MYOOVIRUSES IN ISOLATES OF GAEUMANNOMYCES AND
PHIALOPHORA SPECIES
BY
Rose Mary McGinty
Thesis submitted to the University of London for the degree of Doctor of Philosophy
Biochemistry Department, Plant Pathology Department, Imperial College of Rothamsted Experimental Station, Science and Technology, Harpenden, London, Hert fordsh ire, SW7 2AZ. AL5 2JQ.
1981 To my Mother 3 MYCOVIRUSES IN ISOLATES OF GAEUMANNOMYCES AND PHIALOPHORA SPECIES
By
Rose Mary McGinty
ABSTRACT
Characterisation of virus particles from the avirulent or weakly pathogenic parasites of cereal roots, Phialophora graminicola (Pg), Phialophora species with lobed hyphopodia (P.sp.(lh)) and Gaeumannomyces graminis var. graminis (Ggg) is described. These fungi can cross protect cereal roots from damage caused by Gaeumannomyces graminis var. tritici (Ggt), the causal agent of "take-all" disease of wheat and barley. Total nucleic acid was extracted frcm the mycelium of each of 12 field isolates of the weakly pathogenic fungi. Extracts were treated with DNAse and RNAse under appropriate conditions to remove DNA and single-stranded RNA respectively, and the remaining dsRNA was then analysed by polyacrylamide gel electrophoresis. Eight isolates were found to contain dsRNA, with components varying from two to five in number for different isolates. Molecular weights of the dsRNA components ranged from 1x10 to greater than 6 x 10 . Isometric virus particles were extracted and purified from three of the isolates found to contain dsKNA, one each from the species Pg, P.sp. (lh) and Ggg. Evidence was obtained for the presence of two sero- logically unrelated viruses in P.sp. (lh), both of which were 35 nm in diameter. Each virus had a capsid composed of one type of polypeptide (molecular weights 65,500 and 60,700 respectively) and each had three c dsRNA components with molecular weights in the range 1.03 x 10 c to 1.32 x 10°. A magnesium dependent RNA polymerase was found associated with the virus particles in P.sp.(lh) and a single stranded RNA product was identified. Nucleic acid hybridisation between the single stranded transcripts of one virus and the double stranded RNA genomes of each virus suggested that the two viruses have a dsRNA component in common. Single viruses were extracted from both Ggg and Pg. The former virus had a diameter of 35 nm and had a major capsid component with a molecular weight of 64,000 and four dsRNA components with molecular weights in the range 1.16 x 10^ to 1.63 x 106. The virus of Pg_ had a diameter of 30 nm and four dsRNA components with molecular weights in the range 1.14 x 106 to 1.32 x 106. 4
Interactions of the three isolates of P.sp.(lh), Pc[ and Ggg with seven isolates of Ggt on three different solid media were examined. No evidence for inhibition of Ggt by the three weakly pathogenic isolates was obtained; on the contrary all seven Ggt isolates were found to inhibit the growth of the weakly pathogenic isolates over the pH range 3.5 to 7.0, with optimal pH for this effect at pH 4.0 to 5.0. Inhibition at a distance was observed, although no killing of hyphae occurred. Culture filtrates were shown to contain a stable, diffusible but non- proteinaceous inhibitor. The significance of these results in relation to the biology of the Gaeumannomyces-Phialophora complex and prospects for the practical control of the "take-all" disease are discussed. 5
ACKNOWLEDGEMENTS
I would particularly like to thank Dr. K.W. Buck and Dr. C.J.
Rawlinson for their help, advice and support during the course of this
work.
I thank Professor B.S. Hartley and Mr. E. Lester for the use of the
facilities of the Biochemistry Department, Imperial College and the Plant
Pathology Department at Rothamsted Experimental Station respectively.
I thank Dr. C.J. Rawlinson and Dr. M. Almond for the generous gifts
of antisera to Phialophora and Gaeumannomyces viruses and Dr. R.F.
Bozarth for the gift of Helminthosporium maydis dsRNA.
I thank Dr. K.W. Buck for permission to copy Tables 1 and 2 (Buck,
1980) and Table 9 (Buck et al., 1981) and Dr. T.W. Young for permission
to copy Table 3 (Young and Yagui, 1978).
I thank Mr. I.P. Blench and Dr. G. Ratti for helpful advice and discussions, Mr. R. Woods for electron microscopy, Mr. G. Millhouse, Mr.
G. Higgins and Mr. F. Cowland for photography and Mrs. V. Souilah for her excellent typing.
The receipt of a Science Research Council studentship is gratefully
acknowledged. CONTENTS PAGE
TITLE PAGE 1
ABSTRACT 3
ACKNOWLEDGEMENTS 5
CONTENTS i - viii
INTRODUCTION 14
Discovery and occurrence of dsRNA 14 mycoviruses
Transmission of mycoviruses 14
Properties of dsRNA mycovirus particles 17 in vitro
Effect of dsRNA mycoviruses on their hosts 22
DsRNA and the killer phenomenon 25
Mycoviruses in the take-all fungus 34
The take all disease 34
Control of take-all disease 35
Cross-protection of cereals against take-all disease with weakly pathogenic or avirulent 38 root parasites : possible role of viruses
Aims and approach to the investigation 42 ii
PAGE
ABBREVIATIONS 44
MATERIALS 47
GENERAL METHODS 48
G.M.1. ORGANISMS AND MICROBIOLOGICAL METHODS 48
G.M.2. ISOLATION AND PURIFICATION OF VIRUS 51
a. Small scale preparation of crude virus 51
b. Large scale preparation of crude virus 51
c. Purification of large scale virus 52
preparations by sucrose density gradient
centrifugation
d. Purification of small scale virus 53
preparations
G.M.3. ELECTRON MICROSCOPY 53
G. M. 4. SPECTROPHOTOMETRY 54
G.M.5. PREPARATION OF ANTISERA 54
G.M.6. OUCHTERLONY GEL IMMUNODIFFUSION TEST 55
G.M.7. PREPARATION OF DEAE CELLULOSE 55
G.M.8. PREPARATION OF CF11 CELLULOSE 56
G.M.9. PREPARATION OF SP C-50 SEPHADEX 56 iii
PAGE
G.M.10. ANALYTICAL ULTRACENTRIFQGATION 56
a. Determination of sedimentation coefficient 56 of virus
b. Determination of buoyant density of virus 57
G.M. 11. DETERMINATION OF DENSITY OF CAESIUM 63 CHLORIDE SOLUTIONS
G.M.12. NUCLEIC ACID EXTRACTION 64
a. Preparation of total nucleic acid from 64 mycelia
b. Preparation of virus RNA 64
G.M.13. ACTION OF NUCLEASES ON TOTAL FUNGAL 65 NUCLEIC ACID AND VIRAL RNA
a. Action of DNAse I 65
b. Action of RNAse A 65
G.M. 14. PREPARATION OF EDTA - WASHED PHENOL 65
G.M.15. PREPARATION OF DIALYSIS TUBING 66
G.M. 16. ANALYTICAL GEL ELECTROPHORESIS 66 a. Agarose gel electrophoresis for analysis of 66 intact virus
b. Agarose gel electrophoresis for analysis of 67 nucleic acid
c. Polyacrylamide gel electrophoresis for analysis 67 and molecular weight determination of nucleic acid
i. For dsRNA analysis 67
ii. For dsRNA and ssRNA analysis 68 iv
PAGE
d. Polyacrylamide gel electrophoresis for 69
analysis and molecular weight deter-
mination of polypeptide.
e. Combined analysis of intact virus and polypeptide 70
G.M. 17. GLYOXALATION OF DSRNA 71
G.M. 18. SNA-DEPENDENT FNA POLYMERASE ASSAY 71
G.M. 19. PREPARATION OF VIRUS SSRNA TRANSCRIPTS 72 IN VITRO
G.M. 20. RNA-RNA HYBRIDISATION ASSAY 72
G. M. 21. FLUOROGRAPHY 73
G.M. 22. SLICING AND SOLUBILISATION OF POLYACRYLAMIDE 74 GELS FOR RADIOACTIVE ANALYSIS
G.M.23. INCOMPATIBILITY TESTS OF FUNGAL ISOLATES 74
G.M. 24. PRODUCTION OF INHIBITOR IN LIQUID CULTURE 74
G.M.25 WELL TESTS FOR INHIBITORY ACTIVITY OF CULTURE 75 FILTRATES
G.M. 26. TESTS FOR SENSITIVITY OF INHIBITOR TO PROTEASES 75
SECTION 1 - Screening of field isolates of P.sp.(lh)f 77 Ggg and Pg_ for the presence of dsRNA
RESULTS 77
DISCUSSION 171 PAGE
SECTION 2 - Isolation and properties of virus 89 particles from P.sp.(lh)
RESULTS 89
A. Preparation of crude virus 89
B. Purification of virus by sucrose 90 density gradient centrifugation
a. Small scale purification of virus 90
b. Large scale zonal purification of virus 90
C. Properties of purified particles 93
a. U.V. spectrum 93
b. Particle morphology 93
c. Sedimentation coefficient 93
d. RNA components 99
e. Polypeptide components 102
D. Preliminary evidence that purified 107 virus preparations contain two distinct viruses in P.sp.(lh)2-2
a. Serology of P.sp.(lh)2-2 virus 107
b. Electrophoretic separation of the 107 viruses on agarose gels
c. Fractionation of RNA and polypeptide 110 components across a sucrose gradient.
E. Separation of the viruses of P.sp.(lh)2-2 112
F. Properties of each individual virus 117
a. Particle morphology 117
b. U.V. spectra 123 vi
PAGE
c. Sedimentation coefficients for virus A 123 and virus B
d. dsRNA components 123
e. Polypeptide components 128
f. Serology 128
g. Buoyant density 128
G. Virion associated RNA polymerase activity 131
a. Properties of the RNA polymerase reaction 131
b. Analysis of reaction products 136
c. Cross hybridisation experiments 144
H. Alterations of dsRNA migration patterns 145 of viruses A and B repeated fungal subculture
DISCUSSION 147
SECTION 3 - Isolation of virus particles from Ggt 159 isolate 45/10
RESULTS 159
A. Comparison of two media for fungal growth 159 prior to virus preparation by the PEG method
i. Basal and CSL medium 159
ii. Weste and Throwers medium 160
B. Comparison of four different buffers for 163 virus preparation by the PEG precipitation method and by direct pelleting
i. Direct pelleting method 165
ii. PEG precipitation method 165
DISCUSSION 171 VI1
PAGE
SECTION 4 - Isolation and properties of virus 174 particles from Ggg isolate G1
RESULTS 174
A. Preparation of crude virus 174
B. Purification of virus by sucrose 174 density gradient centrifugation
C. Purification of virus by caesium 176 chloride density gradient centrifugation
D. Properties of purified virus particles 176
a. U.V. spectrum 176
b. Particle morphology 179
c. Agarose gel electrophoresis 179
d. Serology 179
e. Sedimentation coefficient 179
f. Buoyant density 184
g. RNA components 184
h. Polypeptide components 184
DISCUSSION 190
Section 5 - Isolation and properties of virus 194 particles from Pg_ isolate 1348-2
RESULTS 194
A. Preparation of crude virus 194
B. Properties of Pg_ 1348-2 virus particles 196
a. Particle morphology 196
b. RNA components 198
c. Serology 198
DISCUSSION 203 viii
PAGE
Section 6 - Interactions between isolates 204
RESULTS 204
A. Growth characteristics of isolates 204 in control experiments
a. P.sp.(lh)2-2 204
b. Ggg G1 207
c. Pg 1348-2 207
d. Ggt OgA 207
e. Ggt 019/6 207
f. Ggt F3, 45/10, 38-4, 3b1a and T1 212
B. Growth characteristics of isolates 212 showing inhibitory activity on solid media
a. PDA medium 212
b. Glucose/asparagine (LB) medium 215
c. Glucose/urea (H) medium 221
C. Nature of the inhibition reactions 224
D. Diffusion of inhibitor in liquid culture 227
E. Effect of proteases on the inhibition 227 reaction
F. Stability of inhibitor in culture filtrates. 229
DISCUSSION 230
CONCLUSIONS 233
REFERENCES 237 6
TABLES
PAGE
1 Physical properties of isometric dsRNA mycoviruses 18 with undivided genomes
2 Physical properties of isometric dsRNA mycoviruses 19 with segmented genomes
3 Interaction between isolates 26
4 Table of field isolates of Phialophora and 78 Gaeumannomyces species
5 Molecular weights of dsRNA components extracted from 85 field isolates of P.sp.(lh),Ggg and Pg
6 Physicochemical characteristics of P.sp.(lh) 2-2 131 viruses q 7 Incorporation of [ H] into TCA-insoluble material in 134 RNA polymerase assays 8 Calculated coding capacity of virus A and B dsRNA 149 components
9 Physicochemical properties of G. graminis viruses 152
10 Molecular weights of the dsRNA and polypeptide 162 components of Ggt 45/10 viruses
11 RNA components extracted from Ggt 45/10 viruses, 167 prepared by the direct pelleting method
12 RNA components extracted from Ggt 45/10 viruses, 169 prepared by the PEG-precipitation method
13 Physicochemical characteristics of Ggg G1 virus 189
14 Calculated coding capacity of Ggg G1 virus dsRNA 192
15 Morphological appearance of isolates of P.sp.(lh), 206 Ggg, Pg, and Ggt on different media
16 pH range dependence of inhibition between isolates 225 cultured on LB and H media FIGURES
PAGE
1 Typical photoelectric scanner trace for the 58 determination of sedimentation coefficient
2 Schematic representation of a trace used for the 60 determination of buoyant density
3 Polyacrylamide gel electrophoresis of total nucleic 82 acid extracts from field isolates, after treatment with DNAse 1
4 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 83 RNA directly extracted from mycelium
5 Relationship between log molecular weight and 84 electrophoretic mobilities of RNA components of isolates (a) P.sp.(lh) 2-2, (b) P.sp.(lh) 74/1007-2, (c) Ggg G1 and (d) Pg 1348-2
6 Purification of P.sp.(lh) 2-2 virus by rate 91 centrifugation in a sucrose density gradient: ISCO u.v. scanner tracing
7 Purification of P.sp.(lh) 2-2 virus by rate 92 centrifugation in a zonal sucrose density gradient
8 U.v. spectrum of P.sp.(lh) 2-2 virus, following 94 purification on a sucrose gradient
9 Electron micrograph of P.sp.(lh) 2-2 virus 95
10 Histogram plots of VIP diameters measured randomly from 96 electron micrographs of P.sp.(lh) 2-2 virus by (a) direct measurement from negatives or (b) from prints
11 Electron micrograph of 'empty' P.sp.(lh) 2-2 virus 97
12 Photoelectric scanner tracing for the determination of 98 the sedimentation coefficient of P.sp.(lh) 2-2 virus
13 Plot of In x versus t, from which the sedimentation 100 coefficient of P.sp.(lh) 2-2 is obtained
14 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 101 virus RNA in crude and purified virus preparations 8
PAGE
15 Ethidium bromide-agarose plate analysis of P.sp.(lh) 103 2-2 isolate DNA and dsRNA-containing fractions, separated by CF11 chromatography
16 Polyacrylamide gel electrophoresis of P.sp.(lh)2-2 104 isolate DNA and dsRNA-containing fractions,separated by CF11 chromatography
17 Polyacrylamide gel electrophoresis of P.sp.(lh)2-2 105 virus polypeptides
18 Relationship between log molecular weight and 106 electrophoretic mobilities of P.sp.(lh)2-2 virus polypeptides in SDS/polyacrylamide gels
19 Polyacrylamide gel electrophoresis of P.sp.(lh)2-2 108 virus polypeptides: comigration of P2 with pyruvate kinase
20 Imnunodiffusion test of P.sp.(lh)2-2 virus with 109 antiserum raised against this virus
21 Agarose gel electrophoresis of P.sp.(lh)2-2 virus 109
22 Polyacrylamide gel electrophoresis of P.sp.(lh) virus m RNA (a) and polypeptides (b) from sucrose density gradient fractions following rate centrifugation
23 Ethidium bromide-agarose plate analysis of P.sp. (lh) 114 2-2 viruses A and B separated by DE52 cellulose ion exchange chromatography
24 .U.v.profile of P.sp.(lh)2-2 viruses A and B eluted 115 from a DE52 cellulose column
25 Electron micrographs of P.sp.(lh)2-2 viruses A (a) and 116 B (b)
26 Ethidium bromide-agarose plate analysis of P.sp.(lh) us 2-2 virus B purified on SP-Sephadex ion exchange chromatography
27 Agarose gel electrophoresis of P.sp.(lh)2-2 'empty' 119 viruses A and B
28 Electron micrograph of 'empty' virus A particles 120
29 Electron micrograph of 'empty' virus B particles 121 30 Histogram plots of VLP diameters measured randomly from electron micrographs of 'full' virus A (a) and virus B (b) and 'empty' virus A (c) and virus B (d) particles
31 U.v. spectra of P.sp.(lh) viruses A (a) and B (b) following purification by DE52 and SP-Sephadex ion exchange chromatography
32 Photoelectric scanner traces for the determination of sedimentation coefficients of P.sp.(lh)2-2 viruses A (a) and B (b)
33 Polyacrylamide gel electrophoresis of P.sp.(lh) virus A RNA (a) and virus B RNA (b)
34 Polyacrylamide gel profiles of RNA released from P.sp. (lh)2-2 viruses A (a) and B (b)
35 Polyacrylamide gel electrophoresis of P.sp.(lh)2-2 virus A polypeptide (a) and virus B polypeptide (b)
36 Imnunodiffusion test of P.sp.(lh)2-2 viruses A (VA) and B (VB) against antiserum raised against P.sp.(lh) 2-2 virus
37 Photoelectric scanner trace for the determination of buoyant density of P.sp.(lh)2-2 viruses A( a) and B( b)
38 Dependence of the RNA polymerase reactions of virus A (a) and virus B (b) on magnesium ion concentration
39 RNA synthesis catalysed by P.sp.(lh)2-2 viruses A and B RNA polymerases
40 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 viruses A (a) and B (b) dsENA and ssRNA products
41 Polyacrylamide gel analysis of RNA synthesised by Phialophora viruses A (a) and B (b) measured by incorporation of [%] UMP into ssRNA
42 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus A dsRNA and ssRNA product
43 Fluorograph of P.sp.(lh) 2-2 virus A dsRNA and ssRNA product following polyacrylamide gel electrophoresis 10
PAGE
44 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 142 virus A dsRNA and ssRNA product: comigration with glyoxalated dsRNA
45 Hybridisation of virus A RNA polymerase product to 143 (a) virus template dsRNA and (b) virus B dsRNA
Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 146 46 virus A RNA (a) and virus B RNA (b): alteration of migration pattern after repeated subculture
Polyacrylamide gel electrophoresis of Ggt 45/10 virus 161 47 RNA from a crude virus preparation: the isolate was grown on basal and CSL medium
Polyacrylamide gel electrophoresis of Ggt 45/10 virus 161 48 RNA from a crude virus preparation: the isolate was grown on Weste and Throwers medium
Polyacrylamide gel electrophoresis of Ggt 45/10 virus 161 49 polypeptides: isolate grown on basal and CSL (a) or Weste and Throwers (b), media
Investigation of methods of virus isolation from Ggt 164 50 isolate 45/10
51 Polyacrylamide gel electrophoresis of Ggt 45/10 virus 166 RNA obtained by the direct pelleting method (a) or PEG precipitation methods (b) and (c)
52 Polyacrylamide gel electrophoresis of Ggg G1 virus RNA 175 fron a crude virus preparation
53 Polyacrylamide gel electrophoresis of Ggg G1 virus RNA 175 (a) and polypeptides (b) in fractions 15 to 17, following sucrose density gradient centrifugation
54 Ethidium bromide-agarose plate analysis of Ggg G1 virus 177 in fractions from a caesium chloride density gradient
55 U.v. spectrum of Ggg G1 virus 178
56 Electron micrograph of Ggg G1 virus 180
57 Histogram plot of particle diameters measured randomly 181 fran electron micrographs of Ggg G1 virus
58 Agarose gel electrophoresis of Ggg G1 virus 182 59 Immunodiffusion test of Ggg G1 virus with antiserum raised against this virus
60 Photoelectric scanner trace for the determination of sedimentation coefficient of Ggg G1 virus
61 Photoelectric scanner trace for the determination of buoyant density of Ggg G1 virus 62 Polyacrylamide gel profile of RNA released from Ggg G1 virus
63 Relationship between log molecular weight and electrophoretic mobilities of Ggg G1 virus RNA components
64 Relationship between log molecular weight and electrophoretic mobilities of Ggg G1 virus polypeptide components
65 Polyacrylamide gel electrophoresis of Pg virus RNA: the isolate was grown in basal and CSL medium
66 Polyacrylamide gel electrophoresis of RNA extracted from Pg 1348-2 virus purified in four different buffers
67 Electron micrograph of Pg 1348-2 virus
Histogram plot of particle diameters measured randomly 68 from an electron micrograph of Pg 1348-2 virus
Polyacrylamide gel profile of RNA released from Pg 69 1348-2 virus
Relationship between log molecular weight and 70 electrophoretic mobility of Pg virus RNA
71 Rate of growth of P.sp.(lh) isolate 2-2 on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0
72 Rate of growth of Ggg isolate G1 on (a) PDA medium buffered to pH 3.0 and 6.0;(b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0
73 Rate of growth of Pg isolate 1348-2 on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0 74 Rate of growth of Ggt isolate OgA on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0
75 Rate of growth of Ggt isolate 019/6 on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0
76 Colony diameters of isolates Ggt OgA (•-•) > 019/6 (V-V) 45/10 (A-A),T1 00,F3 (#-*), 38-4 (OO) and 3bla (•-•) after 5 days growth on LB medium buffered from pH 3.0 to 7.0
77 Colony diameters of isolates P.sp.(lh) 2-2 (+-+),Ggg G1 (#-#) and Pg 1348-2 (••) after 5 days growth on LB medium buffered from pH 3.0 to 7.0
78 Interaction between P.sp.(lh) isolate 2-2 and Ggt isolate 38-4 on PDA medium buffered to pH 3.0, at 24°C
79 Interaction between Pg isolate 1348-2 and Ggt isolate T1 on PDA medium buffered to pH 6.0, at 24°C
80 Interaction between P.sp.(lh) isolate 2-2 and Ggt isolate T1 on PDA medium buffered to pH 6.0,at 24°C
81 Interaction between isolates P.sp.(lh) 2-2 and Ggg G1 and Ggt isolate OgA on LB medium buffered to pH 3.0, at 24°C 82 Interaction between isolates P.sp.(lh) 2-2, Ggg G1 and Pg 1348-2 and Ggt isolate OgA on LB medium buffered to pH 4.0, at 24°C
83 Interaction between isolates P.sp.(lh) 2-2, Ggg G1 and Pg 1348-2 and Ggt isolate OgA on LB medium buffered to pH 5.0,at 24°C
84 Interaction between isolates P.sp.(lh) 2-2 and Ggg G1 and Ggt isolate OgA on LB medium buffered to pH 6.0, at 24°C
85 Interaction between isolates P.sp.(lh) 2-2, Ggg G1 and Pg 1348-2 and Ggt isolate OgA on LB medium buffered to pH 7.0, at 24°C 86 Interaction between isolates P.sp.(lh) 2-2, Ggg G1 and Pg 1348-2 and Ggt isolate OgA on LB medium buffered to pH 8.0, at 24°C 13
PAGE
87 Interaction between isolates Ggg G1 and Ggt OgA on 220 LB medium buffered to pH 3.5, at 24°C
88 Interaction between isolates P.sp.(lh) 2-2 and Ggt 220 38-4 on LB medium buffered to pH 8.0, at 24°C
89 Interaction between isolates P.sp.(lh) 2-2 and Ggt 220 OgA on H medium buffered to pH 3.0, at 24°C
90 Interaction between isolates P.sp.(lh) 2-2 and Ggt 222 OgA on H medium buffered to pH 4.0, at 24°C
91 Interaction between isolates P.sp.(lh) 2-2 and Ggt 222 OgA on H medium buffered to pH 5.0, at 24°C
92 Interaction between isolates P.sp.(lh) 2-2 and Ggt 223 OgA on H medium buffered to pH 6.0, at 24°C
93 Interaction between isolates P.sp.(lh) 2-2 and Ggt 223 OgA on H medium buffered to pH 7.0, at 24°C
94 Interaction between Ggg isolate G1 and (a) Ggt 38-4 226 and (b) Ggt 019/6 on LB medium buffered to pH 4.0 and incorporating methylene blue
95 Interaction between Pg isolate 1348-2 and Ggt isolate 226 OgA on LB medium buffered to pH 4.0 and incorporating methylene blue
96 Well tests of isolates Ggg G1 and Pg 1348-2 with Ggt 228 45/10 culture filtrate 14
INTRODUCTION
Discovery and occurrence of dsRNA mycoviruses
The first report of a virus infecting a fungus appeared less than 20 years ago, with the discovery by Hollings (1962) of virus particles in the cultivated mushroom Agaricus bisporus. Interest in the search for mycoviruses was further stimulated by the study of two antiviral and interferon-inducing agents, helenine and statolon. The first of these, associated with a strain of the fungus Penicillium funiculosum, was shown to protect mice from swine influenza virus, while statolon, associated with a strain of Penicillium stoloniferum, was active against murine meningoencephalomyocarditis virus and Semliki Forest virus (Shope, 1948;
Powell et al., 1952). The finding that these activities were due to the presence of double-stranded RNA (dsRNA), derived from virus particles infecting P.stoloniferum and P.funiculosum (Ellis and Kleinschmidt,
1967; Lampson et al., 1967; Banks et al., 1968) has been followed by numerous reports of viruses and virus-like particles in fungi.
Mycoviruses are now known to be present in more than 100 species, representing all the main taxonomic groups of fungi. Random sampling of a number of fungal species has indicated that up to 20% of fungi may be infected with virus (Bozarth, 1972; Moffitt and Lister, 1975).
Several morphological types of particle have been discovered
(Hollings, 1978) but by far the most commonly occurring are isometric particles, with diameters in the range of 25 to 45 nm and, in all cases so far examined, with a genome of dsRNA. Virus particles have been found in more than 40 plant pathogenic fungi (Lemke, 1977, C.J. Rawlinson, personal communication).
Transmission of mycoviruses
In nature, virus particles have been shown to be transmitted 15 only intracellularly e.g. by cytoplasmic exchange between cells, by hyphal anastomosis and by means of spores (Buck, 1979).
Transmission during hyphal growth occurs by protoplasmic streaming.
In Agaricus bisporus, for example, particles are found scattered in the cytoplasm, often near the septae of cells, indicating their intracellular migration. Crystalline arrays of 34 nm particles in dense aggregates or linear arrangements, similar to those found in Penicillium stoloniferum and Penicillium chrysogenum, have often been observed but in general particles have not been found at the hyphal tips (Border et al., 1972).
Mycoviruses have been found in oonidiospores, basidiospores, uredinospores, zoospores and chlamydospores, but, so far, only in the ascospores of Saccharomyces cerevisiae. Ascospores from Gaeumannomyces graminis (Rawlinson et al., 1973), Helminthosporium maydis (Bozarth,
1977) and virulent strains of Endothia parasitica (Day et al., 1977) have been cultured to produce virus- or dsRNA-free isolates.
A recent theory (Buck, 1980) suggests that virus may arise in ascospore culture through being carried as a DNA provirus. Support for this hypothesis comes from a number of observations. Rawlinson eit al.,
(1973) found that isolates of Gaeumannomyces graminis from first year cereals are generally virus-free, consistent with the theory that initial colonisation was by virus-free ascospores (Brooks, 1965). However, in a study on one particular field, no virus particles were found in the first three years of cereal cultivation, but over a further period of 3 months, isometric particles 27 nm in diameter, mainly empty, appeared in 17 out of 38 randomly selected isolates. In the next 9 months, larger particles
(35 nm in diameter) appeared in many isolates. A possible explanation is that G. graminis carries endogenous DNA provirus, which can be induced to produce dsRNA and virus particles under certain conditions. Further evidence is obtained from the occurrence of empty viral capsids in
Aspergillus flavus, in which no viral nucleic acid can be detected. The 16 condition is a stable characteristic, transmissible through conidia, and the genetic information for capsid protein production could therefore reside in DNA provirus (Wood et al., 1974).
The presence of serologically related viruses in closely related species of fungi e.g. the 30 - 40 nm particles in Penicillium chrysogenum, Penicillium cyaneo-fulvum and Penicillium brevi-compactum
(Wood and Bozarth, 1972; Buck and Girvan, 1977) and the F and S viruses of Aspergillus foetidus (Ratti and Buck, 1972) and Aspergillus niger
(Buck, Girvan and Ratti, 1973; Buck and Ratti, 1975) may be explained on the basis that hyphal anastomosis between the species is possible, and the presence of isometric particles with the same diameter and similar capsid polypeptide amino-acid composition, suggests that this is likely.
However the presence of serologically related viruses in species between the isometric particles of Diplocarpon rosae and Penicillium stoloniferum virus S (Bozarth et al., 1972) is more difficult to explain. One theory
(Buck, 1980) suggests that divergent evolution of a fungal species could have occurred, while a virus infecting the fungus, with an intracellular mode of transmission, was conserved and persisted in the developing species.
Since the majority of mycovirus infections are latent, the demon- stration of transmission of viruses e.g. by electron microscopy, serology and dsRNA analysis, is laborious since the methods used, such as transmission of viruses in protoplasts and between genetically marked strains by heterokaryosis, produce a very low percentage of infected colonies. In a few cases, where symptoms of disease in the fungus are involved, the transmission of virus particles can be clearly demonstrated. While good evidence for the involvement of a cytoplasmic determinant in disease transmission has been obtained by heterokaryosis, proof of virus as the causative agent of a fungal disease has been shown only in a few cases, including that of the cultivated mushroom, Agaricus bisporus (Hollings,
1962). The viruses and the disease are transmitted by heterokaryosis and 17 in basidiospores, but the infection of healthy mushrooms by inoculation with cell-free virus preparations has also been achieved (Hollings, 1962;
Dieleman-van Zaayen and Temmink, 1968). The infection obtained was,
however, inefficient and difficult to reproduce. The infection of fungal
hyphae is presumed to be difficult because the fungal cell wall impedes
virus uptake. This problem has been overcome, in a few cases, by
infecting isolated protoplasts in vitro (Lhoas, 1971; Pallett, 1976).
Successful transmission of a disease causing stunted growth of the oat pathogen, Helminthosporium victoriae, has recently been reported, where
infection was achieved by fusion of normal virus-free protoplasts in the presence of a cell-free virus preparation (Ghabrial, 1980). However
other workers have found this a difficult and generally unsuccessful
technique (Bozarth, 1975).
Properties of dsRNA mycovirus particles in vitro.
DsRNA viruses can be conveniently divided into two groups, type 1
viruses containing a single polycistronic dsRNA genome and type 2 viruses
containing 2 or more monocistronic dsRNA molecules. The physical
properties of sane of the better characterised viruses of these two
groups are given in Tables 1 and 2). Up to eight dsRNA components, c c. ranging in molecular weight from 0.27 x 10 to 6.3 x 10 (Buck, 1980)
are associated with a given virus.
It is not certain whether all these components are required for
replication of the virus or whether some are derived from others by, for
example, deletion mutation resulting in defective interfering particles
(Huang and Baltimore, 1977). It is also possible that some components
are satellite dsRNAs, using the capsid protein of a helper virus for a
coat protein and making use of the replicative functions of the helper.
The majority of mycoviruses studied to date, for example, those of c. Aspergillus foetidus (except for the 0.27 x 10 dalton component of 18
Table 1 Physical properties of isometric dsRNA mycoviruses with undivided genomes
Mol. wt. Calculated Virus Particle of major Mol. wt. coding diameter capsid of virus capacity Ref, (nm) polypeptide dsRNA^ of dsRNA (x 1(n (x 10b) (daltons xlO )
Helminthosporium 48 121 6.3 350 1 maydis virus
Gaeumannomyces 40 94 6 330 2 graminis virus F10A
Saccharomyces 40 88b 3.4 189 3 cerevisiae virus3
Ustilago maydis 41 75 4.1 229 4 virus c
Mycogone perniciosa 42 69 4.3 239 5 virus
Helminthosporium 6 d victoriae virus 35-40 88 (+83?) 3.0 - (190S particles)
Virions in sensitive yeast strains contain a single dsRNA component (L dsRNA)
Oliver et al. (1977) reported 75,000 but this value is considered to be less reliable, because the determination involved an extrapolation from known standards; minor virion polypeptides of 53000 and 37,000 daltons (Oliver et al., 1977) and 140,000, 82,000 and 78,000 (Hopper et al., 1977) have been reported for this virus.
Virions in certain non-killer Ustilago maydis strains, derived from killer strains by mutation, contain a single dsRNA component (Koltin, 1977).
Almost certainly an underestimate, since the value was obtained by linear extrapolation from reovirus dsRNA standards (see Bozarth and Harley, 1976).
1. Bozarth (1977); 2. Almond (1979); 3. Hopper et al. (1977);
4. Bozarth (1979); 5. Barton (1978); 6. Sanderlin & Ghabrial (1978). Table 2 Physical properties of isometric dsKNA mycoviruses with segmented genomes.
