i LS: . mxggwgggJééskght a I... r t gum. KELS‘K‘!‘ . . . . . .- s . ‘ . 5....21r .. ._. .‘Trftfily‘... . a5. 33....“ 39...... z. 5....unhnv....¢mw ....z.t...§1,.:n..z ..b.z..§.l.itr . 17.. war... .. . . v3. . rm“ 4...} e I .‘r :3 2’: u o...u5.....fi. . .. c. u. s... .n..: mfififi- . «E... 4 .........,.........%a.... a... £33m fix... a. 3% awn. h. a... a. wwNmJnkm . u . 3 v a. .. . ‘ . r . . ..L.%fin Y... we»... W35 LVvte. .5. sfiavalia.rnh N7. 1" Vic val: $2.. m... ....m..,..;.. my... : .1 $.33... :. 1:. o t e . .aabfi.nn.3nq...wwn.mfiz§ “QNBVT swynr'Qr’Qw’v‘V‘vo‘ .7 ’ f s- 0. $1.21- ‘g’v .. pig-'6 6;, R“: 9v, i B, in 'l 1 54; I Eran» . it i E I g . Emu“ . fig} fir n In». 3’. .. v... . . tttptxn: 5.9. . .v' 7' . - uuv . No.1 _. :1... ad... w%..»h.u.n.mw.mh$kw Paw... .5 Ebmrrflfi L ..., at. 1. . E..m.nmm.( "Winn : .. :3 l v: any 0.. v u! . . . it .tl Muhfl~xf~¢ilo4£l wir . :3. r . 3? It: hithnnwflvw Mflcflc$1ufiomfi :41“. own... It». % 315:»; ‘ vtttih 1a (in 3:. ‘2... . .. k! _ . . . Q. . t T.-n.. S ~49. . a .3. .. ,Z.‘ n}; {11.1 _ : v? A . I 5!. D9 . ‘t a» . . . . . . . . 2 ~\. t. a... .5... . . 1...»...w .aqyfidfi. int..- m... 3%.... «$4.33... um. , $9.. . , , , . . azyvtihkbanh:.LW,X...Y ....;........Lfin# t1.l%vah..:1: fivde VI: r:- .. an“: .3 a 7v . .s .3. if. .fixu.~t:~...3.fl\wfu{4ks 3+... , .. itfltfirvfiflfiv 7% M “10.3% t r E g. 31!. IV 9. .H1 w . v . . _ :5 .3”: Nfiflfiflsio» iwsft..fi rift. v-fibmnbflu 19 . . . 7, wink . . . . . a?) . . 9 é. -. innit _ .3: .. 1.... .a .3 z. . an... hammmum s. «aft; :. I. . . . , , 24 hr: .1!!fl~.¥ti.. t‘fi..u..z.b$r,.kit at if .r ’v .320 .6 r E. 5. . .1 a . \. ht: P: Exxsh 1 L125 . .4 .trrrt. _ a v7 .5 .2. . 1... : fl . . Ivgivonrt v IV?! 61V ‘ I“ a $3. » It. '2‘ t E. ill!!~0£\5t; v o... a: ..? Kw “5.4Vth .6 46 V5 v ». t. v u r h m .3 #fiztacrtp...‘ . {Its}?! v...) 1.3.5. it .. 1 , m‘unfinv 039.53% it a 13. a lung... ~1¢¢§$~$§ untufiknafiir... .. . fie...” «aduwwugfi. . $1.0m» . .. ‘ .. ~ . . ufiE. vs»: an fiffififfiifivfififlbd. . n r 9... . . xx» .11. i‘ts I . .. ¥ 3 ill“?! b £312.a\¢2.§:6s5t1&hunu. o. a o‘f’ g «z. a: I? : l ‘ ~ .23? ‘73 .fi 3: $4; I {59 ii? , . 3 1L ‘ 33;. '$ 7 i? 1 ‘I 3 . 3‘ .K try. ‘ t.‘ . : .. . . .tfiavtltla-fi 5.. I»;.‘: . . .. : t. 19.34. :9 u. It‘d. auxin? vii-7b . . 0’9 t Litu é... sl.‘ ERIK-1.x.” r. r::...r3b 1.1.53... . _ 4.1.. . N13 (9 .. A . . o. ‘52 _ His.» 5. 3.3.. ivhfinl v.3» u INN“; . 1!..wku...» «“1483: . . a aflfiifilfint ‘ .. . 11.1). .4n9093. .. y .. | .- In... . . I . .. Virevr ilhrxvxgivnbh Vivackw‘ £5. 0.)th .v v.11}. . t t . . . . . A . ' . r. In I. vii-1...: .i I ‘ . . 14 l- . . t 5.!1. 1. 1!!! , .mkhz. «3.x.%&..b.sxm§tt. hum?“ ...z£k......mv¥,wm.m:w a. . a .. «.3... EB». _ 1..- uoarrva-a +. v . . . . x . Hg! t. 9. 1L} :1! Entry}: a—Qnaywnd. Huron... y .2. Wing :22! 21s . , . in»... , 4!. , in: Va. 13.0323y7! [IR-.2. 1. . *I‘Ivc‘cot vvonflig z... nmflkbrvt I a a 5 t. e a! . .rV4 {glinxb AWN-hi; Atnlbag {fund .1 l..u.ih:iv...:..::v.¥¥z ...£¥.....Sw.f..:¢n¢.vu;z¥ , x. .. u .I :bmaufltnuafin a». . Vi...h.5...${..n:.-25. .fi «$1432.5a9 1.4.2... i. . .2 a. 3.3.1:}: 3.... . $2.1}th v... .e . 3. .N...........u..§...‘..~.b... 2.....- a 1.5.7:“! .. or.“ . . . . 2. ii... 1. 2.6? 1.! 31......“ it... . E3. ~§f Efi¢ V!‘.YIV.5FIECI9OV ! . . . . , . . _. Lu!“ 3‘ . AN. ll». ‘5 . . r. ‘3‘...» x o 73‘ this; .. . fr. \V [Ni .. a; Zr. . a 2:1- 1 . . ...... .. . . .. .. .z. E. . 6...... fifinfi..siiunu¢$m§nm gun‘s; Emu ...w.0n€.»r,vh.ru&i 1.1... 2332.1. {9% Md»? . r .612... , .5. «yr... “32. .1..qk.>iz!utz&. $1.11. 3:34.12: .2..and.nnanu¢i..rndu Rand; . . a a this: - ,l .i; .. (Eguxndnlzn. REEL». _ ”pr Whit u...‘ {P «5:19. 7.519.? t h... to. .0122... .000... uni-.3)" zgfi. “flatlflflviorhukr’i x .L .i a . 1 .u .s. .u. 1.2 .9 . . .IIELL v1..w~0.~i.n...ru at”; Ive 2.19: '6‘: l. , ‘ . I? “k“..{IlCS . u $039396} .q 3:35;: 1 $1. . 5 Q36 . “35...; .. , .xuilhh.‘ts.r.!.€:vvntlpfr\. 19.. . . r. 125. .. fl 565 $3.25.. [the-.1. . . ._ €137...- Ix..r.zurviv.v.ahtin ufiqIXtanV. 0t Tg t {veil-.0"! tlvtfhfi Jag... . . OK. . ‘ quuvdtitsé ‘- L1 .3 «r Luv. £24.“...er .23.. d}. awnfiYK: .... uni... ,2; .1. .xzdhu: ‘ 2 . a ,. . «.03.; 5%.»): i ail? u. . Lf.l’:kr....x§t:!i\ .. .rt 61!. . l. iLtt I. r . , . . . «Skin. .. :3. LE... if»; M .ii . . : i.‘§oc¢{.iva¥oo.+r§ies It": x l.- g .. . I. rial... 5.25.11.10ch n! all .Ui .r!€.t..fle)v§7v£ » It! 9:11 1:. .. vo .. it ‘0: . . . . 4.1.... a... 9'1; 36.4.1.3: 1.1.... 6.5.7. .1 ifv‘. b‘n-beVtD. v!t$§!~l77:‘~ Y1: $9 luv a: U.- hé‘ztw.‘ » fig. .kazigrfizi ATE.I¢Z¢IO. 1|.)an58)..7-7|?01 VT. 1.50.53.35.th I 00.11. . . . bx...» .vOlEALfI. 1.5.... 3. . ...... ‘I11‘1‘Ifi‘90ftlvdvl. Q's-it"s; :2 ;. It... . . t§i53 l «v- suE... titratil50o¢ii§h I!!! .besD... ‘1‘; . . , . . . rituuixztkrvsnuxv g... .tv'lltlactosrrllivuiiovtv. 7'1... fitcfié :1..le . $959!}. 7:3...” .:. pnv..v)\.'or. . [illsxlrv¢ll .Qrfia‘: )0: u... s. . . . in... 011%.!A6‘TZK . . l. :vvrfv......:d: r Icfifil‘flflvhvvhray. : !§§ah§ . . }%tltr 76.2! .Ncu..vzv.rrn.w.u7t....os. 51.5 1.... vltoiti!’ .iletotthiEwttavt‘v n. .3; . . tumufiuhwfi I. v . . .. 'EF..7N9I . ‘ti‘lv'l > 'Blfl’i'1’O‘Qtl’r'.'3«n’n 'E.§ . .h .. val." v3? u... . )tllttnit .N’. . .o‘bfi 4 .ft.% §h|srz rs . Ififl fil‘Hff‘lfl. Lhtfit~AnythsW£tutiu \" ’A . . . 0.6 e 7.. kc". . .. «L Lg! 5.? ELEQNuaNh. x. flukuuzztdj‘; .5... . . u~fuf .- . hm»... . it .2460. .. 7,33%; . a Hittite...“ l. l I . 3G! .5): \ s ¢ .1 . A . .. ‘lhzi {Etiizzugutfi it... . flank... $5». fl, abhfiuugthutkivfz, . . u q to r: Ettiouaivrthnu «xiii-it. I z~ t &... 166 I3" 534%... EA. .m n 5 VI LL-‘ ,. Af m I l l |< < + + I (C) (D) Figure 2. Patterns of a gene—for-gene interaction under the hypothesis that a virulence gene must specifically interact with a susceptibility gene. (a) Resistance and avirulence dominant. (b) Resistance and virulence dominant. (c) Susceptibility and avirulence dominant. (d) Susceptibility and virulence dominant. "+" = a compatible interaction. ”—" = an incompatible interaction. issue of dominance is of no importance in determining which combinations are specific. Note that in Figure 1, regardless of dominance, only when an interaction between a resistance gene and an avirulence gene occurs, is incompatibility specified. All other combinations result in compati— bility, which is therefore nonspecific. Note that in Figure 2, regardless of dominance, only when an interaction between a virulence gene and a susceptibility gene occurs, is compatibility specified. All other combinations result in incompatibility, which is therefore nonspecific. The specificity pattern shown in Figure 1 was found by Flor to successfully predict the outcome of all combinations of host lines and pathogen isolates. The specificity pattern is the basis for the gene—for—gene hypothesis (9,10). The genetic pattern presented in Figurel implies that a physiological mechanism for incompatibility involves the positive function of a gene in the host for resistance and a gene in the pathogen for avirulence. Since resistance and avirulence are usually dominant (Figure la), positive physiological functions for susceptibility or for Virulence in the specificity interaction are unlikely. That a gene should function for avirulence in the pathogen rather than for Virulence is not intuitive, however, and alternative explanations in terms of physiological functions are often suggested. For example, a common idea is that a resistance gene is general in function, and that a virulence gene specifically overcomes a particular resistance gene. There are several problems with this idea. The first is the fact that avirulence genes are usually fully dominant. If virulence had a specific function, then a large number of ER heterozygotes should be less incompatible than EE homozygotes. This is not observed. The main problem with the alternative explanation is that genes which determine resistance can exist as multiple alleles. Some examples are: flax to flax rust (9), maize to maize rust (12), and barley to barley mildew (16). Each allele for resistance has a different E_gene specificity. Yet there are no examples of multiple allelism in E_genes. If the function of a E_gene was for virulence, then a corresponding multiple allelic series of E_genes would be expected in cases where such alleles occur in.E genes. Instead each E_gene in the pathogen which "overcomes" an.E allele (in the host) maps to a different locus in the pathogen. There are at least 12 identified alleles at the barley Ela locus (16). The 12 different alleles of the one Ela locus in different host lines and the 12 different E_genes of a single isolate of barley mildew are shown in Figure 3. If a P gene is specifically recognized by an.E gene to give incompatibility, then the E_genes could function for any purpose in the pathogen, yet still be pleiotropically recognized for incompatibility. 0n the other hand, if the E_genes must function to specifically overcome an E_alle1e, then 12 different loci in the pathogen must expend metabolic energy to produce products which are all targetted for one E gene locus. All twelve different E genes would have to be present, and perhaps all expend metabolic energy, whether the specific E allele was present or not, and only two E_a11e1es can be present at one time. The simplest explanation for the interactions depicted in Figure 3 is that only £21 and EEEI specifically interact. A complicated explanation is that the pathogen actively conditions 12 specific inter— actions for one locus when only two are possible in any given interaction. Essentially all variability in pathogenicity which has been found in nature in any host—parasite interaction not involving toxins fits the .sOHuomnoudH oaafiumaaoocfi cm H :I : .noauowumuua wanfiquEoo .oamm m mawcfiw m mo moaoaam o>aosu ecu mo mno Mom msowmuoaos ma mafia umon comm .mmcHH H :+: .m mpdwflm umom ufimpwwwfic o>HmBu nufi3 wumaomfi nowoauma medam m mo soauumuwuaa mo nuwuuma use SM .2m .3m .am .wm .3 6m. .3 + + + + + + + + + + + I ram .3 .Nm .HM .. madam Swim Smoam swam mmwm swam omom mdwmm «Mam mmmm NMNM %& 838.8 cowonuwm mahuoamw umom | 10 predictions outlined in Figure 1. (There are examples of suppressor genes which exhibit more complicated patterns, but their behavior is consistent with the gene-for—gene hypothesis (7,13). Since there are specific reactions between host and parasite for incompatibility, there must be other genes than these involved in promoting pathogenicity. In other words, gene-for—gene incompatibility must be superimposed on a basic compatibility (4). With the exception of host—specific toxins (23), tumor—inducing plasmids (20), and possibly pectic enzymes (17), there are no known examples of naturally—occuring variability in genes which promote pathogenicity. This is surprising. Variability in these genes would presumably resemble that of E_genes, except that they would either be non—specific (no requirement for a particular host gene for function) or their specificity would be for compatibility. Yet there are no reports of this in the literature. If genes involved in the promotion of pathogenicity exhibited a gene— for—gene specificity, they would fit the expectations of the pattern presented in Figure 2. The host-specific toxins may fit the pattern, since the toxin molecules are specific for host genes (19). But one of the assumptions of the pattern in Figure 2 is that one and only one gene in the host is specific for one and only one gene in the pathogen. A toxin is a small molecule, made by enzymes, and not itself a primary gene product. Toxin production would therefore not be expected to be controlled by a single gene, a necessary requirement for the pattern in Figure 2. The genetic data on gene—for—gene specificity