrations of toxin. Measurement of electrolyte leakage from sheaths was examined because of the necessity for in- .39 tact leaf-sheaths in the leaf-sheath bioassay. Although the slopes of the lines often appeared different, linear re- gression analysis did not demonstrate any significant differences. It can be concluded that the plasmalemma of the host is probably not a site of action for GRA. Phytotoxic activity of GRA was shown by Creatura et al. (1981) . In her study stomates of toxin treated leaves were to open wider and respond more slowly to water potential changes than stomates not treated with the compound. She showed differences in stomatal activity were not a function of the leaf water status. In light of her work, it would be of interest to see if GRA has the same action on guard cells as does fusicoccin (Turner, 1973). Fusicoccin is a toxin produced by Fusicoccum amygdali which opens stomates both in the light and in the dark in a wide range of species (Turner and Graniti, 1969). In an attempt to determine the role of GRA in patho- genesis, the guidelinestroposed for the evaluation of the significance of toxins in disease were followed (Rudolf, 1976, Yoder, 1980). A summary of the results based on commonly used criteria follows. Toxin was not isolated from diseased tissue (Kobayashi, 1980), although Robayashi (Rob- ayashi and Ui, 1979, Kobayashi, 1980) reported reproduction of typical disease symptoms when toxin was applied to healthy plants or excised leaves. In contrast, the repro- duction of typical disease symptoms in the leaf-sheath assay 90 was not found in this study. There was no correlation of virulence with the ability of the fungus to produce the toxin in culture. Elimination of the toxin from the bio- locical system did not eliminate pathogenicity or greatly affect virulence. Although GRA has antimicrobial and phyto- toxic activity, based on this study it does not appear to be a significant determinant of disease. PART II: SCREENING WHEAT LINES FOR RESISTANCE TO CEPHALOSPORIUM GRAMINEUM WITH GRAMININ A AND WITH ISOLATES VARYING IN VIRULENCE 9i 92 I NTRODUCTI ON Cephalosporium leaf stripe (CLS), caused by Cephalo- spgrium gramineum , is the only known vascular pathogen of wheat. Presently, resistance in Triticum aestivum L. has not been identified and field tolerance has only been ob- served in a few lines (Mathre et al., 1977). Using fungal inoculum in field plots Mathre et al. (1977) observed varia- tion in yield components (seed size, kernel weight, and seed number per head) in different wheat lines. Morton and Mathre (1980a) suggested that the intercrossing of wheat lines showing some tolerance in yield components may give rise to lines with resistance to both decreased kernel weight and decreased seed number per head. They suggested this could be an effective means for identifying and eval- uating resistance in infected plants of winter wheat germplasm lines. Morton and Mathre (1980b) also identified three types of resistance to CLS: 1) a reduction in the num- ber of diseased plants in a population: 2) a reduction in number of diseased tillers within a plant: and 3) a reduc- tion of the rate and severity of disease symptom development within a plant. With type 1 resistant wheat varieties symp- toms appear when the pathogen is inoculated above the root. Screening plant lines for reistance to disease with toxins has been performed primarily with host-specific toxins. Wheeler and Luke (1955) used HV toxin to screen for ’93 victoria blight resistance in oat lines. Schertz and Tai (1969) used a toxin isolated from Periconia circinata to identify resistance in sorghum. A procedure for identifying resistant clones of sugarcane to eyespot disease was devel- oped using the host-specific toxin isolated from Helmintho- sporium sacchari (Steiner and Byther, 1971). Graminin A (GRA), a toxic metabolite of g; gramineum , has been shown to possess phytotoxicity and to be specific for wheat (Kobayashi and Ui, 1979), and thus may be useful in screen- ing winter wheat lines for resistance to CLS. The purpose of this study was to screen wheat lines for resistance to g; gramineum by fungal inoculation and to det- ermine if the use of GRA to identify resistant germplasm is a feasible alternative to inoculation with the pathogen. This was approached by screening, in the seedling stage, wheat lines known to be susceptible or tolerant in the field, for disease severity in a pathogenicity assay. The same wheat lines were also examined for the extent of chlor- osis and wilting in a toxin-leaf-sheath assay. MATERIALS AND METHODS Isolates of Cephalosporium gramineum These are described in Table 9. Cultures of g; gramineum were grown and maintained on potato dextrose agar (1.5% agar) plus streptomycin (100 pg/ml) (PDA + str), or 94 grown in potato dextrose broth (PDB) (Tuite, 1969). TABLE 9. Isolates of Cephalosporium gramineum Isolate Virulence Source CG-82 very high Michigan field isolate M-l3 very high Michigan field isolate CG-18 moderate UV mutant of M-13* N20 high NTG mutant of CG-l8 E53 low EMS mutant of CG-18 E67 _ very low EMS mutant of CG-18 *Fulbright and Ravenscroft, 1981 Toxin Refer to Part 1. Plants Winter wheat lines, other than Yorkstar, were provided by D. E. Mathre, Montana State University, Bozeman, MT 59717. Wheat lines are described in Table 10. 