Virus Particle diameter MOl.wt of virion NO. of copies of mol.wt. of virion Coding capacity of (nm) polypeptides polypeptides per dsRNA components dsRNA. components (x10 ) virion (x10 ) (daltons x10 )
Penicillium 2.2 123 chrysogenum 35-40 130 60 2.1 117 virus 2.0 111
Aspergillus 2.7 150 foetidus , 35-40 125 1 1.87 104 — t virus F 100 1 1.70 94 87 120 1.44 80 1.24 69
Mushroom 35 64 ND 1.5 83 virus 4 c 1.4 78
Penicillium 59 1 0.99 55 stoloniferum. 30 47 120 0.89 49 virus F c 0.46 27
Penicillium 56 1 1.0 61 stoloniferum. 30 42 120 0.94 52 virus S c
* Wood and Bozarth (1972); Buck and Girvan (1977) b Ratti and Buck (1972); Buck and Ratti (1975) ® Barton and Hollings (1979) Bozarth et al. (1971), Buck and Kempson-Jones (1974), Bozarth and Harley (1976)
ND not <3etermine<3. Reproduced with kind permission of Dr. K.W. Buck (Buck, 1980) 20
A.foetidus virus S), Penicillium chrysogenum and Penicillium stoloniferum
(Buck, 1979b) encapsidate each segment of dsRNA in a separate protein coat. In general the protein coats are constructed from one major polypeptide, ranging from 25,000 to 130,000 in molecular weight, for different viruses. Using in vitro translation systems the capsid proteins of P. stoloniferum virus S, P. chrysogenum virus and Saccharomyces cerevisiae virus have been shown to be encoded by dsRNA (Buck, 1979;
Bostian et al., unpublished and Hopper et al., 1977). In those viruses with undivided genomes, e.g. Ustilago maydis virus, Helminthosporium maydis virus and S. cerevisiae virus, the coding capacity of the dsRNA has been shown to be in excess of that required for the major capsid polypeptide (Bozarth, 1977, 1979; Hopper et al., 1977), only one third of the capacity being utilised. Since the minimum requirement for such a virus to replicate within a host is the encoding of virus capsid poly- peptide and an RNA-dependent RNA polymerase to replicate the genome, it
is likely that these single dsRNA genomes are dicistronic at least.
Viruses are assumed to have multipartite genomes when the coding capacity of a single dsRNA component is close to that required for the major capsid polypeptide i.e. it is likely to be monocistronic.
All the known dsRNA viruses of other hosts have divided genomes e.g. reovirus has ten monocistronic segments enclosed in a complex double- shelled capsid (Joklik, 1974), bacteriophage 0(6 has three polycistronic dsRNAs encapsidated in a lipoprotein shell (van Etten et al., 1976) and
Drosophila X and similar animal viruses have two segments of encapsidated dsRNA (Dobos et al., 1978). However, all these systems differ from the mycoviruses as each virion encloses one copy of all the dsRNA segments of the genome. Isometric dsRNA. mycoviruses are generally smaller than the other dsRNA viruses, with capsids which each enclose a separate dsRNA component. This multicomponent system is no disadvantage to viruses transmitted intracellularly and may have evolved as a convenient method 21 of dividing the genome into monocistronic segments. It may be an advantage in allowing variation in the numbers of individual segments in the virions, allowing, for example, the accommodation of satellite dsRNAs, and control, in the case of viruses with associated RNA polymerases, of production of individual messenger RNAs. Virions of
Penicillium stoloniferum viruses S and F, Penicillium chrysogenum virus,
Aspergillus foetidus virus F and Saccharomyces cerevisiae virus have been shown to contain associated RNA polymerase activity (Chater and Morgan,
1974; Nash et al., 1973; Ratti and Buck, 1975; Herring and Bevan, 1977), although formal proof that the enzyme is encoded by virus dsRNA is lacking. However, the dsRNA genome of each of these viruses has the coding capacity for both capsid polypeptide and polymerase enzyme.
Double stranded RNA does not act as messenger RNA, and an enzyme is required to transcribe the dsRNA genome to produce single-stranded RNA.
Although cellular enzymes have been shown to be able to transcribe dsRNA in vitro e.g. transcription of P. chrysogenum virus dsRNA by a
DNA-dependent RNA polymerase of Escherichia coli (Sugiura and Miura,
1977), and in vivo e.g. the infectivity of the replicative form of poliovirus suggests transcription of this dsRNA in vivo (Baltimore,
1969), the efficiency of transcription is very poor and in general the presence of virus functions has been found necessary for viral RNA replication. All the dsRNA viruses examined so far, including reovirus and'phage 06 (Silverstein et al., 1979; Partridge et_ al., 1978) have been shown to contain a virion-associated RNA polymerase. Genetic evidence has shown the enzymes associated with reovirus and 'phage 06 to be virus encoded (Cross and Fields, 1977; Rimon and Haselkorn, 1978).
Virions of Penicillium stoloniferum virus S possess an RNA polymerase activity whereby the dsRNA is transcribed semiconservatively, i.e. one daughter strand is synthesised on the dsRNA, thus displacing one of the parental strands, and a second daughter strand is subsequently 22 synthesised on the displaced parental "template" (Buck, 1978).
Aspergillus foetidus virus S virions, isolated from stationary phase culture, have also been shown to possess a transcriptase acting with a semiconservative mechanism (Ratti and Buck, 1978). Full length single- stranded RNA copies of dsRNA 2 are released from the virus particle. In
Saccharomyces cerevisiae virus, there are two types of associated polymerase activity, one a ss to dsRNA activity isolated from log phase cells (Bevan and Herring, 1976) and the other a transcriptase (ds to ssRNA) activity isolated from stationary phase cells (Herring and
Bevan, 1977; Bruenn et al., 1980). Herring and Bevan have postulated that replication of yeast virus dsRNA may occur asynchronously, in a similar way to that of reovirus. Thus the dsRNA mycovirus group show a diversity of genomic organisation and may not all employ the same mode of replication. In comparison, viruses of Reoviridae show conservative replication (Silverstein et al., 1976) and 'phage 06 appears to have a semiconservative displacement mechanism (Coplin et al., 1975; Van Etten et al., 1980) similar to that of A.foetidus virus S (Ratti and Buck,
1978).
Effects of dsRNA mycoviruses on their hosts
One of the typical features of many of the fungal viruses is their latent character. It was probably the absence of any apparent effect by these viruses on their host fungi that prevented their early recognition.
Despite, in some cases, accommodating large numbers of particles, fungal
cells generally develop without impairment to growth or synthesis of DNAr
RNA and protein and without cell lysis. This is likely to be due to the fact that virus replication is too slow to adversely affect host replication by competition, and that no toxic products are released by the virus to which the fungus is susceptible (Buck, 1980). It has been calculated for Penicillium stoloniferum, which produces one of the highest yields of virus particle, that only 0.5% of the dry weight of the fungus is virus (Buck, 1977) and it has been shown that only 20% of the particles are synthesised during the exponential stage of the fungus growth. It may be that growth is limited by competition between host and virus in the exponential phase in some way. Thus replication is controlled in the apical hyphal compartment where host growth is occurring, but virus particles build up in older cell compartments where growth is slowed. This has been demonstrated by electron microscopy of sections of fungal hyphae by Border et al., (1972). No particles were found at the hyphal tips.
The disease of the cultivated mushroom Agaricus bisporus, called
'die-back' disease, arose spontaneously in the U.S.A. in 1949, and in
Europe shortly afterwards (Sinden and Hauser, 1950) and, it has been suggested (Buck, 1980), could have developed as a result of a virus mutation within the fungus. There have been some reports of low tit res of particles serologically related to viruses from diseased mushrooms in many apparently disease-free mushroom spawns (Del Vecchio, Dixon and
Lemke, 1978) and one report (Passmore and Frost, 1974) of high titre virus in symptomless mushrooms, which throw doubt on the virus cause of the disease. Generally, however, the disease is thought to be trans- mitted by a cytoplasmic determinant, which has been shown to be a dsRNA virus. Which single virus is the cause of the disease, out of the five mushroom viruses which have been described (Van Zaayen, 1979), is still in doubt (Barton and Hollings, 1979; Hollings, 1962).
In Rhizoctonia solani, a degenerative disease of the fungus called
Rhizoctonia decline, results in a marked reduction in the level of virulence of this pathogen, in addition to the morphological and physio- logical changes observed such as hyphal lysis, slow growth and reduced sporulation (Castanho and Butler, 1975). Recent evidence has lent support to the theory that hypovirulent strains i.e. diseased strains, could be used in the biological control of the fungus. Protection of sugar beet and cabbage seedlings by hypovirulent strains has been demon- strated both on agar and in soil (Castanho and Butler, 1978). DsRNA was found in one of these strains and was shown to be transferred to the pathogenic variant by anastomosis. Isolates were found to be either healthy and to contain no detectable dsRNA or hypovirulent and containing unique segments of dsRNA (Castanho et al., 1978).
Helminthosporium victoriae, the cause of victoria blight of oats, has been found to be associated with two serologically distinct virus particles of 145S and 190S. A disease of H.victoriae was first noted by
Lindberg (1959) who found several abnormally stunted colonies of the fungus among cultures newly isolated from diseased oats. The disease of the fungus could be transmitted by anastomosis and resulted in severely stunted isolates markedly reduced in their ability to sporulate and occasionally suffering lysis of hyphae. Nearly 20 years after Lindberg suggested that the disease was caused by a virus, evidence supporting this hypothesis was obtained. The larger 190S particle was found in both healthy and diseased isolates (Sanderlin and Ghabrial, 1978), but the
145S particle was found only in the diseased strains where good correlation was found between the levels of this virus and the severity of the disease in the colony. Distinct dsRNA components were present in both particles, a single species in the 190S particle, and four smaller dsRNA species in the smaller particle (Ghabrial et al., 1979). It is thought the 190S particle may be a helper virus, similar to the L species of yeast virus (Hopper et al., 1977) as 145S particles have never been found alone in a strain of H.victor iae (Ghabrial et al., 1979). Virus-free isolates have been infected with VLPs by cell fusion in the presence of a preparation of the two viruses, but proof of the role of the 145S species awaits infection of protoplasts, already carrying the 190S particle, with a pure virus preparation of 145S particles. Comparison with protoplasts 25 infected with the 190S particle alone, presumably causing no disease, would then be possible.
Chestnut blight, caused by the pathogen Endothia parasitica, has resulted in the virtual elimination of the chestnut as a major species in North America. However, the spread of hypovirulent variants, particularly in Italy, is thought to have prevented chestnut blight remaining an important disease in Europe. Van Alfen et al. (1975) showed this hypovirulence to be due to a transmissible cytoplasmic factor and suggested that these strains could be used in the biological control of the disease. Ihe cytoplasmic factor was shown to be transmitted together with dsRNA, and a club-shaped particle has now been purified from hypo- virulent strains (Dodds, 1979) and has been shown to contain high 6 6 molecular weight dsRNA in the range 5.Ox 10 to 7.Ox 10. A non-pathogenic strain of E.parasitica has been used to control blight cankers on trees in America, but some were found to be ineffective. This was thought to be caused by vegetative incompatibility between the interacting strains, and since hyphal anastomosis is required for the transfer of the determinants for hypovirulence, this presented a problem for the use of these strains as controlling agents (Anagnostakis, 1977).
However more recently Anagnostakis and Day (1979) found that some vegetatively incompatible pairings allow the transmission of hypovirulence determinants. It was suggested that either the presence of the dsRNA may override vegetative incompatibility, or that even in an incompatible interaction there may be sufficient time to permit transfer of dsKNA before cell death occurs.
DsRNA and the killer phenomenon
Certain strains of the yeast Saccharomyces cerevisiae have been shown to secrete a toxin or 'killer factor' lethal to sensitive strains of the same or closely related species (Woods and Bevan, 1968; Philliskirk and Young, 1975). Neutral strains are also found which are insensitive Table 3 Interaction between killer yeasts
S. cerevisiae A8209B K1 - - - + + + 232,235 S.hybrids 631.663
1001 S.cerevisiae 738, K2 + + + + S.diastaticus 713
S.capensis 761 K 3 + + + -
Torulopsis glabrata K 388 4 + — — Debaromyces vanriji 577
Hansenula anomala 434 K - - - - 5 - + + + H.subpelliculosa 16
Kluyveranyces fragilis 587 K — 6 -• + + + + +
Candida valida 327 K7 + + + + + Pichia membranaefaciens333
H. anomala K — 435 8 + + + + + — H. mrakii 500 K9 + + + + + + +
K. drosophilarum 575 K 10 + + + + + + + + +
Reproduced with kind permission of Dr. T.W. Young (Young and Yagiu, 1978) 27 to the killer toxins and which do not produce a toxin. A classification was recently made by Rogers and Bevan (1978) on the basis of immunity and killing ability in fourteen strains of different species. Using buffered agar containing methylene blue to determine the ability of these strains to kill seven stock sensitive strains, they found the killers all expressed their ability over a narrow pH range. The strains fell into four groups dependent on their killing and immunity cross-reactions:
TOX 1 Laboratory stock strains of Saccharomyces cerevisiae,
Candida albicans and a neutral strain of S.cerevisiae
TOX 2 Brewery and wine fermentation strains of Saccharomyces
species and a neutral strain of S.cerevisiae
TOX 3 Torulopsis glabrata
TOX 4 Saccharomyces drosophilarum
Young and Yagiu (1978) investigated killing and immunity inter- actions between twenty killer yeasts of different genera and species and found ten killer classes (K1 to K10, Table 3), the K1, K2 and K4 classes corresponding to the TOX 1, 2 and 3 groups of Rogers and Bevan. The majority of killer strains were found to be stable to pH up to pH 5.0.
The Sake yeast killer of Kotani et al. (1977) was stable up to pH 7.0, but in other characteristics fell into the K2 class. The optimum pH for these toxin activities falls within a narrow range of pH 4.2 to
4.7. Killer groups could be further distinguished by their sensitivity to proteolytic enzymes, showing a range of susceptibility to papain, pronase or pepsin at pH 4.2. By means of these tests and by gel chroma- tography of crude killer factor, the toxins of different groups were shown to be distinct, although all were shown to be proteins. Toxin from a single strain of Saccharomyces cerevisiae has been purified and has been shown to be a protein of molecular weight 11,500 (Palfree and 28 Bussey, 1979). Most of the toxins were found to be produced by yeasts from two genera, Saccharomyces and Hansenula, both of which undergo simple heterothallic mating. The killer strains of yeast are generally autoimmune, but strains within the genus Saccharomyces can kill each other. The toxin molecules are presumed to be distinct in this case since the immunity of the 'sensitive' killer to its own toxin does not extend to the killing toxin.
The mechanism of action of the Saccharanyces cerevisiae K1 toxin involves leakage of ATP and cellular potassium (K+) ions from the cell and inhibition of macromolecular synthesis (Skipper and Bussey, 1977);
Kotani et al., 1977) suggesting an alteration in the cell membrane permeability. De la Pena et al. (1980) report a rapid disturbance of the electrochemical proton gradient across the plasma membrane, as a result of toxin action on sensitive cells. Bussey (1972) suggested that early glycolysis or some catabolite-sensitive step was the site of killer action, but the exact nature of the reaction is unclear.
Genetic analysis of Saccharonryces cerevisiae cells segregating for killer and immune characteristics, showed that, although inherited cytoplasmically, the phenomenon is under the influence of both cytoplasmic and nuclear determinants. The elements associated with toxin production and immunity have been identified as dsRNA plasmids (Wickner, 1976).
Sensitive laboratory strains are infected with virus particles which contain a dsENA component (L-dsENA) with a molecular weight in the range
2.5 to 3.4 x 106 (Hopper et al., 1977; Young and Yagiu, 1978). Killer strains also contain these L virions and an additional dsRNA component, termed M-dsENA, with a molecular weight in the range 1.0 to 1.5 x 10
(Bevan et al., 1973; Toh-E et al., 1978; Bruenn, 1980). The two virions have a particle diameter of 35 to 40 nm and have identical capsid proteins.
It is thought, since no strain has been found containing M-dsKNA alone, that this is a satellite dsRNA, with the L-virion synthesising capsid 29 protein for both particles and an RNA polymerase for replication of the two species of dsRNA. This is analogous to the case for satellite
RNA-5 of cucumber mosaic virus (Mossop and Francki, 1978). Evidence consistent with this theory is provided by the finding that the major
L-virion capsid protein is identical with the major m vitro translation product of the L-dsRNA species (Hopper et al., 1977), and the cross- antigenicity of the M and L virions suggests that at least one polypeptide is identical (Herring and Bevan, 1975). Some of the M-dsRNA species observed have insufficient coding capacity to code for a polypeptide of molecular weight 88,000. It is unlikely that nuclear genes code for this species as, after extensive searching, no chromosomal mutants for this function have been found and it is unlikely to be essential for cell survival. The M-dsRNA species is not a cleavage product of L-dsRNA as no sequence homology has been found between the two species by fingerprinting or hybridisation (Bruenn and Kane, 1978). Other than one report to the contrary (Vodkin, 1977), no homology between yeast chromosomal DNA and dsRNA from several yeast strains has been demonstrated. The kinetics of renaturation of L-dsRNA in the presence of yeast DNA were as predicted for the species individually and homology was found to be less than one
L-dsRNA copy per forty yeast genomes (Hastie et al., 1978).
In strains of Saccharomyces species cured of killer activity and immunity to killer, for example by treatment with cycloheximide, heat or
5-fluorouraci1 (Fink and Styles, 1972; Wickner, 1974; Mitchell et al.,
1973), only the M-dsRNA species disappears. Recent evidence, derived from in vitro translation studies, has proven that the M-dsRNA species encodes the toxin (Bostian et al., 1980).
Only Saccharomyces species and those in killer groups K1 to K3 could be cured of killer activity by cycloheximide or by heat, with the loss of the M-dsRNA species. Other strains in groups K4 to K10 could not be cured and were found not to contain dsRNA. The genetic basis for the 30 killer activity of these strains is uncertain (Young and Yagiu, 1978).
However a recent report links the occurrence of two linear DNA plasmids with killer activity in the yeast Kluyveromyces (Copper, 1981).
Although nuclear genes do not code for the killer activity of
Saccharomyces strains a number of these genes have been found essential for the stable expression of the killer phenotype in the cell. These fall into two types: those genes required for the maintenance of the killer function, i.e. the replication of M-dsRNA (genes mak 1 to 27, pet
18 and spe 2) (Bevan et al., 1973; Wickner, 1978; Cohn et al., 1978) and those genes required for the expression of the killer function i.e. the synthesis and secretion of killer toxin (kex genes) and the expression of resistance to toxin determined by M-dsRNA (rex genes) (Wickner and
Leibowitz, 1976). There are more than 20 distinct loci for these genes, distributed between 15 of the 18 yeast chromosomes. Recessive mutations
in the maintenance genes affect the presence of dsRNA, while those in kex and rex genes do not. Other types of host chromosomal mutation can affect the killer phenotype, e.g. ski mutations, in which the genes recessive mutations result in a 'superkiller' phenotype. On heating
to 30°C, mutants with this phenotype produce a larger zone of killing
than wild type strains. It is thought possible, therefore, since the killer toxin is normally rapidly degraded at 30°C, that the ski mutation results in a more heat resistant toxin. Kex 1 and kex 2 mutations are epistatic to ski mutations. Mutations in rex genes affect the expression of resistance to killer toxin determined by M-dsRNA (Wickner, 1974).
In general growth of killer strains at elevated temperatures
results in mutation to sensitive non-killers (Wickner, 1974), with the
loss of the M-dsRNA species. Suppressive non-killer mutants have also been obtained by mutation. These contain a small species of dsRNA, termed c S-dsRNA (molecular weight 0.5 x 10 ) but lack the M-dsRNA molecule (Bruenn and Kane, 1978). The mutants, when crossed with wild-type killer strains, result in nonkilling diploids and sensitive meiotic products lacking M-dsRNA. Electron microscopy heteroduplex mapping has shown that one S dsRNA species has arisen by internal deletion of the wild type
M-dsRNA (Fried and Fink, 1978). These mutant dsRNAs have been compared with the defective interfering particles of animal viruses, as the mutant species prevent the propagation of the wild type species of dsRNA when both are present in the same cell. The mutant and wild-type dsKNAs share the same terminal sequences, suggesting these are necessary for replication.
None of the products of the 29 genes required for M-dsKNA replication are needed for L-dsRNA replication. In fact no mutants of yeast have been found unable to replicate L-dsRNA.
A 'killer' system has also been found in the cereal smut pathogen
Ustilago maydis, strains of which release a toxin lethal to sensitive strains of the same or related species (Koltin and Day, 1975). Three different killer specificities, P1, P4 and P6, have been identified, each associated with different but related viruses, and each having a distinct pattern of dsFNA components (Koltin and Day, 1976a). The virus particles are 41 nm in diameter and contain four dsRNA components with molecular weights in common. Each particle has a single capsid protein of 75,000 molecular weight, vAiich is probably identical in the three cases.
The three toxins are low molecular weight proteins (10,000) and can be distinguished in several ways. They are unrelated serologically, have different migration patterns on polyacrylamide gel electrophoresis and show varying dependence on pH and temperature. Also, while the P6 toxin appears to be a true 'killer' toxin, capable of killing 100% of sensitive cells, the P1 and P4 toxins are far less effective, even when several-fold more concentrated than the P6 protein. The biological distinctions made between these toxins were confirmed through heterokaryon formation and cross testing. All attempts to include two killer specificities in a single cell have failed. The introduction of more than one virus type into the same cell results in the exclusion of one or both viruses and removal of certain of the dsRNA components. This is termed the exclusion phenomenon (Koltin and Day, 1976b), also found in Saccharomyces cerevisiae.
The killer function and immunity to toxin are cytoplasmically transmitted, although immunity (resistance to toxin) can also be determined by nuclear genes as in P2 cells. These cells are sensitive or resistant to toxin depending cn the presence of one or two alleles at a chromosomal locus. P3 cells carry resistance determined by dsRNA and produce no toxin, like neutral strains of Saccharomyces cerevisiae, but can be distinguished from P2 cells by heterokaryon analysis (Koltin and Day,
1976a; Wood and Bozarth, 1973). The genes for resistance to toxin are distinct for the three toxins and are recessive. In contrast to yeast, no nuclear genes for the maintenance and expression of killer functions have been detected in Ustilago.
Using mutant cells containing partial viral genomes, Koltin (1977) was able to assign genetic functions to specific dsKNA components through genetic crossing. A single species, H1 dsRNA, was found to have sufficient information for capsid protein formation and replication of dsKNA, and is thought a possible candidate for a helper function, as for the L-dsKNA species in yeast. Two other dsKNA molecules are associated with killer expression and immunity to killer (Koltin, 1977; Koltin and Kandel,
1978).
The rapidity with which cells were killed and the lethal substance removed by both living and dead cells, led Hankin and Puhalla (1971) to propose that the Ustilago toxin acted, at least initially, at the cell surface. It has been suggested that the toxin exerts its effect via the exclusion of dsKNA molecules in sensitive cells, by means of a nuclease activity, (Levine et al., 1979). Using mutants lacking dsRNA for killer 33 expression in crosses with sensitive cells, product cells were obtained containing dsRNA molecules from both parents. This led to the theory that the killer protein acted as a nuclease enzyme, attacking the nucleic acid of a super infecting species. These workers were able to demonstrate an endonucleolytic activity against ssENA and DNA, but not dsENA.
However, the toxins showed a specific cell recognition of which nucleases of a similar size were not capable. (Levine et al., 1979).
Koltin and Day (1975) have suggested that the Ustilago maydis killer system could be used to control the smut disease of cereals if the genetic information for the killer protein could be introduced into the host plant (as dsRNA or DNA), where it could be expressed as resistance
to toxin-sensitive cereal smuts. Despite the fact that some strains of
U.maydis itself are known to be resistant to all three killer strains, most other cereal smuts appear to be sensitive to at least one of them
(Koltin and Day, 1975).
Fungal killer systems have some similarity to the bacteriocins
(Hardy, 1975). Killer toxins and bacteriocins are both proteins lethal
to sensitive strains of the same or related species and are encoded by dsRNA and DNA plasmids respectively. Immunity in the oolicins, i.e.
synthesis of a product to neutralise the toxin, is under the control of
the plasmid, while resistance, i.e. absence of a cell surface receptor
for the toxin, falls under the control of nuclear genes. Colicin
E1 appears to act initially at the cell surface, and results in an + efflux of K ions from the cell and disruption of the energised state
of the membrane. Colicin E2 has been shown to inhibit cell division,
probably by direct cleavage of DNA, and colicins D and E3 cause one
species of riboscmal RNA (16S) to be cleaved into fragments, resulting in
the inhibition of protein synthesis. Hankin and Puhalla (1971) found
that the conditions for maximum production of P1 inhibitor in Ustilago 34 maydis did not coincide with those for maximum growth, which observation has been confirmed by studies on buffered and unbuffered media (Day and
Anagnostakis, 1973? Puhalla, 1968), where stabilisation of pH promoted inhibitor production. A difference in the temperature dependence was also noted for optimal growth and inhibitor production in the bacterial colicin system.
Colicinogeny and the killer systems of Ustilago and Saccharomyces are a few of the many factors contributing to the relative competitiveness of pathogenic strains of bacteria and fungi respectively. The significance of these systems in the establishment of the pathogens and in the prevention of infection requires further study, and advances in the molecular biology and genetics of these organisms (Hardy, 1975) may help not only to increase our understanding of their ecological significance but also to establish the relationships between the killing and inhibitory functions and the cell plasmids as units of selection.
Mycoviruses in the take-all fungus
The take-all disease
Take-all disease of cereals is caused by varieties of Gaeumannomyces graminis (Sacc.) v. Arx and Olivier, an ascomycete fungus. The disease, a serious root rot, is the result of infection by Gaeumannomyces graminis
(Sacc.) v. Arx and Olivier var. tritici (Ggt) in wheat and barley, and by
G.graminis (Sacc.) v. Arx and Olivier var. avenae (E.M. Turner) (Gga) in oats. The root rot together with seme stem-base disease in the cereal plant can result in dead patches in the field, or stunting of older plants. Inflorescences may become bleached and later contain shrivelled grain, resulting in the 'whiteheads' typical of severe take-all infection.
Less severe generalised infection throughout a crop can reduce a yield, despite lack of obvious symptoms on aerial parts of the plant (Polley and Clarkson, 1980). 35 In both varieties of the take-all fungus, hyphae and asoospore germ tubes penetrate root hairs and epidermis in the region of the meristematic zone and result in the plugging of the xylem and phloem and root death.
As these fungi are able to spread readily over the surface of the host root system even a relatively low soil population of infective fragments can result in severe disease development.
Control of take-all disease
There are no resistant cultivars to this disease and reasonable control has only been possible, so far, by crop rotation. Chemical control using various fertilisers has not proved successful. Although application of NH^+N in the field was shown to reduce take-all, an effect thought to be partially related to pH as it was negated by liming
(Brown, Hornby and Pearson, 1973), in cases where take-all development was favoured, application of nitrogen was observed to favour the disease also (Garrett, 1948). As no conventional control method proved simple in its application, alternative methods were explored. cu\HvoJCion of
Under continuousA susceptible cereal crops a decline in the severity of take-all, a phenomenon called 'take-all decline' (TAD) has been noted after the second or third cereal crop and following peak disease. There have been many suggestions as to the possible cause of TAD. (Hornby,
1979). These include the development of an antagonistic soil microflora containing e.g. bacteria and actinomycetes antagonistic to Gaeumannomyces graminis var. tritici (Lester and Shipton, 1967) and nutritional changes in the rhizosphere (Brown, Hornby and Pearson, 1973). With the finding of virus particles in two weakly pathogenic take-all isolates, from wheat fields showing take-all decline (Lapierre et al., 1970; Lemaire et al.,
1970) another possibility was added to the list. These workers distinguished two types of isolate: abnormal virus containing isolates which were hypovirulent, and normal isolates without virus, which were strongly pathogenic. The virus particles in the 'abnormal' hypovirulent 36 isolates were shown to be 35 nm in diameter and isometric in morphology
(Lapierre, 1973).
Two possible methods for controlling the disease of take-all were proposed by Lemaire et al. (1971). The first method involved the stimulation of the antagonistic soil microflora by incorporation of nutrient proteins into the soil, and the second involved the introduction of hypovirulent, virus-containing isolates into take-all infested soil to protect the host plant. A series of experiments using the latter method showed that take-all severity was greatly reduced in the presence of the hypovirulent isolates, and this was held to constitute the first report of biological control of a plant pathogen by a mycovirus (Bozarth,
1972).
However, Rawlinson et al. (1973) were able to dispute the theory that virus infection effected take-all decline, after an examination of
156 isolates, 114 of which contained virus. They found two sizes of isometric particle of diameters 27 nm and 35 nm and showed that particles of either size, or both, could occur in isolates, but that there was no evidence that virus infection impaired pathogenicity, morphology or saprophytic survival of the fungus. Nor were particles confined to soil showing TAD.
Later work has indicated a more complex situation. Blanch (1977) has shown that pathogenicity in Gaeumannomyces graminis var. tritici is under polygenic control and has suggested that host genotype is a primary factor influencing pathogenicity in the field. Any additional effect of virus on pathogenicity would therefore best be tested using genetically defined strains of the fungus.
Frick and Lister (1978) in an examination of 22 isolates of
Gaeumannomyces graminis for dsRNA, using antisera to synthetic double- stranded polyribonucleotides, and for virus, using electron microscopy and antisera to partially purified virus particles, found evidence for 37 serotype variation among purified particles. They showed degrees of relatedness among viruses in G.graminis var. avenae and in some isolates of G.graminis var. tritici in wheat.
Almond et al^. (1977, 1978) purified virus particles from 12
Gaeumannomyces graminis var. tritici (Ggt) isolates, and analysed dsRNA and polypeptide components by polyacrylamide gel electrophoresis. No two isolates were found to contain the same distribution of dsRNA and poly- peptide components. Serological investigations, shewing degrees of relatedness in various groups of virus particle, confirmed and extended the proposal of Frick and Lister (1978) that Ggt viruses commonly occur in a wide range of serotypes, which suggested biotype variation as an explanation for inconsistencies in reports of association of G.graminis with pathogenicity.
The observation was also made that, while some of the viruses from different isolates were serologically closely related, had similar capsid molecular weights and a number of dsRNA components in common, yet in two of these isolates additional dsRNA components, without apparent function in virus structure or replication, were present (Buck et al., 1981).
These were thus implicated as possible satellite dsRNAs. As a number of other workers (Chambers and Flentje, 1967; Rawlinson et al., 1973 and
Nilsson, 1969) had observed heterokaryon (post fusion) incompatibility reactions between isolates of Ggt, and a few instances of inhibition or killing at a distance, it was thought that these possible satellite dsRNAs might act in a similar manner to those of Saccharomyces cerevisiae and Ustilago maydis, i.e. in the production of killer toxin. A search was made for such a factor in Ggt and a diffusible growth inhibitor was subsequently found in three Ggt isolates (Romanos et al., 1980). Although resembling the killer toxins in their requirement for low pH for activity, this inhibitor did not appear to be lethal to sensitive cells at the concentration produced, had a broad spectrum of activity uncharacteristic 38 of bacteriocins or killer factors and did not appear to be associated with specific virus dsENA components. This does not exclude the possibility of a true killer toxin being found in Ggt, since only three killers have been found out of more than 200 Ustilago strains investigated, nor that another function of the satellite dsRNAs may be discovered.
This work indicated the strong necessity of using genetically defined material, freed from virus, e.g. by production of perithecia and isolation of ascospores, or through hyphal tip culture, to assess the effect of virus infection on pathogenicity.
In their survey of killer strains in various yeast species, Young and Yagiu (1978) found that a number of these isolates produced killer toxin in the absence of virus particles or dsRNA (K4 to K10 killer groups). Woods et al. (1974) also found a spontaneous mutant from a sensitive strain of Saccharomyces cerevisiae, which was able to 'kill1 K1 group killers (Young and Yagiu, 1978) and vice versa, yet contained only the L-dsRNA species. It is possible, therefore, that the killer toxin may be encoded by a nuclear gene, suggesting the possibility of a DNA provirus integrated into the host genome. However, a recent report
(Cooper, 1981) showed that strains of Kluyveromyces containing two linear
DNA plasmids were able to kill both S. cerevisiae killer and sensitive strains. These represent the first non-circular DNA plasmids reported in yeast or any other organism.
Cross-protection of cereals against take-all disease with weakly pathogenic or avirulent root parasites; possible role of viruses.
The weakly or non-pathogenic parasites of cereal and grass roots,
Gaeumannomyces graminis var. graminis (Ggg), Phialophora species with lobed hyphopodia (P.sp.(lh) and Phialophora graminicola (Pg) are closely related to Ggt i.e. Ggg is a variety of Gaeumannomyces graminis,
P.graminicola is likely to be the oonidial state of Gaeumannomyces 39 cylindrosporus (Hornby et al., 1977) and Phialophora species with lobed
hyphopodia is probably the same fungus as Gggf but without the sexual state (Walker, 1981). These parasites in the Gaeumannomyces-
Phialophora complex have been of increasing interest in recent years as many isolates have been shown to restrict the mycelial spread of Ggt and so protect their hosts against take-all disease (Scott, 1970; Balis,
1970; Wong, 1975; Deacon, 1974). G.graminis var. graminis is common on grasses but also causes brown sheath rot disease of rice. Phialophora sp.
(with lobed hyphopodia) is associated with maize and other cereals, and
Phialophora graminicola is common on grasslands of several countries. It is because these species are common on grassland, and less so on cereal crops, that short term grass leys have proved useful in minimising take-all in intensive cereal crop rotations (Deacon, 1973).
Due to recent confusion over the taxonomy of these species in the literature, and in order to introduce sane uniformity to references to these fungi, a standard usage has been proposed (Walker, 1981). For this reason, the Phialophora radicicola var. graminicola of Balis (1970) and
Scott (1970) has been raised to the species Phialophora graminicola and the Phialophora radicicola var. radicicola of Deacon (1974) to Phialophora species with lobed hyphopodia. The other species mentioned here, namely
Phialophora fastigiata, Phialophora hoffmannii and Phialophora malorum are unaffected.
Scott (1970) isolated the avirulent Phialophora graminicola
(then termed Phialophora radicicola Cain var. graminicola (Deacon)
Walker) from British grasslands, and showed that it inhibited the spread of Gaeumannomyces graminis var. tritici when used to precolonise wheat roots. Subsequent experiments showed that P.graminicola was able to reduce take-all on wheat roots, if these were allowed to grow through a layer of the avirulent fungus before encountering the pathogen (Balis,
1970; Deacon, 1973). 40
Phialophora species with lobed hyphopodia was also shown to protect wheat roots against take-all, when potted wheat plants were able to put out roots through layers of first the avirulent strain and then the pathogen (Deacon 1974). Since this species grows much more quickly than
Phialophora graminicola at 20 °C, and invades root cortices more extensively, it was hoped it would prove more useful as a cross-protectant. However, it was shown to be ineffective when used in an experiment with soil containing large amounts of take-all inoculum (Deacon, 1976).