offer no examples of more than single gene involvement in either host or parasite for any given 11 incompatibility reaction. Any proposed physiological mechanism of pathogenic restriction must be consistent with this data. A particular group of hypotheses, involving recognition signals triggering a common defense mechanism--~the phytoalexin hypotheses (2)——-have dominated research and thinking in this area (1). Yet the genetic data are inconsistent with the phytoalexin hypotheses (5). Different genes for resistance in highly isogenic backgrounds are known to have different effects on pathogen development (3). Even though inhibitory compounds may be produced by a plant in response to incompatible pathogens, such responses can be elicited by a wide variety of stimuli, including the injection of aspirin (21). The role of phytoalexins in disease resistance appears to be secondary to the incompatibility response. Since the genetic data offer no examples of more than single gene involvement in either host or parasite for specificity, the direct interaction of two primary gene products, one from a resistance gene and one from an avirulence gene, to determine incompatibility is a reason— able hypothesis. Proteins are the primary products of genes, and they have the requisite capacity for variation. Furthermore, proteins can be very sensitive to small changes in temperature. There are no examples of small molecules losing function or specificity with a 5 C rise in temperature. A loss of specificity and function by proteins and enzymes with a 5 C rise in temperature is a well documented characteristic of these molecules which is related to their tertiary folded structure. Certain gene—for—gene incompatibility interactions appear to be temperature— sensitive. For example, the E36 gene in wheat for resistance to Puccinia graminis is specific for the E6 gene in the pathogen to give incompatibility of the interaction at 20 C. Yet a 5 C rise in temperature completely 12 negates the incompatibility and leaves the nonspecific compatible reactions unchanged (see Figure 4)(15). Gene—for-gene specificity appears to involve a protein: protein or a protein: nucleic acid interaction (5). If a protein: nucleic acid interaction determines incompatibility, a resistance or an avirulence gene product might be a nucleic acid regulatory protein, such as a non—histone chromosomal protein. These proteins are thought to function as specific regulators of gene trans— cription in eukaryotic cells. Two—dimensional electrophoresis of these proteins from He—La cells reveals more than 450 proteins which are rare (less than 10,000 copies/ cell) and are not detectable in the cytoplasm (18). Such proteins, if involved in disease resistance, would be difficult to detect. 13 Host Host Sr6 sr6 Sr6 sr6 S I: co 36 - + 3}, E6 + + o o .5 .1: g p_6 + + g 36 + + m m 20 C 25 C Figure 4. The interaction of alleles at the E6 locus in Puccinia graminis with alleles at the E36 locus in wheat at two slightly different temperatures. "+" = a compatible interaction. "—" = an incompatible interaction. LITERATURE CITED 10. ll. 12. 13. LITERATURE CITED Daly, J.M. 1979. Basic mechanisms of host/pathogen interactions. 33 Recent Advances in Tobacco Science, Vol. 5. W.F. Kuhn, ed. pp3—25. Deverall, B.J. 1976. Current perspectives in research on phytoalexins. IE. Biochemical Aspects of Plant Parasitic Relationships. J. Friend and D.R. Threlfall, eds. Academic Press, N.Y. pp208—23. Ellingboe, A.H. 1972. Genetics and physiology of primary infection by Erysiphe graminis. Phytopathology 62: 401—6. Ellingboe, A.H. 1976. Genetics of host—parasite interactions. 33 Physiological Plant Pathology. R. Heitefuss and P.H. Williams, eds. Springer—Verlag, N.Y. 890 pp. Ellingboe, A.H. 1981. Genetical Aspects of Active Defense. Manuscript submitted for publication. Ellingboe, A.H. and D.W. Gabriel. 1977. Induced conditional mutants for studying host—pathogen interactions. 33 Use of Induced Mutations for Improving Disease Resistance in Crop Plants. A. Micke, ed. IAEA, Vienna. pp35—46. Flor, H.H. 1946. Genetics of pathogenicity in Melampsora lini. J. Agr. Res. 73: 335—57. Flor, H.H. 1947. Inheretance of reaction to rust in flax. J. Agr. Res. 74: 241—62. Flor, H.H. 1956. The complementary genic systems in flax and flax rust. Adv. Genet. 8: 29—54. Flor, H.H. 1971. The current status of the gene—for—gene concept. Ann. Rev. Phytopathol. 9: 275—96. Heitefuss, R. and P.H. Williams. 1976. Physiological Plant Pathology. Springer—Verlag, N.Y. 890 pp. Hooker, A.L. and K.M.S. Saxena. 1971. Genetics of disease resistance in plants. Anna Rev. Genet. 5: 407—24. Lawrence, G.J., G.M.E. Mayo and K.W. Shephard. 1981. Interactions between genes controlling pathogenicity in the flax rust fungus. Phytopathology 71: 12—19. 14 14. 15. l6. l7. 18. 19. 20. 21. 22. 23. 15 Loegering, W.Q. 1978. Current concepts in interorganismal genetics. Ann. Rev. Phytopathol. 16: 309—20. Loegering, W.Q. and J.R. Geiss. 1957. Independence in the action of three genes conditioning stem rust resistance in Red Egyption Wheat. Phytopathology 47: 740—1. Moseman, J.G. 1966. Genetics of powdery mildews. Ann. Rev. Phytopathol. 4: 269—90. Mount, M.S. 1978. Tissue is disintigrated. 13 Plant Disease, an Advanced Treatise, Vol III. J.G. Horsfall and E.B. Cowling, eds. Academic Press, N.Y. pp279-97. Peterson, J.L. and E.H. McConkey. 1976. Non—histone chromosomal proteins from HeLa cells. J. Biol. Chem. 251: 548—54. Scheffer, R.P. 1976. Host—specific toxins in relation to pathogenesis and disease resistance. E§_Physiological Plant Pathology. R. Heitefuss and P.H. Williams, eds. Springer—Verlag, N.Y. pp247—69. Watson, B., T.C. Currier, M.P. Gordon, M.D. Chilton and E.W. Nester. 1975. Plasmid required for virulence of Agrobacterium tumefaciens. J. Bacteriol. 123: 255-64. White, R.F. 1979. Acetylsalicylic acid (aspirin) induces resistance to tobacco mosaic virus in tobacco. Virology 99: 410—12. Wood, R.K.S. 1967. Physiological Plant Pathology. 33 Botanical Monographs, Vol. 6. W.Q. James and J.H. Burnett, eds. Blackwell Scientific Publications, Oxford. 570 pp. Yoder, 0.C. 1980. Toxins in pathogenesis. Ann. Rev. Phytopathol. 18: 103—29. CHAPTER I THE INDUCTION AND ANALYSIS OF TWO CLASSES 0F MUTATIONS AFFECTING PATHOGENICITY IN AN OBLIGATE PARASITE 16 CHAPTER I THE INDUCTION AND ANALYSIS OF TWO CLASSES 0F MUTATIONS AFFECTING PATHOGENICITY IN AN OBLIGATE PARASITE Abstract Twenty—nine mutations to increased virulence were induced in the pathogen Erysiphe graminis f. sp. tritici. Selection was made for a change from infection type 0 to infection type 4 on congenic lines of Triticum aestivum with either of two unlinked E genes, Egla or Em4a. Since no other changes in pathogenic specificity were detected, the mutations were interpreted as affecting the corresponding gene loci Ela and E4a in the pathogen. When the mutants were inoculated onto congenic susceptible plants, the rate of infection and the final infection type was indistinguishable from the wild type. The pathogen can apparently sustain a mutation at the Ela or the E48 locus without affecting its parasitic capability on plants with pmla and pg4a. These loci therefore appear to function for avirulence and not for virulence. Nineteen temperature—sensitive mutations were also induced in the pathogen. Growth and development of the mutant isolates could be stopped by raising the temperature from 20 C to 25 C. Growth could be resumed by lowering the temperature, even after a week at the high (restrictive) temperature. The infection types on plants with E3 genes were unaffected by the temperature—sensitive mutations. The use of conditional mutations to generate variability in genes essential for parasitism and to manipulate pathogen growth and development is discussed. 17 Introduction Erysiphe graminis f. sp. tritici, an obligate parasite of wheat, can be induced to infect a plant with high efficiency and with reasonable synchrony during the first 58 hours after inoculation (8). Genetic studies of the variability in host—parasite interactions with powdery mildew have been made in both host and parasite. The interactions have been shown to follow the gene—for-gene pattern (11). E, graminis f. sp. tritici was chosen for this study because it has morphologically well— defined stages of development and the effects of variation in genes controlling early events crucial to parasitism can be studied in detail. Induced mutations to increased virulence have been reported in at least seven different pathogens (1). Mutations to avirulence on host plants with specific E;genes are rare. This may be due to the relative ease with which mutations to virulence can be selected using host plants with known E genes. Mutations to specific avirulence would be more difficult to select. Since mutations almost always have deleterious effects on genes, the frequency of forward mutations (losses of function) is always higher than that of reverse mutations (regaining of function). If genes function for specific avirulence, then mutations to increased virulence should be more frequent than mutations to decreased virulence. The gene—for—gene hypothesis successfully predicts the patterns of interactions between E genes in a host and E_genes in a pathogen. Although this genetic hypothesis offers limited insight into physiological mechanisms for pathogenic restriction, it suggests that certain hypotheses of mecha— nisms of interaction are improbable (3). For example, the genetically specific interaction between a host E gene and a pathogen_E gene is for incompatibility. Only the_E and E alleles are involved. The alternate 18 alleles, 3 and p, are not genetically specific. The simplest inter— pretation of the genetic data is that the E_gene codes for a product active in promoting avirulence. Since E_genes do not appear to actively promote pathogenicity, then other genes must be involved in that function. Yet there are no examples (other than the host—specific toxins) of naturally—occurring variability in genes whose function is necessary for pathogenicity (2). Perhaps variability of those genes is lethal or gives a selective disadvantage to the pathogen. This investigation was undertaken with three objectives in mind. The first was to induce mutations in the pathogen to an increase in virulence on host plants with single E genes. We also wanted to determine if mutations of this type affected the interactions between the mutant parasite and host plants with recessive 3 alleles. The second objective was to induce temperature—sensitive mutations that allow the development of the parasite at the normal (permissive) temperature but not at a higher (nonpermissive) temperature. The latter are presumed to be mutations of genes whose function is crucial to successful parasitism. The third objective was to observe the characteris— tics of any mutations obtained with a View towards distinguishing between alternative physiological hypotheses for_E gene function. If E_genes function for avirulence, then: 1. mutations to increased virulence on plants with single_E genes should be easily obtained; 2. mutants should still be able to parasitize plants with 3 genes as well as the wild type; and 3. mutations of genes which are crucial for parasitism should not show E gene specificity. On the other hand, if E_genes function for virulence, then: 1. mutations to wider virulence should be difficult to 19 obtain; 2. at least some of the mutants should be less virulent than the wild type on host plants with the recessive 3 alleles; and 3. mutations of genes which are crucial to parasitism should show E gene specificity. 