95 Table 10 Wheat lines and their gisease reaction to Cepahalosporium gramineum in the field Wheat Line ' Disease Reaction2 Yorkstar 38, 44 Marias 28 Agrotritichum 0 CIO7638 4 C111222 5 P1178383 l3 F6-870 14 P1278212 21 Lenore 14 UT89099 10 P1347738 43 MT77077 28 LRC 40 30 Lancer 34 P1094424 42 1. Disease reaction for the 1980-81 growing season, Mich- igan State University Field Plots. Personal communi- cation, A. Ravenscroft. 2. Percentage of tillers with striping symptoms at growth I stage 5 on the Feekes' scale. Stage 5 is characterized by strongly erect pseudo-stems formed by sheaths and leaves. 96 Seedling assay Assays were performed as described in Part 1. The screening procedure for the seedlings was repeated 3 times. Data from the 1st replication were eliminated as the symptom rating values varied significantly from the 2nd and 3rd rep- lications. The experiment was set up and analyzed as a fac- torial design (Steele and Torrie, 1980) with wheat lines and isolates as factors with two levels of the isolate factor: nonautoclaved and autoclaved shake cultures. A total of 15 wheat lines and 7 isolates, including a PDB control, were used. Another experiment used millipore filtered (0.4stum filters) and autoclaved shake cultures was performed. Analysis was similar to the above analysis. The symptom rating (1-15) (Part 1, Table 2) for each individual plant was transformed to the disease severity index (6-100) by multiplying by 6.66. Each plant was considered a replicate, making a total of 6 replicates per treatment type. Leaf-sheath assay The leaf-sheath assay was performed as described in Part 1. The assay was performed 3 times by placing cut wheat seedlings in 1 m1 solutions of toxin of different con- centrations. The assay was performed twice by placing cut wheat seedlings in 100 pl solutions of toxin of different concentrations, allowing the solution to be taken up and '97 then filling the vials with distilled water. Fresh weights prior to toxin treatment of five cuttings (l4-day-old) for each wheat line were averaged to yield a mean fresh weight for each wheat line to which known concentrations of toxin were applied. RESULTS Disease severity of wheat lines inoculated as seedlings with isolates of Cephalosporium gramineum Disease severity of wheat lines inoculated with wild type isolates and with mutants is summarized in Table 11 and Figures 14 and 15. For all wheat lines except LRC 40 treat- ment with mutant 867 did not differ significantly from PDB. This holds true for the mutant 853 in every wheat line ex- cept Yorkstar, Marias, P1178383, LRC 40 and P1094424. Mutant N20 often appeared to be more virulent than the iso- late CG-18, but the difference may not have been signifi- cant. Isolate CG-82 was usually the most virulent, but it may not have differed significantly from isolate M-13 and mutants CG-18 and N20. Isolate M-13 was more severe than siolate CG-82 on the wheat line Yorkstar, F6-870 and P1094424, but not significantly so on wheat line P1094424. Agrotritichum (a cross between Agropyron elongatum and an unknown hard, red winter wheat) is resistant to g; gramineum (personal communication, Mathre, Table 10). TABLE 11 . 98 Disease severity of winter wheat lines UT89099, Lenore, P1347738, MT77077, LRC 40, Lancer and P109924 inoculated as seedlings with isolates of Cephalsoporium gramineum. Inoculated seedlings were incubated in the growth chamber for 14 days before assessing disease severity. Isolates with the same letter do not differ significantly (p - 0.05) within a wheat line according to the LSD test (Steele and Torrie, 1980). 99 .AommH .oHHHoH cam «Hmoumv ummu awn mnu ou wdapuooom mafia ummsz m dfinuwa Amo. o u av hauamoamacwwm Hmmmau uo: op Houuma mama osu nua3.moumHomH .H Houuzoo a no.0 m mm.a a oa.a m ao.a m ma.a an m~.aa m mm.m mam nu HH.H~ a NH.HH on om.mo a NN.NH nu om.m~ a am.m an ao.am New 0 a~.mo an ~¢.ea on qo.aa an em.e~ an ao.a~ one mo.a~ _o Ne.a mam n m¢.¢m n ma.~m o am.me on mm.m¢ a mo.He co mh.~¢ a No.0m oaz n ao.om n om.me on aa.~o no Hm.mn o H~.- can ma.oa n m~.mm manuo u mm.a~ a mn.mm n aa.