Weakly pathogenic strains of Ggt isolated from take-all decline soils in France, were shown to reduce disease in pot experiments, particularly if they colonised wheat roots 3 to 4 days before inoculation of take-all (Tivoli et al., 1974).
Wong (1975) showed that precolonisation of wheat roots with non- pathogenic isolates of Gaeumannomyces graminis var. graminis, cross- protected the roots from infection with G.graminis var. avenae (E.M.
Turner) Dennis, as well as from Ggt. He proposed that the restriction of
Ggt was not due to direct suppression of the pathogen, since this infects distal portions of root already infected with Ggg, but was probably host mediated, possibly by phytoalexin production.
Sivasithamparam (1975) demonstrated that three Phialophora species,
P.fastigiata (Lagerb. and Melin) Conant, P.hoffmannii (Beyma) Schol-Schwarz and P.malorum (Kidd and Beaumont) McCalloch, were able to reduce take-all in pot experiments where seminal wheat roots grew through one of these species before encountering Ggt.
Thus restriction of Ggt on cereals is most effective using avirulent strains which are closely related to the pathogen, and suggests that they may be successful through occupying the same sites of infection as Ggt, and thus displacing the pathogen by prior occupation. However, an unrelated fungus Fusarium oxysporum has also been shown to effectively cross protect cereals from take-all (Sivasithamparam, quoted in Wong,
1981). 41
It has been suggested that cross-protection could involve the induction of host responses by the weakly or non-pathogenic fungi, since there is a requirement for prior colonisation of several days by these species and direct suppression of take-all does not seem to occur.
However the response to these avirulent species is localised and seems to depend on the relative inoculum level (Deacon, 1974) and extent of colonisation of the root. Speakman and Lewis (1977) have shown that a significant part of the interaction between Ggt and Phialophora graminicola
(Pg) depends on the ability of the latter to restrict the colonisation of
Ggt in the oortex and so prevent necrosis of the phloem. Thus, relative proportions of Ggt and Pg transferred from grass to wheat determine not only the outcome of direct competition between the two fungi, but also the host-parasite interaction throughout the root segment infected.
These workers later showed (1978) that despite differences in the levels of penetration of grass roots by Phialophora graminicola, Phialophora sp. with lobed hyphopodia and Ggt, all these fungi stimulated the lignificaticn and suberisation of the stele and inner tangential wall of the endodermis.
This was found to be true in maize roots also, even though fungi had only infected cells at some distance from the endodermis tangential walls.
However in wheat roots only P.graminicola and P.sp. with lobed hyphopodia showed stimulation of lignification and suberisation, and Ggt did not and these workers suggested that in wheat Ggt was prevented from spreading longitudinally in the stele following these induced changes in the host.
Cowan (1978) also found that colonisation of wheat roots with
Phialophora graminicola resulted in a qualitative change in the root lignin. Metabolic precursors of lignin, particularly caffeic acid, were shown to inhibit Ggt in Petri-dish culture, and Cowan suggested that caffeic acid could accumulate in roots infected with P.graminicola and so inhibit Ggt. 42
Another possible mechanism for cross-protection by these species has been proposed by Speakman and Lewis (1980) who have suggested that reduction in take-all brought about by P.graminicola results from competition between different isolates of the two species, both of which have a partial requirement for two vitamins, biotin and thiamine. Since roots of wheat plants excrete quantities of both these vitamins, and since a number of P.graminicola isolates have a relatively low requirement for the vitamins compared with Ggt, the avirulent fungus would have a selective advantage over Ggt should the vitamin content of the rhizosphere become a limiting factor.
Lemaire et al. (1975) demonstrated that protected roots of some wheat species produce a substance, diffusible in agar, which inhibits mycelial growth of aggressive isolates of Ggt. Sivasithamparam (1975), while demonstrating the ability of Phialophora species to cross-protect wheat seedlings against Ggt, screened the non-pathogenic isolates for their ability to produce a diffusible 'antibiotic1, when paired with the pathogen on agar plates. While P.malorum and P.fastigiata were very effective in this test, P.hoffmannii apparently was able to cross-protect the seedlings while producing no inhibitor on agar media. These inhibitors were unlike the volatile factor produced spontaneously in cultures on potato dextrose agar (Sivasithamparam et al., 1975) which induced transmissible lysis in Ggt.
Aims and approach to the investigation.
Isolates of phialophora graminicola, Phialophora species with lobed hyphopodia and Gaeumannomyces graminis var. graminis have been shown to contain viruses very similar in morphology to those of Ggt (Rawlinson and
Muthyalu, 1976). Since crossprotection of cereal roots by these species
is well documented and since the possibilities for the existence of
satellite dsRNAs coding for protein toxins, analogous to those of 43
Saccharomyces cerevisiae and Ustilago maydis, are the same as discussed for Ggt, an investigation was undertaken to determine whether such a system played a role in the cross-protection of cereal roots from take-all.
It was therefore decided to screen a number of virus containing isolates of the three species for the presence of dsRNA, and, from those containing a good yield of virus and several dsRNA components, to compare their RNA and polypeptide composition and determine the number of viruses present in each isolate. This would allow the selection of possible satellite dsRNA-containing viruses. It was also decided to investigate the interactions of these weakly pathogenic strains with Ggt isolates on artificial media to determine whether isolates with possible satellite dsRNAs produced diffusible toxins which could inhibit Ggt and thereby contribute to the mechanism through which these species cross-protect cereals from take-all.
This investigation of viruses in the Phialophora-Gaeumannomyces complex was considered to be valuable in furthering work on the control of fungal pathogens, in particular the take-all fungus, and in providing information on the biochemistry, taxonomy and ecology of the rapidly enlarging group of dsRNA mycoviruses. 44
ABBREVIATIONS
Buffers are abbreviated:
Ac buffer 0.2 M-sodium acetate titrated with 0.5 M-acetic
acid to pH 5.0 and diluted to 0.1 M-acetate
Bor buffer : 0.0125 M-disodium tetraborate, pH 8.6
CP buffer 0.1 M-citric acid titrated with 0.2 M-disodium
hydrogen phosphate to the required pH
P buffer 0.03 M-sodium phosphate, pH 7.6
PE buffer : 0.01 M-sodium phosphate, 0.005 M-EDTA, pH 7.0
PK buffer 0.03 M-sodium phosphate, 0.15 M-potassium
chloride, pH 7.6
PN buffer 0.03 M-sodium phosphate, 0.15 M-sodium chloride,
pH 7.6
SET buffer : 0.1 M-sodium chloride, 0.001 M-EDTA, 0.05
M-Tris HC1, pH 6.9
SSC buffer 0.15 M-sodium chloride, 0.015 M-sodium citrate,
pH 7.0
STE buffer 0.1 M-sodium chloride, 0.005 M-Tris HC1, 0.001 M-
EDTA, pH 6.85
STM buffer : 0.1 M-sodium chloride, 0.01 M-Tris HC1, 0.01 M-
magnesium chloride, pH 7.3
TAE buffer : (concentrated buffer) 0.04 M-Tris, 0.2 M-sodium
acetate, 0.0013 M-EDTA, pH 8.0
(TAE running buffer 1:25 dilution? 2 x TAE 45
running buffer 1:12.5 dilution)
TBE buffer : 90 mM-Tris, 0.1 M-boric acid, 2.5 mM-EDTA, pH 8.3
TK buffer : 0.05 M-Tris HC1, 0.15 M potassium chloride, pH 8.0
1NE buffer : 0.05 M-Tris Hcl, 0.15 M sodium chloride, 0.1 mM
EDTA, pH 7.9
TSE buffer : 0.1 M-Tris, 0.1 M-sodium chloride, 0.01 M-EDTA,
pH 7.0
Ver buffer : 0.05 M-sodium diethyl barbiturate, pH 8.6
Other abbreviations used are:
Gga Gaeumannomyces graminis (Sacc.) v. Arx and
Olivier var avenae (E.M. Turner) Dennis
Ggg Gaeumannomyces graminis (Sacc.) v. Arx and Olivier var. graminis
Ggt Gaeumannomyces graminis (Sacc.) v. Arx and
Olivier var. tritici (Walker)
Pa Phialophora graminicola (Deacon )Walker.
Phialophora species with lobed hyphopodia. P.sp. (lh)
double-stranded RNA dsRNA
single-stranded RNA ssRNA
deoxyr ibonuclease DNAse
ribonuclease RNAse
polyacrylamide gel electrophoresis p.a.g.e. 46
CsCl caesium chloride
DMSO dimethyl sulphoxide
EDTA ethylenediaminetetracetic acid
LiCl lithium chloride
Butyl PBD 2-(4'-t-butyl phenyl)-5-(4'-biphenylyl)-1,3,4
oxadiazole
PDA potato dextrose agar
PEG polyethylene glycol
PFO 2,5 diphenyl oxazole
SDS sodium dodecyl sulphate
TCA trichloroacetic acid
VLP virus-like particle sp.act. specific activity. 47
MATERIALS
Glass distilled water was used throughout.
Materials were obtained as follows:
Coomassie Brilliant Blue R from the Sigma Chemical Co. (St.Louis,
Missouri, U.S.A.). Toluidine blue from Hopkin and Williams (Chadwell
Heath, Essex, England); Ethidium bromide frcm BDH Biochemicals (Poole,
Dorset, England); DEAE cellulose and CF11 cellulose from Whatman
Laboratory Sales (Maidstone, Kent, England) and SP Sephadex from
Pharmacia (G.B.) Ltd. (Uxbridge Road, London, England); JH-UTP from the Radiochemical Centre (Amersham, Bucks, England); unlabelled nucleoside triphosphates from P-L Biochemicals (Milwaukee, Wisconsin, U.S.A.) and
Boehringer-Mannheim (Uxbridge Road, London, England); actinomycin D from
Merck, Sharp and Dohme (West Point, Pennsylvania, U.S.A.); RNAse A and
DNAse 1 (RNAse free) frcm Worthington Biochemical Corporation (New
Jersey, U.S.A.); acrylamide and N.N'-methylene bis acrylamide from BDH
Biochemicals (Poole, Dorset); PDA, Purified agar (L28) and malt extract
(L39) from Oxoid Ltd. (London, England). 48 GENERAL METHODS
G.M.I. ORGANISMS AND MICROBIOLOGICAL METHODS
Cultures were grown at 24°C and maintained at 4°C on unbuffered PDA slants in screw-top vials, loosely capped to provide aerobic conditions.
Subcultures were made at 4 to 6 monthly intervals.
PDA was prepared by suspending 39g dehydrated medium in 1,000 ml. distilled water and autoclaving at 15 lb/in2 for 15 min (121°C).
Cultures were grown in liquid media (100 ml medium in 500 ml conical flasks) by inoculating with 3 ml of mycelial suspension (in sterile physiological saline) obtained from PDA slant cultures. These cultures were shaken at 200 rev/min. (4.5 cm throw) at 24°C for 3 to 142 days.
Three liquid media were used for preparation of the cultures for nucleic acid or virus preparation. Each medium was sterilised by autoclaving at 15 lb/in2 for 20 min (121°C):
1. Basal and corn steep liquor. (Banks, Buck and Fleming, 1971).
% w/v
Glucose monohydrate 1.00
KH2PO4 0.10
MgS04. 7H20 0.05
FeS04. 7H20 0.001
Corn steep liquor 3.00
The pH was adjusted before sterilisation to pH 6.0 with sodium hydroxide; after sterilisation the pH was 5.5 to 6.0. 49
2. 2% Malt Extract
% w/v
Dehydrated malt
extract (Oxoid L39) 2.00
3. Weste and Throwers liquid Medium (Weste and Thrower, 1963)
% w/v
Glucose monohydrate 1.00
L-Asparagine 0.20
KH2PO4 0.10
MgS04. 7H20 0.05
FeS04. 7H20 0.0001
MnCl2. 4H20 0.0004
ZnSO„. 7H20 0.0001 4 Biotin 5.0 x 10
Thiamine 1.0 x 10"
The pH was adjusted, before sterilisation, to pH 6.0 with sodium hydroxide.
For small scale preparations of virus or nucleic acid, mycelia were harvested from liquid culture by filtration.
For large scale fermenter preparations the primary culture flasks were used to inoculate secondary cultures (1 litre in 3 litre conical flasks), which were grown at 24°C for 3 to 4 days and were then in turn used as inoculum for 30 litre fermenters. Fermenters were maintained at 2 an air pressure of 15 lb/in with agitator shaft speeds (183 to 367 rev./min.) and air flow rates (30 to 60 litres/min) adjusted to facilitate heat transfer and culture aeration. Mycelium was harvested after one week by filtration through a filter press. 50
Cultures were tested for incompatibility (G.M. 23) on three solid media buffered between pH 3.0 and pH 8.0 in 0.5 pH unit steps with CP buffer. The three solid media used were:
(a) PDA (3.9% w/v).
(b) glucose/asparagine medium (Lilly and Barnett, 1951).
% w/v
Glucose 1.00
L-Asparagine 0.20
KH2PO4 0.10
MgS04. 7H20 0.05
Fe2(S04)3 0.02
ZnS04. 7H20 0.02
MnS04 0.01
Biotin 5.0 x
Thiamine 1.0 x
Purified agar 2.00
glucose/urea medium (Haskins,
% w/v
KH2PO4 0.10
MgS04. 7H20 0.04
FeS04. 7H20 0.003
Yeast extract 0.06
Urea 0.06
Sucrose 5.00
Purified agar 2.00
Equal volumes of twice concentrated medium and 2xCP buffer, at the 2 appropriate pH, were autoclaved separately at 15 lb/in and 121°C for
15 min. The solutions were mixed aseptically and poured into Petri dishes.
The stock cultures on unbuffered PDA were used as a source of inoculum for these tests. 51
A solid medium for discrimination between living and dead hyphae was prepared by incorporating 0.003% w/v methylene blue into Lilly and
Barnetts1 solid medium buffered to pH 4.0.
G.M.2 ISOLATION AND PURIFICATION OF VIRUS a. Small scale preparation of crude virus
Fungal mycelium was grown in liquid culture as described in G.M.1, and was collected by filtration. Filtered mycelium was treated in one of two ways: i. Mycelium was homogenised by passage twice through the
Pascall rowmill (The Pascall Engineering Co. Ltd., Crawley, Sussex,
England) and homogenised mycelium was collected in 20 ml/g wet weight of
P buffer at 4°C. The degree of lysis of mycelium was checked under the
Leitz Wetzlar light microscope (E. Leitz (Instruments) Ltd., Luton,
England) (magnification x 25). ii. Filtered mycelium was suspended in 3 to 4 ml/g wet weight of P buffer at 4°C and was homogenised by passage twice through the A.P.V. Manton-Gaulin homogeniser (A.P.V. Co., Crawley, 2 Sussex, England) at 7,500 lb/in pressure. The degree of lysis of the mycelium was determined using the Leitz microscope as before.
Insoluble debris was removed from the suspension of homogenised material by centrifugation. Sodium chloride was added to the supernatant and dissolved with stirring, followed by PEG 6,000 (30% w/v in P buffer) added dropwise with stirring, to bring the final concentrations of these compounds to 1M and 10% w/v respectively. The precipitate formed overnight was collected by centrifugation at 23,000 g for 30 min. and was then resuspended in P buffer (1ml/g wet wt. fungal material) by stirring for 4 to 5 h. Insoluble material was removed by centrifugation at 30,000 to
38,000 g. Virus pellets were resuspended in P buffer (volume as required) and were given a short cleaning spin (26,000 g for 15 min). b. Large scale preparation of crude virus
Mycelia, grown in fermentation culture as in G.M.1., were harvested in a filter press and resuspended in 20 litres of P buffer at 4°C and 52 were homogenised by passage through the A.P.V. Manton-Gaulin homogeniser as before. The homogenate was further diluted with P buffer and checked for lysis under the Leitz microscope. Insoluble debris was removed by centrifugation in a Sharpies Super centrifuge (Sharpies Centrifuges
Ltd., Camberley, Surrey, England), at 13,200 gav and a flow rate of 4 litres/min. The supernatant was adjusted to 1.0 M sodium chloride with solid salt and to 10% w/v PEG 6,000 by addition of 30% w/v PEG dropwise, with stirring for 2 to 3 h. The suspension was left to precipitate overnight. The precipitate formed was collected through the Sharpies centrifuge and was resuspended in P buffer, left to stir for several hours at 4°C, and was given a cleaning spin at 23,000g for 1h. Virus was pelleted from the supernatant overnight at 30,000 to 38,000 g. The pellets were resuspended in chilled P buffer (1 to 3 ml). The crude virus suspension was given a further clarifying spin. Further purifi- cation of crude virus preparations are noted in the next method. c. Purification of large scale virus preparations by sucrose
density centrifugation
Crude virus samples (1,000 to 1,500 A260 units) were purified by zonal preparative density gradient ultracentrifugation in an MSE
Superspeed 65 centrifuge (MSE, Manor Royal, Crawley, Sussex, England). A sucrose density gradient of 20% to 50% w/v sucrose in P or PN buffer was loaded into a B XIV rotor at 3,000 rev/min loading speed, followed by the virus sample into the centre of the rotor. The preparation was centrifuged at 47,000 rev/min for 3 h. and the gradient was pumped out of the rotor with 55% w/v sucrose in P or PN buffer at the loading speed. The gradient was scanned at 254 nm with the ISCO Model D density gradient fractionator
(Instrument Specialities Inc., Lincoln, Nebraska, U.S.A.). Alternatively,
2 to 10 ul samples of each fraction were spotted on to an agarose gel
(0.5% w/v) plate containing 2 ug/ml ethidium bromide. On irradiation with U.V. light (254 nm) fluorescence of the nucleic acid-ethidium 53 bromide complex was used to locate the virus sample in the fractionated gradient. Distribution of dsKNA and polypeptide virus components across the gradients was determined by p.a.g.e. of the fraction samples as in
G.M.16. d. Purification of small scale virus preparations
Small samples (1.0 ml, A26q = 1.0 to 5.0) of crude virus were purified by sucrose density gradient centrifugation in 35 ml gradients of
10 to 45% w/v sucrose (in P or PK buffer) at 24,000 rev/min in the Spinco
SW25 rotor.
Smaller samples than this (0.1 ml, A26q = 0.5 to 1.5) were centrifuged in 5 ml gradients of 10 to 45% w/v sucrose in buffer, at
45,000 rev/min in the Spinco SW 50.1 rotor. Larger gradients were fractionated in the ISCO Model D fractionator as before, while the small gradients were fractionated in a Buchler gradient fractionator (Buchler
Instruments Inc., Port Lee, New Jersey, U.S.A.) surrounded by ice.
Fractions were analysed for the presence of virus with agarose - ethidium bromide plates, and for distribution of virus components by p.a.g.e.
G.M.3. ELECTRON MICROSCOPY
A few drops of the virus sample were mixed with an equal volume of negative stain (2% w/v potassium phosphotungstate, pH 7.0) as described by Brenner and Home, 1959. A drop of a 0.2% w/v solution of bovine serum albumin was added to spread the sample on the grid. The mixture was poured into a high velocity spray gun (Aerograph Super 63) and droplets were sprayed (from a distance of 30 cm) onto the surface of 200 mesh copper grids coated with collodion (stabilised with carbon). Virus samples were examined in a Siemens 1A Elmiskop electron microsope. The electron microscope was calibrated by direct internal calibration as it possessed a suitably long focal length. The microscope was calibrated 54 externally with tobacco mosaic virus. Photographs, at a magnification of x 100,000 were taken with Ilford EM plates.
G.M.4. SPECTROPHOTOMETRY
Ultraviolet spectra were measured with a Cary 15 recording spectro- photometer (Applied Physics Corp., Moravia, California, U.S.A.) or with a
Gilford spectrophotometer (Gilford Instrument Laboratory Inc., Ohio,
U.S.A.) for point readings, using silica cells of 1 cm path length. An
absorbance unit (A26QU^ is defined as the absorbance at 260 nm multiplied by the sample volume in ml. An approximate correction for light scattering was made by subtraction of the reading at 350 nm.
Polyacrylamide gels of nucleic acid samples stained in toluidine blue (0.01% w/v) were scanned (at 550 nm, 0 to 2.00 absorbance range) using the scanning attachment of the Gilford spectrophotometer. Relative proportions of nucleic acid species in the sample were determined from the area under the trace of each peak.
G.M.5 PREPARATION OF ANTISERA
Preimmune sera were removed from rabbits prior to immunisation with virus. Antisera were raised to pure virus preparations in rabbits
by injection of 1.5 to 2.0 ml of virus (A2^q = 0.60^ in PN buffer, according to the following immunisation schedule. An intravenous injection of 0.8 ml virus preparation was given first, followed after 9 to 10 days by an intramuscular injection of virus preparation (0.5 ml) homogenised with an equal volume of Freund's incomplete adjuvant, into the hind leg of the rabbit. A second intramuscular injection was given after an additional two weeks. Serum was collected 2 to 3 weeks after the last injection.
After collection the blood was allowed to clot at 37°C for 2 h and 55 was then cooled at 4°C overnight and centrifuged to remove the fibrin and blood cells the following day.
Antisera were stored liquid at 4°C with added glycerol (50% v/v) as bo^cb^rCos&xt -
G.M.6. OUCHTERLONY GEL IMMUNODIFFUSION TEST
Plates for double diffusion serological analysis were prepared with solutions of ion agar no.2 (0.5 to 1.0% w/v) or agarose (1.0% w/v) in PB buffer, boiled to dissolve the matrix. When the solution had cooled to
50 to 60°C sodium azide was added to 0.01% w/v final concentration, and 2 to 3 ml of the solution was poured onto each microscope slide (3" x
1") using a Pasteur pipette. The plates were allowed to set for more than 1 h, after which time wells (5 mm diameter) were cut in the gel.
The agar discs were removed by vacuum and the wells were sealed with a solution of ion agar no.2 (0.2% w/v) cooled to 55°C. Up to 20 ul of antiserum or antigen could be loaded into a well of 0.1 cm depth.
Dilutions of antiserum or antigen were made with physiological saline
(0.85% w/v sodium chloride). The plates were allowed to develop for at least 48 h (in a Petri dish containing moist filter paper). Slides were washed in P buffer overnight (several changes ) and were then stained in
Cocmassie blue/methanol/ acetic acid/ distilled water (0.1% w/v in
45:10:45 solvent) for 1 to 2 h and were then destained in the same solvent mixture. (The staining procedure greatly increases the sensitivity of location of the preciptin lines).
G.M.7. PREPARATION OF DEAE CELLULOSE
DEAE cellulose DE52, was suspended in a dilute solution of phosphoric acid (pH 2.0). The suspension was evacuated to remove carbon dioxide and the pH was brought to 7.6 with sodium hydroxide. After 10 min this 56 buffer was decanted and the ion-exchanger was washed with several changes of P buffer until the pH of the filtrate was the same as that of the washing buffer.
G.M.8. PREPARATION OF CF11 CELLULOSE
CF11 cellulose was suspended in 10 volumes of SET buffer, allowed to stand overnight and was rewashed several times with fresh buffer prior to use.
G.M.9. PREPARATION OF SP C-50 SEPHADEX
SP Sephadex cation exchanger was suspended in an excess of Ac buffer. After 10 min this buffer was decanted and the Sephadex was rewashed and allowed to swell at room temperature for 3 days in the same buffer.
G.M. 10 ANALYTICAL ULTRACENTRIFUGATION
Analytical ultracentrifugation was performed in a Beckman Model E ultracentrifuge equipped with a monochromator and double beam ultraviolet absorption optical system with photoelectric scanner. a. Determination of sedimentation coefficient of virus
Sedimentation coefficients were determined using a cell with a 12 mm, 2.5° aluminium filled Epon double sector centrepiece and quartz windows, in the AN-H rotor at 12,000 or 16,000 rev/min in a Beckman
analytical ultracentrifuge. Virus samples (A2gQ = 0.20 to 0.90) in P buffer (0.45 ml) were used and detection of the moving boundary was made with the ultraviolet scanner at 265 nm. Cells were scanned at 4 min intervals over a period of 50 min. 57
By definition sedimentation coefficient, S = velocity (dx) (dt) per unit centrifugal field (xw 2 ) i.e• . dx = xw 2 s dt where x = distance of the moving boundary frcm the centre of rotation
(cm) at time t (sec), x is calculated frcm the scan (a typical
scan is shown in Fig. 1) by the equation:
reference point to the middle of the boundary + 5.70 cm magnification factor
The magnification factor is the ratio of the traced distance
between the inner edges of the reference holes to the actual
distance (1.6 cm as specified by the manufacturer).
w = angular velocity (radians/sec).
Integration of this equation gives: 2
In x = w s t + constant.