20 Materials and Methods Stock cultures. The wild—type culture of Erysiphe graminis f. sp. tritici (MS-1) was maintained as follows. Conidia from plants inoculated seven days earlier were inoculated onto 6—7 day old wheat seedlings (cv. Little Club) grown in four inch pots. Inoculated plants were kept in a growth chamber at 18—20 C during the light period and 16—18 C during the dark period. Light intensity was 700—800 ft.-c. on a 15 hour photo— period. Relative humidity ranged from 70-90% during the light period to 90—100Z during the dark period. These conditions will be referred to as "standard conditions” throughout the paper. Mutant cultures were maintained as described above in separate growth chambers, or under 8.5 inch glass chimneys (#845310, Corning Glass Works) topped with four layers of cheesecloth. Stock cultures were also main— tained for up to four weeks in 4 inch test tubes on cut leaves with the base of the leaf immersed in 1 m1 of benzimidazole (125 ug/ml) in water. The tubes were stored in growth chambers under standard conditions and were capped with ribbed test—tube caps to allow for gas exchange. Isolates were also stored for up to two months on 1—3 seedlings grown in 12 inch test tubes containing 2 inches of vermiculite, capped with a cotton plug. Seven days after inoculation, these were placed in a refrigerator at 4—10 C, in continuous light from a single 24 inch fluorescent light. Induction of Mutations. Seven—day—old leaves which had been lightly inoculated with conidia 15 hours previously were cut at the base and placed in 15 cm test tubes containing benzimidazole (100 ug/ml) and N—Methyl—N'—nitro—N—nitrosoguanidine (NTG) (125 ug/ml) (Aldrich Chemical Company, Milwaukee, WI 53233) in 2 ml of water. The tubes were placed uncapped in a growth chamber with the lights on for 8 hours under standard 21 conditions to promote transpiration. Isolation of Mutants with Increased Virulence. After 7—10 days of NTG treatment, the leaves were removed from the tubes and conidia from treated mildew colonies were rubbed onto 7—day—old seedlings of Chancellor wheat. Chancellor contains no known_E genes for resistance to culture MS—l. The inoculated plants of Chancellor were then placed under standard conditions used for induction of synchrony (10). An uninoculated 4 inch pot of wheat seedlings with gene Egla and another pot of seedlings with gene E34a were placed next to the inoculated Chancellor plants in each experiment. The wild—type culture MS-l contains the corresponding genes Ela and E4a, and on both the E913 and E34a isolines gives an infection rating of 0 on a 0—4 infection type scale. MS-l is compatible (infection type 4) with Chancellor. After seven days of growth on Chancellor, the conidia produced on Chancellor were daily dusted onto the Egla and E34a isolines in the same growth chamber. After a further seven days, the Egla and Eg4a isolines were examined for mildew development. Any colonies growing on these lines were increased on the same line, and isolates which could be continuously maintained on plants with Emla or Eg4a were isolated and considered to be mutants. Only one mutant from each host line was isolated during each experiment to insure the independence of origin of each mutant culture. I Several precautions were taken to guard against the possibility of contamination by other virulent isolates. First, all experiments were conducted in a building physically isolated from any known sources of mildew. Some of the mutants were obtained in the winter, when contamina— tion from outdoor sources was impossible. Second, all mutants were tested 22 for their reactions on four host lines with Egla, Eg2a, E33a, or EE4a. These differentials could distinguish the isolates of mildew kept in culture in our laboratory. Two of the isolates, one virulent on Egla and the other virulent on Eg4a, were also tested against isolines Eg4b, E36,.Em7, EEE, Lehmi, 15889, 17339, and Triticum sphaerococcum. We detected no contaminants in our tests of the mutants during the course of this work. Third, different field isolates of mildew show unique polypeptide maps by two—dimensional electrophoresis (6). Three of the mutants were extensively examined by two—dimensional electrophoresis and their polypeptide maps were indistinguishable from each other and from the wild—type culture MS—l (6). Isolation of Temperature—Sensitive Mutants. After four days of NTG treatment, the leaves were removed from the tubes and as few conidial chains as possible (usually 1—5) were removed from well—separated single pustules. A fine platinum wire was used to transfer the chains of spores onto a leaf of Chancellor. The leaf was then cut at the base and placed in benzimidazole (125 ug/ml) in 1 ml of water in 4 inch test tubes. The tubes were capped and placed in a growth chamber under standard conditions. After eight days spores from individual pustules were transferred by gently brushing the pustule against two leaves. Each leaf was cut at the base and placed in the benzimidazole solution in 4 inch test tubes. One of the inoculated leaves was kept at 20 C under standard conditions, and the other kept at 25 C under otherwise standard conditions. After one week the isolates were scored for growth. Isolates were considered temperature—sensitive if growth appeared normal or nearly normal at 20 C (the permissive temperature) and if no growth was observed at 25 C (the restrictive temperature). These tests were performed three times in test 23 tubes. Isolates which were scored as temperature-sensitive in 3 trials were transferred on Chancellor wheat in 4 inch pots under chimneys each week for the duration of the experiment. 24 Results Mutants with Increased Virulence. Twenty—nine constitutive mutations with increased virulence were induced. Sixteen had increased virulence on plants with Egla and 13 had increased virulence on plants with Eg4a. The wild-type culture MS—l gave an infection type 0 on these isolines, and all mutants gave an infection type 4 on the isoline on which they were selected. All mutants were tested for infection type against host lines with other Em genes and all gave reactions indistinguishable from MS—l except for the selected change in virulence. When the mutants were grown on Chancellor wheat, which has the recessive 331a and 334a alleles, the rate of infection and the final infection type were indistinguishable from the results with MS—l. Temperature—Sensitive Mutants. A total of 530 single pustule isolations were made following treatment with NTG. After at least three tests for temperature-sensitivity in test tubes, 46 mutants (8.6%) were isolated for increase of inoculum and further study. Of these, 19 (3.6%) proved to retain their temperature—sensitivity when grown on plants in 4 inch pots under chimneys. Growth and development of the isolates could be stopped by raising the temperature from 20 C to 25 C. Growth resumed after the temperature was lowered, even when plants were held at the restrictive temperature for a week. Microscopic observations of the mycelium of these mutants after a week at 25 C revealed that most of the hyphae, but not all, were lysed. All 19 temperature—sensitive mutants were evaluated for their reactions with the four host isolines Egla, EEZa, Eg3a, and Eg4a. No changes in virulence from that of MS-l were found. 25 Discussion In all gene—for—gene systems, pathogenicity affected by a host E gene is completely dependent upon the presence of a specific corresponding parasite E gene. The simplest genetic interpretation of the interactions with all possible combinations of host and parasite alleles involved in a gene—for—gene interaction is that specific recognition is for an incompatible relationship. Resistance is an active function. Avirulence is an active function. (For reviews see 2 and 5). Therefore a mutation from avirulence to virulence would be from one particular specificity to any other———or no———specificity. Mutations of any particular E gene should be easily obtained, unless the gene has some other function essential to the organism (7). Twenty—nine mutations of independent origin to increased virulence on host lines with single E3 genes were obtained. Each mutant had increased virulence only on plants with the selected E3 gene. Because the change in reaction is so specific to plants with only one E3 gene, it is considered likely that the mutation in the parasite is at the corresponding E locus and not at suppressor loci. Also, had suppressor genes been involved, a great many polypeptides might have been affected, and this would be expected to result in greatly alterred two—dimensional polypeptide maps (6). The polypeptide maps of three of the mutants were indistinguishable from maps of MS—l, and from each other (6). We tentatively conclude that the mutational event partially or completely destroyed the specificity of a E gene (Figure 5). Isolate MS—l, with genes Ela and E4a, gave an infection type 0 on plants with either Egla or E§4a. MS—l was easily mutated to give an infection type 4 on plants with either of these two genes. If the 26 Host Genotype Inferred Egla Emla 391a 339a 331a 331a Pathogen Genotype pg4a 334a Em4a EE4a 334a pg4a _la E4a (MS—l) — - + pla p4a (mutant) + — + Figure 5. Reactions of MS—l and a mutant derived from MS-l on three congenic isolines and the inferred mutational event. "+" = a compatible interaction. "—" = an incompatible interaction. 27 recessive E_alleles functioned to specifically recognize and overcome particular E3 genes, then a mutation from avirulence to virulence would be from no function to a gain of specific function. Such mutations should be difficult to obtain, because the creation of a specific function requires non—random, highly specific changes. Mutations which by chance completely negate the effect of the E3 gene and restore maximum, as opposed to partial, compatibility should be rare indeed. At least some of the mutants might have been expected to give an intermediate infection type, but none were recovered. The ease with which MS—l was mutated 29 independent times from incompatibility (infection type 0) to compatibility (infection type 4) with host lines with either of two unlinked E3 genes argues against the idea that the p_allele actively interacts with the E3 allele to give a compatible reaction. Since the mutations to increased virulence do not measurably affect compatibility of the mutants with plants with recessive 33_genes, it appears that the E_loci do not affect basic ability to parasitize. Parasite strains with the mutant 3% gene appear to grow and develop on plants with the recessive 33 genes similar to parasite strains with the wild—type E gene. A loss in specific avirulence to particular E3 genes, with no loss in parasitic capacity in comparison to the wild—type, was readily obtained by mutation. The simplest interpretation is that E genes can be altered without loss of virulence. Since E_genes did not appear to be involved in the basic ability of E. graminis f. sp. tritici to parasitize, a search was initiated to identify genes that might be. With the notable exception of the host—specific toxins, we know of no examples of naturally—occurring variability in genes whose function(s) promote basic parasitic capability. This may be because 28 variability in genes whose function is necessary for parasitism is selected against in nature. For example, several researchers studying stabilizing selection have discovered that there are differences in fitness between pathogen isolates which are not considered attributable to gene— for-gene interactions (9). These "fitness" genes may be examples of naturally-occurring variability in genes crucial to the basic ability to parasitize, crucial to basic life functions unrelated to