~e co ~¢.om o an.aa a Hm.mn o am.ee mH-z a ow.am o mo.oa u a¢.om a me.me o m~.wm m oo.oa o o~.ma ~m-¢o_ «qumoam House; as 0mg finches: mmea¢m- «you»; maoamee mumaoma HmmaHA amass mo huwum>mm «madman Ha HAm 30%) wheat lines, but differentiation was question- able for wheat lines showing intermediate resistance. The screening procedure revealed F6-870 (field disease reaction 112 14%) as intermediate in resistance and P1278212 (field dis- ease reaction 21%) as low-intermediate in‘resistance, as might be expected from their field disease reactions. How- ever, PIl78383 (field disease reaction 13%) appeared to be susceptible. In this case, if the seedling assay were being used to screen winter wheat lines for resistance, a wheat line with intermediate field resistance may have been elim- ianted prior to field trials. This illustrates a possible pitfall; the resistance expressed by a wheat line in the field trials may differ from that expressed in the seedling assay. All the wheat lines showed about the same reaction from toxin in the leaf-sheath assay. regardless of their reac- tion to the fungus. However, Agrotritichum and CIO7638 appeared to be the most tolerant to toxin. This correlated well with resistance determined by the seedling assay and field disease reaction. In contrast, PIl78383 (field dis- ease reaction 13%) appeared to be more affected by GRA than was Yorkstar (field disease reaction 38, 44%). The fresh weight of cut seedling to which the GRA was applied did not appear to influence symptom production. The Agrotritichum tissues had a mean fresh weight approximately two-thirds that of Marias tissues, but Agrotritichum was less affected by toxin treatment than was Marias. TisSues of wheat lines PIl78383 and F6-870 had the same mean fresh weight but the former was more affected by toxin than was the latter. 113 Screening wheat lines for resistance to g; gramineum with GRA does not appear to be a feasible alternative to inocula- tion with the fungus since: 1) the the leaf-sheath assay with toxin was very subjective; and 2) distinctive differ- ences were not evident. This supports the tentative conclu- sion that GRA is not a significant determinant of disease (Part 1). The seedling assay might be used as a rapid means of identifying winter wheat germplasm of intermediate and high resistance to g; gramineum, The resistant lines should then be screened in the field to determine field disease reaction. 9; gramineum isolates of different, but known virulence should be included in both laboratory and field trials. This practice will aid in identification of germ- plasm resistant to a potentially greater number of isolates and races of the pathogen in question. APPENDI CBS 114 115 APPENDIX A The effect of temperature and_pH on graminin A Samples of graminin A'(GRA) in CHCl3 (1 Pg/ml) were auto- claved (20 min, 15 psi), evaporated to dryness under reduced pressure at 50 C, and acidified with HCl to pH 1.2. The acidified preparation was then extracted with ethyl acetate and held at room temperature for two days. All treated sam- ples were dissolved in 100 pl CHClB. Treated samples were quantified by gas chromatography (GC) and tested for bio- logical activity by the thin layer chromatography (TLC) plate assay using Cladosporium cucumerinum as the activity indicator. Summary Autoclaving GRA resulted in the quantitative loss of 75% of the compound due to destruction by heat. Evaporation of distilled water-GRA solutions resulted in the quantita- tive loss of 50% of the compound. GRA may have been lost by heat destruction or volitility or from inefficient recovery of the compund from the inside of the evaporating flask. Treating GRA with acid resulted in the quantitative loss of 60% of the compound. This could be due to destruction of the compound by acid or a poorer efficiency of extaction with ethyl acetate from acidified solutions. GRA did not 116 break down at room temperature. Inhibition of growth of g; cucumerinum on TLC plates was complete when 10 pg of GRA were spotted and incomplete when 5 pg of GRA wse spotted. 117 TABLE A.l Quantification and biological activity of gram- inin A samples after exposure to heat or pH change. Quantification Biological Activity by GC Sample p1 GRA g GRA . . . 2 Treatment pg GRA/pl CHCl3 spotted spotted Inhibition autoclaved 0.245 15 2.94 - 20 4.90 +/- 25 6.12 +/- evaporated 0.510 10 5.10 - 15 7.65 - 20 10.20 + acidified 0.383 ND room temp. 0.989 5 4.95 + 10 9.89 + 15 14.83 + 20 19.78 + standard 1.0241 5 5.12 - 10 10.24 + 20 20.48 + 1. Weight of GRA present was determined by ultra- violet spectral analysis. 2. Inhibition of growth of C. cucumerinum : + complete, +/- incomplete, - absent, ND not determined. 118 APPENDIX B Efficiency of extraction of graminin A from water by organic solvents Samples of authentic graminin A (GRA) (1.63 pg/pl CHC13) were added to 10 ml distilled water. Solutions were ex- tracted four times, each time with 10 m1 of ethyl acetate (EtOAc), CHC13, or CH2C12. Organic extracts and remaining distilled water were evaporated to dryness under reduced pressure at 38 C. Compounds were redissolved in 100 pl CHC13and quantified by gas chromatography (GC). The percent GA recovered from distilled watwe was determined as compared to a non extracted sample. The results indicated that CMZCl2 was the best solvent for extracting GRA from culture filtrate of g; gramineum. 