The sedimentation coefficient was found by plotting In x vs. t. 9 The gradient of the slope = w s. At the low concentrations of virus used, the value of s was taken to approximate to s° (Shumaker and
Schachman, 1957). S°2( qn .wa )s calculated from the equation. S° 20 = q°
< n20} where = viscosity of water at the temperature of the
determination
n20 = vi*3008*^ water at 20°C b. Determination of buoyant density of virus
i. Samples of virus (0.1 ml, A26Q = 0.90) were analysed for buoyant density by equilibrium centrifugation in CsCl solutions (in PK 58
! ! r: -1—1-—
! i i I I -i—c cell bottom Figure 1 Typical photoelectric scanner trace for the determination of sedimentation coefficient 59 buffer, 5 ml) at 40,000 rev/min in the Spinco SW 50.1 rotor for 60 h. Caesium chloride solutions of approximately 1.40 g/ml density were used, and after oentrifugation, the density gradient was fractionated (0.2 ml fractions). The refractive index distribution across the gradient generated in the centrifuge was determined and the location of the virus sample in the gradient found by agarose-ethidium bromide plate analysis (G.M. 2c). This method provided an estimate for the buoyant density of virus particles and more accurate analysis of density was made by method ii. ii. Equilibrium density centrifugation in the Beckman Model E analytical ultracentrifuge was used to determine buoyant density of virus particles. 12 mm, 2.5° double sector, charcoal filled centrepieces were employed and virus preparations (^SO to "®"n P buffer, containing caesium chloride of average density 1.365 to 1.425 g/ml were centrifuged for 18 to 24 h at 30.000 to 24,000 rev/min in the AN-F rotor. Scans of the cells were made using the ultraviolet scanner at 265 nm and the Multiplexor accessory. Average densities of the caesium chloride solutions were calculated from their refractive indices as shown in G.M.11. Buoyant density was determined for a single virus or a mixture of two viruses by banding these at two or three different positions in the cell by adjustment of the initial caesium chloride density. After centrifugation to equilibrium, scans were taken and the magnification factor (Mf) was determined as for the sedimentation coefficient analysis. To calculate the buoyant density of the virus particles in the cell the following values were determined: (see Fig. 2. for a schematic representation of a cell containing two virus bands). K = centre of cell cavity F = distance from top to bottom of cell cavity (cm) 60 Figure 2 Schematic representation of a trace used for the determination of buoyant density 61 Rm = inner reference hole, 5.70 cm from centre of rotation (inside edge) Rg = outer reference hole, 7.30 cm from centre of rotation (inside edge) R = distance between inside edges of reference holes (cm) P = distance between bottom edge of cell cavity and inside edge of outer reference hole (cm) M = meniscus of solvent N = distance between bottom edge of cell cavity and meniscus M (cm) D = distance between virus bands X and Y (cm) A = midpoint between bands X and Y E.j = distance between A and cell centre K (cm) C = point of isoconcentration (centre(mean) between squares of the distances from the centre of rotation to the cell bottom rB and to meniscus rM). H = distance between band X and C (cm) = distance between cell centre K and midpoint between X and C (cm) G = distance between bottom edge B of cell cavity and point C Mf = R (cm) the magnification factor for the scan 1.60 62 The procedure for calculation of buoyant density was as follows: 1. The relative length __F of each centrepiece and the distance of Mf the bottom of the cell from the centre of rotation (7.30 - P_) were Mf determined. 2. C was determined as the centre between the squares of the distances from the centre of rotation to the cell bottom rg and 2 to the meniscus rM: rB2 + rM 2 3. G was found for the measured distance N by the equation: 7.3 Mf-P-G = (7.3 Mf-P)2 + (7.3 Mf-P-N)2 2 4. Distances H were measured for the scans made. These were corrected for the relative increment incurred in the distance between the two bands due to their position in the cell (the density gradient as a function of the radius becomes steeper near the bottom of the cell) using a correction factor k (frcm a diagram giving known values of k relative to the density gradient distribution in the cell): Ho = H + kH This gives values of HQ' and H0" for two cells (W. Szybalski, 1968). 5. The actual density gradient a was then calculated: a = Po" - Po1 H ' - H " o o and the density of the virus found from: P = PQ. + aH0' (It was assumed that the buoyant density at point C 63 corresponded to that determined by refractive index for the total caesium chloride solution. This should give identical values of a and p at different positions in the cell). For determination of the buoyant density of a virus in relation to a reference peak of known density, the measured distance between A (centre point between the two bands) and cell centre K (= E^ )f was corrected for the distance above or below K (correction factor k) and the density of the unknown virus found from: pM = Kp + aD o — o where p _ is the density of the reference band and D is the o o corrected distance between the two bands. G.M. 11. DETERMINATION OF DENSITY OF CAESIUM CHLORIDE SOLUTIONS Density of caesium chloride solutions was determined refracto- metrically (Brakke, 1967). The refractive indices of samples were measured using the Abbe High Accuracy "60" refractometer (Bellingham and Stanley, Ltd., London N.5. England) provided with a sodium D^ line source (Gallenkamp and Co. Ltd., London E.C.2. England). The refractometer was calibrated with a quartz block of known refractive index (1.54422). The refractive index was converted to density using the expression: _ a „ 25°C , 25°C ~ a °D D where 250^ = density of the solution at 25°C 25 °C r^ = refractive index of the solution at 25°C at the wavelength of the sodium D^ line (589.6 nm) a and b are coefficients which for the density range 1.25 to 1.90 g/cm3 are: a = 10.8601 b = 13.4974 64 G.M.12. NUCLEIC ACID EXTRACTION a. Preparation of total nucleic acid from mycelia Mycelium was grown as described in G.M.1., filtered from liquid culture, homogenised in the Pascall rowmill and resuspended in chilled TSE buffer. SDS was added to the homogenate to bring the final concentration to 1% w/v, and the mixture was stirred at 4°C for 10 min. The hcmogenate was then extracted twice by addition of 90% v/v phenol containing 9.9% v/v TSE buffer and 0.1% w/v 8-hydroxyquinoline (100ml). The emulsion was stirred for 15 min and the two layers were separated by centrifugation at 18,000 g for 10 min. The twice extracted aqueous layer was made 0.2 M with respect to sodium acetate and two volumes of absolute ethanol, chilled to -20°C, were added. After storage at -20°C overnight, the nucleic acid precipitate was collected by centrifugation at 23,000 g for 15 min and was then resuspended in the required buffer and dialysed against this buffer (usually SIM buffer). b. Preparation of virus RNA RNA from purified virus preparations was prepared essentially as described by Marcus et al. (1974). Virus suspensions were made 5 mM with respect to EDTA and 0.5% w/v with respect to SDS and were heated to 60°C for 20 min. The suspension was extracted with EDTA - washed phenol (prepared as in G.M. 14), once at 60°C and twice at room temperature. Aqueous and phenol layers were separated by low speed centrifugation (8,000 g, 15 min). Potassium acetate solution (pH 5.5) was added to 0.1 M concentration and the RNA was precipitated by addition of 2.5 vol. of ethanol at -20°C and storage for 18 h at this temperature. The RNA precipitate was collected by centrifugation, dissolved in buffer and dialysed against 65 buffer prior to storage at -20°C. Occasionally RNA was not precipitated with ethanol and RNA in aqueous phase was dialysed directly against 6 changes of distilled water. G.M.13. ACTION OF NUCLEASES ON TOTAL FUNGAL NUCLEIC ACID AND VIRAL RNA a. Action of DNAse I Nucleic acid samples (10 to 50 ug) in SIM buffer were incubated with 1Oug/ml RNAse free DNAse 1 for 2 h at 37°C. Samples were then made 1% w/v with respect to SDS and were heated to 60°C for 20 min. Solutions were loaded onto standard polyacrylamide gels and were analysed by electro- phoresis as in G.M. 16. c.i. DNAse 1 activity was tested by incubation with DNA and examination by p.a.g.e. b. Action of RNAse A Nucleic acid samples (2 to 10ug) in 0.1 x SSC or 2 x SSC buffer were incubated with 0.2 and 2.0ug/ml RNAse A at 24°C for 2 h. Samples were then made 1% w/v with respect to SDS and were heated to 60°C for 20 min. Solutions were examined by p.a.g.e. on semi micro gels as in G.M.16. c.i. RNAse A activity was tested by incubation with MS2 ss RNA and examination by p.a.g.e. G.M. 14. PREPARATION OF EDTA-WASHED PHENOL EDTA-washed phenol for use in nucleic acid extractions was prepared as described by Marcus et al.(1974). Solid phenol was melted in water (500 g in 200 ml) at 30°C. To the suspension 0.1M EDTA, pH 7.4 (50ml) and 1N KOH (8ml) were added. The mixture was stirred at room temperature for 10 min and was then allowed to separate. The upper layer was removed 66 by aspiration and distilled water (400 ml) was added to the phenol phase. The suspension was stirred for 2 min. at room temperature and, after standing in the dark for 30 min, the upper layer was again removed by aspiration. The water wash was repeated and 0.05 M EDTA, pH 7.4 was added to the separated phenol phase. After stirring for 10 min. and storing in the dark for 30 min, the upper layer was removed by aspiration. The EDTA-washed phenol was stored at 4°C (in a dark brown bottle). G.M. 15. PREPARATION OF DIALYSIS TUBING Sections of dialysis tubing were prepared by boiling in six successive changes of 0.1% w/v sodium bicarbonate, rinsing between each change with cold distilled water. The tubing was then rinsed once with absolute ethanol and boiled twice in 1mM EDTA, with cold water rinsing as before. After a final boiling in EDTA the tubing was allowed to cool in the solution and was stored at 4°C. Pieces were rinsed in the required buffer before use. G.M.16. ANALYTICAL GEL ELECTROPHORESIS a. Agarose gel electrophoresis for analysis of intact virus Agarose (0.5% w/v) in TK buffer was boiled for 5 min. and, after cooling to between 50 and 55°C was poured either into glass electro- phoresis tubes (8.0 x 0.5 cm) sealed with parafilm at the lower end, or onto a horizontal plate fitted with a slotted comb to produce a slab gel (10 x 10 x 0.5 cm). Occasionally ethidium bromide was added to the gel solution just before dispensing, to give a final concentration of 2 ug/ml. In the case of tube gels, when the gels had set, nylon mesh was secured over the top of the tubes and the gels inverted, so that a flat surface was provided at the upper end. For disc gels, virus samples (50 ul, A26Q = 0.3 to 1.0) containing 10% w/v sucrose and 2 ul saturated bromophenol blue as electrophoresis marker, were electrophoresed in TK buffer at 6 mA/tube for 1 to 2.5 h. Gels containing ethidium bromide were examined under U.V. light (254 nm) for fluorescence. Other gels were stained with 0.1% w/v Coomassie blue R containing 10% w/v 5-sulphosalicylic acid and 10% w/v trichloroacetic acid for 1 to 3 h and were destained in 5% w/v acetic acid. Similar volumes and concentrations were loaded into the slots in the slab gel, and these were electrophoresed at 50 V for 3 to 5 h, until the bromophenol blue tracker dye had reached the end of the gel. Gels were stained and analysed as above. b. Agarose gel electrophoresis for analysis of nucleic acid Agarose (1.25% w/v) in TBE buffer, was boiled for 5 min, and, after cooling to 50 to 55°C, ethidium bromide was added to a final concentration of 2 ug/ml. The mixture was poured into a horizontal slab apparatus as in G.M. 16 a. Nucleic acid samples (30ul, = 80) containing 1% w/v SDS, bromophenol blue and sucrose (10% w/v), were loaded into the slots through TBE running buffer (containing 2 ug/ml ethidium bromide) and were electrophoresed at 15 V for 16 h. Nucleic acid components were visalised under U.V. light (265 nm, long wavelength for preservation of intact dsRNA). c. Polyacrylamide gel electrophoresis for analysis and molecular weight determination of nucleic acid i. For dsRNA analysis 4% w/v polyacrylamide gels were prepared as follows: 20 ml of gel solution was made up containing : 2.0 ml of 39.6% w/v acrylamide containing 0.4% w/v methylene bis acrylamide, 2.0 ml of 1% (v/v) N,N,N',N'-tetra- methylethylenediamine (TEMED), 2.0 ml of 1.5% w/v ammonium persulphate, 1.6 ml of 25 x TAE buffer and 12.4 ml degassed distilled water. The mixture was cooled to 4°C, degassed for 2 min and poured into glass 68 tubes (standard 8 cm x 0.5 cm or semi micro 8 cm x 0.25 cm) sealed at the lower ends with parafilm. Water was layered on top of the gels with a finely drawn out Pasteur pipette, before the gels had set, to provide a flattened upper surface to the gel. The gels were preelectrophoresed for 0.5 h (semi micro gels) or 1 h (standard gels) in 2 x TAE buffer. Virus preparations were dissociated by addition of SDS (1% w/v final concentration) and incubation at 60°C for 20 min. Samples containing bromophenol blue and sucrose (RNAse free) were loaded onto the gels (50 ul maximum on standard gels, 12 ul maximum on semimicro gels). Samples were electrophoresed in 2 x TAE for 2 to 7 h for standard gels and 0.5 to 2 h for semimicro gels. RNA was visualised by staining gels with 0.01% w/v toluidine blue containing sodium azide (0.01% w/v). Molecular weight determinations were made by coelectrophoresis of nucleic acid samples with dsRNA markers. The molecular weights of the c dsRNA markers were: Penicillium chrysogenum virus 2.21 x 10 , 2.08 x 6 6 10 and 1.98 x 10 (Wood and Bozarth, 1972); Penicillium stoloniferum virus S: 1.10 x 10 6 and 0.94 x 10 6 (Buck and Kempson-Jones, 1973); 6 6 Aspergillus foetidus virus S: 4.1 x 10 and 2.6 x 10 ; A. foetidus 6 6 6 6 virus F: 2.9 x 10 , 1.87 x 10 , 1.70 x 10 and 1.44 x 10 c (Buck and Ratti, 1977); Helminthosporium maydis virus: 6.3 x 10 . (Bozarth, 1977). Molecular weights were detemined frcm plots of log molecular weight versus mobility (cm) (Bozarth and Harley, 1976). ii. For dsRNA and ssRNA analysis 4% polyacrylamide gels were prepared as follows. 20 ml of gel solution was made up containing : 15.0 ml 10M urea (filtered or A.R.) containing 1% w/v SDS, 2.0 ml 40% w/v acrylamide containing 1.1% w/v bis acrylamide, 1.0 ml 2% w/v TEMED, 1.2 ml 3% w/v ammonium persulphate and 0.8 ml 25 x TAE buffer. The mixture was made up on ice, degassed under 69 vacuum and poured into standard gel tubes (8 x 0.5 cm) or slab apparatus (10 x10 cm) and the surface was overlayed with 5 to 8 M urea/TAE/0.1% w/v SDS in the standard gels, and a slotted comb was inserted into the top of the gel solution in the slab. Gels were allowed to set overnight and were then prerun for 1 h in TAE/0.1% w/v SDS running buffer. Samples (dsRNA, RNA polymerase product or glyoxalated dsRNA) were prepared for electrophoresis by addition of an equal volume of 10M urea containing 0.2% w/v SDS in TAE buffer, and were loaded onto the gel after addition of sucrose (to 10% w/v) and bromophenol blue as tracker dye. Samples were layered over with 5 M urea containing 0.1% w/v SDS in TAE, and were subjected to electrophoresis in TAE containing 0.1% w/v SDS for 5 to 7 h at 6mA/tube for standard gels and for 15 h at 60V for slab gels. Gels were rinsed twice in distilled water (0.5 h each) and were stained overnight in 0.01% w/v toluidine blue and were destained in distilled water. d. Polyacrylamide gel electrophoresis for analysis and molecular weight determination of polypeptide 8% w/v polyacrylamide gels were prepared as follows. 20 ml of gel solution was made up containing: 40 ml of 39.6% w/v acrylamide containing 0.4% w/v methylene bis acrylamide, 2.0 ml of 1% w/v TEMED, 2.0 ml of 1.5% w/v ammonium persulphate, 2.0 ml of 1.0% w/v SDS and 10.0 ml of 0.2M sodium phosphate buffer containing 0.02M disodium EDTA, pH 7.2. Gels were made up as standard, semimicro or slab gels, as before, and were preelectrophoresed in running buffer (0.1 M sodium phosphate buffer, containing 0.01M disodium EDTA and 0.1% w/v SDS, pH 7.2. Virus samples were dissociated by addition of SDS (1% w/v final concentration) and 2-mercaptoethanol (1% w/v final concentration) and heating to 100°C for 3 min. Samples were loaded under buffer as before, and were electrophoresed 70 at 6tnA/tube for 0.5 to 1 h for semimicro gels, and 5 h for standard gels, and at 30 to 60 V for 15 h or more for slab gels. Polypeptides were located with Coomassie blue R stain (0.1% w/v) in 45% methanol, 45% water and 10% acetic acid and gels were destained in the solvent mixture alone. Molecular weight determinations were made by coelectrophoresis of polypeptide samples with protein markers. The following standards were used; phosphorylase a (rabbit muscle), subunit mol. wt. 100,000; bovine serum albumin, mol. wt. 68,000; lactate dehydrogenase (hog muscle), subunit mol. wt. 36,000 (Weber et al., 1972); glutamate dehydrogenase (bovine liver), subunit mol. wt 55,390 (Smith et al., 1975). Graphs were drawn of log molecular weight versus mobility (cm) and molecular weight of the test polypeptides estimated by means of least squares analysis. e. Combined analysis of intact virus and polypeptide Virus samples were subjected to electrophoresis in semimicro agarose gels as in G.M. 16 a. and 1 of a set of 3 identical gels was stained with Cocmassie blue R as before. The other two gels were sliced (0.2 cm slices) using the Mickle gel slicer (The Mickle Laboratory Engineering Co., Gomshall, Surrey) and gel slices were boiled with 1 ul 10% v/v 2-mercapto- ethanol, 1 ul 10% w/v SDS, together with bromophenol blue and sucrose at 100°C for 3 min. The solutions were allowed to cool to 50 to 55°C and were then loaded onto semimicro polypeptide gels (8% w/v polyacrylamide) and allowed to cool and set. Running buffer was layered on top of the gels and the samples were electrophoresed at 6mA/tube for 1 to 2 h in 0.1 M sodium phosphate buffer containing 0.01 M disodium EDTA and 0.1% w/v SDS, pH 7.2. Gels were stained in Coomassie blue R/methanol/acetic acid stain as before, for 0.5 h and were destained in methanol/acetic acid solution. 71 Molecular weights of the polypeptides were determined by coelectrophoresis with marker proteins as in G.M.16 d. G.M. 17. GLYOXALATION OF DSRNA DsRNA samples were denatured and glyoxalated by a modification of the method of McMasters and Carmichael (1977) (R. Empson, personal communication). Samples were incubated in a mixture containing 1.0 M glyoxal, 10 mM P buffer and 60% DMSO at 50°C for 1 to 1.5 h. Intact virus samples were glyoxalated immediately after disruption of particles with 1% w/v SDS at 60°C for 20 min. For electrophoresis, glyoxalated samples were loaded directly onto 8M urea, 0.75% SDS polyacrylamide gels (G.M.16. c. ii). G.M. 18. RNA-DEPENDENT RNA POLYMERASE ASSAY The reaction mixture used contained the following: Virus (A26q 0.25 to 5.0) 0.05 M-Tris HC1, pH 7.9 0.15 M-NaCl 0.10 mM-EDTA 0.15 mM-ATP 0.15 mM-CTP 0.15 mM-GTP 0.15 mM-[3H] UTP (sp. act. 17 to 68 mCi/mmole) 2.5 mM-or 5.0 mM-MgCl2 Bentonite 800 ug/ml Actinomycin D (C1) 125 ug/ml Samples were incubated at 30°C for up to 72 h. Bentonite was removed from the preparation by pelleting at 8,700 g in a Beckman microfuge 72 B (Beckman-RIIC Ltd., Cressex Industrial Estate, High Wycombe, Bucks). Incorporation of [^HJ-UTP into acid - insoluble material was determined by precipitation from 10 to 1,000 ul of reaction mixture with 3 ml of 10% w/v TCA at 0°C for 30 min. The precipitates were collected on glass fibre filters (Whatman GF/F) and were washed with 80 ml of 2% w/v TCA and then with ethanol. Filters were dried at 60°C for 1 h and radioactivity was determined by liquid scintillation counting following addition of 10 ml scintillation fluid (0.6% w/v Butyl PBD, 5% naphthalene in toluene) to each vial. G.M. 19. PREPARATION OF VIRUS SSRNA TRANSCRIPTS IN VITRO RNA polymerase reaction mixtures, following incubation at 30°C, were made 1% w/v with SDS and heated to 60°C to disrupt the virus particles and bentonite was removed by centrifugation in the Beckman microfuge (8,700 g, 5 min). The suspension was extracted with phenol as in G.M. 12 b. and the nucleic acid was precipitated with ethanol. After collection by centrifugation the precipitated nucleic acid was taken up in PE buffer and dialysed against this buffer. SsRNA was separated from the dsRNA by selective precipitation with LiCl according to the method of Baltimore (1966). RNA solutions were made 2M with respect to LiCl and incubated at 4°C for 24 h. The precipitate was collected by centrifugation, washed in 4M-LiCl and dissolved in PE buffer and dialysed against this buffer. Ethanol precipitation was repeated on this preparation once more and a second cycle of LiCl precipitation performed. G.M.20. RNA-RNA HYBRIDISATION ASSAY This was carried out by a modification of the method of Vandewalle and Siegel (1976). Virus dsRNA was denatured by heating at 100°C for 3 min in PE buffer and was allowed to anneal with radioactively labelled 73 ssRNA (Ratio dsRNA/ssRNA = 0.25 to 80) by increasing the salt concentration to 0.2M - sodium chloride by addition of 8 x STE buffer and "3 incubating at 70°C for 4 h. The amount of [ H] -dsRNA formed by annealing was determined by treating the annealing mixture with RNAse A (Ratio dsRNA/enzyme = 10 to 20) for 1 h at 37°C and precipitating the remaining labelled material with TCA (3 ml of 10% w/v solution added to 50 to 200 ul samples). Control samples were assayed at the same time: i. where ssRNA was added and samples were processed in the same way, but the RNAse step was omitted, and ii. where ssENA was added but was not processed through the annealing procedure, but RNAse was added. TCA precipitated samples were collected on glass fibre filter papers (Whatman GF/F), ethanol dried at 60°C for 1 h and counted in 10 ml scintillation fluid as in G.M.18. G.M.21. FLUQRQGRAPHY 3 Nucleic acids, radioactively labelled with [ H]-UTP as described in G.M.18. were detected, after separation on polyacrylamide gels, by fluorography caried out by the method of Laskey and Mills (1975). The gels after electrophoresis, either directly or after staining, were equilibrated with DMSO by immersion for two 30 min intervals in 20 times their volume of DMSO. Gels were then immersed in 4 volumes of 20% w/w PPO in DMSO for 3 h and were finally soaked in 20 x their volume of water for 1 h. The gels were dried onto 3 MM paper under vacuum and were placed in contact with Fuji RX film and stored for -70°C for 1 to 48 h. Films were developed by passage through the Kodak Industrial X Omat processor. G.M.22. SLICING AND SOLUBILISATION OF POLYACRYLAMIDE GELS FOR RADIOACTIVE ANALYSIS Radioactivity of nucleic acids, after separation by polyacrylamide gel electrophoresis (G.M.16 c.ii), was determined by scintillation counting after slicing and solubilisation of gels. Gels were frozen at -20°C, were kept frozen with crushed solid carbon dioxide and slices of regular thickness were made with the Mickle gel slicer, each slice being placed in a separate scintillation vial. These were heated at 60°C for 1 h to dehydrate the gel slices, which were then solubilised by soaking in 0.25 ml of hydrogen peroxide (30% w/w) containing 1% by volume of aqueous ammonia at 37°C overnight. 2-Methoxy ethanol (5ml) and scintillant (10 ml) consisting of 0.6% w/v butyl PBD and 5% w/v naphthalene in toluene, were added to each vial and samples were counted in a Beckman LS-230 scintillation counter. Graphs were plotted of cts/min against distance travelled down the gel. G.M.23. INCOMPATIBILITY TESTS OF FUNGAL ISOLATES Discs (6mm) of the pair of isolates under test were placed 2 to 2.5 cm apart in the centre of Petri dishes containing one of three buffered solid media, as described in G.M.1. The dishes were incubated in the dark at 24°C and developing cultures were observed microscopically at regular time intervals. Discs of the pairs of isolates were also plated onto medium containing methylene blue (G.M.1.) to determine the viability of the hyphae in the developing cultures. G.M.24. PRODUCTION OF INHIBITOR IN LIQUID CULTURE Mycelium was cultured in liquid medium to determine the inhibitory activity of culture filtrate. A mycelial suspension of fungal isolate in physiological saline (approximately 3 ml per flask)was made from stock 75 cultures on PDA slopes and was added to a flask of 2% w/v malt extract (100 ml) containing 0.1 mg litre thiamine hydrochloride and was shake cultured at 24°C for one to two weeks. The primary mycelium was first filtered from the suspension, gently homogenised in a Sorvall Omnimix blender (Dupont (U.K.) Ltd., Hitchin, Herts., England; low speed, 10 sec) and used to inoculate 2 x 100 ml secondary cultures of the same medium or Lilly and Barnetts' liquid medium buffered to pH 4.0 with CP buffer. These were shake cultured for up to 30 days. Samples of culture filtrate (5 ml) were removed at various time intervals and were sterilised by passage through several layers of sterile filter paper and exposure to bacteriocidal ultraviolet light (95% of radiated energy is at 253.7 nm wavelength, near maximum for germicidal effectiveness) for 30 min. Culture filtrates were stored at 4°C. G.M.25. WELL TESTS FOR INHIBITORY ACTIVITY OF CULTURE FILTRATES Isolates under test for sensitivity to culture filtrate were grown on Lilly and Barnetts1 solid medium (buffered to pH 4.0 with CP buffer) to a diameter of 2 to 3 cm. Pour wells, 6 mm in diameter and approximately 1 cm apart, were cut on one side of the colony at 0.5 cm from the edge of the colony. Wells were filled with the culture filtrate being tested and dishes were incubated at 24°C for 5 to 7 days. Wells were each refilled with the same sample of culture filtrate twice a day for 3 days. G.M.26. TESTS FOR SENSITIVITY OF INHIBITOR TO PROTEASES Petri dishes containing Lilly and Barnetts' solid medium, incorporating Millipore (0.22 um) sterilised enzymes, papain (0.5 mg/ml) pepsin (0.1 and 0.5 mg/ml) and proteinase K (0.01 mg/ml) were prepared by addition of the enzyme solutions to medium cooled to 50°C just prior to setting. The 76 media were buffered to pH 4.0 (optimal pH for pepsin's activity) and pH 7.0 (optimal pH for the activity of papain and proteinase K) with CP buffer as in G.M.1. Isolates under test for sensitivity to culture filtrate inhibitor were grown on the solid media containing proteases and well tests were performed as in G.M.25. Culture filtrates from sensitive isolates, and also medium containing no mycelium but shake cultured under the same conditions, were employed as controls. Enzyme activity in the plates was determined, after one week, by incubating homogenised plugs (6mm) from each plate with a solution of bovine serum albumin (0.25 mg/ml, 100 ul) in CP buffer (pH 4.0 or 7.0 as appropriate). After incubation at 24°C for 24 to 48 h samples were analysed on SDS p.a.g.e. for the presence of bovine serum albumin. 77 SECTION I Screening of field isolates of P.sp.(lh) Ggg and Pg for the presence of dsRNA RESULTS P.sp(lh) isolates 2-2, 5-4, 12-2, 17-3, 21-3, 24-4, 55-1, and 74/1007-2, Ggg. isolate G1 and Pc[ isolate 1348-2 (Table 4) were grown to stationary phase in liquid culture, in one of three media (G.M.1.): (a) basal and corn steep liquor medium (Banks et al., 1971); (b) glucose/ asparagine medium (Weste and Thrower, 1963); and (c) 2% (w/v) malt extract. Nucleic acids extracted frcm the mycelium of each isolate (G.M.12a) were analysed by p.a.g.e. (G.M. 16.c.i.). Double stranded RNA was identified by the properties of staining pink with toluidine blue (single- stranded nucleic acids stain blue; Berry and Bevan, 1972), by its resistance to DNAse 1 and by its susceptibility to RNAse A at low salt concentrations (0.1 x SSC, 0.5 ug/ml RNAse) and relative resistance to this enzyme at high salt concentrations (2 x SSC) when compared with single-stranded RNA (bacteriophage MS2 RNA was degraded by incubation with 0.5 ug/ml RNAse at 2 x SSC for 1 h). In accordance with the findings of Loviny and Szekely (1973) the double stranded RNA was degraded by pancreatic RNAse, at high salt concentrations, when the enzyme/dsRNA ratio was increased (RNAse concentration increased to 8 ug/ml, enzyme/dsRNA > 1). All preparations contained a diffuse band of DNA which was completely degraded by the action of DNAse 1. Comparative polyacrylamide gel electrophoresis, using internal dsRNA standards isolated from Penicillium chrysogenum virus, P. stoloniferum virus S and occasionally Aspergillus foetidus virus F (G.M.16c.i.), was used to determine the molecular weights of the dsRNA components extracted Table 4 Table of Field Isolates of Phialophora and Gaeumannomyces species Isolate Host Source Date isolated Virus particle diameter (nm) P.sp(lh)2-2 Continuous barley D.B. Slope, 1973 35 and 40 Hoosfield, Rothamsted, England. P.sp(lh)5-4 Continuous barley 1973 35 D.B. Slope, P.sp(lh)12-2 Continuous barley Hoosfield, Rothamsted, England. 1973 35 and 40 D.B. Slope, P.sp(lh)17-3 Continuous barley 1973 HOosfield, Rothamsted, England. P.sp(lh)21-3 Continuous barley D.B. Slope, 1973 35 Hoosfield, Rothamsted, England. P.sp(1h)21-4 Continuous barley 1973 35 D.B. Slope, P.sp(lh)24-4 Continuous barley Hoosfield, Rothamsted, England. 1973 35 D.B. Slope, P.sp(lh)29-5 Continuous barley 1973 35 Hoosfield, Rothamsted, England. D.B. Slope, Hoosfield, Rothamsted, England. HoosfieldD.B. Slope, ,Rothamsted , England. Isolate Host Source Date isolated Virus particle diameter (nm) 1973 35 P.sp (lh) 55-1 Continuous barley D.B. Slope, Hoosfield, Rothamsted, England. 1973 35 P.sp (lh) 55-3 Continuous barley D.B. Slope, 1973 35 P.sp (lh) 55-4 Continuous barley Hoosfield, Rothamsted, England. D.B. Slope, 1974 35 P. sp (lh)74/951 -1A Continuous barley Hoosfield, Rothamsted, England. 1974 35 P. sp(lh) 74/1007-2 Continuous barley D.B. Slope, Hoosfield, Rothamsted, England. 1974 30 Pg_ 1348-2 1st wheat after ley D.B. Slope, Ggg G1 Prairie grass Hoosfield, Rothamsted, England. 1973 55 D.B. Slope, Ggt T1 Wheat 1974 35 Summerdells, Rothamsted, England Ggt F3 2nd wheat P.T.W. Wong, 1974 40, 35 and 27 Rydalmere, N. S .W., Australia. Ggt OgA 2nd wheat 1972 35 and 27 P.T.W. Wong, CarnaC.JLe Rheu. ,Rawlinson Wester, nearn Rennes,Australia , Franc. e J.E.E. Jenkins, Gcjft 3b1a 3rd wheat 1970 35 and 27 Selby, England M. Chu Chou, Little Knott Field Rothamsted, England. Isolate Host Source Date isolated Virus particle diameter (nm) Ggt 38-4 Continuous barley D.B. Slope, 1973 35-27 Hoosfield, Rothamsted, England Ggt 019/6 Barley after wheat C.J. Rawlinson, Highfield, 1972 35 Rothamsted, England Ggt 45/10 12th wheat C.J. Rawlinson, Little 1972 35 Knott field, Rothamsted, England. * Determined by electron microscopy (C.J. Rawlinson, personal communication) from each isolate (Fig. 3). Two isolates, P.sp.(lh) 2-2 and 12-2, were found to contain high molecular weight dsRNA species, which fell outside the range of these internal standards. The molecular weights of the larger species were determined using Helminthosporium maydis dsRNA and A. foetidus virus S and F dsRNAs (G.M.16c.i.); Fig. 4 for P.sp(lh) 2-2 nucleic acid species). Molecular weights were interpolated frcm plots of log molecular weight versus mobility of the internal standards as shown in Figs. 5(a) to 5(d). As demonstrated by other workers (Bozarth and Harley, 1976; Buck and Ratti, 1977) the plots were linear over a narrow range of molecular weight only. Four isolates, P.sp(lh) 5-4, 21-3, 29-5 and 55-1 were found to contain no detectable dsRNA. Eight isolates, P.sp(lh) 2-2, 12-2, 17-3, 24-4, 55-4 and 74/1007-2, Ggg_ G1 and P^ 1348-2, were shown to contain dsRNA. Results are given in Table 5. 82 (a) (b) (c) (d) (e) (f) (g) (h) 9 8 7 6 5 4 3 2 1 6 i- i i i i 1— J x 10 molecular weight Figure 3 Polyacrylamide gel electrophoresis of total nucleic acid extracts, from field isolates, after treatment with DNAse 1: (a)P.sp. (lh)2-2; (b)P.sp. (lh)12-2; (c)P.sp. (lh)17-3; (d)P.sp. (lh)24-4; (e)P.sp. (lh)55-4; (f)P.sp. (lh)74/1007-2; (g)Ggg Gl; (h)Pg 1348-2 83 P.sp. (lh) P .sp. (lh) 2-2 total 2-2 total nucleic nucleic acid acid plus dsRNA standards dsRNA 1 DNA dsRNA ~ 5 Figure 4 Polyacrylrunide gel electrophoresis of P.sp.(lh) 2-2 ~ - RNA directly extracted from mycelium 84 6.5 6.0 075 T^5— 5,51 0^5 TtO 1^5" Relationship between log molecular weight and electrophoretic mobilities of RNA components of isolates (a) P.sp.(lh) 2-2, (b) P.sp.(lh) 74/1007-2, (c) Ggg G1 and (d) Pg 1348-2 85 Table 5 - Molecular weights of dsRNA components extracted from field isolates of P.sp(lh) Ggg and Pg Field isolate Description Mol. wt. Approximate* (x10 6) yield (ug/g wet weight of mycelium) P.sp(lh)2-2 RNA 1 6.1 4 RNA 2 1.31 4 RNA 3 1.28 4 RNA 4 1.23 4 RNA 5 1.04 4 P.sp(lh)12-2 RNA 1 6 1 RNA 2 1.57 0.1 P.sp(lh)17-3 RNA 1 1.19 0.1 RNA 2 1.14 0.1 P.sp(lh)24-4 RNA 1 1.69 0.8 RNA 2 1.62 0.8 P.sp(lh)55-4 RNA 1 1.19 0.1 RNA 2 1.14 0.1 P.sp(lh)74/1007-2 RNA 1 1.35 4 RNA 2 1.28 4 RNA 3 1.23 4 RNA 4 1.04 4 Ggg G1 RNA 1 1.65 RNA 2 1.54 RNA 3 1.31 RNA 4 1.19 PC[ 1348-2 RNA 1 1.32 RNA 2 1.23 RNA 3 1.17 RNA 4 1.14 * Estimated from intensity of bands on polyacrylamide gels, after staining with toluidine blue, by comparison with known dsRNA standards. 86 DISCUSSION In a recent study of virus particles in isolates of Gaeumannomyces graminis, Frick and Lister (1978) found that serotype variability existed between the particles of different isolates and suggested that this might reflect similar virus biotype variability, and hence offer an explanation for the apparent lack of association between the presence or absence of any class or classes of virus and fungal pathogenicity. Although trad- itional analysis of viral strains has depended on serological studies, the availability of electrqphoretic analysis of nucleic acids has begun to play an important role and the method has permitted precise analyses of several viruses with segmented RNA genomes, such as reovirus (Hardy, Rosen and Field, 1979). The importance of genetic variation of dsRNA has been emphasised by the finding of different electrophoretic patterns of dsRNA and marked variation in the degree of reduction of pathogenicity by dsRNA in strains of Endothia parasitica (Day and Dodds, 1979; Dodds, 1980). Comparative electrophoretic analysis of dsRNA, directly extracted from fungal mycelium and following enzymic degradation of ssRNA and DNA, has 3 main advantages over serological comparisons of viruses for examining possible correlations of virus infection with a particular fungus phenotype. Firstly, it is a rapid analysis as it can be applied to small amounts of fungal mycelium (ca. 1g wet weight), which require only a short time to grow. With the use of semimicro gels the analysis can be scaled down further, by a factor of ten-fold, for rapid determination. The method does not require the use of an ultracentrifuge or time-consuming antisera to be prepared. Secondly, comparative p.a.g.e. can indicate differences not revealed by serological comparisons e.g. in Saccharomyces cerevisiae where both killer and sensitive strains contain serologically identical virus particles, killer strains contain an additional satellite dsRNA not found in sensitive strains (Berry and Bevan, 1972). Thirdly, the method 87 allows all dsRNA components to be detected, whether they are encapsidated or not, and relative proportions to be determined, without the problems of selective loss of some of the species through virus insolubility or instability. The results reported here show that there is considerable variation in the yields of dsRNA/g wet weight of mycelium (between 0.2 ug and 20 ug). Indeed in some isolates (P.sp(lh) 5-4, 21-3, 29-5 and 55-1), previously shown to contain virus particles (Table 3), no dsRNA could be detected by this method. The limit of detection of dsRNA on standard size gels is approximately 2.5 ug/ml (Almond, 1979) corresponding to a virus concentration of 12.5 ug/ml (assuming dsRNA to be 20% of the mass of the virus; Rawlinson et al., 1973). However the limit of detection of O virus particles by electron microscopy is estimated at 10 particles/ml (Anderson et al., 1966). Assuming that the molecular weight of the virus is 9 x 10 , the minimum concentration of virus that could be detected _3 by electron microscopy is 1.5 x 10 ug/ml, It is more than likely therefore that virus is present in these isolates, but at too low a concentration for possible dsRNA analysis. In the eight isolates found to contain dsRNA, considerable variation was found in the patterns of dsRNA components obtained following p.a.g.e. Only two of the eight isolates (55-4 and 17-3) had the same pattern of dsRNA components. Numbers of these components varied from two to five 6 fi and molecular weights varied from 1 x 10 to 6 x 10 in different isolates. Five different dsRNA patterns were obtained from isolates of phialophora sp. with lobed hyphopodia collected from a single field at Rothamsted Experimental Station (although some components frcm different isolates had molecular weights in common). As dsRNA mycoviruses are only transmitted intracellularly and do not lyse their hosts, they become a relatively permanent cytoplasmic feature of particular fungal strains, rather like DNA plasmids in bacteria. The results therefore indicate a 88 very heterogeneous population of the fungus P.sp.