parasitism, or both. A means has been suggested to distinguish genes necessary for parasitism from those which are not (4). The effects of fitness genes are slight, and the assay is tedious. One way to overcome the problem may be to artifically generate major variability in genes essential for parasitism. One of the best ways to study variation in essential genes is to use conditional, temperature—sensitive mutations. The mutant phenotype is expressed only at the restrictive temperature———usually slightly elevated above normal-——and is not expressed at the permissive temperature, usually a temperature normal for the organism. Thus a mutation affecting parasitic growth of even an obligate parasite can be maintained, and the mutation will be expressed by raising the temperature. The temperature-sensitive mutations isolated in this study did not affect the gene-for—gene specificity. At the restrictive temperature, growth of the pathogen on wheat lines with 33 genes was blocked or slowed, indicating that the primary defect was in a gene whose product was necessary for growth and development of the parasite. It is possible that none of these mutations affected the machinery involved in parasitism 333 33, but rather affected only basic life functions, such as protein or DNA synthesis. This possibility is unlikely, however, since observations 29 of all mutants revealed that germination and limited mycelial growth did, in fact, occur in cultures continuously exposed to the restrictive temperature immediately after inoculation. Prolonged exposure to restrictive temperatures caused lysis of most of the mycelium of the temperature—sensitive mutants. Lowering the temperature to permissive levels, even after a week at restrictive levels, allowed resumption of normal growth of at least some of the parasite units. Temperature—sensitive mutants with the ability to resume normal growth after cessation of growth should prove to be of immediate use in applications where regulating the growth of the pathogen is desired. For example, hypotheses involving rates of accumulation and efficacy of phytoalexins at disease sites could be tested using these conditionally expressed mutants. These experiments demonstrated the feasibility of artifically generating variability in the pathogen affecting two distinctly different aspects of pathogenicity. The mutations which were induced to increased virulence were involved in gene—for—gene relationships, but did not affect basic ability to parasitize. The temperature—sensitive mutations affected basic ability to parasitize, but were not involved in gene-for—gene specificity. The functional role these genes play in mechanisms of pathogenic restriction or capability has not been determined, but certain expectations as to their general roles have been hypothesized. The ready availability of variation in these genes should facilitate the isolation of their gene products. 30 Literature Cited 10. 11. Day, P. 1974. Genetics of Host-Parasite Interaction. W.H. Freeman, San Francisco. pp140—1. Ellingboe, A.H. 1976. Genetics of host—parasite interactions. 33 Encyclopedia of Plant Physiology, New Series, Vol. 4, Physiological Plant Pathology, R. Heitefuss and P.H. Williams, eds. Springer— Verlag, N.Y. Ellingboe, A.H. 1981. Genetical aspects of active defense. Manuscript submitted for publication. Ellingboe, A.H. and D.W. Gabriel. 1977. Induced conditional mutants for studying host—pathogen interactions. 33 Use of Induced Mutations for Improving Disease Resistance in Crop Plants, A. Micke, ed. IAEA, Vienna. Flor, H.H. 1971. Current status of the gene—for—gene concept. Ann. Rev. Phytopathol. 9: 275-96. Gabriel, D.W. and A.H. Ellingboe. 1981. Polypeptide mapping by two— dimensional electrophoresis and pathogenic variation in field isolates and induced mutants of Erysiphe graminis f. sp. tritici. Manuscript submitted for publication. Gabriel, D.W., A.H. Ellingboe, and E.C. Rossman. 1979. Mutations affecting virulence in Phyllosticta maydis. Can. J. Bot. 57: 2639—43. Haywood, M.J. 1975. Genetic control of the development of haustoria of Erysiphe graminis f. sp. tritici on wheat. Ph.D. thesis, Michigan State University. 69 pp. Leonard, K.J. 1977. Selection pressures and plant pathogens. Ann. N.Y. Acad. Sci. 287: 207—22. Masri, 8.3. and A.H. Ellingboe. 1966. Germination of conidia and formation of appressoria and secondary hyphae in Erysiphe graminis f. sp. tritici. Phytopathology 56: 304—8. Powers, H.R., Jr. and W.J. Sando. 1960. Genetic control of the host—parasite relationship in wheat powdery mildew. Phytopathology 50: 454—7. CHAPTER II POLYPEPTIDE MAPPING BY TWO—DIMENSIONAL ELECTROPHORESIS AND PATHOGENIC VARIATION IN FIELD ISOLATES AND INDUCED MUTANTS OF ERYSIPHE GRAMINIS F. SP. TRITICI 31 CHAPTER II POLYPEPTIDE MAPPING BY TWO—DIMENSIONAL ELECTROPHORESIS AND PATHOGENIC VARIATION IN FIELD ISOLATES AND INDUCED MUTANTS 0F ERYSIPHE GRAMINIS F. SP. TRITICI Abstract Methods were developed to extract proteins from the mycelium and conidia of the obligate parasite Erysiphe graminis f. sp. tritici and the facultative parasite Colletotrichum lindemuthianum. These methods resulted in preparations suitable for separation by two—dimensional electrophoresis; over 600 peptides were routinely visualized when stained with silver. Proteins were extracted from E, lindemuthianum grown in shake culture. Polypeptide maps of a E, lindemuthianum mutant and the wild type showed differences in the positions of as many as 10% of the polypeptides. Proteins were extracted from E, graminis f. sp. tritici grown on wheat, without any detectable extraction of the wheat leaf proteins. Polypeptide maps of three E, graminis f. sp. tritici mutants with increased virulence were indistinguishable from each other and from the wild type culture MS—l. Two field isolates, MS-l and MO—lO, had unique polypeptide maps differing by five polypeptides. Independent segregation of the polypeptide differences and three known_E gene differences was observed in progeny of a cross between MS-l and MO—lO. 32 Introduction The inheritance of disease resistance has been studied in almost every commercially grown crop because host plant resistance is an effective means of disease control. The inheritance of variability in the pathogen as well as in the host has been made in a smaller number of host—parasite combinations (2). These latter studies have demonstrated that genes for resistance in the host require specific genes in the pathogen for expression. the gene in the pathogen specifically required for an.E gene to be expressed is called a E_gene. Although the genetics of host-parasite interactions are well-known, the molecular basis of the restriction of development of a pathogen is unknown. Certain characteristics of the interactions, particularly temperature—sensitivity, support an hypothesis that the active products of E_genes in a pathogen and E genes in a host are proteins. It has been suggested that the protein product of a E gene and the protein product of the corresponding E gene interact to form a dimer and that the dimer is the active molecule which determines the fate of the interaction (1). Two—dimensional (2-d) electrophoresis is a very powerful analytical tcxnl for the separation of proteins. The technique is potentially Ciipable of resolving over 7,000 proteins on a single gel (7). Proteins alfe separated by isoelectric focusing in one dimension, and by molecular Sieving in acrylamide gels in the second dimension. Small variations in (truarge or in molecular weight are detectable with this technique. IVIutations that lead to an amino acid substitution can lead to a change :111 the net charge of a polypeptide. Such variants will lead to a disPlacement of the polypeptide in the isoelectric focusing dimension. ¥ 33 Mutations of codons affecting chain termination can lead to a polypeptide of different size. Such variants will lead to a displacement of the polypeptide in the molecular sieving dimension. Mutations of E genes in the pathogen, whether artificially induced or naturally occurring, should be accompanied by detectable changes in the electrophoretic mobilities of the proteins responsible for the observed phenotypic changes. This paper reports the extraction, separation, and visualization total protein from the surface mycelium and conidia of the obligate parasite E, graminis f. sp. tritici and from the mycelium of E, lindemuthianum. The objective was to determine whether or not any the peptides could be associated with particular E genes. 34 Materials and Methods Isolates of Erysiphe graminis f. sp. tritici. Two isolates of Erysiphe graminis f. sp. tritici, MS—l and MO—lO, were maintained on Triticum aestivum L "Little Club" or "Chancellor" in separate growth chambers under standard conditions as described previously (4). MS—l and MO—lO differ by at least three E_genes and are of opposite mating type. They have been crossed and offspring were recovered with all eight possible combinations of the three E_genes. (Offspring were provided by Ms. C. Bronson, Michigan State University). Progeny were either kept in separate growth chambers or more than one culture was kept in a chamber but with each isolate maintained on a wheat line susceptible to only one parasite culture. All wheat lines used were congenic. Each contained a known E3 gene obtained by eight backcrosses to the recurrent parent Chancellor. A host line containing, for example, the E3la gene will be referred to as the E3la isoline. (Seed of the E3 isolines and cv. Chancellor were supplied by Dr. John G. Moseman, USDA, Beltsville, MD 20705). Growth of powdery mildew on wheat is limited by an incompatibility system which exhibits genetic behavior of the usual gene—for—gene pattern (9). For example, strains found to be avirulent on the E3la isoline were found to have the Ela gene and isolates virulent on the Pmla isoline were shown to have a pla allele. Chemicals and Abbreviations: Tris, Tris(hydroxymethyl)aminomethane; SDS, Sodium Dodecyl Sulfate; DTT, (—) 1,4-Dithio—L-Threit01; EDTA, (Ethylenedinitrilo)~tetraacetic acid; NP—40, Nonidet P40 from Particle Data Laboratories, Elmhurst, IL; ultrapure urea from Schwarz/Mann, Orangeburg, N.Y.; Ampholytes from LKB Instruments, Rockville, MD; NTG, N—Methyl—N'~nitro—N-nitrosoguanidine. 