119 TABLE B.l Quantification of graminin A in organic solvent extracts. Sample pg GRA/pl CHC131 % Recovery c—n Cl 1.35 ST. 2 aqueo s Eesidue 0.11 CHCl3 0.36 21.84 aqueous residue ND EtOAc extracted 0.96 58.65 aqueous residue ND unextracted standard2 1.63 1. ND - not detected by GC. 2. Weight of GRA present was determined by ultra- violet spectral analysis. 120 APPENDIX C Migration of graminin A on silica thin layer chromatography using different mobile phases 1 TABLE C.l Migration of graminin A layer chromatography plates =mehile'phases. on silicic acid thin using different Solvent Dielectric Constant (25 C) sz acetonitrile 38.80 0.89 methylene chloride (CH2C12) 8.93 NM isopropyl alcohol 19.90 0.84 n-propanol 20.30 0.84 ethyl acetate (EtOAc) 6.02 0.81 butanol 17.51 0.93 chloroform (CHC13) 4.73 NM ethanol (EtOH) 24.55 0.85 methanol (MeOH) 32.70 0.98 n-hexane 1.88 NM dioxane 2.21 0.64 distilled water 78.54 NM acetone (Ac20) 20.70 0.90 0.78 CHZC12-EtOAC (75:25) 121 TABLE C.1 con't. Solvent Rf CHCl3-EtOAc (75:25) 0.89 CHC13-AC20 (75:25) 0.80 CHZC12-EtOAc (98:2) 0.21 CHC13-EtOAC (95:5) 0.48 C32C12‘AC20 (95:5) 0.65 CHC13-AC20 (95:5) 0.68 CHZClz-MeOH (95:5) 0.70 CHCl3-MeOH (98:2) 0.54 CHCla-butanol (98:2) 0.33 1. 2. An authentic sample of graminin A from Japan was used. NM - no migration. 122 APPENDIX D High pressure ligyid chromatography of graminin A Column: LiChrosorb C18, 10 m, 4.6 x 250mm, Altex Inc. HPLC: varian Model 5000 liquid chromatograph with a Hitachi Model 100-40 Spectrophotometer. Settings: UV 210nm, range 0.05. Solvent System: distilled water:methanol (50:50), isocratic. Samples: Graminin A in methanol, 2ppm. Retention Time: 6 min 7 sec - 6 min 15 sec. 123 APPENDIX E Flash chromatography of graminin A Flash chromatography was tried to see whether some of the yellow contaminating pigments associated with organic ex- tracts could be removed from the extract sample proir to thin layer chromatography (TLC). A slurry of superfine silica (30 g, Chrommedia for column chromatography, LPS-2, Whatman, Inc.) was made with CHZC12 and poured into glass column (2.5 cm GD, 19 cm length of packing). A dirty cul- ture extract of graminin A (GRA) (0.6 ml) plus 200 p1 authentic GRA (200 pg) was loaded onto the column and ex- tracted with four different mobile phases containing CHZCL2 and/or ethyl acetate (EtOAc). A total of 80 frac- tions (10 ml each) were collected. Mobile phases and corre- sponding fractions eluted from the column are listed in Table E.l. Fractions were combined, evaporated, redissolved in 100 pl CHCl3 and run on TLC plates. TLC plates were illuminated with shortwave (300 nm) ultra-violet light to visualize GRA. Combined fractions were also assayed for GRA by gas chromatography (GC). GRA was quantified by the cut and weigh method. The presence of GRA in samples run on TLC plates and the quantity of GRA present as determined by GC are presented in Table E.2. Biological activity of GRA in 124 combined fractions was determined by the TLC plate assay (Table E.3). Summary: Most of the yellow pigmentation was found in the first 25 fractions and the last 20 fractions. As was the case with all other column chromatography packings tried (silicic acid, LH-20, SX-3: data not reported), GRA was not confined to discrete fractions and could be detected in most frac- tions. Flash chromatography was the methood of column chromatography which best removed contaminating pigments. GRA was quantified by GC before it was known that a contam- inating compound comigrated with GRA (Part 1). 125 TABLE 8.1 Mobile phases used and corresponding fractions eluted from column. - Solvent Fractions 100% CHZC12, 100 m1 1-10 CHZC12:EtOAc (97:3), 250 m1 11-42 CHZC12:EtOAc (92:8), 200 ml 42-67 100% EtOAc, 100 ml 68-80 TABLE E.2 The presence and quantity of graminin A in sam- ples as determined by thin layer chromatography and gas chromatography, respectively. . . GC peak2 Combined Fractions TLC area (cm ) 11-25 - 0.48 26-40 + 8.29 41-50 ? 0.29 51-60. - 0.68 61-67 - 0.07 68-72 - ' - 73-80 - ’ authentic GRA + 2.19 2 pg 126 TABLE E.3 Biological activity of graminin A in combined fractions as determined by the thin layer chrom- atography plate assay. Inhibition of Combined Fractions pl spotted' Cladospgrimm cucumerinum 26-40 5 - 10 10 127 APPENDIX E The optimun number of days of growth, the optimum tempera- ture for growth and the optimum growth medium for M-l3 for maximun production of toxin. F.l The optimum number of days of growth of M-l3 in cul- ture for the maximum production of toxin Harvest of 7 liters of fungal cultures took place 13, 21, 28, 32 and 35 days after inoculation of medium (Kobayashi and Di medium, 1979) in diptheria and roux bottles with a 4-day-old spore suspension of M-l3. Only the cultures grown for 35 days.were grown in roux flasks. Cultures were grown at 25 C, photoperiod 12 hours. The mycelial dry weight after 3 days of drying at 45 C was measured. Toxin was ex- tracted with CHCl3 after proteins and carbohydrates were precipitated with MeOH. Dried extracts were redissolved in 200 pl CHCl3 and 5 pl of each were run on a thin layer chromatography (TLC) plate (solvent system, 98 CHCl3 : 2 MeOH). Darkness of bands visualized with shortwave ultra- violet light (300nm) with Rf similar to authentic GRA was used to determine toxin production. Results are given in Table F1.l. 128 TABLE F.1.l Tissue dry weight of M-13 and visualization and migration of graminin A from culture filtrate extracts of M-13, with different harvest times, on thin layer chromatography plates. 