(lh) colonising barley roots. The observed variation in dsENA components could be due to infection of the fungal isolates with mixtures of viruses. Several fungi have in fact been shown to be infected with a mixture of individual viruses, e.g. Thielaviopsis basicola (Bozarth and Goenaga, 1977), Agaricus bisporus (Barton and Hollings, 1979), Penicillium stoloniferum (Bozarth et al., 1971) and Aspergillus foetidus (Ratti and Buck, 1972). Consideration of the size of the virus particles visualised by electron microscopy (Table 3) indicates the presence of at least two viruses (35 nm and 40 nm in diameter) infecting isolates P.sp(lh) 2-2 and 12-2. A striking observation was that dsRNA components of molecular weight > 6 x 10^ were found only in these isolates containing 40 nm particles. Another possibility to explain the variation in dsRNA patterns is the presence of additional dsRNA segments to those required for virus replication (encoding a polymerase enzyme) or encapsidation (encoding capsid polypeptides). These additional components might be satellite dsRNAs encoding specific proteins, e.g. M dsRNA of yeast which encodes killer toxin (Bostian et al., 1980) or might be deletion mutants, e.g. S dsRNAs of yeast (Bruenn and Kane, 1978). 89 SECTION 2 Isolation and properties of virus particles from P.sp(lh) RESULTS A. Preparation of crude virus P.sp. (lh)2—2 was grown in basal and CSL culture medium (G.M.1.) for 8 to 12 days, and was harvested, at the primary stage, by filtration under reduced pressure. Cultures were grown in fermenters by inoculation of medium with mycelium from 5 days primary and 5 days secondary culture and growth for an additional 2 to 3 days before harvest. The wet weight of mycelium obtained from three separate small scale preparations was between 6.0 and 6.5 g/100ml culture medium, and from 601 fermenter preparations was between 11.5 and 13.5 g/100ml culture medium. Virus preparations were obtained from P.sp.(lh)2—2 mycelium by a method involving PEG precipitation and differential ultracentrifugation (G.M. 2.). Samples of the crude virus preparations before and after the final clarifying spin, were incubated with SDS to disrupt the virus particles and were then analysed by electrophoresis in RNA gels (G.M.16c.i.). Results show that the high molecular weight species of dsRNA (RNA 1 in total nucleic acid preparations, see table 5) is present in the virus extracts, but is largely removed by the final clarifying spin given to the preparation. The other RNA species (RNAs 2 to 5, Table 5) are apparently unaffected by this procedure. Yield of virus was estimated from the intensity of toluidine blue staining of dsRNA bands, after p.a.g.e. of serially diluted samples of crude virus, and was calculated on the assumption that 20% of the mass of the virus is nucleic acid (Rawlinson et al., 1973). The yield of virus was found to be between 95 and 108 ug/g wet wt. mycelium for small scale preparations and between 78 90 and 90 ug/g wet wt. mycelium for fermenter preparations. B. Purification of virus by sucrose density gradient centrifugation a. Small scale purification of virus. The crude virus preparation was further purified by density gradient centrifugation of samples in a 10% to 45% (w/v) sucrose gradient (5ml, G.M.2). Gradients were fractionated into 0.2ml samples. Virus particles containing dsRNA were located by spotting on to an ethidium bromide-agarose plate (G.M.2c). Fractions containing dsRNA fluoresced under U.V. light in the presence of ethidium bromide which intercalates between the bases in double stranded nucleic acids (Douthart et al., 1973). Larger volumes of crude virus were purified by centrifugation in a 50 ml 10 to 45% (w/v) sucrose density gradient. The trace obtained by passage of a gradient through a photocell in the ISCO Model D apparatus (G.M.2) is shown in Fig.6. b. Large scale zonal purification of virus Crude virus samples were purified by sedimentation through 20 to 50% (w/v) sucrose gradients in an MSE B XIV zonal rotor (G.M.2c). Fractions (15ml) were collected and analysed for the presence of virus by examination in the Cary '15' spectrophotometer (G.M.2). A U.V. profile of the gradient is shown in Fig.7. Fractions from the central portion of each of these gradients, shown to contain dsFNA by reaction with ethidium bromide under U.V. light, were analysed by electron microscopy to confirm the presence of intact virus particles. Electron micrographs showed the presence of isometric virus particles with electron dense cores, indicating encapsidation of nucleic acids (in fractions 18 to 30 of the zonal gradient, Fig. 7). Evidence was also obtained for the presence of virus particles empty of Volume (ml) Figure 6 Purification of P.sp.(lh) 2-2 virus by rate centrifugation in a sucrose density gradient: ISCO u.v. scanner tracing CD A 260 5 10 15 20 Fraction number CO Figure 7 Purification of Ptsp.(lh) 2-2 virus by rate centrifugation in a zonal sucrose density gradient to 93 nucleic acid, which did not sediment as rapidly in the sucrose gradients as full particles and were thus detected nearer the upper end of the gradient (fractions 15 to 17, Fig.7K C. Properties of purified virus particles a. U.V. spectrum A U.V. spectrum of P.sp(lh)2-2 virus sample is given in Fig.8. The spectrum is typical of nucleoprotein with a maximum at 260 nm and minimum at 254 nm and ~ 1*4, b. Particle morphology Electron micrographs of the purified virus preparation were taken after negative staining with phosphotungstate (G.M.3). A typical example is shown in Fig.9. The size of the particles was determined from the micrographs (x 100,000 magnification) by direct measurement and frcm the negatives (x 40,000 magnification) using an eye piece graticule. Histograms constructed from both these measurements, Fig. 10 a and b, show that the modal diameter of P.sp(lh)2-2 virus particles is 35 nm (with a very few 40 nm particles being present). Examination of fractions containing "empty" virus particles, as judged by the electron-transparent core, Fig.11, show the empty particles to have a modal diameter of 35 nm also. c. Sedimentation coefficient The sedimentation coefficient of virus particles purified frcm P.sp. (lh)2-2 was determined by measurement of the rate of sedimentation in P buffer (G.M.10a). A scan of the cell showing the moving boundary after 20 min. of centrifugation is shown in Fig.12. The plot of In x^ against t (min) made as described in G.M. 10a, 230 260 290 Wavelength of u.v.light (on) Figure 8 U.v.spectrum of P.sp.(lh) 2-2 virus, following purification on a sucrose gradient Figure 9 Electron micrograph of P.sp.(lh) 2-2 virus (x 100,000 magnification) 96 60 (a) 40 N-T-I 30 35 40 Particle diameter (nm) Figure 10 Histogram plots of VLP diameters measured randomly from electron micrographs of P.sp.(lh) 2-2 virus by (a) direct measurement from negatives or (b) ' from prints Figure 11 Electron micrograph of 'empty' P.sp.(lh) 2-2 virus (x 100,000 magnification) cell top Figure 12 Photoelectric scanner tracing for the determination of sedimentation coefficient of P.sp.(lh) 2-2 virus CO 00 99 2 provided a slope of w s (Fig. 13) from which the sedimentation coefficient 20 was obtained. Following correction for temperature two values of Sq were obtained. These were 121.6 S and 120.6 S (average 121.1 S). d. RNA components RNA components present in pure virus preparations were analysed by disruption of the particles with SDS, followed by p.a.g.e. (G.M.16c.i). In every case it was found that purification of the virus preparations in sucrose density gradients resulted in the loss of the RNA species 1 (Fig. 14). RNA 1 was the component in crude virus preparations partially removed by a final clarifying spin given to the extract. The other four RNA species, numbers 2 to 5, were unaffected by this procedure. Relative proportions of the four species remained the same as in total nucleic acid and crude virus preparations. On analysis with internal dsRNA markers (G.M. 16) the molecular weight values of the four components were found to be the same as determined for total nucleic acid preparations (Section 1) Molecular weight RNA2 1.31 x 106 RNA3 1.28 x 106 RNA4 1.23 x 106 RNA5 1.04 x 106 RNA from total nucleic acid preparations was compared with pure virus RNA for double strandedness. This was made possible by separation of the mycelial RNA from DNA by CF11 chromatography (G.M.8.) Mycelial nucleic acid samples were loaded onto a CF11 column (4 A£gQ units/ml bed volume) and the column was eluted batchwise with 65% SET buffer/35% ethanol, 85% SET/15% ethanol and 100% SET buffer. Fractions were collected (0.5 ml) and nucleic acid detected by spotting samples on an ethidium 100 20 40 t (min) Figure 13 Plot of In x versus t, from which the sedimentation coefficient of P.sp.(lh) 2-2 virus is obtained 101 P.sp.(lh) P.sp.(lh) P.sp.(lh) PcV,PsVS 2-2 crude 2-2 pure 2-2 pure dsRNA virus virus virus plus standards dsRNA dsRNA PcV,PsVS dsRNA stds. * PcV-Peni ci11ium chrysogenum dsRNA PsVS-Penicillium stoloniferum virus S dsRNA Figure 14 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus RNA in crude and purified virus preparations 102 bromide agarose plate (Fig. 15). It appeared that DNA was eluted frcm the column by 85% SET/15% ethanol and dsRNA by 100% SET. The separation was further confirmed by gel electrophoresis (G.M.16c.i) as shown in Fig. 16. Hie five species in P.sp(lh)2-2 were shown to be well separated frcm DNA, a diffuse band shown to be susceptible to DNAse in the screening procedure (section 1). This separation of RNA and DNA was necessary so that the required RNAse/RNA ratio could be determined for both mycelial and viral RNA samples. The purified viral and mycelial RNA species were analysed for their differential susceptibility to RNAse A at high and low salt concentration (G.M. 13b; dsRNA/PNAse = 22.5 to 40) using MS2 RNA as a ssRNA control, and were found to be double stranded. e. Polypeptide components Samples of purified virus particles were examined, after denaturation by boiling with SDS and p mercaptoethanol, by electrophoresis in SDS polyacrylamide gels (G.M. 16d.) in order to determine the number of polypeptide species present and their molecular weight. Two species of polypeptide were found to be present in the pure virus preparation (Fig.17). Values for the polypeptide molecular weights were interpolated from plots of log^Q molecular weight of the internal markers versus their mobility in polyacrylamide gels, as shown in Fig. 18. Molecular weight values obtained for the species were corrected by the method of least squares analysis and were found to be: Molecular weight* Polypeptide 1 65,500 Polypeptide 2 60,700 * Mean of 5 determinations each. 103 Figure 15 Ethidium brorrdde-a~rose plate analysis of P.sp.(lh) 2-2 isolate DNA and dsRNA-containing fractions, separated by CF11 chromatography 104 fraction no. 13 14 15 16 17 18 19 20 21 22 23 \ /JO% total DNA SET dsRNA I nucleic acid Figure 16 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 isolate DNA and dsRNA-containing fractions,separated by CF11 chromatography 105 phosphorylase a bovine serum albumin P1 P2 glutan~te dehydrogenase lactate dehydrogenase polypep. P.sp.(lh) P.sp. (lh) stds. 2-2 virus 2-2 virus polypeps. polypeps. plus stds Figure 17 PolyacrYlamide gel electrophoresis of P.sp.(lh) 2-2 virus polypeptides 106 5.0 -p a •H 4.75 a 1 b£> O 4.5 1 2 3 Mobility (cm/h) • Protein standards in order of increasing mobility: phosphorylase a; bovine serum albumin; glutamate dehydrogenase and lactate dehydrogenase O P.sp.(lh) 2-2 polpeptides as numbered Figure 18 Relationship between log molecular weight and electrophoretic mobilities of P.sp.(lh) 2-2 virus polypeptides in SDS/polyacrylamide gels 107 The polypeptide species were also analysed with another protein marker - pyruvate kinase (molecular weight, 60,000), with which polypeptide 2 was found to oomigrate (Fig. 19). This value is in good agreement with that obtained using the other markers. Literature values for the molecular weight of standard polypeptides often show differences 3 . of up to 1 x 10 in size, for example bovine serum albumin has been reported to have a molecular weight of 67,000 (Squire et al., 1968; Castellino and Barker., 1968) and 68,000 (Weber et al., 1972). In these determinations the highest and lowest molecular weight values obtained differed by up to 0.8 x 10 3 for a single polypeptide. D. Preliminary evidence that purified virus preparations contain two distinct viruses in P.sp(lh)2-2. a. Serology of P.sp(lh)2-2 virus. Purified virus from P.sp(lh)2-2 was used to raise antisera in a rabbit (G.M.5). Preimmune and test (post immunisation) sera were collected and prepared as in G.M. 5. and were tested against P.sp (lh) 2-2 virus preparations by the Ouchterlony double diffusion method (G.M.6). Samples of the antiserum were diluted stepwise with PK buffer and were tested against an undiluted sample of P.sp (lh) 2-2 pure virus ^gg = 0.5) to determine the antiserum end titre. Evidence for the presence of two viruses in the pure virus preparation was shown in the development of two preciptin lines between wells containing the virus antigen and the antiserum, as shown in Fig.20. The two precipitin lines showed distinct antiserum end titres (the dilution of antiserum at which no antigen- antiserum complex formation was visible) of 1 to 512 dilution and 1 to 2048 dilution with PK buffer. b. Electrophoretic separation of the viruses on agarose gels. PI pyruvate kinase P2 Figure 19 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus polypeptides: comigration of P2 with pyruvate kinase 109 v AS v Figure 20 Immunodiffusion test of P.sp.(lh) 2-2 virus with antiserum raised against this virus A B Figure 21 Agarose gel electrophoresis of P.sp. (lh) 2-2 virus 110 Virus preparations (A2g0 2.0 to 4.5) were examined by agarose gel electrophoresis (G.M. 16a) and showed the presence of two bands, staining with Coomassie blue, as in Fig.21. The two viruses were further examined by excising regions corres- ponding to the stained bands, from unstained gels and solubilising the slices by heating with SDS and mercaptoethanol (G.M. 16e.). Samples were cooled to between 50 and 55°C and were loaded on top of semi-micro polypeptide gels prior to setting. Virus samples were then analysed for their polypeptide constitution by electrophoresis in the usual way (G.M. 16d) for 75 min. The results showed that the polypeptide component with the slowest mobility on these gels, called polypeptide 1 (section 2C.e.) was contained in the virus with the fastest mobility on agarose gels, and vice versa. The virus species were termed virus A and virus B, these being the slower and faster moving species on agarose gels, respectively. Thus virus A contained polypeptide 2, and virus B contained polypeptide 1. c. Fractionation of RNA and polypeptide components across a sucrose gradient. Examination of virus-containing fractions by p.a.g.e., as in Section 2 C.d (Fig.22) demonstrated four distinct RNA bands, RNAs 2 to 5, but showed fractionation of these species across the sucrose density gradients on which they were purified. This fractionation of components was also demonstrated in the polypeptide gels, as shown in the same figure (Fig.22^. RNA component 5 and polypeptide 2 appeared closest to the top of the gradient, i.e. closest to the centre of rotation, and were the predominant species in fraction 14, where RNA components 3 and 4 showed only faintly and polypeptide 1 appeared to be absent. RNA component 2 and polypeptide 1 appeared in the next fraction, fraction 15, but whereas RNAs 3 to 5 peaked in this fraction, RNA 2 peaked in the next fraction, number 16, as in the figure. In further fractions all four RNA components decreased in concentration together, with RNA 2 being the last to disappear closest to the lower end of the gradient. Ill Fraction no. Fraction no. 14 15 16 17 18 19 20 14 15 16 17 18 19 20 C3 O •H t( «H O -p fl CO •H O Q •H $ CD •H Q (a) (b) Figure 22 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus RNA (a) and polypeptides (b) from sucrose density gradient fractions following rate centrifugation (15 V/cm,l h) 112 The polypeptides 1 and 2, although first appearing in different fractions, peaked together in fractions 16 and 17 and then decreased in intensity together, towards fraction 21, where the components were no longer visible. These results indicated that the RNA components 5, and possibly 3 and 4 and the polypeptide component 2, were retarded on the sucrose density gradient relative to the components with lower mobility on p.a.g.e. RNA 2 and polypeptide 1. Consideration of the relative proportions of these species in the various fractions from the gradients, indicated that polypeptide 2 encapsidated RNA 5 and possibly RNAs 3 and 4, and suggested that polypeptide 1 probably encapsidated RNA 2. It was noted, from a comparison of the fractionation patterns across the gradients of both small and large scale (zonal) preparations that the former methods, on 5 ml and 50 ml gradients, gave better separation of the RNA and polypeptide components across the density gradient. E. Separation of the viruses of P.sp(lh)2-2 Electrophoretic separation of the two viruses (section 2 Db) had been made possible due to differences in surface charge and mass of the virus particles. The two virus species moved at different rates in an electric field, mainly due to differences in the proportion of basic and acidic groups present in the virus capsid proteins at a given pH. The same differences between the virus species were thought likely to allow their separation by ion exchange chromatography, whereby affinity for particular charged species bound to a column matrix is utilised. The two major classes of ion exchange resin used are cation exchangers (with fixed anionic groups) and anion exchanges (with fixed cationic groups). DEAE cellulose, an anion exchanger, had previously been used to separate the viruses of Penicillium stoloniferum and those of Aspergillus foetidus (Buck and Kempson-Jones, 1973; Buck and Ratti, 1975) and it was therefore thought that this resin would be useful in separating viruses of P.sp.(lh)2-2. 113 DE52 cellulose in P buffer was prepared (G.M.7) and virus samples (loading volume max. ^^q = 2/ml bed volume) in P buffer were allowed to enter the column bed. Samples were washed into the column with P buffer (half the column volume) and the column was then eluted with a batchwise or continuous gradient of 0 to 0.6 M KC1 in P buffer at a maximum flow rate of 1.2 ml/min. The presence of virus in collected fractions was detected by ethidium bromide-agarose plate analysis (Fig.23) and was confirmed by electrophoresis in agarose gels. An A2^q profile of one of these gradients (Fig.24), eluted by batchwise addition of P buffer followed by P buffer containing 0.6 M KC1, showed that virus A, the virus with slowest mobility in agarose gels, was eluted immediately after the void volume, and virus B, the virus with the fastest mobility in agarose gels, was eluted with the higher salt concentration. Elution with a continuous gradient showed that virus B eluted between 0.39 and 0.44 M KC1. The presence of virus in these fractions was confirmed by electron microscopy (Fig.25a and b). Thus, while in P buffer, the more highly charged virus B (as shown by its greater mobility in agarose gels) was retained by the opposite charge of the anion exchanger. (The pH of the buffer for the agarose gels (pH 8.0) and ion exchange column (pH 7.6) were similar). It was clear that virus A was the predominant species extracted from the P.sp(lh)2-2 isolate, contributing more than 75% of the total number of particles. However it was possible that virus B was more unstable under the experimental conditions used. As the B fraction was eluted second from the column, and the A fraction was so large, it was possible that the B fraction might be contaminated with very small amounts of virus A. Further separation of the virus B fraction was therefore attempted on a cation exchanger of SP sephadex, an exchanger with weak affinity for cations. Dialysis of virus particles against an acidic buffer increases the proportion of positively charged species and allows 114 virus A virus B Figure 23 Ethidium bromide-agarose Vlate analysis of P.sp.(lh) 2-2 viruses A and B separated by DE52 cellulose ion exchange chromatography Fraction number Figure 24 U.v. profile of P.sp.(lh) 2-2 viruses A and B eluted from a DE52 cellulose column Figure 25 Electron micrographs of P.sp.(lh) 2-2 viruses A (a) and B(b) (x 100,000 magnification) 117 the separation of virus particles by virtue of their affinity to anionic groups on the Sephadex beads. Virus B fractions from the DE52 column were therefore dialysed against Ac buffer and were absorbed on to a column of SP Sephadex equilibrated with Ac buffer (G.M.9). The virus sample was washed into the column with Ac buffer and was then eluted with a continuous gradient of 0 to 1.0 M KC1 in Ac buffer. Fractions frcm the column were analysed on an ethidium bromide-agarose plate (Fig.26) Virus B was shown to be eluted first between 0.04 and 0.1 M KC1 followed by traces of virus A with 0.3 M KC1. Separation of the "empty" fraction of virus particles (Fig.11) prepared by sucrose density gradient centrifugation, was attempted by DE52 cellulose ion exchange chromatography. Examination of the separated virus fractions in agarose gels, showed that while virus A empty virus particles had the same mobility as intact virus A, virus B seemed to have broken down into two species with higher mobility than intact virus B (Fig.27). Electron microscopic examination of the two fractions showed the presence of intact, but empty virus A capsids (Fig.28) but a number of disrupted virus B capsids with a few intact, though empty, particles (Fig.29). F. Properties of each individual virus a. Particle morphology Electron micrographs were taken of samples of both intact and "empty" virus particles A and B (Figs.25a and b, 28 and 29). Histograms, constructed to show the size distribution of particles in these samples (Fig. 30a to d) showed that intact virus A and virus B were 35 nm in diameter. Virus A particles, containing no nucleic acid, as demonstrated by their electrontransparent cores, were also 35 nm in diameter. Virus B particles appeared to be disrupted by the DEAE cellulose exchange procedure, and showed a range of "defective" particles, of two main types. There 118 Figure 26 Ethidium bramide-~arose plat~ analysis of P.sp.(lh) 2-2 virus B purified on SP-Sephadex ion exchange chromatography 119 virus B virus A Figure 27 Agarose gel electrophoresis of P.sp.(lh) 2-2 'empty' viruses A (a) and B (b) Figure 28 Electron micrograph of 'empty' virus A particles ( x 100,000 magnification) 121 Figure 29 Electron micrograph of 'empty' virus B particles (x 100,000 magnification) 122 60 30 30 35 40 30 35 40 70 ft (d) «oH 35 40 20 25 30 35 40 Particle diameter (nm) Figure 30 Histogram plots of VLP diameters measured randomly from electron micrographs of 'full1 virus A (a) and virus B (b) and 'empty' virus A (c) and virus B (d) 123 were the expected "empty" capsids of 35 nm in diameter and also smaller particles with apparently electron dense cores of mean diameter 29 nm and small "stellate" particles, which possibly consist of fragments of viral capsid reassembled in a regular array of 21 nm diameter. b. U.V spectra U.V. spectra of samples of purified virus A and B (dialysed against P buffer) were similar to that of the virus mixture (Fig.31a and b) with A260/280 = 1,50 c. Sedimentation coefficients for virus A and virus B Samples of the viruses A and B, in P buffer, were centrifuged to sediment (G.M. 10a) and the rate of movement of the virus boundary 20 (Fig.32a and b) determined for each virus. Two values of Sq were obtained for each virus, 115.9 S and 116.2 S (average 116.05 S) for virus A, and 121.8S and 123.OS (average 122.4 S) for virus B. d. DsRNA components DsRNA components from each virus were analysed by p.a.g.e. (G.M.16.c.i. Fig.33). and scans were taken of the gels (G.M.4.) as shown in Fig. 34a and b. The proportions of the dsRNA components in each virus, as measured from the areas of the peaks in the gel scans, were not equal, suggesting that the three dsRNA components were not enclosed in a single particle, but were probably encapsidated separately as has been found for several other dsRNA mycoviruses (Hollings, 1978K Coelectrophoresis of the dsRNA components with dsRNA markers (G.M.16c.i.) demonstrated that virus A particles contained dsRNA components of molecular weight 1.29 x 106, 1.22 x 106 and 1.03 x 106 and virus B particles contained dsRNA components of molecular weight 1.32 x 10 , 1.25 x 106 and 1.03 x 106. 124 •8 O W 3 230 260 290 320 wavelength of u.v.light (nm) > Figure 31 U.v. spectra of P.sp.(Ih)2-2 viruses A (a) and B (b) following purification by DE52 and SP-Sephadex ion exchange chromatography Figure 32 Photoelectric scanner traces for the determination of sedimentation coefficients of Psp.(lh) 2-2 viruses A (a) and B (b) 126 (a) (b) Direction of migration » Figure 33 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus A RNA (a) and virus B RNA (b) Distance in gel (cm) Figure 34 Polyacrylamide gel profiles of RNA released from P.sp.(lh) 2-2 viruses A (a) and B (b) 128 e. Polypeptide components When purified preparations of virus A and virus B were analysed by SDS p.a.g.e. (G.M. 16d) it was shown that virus A contained one major polypeptide of molecular weight 60,000 and virus B contained one major polypeptide of molecular weight 66,000 as shown in Fig.35. f. Serology In gel immunodiffusion tests, the separated viruses A and B were allowed to diffuse against an antiserum prepared against purified virus fractions from the sucrose density gradient, i.e. that contained both virus particles and had been shown to produce two precipitin lines against these particles (section 2D.a.). Each virus gave rise to a single precipitin line and when the viruses were placed in adjacent wells, crossing precipitin lines were obtained (Fig.36) indicating that the two viruses were serologically unrelated. The antiserum was titrated in two-fold serial dilution against each virus and the antiserum end titre was found to be 1:2048 against virus A and 1:512 against virus B. A mixture of P.sp.(lh)2-2 viruses A and B was tested against antisera raised against Pg 1348-2 virus (see section 5), Ggg G1 virus (see section 4) and viruses of Ggt isolates OgA, 019/6, F3, T1, 3b1a and 38-4 using the Ouchterlony double diffusion method (G.M.6). No sign of reaction to any of these antisera was noted. Ggt 45/10 virus (preparations in Bor and PK buffers, see section 3) was found to react with antisera raised to the P.sp (lh) 2-2 viruses. Two precipitin lines developed, suggesting that the Ggt 45/10 crude virus preparation contained two viruses with antigenic determinants related to those of the P.sp(lh)2-2 viruses. g. Buoyant density The buoyant densities of virus particles A and B in caesium chloride were calculated by centrifuging to equilibrium sanples of each of the 129 A VvvWb pq^pep siAT^ W-T stds T I1 (a) (b) Figure 35 Poly acrylamide gel electrophoresis of P.sp.(lh) virus A polypeptide (a) arid virus B polypeptide (b) 130 MBM VB Figure 36 Inmunodiffusion test of P.sp. (lh) 2-2 viruses A (VA) and B (VB) against antiserum raised against P.sp.(lh) 2-2 virus 131 viruses, and a mixture of them both, at different starting densities of caesium chloride in the Beckman Model E centrifuge, (G.M. 10 b). A scan of the cell containing both viruses is shown in Fig.37. Virus A was found to have a buoyant density of 1.347 g/ml and virus B a buoyant density of 1.363 g/ml. A summary of the physicochemical characteristcs of the viruses A and B is shown in Table 6, including the new nomenclature used for P.sp(lh)2-2 dsRNA species. Table 6 Physicochemical characteristics of P.sp.(lh)2-2 viruses Virus A Virus B Particle diameter (nm) 35 35 20 SQ 116S 122S Buoyant density (g/ml) 1.347 1.363 Molecular weights of RNA 1 1.29 x 106 RNA 1 1.32 x 106 dsRNA components ^ RNA 2 1.22 x 10 RNA 2 1.25 x 10 6 RNA 3 1.03 x 106 RNA 3 1.03 x 106 Molecular weights of P 3 6.0 x 104 P 1 6.6 x 104 polypeptide components G. Virion associated RNA polymerase activity a. Properties of the RNA polymerase reaction Both viruses A and B were shown to catalyse the incorporation of [ H] UMP into TCA-insoluble material in RNA polymerase reaction mixtures (G.M. 18). Polymerase assays were set up with varying concen- trations of magnesium ions from 0.625 mM to 10.00 mM and the radioactivity incorporated into the TCA-insoluble material was determined (Fig 38a and b). Viruses A and B were found to have optimal magnesium ion concen- trations of 2.5 mM and 5.0 mM respectively. No pretreatment of the virions with a proteinase or by heat shock was found necessary to activate (a) (b) Density of CsCl (g/ml) Figure 37 Photoelectric scanner trace for the determination of buoyant density of P.sp.(lh) 2-2 viruses A (a) and B (b) 133 Figure 38 Dependence of the RNA polymerase reactions of F.sp.(lh) virus A (a) and virus B (b) on magnesium ion concentration 134 the polymerase, as is the case with other dsRNA viruses such as reovirus (Shatkin and Sipe, 1968). Omission of magnesium ions in the polymerase mixture resulted in the abolition of [ H]-UMP incorporation, and emission of one of the four nucleoside triphosphates, ATP, resulted in the incorporation of a very low level of radioactivity into the TCA insoluble material. This is shown in Table 7. Table 7 Incorporation of [3H] into TCA insoluble material in RNA polymerase assays 3 Polymerase reaction [ H] ct/min incorporated mixture after 10 h. Virus A (2.5mM[Mg+ +D 124336 + Virus B (5.0mM[Mg +n 10939 - Virus A 3432 - Virus B 3055 I |. - Mg , virus A 1018 - Mg4"*", virus B 1577 - ATP, virus A 4977 - ATP, virus B 2644 A time course of the polymerase reaction was plotted for both viruses (Fig. 39a and b), by incubation of reaction mixtures containing 2 A26q units of virus at 30°C (G.M. 18). This showed that, while the virus B reaction slowed down after 3 to 4 h, the virus A reaction continued 3 with a linear [ Hl-UMP incorporation for more than 55h. This difference was probably due to the greater instability of virus B under the reaction conditions. 135 Figure 39 RNA synthesis catalysed by P.sp.(lh) 2-2 viruses A and B RNA polymerases 136 b. Analysis of reaction products The products of a 5h polymerase reaction with each virus (A2gQ = 1.0) were isolated by phenol/SDS extraction (G.M. 12b) and examined by electrophoresis in semi micro polyacrylamide gels containing 8M urea (G.M.16c.i.). Staining of these gels with toluidine blue failed to demonstrate any difference between the virus B pattern before and after reaction, but showed two extra bands in the virus A polymerase product gel (Fig. 40). The gels were frozen, sliced (0.1 cm slices) and the slices solubilised and counted in liquid scintillant (G.M.22) to determine the pattern of radioactivity along the gel. The results of this analysis are shown in Fig.41 a and b. The approximate positions of the double stranded and putative single stranded components, as determined from the stained gels before slicing, are marked on the figures. (Taking into account the expansion of the gel on freezing, it is likely that the two dsRNA peaks marked on Fig.41 a, correspond to the second large peak on the graph (3.0 to 3.8 cm in gel). This did not reveal much detail about 3 the reaction products, although results do indicate that [ H]-UMP was incorporated into dsKNA as well as into areas corresponding to the putative ssKNA components of virus A and virus B. Samples of polymerase reaction mixtures (5h) of viruses A and B were extracted with phenol and SDS. RNA was then precipitated with ethanol and the precipitate was resuspended in PE buffer. Samples were examined by semi micro gel electrophoresis and were compared with samples containing virus A or virus B, but lacking the nucleoside triphosphate ATP, and prepared in the same way. Staining of the gels of the virus Bg product revealed only the largest dsRNA component (mol. wt. 1.32 x 10 ) in both ATP+ and ATP- samples. Virus A gel analysis again showed the presence of two dsRNA bands (in both ATP+ and ATP- samples) and two extra bands staining in the ATP+ sample. These gels were processed for fluorography (G.M.21) but it was difficult to distinguish the number of new bands 137 virus A ssENA and dsRNA virus A dsRNA virus B ssRNA and dsRNA virus B dsREK Figure 40 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 viruses A (a) and B (b) dsRNA and ssRNA product 138 Polyacrylamide gel analysis of RNA synthesised 1 Phialophora viruses A (a) and B (b) measured by incorporation of PH] UMP into ssRNA 139 present in the gel pattern of virus A or virus B. However none were visible in the ATP- samples. It was evident again that the highest level 3 of incorporation of [ H]-UMP was into the slower moving components on the gel. There were a number of problems with the semi micro gels, both in the intensity of band staining and in the deformability of the gels during the fluorography process. The experiment was repeated using a higher concentration of virus A (A260 = 2.0) in an 18 h polymerase reaction mixture. Following phenol/SDS extraction, TCA and ethanol precipitation, dsRNA genome and products were separated on slab gels. After electrophoresis (G.M. 16c.ii) and staining with toluidine blue, six bands were revealed (the three slow moving components were blue compared to the pink of the three faster moving dsRNA components). The three faster moving components corresponded to virus template dsRNAs, fi fi fi molecular weights 1.29 x 10 , 1.22 x 10 and 1.03 x 10 and three bands of newly synthesised RNA of slower mobiltiy. The RNA products were incubated with RNAse A at concentrations of 0.1, 0.25, 0.5, 1.0 and 2.0 ug/ml. The three newly synthesised bands were partially degraded at 0.1 ug/ml RNAse A and completely degraded at 0.25 ug/ml RNAse A. The dsRNA bands were unaffected at any concentration up to 1.0 ug/ml enzyme, when a decrease in intensity of the three bands indicated partial degradation (these samples were in PE buffer, a low salt buffer which would allow degradation of dsRNA at this concentration of RNAse, cf 0.1 x SSC buffer). All bands were absent from the gel when treated with 2.0 ug/ml RNAse A. Thus the three newly synthesised RNA components are single stranded. It was noted that, while the three virus A dsRNA components had the 6 6 same intensity as before, i.e. the 1.29 x 10 and 1.22 x 10 molecular weight components of equal intensity and the 1.03 x 10 species of higher intensity (Fig.42), the single stranded components showed a 140 ssRNA 1 -=- 2 dsRNA 3 Figure 42 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus A dsRNA and ssKNA product Figure 43 Fluorograph of P.sp.