35 Three isolates with increased virulence (from infection type 0 to infection type 4 on a 0—4 scale) against single E3 genes were derived from the wild type culture MS—l by mutagenesis as previously described (4). The mutations recovered in these isolates rendered the corresponding host genes E31a or E34a ineffective in promoting incompatibility. Mutant isolate lpl gave an infection rating of 4 on the E31a isoline, whereas MS—l gave an infection type 0. Mutant isolates 2p4 and 3p4 gave an infection rating of 4 on the E34a isoline, whereas MS—l gave an infection type 0. We believe that isolate lpl had a mutation at the E1 locus (from E1 to 31), and that isolates 2p4 and 3p4 had independent mutations at the E4 locus (from E4 to 34). No other phenotypic changes between the mutants and the wild type were detected (4). Protein extraction from Colletotrichum lindemuthianum. A tempera- ture sensitive mutant of E, lindemuthianum and the wild type isolate from which it was derived were grown in shake culture in a complete medium containing, per liter: 20 g glucose, 2 g Bacto—peptone, 0.46 g KHZPOA’ 1 g KZHPO4’ and 0.5 g MgSO4‘7H20. The mycelium of each isolate was collected in cheesecloth, rinsed twice with distilled water, frozen, and lyophilized. The lyophilized samples were ground to a dry powder with glass beads. About 0.5 g of ground mycelium was added to 10 m1 of ice cold SDS extraction buffer (0.05 M Tris—HCI, pH 6.8; 10 mM DTT; 2% SDS; 0.5 mM MgC12; 1 mM EDTA). The samples were held in a boiling water bath for five minutes with constant stirring, then cooled on ice and centrifuged at 4 C for one hour at 100,000 X g. A volume of supernatant containing approximately 400 ug of protein was precipitated in 9 volumes of ice cold acetone. The pellet was resuspended in 100 ul of SDS extraction buffer and sonicated until the pellet was dissolved. Solid urea was 36 added to a 9 M concentration, 200 ul of sample dilution buffer (9.5 M ultrapure urea; 2% ampholytes; 10 mM DTT; 1 mM EDTA; 8% (w/v) NP-40) was added, and the solution was vortexed until clear. Protein extraction from Erysiphe graminis f. sp. tritici. Extractions were made from very dense growths of E. graminis f. sp. tritici seven days after inoculation on wheat seedlings. The mildew—covered leaves of 50 to 100 seedlings were immersed in 10 ml of degassed, ice cold mildew extraction buffer (0.05 M Tris—HCl, pH 6.8; 0.04% SDS; 1 mM DTT; 0.5 mM MgC12; 1 mM EDTA; 2% (w/v) NP—40). A number 2 artist's brush was used to remove the surface mycelium and conidia from the immersed leaves, with minimal damage to the leaves. Buffer containing the fungus was then centrifuged for 30 minutes at 120,000 X g at 4 C. The supernatant was assayed for protein content and then precipitated in 9 volumes of ice cold acetone. The precipitate was dried under purified N resuspended 2, in sample dilution buffer, and sonicated. It was crucial to keep the time required at all steps of the extraction to a minimum. Electrophoresis and staining. Samples containing approximately 100 ug of protein were loaded without delay onto prefocused isoelectric focusing (IEF) gels prepared and run essentially as described by O'Farrell (7). Ampholytes in the desired pH range were used to fractionate the proteins in the first dimension. In these experiments the pH gradient ranged from 3.7 to approximately 7.2, a range which was found to reveal the most proteins from the extracts. This pH range was achieved by using 1.6% pH 5—7 ampholytes and 0.4% pH 3.5—10 ampholytes in both the IEF gels and the sample dilution buffer. Areas of ghe gel which were saturated by polypeptides were resolved by using ampholytes which spread out the pH range of interest. In some experiments, fractionation of proteins was 37 accomplished with ampholytes focusing in the pH ranges of 3.5—5, 5—7, and 7—10. Focusing in the range of 7—10 required the use of nonequilibrated gels in the first dimension (8). The second dimension gels were 11.25% acrylamide, and were run immediately following unloading from the IEF gels and equilibration in O'Farrell's buffer "0". In our experience, neither the samples nor the rod gels could be stored frozen for any length of time without significant loss of resolution. After electrophoresis, gels were stained with Coomassie blue (0.125% in 50% methanol/ 12% acetic acid) overnight and then destained in two changes of 10% ethanol/ 5% acetic acid. Gels which were judged to be slightly underloaded and of high resolution were then stained with silver exactly as described by Merril (6). Gels were photographed with trans- mitted light after staining with Coomassie blue and after silver staining. Photographs were taken without a filter on Kodak technical pan 2415 film, developed for maximum contrast. At least 10 replicate analyses were made of extracts from isolates MS—l, MO—lO, lpl, and 2p4. At least 5 replicate analyses were made of extracts from isolate 3p4 and of each of the progeny of isolates MS—l and MO—lO. 38 Results The preliminary experiments with E, lindemuthianum grown in shake culture demonstrated the feasibility of extracting fungal mycelium in preparations suitable for separation by 2—d electrophoresis (Figure 6). Over 300 spots were seen when the gels were stained with Coomassie blue. When preparations of the wild type culture were compared with preparations from the temperature-sensitive mutant grown at permissive temperatures, changes in the electrophoretic mobilites of at least 30 polypeptides were observed (Fig.6, a and b). This demonstrated that mutagenesis may affect a great many polypeptides, even when the phenotype has changed in only one detectable way. All growth of this temperature—sensitive mutant ceased at 28 C, although growth at 20 C was normal. Temperature—sensitivity was the only macroscopically observable phenotypic change. Since E, graminis f. sp. tritici is an obligate parasite, the mycelium was removed from the leaf surface of its natural host. Thus, there was a possibility that plant proteins were mixed with fungal proteins in the extracts. The mildew extraction buffer effectively lysed and extracted the fungal mycelium, but it did not appear to penetrate the waxy cuticle of fresh, undamaged leaves and extract the leaf proteins. When the extracts were electrophoresed and stained with Coomassie blue, at least 300 peptides of E, graminis f. sp. tritici were routinely observed, with a maximum of about 400 on certain gels (Figure 7, a and c). When silver stain was used, at least 600 spots were always visible, with a maximum of around 900 on the best gels (Figure 7, b and d). Uninoculated leaves were treated the same as the infected leaves, and the extracts were electrophoresed as a control to determine if any plant V"— .qa- 39 Figure 6. Two—dimensional gels of denatured proteins extracted from Colletotrichum lindemuthianum grown in shake culture. Proteins were stained with Coomassie blue. (a) the wild type. (b) a temperature— Sensitive mutant of the wild type. 40 Figure 7. Two—dimensional gels of denatured proteins extracted from Erysiphe graminis f. sp. tritici. (a) MS—l, stained with Coomassie blue. (b) the same gel as in ”a", stained with silver after staining with Coomassie blue. (c) MO—lO, stained with Coomassie blue. (d) the same gel as in ”c", stained with silver after staining with Coomassie blue. .4 2% 4: Figure 7. 42 Figure 7. 43 proteins were extracted from the leaf by this method. Ten very faint spots and 2 streaks appeared in these gels, when stained with Coomassie blue. When extracts were made from seedlings grown under sterile conditions (3), the spots were not visible but the two faint streaks remained. Thus, the faint spots on uninoculated leaves may have been from bacterial cells growing on the leaf surface. The streaking was eliminated by using less brush pressure in removing the mycelium from the leaf. The extracts from isolates MS—l and its 3 mutants were compared; no differences in electrophoretic mobility of any of the polypeptides ,_ could be found in 10 different experiments. The polypeptide maps were apparently identical for any given pH range. Although apparent differences could occasionally be found between isolates, especially when the isolates were run on different days, such differences were always found to be quantitative, and not qualitative. Proteins were also extracted and separated from another field isolate, MO—lO. MO—lO differed from MS—l by at least 3.E genes. Five qualitative differences in peptide positions were found in comparisons of the peptide maps of the two isolates. The peptide maps were unique to each isolate and highly reproducible. The positions of 4 of the 5 peptides that differentiate MS—l from MO—lO are indicated by numbered arrows in the electrophoretograms shown in Figure 7. The position of peptide number 4 cannot be clearly distinguished on these particular gels and so is not indicated in Figure 7. The alternate positions of peptide variants 1, 3, and 4 represent a change in the isoelectric focusing dimension, with no apparent change in molecular weight. These three changes are consistent 44 with the hypothesis that the alternate alleles present in MS—l and MO—lO code for one or a small number of amino acid substitutions that result in a net change in charge of the polypeptide but not in a size change. A polypeptide corresponding to polypeptide # 2 could not be found in MS—l, and a polypeptide corresponding to polypeptide # 5 could not be found in M0—10. MS—l is avirulent on isolines containing the genes E32a, E33a, or E34a, and its genotype may be written E2 E3 E4. MO—lO is virulent on these same isolines, and its genotype may be written 32 33 34. A cross between MS—l and MO—lO produced progeny of 8 types, as expected with unlinked genes at three loci. The inferred genotypes of MS—l, M0—10, and their progeny, as well as the position each of the 5 variant poly— peptides takes in the parental and progeny isolates are given in Table 1. In each of the progeny, peptide spots 1, 3, and 4 appeared in either the position characteristic of one parent or the other, but never both. The positions of peptides 3 and 4 could not be confidently determined for progeny 3 and 6, respectively, even though the entire procedure was performed at least 5 times for each culture. None of the 5 peptide spot positions could be correlated with a particular E gene. Instead, independent assortment of the genes responsible for the infection types and the spot position phenotypes was found. 45 Table l. The Segregation of Three Genes Determining Pathogenicity in the Progeny of MS—l x MO—lO Compared to the Gel Position Phenotypes of Five Polypeptides that Differentiate MS—l and MO—lO. Gel Position Taken By** Mildew Peptide Peptide Peptide Peptide Peptide Isolate Genotype* 1 2 3 4 5 MS—l E2 E3 E4 MS—l Absent MS—l MS—l MS—l MO—lO 32 33 34 MO—lO MO—lO MO—lO MO-lO Absent Prog. 1 E2 E? E4 MO—lO Absent MS—l MS-l Absent Prog. 2 E2 33 E4 MO—lO MO—lO MO—lO MS—l MS—l Prog. 3 E? 33 34 M0—10 MO—lO ? MO-lO Absent Prog. 4 32 33 34 MO-lO MO—lO MO—IO MS—l MS—l Prog. 5 32 33 34 M0—10 MO—lO MO—lO MS—l MS—l Prog. 6 32 33 E4 MO—lO MO—lO M0-10 ? Absent Prog. 7 32 E3 34 MO—lO M0—10 MO—lO MS-l MS-l Prog. 8 32 33 34 MS—l MO—lO MO—lO MO—lO Absent *Genotypes are inferred from the reactions of each isolate on three differential host isolines: E32a, E33a, and E34a. For example, an isolate avirulent on the E32a isoline is inferred to carry the corresponding E2 gene. An isolate virulent on E32a is inferred to carry the 32 allele. **Peptides l, 3, and 4 were always found in a position corresponding to that of one of the parents, MS—l or MO—IO, but never in both positions. Peptide 2 was either present, as in MO—IO, or absent, as in MS—l. Peptide 5 was either present, as in MS-l, or absent, as in MO—lO. 46 Discussion Two—dimensional electrophoresis can theoretically resolve over 7,000 polypeptides on a single gel (7). In practice, however, comparative analyses of protein spots on crowded gels is much more limited. In certain areas of the gels examined in this study, the number of polypeptides was so great that the resolving power of the gel system was beyond its limit. When silver was used to stain a gel that had already been stained with Coomassie blue, about two times more spots were revealed than were originally observed (compare a to be and c to d in Fig. 7). The silver stain often revealed spot clusters which had the appearance of single spots when stained with Coomassie blue alone. The presence or absence of a spot within such a cluster was often difficult to determine. In areas of overcrowding, it is possible that peptides of the same or very similar isoelectric point and molecular weight were mistaken for poly— peptides which were actually missing. Even for the five polypeptide differences detected between MS—l and MO—lO, replications were essential for interpretation. The polypeptide maps obtained from a wild type isolate of E. lindemuthianum and its mutant showed that a mutant induced with NTG can differ in many polypeptides. The great variation suggests that the mutation affected a basic part of