2119 Days until Fungal Tissue 5 x28 Darkness harvest dry wt (9) band of band 131 2.44 0.61 light 21 2.25 0.61 medium 28 3.25 0.60 dark 32 4.69 0.59 dark 352 8.98 0.60 medium authentic GRA 0.63 very dark The 13 day fungal dry weight was measured on a different balance. Growth in the roux flasks was greater than in the dip- theria bolltes. 129 E.2, The optimum temperature for culture growth of M-13 for the maximum production of toxin Fungal cultures of M-l3 were grown in diptheria bottles (Kobayashi and 01 medium, 1979) at 15, 19 and 25 C for 28 days prior to harvest in growth chambers with a 12 hour photoperiod. Toxin was extracted, prepared, run on TLC plates and visualized as in Appendix F.l. Results are given in Table F.2.1. TABLE F.2.1 Migration and visualization of graminin A from culture filtrate extracts of M-13, grown at different temperatures, on thin layer chrom- atography plates. Temperature (C) Rf Darkness of Spot 15 0.58 - 19 0.5756 light 25 0.58 dark authentic GRA OUbO very dark 2 pg 130 l?’.3 The optimum culture medium for growth of M-13 for maximum production of toxin (:iiltures of M-13 were grown in diptheria bottles for 32 days ant.25 C, photoperiod 12 hours. Three types of culture med- ium were used: Kobayashi and U1 medium (1979), Pool and Sharp medium (1969) and modified Fries No. 3 basal medium (Pringle and Scheffer, 1963). Cultures were extracted, pre- pared, run on TLC plates and visualized as in Appendix F.l. Results are given in Table F.3.l. TABLE F.3.l Migration and visualization of graminin A from culture filtrate extracts of M-13, grown in different media, on thin layer chromatopgraphy plates. Medium Rf Darkness of spot Kobayashi and Di 0.46 dark Pool and Sharp 0.48 light modified Fries 0.47 very light authentic GRA 0.49 very dark 2 pg 131 Conclusion: The optimum number of days of growth of M-13 for great- est toxin production was from 28-32 days. The optimum temp- erature for growth of M-l3 for greatest toxin production was 25 C. The optimum culture medium for growth of M-l3 for greatest toxin production was Kobayashi and Di medium. These conclusions were made before it was known that a contaminating compOund comigrated with GRA (Part 1). Quantifying the actual amount of GRA present by gas chrom- atographyhass spectroscopy may alter this conclusion. If the contaminant were removed prior to gas chromatography then this method of quantification would be adequate. 132 APPENDIX G Growth kinetics of CG-18 and 7-54 The growth curves obtained for CG-18 and 7-54 in potato dex- trose broth (PDB) and PDB plus methionine (20 pl/ml) are in Figure G.l. When grown in PDB, CG-18 remained in lag phase for approximately 30 hours before entering the exponential growth phase. Approximate generation time was 10 hours. The mutant 7-54 remained in lag phase for approximately 75 hours when grown in PDB. The generation time for 7-54 was appoximaely 12 hours. The addition of methionine to the PDB decreased the time CG-18 was in lag phase by approximately 6 hours and increased the generation time by approximately 5 hours. The time in lag phase for 7-54 was decreased by approximatley.1l hours with addition of methionine to PDB. The generation time for 7-54 was reduced to approximately 4 hours. This was one third of the generation time found in PDB. 133 Figure F.l Growth kinetics of CG-18 and 7-54 in potato dextrose broth and potato dextrose broth plus methionine (20 pl/ml). H.h MMDUHM Amer. ut~h am. no no we do on a. on o3 o: :3 ca— 0: C, . _ p b . _ . C . . _ r x. 1134 - 1 05.8.5.5 + talk . talk 05.8.5.5 + o—loo 07.8 ’8-‘ 31-1» . 1 . ‘00:? . . .052. . . roét.s . 1 SllNfl 1131M 1 007 'irj‘T—fj "" 009 BI BLI OGRAPI-IY Allen, E. H., and J. Kuc. 1968. a-Solanine and¢z-chocomine as fungitoxic compounds in extracts of irish potato tubers. Phytopathology 58:776-781. Anke, H., H. Schwab, H. Achenbach. 1980. Tetronic acid der- ivatives from As r illus panamensis. Production, isolation, characterization and SioIogical activity. J. Antibiot. 33:931-939. Ayers, P. G. 1978. Water relations of diseased plants. pp. 1-60 in: Water Deficits and Plant Growth. Vol. 5. T. T. Roz- lowski, ed. Academic Press, New York. Bailey, J. E. 1980. Cephalosporium leaf stripe of wheat: De- cline and mode of infection. Ph.D. Thesis, Michigan State Univ. Bailey, J. E., J. L. Lockwood, M. v. Wiese. 