(lh) 2-2 virus A dsRNA and ssKNA product following polyacrylamide gel electrophoresis 141 different distribution of intensity. However, the distances between the three bands were equal to those between the three dsRNA bands. Samples of the same dsRNA/ssRNA product preparation were separated by electrophoresis on a polyacrylamide slab gel as before, and the gel was then subjected to the fluorography process (G.M. 21). The result is 3 shown in Fig.43. Incorporation of [ H]-LJMP into single stranded RNA appeared to be less efficient into the virus A dsRNA 1 product (the largest dsRNA species, molecular weight 1.29 x 10**), than into the products fran the other two virus A dsRNA species, RNAs 2 and 3, as judged by the intensity of film-darkening. Coelectrophoresis experiments on the virus A polymerase product with denatured and glyoxalated (G.M. 17) dsRNA from virus A, showed that the three ssRNA products had the same mobilities as the products of glyoxalation of the dsRNA species (Fig.44). This demonstrated that the ssRNA products had molecular weights about half those of the dsRNA components. This analysis makes the assumption that in the 8M-urea buffer the unglyoxalated ssRNA had little secondary structure, which might affect its mobility. In order to prove that the ssRN3A product arose by transcription of the virus template dsRNAs, the [ H] ssRNA products of an 18h polymerase reaction were isolated by precipitation with 2M-LiCl (G.M.19) and were annealed with increasing amounts of unlabelled virus A dsRNA (ug dsRNA/ssRNA = 1.25, 2.5, 5.0, 10.0, 20.0 and 40.0). The extent of hybridisation which occurred was measured by the amount of label which remained TCA-insoluble after treatment with RNAse A in hybridisation buffer, i.e. that was annealed into dsRNA form. The proportion of product capable of hybridising to the virus template RNA increased with the amount of virus dsRNA to 85 to 95% (Fig.45 a). Little self annealing of the product (25 to 125 ug/ml), in the absence of added denatured dsRNA, could be detected following treatment with RNAse A. (Only 2.8 to 142 dsRNA 1— 3 ds/ss ds/ss GLY RNA RNA plus dsRNA product GLY dsRNA GLY dsRNA - glyoxalated dsRNA 3 ds/ss RNA _ H labelled dsRNA template and ssRNA product product Figure 44 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus A dsRNA and ssRNA product: comigratiOn with glyoxalated dsRNA 143 80 xi Q) W 3 60 •H .d CD 9 40 - •P 8O t* 0 ft 20 10 20 30 40 Weight ratio dsRNA:ssRNA Figure 45 Hybridisation of virus A RNA polymerase product to (a) virus template dsRNA and (b) virus B dsRNA 144 5.4% of the incubated ssRNA remained TCA precipitable after RNAse treatment). The products of the polymerase reaction are therefore full length ssRNA transcripts of each of the three virion dsRNA molecules. The amount of ssRNA product synthesised after a 24h RNA polymerase reaction was calculated (a) from absorbance at 260 nm, assuming of 1 is equivalent to 40 ug/ml of ssENA and 50ug/ml dsRNA and (b) [JH] UMP incorporation, assuming a UMP content of 25% for ssRNA. Both methods indicated that a weight of ssRNA approximately equal to that of the template dsRNA had been synthesised, i.e. on average two rounds of transcription per dsRNA molecule had occurred, indicating that re- initiation of transcription can occur in the iji vitro system. c. Cross hybridisation experiments Tritium labelled virus A ssRNA transcripts were hybridised to unlabelled virus B dsRNA in an attempt to determine if there was any sequence homology between the virus A and B templates. The proportion of products capable of hybridising to the virus B template increased with the amount of virus B dsRNA (ug dsRNA/ssRNA as before). However the level of hybridisation of the ssRNA to the virus B RNA was much lower (Fig.45 b) than that achieved with the equivalent amount of virus A dsRNA and did not reach a saturation level. This is possibly because one of the three dsRNA species, dsRNA 3 of both virus A and virus B, is common to the two viruses. This would allow ssRNA product frcm virus A dsRNA 3 to anneal to the denatured virus B dsRNA 3, but not to unrelated dsRNA species. Thus a lower level of incorporation (of the same amount of labelled product used in hybridisation to virus A dsRNA) would take place into the virus B dsRNA preparation and saturation of the virus A product with the virus B nucleic acid would not take place as was shown. This dsRNA 3 component is unlikely to be a contaminant of the virus B preparation from virus A because of the rigorous purification procedure followed (section 2E) and because no virus A could be detected in virus B 145 preparations by immunodiffusion analysis, p.a.g.e of virus polypeptides and neither of the other two dsRNAs of virus A could be detected in virus B preparations. H. Alterations of dsRNA migration patterns of viruses A and B after repeated fungal subculture The pattern of dsRNA mobility in RNA gels was found to be notably altered (Fig.46) following repeated subculture of the P.sp(lh) 2-2 fungus after a period of three years. No alteration in this property had been noted before this time, when the previous experiments in this section had been completed. There was a gradual decrease in the proportion of virus B in the virus extracts after the first year and this virus showed greater lability, after separation on DEAE cellulose, than virus A, but no alteration was noted in these features after a further six months. The pattern of stained RNA bands of virus A was the same as before, although the proportion of RNA 3 g (molecular weight 1.03 x 10 ) was increased. However it appeared that in virus B, RNA 3 was lost from the virus preparation and an additional component, which comigrated with virus A RNA 1 (molecular weight 1.29 x £ 10 ) when a mixture of the two viruses was analysed on RNA gels, arose in the virus B preparation. The serological and column properties of the two viruses appeared unaltered by this property. 146 Figure 46 Polyacrylamide gel electrophoresis of P.sp.(lh) 2-2 virus A (a) and virus B (b) RNA: alteration of migration pattern after repeated subculture 147 DISCUSSION Analysis of virus extracts frcm Phialophora sp. (with lobed hyphopodia) isolate 2-2 by electron microscopy (demonstrating isometric particles of 35 nm and 40 nm) and by isolation and separation of the viruses, suggested that three viruses were present in this isolate. These were a 40 nm virus which probably encapsidated a dsRNA component of molecular weight 6.1 x 10 , and two 35 nm viruses containing 3 dsRNA components each (virus A dsRNA, molecular weights 1.29 x 10 , 1.22 x 106 and 1.03 x 106; virus B dsENA, molecular weights 1.32 x 106, 1.25 x 106 and 1.03 x 106) Comparison of the methods of dsRNA isolation, by direct extraction of nucleic acids from fungal mycelium (followed by removal of ssRNA and DNA, section. 1) or by nucleic acid extraction from partially purified virus preparations, demonstrated that the yields of dsRNA were essentially the same (approximately 20 ug/g wet weight mycelium). The scale up of virus preparation from shake culture of 2000 ml (100 ml/flask ) to fermenters of 60 litres was entirely satisfactory, allowing purification of virus with the same efficency. Fermenter preparations showed an advantage over shake culture, as a greater yield of fungal mycelium, and hence virus and dsKNA, for the same volume of medium was obtained. The selective loss of the 40 nm virus present in crude virus preparations but lost on clarifying centrifugation and on purification by sucrose density gradient centrifugation, suggested that this virus may have precipitated out of solution, a property shown by a virus, designated 45/9-A, from an isolate of Gaeumannomyces graminis var tritici (Buck et al., 1981 U Alternatively the 40 nm virus may be unstable in sucrose. This also confirms that, of the two methods of dsRNA isolation, direct extraction from mycelium is the better method for screening of isolates for dsRNA content. However the latter method provides no information on 148 the number of viruses present, which would allow assignment of particular dsRNA components to each mycovirus and possible correlations between virus/dsRNA infection and fungus phenotype to be determined. There was some evidence for selective loss of virus B from these preparations also, probably due to two reasons. There was a gradual decline in the proportion of virus B particles in the virus extract, which did not seem due to virus insolubility as it took place over some months, but rather to a decrease in the concentration of the virus within the isolate. Virus B also showed greater instability during the purification procedure than virus A, e.g. as shown by the disruption of virus B particles following DEAE cellulose chromatography and by the gradual 'tailing off' of the virus B polymerase reaction. It was considered possible, in the early stages of the work with the virus particles from isolate 2-2, that either two distinct viruses with the same diameter were present, as has been found in several fungi (Hollings, 1978), or that the virus particles may have been partially degraded by a protease giving rise to two electrophoretic forms, analogous to those of oowpea mosaic virus (Geelen et al., 1973), in agarose gels. The particles were both 35 nm in diameter, had capsid polypeptides of similar molecular weight (66,000 and 60,000) and both ion exchange and electrophoretic properties could be accounted for by this theory. However results obtained with the antiserum raised against both purified viruses, showed that the two viruses were serologically unrelated. Each virus gave rise to a single precipitin line, which crossed over when the two viruses were placed in adjacent wells. The antiserum end titres also differed significantly. This distinction was later confirmed by the demonstration of the distinct dsRNA components and polymerase specificities of the two viruses. The proportions of the dsRNA. components, as measured frcm the area under the peaks in gel scans, were not equal, suggesting that the 3 dsEN& 149 opponents were not enclosed in a single particle, but were probably encapsidated separately as has been found for several other dsRNA mycoviruses (Hollings, 1978). Further evidence to support this view was obtained by p.a.g.e. of dsRNA prepared frcm virus fractions from sucrose density gradients. The slower sedimenting side of the peak contained a higher proportion of the 1.03 x 10 dalton RNA species, whereas the faster sedimenting side contained a higher proportion of the higher molecular weight RNAs. It seemed likely, therefore, that these isometric viruses were type 2 viruses, i.e. those with a genome divided into two or more segments of monocistronic dsKNA, rather than type 1 viruses which have an undivided polycistronic genome (see introduction). Consideration of the size of the dsRNA segments, calculated on the basis that (a) only one strand codes, (b) three nucleotides encode one amino acid (average molecular weight of nucleotide and amino acid residues, 330 and 110 respectively, (c) the entire RNA strand codes (non coding sequences are neglected), gives the following figures for their coding potential (Table 8). Table 8 Calculated coding capacity of virus A and B dsRNA components Mol. wt. of Coding capacity of virion dsRNA _fi dsRNA_components components (x10 ) (x10 Virus A dsRNA 1 1.29 71 2 1.22 68 3 1.03 57 Virus B dsRNA 1 1.32 73 2 1.25 69 3 1.03 57 150 This further outlines the likely monocistronic nature of these segments, as the molecular weights for virus A and B capsid polypeptides are 60,000 and 66,000 respectively and the polymerase enzyme is likely to be encoded by only one additional dsRNA component (The putative RNA polymerases of Penicillium stoloniferum viruses S and F have molecular weights of 55,500 and 59,000 respectively; Buck and Kempson-Jones 1974). It may be expected that, as a minimum, a virus genome will encode a polypeptide for capsid construction and an RNA dependent RNA polymerase for replication. This leaves one component, in both viruses, as a possible candidate for a satellite or defective dsRNA species. The 40 nm virus, lost during the purification procedure, is more likely to be polycistronic, since the large dsRNA species (molecular weight 6.1 x 10 ) appears to have the capacity to code for both capsid polypeptide and replicative function. This is similar to the Saccharomyces cerevisiae virus (40 nm) present in sensitive strains of the fungus and containing a single large dsRNA component (L-dsRNA; molecular 6 —3 weight 3.4 x 10 ). This species has a coding capacity of 189 x 10 3 and encodes a capsid polypeptide of molecular weight 88 x 10 (Hopper et al., 1977). The Phialophora 40 nm virus would therefore be a type 1 virus. Since mycoviruses probably have an entirely intracellular existencer separate encapsidation is probably an advantage, since it allows for variation in the proportions of dsRNA species, which may be inportant in gene expression, allowing gene amplification or suppresssion. It also eliminates the need for selection of each RNA component. In reovirus, on the other hand, with large numbers of dsRNA components and the necessity of infection from outside the cell, the encapsidation of all ten dsRNA components within one virus particle can be seen as a necessity for an efficient infection process. The properties of the Phialophora viruses can be compared 151 with viruses from the closely related fungus Gaeumannomyces graminis. In a recent study of thirteen viruses obtained from eight G.graminis var. tritici isolates (Buck et al., 1981) three groups were obtained based on the physicochemical and serological properties of the viruses (Table 9). Various degrees of serological relationship were found within each group, but not between viruses frcm different groups. On the basis of its physicochemical properties, virus A from Phialophora sp.(lh) isolate 2-2 could be classified with the G.graminis group I viruses. Virus B, on the other hand, has a sedimentation coefficient and dsRNA molecular weight similar to the group I viruses but its polypeptide molecular weight is closer to that of the group II viruses. In gel immunodiffusion tests (K.W. Buck, personal communication) none of the thirteen viruses eleven of which were in groups I and II, reacted with antiserum to the Phialophora viruses A and B, nor did viruses A and B react with any of eight antisera to a total of eleven of the G.graminis viruses. However, the absence of serological cross reactions does not preclude other relationships between these viruses. The antigenic determinants of a virus are encoded by a comparatively small portion of its genome and a fairly small number of mutations may lead to a marked change in its serological properties. However, the serological relationship noted between viruses from Ggt isolate 45/10, which recently have been shown to belong to Groups I and II (K.W. Buck, personal communication), and the Phialophora viruses A and B, suggests that the inclusion of the latter viruses in this mycovirus grouping system is well founded. Studies of the nucleotide sequence homology of the virus dsRNAs, preferably carried out by direct comparison of sequences, may provide a greater insight into possible relationships between the viruses of these two fungal species. Because of the existence of possible satellite and/or defective RNAs in fungal viruses, the number of dsRNA components is probably not a useful property for the grouping of mycoviruses. This is amply exemplified by Table 9 Physicochemical properties of G. graminis viruses Group Virus Particle S^ Mol. wt. of dsRN| Mol. wt. of Electrophoretic diam. (nm)* (Sveaoergs)+ components (x10 ) capsid poly- movements (nw/h) peptides towards anode (x 10 )++ 019/6-A 35 126 1.27,1.19 60 1.25 38-4-A 35 115 1.27,1.19,1.09 55 1.18 01-1-4-A 35 109 1.22,1.14 55 P 45/9-A 35 117 1.30,1.22,1.44,1.11 55 NT OgA-B 35 125 1.30,1.22 55 0.58 3b1a-C 35 115 1.27,1.19,1.11 55 1.10 F6-C 35 128 1.27,1.19 54 2.20 II T1-A 35 133 1.49,1.47 73 P F6-B 35 133 1.60x,1.56,1.45 73 1.10 OgA-A 35 135 1.49,1.39 68 2.40 3b1a-B 35 140 1.60,1.54,1.45,1.43 68 1.38 III 3b1a-A 40 163 4.1,3.5 87,83,78 2.10 F6-A 40 159 4.3,3.2 87,83,78 1.14 * Modal values. The diam. of between 75 and 250 particles of each virus were measured and histograms were constructed of particles of each diameter in 1nm intervals. Clear modes were obtained in every case. Rest of footnotes on next page (reproduced by kind permission of Dr. K.W. Buck) M cn CO 153 to + Measured in PK buffer at very low concentrations (A2gQ O-?)- Values are means of three determinations. In no case did any of the three values differ by more than 2 units from the mean value. ++ All dsRNA and polypeptides have been tested for identity of non- identity of mol. wt. by coelectrophoresis. The dsENA species which g differ in mol. wt. by 0.02 x 10 and polypeptide species which differ 3 in mol. wt. by 1 x 10 have been shown to be separated in gel electrophores is. NT NOt tested because virus was obtained as a precipitate. x This dsFNA component was present in virus isolated from the G.graminis isolate F6 and from nine single microconidial cultures (SM1 to SM7, SM9 and SM10) derived from F6; however, it was absent in virus from one derived microconidial culture (SM8). 154 serologically related viruses in Groups I and II (Table 9) which, contain, two, three or four dsRNA molecules. Fungal isolates containing more than one virus are well documented e.g. in the cultivated mushroom (Barton and Hollings, 1978), Aspergillus foetidus (Ratti and Buck, 1977) and Penicillium stoloniferum (Buck and Kempson-Jones, 1970, Bozarth et al., 1971). Out of the eight Gaeumannomyces graminis isolates studied by Buck et al. (1981) one isolate was found to contain two viruses and two isolates each contained three viruses. As field isolates were used in this study it was uncertain whether all viruses were present in a single cell, or whether these isolates were mixtures of G.graminis strains growing in close asociation, each with one virus. (The same argument applies to the viruses of Phialophora sp. (lh) isolate 2-2 and could be tested in the same way). Examination was made of the virus content of ten microconidial cultures derived from one of the isolates of G.graminis containing three viruses. All ten single microconidial cultures contained the three viruses found in the original isolate, indicating that these viruses were all present together in the same cell. Production of single microconidial isolates derived frcm the Phialophora isolate 2-2 might also indicate whether the gradual decline in virus B presence in the isolate were due to the effect of subculture on the virus or on a strain of Phialophora in a mixed-strain culture. The alteration in the pattern of virus B dsRNA component mobility on RNA p.a.g.e., following repeated subculture of the Phialophora sp.(lh) 2-2 fungus for two years, suggests that dsRNA variability may arise in this virus through deletions in virus genomic dsRNA. In a recent study of defective interfering particles in Saccharomyces cerevisiae (containing fragments of M-dsRNA arising by internal deletion of this molecule) Bruenn and Brennan (1980) recognised five possible selection pressures operating on dsRNA molecules in these particles which might result in 155 sequence changes in the genomic dsRNA. These were replicase recognition of the genome, transcriptase recognition, packaging recognition, coding requirements and ribosomal recognition (L-dsRNA encodes the capsid protein). It is possible, therefore, that selection pressures, arising through continued subculture of the host fungus could result in the deletion or 'editing' of some region of the genomic dsRNA molecules in phialophora virus B. The loss of the dsRNA 3 species and the increase in another species, may have resulted from a deletion arising at a different site in a genomic dsRNA molecule, giving rise to an apparently new dsRNA component in virus B. Alternatively, gene redundancy or amplification of one dsRNA species may result from selective gene duplication in the virus and gene suppression of a second dsRNA species, allowing the concentration of the individual RNA species to rise or fall within the virus. Analysis of the dsRNA species by nucleotide sequence homology would allow this question to be resolved. Virion-associated RNA polymerases have been found in isometric dsRNA viruses of both type 1 and 2 'see Introduction). In Saccharomyces cerevisiae L virus, a type 1 virus, the RNA polymerase in mature virions was found to be a transcriptase, catalysing the synthesis and release of full length ssRNA copies of one of the strands of the dsRNA genome (Herring and Bevan, 1977; Bruenn et al., 1980). Transcriptase activity was also detected in two other mycoviruses with polycistronic dsRNA genomes, Aspergillus foetidus virus S (Ratti and Buck, 1979) and Allomyces arbuscula (Khandjian and Turian, 1977). In contrast, the virion associated RNA polymerase of Penicillium stoloniferum virus S, a type 2 virus, was found to be a replicase, synthesising a new molecule of dsRNA in vitro, giving rise to diploid virions (Buck, 1975). Since the Penicillium virus replicase was the only RNA polymerase of a type 2 dsRNA mycovirus to have been characterised, it was of interest to know if replicase activity was 156 characteristic of other type 2 dsRNA mycoviruses. Both viruses A and B of the Phialophora sp(lh) isolate 2-2 have distinct virion associated polymerase activity, as noted by their different magnesium ion optima and rates of activity. The virus A polymerase was shown to be a transcriptase, which catalyses the synthesis of ssRNA from each of the virus dsRNA components. This activity differs from that of P.stoloniferum virus S and is the first report of transcriptase activity associated with particles of a type 2 dsRNA mycovirus. Another difference is that transcription can be initiated iri vitro by the Phialophora virus enzyme, whereas the Penicillium virus replicase requires the addition of ssRNA primers to the dsRNA template to initiate RNA synthesis. These results suggest the subdivision of type 2 viruses into type 2a, those with virion-associated replicase activity and type 2b, those with vir ion-associated transcriptase activity. 3 The incorporation of [ H]-UMP into dsRNA as well as into ssRNA transcripts during the Phialophora virus A polymerase reaction suggsts that transcription takes place by a semiconservative displacement mechanism. Transcription of Aspergillus foetidus virus S takes place by such a mechanism (Ratti and Buck, 1978) where the products of reaction are full length ssRNA copies of one of the strands of dsRNA-2, which are released from the virus particle. This can be distinguished from the semi- conservative replication reaction in Penicillium stoloniferum virus S (Buck, 1978) where transcripts are retained in the particles. Thus there are two mechanisms for semi-conservative RNA polymerase activity, semi-conservative replication where transcripts are retained and semi- conservative transcription where transcripts are released. Semi- conservative replication has also been found in phage 06 in vivo (Coplin et al., 1975) and it is likely that transcription of f)6 dsRNA by the vir ion-associated RNA polymerase (Partridge, Vidaver and van-Etten, 1978) also takes place by a semi-conservative displacement mechanism. 157 However in reovirus (Silverstein et al., 1976) transcription is completely conservative. Both virus A and virus B contained a dsRNA of molecular weight 1.03 x 10 which was not due to contamination of one virus by the other. The occurrence of a dsRNA of the same molecular weight in the two viruses was possibly due to genomic masking, the encapsidation of FNA components of one virus by the capsid of another. In Phialophora viruses A and B all the dsRNA components of each virus were in the same size range, so that genomic masking would not be prevented by gross differences in genome size. This theory was lent credence by the degree of homology found between the viruses, probably due to dsRNA 3, in nucleic acid hybridisation tests. No genomic masking has been observed in mixed infections with serologically unrelated fungal viruses before, although it was originally thought that one RNA component was common to both Aspergillus niger viruses S and F (Buck et al., 1973). However this was later disproved when it was found that the dsRNA components had identical molecular weights but different base composition (Buck, 1977). Phenotypic mixing (mixing of one virus polypeptide in another virus capsid) and genomic masking usually occur readily in vivo between related viruses and genomic masking, but no phenotypic mixing has been reported between structurally dissimilar unrelated bacterial viruses, animal viruses and plant viruses (Dodds and Hamilton, 1976). Testing of the dsRNA 3 species from phialophora viruses A and B by nucleotide sequence analysis would establish whether genomic masking has occurred. However two other possibilities exist for the presence of the same RNA 3 in the Phialophora viruses A and B. Firstly the RNA 3 species may be a variant or deletion mutant of RNAs 1 or 2 of one virus. However this is unlikely, since genomic masking of the RNAs 1 and 2 by the capsid of the other virus would also be expected in such a case. Secondly the 158 dsRNA 3 species may be a satellite RNA (Mossop and Francki, 1978). The possible function of such a satellite species includes encapsidation of an RNA polymerase for its own replication (selfish "dsRNA"), a toxic protein analogous to the killer factors of Saccharomyces cerevisiae and Ustilago maydis and plasmid-coded bacteriocins of enteric bacteria, or proteins with other functions. However if dsRNA 3 is a satellite RNA then the phenomenon is not genomic masking but merely encapsidation of the same satellite molecule in two different capsids. 159 SECTION 3 Isolation of virus particles from Ggt isolate 45/10 RESULTS Several attempts were made to purify viruses from Ggt isolate 45/10 to provide a marker system containing twelve dsRNA components of known molecular weight (Table 10; Almond, Buck and Rawlinson, 1977), to facilitate the comparison of the molecular weights of dsRNA components frcm other species. This was never successfully achieved, for the reasons outlined below, but the procedure finally used to purify the Ggt 45/10 viruses provided the basis for purification of viruses of Ggg isolate G1 and Pc[ isolate 1348-2. A. Comparison of two media for fungal growth prior to virus preparation by the PEG method i. Basal and CSL medium Mycelium grown for 5 days was harvested by filtration and gave a yield of 4.6 g/100 ml medium. Crude virus was prepared by lysis of mycelium (69g), PEG precipitation and high speed ultracentrifugation (G.M. 2a) in P buffer and virus pellets were resuspended in the same buffer (3 ml). Samples of crude virus were examined by RNA p.a.g.e. (G.M.16c.i) and were found to contain 5 dsRNA components (when diluted 1:10, only 2 dsRNA components were visible) when stained with toluidine blue. When samples of the crude virus preparation were purified by sucrose density gradient centrifugation (10 to 45% w/v sucrose in P buffer) a narrow peak of virus dsRNA was detected in the centre of the gradient by ethidium bromide-agarose plate analysis. However, on analysis by RNA p.a.g.e. only two dsRNA components were visible in gels stained with 160 toluidine blue and two polypeptide components were visible on SDS/ polyacrylamide gels stained with Coomassie blue R stain. This indicated that the virus had been well purified as 13 polypeptide components were found in the crude virus preparation on SDS p.a.g.e. analysis. An attempt was made to concentrate the virus sample by dialysis against 20% w/v PEG 6,000, but this failed to increase the number of dsRNA bands stained in the gels. The same result was found when a larger volume of crude virus was purified by centrifugation in a 50 ml sucrose gradient. No further dsRNA bands were detected by nucleic acid extraction and concentration from the purified virus preparations either. The molecular weights of the dsRNA species in the crude virus preparation, determined using Penicillium chrysogenum virus dsRNA and Penicillium stoloniferum virus S dsRNA as internal markers, were 1.57 x 106, 1.39 x 106, 1.23 x 106, 1.14 x 106 and 0.98 x 106 (Fig.47). The molecular weights of the polypeptide components determined using the internal markers phosphorylase a, bovine serum albumin, glutamate dehydrogenase and lactate dehydrogenase, were 55,000 (comigrated with glutamate dehydrogenase) and 73,000. ii. Weste and Throwers medium Growth was much slower in this medium and after 22 to 23 days a yield of only 1.3 to 1.9 g/100 ml was obtained. Crude virus was prepared from mycelium (26 to 38g) as before. Virus pellets were resuspended in 1ml of P buffer. Preparations of crude virus, when analysed by RNA p.a.g.e. showed some variation in dsRNA pattern. Four or six bands were visualised in the gels by staining with toluidine blue (Fig.48). By comparison of these RNA migration patterns with that expected from the full complement of Ggt 45/10 viruses (Almond, Buck and Rawlinson, 1977; see Table 10), it was shown that extraction of viruses 161 RNA 1,4 or 5,8,9 and 11 Direction of migration Figure 47 Polyacrylamide gel electrophoresis of Ggt 45/10 virus RNA from a crude virus preparation: the isolate was grown on basal and CSL medium RNA 1,5,6,7,9 or 10 and 11 Direction of migration > Figure 48 Polyacrylamide gel electrophoresis of Ggt 45/10 virus RNA from a crude virus preparation: the isolate was grown on Weste and Throwers medium Polypeptides 1,2,3,5 Direction of migration » polypeptides 1 and 5 Figure 49 Polyacrylamide gel electrophoresis of Ggt 45/10 virus polypeptides : isolate grown on basal and CSL (a) or Weste and Throwers (b) media 162 RNA component Molecular ^ Polypeptide Molecular ~ weight (x 10 ) component weight (x 10 ) * RNA 1 1.56 PI 72 * RNA 2 1.53 P2 67 * RNA 3 1.52 P3 63 RNA 4 1.43 P4 60 * RNA 5 1.38 P5 52 RNA 6 1.33 RNA 7 1.29 RNA 8 1.24 RNA 9 1.22 RNA 10 1.15 RNA 11 1.11 RNA 12 1.01 * Polypeptide 1 encapsidates RNA 1 Polypeptide 2 encapsidates RNA 3 Polypeptide 3 encapsidates RNA 7 Polypeptide 5 encapsidates RNA 5 Table 10 Molecular weights of the dsRNA and polypeptide components of Ggt 45/10 viruses 163 from basal and CSL medium isolated RNA components 1, 4 or 5, 8, 9 and 11, and fran Weste and Throwers' medium, RNA components 1, 5, 7 and 11 in one preparation and components 1, 5, 6, 7, 9 or 10 and 11 in the other. From the polypeptide analysis it is likely that polypeptides 1 and 5 (encapsi- dating RNAs 1 and 5 respectively) are present in the basal and CSL purified virus preparation, while four polypeptide species were present in the Weste and Thrower preparations, probably polypeptides 1,2, 3 and 5 (Fig.49). There were two possible explanations for the difficulty experienced in preparing large quantities of virus, with the full complement of dsRNA components expected. Firstly, it was possible that the Ggt 45/10 virus population had changed in culture, as a result of repeated subculture made from the original stock. This could either be due to an alteration in relative proportions of certain strains, harbouring different viruses, in a mixed-strain isolate of the fungus, or by a change in the proportion of individual viruses. B. Comparison of four different buffers for virus preparation by the PEG precipitation method and by direct pelleting. In order to attempt to overcome the problems described in the previous Section, the following approaches were used: a. A PDA slope of Ggt 45/10 which had been subcultured only once from the original culture, was used as a source of inoculum. This was to minimise any changes due to repeated subculture. b. Two different methods of virus preparation employing four different extraction buffers were compared, in order to find a system which minimised virus loss due to insolubility or adsorption on host debris (Fig.50). Basal and corn steep liquor medium was chosen for liquid culture of the mycelium since this had been shown to produce the greater yield of mycelium (part A), ffycelium was cultured at 24°C for 13 days and the wet weight obtained was 220 g (7.3 g/100 ml). 164 Ggt 45/10 grown for 13 days in basal and CSL medium. Yield 7.3 g/IOOml) Mycelium resuspended in extraction buffer P buffer Bor buffer Ver buffer Direct pelleting PEG precipitation of of virus from virus from mycelial mycelial homogenate homogenate Resuspension in Resuspension Resuspension extraction buffer in equivalent in 1/10 volume volume of Of buffer used buffer used in in direct direct pelleting pelleting RNA. analysis - i iRNA analysi/ s Clarification •1 Clarification RNA analysis RNA analysisI . Fig.50 Investigation of methods of virus isolation from Ggt isolate 45/10 165 i. Direct pelleting method. Four samples of Ggt 45/10 mycelium (6.5 g wet wt. each) were homogenised and each taken up into one of four buffers as shown in the flow diagram (Fig.50, 120 ml buffer). Following low speed oentrifugation to clarify the preparation, virus was directly pelleted from the supernatant by ultracentrifugation at 20,000 rev/min for 15h. Pellets were each resuspended in the relevant buffer (1.5 ml) and samples taken for RNA analysis. Further samples were taken after clarifying the preparations by centrifugation. Results are shown in Fig.51. Direct pelleting of virus appeared to yield more RNA components than previous PEG precipi- tation procedures had done. By comparison with the known migration pattern of Ggt 45/10 virus dsRNA on gels, the RNA components present in the four preparations were determined and the approximate yield estimated from the relative intensities of the stained bands (Table 11). The greatest amount and number of dsRNA components were present in the veronal buffer preparation, indicating that this was the most efficient extraction buffer (extracting viruses containing RNA components 1, 2, 5, 6, 7, 9 and 11). However, following clarifying centrifugation, none of these bands were detectable in the gels, indicating that these viruses were insoluble in the buffer but were probably present in sufficiently high concentration, following pelleting of virus for the RNA to be detected. The viruses were then centrifuged down with the debris material (15,000 rev/min, 10 min). 90% or more of virus was lost from the preparations, as judged by the decrease in band intensity on stained gels. Moreover virus RNA bands were only visible in the PK preparation (RNAs 5, 6, 7 and 9) and borate preparation (RNA 9). ii. PEG precipitation method Four samples of mycelium (48.5 g wet. wt. each) were homogenised and 166 Buffer P RNA 5,6 7,9,11 PK RNA 5,6,7,9,10,11 RNA 5,6,7,9 (a) Bor RNA 1,5,6,7,9,11 RNA 9 ? Ver RNA 1,2,5,6,7,9,11 Direct pelleting RNA 1 ! PK RNA 1,2,3,5 or 6 RNA 1,3,5 or 6 (b) Bor RNA 1,2,5 m 1:2,5 Ver PEG-equivalent volume RNA 1,2,3 RNA 1 j PK RNA 1 ,i2,3, 5or6, 7or8,9 RNA 1,2,3,5or6,7or 8 Bor RNA 1,2,3,5 M 1/2,3,b Ver RNA 1,5 " KNA '-r'rr—,,J' PEG-1/10 volume Figure 51 Polyacrylamide gel electrophoresis of Ggt 45/10 virus RNA obtained by the direct pelleting method (a) or PEG precipitation methods (b) and (c) (Migration was from left to right) 167 Extraction RNA components Approximate yield buffer present in the preparation before cleaning after cleaning P buffer RNA 5 + 6 ++++ - 7 ++ - 9 +++ - 11 ++ - PK buffer RNA 5 ++++ ++ 6 ++++ + 7 +++ ++ 9 +++ ++ 10 + - 11 +++ - Bor buffer RNA 1 + - 5 + - 6 ++++ - 7 +++ - 9 ++++ ?+ 11 ++ - Ver buffer RNA 1 + - 2 + - 5 ++ - 6 ++++ - 7 +++ - 9 ++++ - 11 +++ - Estimated from relative intensity of bands,after p.a.g.e., stained with toluidine blue Table 11 RNA components extracted from Ggt 45/10 viruses,prepared by the direct pelleting method 168 taken up into one of the four buffers (905 ml). Each preparation was given a cleaning spin and was subjected to PEG precipitation and ultra- centrifugation (G.M.2a). Two pellets were obtained from each preparation, one of which was resuspended in an equivalent (ml/g) volume of buffer to that used in direct pelleting and the other in approximately one-tenth of this volume, the latter step used to gauge the effect of buffer volume on solubility. Results are shown in Table 12. Solubilisation of virus pellets in an equivalent volume to that used for directly pelleted virus, showed that fewer RNA components were extracted in the preparation. However, rather than decreasing the concentration of virus, and therefore RNA in the preparation, a decrease in the volume of extraction buffer used to resuspend the virus, allowed more concentrated virus preparations to be obtained. Centrifugation of the preparations to clarify them did not result in the dramatic decrease in RNA components found in directly pelleted preparations, although some RNA components were lost. In general, however, results with the four buffers correlated with those obtained by the direct pelleting method, with PK and borate buffers the most efficient at virus solubilisation. Direct pelleting was the more efficient process for the extraction of virus from mycelium. However, the solubility of viruses in these preparations remained low, probably due to the large amount of debris present, to which virus particles may have adhered. PEG precipitation was the more efficient method for preparation of crude Ggt 45/10 virus following the final clarifying spin. This was probably because most of the mycelial debris was removed at the earlier stages (clarification, precipitation, clarification) of the preparation. Sucrose density gradient centrifugation resulted in purfication of virus containing RNAs 1, 2, 3 and 5 from both pK and borate preparations of PEG-precipitated virus, indicating that the solubility of these viruses remained relatively constant in these buffers. 169 Extraction RNA components Approximate yield buffer present in the preparation before cleaning after cleaning equivalent 1/10 equivalent 1/10 vol. vol. vol. vol. P buffer RNA 1 ++ +++ — ++ 2 - ++ - - 3 - +++ - - PK buffer RNA 1 +++ ++++ + ++ 2 +++ +++ - +++ 3 +++ ++++ + ++ 5 or 6 +++ ++++ + +++ 7 or 8 - ++ - ++ 9 - +++ - - Bor buffer RNA 1 ++ +++ ++ +++ 2 ++ +++ + +++ 3 - ++ - ++ 5 ++ ++++ + +++ Ver buffer RNA 1 + + - + 5 + +++ — + Equivalent buffer volume / mycelium ratio as used for direct pelleting method Table 12 RNA components extracted from Ggt 45/10 viruses, prepared by the PEG precipitation method 170 Thus, while the final yield of RNA components remained low, valuable information on the solubility of these viruses in extraction buffers was obtained and proved useful in the extraction of viruses from P.sp.(lh)2-2, Ggg G1 and P^ 1348-2. Samples of Ggt 45/10 crude virus preparations (extracted by PEG precipitation in borate and PK buffers) were tested against samples of P.sp.