the protein synthesis apparatus. For example, a mutation in a gene affecting the specificity of an amino—acyl t—RNA synthetase would be expected to affect many proteins. Because the sexual stage of this fungus is unknown, it was not possible to readily determine whether the mutation(s) induced with NTG affected one gene, or many. 47 The few differences observed in polypeptide maps for different cultures of E, graminis f. sp. tritici was surprising. Mutagenesis with NTG was expected to affect many genes. Yet when 3 mutants to increased virulence, obtained by NTG treatment, were compared to the wild type, no differences were detected amongst approximately 600 polypeptides. Furthermore, only 5 polypeptide differences were observed between 2 different field isolates, MS—l and MO—lO, MS-l was isolated from wheat in East Lansing, Michigan in 1961 and has been maintained on Little Club wheat in a constant environment chamber for 19 years. MO—lO was isolated from cleistothecia on wheat carrying the E34 gene in Monroe County, Michigan in 1977. That cultures of such diverse history have so little electrophoretic variation is interesting. Field isolates of E, 3333 commonly exhibit electrophoretic variation in about 5% of their proteins (various personal communications). Similarly, conspecific plant populations have been found to exhibit electrophoretic variation in about 5% of the proteins observed (5). Is it possible that an obligate parasite such as E, graminis is so finely tuned to its environ— ment (a host plant) that almost any variation is lethal? The low level of variation found in E, graminis f. sp. tritici was advantageous because it eliminated the need for following the segregation patterns of a large number of polypeptides. Progeny from a cross between MS~1 and MO-lO showed segregation of only 5 polypeptide variants. E, graminis f. sp. tritici is haploid; therefore, the genes responsible for coding these proteins are present in one dose. When a difference in position of a peptide is found between two isolates, progeny of a cross between the two should exhibit one parental type or the other, but never both if the two forms are the product of each of two alleles of one locus. 48 This is what was found. No latent differences appeared as a result of segregation. The segregation pattern in the progeny of MS-l x MO—lO showed that the peptide differences between the parents were not the the products of the E_genes examined (E2a, E3a, or E4a). The 3 independently derived mutants to increased virulence were found to give polypeptide maps indistinguishable from that of the parent MS—l. The failure to find a polypeptide whose net charge was changed concommitantly with a mutation to increased virulence was not surprising because only about 600 polypeptides were observed; this probably is only a small portion of the total number of polypeptides present in the pathogen. The methods used revealed only the more abundant proteins. Furthermore, only approximately 30% of any mutational changes in nucleotides can be expected to lead to amino acid substitutions that give the peptide a net change in charge (10). Perhaps the three mutants examined did not contain a charge change in the peptides responsible for the changes in virulence. The extraction procedures outlined here give highly reproducible polypeptide patterns on 2—d gels. We have other results which suggest that the gel patterns can be quite different if other extraction procedures are used. The number of peptides visualized can be increased by expanding the pH gradient and thus the resolution of the 2—d gels. Since at least 10 E_loci are available for mutational analysis in E, graminis f. sp. tritici, the possibility of finding a product of a E gene with many mutants at each E_locus is not unrealistic. 49 Literature Cited 10. Ellingboe, A.H. 1981. Genetical aspects of active defense. Manuscript submitted for publication. Flor, H.H. 1971. The current status of the gene—for—gene concept. Ann. Rev. Phytopathol. 9: 275-95. Gabriel, D.W. and A.H. Ellingboe. 1981. High resolution two— dimensional electrophoresis of proteins from congenic wheat lines. Manuscript submitted for publication. Gabriel, D.W., N. Lisker, and A.H. Ellingboe. 1981. The induction and analysis of two classes of mutations affecting pathogenicity in an obligate parasite. Manuscript submitted for publication. Gottlieb, L.D. 1977. Electrophoretic evidence and plant systematics. Ann. Missouri Botanical Garden 64: 161—180. Merril, C.R., R.C. Switzer, and M.L. VanKeuren. 1979. Trace polypeptides in cellular extracts and human body fluids detected by two—dimensional electrophoresis and a highly sensitive silver stain. Proc. Nat. Acad. Sci. 76: 4335—9. O'Farrell, P.H. 1975. High resolution two—dimensional electrophoresis of proteins. J. Biol. Chem. 250: 4007-21. O'Farrell, P.Z., H.M. Goodman, and P.H. O'Farrell. 1977. High resolution two—dimensional electrophoresis of basic as well as acidic proteins. Cell 12: 1133—42. Powers, H.R., Jr. and W.J. Sando. 1970 Genetic control of the host—parasite relationship in wheat powdery mildew. Phytopathology 50: 454—7. Shaw, C.R. 1970. How many genes evolve? Biochem. Genetics 4: 275—83. CHAPTER III HIGH RESOLUTION TWO—DIMENSIONAL ELECTROPHORESIS OF PROTEINS FROM CONGENIC WHEAT LINES 50 CHAPTER III HIGH RESOLUTION TWO—DIMENSIONAL ELECTROPHORESIS OF PROTEINS FROM CONGENIC WHEAT LINES Abstract Two methods were developed for the extraction of proteins from green wheat leaves. The methods result in distinct, highly reproducible patterns of polypeptides when separated by two—dimensional electrophoresis. Problems associated with phenolic compounds and proteinases were minimized by these methods, and over 300 polypeptides were observed when visualized with Coomassie blue stain. Comparisons of 12 different congenic host lines, each differing by a single gene for resistance to Erysiphe graminis f. sp. tritici (E3 isolines), failed to reveal any differences in poly— peptide mobility which could be attributed to, or associated with, a gene for resistance. Some seedlings of Chancellor wheat, the common background parent cultivar from which all of the E3 lines were derived, exhibited a unique polypeptide variant. The variant polypeptide was not found in all seedlings of cv. Chancellor, nor was it found in any of the E3 lines tested. The variant polypeptide appears to reflect a change in molecular weight from about 63,000 to about 68,000, with a very slight change in charge (pI of about 4). The variant polypeptide is among the 300 most visible polypeptides in seedling extracts that were subjected to 51 electrophoresis and stained with silver. Thus, the variant polypeptide appears to be made in a large number of copies. Chancellor appears to be a mixture of the original type and a variant type which appeared subsequent to the breeding of the E3 lines. 52 Introduction There are numerous reports in the literature of changes in proteins in a plant after infection with a pathogen (10,30). There are also numerous reports of protein differences between host lines that are resistant or susceptible after infection by a given pathogen (5,26). Attempts have been made to associate these protein changes with resistance and susceptibility. In most instances the procedures have been to examine differences in proteins (usually about one to three dozen proteins), in the relative enzymatic activities of particular proteins, and in the number and abundance of isozymic forms of particular enzymes. These procedures have allowed the examination of only a relatively small number of individual proteins, and no conclusive results are evident. High resolution two—dimensional (2—d) electrophoresis is a method which separates proteins in one dimension on the basis of charge and in the second dimension by size. It offers three distinct improvements over earlier separation methods. First, it can routinely resolve over 1,100 polypeptides on a single gel (22). Second, it separates denatured, individual gene products, allowing the identification of polypeptide products of genetic variants in the absence of a specific enzymatic or immunoreactive assay. Third, it separates by two different parameters with very high resolutions yielding a combined resolution better than 0.1 charge units and 1,000 daltons for an average 50,000 dalton protein (2). The technique has been used to separate a large number of polypeptides in bacteria and animal cells. In some cases it has been possible to identify, with a high degree of certainty, the products of a given gene (20,24). The principal procedure has been to Show a change in position of a peptide in the isoelectric focusing dimension concommitant with a mutation of 53 a particular gene. The genetics of resistance in wheat to pathogens has been studied extensively. At least eight loci (the E3 genes) are involved in resistance to Erysi3he graminis f. sp. tritici. Our approach was to compare proteins extracted from noninoculated, congenic host lines with or without individual dominant E3 genes. There are four assumptions to these experiments. One is that the E3 loci are transcribed and translated in noninoculated plants. Second is that the products of the E3_loci can be extracted. Third is that the products are transcribed and translated in sufficient copy number to be detected by the procedures used. Fourth is that the alternate alleles yield products that differ in net charge or size. Our first attempts to apply 2—d electrophoresis to plant extracts were disappointing. Extracts from plant cells tended to give few or no discreet spots with Coomassie blue staining whereas control extracts from fungal cells gave large numbers of discreet, non-diffuse spots. The problem with extracts from plant cells was in the preparation of the extracts. The procedures reported herein for the extraction and processing of total proteins from wheat leaves have made it possible to visualize several hundred polypeptides with both Coomassie blue and silver staining. 54 Materials and Methods Wheat seedlings. All wheat lines used were congenic. Each contained a known, homozygous E3 gene for resistance to Erysiphe graminis f. sp. tritici. The E3 lines were obtained by 8 backcrosses to the recurrent parent Chancellor (4). (Seed of all E3 lines, Chancellor, and derivatives were provided by Dr. John G. Moseman, USDA, Beltsville, MD 20705). Cultivar Chancellor has no known genes for resistance to powdery mildew. A host line containing, for example, the homozygous E3la gene will be referred to as the E3la isoline. The twelve isolines used for comparisons to cv. Chancellor and to each other were: E3la, E31b, E3lc, E3ld, E32a, E32b, E33a, E33b, E33a, E34a, E34b, and E36. All wheat seedlings were grown under sterile conditions. Seeds were surface sterilized in 10% Chlorox containing 0.1% SDS for 10 minutes, then rinsed 3 times in sterile distilled water before sowing on the surface of a sterile 1% agar medium. The agar contained, per liter: 3.3g NH NO , 4 3 3.8 g KNO 0.9 g CaClz-ZHZO, 0.6 g MgC12‘6H20, 0.3 g KHZPO 0.1 g NaZEDTA, 3’ 4, and 0.06 g FeSOa'7H20. Sterile hurricane lamp chimneys (Corning Glass Works # 845310) plugged at the top with cheesecloth and cotton, were placed over the surface—sterilized seeds planted in sterile agar. Plants were grown in the laboratory under Gro—Lux lights on a 14 hour photoperiod. Temperature under the lamps varied from 25 C with lights on to 21 C with lights off. After 7 days the seedlings were either extracted or aged in a refrigerator (in the dark at 4 C) for two weeks before extraction. Chemicals and Abbreviations: Tris, Tris (hydroxymethyl) aminomethane; SDS, Sodium Dodecyl Sulfate; DTT, (-)1,4—Dithio—L—threitol; EDTA, (Ethylenedinitrilo)~tetraacetic acid; NP—40, Nonidet P—40 from Particle Data Laboratories, Elmhurst, IL; ultrapure urea from Schwarz/Mann, Orangeburg, N.Y.; Ampholytes from LKB Instruments, Rockville, MD; PVP, Polyvinyl—polypyrrolidone, Insoluble. 