1982. Infection of wheat by Cephalospgrium ramineum as influenced by freezing of roots. P ytopatgoIogy 72:1324-1328. Ballantine, J. A., V. Fernto, C. H. Hassall, v. I. P. Jones. 1969. Asptertonin A and B, two novel tetronic acid deriva- tives produced b a blocked mutant of Aspgrgillus rugulosus. J. Chem. Soc. (C 1969:56-61. Ballio, A. 1981. Structure-activity relationships. pp. 395-441 in: Toxins in Plant Disease. R. D. Durbin, ed. Aca- demic Press. Barron, G. L. 1968. The Genera of Hyphomycetes from Soil. Williams and Wilkins Co., Baltimore, MD. Bockus, W. W., and T. Sim Iv. 1982. Quantifying cephalospor- ium stripe disease severity on winter wheat. Phytopathology 72:493-495. Braun, A. C. 1937. A comparative study of Bacterium ggpgggm Wolf and Foster and Bacterium angulatum Fromme and Murray. Phytopathology 27:283-304. Bruehl, G. W. 1956a. Cephalosporium stripe disease of wheat in Washington. Phytopathology 48:178-180. Bruehl, G. W. 1956b. Prematurity blight phase of cephalo- sporium stripe disease of wheat. Plant Dis. Reptr. 40:237-241. Bruehl, G. W. 1957. Cephalosporium stripe disease of wheat. Phytopathology 47:641-649. 135 136 Bruehl, G. W. 1963. Hymenula cerealis, the sporodochial stage of Cephalospgrium gramineum. Phytopathology 53:205-20 . Bruehl, G. W. 1968. Ecology of cephalosporium stripe disease of winter wheat in Washington. Plant Dis. Reptr. 52:590-594. Bruehl, G. W., B. Cunfer, M. Toiviainen. 1972. Influence of water potential on growth, antibiotic production, and survi- val of Cephalospgrium gramineum. Can. J. Plant Sci. 52:417-423. Bruehl, G. W., and P. Lai. 1968a, Influence of soil pH and humidity on survival of Cephalosporium gramineum in infested wheat straw. Can. J. Plant Sc1. 4 :2 -2 . Bruehl, G. W., and P. Lai. 1968b. The probable significance of saprophytic colonization of wheat straw in the field by Cephalospgrium gramineum. Phytopathology 58:464-466. Bruehl, G. W., P. Lai, O. Huisman. 1964. Isolation of Ceph- alos rium gramineum from buried, naturally infested host e ris. P ytopathology 54:1035-1036. Bruehl, G. W., R. L. Millar, B. Cunfer. 1969. Significance of antibiotic production by Cephalosporium ramineum to its saprophytic survival. Can. J. P ant Sc1. 49:235-246. Buchanan, R. E. 1911. Morphology of the genus Cephalospgrium, with description of a new species and a var- iety. Myco ogia 3:170-174. Ceigler, A., R. W. Detroy, E. B. Lillehoj. 1971. Patulin, penicillic acid, and other carcinogenic lactones. pp. 409-434 in: Microbial Toxins V1 Fungal Toxins. A. Ciegler, S. Kadis, S. J. Ajl, eds. Academic Press, New York. Creatura, P. J., G. R. Safir, R. P. Scheffer, Y. D. Sharkey. 1981. Effects of Ce halos rium ramineum-and a toxic metab- olite on stomatal conductance of wheat. Physiol. Plant. Path. 19:313-323. Dimond, A. E. 1972. The origin of symptoms of vascular dis- eases. pp. 289-309 in: Phytotoxins in Plant Disease. R. K. S. Wood, A. Ballio, A. Graniti, eds. Academic Press, New York. Duniway, J. M. 1976. Water status and imbalance. Physiol. Plant Path. 4:403-409. 137 Ellingboe, A. H. 1979. Inheritance of specificity: The gene-for-gene hypothesis. pp. 3-17 in: Recognition and Specificity in Plant Host-Parasite Interactions. J. M. Daly and I. Uritani, eds. Jpn. Sci. Soc. Press, Japan. Ellis, J. R., and T. M. Mc Calla. 1973. Effects of patulin and method of application on growth stages of wheat. App. Microbial. 25:562-566. Fulbright, D. W., and A. v. Ravenscoft. 1981. Mutants of Cephalospgrium ggamineum . (Abstr.) Phytopathology 71:217. Gabriel, D. W., A. H. Ellingboe, E. C. Rossman. 1979. Muta- tions affecting virulence in Phyllosticta maydis. Can. J. Bot. 577:2639-2643. Glover, S. W. 1968. Mutation. pp. 13-26 in: Experiments in Microbial Genetics. R. C. Clowes and W. Hays, eds. Blackwell Scientific Pub., New York. Gray, E. G., and M. Nobel. 1960. Cephalosporium stripe in cereals in Scotland. Pl. Prot. Bull. F. A. 0. 8:46. Gray, L. E., and D. W. Chamberlain. 1975. Evidence for toxin production by a strain of Cephalospgrium gregatum. Phyto- pathology 65:89-90. Gudgeon, J. A., J. S. E. Holker, T. J. Simpson, K. Young. 1979. The structures and biosynthesis of multicolanic, multicolic and multicolosic acids, novel tettronic adic met- abolites of Penicillum multicolor. Bioorg. Chem. 8:311-322. Harper, J. R., and N. E. Balke. 1981. Characterization of the inhibition of K absorption in oat roots by salicylic acid. Plant Physiol. 68:1349-1353. Hawksworth, D. L., and J. M. Waller. 1976. H enula cerealis. C. M. 1. Descriptions of Pathogenic Fungi and Bac- teria. No. 501. Hodge, J. E., and Hofreiter, B.T. 1962. Determination of re- ducing sugars and carbohydrates. pp. 380-394 in: R. L. Whistler and M. L. Worfrom,eds. Methods of Carbohydrate Chemistry. Vol. 1 Academic Press, New York. Haynes, L. J., and J. R. Plimmer. 1960. Tetronic acids. Quart. Rev. (Chem. Soc. London) 14:292-315. 138 Holliday, R. 1956. A new method for the identification of 'biochemical mutants of microorganisms. Nature 178:987. Howell, M. J., and P. A. Burgess. 