(lh)2-2 virus antiserum, and were found to cross react. TWo precipitin lines formed (in both cases), suggesting that the Ggt 45/10 virus preparation contained antigenic determinants related to the P.sp(lh)2-2 viruses A and B. 171 DISCUSSION Variation in the number and relative intensity of Ggt 45/10 RNA bands on polyacrylamide gels is undoubtedly due mainly to the loss of individual viruses through their poor solubility in the extraction buffers. Comparison of the methods of PEG precipitation with direct pelleting of virus indicates that although the former method involves initial loss of yield through the number of steps involved i.e. small losses undoubtedly occur in clarifying spins and during the PEG precipitation process, the final yield obtained, particularly in PK and borate buffers, is relatively high and remains roughly constant through the sucrose density gradient purification procedure. Conversely, direct pelleting is more rapid but has a serious drawback in that the solubility of viruses is further impaired by the high concentration of mycelial debris present to which the viruses may also adhere. It is likely with both these methods, that the majority of virus is lost in the initial clarifying spin, immediately following homogenisation. The importance of the choice of buffer is evident however frcm the observation that PK and borate buffer preparations contained more RNA components than the other buffers used in either method. However the proportions of these components varied between the buffer preparations, suggesting that different viruses were better solubilised in each, e.g. following direct pelleting, the virus containing RNA 5 is resuspended more efficiently in PK buffer than in borate buffer, while the virus containing RNA 1 is resuspended in borate buffer but not apparently in PK buffer. It is noteworthy that the same viruses are not necessarily isolated in the same buffer, when the different methods are used. Undoubtedly the direct pelleting method gives a fuller picture of the virus content of the isolate, while the PEG precipitation procedure appears to show some selectivity in the viruses precipitated from solution. 172 Despite the physical losses of virus through solubility problems in these preparations, it is likely that the viruses in this isolate have altered since first isolated from Ggt 45/10, since only a total of four polypeptides and ten RNA components were visualised in RNA gels. (Five polypeptides and twelve RNA components were noted in preparations of Ggt 45/10 viruses made 1 year previously (Almond, 1979)). There are a number of possible explanations for this behaviour. Firstly, there may have been an alteration in the growth of one or more individual isolates in a culture of mixed individuals, resulting from continued storage in culture of the fungus (since inoculum was taken from mycelium that had been subcultured only once in the later experiments). Since the resulting culture could contain a different population of fungal strains to that originally present, through persistence of some, but not all, of the strains in the culture, so too could it contain a different virus population. As discussed in section 2, the possibility of culturing germinable conidia of these fungi (Rawlinson, Muthyalu and Deacon, 1977), which are capable of transmitting virus, would allow the distinction between a single isolate containing a mixture of viruses in a single cell and an isolate consisting of a mixture of G.graminis strains, growing in close association and each containing one virus, to be made (as was done for F6 viruses, Buck et al., 1981). This analysis would also make it clear whether the second possibility, of a selective decrease in one or more of a mixture of viruses infecting the fungus, were valid. Finally, because of their intracellular transmission in these fungi, the dsRNA rnycoviruses are able to encapsidate the segments of their dsRNA genome separately without creating problems for virus infectivity, but allowing the possiblity of gene selection and amplification. In this case it is possible that individual dsRNA components may vary within a virus, and the levels of these components found to be different on analysis after some time in 173 culture. This appears to have been the case in P.sp. (lh)2-2 virus B (section 2) and in the defective interfering virus particles of yeast (Bruenn and Brennan, 1980) and Vesicular Stomatitis Virus (Holland et al., 1979). The production of two precipitin lines in Ouchterlony tests between Phialophora sp.(lh)2-2 antiserum and the Ggt 45/10 crude virus preparations, suggests that there may be two viruses in Ggt 45/10, related to 2-2 viruses A and B. It is noteworthy that polypeptide 4, a capsid polypeptide of one of the four 45/10 viruses, comigrates with pyruvate kinase (Almond, 1979) as did the capsid polypeptide of P.sp(lh)2-2 virus A. Moreover, 45/10 polypeptide 3, with a molecular weight of 65,000 (close in molecular weight to the virus B polypeptide) has recently been shown to encapsidate RNA 7 (K.W. Buck, personal communication) with g molecular weight 1.25 x 10 (the same molecular weight as virus B RNA 2). Thus it is likely that there are two viruses present in the 45/10 complex which are related to the P.sp.(lh)2-2 viruses. Identification of the viruses involved awaits separation of the 45/10 viruses. It is evident nonetheless that serological relationships exist between the viruses of the Gaeumannomyces - Phialophora complex. 174 SECTION 4 Isolation and properties of virus particles from Ggg isolate G1 RESULTS A. Preparation of crude virus The isolate was grown in primary liquid culture in basal and corn steep liquor (G.M.1.) for 10 days. The mycelium was intact on examination for natural (auto) lysis under the light microscope and was harvested by filtration. The wet weight of mycelium obtained was 8.9 g/100 ml medium. Mycelium was homogenised in PK buffer (20 ml/g wet wt. mycelium) in the Manton-Gaulin hcmogeniser (PK buffer had proved more efficient than P buffer in the solubilisation and purification of Ggt 45/10 virus, section 3). Debris was removed from the preparation by centrifugation and virus was extracted from the supernatant by PEG precipitation and high speed ultracentrifugation (G.M.2). Virus pellets were resuspended in PK buffer (0.5ml) and samples were taken for RNA analysis before and after a final clarifying spin (Fig. 52). Cleaning of the virus preparation by centrifugation (15,000 rev/min 15 min) did not appear to affect the number or relative proportions of the dsRNA species when analysed on RNA gels. The yield in the crude virus preparation was estimated from the intensity of band staining of gels with toluidine blue, to be 15 to 20 mg virus (110 to 150 ug/g wet wt. mycelium). B. Purification of virus by sucrose density gradient centrifugation Samples of the crude virus preparation were purified by centrifugation in sucrose density gradients (10 to 45% w/v sucrose in PK buffer, 5 ml) 175 RNA components 1, 2, 3, 4 and 5 Figure 52 Polyacrylamide gel electrophoresis of Ggg G1 virus RNA from a crude virus preparation (a) RNA components 1, 2, 3, 4 and 5 1 (b) Polyacrylamide gel electrophoresis of Ggg G1 virus RNA (a) and polypeptides (b) in fractions 15 to 17, following sucrose density gradient centrifugation 176 at 45,000 rev/min for 1.5 h. Gradients were fractionated and samples from each fraction analysed by spotting on to an ethidium bromide-agarose plate. Fractions from the centre of the gradient, showing evidence of nucleic acid content, as shown by the fluorescence resulting from illumination with U.V. light (254 nm) were dialysed against PK buffer. Samples of the purified virus preparation were examined by electron microscopy to confirm the presence of virus particles. There was no evidence of fractionation of the polypeptide or RNA species across the sucrose gradient, when fractions were analysed by RNA p.a.g.e. Five RNA species and three polypeptides were visible in fractions at the centre of the peak (fractions 15-17; Fig. 53). C. Purification of virus by caesium chloride density centrifugation. Samples of crude virus preparation were loaded on to a solution of caesium chloride in PK buffer (p = 1.413g/ml, 5ml) and were centrifuged at 40,000 rev/min for 62h. The gradients were fractionated (0.2ml fractions) and immediately following fractionation samples were taken from each tube for measurement of refractive index (G.M.11). A second series of samples were then taken for examination by ethidium bromide- agarose plate analysis. Those fractions showing evidence of nucleic acid content (Fig. 54) were dialysed against PK buffer. The ethidium bromide- nucleic acid peak was shown to be contained in three fractions corresponding to a density range of 1.362 g/ml to 1.394 g/ml, with the central fraction corresponding to a density of 1.377 g/ml. D. Properties of purified virus particles a. U.V. spectrum The U.V. spectrum of a Ggg G1 virus preparation (similar vtfiether prepared by sucrose or caesium chloride density gradient centrifugation) is shown in Fig. 55. The A260/A280 ratio was 1#5Q# 177 Figure 54 Ethidium bromide-agarose plate analysis of Ggg G1 virus in fractions from a caesium chloride gradient 178 wavelength of u.v. light (nm) Figure 55 U.v. spectrum of Ggg G1 virus 179 b. Particle morphology Examination of the pure virus preparation in the electron microscope showed that the majority of the purified particles were intact (Fig. 56). A histogram constructed to determine the distribution of size in the particles showed that the virus contained only one size class of particle with a modal diameter of 35 nm (Fig. 57). c. Agarose gel electrophoresis Examination of the virus in agarose gels showed that only one virus band was visible (Fig. 58) after staining with Coomassie blue R. d. Serology Antibody to Ggg G1 virus was raised in a rabbit by injection of a pure virus preparation (G.M.5) and the antibodyantigen reactions between samples of pure virus (in PK buffer, A2^q = 0.9, 20 ul/well) and antiserum (diluted serially with PK buffer) were analysed by Ouchterlony double-diffusion tests (G.M. 6) Results show a single precipitin line (Fig. 59) which developed over 24h, with an antiserum end titre of 1:128. Ggg G1 virus did not react with antisera raised against the P.sp. (1h) 2-2 viruses A and B, a mixed P.sp. (lh) virus preparation, Pc[ virus and six Ggt virus preparations. P.sp. (lh)2-2 viruses A and B did not react with antiserum raised against Ggg G1 virus. e. Sedimentation coefficent Velocity sedimentation of a sample of the virus in P buffer (&260 - 0.8, 0.5 ml sample) was performed in the Model E ultracentrifuge. A 20 scan of the moving boundary is shown in Fig. 60. SQ of the G1 virus was found in two determinations, to be 141.8 S and 144.3 S (average 143 S). 180 (x 100,000 magnification) 181 ed ft fH o 0 32 33 34 35 36 37 38 39 Particle diameter (nm) Figure 57 Histogram plot of particle diameters measured randomly from electron micrographs of Ggg G1 virus 182 Figure 59 Irimmodif fusion test of Ggg G1 virus with antiserum raised against this virus -» « 184 f. Buoyant density The buoyant density was determined by allowing samples of the virus to equilibriate in caesium chloride density gradients (starting densities in three cells, 1.354g/ml, 1.372g/ml and 1.408 g/ml) centrifuged at 30,000 rev/min for 22h. The trace obtained in the third of these cells is shown in Fig. 61. The buoyant density of Ggg G1 virus was found to be 1.382g/ml. g. RNA components The RNA components of the pure virus preparation were separated by RNA p.a.g.e. (G.M.16c.i.) with both external and internal dsRNA markers. Gel scans were taken of these gels (G.M.4). to facilitate the calculation, of molecular weights of dsRNA species and to demonstrate their relative proportions in the preparation (Fig. 62). Molecular weights of the dsRNA. fi fi fi components were found to be 1.63 x 10 , 1.54 x 10 , 1.29 x 10 , 6 6 1.17 x 10 and 1.16 x 10 by interpolation frcm a plot of log molecular weight versus mobility of the internal markers (Fig. 63). These components were found to be dsRNA by the relative resistance to RNAse A at 2 x SSC and susceptibility to the enzyme at 0.1 x SSC of a nucleic acid extract of the virus preparation. h. polypeptide components Coelectrophoresis of pure virus polypeptides with markers (G.M.16 d) on SDS p.a.g.e. gave values of 64,500, 58,200 and 49,100 for the polypeptides, when measured from a plot of log mol. wt. versus mobility (cm/h) for the internal markers (Fig. 64). It was not possible to determine the degree of purification by comparison of crude virus and pure virus polypeptide patterns on SDS p.a.g.e. since the pattern of 3 polypeptides, 1 major, the other two of the same intensity, remains unaltered on purification. (The crude virus preparation appeared very Figure 61 Photoelectric scanner trace for the determination of buoyant density of Ggg G1 virus 186 2.0 2.5 3.0 Distance in gel (cm) Figure 62 Polyacrylamide gel profile of RNA released from Ggg G1 virus 187 0.9 0.2 0.4 0.6 0.8 Mobility (cm/h) • RNA standards in order of increasing mobility O Ggg G1 virus RNA components as numbered Figure 63 Relationship between log molecular weight and electrophoretic mobilities of Ggg G1 virus RNA components 188 Mobility (cm/h) • Protein standards in order of increasing mobility O Ggg G1 virus polypeptides as numbered Figure 64 Relationship between log molecular weight and electrophoretic polypeptide components 189 clean, when seen by eye, following the final clarifying spin, although light scattering material was obviously present. "Dirty" preparations frequently contain grey material, probably mycelial debris, which is usually spun out in cleaning spins). Virus preparations were further analysed by electrophoresis on agarose gels, as in part D.c. Following electrophoresis, 1 of 3 identically treated gels was stained with Coomassie blue R to locate the virus and a sharply defined band was obtained. The other two gels were sliced to obtain the virus-containing region. Slices were then boiled in SDS/ mercaptoethanol (G.M.16 e.) and were analysed by SDS p.a.g.e. for polypeptide. The same three polypeptide species, as where found in the pure virus preparation analysed directly, were found following isolation of the sharp virus band on agarose gels. The proportion of the polypeptide components remained unaltered. A summary of the physical characteristics of Ggg isolate G1 is given in Table 13. Table 13 Physicochemical characteristics of Ggg G1 virus Particle diameter (nm) 35 20 S0 143S Byoyant density in CsCl (g/ml) 1.382 zr Mol. wts. of dsRNA components RNA 1 1.63 x 10 RNA 2 1.54 x 106 RNA 3 1.29 x 106 RNA 4 1.17 x 196 RNA 5 1.16 x 106 Mol. wts. of polypeptide P1* 64,500 components P2 58,200 P3 49,100 * The major polypeptide species 190 DISCUSSION Comparative gel analysis of dsRNA components extracted from Ggg isolate G1 showed that a higher yield of dsRNA was obtained in nucleic acid extracted from partially purified virus preparations (26 ug/g wet wt. mycelium) than in that directly extracted frcm fungal mycelium (16 ug/g wet wt. mycelium) and also demonstrated an extra dsRNA component than had been noted during screening of the dsRNA components. This apparent contradiction is explained by the finding that, in the year following the screening procedure, the concentration of G1 virus in the fungus increased (from an estimated 80 ug/g wet wt. in the screening procedure to 130 ug/g wet wt. in the virus extraction) so bringing the dsRNA component (RNA 5) up to detectable level by the RNA p.a.g.e. procedure. The purification of this virus proceeded with none of the problems of solubility which had been noted with the Phialophora isolate 2-2 40 ran virus, the Ggt 45/10 viruses and the Phialophora graminicola virus (section 5). Purification of the virus by sedimentation in caesium chloride and sucrose density gradients showed that the former method produced a much sharper and more concentrated virus peak, with no evidence of virus instability in caesium chloride. (P.sp. (lh) 2-2 virus B occasionally showed signs of instability in buoyant density tests). There was no sign of fractionation of dsRNA or polypeptide species in fractions across the gradient, indicating that these species sedimented together. The presence of a single virus was confirmed by the production of a single precipitin line in Ouchterlony tests, with antiserum raised against the G1 virus. The virus showed no sign of heterologous reaction to antisera raised against the Phialophora sp. (lh ) viruses A and B, Phialophora graminicola virus (section 5) and the Ggt viruses (Table 9). Nonetheless the same 191 arguments for possible relationships between viruses not showing serological relatedness, outlined in the discussion to section 2, still apply and since this fungus is closely related to Gaeumannomyces graminis var. tritici. a comparison between the G1 virus and those in groups I to III (Buck et al., 1981) can be made. The size (35nm), sedimentation coefficient (143S) major polypeptide component (mol. wt. 64,500) and two c c largest dsRNA components (mol. wt. 1.63 x 10° and 1.54 x 10°) suggest that this virus belongs to group II. Despite the fact that, of the G.graminis viruses analysed to date (Buck et al., 1981) all seem to fall between quite distinct grouping ranges, as more viruses are found it is possible that these distinctions (particularly those between virus particles of the same diameter) will become less acute and "intermediate" groups will arise. Nonetheless there appears to be quite striking conservation of polypeptide molecular weight within these groups of Ggt viruses and the serological relationships shown between viruses in a group but not between groups, suggest that those in a group may be variants of a single virus. The intracellular persistence of mycoviruses in their fungal hosts creates an ideal opportunity for the survival and selection of virus mutants, so that the virus found in a particular isolate of G.graminis may be that which replicates most efficiently against the genetic background of the host. Four possibilities existed for the presence of three polypeptides in this virus system. Firstly, it was possible that the three polypeptides were present as capsid molcules of three distinct viruses. However no evidence for the presence of more than one virus was found by serological analysis, agarose gel electrophoresis, fractionation of RNA and polypeptide species across density gradients or in buoyant density determinations. Secondly, it was possible that one virus was present containing three structural polypeptides (which however would exclude it from the Gaeumannomyces graminis virus group II (Buck et al., 1981). Thirdly, 192 three polypeptides may have been demonstrated in one virus, due to the degradation of one species to form two additional polypeptides. However, analysis of the virus in agarose gel electrophoresis failed to show any evidence for breakdown of the virus capsid, which would undoubtedly have been detected in view of the large molecular weight differences involved. Moreover the finding of the same three polypeptides (and in the same proportion) following polypeptide analysis of a single virus band in agarose gels, indicated that the three proteins were true virus polypeptides. Finally, it was possible that the two additional species were RNA polymerase components, required for the replication and transcription of the virus. However, this is unlikely as in other dsRNA mycoviruses known to have associated RNA polymerase activity e.g. Penicillium stoloniferum viruses S and F (Chater and Morgan, 1974) and Aspergillus foetidus virus F (Ratti and Buck, 1975) one molecule of RNA polymerase is found per virus particle, indicating that the concentration of polypeptide demonstrated here is too large as the enzyme would not normally be visible on polypeptide gels when present in this proportion. The inequal proportions of the five dsRNA components in the virus indicate that the G1 virus is a type 2 virus, encapsidating each dsRNA component in a separate capsid composed of one major (molecular weight 64,000) and two minor (molecular weights 58,200 and 49,100) polypeptides. Calculations based on.the size of the dsRNA components in this virus, making the assumptions outlined in section 2, presented the following figures for the coding capacity of G1 dsRNA (Table 14). Table 14 Calculated coding capacity of G1 virus dsRNA Mol. wt. dsRNA Coding capacity component (x10 -6 dsRNA ..component (x10 ) G1 virus dsRNA 1 1.63 90 2 1.54 86 3 1.29 72 4 1.17 65 5 1.16 64 193 As the three G1 polypeptides would require most of the coding capacity of three of these dsRNA components (which are likely to contain non-coding sequences) it is possible, assuming only one dsRNA component is required to oode for a polymerase enzyme, that a satellite or defective dsRNA is present. The use of nitrosoguanidine to make partial- genome mutants, as shown by Koltin (1977), would seem a very useful technique to determine whether defective or satellite dsRNAs, i.e. those not required for virus capsid synthesis or replication, were present. Fingerprinting analysis of dsRNA components and in vitro protein synthesis from these species, would also help to distinguish these possibilities. The study of a population of natural variants of these viruses, cf. the Gaeumannomyces graminis viruses (Buck et al., 1981) would also allow the detection of some of these molecules. 194 SECTION 5 Isolation and properties of virus particles from Pg isolate 1348-2 RESULTS A. Preparation of crude virus The isolate Pg_ 1348-2 was grown in basal and corn steep liquor primary culture (G.M. 1) for 11 days and the wet weight of mycelium obtained was 7.8 g/100ml. The mycelium was homogenised in PK buffer (10 ml/g wet wt.) and virus was precipitated frcm the clarified supernatant with PEG. Virus was pelleted by ultracentrifugation and was resuspended in PK buffer (0.5 ml). Samples were taken for RNA analysis by p.a.g.e., before and after a final cleaning spin (15,000 rev/min, 15 min). Four components staining with toluidine blue were present in the crude preparation (Fig. 65). However the concentration of these components, as judged by the toluidine blue staining, was much reduced following the clarifying centrifugation. It was thought possible that this preparation was affected by the problems of virus solubility (as noted with Ggt 45/10 viruses, section 3). Therefore the extraction of virus in the four buffers used in Ggt 45/10 virus extraction was compared. A second mycelial culture was set up in basal and corn steep liquor as before. Mycelium was harvested after 7 days growth at 24°C, with a wet weight yield of 9.4 g/100 ml medium. The mycelium was divided into four equal parts, which were each suspended in one of the four buffers (250 ml). Each preparation was subjected to homogenisation, clarification, PEG precipitation and ultracentrifugation as before. Samples were removed for RNA p.a.g.e. analysis before and after a final cleaning spin (12,000 rev/min, 30 min). Similar amounts of virus were extracted by all 195 Direction of migration Figure 65 Polyacrylamide gel electrophoresis of Pg virus RNA: the isolate was grown in basal and CSL medium Buffer before cleaning after cleaning before cleaning PK after cleaning before cleaning Bor after cleaning before cleaning Ver after cleaning Figure 66 Polyacrylamide gel electrophoresis of RNA extracted from Pg 1348-2 virus purified in four different buffers 196 four buffers (Fig. 66), as measured by the intensity of band staining on the gels. However, following the clarifying spin, only PK and veronal buffers showed evidence of virus dsRNA content, with the loss of the RNA component with the lowest molecular weight (probably because of its relatively low proportion to the other components, not through selective loss of this species). There was considerable reduction in the intensity of staining of these bands, indicating a decrease from approximately 4 ug/g wet weight of mycelium in the PK buffer and 2.7 ug/g in the veronal buffer, down to 1.3 ug/g in both cases. When attempting to purify this virus (from PK and veronal buffer preparations) by sucrose density gradient centrifugation (0.15 ml sample, 10 to 45% w/v sucrose in PK buffer, 5 ml) at 45,000 rev/min for 1.5 h, no virus dsRNA components were found in fractions distributed centrally in the gradient. Instead virus sedimented through to the bottom of the gradient. This experiment was repeated many times and confirmed by the existence of appropriate dsRNA species and highly aggregated virus particles (Fig. 67) in the sedimented material. (It is thought that the occasional single particles found in the preparation, shown in Fig. 67, were a result of the procedure for spraying virus on to the electron microscope grids, rather than evidence for a naturally separate species). B. Properties of Pg 1348-2 virus particles a. Particle morphology Examination of the crude virus preparation (that had sedimented through a sucrose density gradient) in the electron microscope showed that the majority of the particles were intact, though in a highly aggregated state. A histogram constructed of numbers of particles versus 197 Figure 67 Electron micrograph of Pg 1348-2 virus ( x 100,000 magnification ) 198 particle diameter (Fig. 68) showed that the virus had a modal diameter of 30 nm. b. RNA components RNA was extracted frcm a preparation of crude virus by the SDS/phenol procedure (G.M.12b), and ethanol precipitation. Nucleic acid samples were resuspended in SIM buffer and were analysed by RNA p.a.g.e. Scans were taken of the stained gels (Fig. 69) and facilitated the calculation of the molecular weight of these species through coelectrophoresis with internal dsRNA markers. The molecular weights were interpolated from plots of log molecular weight versus mobility (cm/h) of the markers (Fig. 70) and were found to be 1.32 x 106, 1.21 x 106 and 1.14 x 106. The fourth dsRNA component appeared to comigrate with the larger of the Penicillium stoloniferum virus S dsRNA species, with molecular weight 1.11 x 106. It proved impossible to determine the polypeptide content, sedimentation coefficient, buoyant density characteristics and ultraviolet spectrum of this virus, because of the difficulties involved in purifying and separating such a highly aggregated species. c. Serology. Samples of crude virus were tested for cross reactivity with antisera to P. sp.(lh)2-2 viruses and mixed viruses, Ggg G1 virus, Pc[ 1348-2 virus, and Ggt viruses (from isolates OgA, 019/6, T1, F3, 3b1a and 38-4). The virus gave only faint reaction to the Pc[ antiserum and also to the P.sp. (lh) (mixed virus) antiserum. It is not known whether the latter reaction was to a crude sample contaminant or to a virus in this preparation, since the virus from Pcj_ could not be purified by the conventional methods used in this study. (The Pc[ virus reacted only with a 'fourth bleed' sample of antiserum from the rabbit in which the P.sp(lh) 199 60 50 40 w 0 H O •H 30 ft H o 20 10 28 29 30 31 32 33 34 35 Particle diameter (nm) Figure 68 Histogram plot of particle diameters measured randomly from an electron micrograph of Pg 1348-2 virus 200 2.0 2.5 Distance in gel (cm) Figure 69 Polyacrylamide gel profile of RNA released from Pg 1348-2 virus 201 1.6 - 1.5 • 1.4 1.3 - 1.2 1.1 - 1.0 0.9 0.2 0.4 0.6 0.8 Mobility (cm/h) # RNA standards in order of increasing mobility O Pg virus RNA components as numbered Relationship between log molecular weigh' electrophoretic mobility of Pg virus RNA 202 antiserum was raised. This was also the only bleed to which the P.sp. (lh) 2-2 viruses reacted). It is possible that the faint reaction of the Pg^ virus to its homologous antiserum and lack of reaction to heterologous antisera, is due to its highly aggregated state and hence difficulty in entering the gel matrix. 203 DISCUSSION Purification of P£ virus involved the problems of virus solubility as shown for the Ggt 45/10 viruses. Comparative gel analysis of dsRNA extracted directly from nYcelium showed that a considerably higher yield (20 ug/g wet weight mycelium) had been obtained than with any of the buffers used in the virus isolation procedure (4 ug/g in PK and veronal buffers), indicating that approximately 75% of the virus had been lost in the first clarifying spin given to the mycelial homogenates. However, serological analysis and the highly regular array of the 30 nm virus particles suggest that only one virus is present, although this hypothesis remains to be proven. Further experimentation with this virus, using buffers with a range of pH and ionic strength, is required to extract and solubilise the virus particles, enabling their characteristics to be determined. Despite the difficulties in making correlations between this virus and those of Ggt, the observation that Pg_ virus (and P.sp(lh)2-2 viruses A and B) reacted with an antiserum to a mixed group of viruses from Phialophora species (with lobed hyphopodia) suggests that biotype variation exists between the viruses in the two groups. The molecular weight f» values of the Pcj_ virus dsKNA components (1.32 x 10 fi , 1.21 x 10 , 6 6 1.14 x 10 and 1.11 x 10 ) show marked similarity to those of group I Ggt viruses (Buck et al., 1981) although the particle diameter is smaller than any of those yet noted. Nonetheless, the possession of four dsRNA components, which are most likely to be each encapsidated in a single virus, suggests that this virus may contain satellite or defective dsRNAs. Because of this, the number and molecular weight of the dsRNA components should be regarded as doubtful properties for the purposes of virus grouping. With the lack of information on capsid polypeptides and polymerase molecules, the possession of satellite dsRNAs remains only a possibility, awaiting RNA fingerprint analysis and cell-free protein synthesis to be resolved. 204 SECTION 6 Interactions between isolates RESULTS Ggt isolates OgAf 019/6, 45/10, F3, T1, 38-4 and 3b1a (G.M. 1., Table 1) were examined for their interaction with P.sp. (lh) isolate 2-2, Ggg isolate G1 and Pc[ isolate 1348-2 on solid media buffered to pH 3.0, 3.5, 4.0, 4.5, 5.0, 6.0, 7.0 and 8.0 with CP buffer. The three media in which tests were made were a. PDA, b. glucose/asparagine medium (LB medium; Lilly and Barnett, 1951) and c. glucose/urea medium (H medium; Haskins, 1950). Plugs (6mm discs from stock PDA cultures) of different isolates were placed 2 to 2.5 an apart in the centre of Petri dishes containing buffered medium. For each isolate, two discs of the same isolate were paired to serve as controls. Colony diameter was measured at intervals of 3 to 5 days to determine rates of linear growth at the different pH values and on the different media. A. Growth characteristics of isolates in control experiments. Rates of growth of colonies on buffered media are shown graphically for isolates grown at pH 3.0 and 6.0 on PDA medium, at pH 4.0 and 7.0 on LB medium and at pH 5.0 on Haskins medium. Isolates grown in control experiments showed no sign of inhibition, each colony growing symmetrically away from the disc inocula and meeting with free intermingling of hyphae. There was no sign of heterokaryon (post fusion) death (Esser and Blaich, 1973) in these paired controls. a. P.sp.(lh)2-2 Rates of growth are given in Fig. 71a, b and c. The morphological appearance of the fungal isolates on the three media is outlined in Table 15. 205 Figure 71 Rate of growth of P.sp.(lh) isolate 2-2 on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0 206 Field pH of medium Medium isolate 3>Q 4>Q 5>Q 6 white-pale grey, lanuginose colonies with LB II entire margins Ggg G1 pale grey, f:.occos e colonies PDA with with entire margins irregular margins white, lanuginose colonies with irregular LB margins white, lanuginose colonies with entire H margins Pg 1348-2 orange-brown pale grey,lanuginose PDA floccose colonies with entire colonies margins with entire margins pale yellow pale grey,lanuginose LB H lanuginose colonies with entire colonies margins with entire margins Ggt OgA pale grey-brown,lanuginose colonies with PDA entire margins white-pale cream,lanuginose colonies with LB entire margins pale grey- white-pale cream, lanuginose H brown colonies with entire margins lanuginose colonies with entire margins Ggt 019/6,F3, white-pale grey, lanuginose colonies with PDA LB H 3bla,38-4, entire margins 45/10 Ggt T1 white-pale cream, lanuginose colonies with PDA LB entire margins white-pale cream white, lanuginose H lanuginose colonies colonies with entire with irregular margins margins Morphological appearance o: and Ggt on different media 207 b. Ggg G1 Rates of growth are shown in Fig. 72a, b and c. c. Pg 1348-2 Rates of growth are shown in Fig. 73a, b and c. Unlike the isolates P.sp (lh) 2-2 and Ggg G1, the isolate of Pc[ did not form growing colonies at pH 3.0 on PDA medium and grew very slowly at pH 6.0. At pH 3.0 it could be seen that the discs of mycelium were alive, as short aerial hyphae were seen emerging from the plugs when observed under a binocular microscope (x10 magnification). In contrast, the isolate grew appreciably on both LB and H media at the low pH values. The appearance of the colonies on the LB and H media were very similar above pH 4.0 (Table 15). d. Ggt OgA In common with Pg^ isolate 1348-2 and the other six Ggt isolates, no growth was observed on PDA medium at 3.0, although under the binocular microscope aerial hyphae were seen emerging frcm the agar discs. This isolate, unlike any of the other Ggt isolates appears able to overcome, possibly by alteration of the pH of the medium, a slow initial rate of growth on LB medium, pH 4.0 (Fig. 74), resulting in paired colonies becoming confluent after 15 days (P.sp.(lh)2-2 colonies became confluent after 12 to 13 days). e. Ggt 019/6 Growth rates are shown in Figs. 75a, b and c. The isolate was slow growing on LB medium pH 4.0 and H medium pH 5.0 with a growth rate of approximately 0.02 cnv/day, but grew more rapidly, as expected, on LB medium buffered to pH 7.0 and PDA to pH 6.0 (0.3 and 0.4 cn\/day respectively). 208 5 10 15 Days (b) * 7.0 /A 4.0 &o rH 8 _ ... .i i i 5 10 15 Days Figure 72 Rate of growth of Ggg isolate Gl- on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0 209 5 10 15 Days &o +-a>) Days •H Q Figure 73 Rate of growth of Pg isolate 1348-2 on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0 210 5 10 15 Days Figure 74 Rate of growth of Ggt isolate OgA on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0 211 (b) 47.0 >> es > o . i t —i i •H Q 5 10 15 Days Figure 75 Rate of growth of Ggt isolate 019/6 on (a) PDA medium buffered to pH 3.0 and 6.0; (b) LB medium buffered to pH 4.0 and 7.0; (c) H medium buffered to pH 5.0 212 f. Ggt F3, 45/10, 38-4, 3b1a, and T1 The seven Ggt isolates showed an increase in growth rate with pH (Fig. 76). This was also demonstrated for Pg 1348-2. The species P.sp.(lh) and Ggg (isolates 2-2 and G1 respectively) showed a pH optimum for growth rate of pH 6.0, as shown for example, by the diameter of colonies grown on LB medium, buffered over the pH range 3.0 to 7.0 after 5 days at 24°C (Fig. 77). All the Ggt isolates grew very slowly at low pH (pH 3.0 to 5.0). The P.sp.(lh) and Ggg isolates were invariably the fastest growing isolates at any pH, e.g. with a growth rate of 0.1 cn\/day or more at 3.0. B. Growth characteristics of isolates showing inhibitor activity on solid media. a. PDA medium Growth rates of paired isolates of the two groups (i.e. the isolates avirulent to cereals, P.sp. (lh)2-2, Ggg G1 and Pg 1348-2, with those pathogenic to cereals, Ggt OgA, 019/6, 45/10, F3, T1, 38-4 and 3b1a) were measured in the same way as controls. Growth rates of the individual colonies did not appear to differ significantly from control colonies. P.sp.(lh) and Ggg isolates showed the same dependence on pH, with optimal growth at pH 6.0, while Pcj and Ggt isolates increased in growth rate from pH 3.5 to 8.0. There was no evidence of inhibition at pH 3.0, when isolates P.sp. (lh)2-2 and Ggg G1 were paired in turn against the seven Ggt isolates (Fig. 78). P.sp. (lh), Ggg or Pc[ inocula are on the left and Ggt inocula on the right in all the figures. Neither Pg_ nor the Ggt isolates produced growing colonies of appreciable size (hyphae were just visible under 213 Figure 76 Colony diameters of isolates Ggt OgA (4-4), 019/6 (W), 45/10 (A-A),T1 (QQ,F3 (»-#),38-4 (Q-Q)and 3bla (H) after 5 days growth on LB medium buffered from pH 3.0 to 7.0 214 !0H •0P 3 >> O rH Figure 77 Colony diameters of isolates P.sp.