55 Seedlings grown on agar were tested for disease reaction by inoculation with E, graminis f. sp. tritici, culture MS-l. Both seven-day-old seedlings and seedlings held in the refrigerator for two weeks as described were inoculated. Inoculated plants were held under Gro-Lux lights for one week. The infection type was determined for each E3_ isoline and Chancellor seven days after inoculation. Protein extraction. Two methods were developed for the extraction of proteins from plants. Both methods utilized 3 inch portions of two freshly cut leaf blades. The first inch of the leaf tip and the leaf sheath were excluded from extraction. The blades were cut into ca. half-inch long pieces and thorougly ground (for about 1 minute) in mortars containing acid-washed sea sand and 10 m1 of ice cold, degassed Water Extraction Buffer (1 mM Tris-HCI, pH 7.8; 10 mM DTT; 2 mM MgC12; 1 mM EDTA; 2% PVP). Vigorous grinding was avoided in an effort to keep foaming to a minimum. The solution was allowed to settle briefly (less than 1 minute) and was then decanted and centrifuged at 120,000 X g for one hour at 4 C. The supernatant was assayed for protein content (3) and a volume of supernatant containing approximately 400 ug of protein was precipitated in 9 volumes of ice cold acetone. The pellet was dried under purified N2 and resuspended with sonication in 100 ul of SDS Extraction Buffer (0.05 M Tris-HCl, pH 6.8; 10 mM DTT; 2% SDS; 0.5 mM MgC12; 1 mM EDTA). Solid urea was added to a 9M concentration, 200 ul of Sample Dilution Buffer (9.5 M ultra pure urea; 2% ampholytes; 10 mM DTT; 1 mM EDTA; 8% (w/v) NP—40) was added, and the solution vortexed until clear. It was important to keep the ratio of the final percentages of NP—40 to SDS in excess of 8, to avoid excessive streaking on the gels (1). 56 In the second method of extraction, leaf sections were ground as before, but SDS Extraction Buffer was used in place of the Water Extraction Buffer. After grinding, the solution was held with stirring at 100 C in a boiling water bath for four minutes. The samples were then chilled on ice, decanted into centrifuge tubes, and centrifuged at 120,000 X g for 30 minutes at 4 C. Samples were then precipitated and resuspended as described for the first method. It was crucial to keep the time required for all steps of the extraction to a minimum to avoid artifacts and a loss of resolution (refer to discussion). Electrophoresis and stain33g. Samples containing approximately 100 ug of protein were loaded without delay onto prefocused isoelectric focusing (IEF) gels prepared and run essentially as described by O'Farrell (22). Fractionation of proteins in the first dimension was accomplished by the selection of ampholytes (which set up the pH gradient) in the desired pH range. In these experiments the pH gradient ran from approximately 3.7 to 7.2, a range which was found to reveal the most proteins from the extracts. This pH range was achieved using 1.6% pH 5—7 ampholytes and 0.4% pH 3.5-10 ampholytes in both the IEF gels and the Sample Dilution Buffer. The second dimension gels were 11.25% acrylamide, and were run immediately following unloading from the IEF gels and equilibration in O'Farrell's Buffer ”O". In our experience, neither the samples nor the rod gels could be stored by freezing for any length of time without significant loss of resolution. After electrophoresis, gels were stained with Coomassie blue (0.125% in 50% methanol/ 12% acetic acid) overnight and then destained in 2 changes of 10% ethanol/ 5% acetic acid. Gels which were judged to be slightly underloaded and of high resolution were then stained with silver (19). Gels were photographed with transmitted 57 light after staining with Coomassie blue and/ or silver. Photographs were taken without a filter using transmitted light and Kodak Technical Pan 2415 film, developed for maximum contrast. These experiments were performed at least three times on all 12 E3 lines, and over ten times on cv. Chancellor, the E31a isoline, and the E34a isoline. 58 Results The reactions of plants grown on agar to infection by E, graminis f. sp. tritici culture MS—l were the same as those plants of the same genotype grown on soil. Each E3 isoline, when grown on agar, expressed the final infection type expected of that E3 isoline when grown on soil. The disease reactions were also independent of aging for two weeks in the dark at 4 C. Over 300 protein spots were visible following staining with Coomassie blue when freshly cut,seven—day—old wheat seedlings were extracted with the Water Extraction Buffer and electrophoresed as described. This buffer contains no detergents and extracts only water—soluble proteins. Few problems with streaking were encountered using this buffer, which was developed to extract the largest possible number of polypeptides from wheat seedlings (Fig. 8). Fewer polypeptide spots were obtained when an equal amount of protein was extracted from older plants. No differences were found in the polypeptide maps of 12 different E3 isolines and cv. Chancellor using this buffer. A buffer containing the detergent SDS was used in an attempt to extract membrane—bound proteins. The polypeptide maps obtained with the SDS Extraction Buffer (Fig. 9) were different from those obtained with the Water Extraction Buffer. Extracts made with the SDS buffer had a tendency to streak on the gels, and only about half as many polypeptides were visible as were seen when the Water Extraction Buffer and Coomassie blue stain were used. The use of silver stain increased the number of visible spots to about 300. The polypeptide maps of all 12 E3_isolines appeared to be identical. A polypeptide difference was, however, detected between the fully susceptible cultivar Chancellor and all of the other E3 59 Figure 8. A two—dimensional gel of denatured proteins extracted from Triticum aestivum cv. Chancellor. Proteins were extracted using the Water Extraction Buffer as described and stained with Coomassie blue. 60 Figure 9. Two—dimensional gels of denatured proteins extracted from Triticum aestivum. Proteins were extracted using the SDS Extraction Buffer as described and stained with silver. The abscissa scale represents typical values of a pH gradient obtained during isoelectric focusing under the experimental conditons described. The ordinate scale represents approximate molecular weights obtained using molecular weight (Mr) standards (not shown). (3) cv. Chancellor. (b) the E31a isoline. (c) a combined sample of cv. Chancellor + the E3la isoline. 61 H MIOH N 2 --155 l4. 6 F _ “'54 -—48 --32 --16 Figure 9a, I: x a: 5 A. A. 8 2 6 l 6 5 4 3 1 L _ _ _ . _ a _ _ _ _ a 62 pH Figure 9b. 63 "155 Figure 9c. Mr x 10-3 64 lines tested (Fig. 9, a & b). The variant polypeptide appears to reflect a change in molecular weight from approximately 63,000 in the E3 lines to about 68,000 in cv. Chancellor. A 1:1 mixture of an extract of cv. Chancellor and one of the E3 lines revealed little or no difference in the pI of the variant and the "normal" polypeptide (Fig. 9c). Another seed lot of cv. Chancellor was tested and the 63,000 dalton polypeptide spot was present, as in the E3 lines. Chancellor wheat therefore appears to be comprised of at least two distinguishable types. It was not determined whether the two variant types are mixed within, or only between, seed lots. 65 Discussion The E3 genes are known to follow the gene-for~gene pattern of interaction with E, graminis f. sp. tritici (25). Extensive genetic analyses of gene—for—gene resistance in a large number of hosts have demonstrated that single genes can be responsible for the resistant phenotype. There are no clear examples in gene—for—gene systems where more than one host gene is required for a specific incompatibility response. This strongly implies that resistance genes are structural genes coding for polypeptides directly responsible for specificity (9). If small molecules were directly involved in gene—for—gene specificity, genetic analyses should have revealed enzymatic pathways, and more genes than one would occasionally be necessary for a given specificity response. The genetic data are consistent with the hypothesis that incompatibility between host and parasite is the result of a protein—protein or a protein—nucleic acid interaction. For example, the E36 gene in wheat interacts with the E6 gene in Puccinia graminis to give incompatibility in the interaction at 20 C. Yet at 26 C, the interaction is compatible (17). This observation is consistent with the concept that the E36 gene product, the E6 gene product, or the interaction product is a temperature— sensitive protein. A loss in specificity and function with a 6 C rise in temperature is a well—documented characteristic of some proteins which is related to their tertiary, folded structure. Small molecules can exhibit the requisite capacity for variation (eg., mixed glycans), but their synthesis requires more than one gene———a fact which is inconsistent with the genetic data on specificity———and they are not sensitive to a 6 C rise in temperature. 66 An hypothesis consistent with all the data is that an_E gene codes for a protein that is the physiologically active host molecule involved in gene—for—gene specificity. Our approach to finding this protein was to compare polypeptides in noninoculated plants. It is possible that E3 genes are not constitutively expressed in noninoculated plants. Information bearing on E gene expression is available on only one E gene: the E32 or E3_gene. The E3 gene, which confers resistance in oats to Puccinia coronata, also confers susceptiblity to Helminthosporium victoriae and to its toxin (16). All attempts to separate susceptibility to E, victoriae toxin from resistance to E, coronata have failed. Thus, these two functions (resistance to E. coronata and susceptibility to E, victoriae) are apparently dependent on the single dominant E3 gene (31). The specific receptor of E, victoriae toxin appears to be the protein product of the E3_gene (14). Furthermore, the E, victoriae toxin affects the plasma membrane of oats with the E3_gene almost instantly, whereas plasma membranes of oats with the 33 allele are not affected (14,27). The speed of the reaction to toxin makes it unlikely that derepression, transcription and/or translation of the E3_gene has occurred after toxin application, and is evidence for constitutive E3_gene expression. Since the E3 gene is a specific_E gene against E. coronata and since its protein product appears to be constitutive, we assumed by analogy that the protein products of the E3 genes constitutively expressed. The separation of a large number of proteins from fresh plant leaves by 2—d electrophoresis is more limited by the extraction technique than by the capacity of the gel system. Unlike protein extraction from bacteria, fungi, or cells in culture, the extraction of proteins from whole leaves is fraught with difficulties arising from the breakdown of compartmentation. 