1967. Ce halos orium ram- ineum causing leaf stripe in grasses, and its spoodochiaI stage, H menula cerealis, on cereals and grasses. Plant Path. 18:67-70. Ikata, S., and I. Kawai. 1938. Studies in the stripe disease of wheat (translated title). Bull. Agr. Exp. Sta., Okayama, Japan, 111 p. (Abstr. in Rev. Appl. Mycol. 17:593). Johnston, R. H., and D. E. Mathre. 1972. Effect of infection by Ce halosporium gramineum on winter wheat. Crop Sci. 12: l7- . Kobayashi, K. 1980. Changes of gregatin A and graminin A in a host plant. J. Fac. Agr. Hokkaido Univ. 59:392-396. Kobayashi, R., and T. Ui. 1977a. Wilt-inducing antibiotic compounds produced by Cephalosporium gregatum. Physiol. Plant Path. 11:55-60. Kobayashi, H., and T. Ui. 1977b. Graminin A, a new toxic metabolite from Cephalospgrium gramineum Nisikado and Ikata. J. C. 5. Chem. Comm. 21: 74. Kobayashi, R., and T. Ui. 1979. Short communication: Phyto- toxicity and antimicrobial activity of graminin A, produced by Cephalospgrium gramineum, the causal agent of cephalo- sporium stripe isease of wheat. Physiol Plant Paht. 14:129-133. Lai, P. 1967. Aspects of the saprophytic life of Cephalo- spgrium gramineum. Diss. Abstr. 26:2219-2220B. Lai, P., and G. W. Bruehl. 1966. Survival of Ce halos rium gramineum in naturally infested wheat straw in soiI in the ie an in the laboratory. Phytopathology 56:213-218. Lai, P., and G. W. Bruehl. 1968. Antagonism among Ce halo- spgrium gramineum, Trichoderma spp. and Fusarium cuImorum. P ytopat o ogy :532-566. Large, E. C. 1954. Growth stages in cereals. Illustration of the Feekes scale. Plant Path. 3:128-129. Latin, R. X., R. W. Harder, M. V. Wiese. 1982. Incidence of cephalosporium stripe as influenced by winter wheat manage- ment practices. Plant Dis. 66:229-230. 139 Lederberg, J. 1950. Isolation and characterization of bio- chemical mutants of bacteria. Methods in Medical Research. 3:5-22. Luke, H. H., and R. H. Biggs. 1976. Phytopathogenic toxins from fungi: An overview. pp. 298- 317 in Mycotoxins and Other - Fungal Related Food Problems. J. v. Rodricks, ed. Adv. in Chem. Series 149. American Chemistry Society, Wash. D. C. McGinnis, M. R. 1980. Laboratory Handbook of Medical Mycol- ogy. Academic Press, London. Mathre, D. E., and R, H. Johnston. 1975a. Cephalosporium stripe of winter wheat: Procedures for determining host re- sponse. Crop Sci. 15: 591- 594. Mathre, D. E, and R. H. Johnston. 1975b. Cephalosporium stripe of winter wheat: Infection processes and host re- sponse. Phytopathology 65: 1244-1249. Mathre, D. E., and R. H. Johnston. 1979. Decomposition of wheat straw infested by Cephalospgrium gramineum. Soil Biol. Biochem. 11:577-580. Mathre, D. E, R. H. Johnston, C. F. McGuire. 1977.Cepha1o- sporium stripe of winter wheat: Pathogen virulence, sources of resistance, and effect on grain quality. Phytopathology 677: 1142- 1148. Miller, J. H. 1972. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, New York. Morrison, R. T, and R. N. Boyd. 1975. Organic Chemistry. 3rd ed. Allyn and Bacon, Inc., Boston. Morton, J. B., and D. E. Mathre. 1980a. Physiological effects of Cephalospgrium gramineum on growth and yeild of winter wheat cu tivars. Phytopat o ogy 70: 807-811. Morton, J. B., and D. E. Mathre. 1980b. Identification of resistance to cephalosporium stripe in winter wheat. Phyto- pathology 70:812-817. Morton. J. B., D. E. Mathre, R. H. Johnston. 1980. Retation between foliar symptoms and systemic advance of Cephalospg - ium ramineum during winter wheat development. Phytopat o ogy 70: 802- 807. 140 NCA-14 Meetings. 1979. Report from Chairman, Kansas State University. Nisikado, Y., H. Matsumoto, K. Yamauti. 1934. Studies on a new Cephalospgrium, which causes the stripe disease of wheat. O ara Inst. Landw. Forsch. Ber. 6:275-306. Otiano, N. C. 1962. Ingress of CephalosporiuI gramineum into winter wheat seedlings and subsequent isto ogic re ation- ships between the suspect and the pathogen. Diss. Abstr. 22- 3804. Patil, S., A. C. Hayward, R. Emmons. 1974. An ultra- violet-induced nontoxigenic mutant of Pseudomonas haeoli- cola of altered pathogenicity. Phtopathology 64:590-595. Pattenden, G. 1978. Natural 4-y1idenebutenolides and 4-ylidenetetronic acids. Fortsch. Chem. Organ. Natst. 35:133-198. Pool, R. A. F. 1967.Cepha1osporium stripe of winter wheat: Disease processes and effects. Diss Abstr. 28: 35573. Pool, R. A. S., and E. L Sharp. 1966. A relationship between autumn root growth of winter wheat and infection by Cephalo- spgrim gramineum. (Abstr. ) Phytopathology 56: 895. Pool, R. A. F., and E. L. Sharp. 1967. Soil moisture as a factor affecting the incidence of cephalosporium stripe dis- ease of winter wheat. (Abstr. ) Phytopathology 57: 1008. PooL R. A. F., and E. L. Sharp. 1969a. Possible association of a polysaccharide and an antibiotic with the disease cycle of Cephalosporium stripe. Phytopathology 59: 1763- 1764. Pool, R. A. F., and E. L. Sharp. 