67 Whenever plant protoplasmic and vacuolar contents are mixed, a variety of compounds are released and oxidized; oxidized phenols are known to react with proteins (18). In addition, the breakdown of compartmentation is thought to produce changes in the conformation of proteins that render otherwise protected peptide bonds susceptible to proteolytic hydrolysis (28). High levels of proteolytic activity are found in cereal seedlings only a few days old (23). Extraction procedures typically utilize low tempera- tures, denaturing agents, competitive adsorbents, etc., to obviate these difficulties (18). In our hands these methods proved inadequate for extraction of proteins from fresh wheat leaves, and we had to empirically develop two methods which gave preparations suitable for analysis by 2—d electrophoresis. The method which gave the greatest number of polypeptides was the one using Water Extraction Buffer and seven—day—old seedlings. Care was taken to insure that seedlings used for extraction would be free of contaminating microoganisms. Equally important was the demonstration that expression of specific resistance in wheat was unaffected by the growth conditions used. Despite these precautions, no differences could be observed between gels from extracts of each of the Em isolines. Seedlings which were aged for two weeks in the dark at 4 C yielded fewer polypeptides on electrophoresis than younger seedlings, though the same amount of protein was loaded onto the gels. Since resistance was still expressed in the older plants, this reduction in polypeptide number was used in later experiments to eliminate a number of proteins apparently superfluous to the search for a resistance gene product. The Water Extraction Buffer contained no detergents and therefore membrane—bound proteins are not extracted with this buffer. The possibility 68 that an_E gene product might be membrane—bound cannot be overlooked. The earliest detectable time for the expression of resistance in powdery mildew on wheat and barley is after the parasite has made contact with the host plasma membrane (8). A major class of incompatibility proteins in mammals—--the histocompatibility proteins—--are known to be membrane- bound (15). Therefore a buffer containing SDS, which can solublize membrane-bound proteins (1,15), was developed and utilized. Fewer peptides were revealed using this buffer than when using the Water Extraction Buffer because of streaking in the neutral pH range in the first dimension (Fig. 9). Dodecyl sulfate will dissociate proteins from other molecules, such as chlorophyl, but only incompletely. The extent of the displacement of these molecules depends on the amount of dodecyl sulfate and the conditions used in the extraction (29). We suspect that random and incomplete removal of variously charged, low molecular weight molecules caused the streaking. A further problem with the use of SDS is that at least some proteinases are highly active in its presence, and proteinase activity may even be enhanced by the denaturing effect SDS has on most polypeptides (13). We had reported finding differences between the Em lines in preliminary experiments utilizing the SDS Extraction Buffer (11). The differences were the faint, but reproducible appearances of acidic, low molecular weight polypeptides on certain gels. However, the differences were not dependent on Em genes. Instead they were found to be artifacts apparently dependent on the amount of time taken to extract the leaves. Large molecular weight proteins tend to be degraded more rapidly than small ones (7), and acidic proteins tend to be more rapidly degraded than basic ones (6). For example, the half—life of ornithine decarboxylase 69 is 12 minutes (7). In our earlier experimental protocol (12), we prepared extracts from several Em lines in sequence, and extracts of the first Em line could remain on ice in the SDS buffer for up to an hour before the last Em_line was extracted. Polypeptide fragments resulting from specific proteolytic clips of acidic, large molecular weight proteins may have been responsible for the appearance of the acidic, low molecular weight polypeptide artifacts. When fresh leaves were extracted and precautions were taken to keep the maximum elapsed time between the extraction of the first and last samples to 15 minutes, no differences in polypeptide maps between the Em lines were detected. Detection of artifacts resulting from proteolytic degradation of proteins becomes more likely as the separation system becomes more sensitive. A polypeptide difference was found between cv. Chancellor and all other Em lines tested. The result was surprising, since we were searching for a polypeptide unique to one of the Em_isolines. The variant poly— peptide in cv. Chancellor appears to reflect a change in molecular weight from 63,000 to 68,000, with a very slight change in charge. The change in molecular weight may not necessarily involve the protein's structural gene, since achange of this magnitude would probably be accompanied by a sizable change in charge. A shift in molecular weight without a change in charge could result from a gain of a polysaccharide comprised of about 25 sugars. An enzyme involved in glycoprotein synthesis or modification may be involved. The polypeptide variant is not found in all Chancellor seedlings. A mutation or an accidental outcross may have occurred in cv. Chancellor subsequent to the breeding of the Em lines. Even though electrophoretic variation was not detected in these experiments in any of the Em lines, 70 such variation may have been missed by chance. Variation has been observed within the Em lines and within cv. Chancellor in the field (21). Whether the electrophoretic variant can be correlated with one of the field phenotypes remains to be examined. The possibility that much variation exists within congenic lines, coupled with the knowledge that a great deal of variation due to linkage exists between the lines, makes interpretations of physiological differences more difficult. Reliance on genetic purity in congenic lines is not warranted. Comparisons in this study were of congenic wheat lines differing by single Em genes and perhaps other genes closely linked to these. We had hoped to detect a polypeptide difference that was at least linked to a resistance gene in at least one of the Em lines examined. The two extraction buffers gave very different polypeptide maps of the Em lines. The use of a large number of different Em lines and different extraction techniques should allow the identification of the product of a gene at least linked to a Em gene. This would provide a handle on the segment of DNA containing the Em gene which might be exploited using recombinant DNA techniques. In the absence of a direct assay for an.E gene product, such a handle may be valuable. It seems unlikely that a mechanism for resistance will be elucidated until the product(s) responsible are isolated and purified. 71 Literature Cited 10. ll. 12. l3. l4. Ames, G.F. and K. Nikaido. 1976. Two—dimensional electrophoresis of membrane proteins. Biochemistry 15: 616—23. Anderson, L. and N.G. Anderson. 1977. High resolution two— dimensional electrophoresis of human plasma proteins. Proc. Nat. Acad. Sci. 74: 5421—5. Bradford, M.M. 1976. A rapid and sensitive method for the quantita— tion of microgram quantities of protein utilizing the principle of protein—dye binding. Anal. Biochem. 72: 248—54. Briggle, L.W. 1969. Near—isogenic lines of wheat with genes for resistance to Erysiphe graminis f. sp. tritici. Crop Sci. 9: 70—5. Daly, J.M. 1976. Some aspects of host-pathogen interactions. Em Physiological Plant Pathology. R. Heitefuss and P.H. Williams, eds. Springer—Verlag, N.Y. 890 pp. Dice, J.F. and A.L. Goldberg. 1975. Relationship between Em_vivo degradation rates and isoelectric points of proteins. Proc. Nat. Acad. Sci. 72: 3893—7. Dice, J.F., P.J. Dahlinger, and R.T. Schimke. 1973. Studies on the correlation between size and relative degradation rate of soluble proteins. J. Biol. Chem. 248: 4220—8. Ellingboe, A.H. 1972. Genetics and physiology of primary infection by Erysiphe graminis. Phytopathology 62: 401—6. Ellingboe, A.H. 1981. Genetical aspects of active defense. Manuscript submitted for publication. Fric, F. 1976. Oxidative enzymes. Em Physiological Plant Pathology. R. Heitefuss and P.H. Williams, eds. Springer-Verlag, N.Y. 890 pp. Gabriel, D.W. and A.H. Ellingboe. 1978. The primary product of an E gene? Phytopathology News 12: 139. (abstr.) Gabriel, D.W. and A.H. Ellingboe. 1978. Extraction of plant leaf and fungal protein by two—dimensional electrophoresis. Phyto- pathology News 12: 138. (abstr.) Gaertner, F.W. and K.W. Cole. 1976. The protease problem in Neurospora: structural modification of the arom multienzyme system during its extraction and isolation. Archives of Biochem. and Biophysics 177: 566—73. Gardner, J.M. and R.P. Scheffer. 1973. Effects of cyclohexamide and sulfhydryl~binding compounds on sensitivity of oat tissues to Helminthosporium victoriae toxin. Physiological Plant Pathology 3: 147—57. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 72 Klein, Jan. 1975. Biology of the Mouse Histocompatibility—Z Complex. Springer-Verlag, N.Y. 620 pp. Litzenberger, S.C. 1949. Nature of susceptibility to Helminthosporium victoriae and resistance to Puccinia coronata in Victoria oats. Phytopathology 39: 300-18. Loegering, W.Q. and J.R. Geiss. 1957. Independence in the action of three genes conditioning stem rust resistance in Red Egyptian wheat. Phytopathology 47: 740-1. Loomis, W.D. 1974. Overcoming problems of phenolics and quinones in the isolation of plant enzymes and organelles. Methods in Enzymology 31: 528-44. Merrill, C.R., R.C. Switzer, and M.L. VanKeuren. 1979. Trace poly- peptides in cellular extracts and human body fluids detected by two-dimensional electrophoresis and a highly sensitive silver stain. Proc. Nat. Acad. Sci. 76: 4335—9. Milman, G., E. Lee, G.S. Ghangas, J.R. McLaughlin, and M. George. 1976. Analysis of HeLa cell hypoxanthine phosphoribosyl transferase mutants and revertants by two-dimensional polyacrylamide gel electrophoresis: evidence for silent gene activation. Proc. Nat. Acad. Sci. 73: 4589-93. Moseman, J.G., personal communication. O'Farrell, P.H. 1975. High resolution two—dimensional electrophoresis of proteins. J. Biol. Chem. 250: 4007—21. Peoples, M.B. and M.J. Dalling. 1978. Degradation of ribulose—l,5— bisphosphate carboxylase by proteolytic enzymes from crude extracts of wheat leaves. Planta 138: 153-60. Piperno, G., B. Huang, and D.J.L. Luck. 1977. Two—dimensional analysis of flagellar proteins from wild type and paralyzed mutants of Chlamydomonas reinhardtii. Proc. Nat. Acad. Sci. 74: 1600-4. Powers, H.R., Jr. and W.J. Sando. 1960. Genetic control of the host- parasite relationship in wheat powdery mildew. Phytopathology 50: 454-7. Sako, N. and M.A. Stahman. 1972. Multiple molecular forms of enzymes in barley leaves infected with Erysiphe graminis f. sp. hordei. Physiological Plant Pathology 2: 217—26. Samaddar, K.R. and R.P. Scheffer. 1971. Early effects of Helminthosporium victoriae toxin on plasma membranes and counter— action by chemical treatments. Physiological Plant Pathology 1: 319-28. 28. 29. 30. 31. 73 Schimke, R.T. 1975. On the properties and mechanism of protein turnover. Em_lntracellular Protein Turnover. R.T. Schimke and N. Katanuma, eds. Academic Press, N.Y. pp173, 186. Schmidt, G.W., S. Bartlett, A.R. Grossman, A.R. Cashmere, and N. Chua. 1979. lm vitro synthesis, transport, and assembly of the constituent polypeptides of the light—harvesting chlorophyll a/b protein complex. Em Genome Organization and Expression in Plants. C.J. Leaver, ed. Plenum Press, N.Y. pp337—51. Uritani, I. 1976. Protein metabolism. Em Physiological Plant Pathology. R. Heitefuss and P.H. Williams, eds. Springer— Verlag, N.Y. pp509—521. Wallace, A.T., R.M. Singh, and R.M. Browning. 1967. Mutation response and spectrum of mutations at the YE locus in Avena byzantina C. Koch. Em_lnduced Mutations and Their Utilization. Erwin—Baur—Gidachtnisvorlesungen IV, Akademie—Verlog, Berlin. pp47—57. 1114}.§éf. t f. .r Art. .-