1969b. Some environmental and cultural factors affecting cephalosporium stripe of win- ter wheat. Plant Dis. Reptr. 53: 898- 902. Pringle, R. B., and A. C. Braun. 1957. The isolation of the toxin of Helminthospgrim mvictoriae. Phytopathology -47: 369- 371. Pringle, R. B., and R. P. Scheffer. 1963. Purification of the selective toxin of Periconia circinata. Phytopathology 53:785-787. Richardson, M. J., and W. J. Rennie. 1970. An estimate of the loss of yeild caused by Cephalosporium gramineum in wheat. Plant Path. 19:138-140. 141 Rudolph, K. 1976. Non-specific toxins. pp. 270-315 in: En- cyclopedia of Plant Physiology. Vol.4. (new sereis). Phys- iological Plant Pathology. Springer-Verlag, Berlin. Scheffer, R. P. 1976. Host-specific toxins in relation to pathogenesis and disease resistance. pp. 247-269 in: Encyclopedia of Plant Physiology Vol. 4. (new sereis). Phys- iological Plant Pathology. Springer-Verlag, Berlin. Scheffer, R. P. 1983. Toxins as chemical determinants of plant disease. in: Toxins and Plant Pathogenesis. J. M. Daly, ed. Academic Press, Australia. (in press). Scheffer, R. P., and R. Livingston. 1980. Sensitivity of sugarcane clones to toxin from Helminthos orium sacchari as determined by electrolyte leakage. PEytopathology Scheffer, R. P., and R. B. Pringle. 1961. A selective toxin produced by Periconia circinata.. Nature 191:912-913. Schertz, K. F., and Y. P. Tai. 1969. Inheritance of reaction of Sorghum bicolor (L.) Moench to toxin produced by Peri- conia c1rcinata (Mans.) Sacc. Crop Sci. 9:621-623. Slope, D. B., and R. Bardner. 1965. Cephalosporium stripe of wheat and root damage by insects. Plant Path. 14:184-187. Smith, N. A., R. P. Scheffer, A. H. Ellingboe. 1966. Ceph- alosporium stripe pf wheat prevalent in Michigan. Plant Dis. Reptr. 50:190-191. Spalding, D. H., G. W. Bruehl, R. J. Foster. 1961. Possible role of pectinolytic enzyme and polysaccharide in patho- genesis by Cephalospgrium gramineum in wheat. Phytopathology 51:227-235. Staskawicz, B. J., and N. J. Panopoulos. 1979. A rapid and sensitive microbiological assay for phaseolotoxin. Phyto- pathology 69:663-666. Steel, R. G. D., and J. H. Torrie. 1980. Principles and pro- cedures of statistics. A biometrical approach. McGraw-Hill Book Co., New York. Steiner, G. W., and R. S Byther. 1971. Partial characteriza- tion and use of a host-specific toxin from Helminthosporium sacchari on sugarcane. Phytopathology 61:691-695. 142 Steiner, G. W. and G. A. Strobel. 1971. Helminthosporoside, a host-specific toxin from Helminthosporium sacchari X. . J. Biol. Chem. 246:4350-4357. Stoessl, A. 1982. Structure and biogenetic relations: Fungal nonhost-specific. pp. 109-219 in: Toxins in Plant Disease. R. D. Durbin, ed. Academic Press, New York. Tang, C. T., and C. Young. 1982. Collection and identifica- tion of allelopathic compounds from the undisturbed root system of Bigalta limpograss (Hemarthria altissima). Plant Physiol. 69:155-160. Tuite, J. 1969. Plant Pathological Methods. Fungi and Bac- teria. Burgess Publishing Co., Minn. Turner, N. C. 1973. Action of fusicoccin of the potassium balance of guard cells of Phaseolus vulgari . Amer J. Bot. 60:717-725. Turner, N. C., and A. Graniti. 1969. Fusicoccin. A fungal toxin that opens stomata. Nature 223:1070-1071. Turner, W. B. 1971. Fungal Metabolites. Academic Press. London. Uraguchi, R., and M. Yamazaki, eds. 1978. Toxicology: Bio- chemistry and Pathology of Mycotoxins. Halsted Press, New York. van Alfen, N. R., and B. D. McMillan. 1982. Macromolecular plant-wilting toxins: Artifacts of the bioassay method. Phytopathology 772:132-135. Wheeler,H. E., and H. H. Luke. 1955. Mass screening for dis- ease resistance in oats. Science 122:1229. Wiese, M. V. 1972. Colonization of wheat seedlings by Cephalospgrium ramineum in relation to symptom development. P ytopat ology 62:1013-1018. Wiese, M. V. 1977. Compendium fo Wheat Diseases. The Ameri- can Phytopathological Society, St. Paul, MN. Wiese, M. V., and A. V. Ravenscroft. 1973. Quantitative detection of propagules of Cephalosporium gramineum in soil. Phytopathology 63:1198-1201. Wiese, M. V., and A. V. Ravenscroft. 1975. Cephalosporium gramineum populations in soil under winter wheat cultiva- tion. P ytopathology 65:1129-1133. 143 Wiese, M. V., and A. V. Ravenscroft. 1976. Planting date affects disease development, crop vigor, and yeild of Mich- igan winter wheat. Res. Rep. 314, Michigan State Univ. Wiese, M. V., and A. V. Ravenscroft. 1978a. Cephalosporium stripe decline in a wheat monoculture. Plant Dis. Reptr. 62:721-723. Wiese, M. V., and A. V. Ravenscroft. 1978b. Sporodochium development and conidium production in Cephalospgrium gram- ineum. Phytopathology 68:395-401. Yoder, O. C. 1980. Toxins in pathogenesis. Ann Rev. Phyto- path. 18:103-129. Yoder, O. C. 1983. Use of pathogen-produced toxins in gen- etic engineering of plants and pathogens. in: Genetic Engin- eering of Plants. A. Hollaender, T. Kosuge, C. Meredith, eds. Plenum Pub. Corp., New York. ( in press ).