(lh) 2-2 (4-+"), Ggg G1 (•-•) and Pg 1348-2 (T^) after 5 days growth on LB medium buffered from pH 3.0 to 7.0 > 215 the binocular microscope at pH 3.0 (Fig. 78)). This was the same when paired against 2-2 or G1. PDA medium buffered to higher pH values allowed the growth of Pg and Ggt colonies, but few signs of inhibition were noted at any values, except when Pc[ was paired with the Ggt isolates at pH 6.0 and 7.0. The faster growing isolates Ggt 019/6, F3 and 45/10 grew over the Pg colonies before the latter had reached a diameter of 0.7 cm. However colonies paired with Ggt isolates T1 (Fig. 79) 38-4 and 3b1a grew to a diameter of approximately 1.8 cm, when it was noted that colonies of Pg_ appeared to be curving away from the Ggt colonies (after 16 days incubation at 25°C). This was the first sign of inhibition of Pg_ noted up to these pH values. The inhibitory effect of Ggt on the Pc[ colony was first noted with isolates 019/6, OgA, F3 and 45/10 at pH 7.0, possibly because at this pH the Pc[ isolate was growing at a faster rate and the effects on the curvature of the colony were more noticeable. The inhibitory effects noted in tests of P.sp. (lh)2-2 and Ggg G1 with Ggt isolates were less marked. In the majority of paired inocula, at every pH value, no effect on colony symmetry was noted, and in cases where a possible effect was found i.e. 2-2 versus Ggt 38-4 at pH 4.0 and 5.0, and G1 versus 38-4, T1 and 019/6 at pH 4.0 and 5.0 it was noted only in one of a duplicated pair of tests. However when colonies became confluent, hyphae appeared to inter- mingle freely, but following storage at 4°C for 4 to 5 days, a darkened area developed between the colonies, similar in appearance to the 'barrage1 reaction, noted by Esser and Blaich (1973) with other fungi, caused by heteroincompatibility or post fusion death (Fig. 80). b. Glucose/asparagine (LB) medium Inocula of isolates of P.sp. (lh), Ggg and P£ were paired with those of the seven Ggt isolates on LB medium, buffered to pH values 3.0, 3.5, 216 Figure 78 Interaction between P.sv.(lh) isolate 2-2 and Ggt isolate 38-4 on PDA medium buffered to pH 3.0,at 24 0 C Figure 79 Interaction between Pg isolate 1348-2 and Ggt isolate T1 on PDA medium buffered to pH 6.0,at 24°C Figure 80 Interaction between P.sp.(lh) isolate 2-2 and Ggt isolate T1 on PDA medium buffered to pH 6.0,at 24°C 217 4.0, 4.5, 5.0, 6.0, 7.0 and 8.0. The combination of the non-pathogenic isolates with Ggt OgA, is illustrated (Figs. 81 to 86) as representative of the seven Ggt isolates. Photographs were taken after 8 to 20 days at 25°C, the time selected being dependent on the requirement to record either growth stage or the optimum demonstration of presence/ absence of inhibition. Results demonstrated, as did control experiments, that Pg isolate 1348-2, though failing to produce measurable colonies on PDA at pH 3.0, grew well on low pH IB medium. Ggt isolates corresponded well to control growth rates, with slow but measurable growth rates on media buffered from pH 3.5 to 8.0. There appeared to be no growth at pH 3.0, but these isolates grew at increasing rate with values above pH 3.5. The Ggt isolate OgA showed the highest growth rate, as in control experiments. However Ggt 38-4 grew more rapidly when paired with Pc[ 1348-2, than expected from control results. Since it was unlikely that the plate had become contaminated (no sign of contamination was noted in the stock culture) and since inocula were all taken frcm approximately 0.5 cm behind the margin of a growing stock isolate colony (and therefore likely to be growing at a uniform rate) this result was regarded as fortuitous. No sign of inhibition was noted at pH 3.0 on LB medium, for any combination of P.sp. (lh)2-2, Ggg G1 or Pg 1348-2 with any one of the seven Ggt isolates. There was scxne sign of inhibition at pH 3.5 in some of the paired combinations i.e. Ggg G1 with Ggt 38-4, 3b1a and OgA (Fig. 87), P.sp. (lh)2-2 with Ggt 38-4 and 019/6 and Pg^ 1348-2 with Ggt 019-6. The difficulty with assessing inhibition at low pH was that colonies did not grow perfectly symmetrically at these low 01 levels. Inhibition was clearly evident at pH 4.0 in all paired combinations, particularly in the case of Ggg G1 which, as shown in control experiments, had colonies with very irregular margins. This irregularity was completely absent at the edge of the colony nearest to the Ggt colony 2 Fi~re 81 Interaction between isolates P.sp.(lh) 2-2 and Ggg G1 and Ggt isolate OgA on LB medium buffered to pH 3.0, at 24°C Figure 82 Interaction between isolates P.sp.Clh) 2-2, Ggg G1 and Pg 1348-2 and Qgt isolate QgA on LB medium buffered to EH 4.0, at 24 0 C Figure 83 Interaction between isolates P.sp.(lh) 2-2, Ggg Gl and Pg 1348-2 and Ggt isolate OgA on LB medium buffered to pH 5.0, at 24°C 219 Figure 84 Interaction between isolates P.sp.(lh) 2-2 and Qgg G1 and Ggt isolate OgA on LB medium buffered to EH 6.0, at 24 0 C Figure 85 Interaction between isolates P.sp.(lh) 2-2, Ggg G1 and Pg 1348-2 and Ggt isolate OgA on LB medium buffered to pH 7.0, at 24 0 C Figure 86 Interaction between isolates P.sp.(lh) 2-2, Qgg G1 and Pg 1348-2 and 9g! isolate 0~ on LB medium buffered to £H 8.0, at 24 0 C 220 Figure 87 Interaction bet\veen isolates Ggg Gl and Ggt 0g on LB medium buffered to pH 3.5, at 24°C Figure 88 Interaction between isolates P.sp.(lh) 2-2 and Ggt 38-4 on LB medium buffered to 8.0, at 24°C Figure 89 Interaction between isolates P.sp.(lh) 2-2 and Ggt OgA on H medium buffered to pH 3.0, at 24°C 221 producing the inhibition. The evident inhibition was greatest from pH 4.0 to 5.0 where growth rate of the Ggt isolates remained limited relative to that of the non-pathogenic isolates (Figs. 82, 83). Inhibition was nonetheless present at pH 6.0 and above on the LB medium. P.sp. (lh)2-2, for example, could be seen clearly to curve around all the colonies of Ggt at pH 6.0 and 7.0, particularly when paired with Ggt isolates 38-4, 3b1a and 45/10. At pH 8.0 inhibition was less clear, but was evident from the fine hyphae growing into the agar on all sides of the colonies of P.sp. (lh)2-2, except on the side adjacent to the Ggt colonies (particularly noticeable when paired with Ggt isolate 38-4, Fig. 88). Ggg isolate G1 paired with Ggt isolates showed essentially the same results (Figs. 83 to 85) at pH 6.0 to 8.0, with inhibition less evident at pH 8.0, when Ggt colonies had grown to a greater extent relative to the Ggg G1 isolates. Nonetheless some evidence of inhibition was present at the junctions of the Ggg G1 colony margin with the Ggt colony (Fig. 85). Little evidence of inhibition is found between pH 6.0 and 8.0 in combinations of Pc[ 1348-2 with Ggt isolates on LB medium. (Figs. 85 and 86). c. Glucose/urea (H) medium Isolates grown on H medium, like those on LB medium, showed good correlation between general growth characteristics and in the dependence of inhibition on pH. (Figs. 89 to 93). There was no sign of inhibition at 3.0 between isolates paired on this medium (Fig. 89) but at pH 4.0 there was evidence of inhibition in all paired combinations (Fig. 90). Inhibition was also noted between P.sp. (lh), Ggg and Pg isolates paired with Ggt 019/6, 38-4, 3b1a, OgA, F3 and 45/10 at 5.0. Above pH 5.0 inhibition was evident in combinations of P.sp.(lh)2-2 with Ggt isolates at pH 6.0, but not at pH 7.0 or 8.0 where colony growth rate became erratic (Fig. 93). Signs of inhibition were evident at all pH 222 Figure 90 Interaction between isolates P.sp.(lh) 2-2 and Ggt 0gA on H medium buffered to pH 4.0, at 24°C Figure 91 Interaction between isolates P.sp.(lh) 2-2 and Ggt OgA on H medium buffered to pH 5.0, at 24°C 224 values on H medium for combinations of Ggg G1 with the seven Ggt isolates, through the appearance of a sharply delineated margin, immediately adjacent to the growing Ggt colony, with curling back of the Ggg hyphae from the Ggt colony margins. There was no sign of asymmetric growth of colonies of Pc[ isolate 1348-2 in any paired combination with Ggt isolates at pH 6.0 or above (except with Ggt 38-4 at pH 6.0). Thus the assessment of inhibition of colony growth by culture in paired combinations on buffered media, demonstrates a pH dependence for the effect. The range of required for each combination is given in Table 16 for LB and H media. The pH optimum for the inhibition effect by isolates of Ggt fell between pH 4 and 5 for all combinations with the non-pathogenic isolates on LB and H media. It was noted that at all pH levels where greatest inhibition of the growth of P.sp. (lh)2-2, Ggg G1 and Pg_ 1348-2 occured, there was minimal growth of Ggt isolates. At higher pH values where less inhibition was observed, Ggt colony growth was increased. C. Nature of the inhibition reaction In order to determine whether the inhibitor caused cell death inocula of the non-pathogenic isolates were cultured with those of the Ggt isolates on LB medium, buffered to pH 4.0 and incorporating methylene blue (G.M. 23). Cell death was determined by blue staining of the hyphae (living cells can take up methylene blue, but decolourise it). Tests made on this medium showed that in the first 4 to 5 days growth at 25°C no staining of hyphae occurred, nor was there any sign of inhibition up to this point. After 7 to 15 days growth, all combinations showed inhibition reactions but only occasional colonies developed faint patches of stain around the margin as in Figs. 94(a) and b. These seemed to occur randomly along the margin of the colonies, not at sites in closest apposition to the Ggt colony, nor did hyphae in the Ggt colonies Table 16 pH range dependence of inhibition between isolates cultured on LB and H media Non pathogenic Medium Ggt isolate isolate 019/6 T1 38-4 3b1a Oga F3 45/10 I 1 P.sp.(lh)2-2 I LB | 4-7 4-7 3.5-8 4-7 4-7 4-7 4-7 I 1 I H I 3.5-6 4-7 4-6 3.5-6 4-6 4-6 4-6 I 1 Ggg G1 I LB I 3.5-7 4-8 3.5-8 3.5-8 4-7 4-8 4-7 1i 1I 1 H | 4-6 4-7 4-7 4-7 4-6 4-8 4-8 I 1i 1 Pg 1348-2 | LB | 3.5-5 3.5-5 4-6 3.5-5 4-5 3.5-5 3.5-5 1i 1I 1 H I 4-5 4-5 4-5 4-5 4-5 4-5 4-5 1 1 1 1 CO UCOl 226 (a) (b) Figure 94 Interaction between Ggg isolate G1 and (a) Ggt 38-4 and (b) Ggt 019/6 on LB medium buffered to pH 4.0 and incorr9rating methylene blue Figure 95 Interaction between Pg isolate 1348-2 and Gg! isolate OgA on LB medium buffered to pH 4.0 and incorporating methylene blue 227 themselves appear to stain with the methylene blue. It is likely that this generalised staining is due to the normal ageing of the colony and localised exhaustion of nutrients causing cell death. When paired on methylene blue-containing LB medium, the combination of Pc[ 1348-2 with Ggt OgA was the only one to demonstrate confluence between two growing colonies leading to staining between the colonies, possibly as a result of post fusion death (Esser and Blaich, 1973? Fig. 95). It appears therefore, that the inhibition reaction occurs at a distance and does not involve cell death, but that this may develop into a heterokaryon incompatibility reaction (post fusion death) when the initial inhibition is overcome and colonies meet. D. Diffusion of inhibitor in liquid culture Culture filtrate was prepared by filtration of mycelium and sterilisation (G.M.24). Culture filtrates were tested for inhibitory actvity by inoculation into wells cut approximately 0.6 cm from growing colonies of the non-pathogenic isolates in LB medium buffered to pH 4.0 (G.M.25). Inhibition, identical to that found for plug inocula on LB medium at pH 4.0, was found after 3 to 4 days incubation at 25°C (Fig. 96). Control wells were either left empty or filled with LB liquid medium, buffered to pH 4.0. (All inocula tested gave the same results. The figure shown of Gc[t 45/10 culture filtrate with Pej 1348-2 and Ggg G1 colonies are representative of the effects of the other Ggt liquid inocula on the three non-pathogenic isolates). E. Effect of proteases on the inhibition reaction Solid LB medium, buffered to pH 4.0 or pH 7.0, incorporating the proteases pepsin (pH 4.0 for optimum activity), papain and proteinase K (pH 7.0 for optimum activity), was used for culture of isolates as in the preceding section. It was found that, although proteinase K caused a 228 control plus culture filtrate Figure 96 Well tests of isolates Ggg G1 and Pg 1348-2 with Ggt 45/10 culture filtrate 229 small reduction in the growth rate of isolates, before the addition of culture filtrates, in no case was the inhibition effect prevented. In fact, inhibition of growing colonies of 1348-2 ocurred in both tests (with papain and proteinase K) on LB medium buffered to pH 7.0. This was possibly because repeated inoculation of wells with culture filtrate, allowed a greater concentration of inhibitor to be used than was possible with paired colonies on agar. Agar removed from Petri dishes of medium containing the proteases was homogenised, and the homogenate in each case found capable of catalysing the hydrolysis of bovine serum albumin. When the agar was incubated with bovine serum albumin (0.5 ug/ml) at 24°C for 2 to 4 days, no protein band, staining with Coomassie blue, could be detected following electrophoresis in SDS/polyacrylamide gels (G.M. 16d). F. Stability of inhibitor in culture filtrates Stocks of culture filtrate were kept for periods of up to one month at 4°C, following which well tests were performed, with several inoculations of each well as before, over a four day period. Cultures were incubated at 25°C for 7 days (G.M. 25). Despite the length of time in storage and frequent U.V. sterilisation (G.M.24) culture filtrates retained their inhibitory activity. Inhibitory activity in culture filtrates was also stable following incubation at 24°C for 16 h. 230 DISCUSSION Contrary to expectation, none of the isolates of Phialophora graminicola, Phialophora sp.(lh) or Gaeumannomyces graminis var. graminis produced diffusible inhibitors active against the seven Ggt isolates tested. There are two possible reasons for this lack of activity. Firstly, killer toxin may be produced but the seven Ggt tested are resistant to it. This is unlikely, however, since over 50 Ggt isolates have been tested against P.sp.(lh)2-2 and other Phialophora isolates and no inhibitor has been observed to be produced by the Phialophora species (K.W. Buck, personal communication). Moreover, there is no basis for dsRNA-encoded resistance in four of the Ggt isolates tested (Ggt 3b1a, OgA, 019/6 and T1) since they contain only sufficient dsRNA for the structure and replication of the viruses contained within each isolate. Secondly, it is possible that no inhibitor is produced by the Pg_, P.sp(lh) and Ggg isolates, and the additional dsRNA components they have been shown to contain, other than those required to encode virus capsid polypeptides or polymerase molecules, may be defective dsRNAs arising by mutation of genomic dsRNA molecules. However, it may be that the additional molecules are satellite dsRNAs, which could merely encode a polypeptide required for the replication of the satellite dsRNA, or for some other biological function. All of the seven Ggt isolates tested produced inhibitors active against isolates of Pg_, P.sp. (lh) and Ggg. These inhibitors are similar to those produced on PDA by three (out of 20) isolates of Ggt (Ggt isolate 45/8 and ascospore isolates of 3b1a and F911-1) tested by Romanos et al. (1980). Inhibitors from either study had activity associated with low pH and were spontaneously produced in cultures on media buffered to pH 3.5 to 4.5. The production of the inhibitors was associated with slow growth of the Ggt isolates, suggesting that they might be sensitive to 231 their own toxins, and all the inhibitors were resistant to the action of proteases. The only differences found between these factors were that, firstly, all isolates of Ggt tested in this study produce inhibitors, but only three out of the twenty tested by Rcmanos et al., (1980) were found to do so. However this is likely to be an effect of the medium used in the tests, since negligible activity was noted on PDA by the seven Ggt isolates used in this study (the majority of the inhibitory reactions on PDA involved interactions similar to the post fusion death described by Esser and Blaich, 1973) and yet all produced inhibitor on LB and H media. The Ggt 45/8 Q inhibitor of Romanos et al. (1980) was far less stable than the Ggt inhibitors reported here but this might again be caused by the different media used (it is possible that the inhibitors are unstable in malt extract culture medium). The final difference, i.e. in the pH range over which the Ggt isolates are effective in this study (up to pH 7.0) may also be accounted for by media effects since it is likely that the inhibitors may show a greater pH range for activity through being present at a higher concentration. Several differences were apparent between the Ggt inhibitors and the killer toxins of Saccharomyces cerevisiae and Ustilago maydis. These include the resistance of the Ggt inhibitors to the action of proteases and their apparent sensitivity to their own toxins. (Except in the case of rare 'suicide* mutants of yeast killer strains of Saccharomyces cerevisiae and Ustilago maydis are resistant to their own toxin). Ability to produce the Ggt inhibitors was not dependent on the presence of specific viral dsRNA components, since only Ggt isolates 45/10, 38-4 and F3 contained dsRNA in excess of that required for the structure and replication of the viruses contained within them. However they are probably also distinct from the K4 to K10 killers of Young and Yagiu (1978), which do not contain dsRNA, as the latter have been shown to be at least partly proteinaceous through their susceptibility to protease 232 activity, or to temperatures of 35°C or more, or through their irreversible inactivation by pH values greater than pH 5.0. The Ggt inhibitors also differ markedly from the volatile factor described by Sivasithamparam et al. (1975). The inhibitors produced by Ggt isolates in this study, caused direct inhibition of the growth of the sensitive fungi, whereas the volatile factor caused stimulation of the growth of seme Ggt isolates, followed by the formation of lytic plaques. The pH-dependent growth characteristics of the Ggt isolates and isolates of P.sp. (lh) and Ggg are likely to be dependent, at least in part, on the production of inhibitor by Ggt colonies, resulting in the apparent slow growth of the 'self-inhibitory' isolates, relative to the more vigorous non-pathogenic isolates. The Pc[ isolate 1348-2 however is likely to be naturally slow-growing and not dependent on the production of inhibitor, causing self-inhibition of the growth of the isolate. Differences in growth rates of Ggt isolates on different media, reported in the literature, are likely to be a result of inhibitor production in some cases. It is, for example, notable that the isolate Ggt 45/10, an inhibitor producer in this study, was shown to be sensitive to a diffusible inhibitor when in combination with Ggt isolate F911-1 on PDA plates. (Romanos et al., 1980). However the choice of medium may be important in this case, since the self-inhibitory nature of these factors may affect the growth rates of the isolates on one medium but not on another. Alternatively there may be a range of different inhibitor molecules produced by these Ggt isolates, which are sensitive to the action of inhibitors from other isolates, as well as to their own factor. The observation that Phialophora isolates and Gaeumannomyces graminis var. graminis do not produce inhibitor (confirmed now for a range of Phialophora sp.(lh) isolates; K.W. Buck, personal communication), whereas 50 Ggt isolates have been shown to produce inhibitor, suggests that inhibitor production may be a useful feature for distinguishing Ggt isolates from those of p.sp.(lh) and Ggg. 233 CONCLUSIONS Characterisation of the viruses extracted from fungal isolates P.sp. (lh)2-2, Ggg G1 and Pc[ 1348-2 has revealed the presence of type 2 isometric mycoviruses, i.e. with segmented dsRNA genomes, in these species. A possible type 1 virus, containing a single polycistronic dsRNA molecule, is present in P.sp. (lh)2-2, and, by consideration of the size of the dsRNA and virus particle, in P.sp. (lh) 12-2 also. It is noteworthy that a 40 nm virus containing a single polycistronic dsRNA molecule has been found in Ggt isolate F10 (Almond, 1979). The serological relationships demonstrated between P.sp(lh)2-2 viruses A and B and Ggt 45/10 viruses, and between P.sp. (lh) viruses and Pg 1348-2 virus, in addition to the marked similarities in physico- chemical characteristics of the Ggg, Pg, P.sp. (lh)2-2 A and B viruses and Ggt viruses studied by Buck et al. (1981), suggest that these viruses may be variants of viruses infecting both Gaeumannomyces and Phialophora species. Since the mycoviruses can persist indefinitely in their hosts, the opportunity exists for the survival and selection of virus mutants. These results imply transmission of viruses between isolates of the Phialophora and Gaeumannomyces species. This is possible between incom- patible isolates, as well as between compatible fungi, since virus exchange may take place by hyphal anastomosis in the isolates, before post-fusion death occurs in the region of cell fusion, e.g. as found in Endothia parasitica (Anagnostakis and Day, 1979). The transmission of virus between Phialophora and Gaeumannomyces species also provides an alternative explanation to provirus for the introduction of virus into Ggt, in fields previously shown to be free of virus-containing isolates (Rawlinson and Muthyalu, 1976). The serological relationships demonstrated between viruses of the non-pathogenic fungal isolates and with those of Ggt, occur in those specimens isolated from fields of barley and wheat at Rothamsted 234 Experimental Station. However the finding of serologically related viruses in isolates of Ggt widely separated geographically (Buck et al., 1981), e.g. in isolates from Japan (Ggt 01-1-4) and England (Ggt 019/6) and from Australia (Ggt T1), England (Ggt OgA) and France (Ggt F6) implies that transmission of infected propagules of these fungi may occur between continents, or that virus infection may have arisen early in one of the ancestors of these species. It is likely, therefore, that virus variants related to Pg, P.sp.(lh) or Ggg viruses will be discovered in isolates geographically widespread from each other. Study of the viruses infecting the isolates of P.sp.(lh)2-2, Ggg G1 and Pc[ 1348-2 revealed possible satellite viral dsRNA molecules in all these species. However no killer, or other inhibitor, analogous to those of Saccharomyces cerevisiae or Ustilago maydis was found in these organisms. It was likely, therefore, that the ability of these species to cross protect cereal plants from the effects of virulent Ggt, was not due to the production of diffusible inhibitors by these isolates. Further, there is no evidence in this study, that viral dsRNA directly affects the hypovirulence of Phialophora or Gaeumannomyces species. Since these dsRNA mycoviruses in nonpathogenic species are shown to contain dsRNA components, other than those required for virus structure or replication, but apparently do not code for inhibitors or killer toxins, explanations for the presence of this "redundant" dsRNA must be found. There are a number of possibilities. a. Some of these RNA components may be deletion mutants, arising by defective replication of existing genomic constituents. Defective particles containing pieces of virus genome are common in animal viruses (Fenner, 1974). However, experiments by Romanos (1981) have shown (by RNA fingerprinting and in vitro translation) that a satellite dsRNA present in a virus infecting Ggt isolate 38-4 and absent in the identical virus infecting Ggt 019/6, was distinct from other dsRNA components present in the virus. 235 b. Some RNA components may code for non-structural polypeptides. Additional peptides may be required as part of a host polymerase holoenzyme in viruses unable to replicate autonomously, or for other functions in virus RNA replication as in reovirus (Zweerink et al., 1971). c. Some dsRNA components may be directly involved in the control of virus replication. The smallest RNA component of Aspergillus foetidus virus S may possibly be involved in the control of the transcription of this virus (Buck, 1977). d. Some of the multiple dsRNA components may be conformational isomers of the same molecule and may not actually differ in molecular weight. However this is extremely unlikely in view of the requirement for at least two of the components for capsid protein and polymerase enzyme information. The effect of glyoxal and the use of urea/SDS polyacrylamide gels on P.sp. (lh)2—2 virus A dsRNA, precludes such a relationship in this species. However the available evidence does not allow any of these possibilities to be completely disregarded. Despite the lack of a direct correlation between the presence of viral RNA species and the production of inhibitor in the nonpathogenic isolates, the wide variation in the dsRNA pattern of mobility obtained from isolates of a single species, P.sp.(lh)2-2 (isolated from a single field at Rothamsted), provides an explanation for the apparent variation in the degree of cross protection from Ggt afforded cereal plants by these nonpathogenic species. Such a virus mediated effect might be expected to be reflected in variation of other host properties. Speakman and Lewis (1980), for example, have shown that different isolates of Phialophora graminicola have differing requirements for the vitamins biotin and thiamine for growth. Further, although levels of take-all in cereals are generally lower after grass than after non-graminaceous brealk crops, which could be ascribed to carry over and cross protection by 236 P.graminicola (Deacon, 1973; Slope et al., 1978), a high incidence of this fungus has also been found in soils coincident with minor attacks of take-all in a wheat crop (Slope et al., 1979). This suggests that populations of P.graminicola may differ in their ability to delay take-all attacks and that this variation may be correlated with variations in virus dsRNA components. It is evident from the results obtained that a system of vegetative incompatibility exists between isolates of Ggt and the non-pathogenic isolates tested, involving the production of a diffusible inhibitor by Ggt, preventing hyphae from the paired isolates from meeting and/or causing necrosis of the hyphal tips (Caten and Jinks, 1966). These vegetative incompatibility interactions have also been demonstrated between isolates of Ggt when hypovirulent strains were shown to be inhibited (Rcmanos et al., 1980). This implies that cases of take-all decline, or the reduction in take-all experienced on treatment of soil with fertilisers designed to lower soil pH, may be explained by an increase in the inhibitory effect of certain Ggt isolates, since the inhibitory effect in all Ggt isolates tested was maximal between pH 4 and 5. It is possible, for example, that, following a rapid build up of Ggt through some years of severe take-all attack, local exhaustion of various nutrients may occur and the pH of the rhizosphere alter, causing a decrease in disease levels through an increase in the inhibitory effective- ness of certain Ggt isolates. Although the pathogenic Ggt isolates in this study were, without exception, shown to produce an inhibitor limiting the growth of sensitive, non-pathogenic isolates, the inhibitors (like those of Romanos et al., 1980) appeared to be independent of the presence of viral dsRNA e.g. Ggt F911-1 was an ascospore isolate and therefore free of virus and since four out of seven of the Ggt isolates tested in this study contained only sufficient dsRNA to encode capsid polypeptide and polymerase molecules (Almond, 1979; Almond, Buck and Rawlinson, 1977) the dsRNA molecules were unable to encode inhibitory molecules. 237 REFERENCES Almond, M.R. (1979). Biochemical characterisation of viruses isolated from the phytopathogenic fungus Gaeumannomyces graminis var. tritici. Ph.D. Thesis. University of London. Almond, M.R., Buck, K.W. and Rawlinson, C.J. (1977). Viruses of Gaeumannomyces graminis var. tritici. Second International Mycological Congress, Tampa, Florida, 1977, p.13, Abstract. Almond, M.R., Buck, K.W. and Rawlinson, C.J. (1978). The virus complex of Gaeumannomyces graminis var. tritici. Bulletin of the British Mycological Society, V2, 115. Anagnostakis, S.L. (1977). Vegetative incompatibility in Endothia parasitica. Experimental Mycology, 306-316. Anagnostakis, S.L. and Day, P.R. (1979). Hypovirulence conversion in Endothia parasitica. Phytopathology, 69, 1226-1229. Anderson, N.G., Harris, W.W., Barber, A.A., Rankin, C.T. and Candler, E.L. (1966). Separation of subcellular components and viruses by combined rate and isopycnic zonal centrifugation. In National Cancer Institute Monograph, 21_, 253-269. Balis. C. (1970). A comparative study of Phialophora radicicola, an avirulent fungal root parasite of grasses and cereals. Annals of Applied Biology, 66, 59-73. Baltimore, D. (1966). Purification and properties of poliovirus double- stranded ribonucleic acid. Journal of Molecular Biology, J8, 421-428. Baltimore, D. (1969). Replication of picornaviruses. In 'The Biochemistry of Viruses', Ed. H.B. Levy, pp. 101-176. New York: Marcel Dekker. Banks. G.T., Buck, K.W., Chain, E.B., Himmelweit, F., Marks, J.E., Tyler, J.M., Hollings, M., Last, F.T. and Stone, O.M. (1968). Viruses in fungi and interferon stimulation. Nature, London, 218, 542-545. 238 Banks, G.T., Buck, K.W. and Fleming, A. (1971). The isolation of viruses and viral ribonucleic acid from filamentous fungi on a pilot plant scale. Chemical Engineer, 251, 259-261. Barton, R.J. (1978). Mycogone perniciosa virus. Report of the Glasshouse Crops Research Institute, 1977, 133. Barton R.J. and Hollings, M. (1979). Purification and some properties of two viruses infecting the cultivated mushroom Agaricus bisporus. Journal of General Virology, 42, 231-240. Berry, E.A. and Bevan, E.A. (1972). A new species of double-stranded RNA in yeast. Nature, London, 239, 279-280. Bevan, E.A. and Herring. A.J. (1976). The killer character in yeast: preliminary studies of virus-like particle replication. In 'Genetics, Biogenetics and Bioenergetics of Mitochondria', Eds. W. Bandelow, R.J. Schweyen, D.Y. Thomas, K. Wolf and F. Kaudewitz, pp. 153-162. Berlin: Walter de Gruyter. Bevan, E.A., Herring, A.J. and Mitchell, D.J. (1973). Preliminary characterisation of two species of dsRNA in yeast and their relationship to the 'killer' character. Nature, London, 245, 81-86. Blanch, P.A. (1977). Pathogenic variation in Gaeumannomyces graminis. Ph.D. Thesis, University of Oxford. Border, D.J., Buck, K.W., Chain, E.B., Kempson-Jones G.F., Lhoas, P. and Ratti, G. (1972). Viruses of Penicillium and Aspergillus species. The Biochemical Journal, 127, 4P-6P (Abstract). Bostian, K.A., Hopper, J.E., Rogers, D.T. and Tipper, D.J. (1980). Translational analysis of the killer-associated virus-like particle dsRNA genome of Saccharomyces cerevisiae: M dsRNA encodes toxin. Cell, J2, 403-414. Bozarth, R.F. (1972). Mycoviruses: a new dimension in microbiology. Environmental Health Perspectives, 2, 23-29. 239 Bozarth, R.F. (1975). The problem and inportance of transmission of mycoviruses using cell free extracts. Third International Congress for Virology, Madrid, 1975, p. 148, Abstract. Bozarth, R.F. (1977). Biophysical and biochemical characterisation of virus particles containing a high molecular weight dsRNA from Helminthosporium maydis. Virology, 80, 149-157. Bozarth, R.F. (1979). The physico-chemical properties of mycoviruses. In 'Viruses and Plasmids in Fungi1, Ed. P.A. Lemke, pp. 43-91, New York and Basel: Marcel Dekker. Bozarth, R.F. and Goenaga, A. (1977). Complex of virus-like particles containing double-stranded RNA from Thielaviopsis basicola. Journal of Virology, 24, 849-846. Bozarth, R.F. and Harley, E.H. (1976). The electrophoretic mobility of double stranded RNA in polyacrylamide gels as a function of molecular weight. Biochimica et Biqphysica Acta, 432, 329-335. Bozarth, R.F., Wood, H.A. and Goenaga, A. (1972). Virus-like particles from a culture of Diplocarpon rosae. Phytopathology, 62, 493. Bozarth, R.F., Wood, H.A. and Mandelbrot, A. (1971). The Penicillium stoloniferum virus complex: two similar double-stranded RNA virus-like particles in a single cell. Virology, 45, 516-523. Brakke, M. (1967). Density gradient centrifugation. In 'Methods in Virology', Eds. Maramarosch, K. and Koprowski, H., vol. 2, pp. 93-97, London and New York: Academic Press. Brenner, S. and Home, R.W. (1959). A negative staining method for high resolution electron microscopy of viruses. Biochimica et Biophysica Acta, 34, 103-110. Brooks, D.H. (1965). Root infection by ascospores of Qphiobolus graminis as a factor in epidemiology of the take-all disease. Transactions of the British Mycological Society, 48, 237-248. 240 Brown, M.E., Hornby, D. and Pearson, V. (1973). Microbial populations and nitrogen in soil growing consecutive cereal crops infected with take-all. Journal of Soil Science, 24, 296-310. Bruenn, J.A. (1980). Virus-like particles of yeast. Annual Review of Microbiology, 34/ 49-68. Bruenn, J.A., Bobek, L., Brennan, V. and Held, W. (1980). Yeast viral RNA polymerase is a transcriptase. Nucleic Acids Research, 8i, 2985-2997. Bruenn, J.A., and Brennan, V. (1980). Yeast viral double-stranded RNAs have heterogeneous 3' termini. Cell, 923-933. Bruenn, J,A., and Kane, W. (1978). Relatedness of the double-stranded RNAs present in yeast virus-like particles. Journal of Virology, 26, 762-772. Buck, K.W. (1975). Replication of double-stranded RNA in particles of Penicillium stoloniferum virus S. Nucleic Acids Research, _2, 1889-1902. Buck, K.W. (1977). Biochemical and biological implications of double- stranded RNA mycoviruses. In 'Biologically Active Substances: Exploration and Exploitation', Ed. D.A. Hems, pp. 121-148. Chichester: John Wiley and Sons. Buck. K.W. (1978). Semi-conservative replication of double-stranded RNA by a vir ion-associated RNA polymerase. Biochemical and Biophysical Research Communications, 84, 639-645. Buck, K.W. (1979). Replication of double-stranded RNA inycoviruses. In 'Viruses and Plasmids of Fungi', Ed. P.A. Lemke, pp. 93-160. New York: Marcel Dekker. Buck, K.W. (1979b). Vir ion-associated RNA polymerases of double- stranded RNA mycoviruses In 'Fungal Viruses', Eds. H.P. Molitoris, H.A. Wood and M. Hollings, pp. 62, 77. Berlin, Heidelberg and New York: Springer Verlag. 241 Buck, K.W. (1980). Viruses and killer factors of fungi. In 'The Eukaryotic Microbial Cell', Eds. G.W. Gooday, D. Lloyd and A.J. Trinci. Society for General Microbiology Symposium, 30, 329-375. Cambridge University Press, Cambridge. Buck, K.W., Almond, M.R., McFadden, J.J.P., Romanos, M.A. and Rawlinson, C.J. (1981). Properties of thirteen viruses and virus variants obtained from eight isolates of the wheat take-all fungus Gaeumannomyces graminis var tritici. Journal of General Virology, 53, 235-245. Buck. K.W. and Girvan, R.F. (1977). Comparison of the biophysical and biochemical properties of Penicillium cyaneo-fulvum virus and Penicillium chrysogenum virus. Journal of General Virology, 34, 145-154. Buck, K.W., Girvan, R.F. and Ratti, G. (1973). Two serologically distinct double-stranded ribonucleic acid viruses isolated from Aspergillus niger. Biochemical Society Transactions, J_f 1138-1140. Buck, K.W. and Kempson-Jones, G.F. (1970). Three types of virus particle in Penicillium stoloniferum. Nature, 225, 945-946. Buck, K.W. and Kempson-Jones, G.F. (1973). Biophysical properties of Penicillium stoloniferum virus S. Journal of General Virology, 18, 223-235. Buck, K.W. and Kempson-Jones G.F. (1974). Capsid polypeptides of two viruses isolated from Penicillium stoloniferum. Journal of General Virology, 22, 441-445. Buck, K.W. and Ratti. G. (1975). Biochemical and biophysical properties of two viruses isolated from Aspergillus foetidus. Journal of General Virology, 27, 211-224. Buck, K.W. and Ratti, G. (1977). Molecular weight of double-stranded RNA: a re-examination of Aspergillus foetidus virus S RNA components. Journal of General Virology, 37, 215-219. Bussey, H. (1972). Effects of yeast killer factor on sensitive cells. Nature New Biology, 235, 73-75. 242 Castanho, B. and Butler, E.E. (1975). A transmissible disease and infectious disorder of Rhizoctonia solani. Proceedings of the American Phytopathology Society, 2, 112 (Abstract). Castanho, B., and Butler, E.E. (1978) Rhizoctonia decline: Studies on hypovirulence and potential use in biological control. Phytopathology, 68, 1511-1514. Castanho, B., Butler, E.E. and Shepherd, R.J. (1978). The association of double-stranded RNA with Rhizoctonia decline. Phytopathology, 68, 1515-1518. Castellino, F.J. and Barker, R. (1968). Examination of the dissociation of multichain proteins in guanidine hydrochloride by membrane osmometry. Biochemistry, 2207-2217. Caten, C.E. and Jinks, J.L. (1966). Heterokaryosis: Its significance in wild heterothallic Ascomycetes and Fungi Imperfecti. Transactions of the British Mycological Society, 49, 81-93. Chambers, S.C. and Flentje, N.T. (1967). Studies on oat-attacking and wheat-attacking isolates of Ophiobolus graminis in Australia. Australian Journal of Biological Science, 20, 927-940. Chater, K.F. and Morgan, D.H. (1974). Ribonucleic acid synthesis by isolated viruses of Penicillium stoloniferum. Journal of General Virology, 24, 307-317. Cohn, M.S., Tabor, C.W. and Tabor, H. (1978). Isolation and character- isation of Saccharomyces cerevisiae mutants deficient in S-adenosyl methionine decarboxylase, spermidine and spermine. Journal of Bacteriology, 134, 208-213. Cooper, T.G. (1981). Yeast genetics and molecular biology. Nature, London, 289, 119-120. Coplin, D.L., van Etten, J.L., Koski, R.K. and Vidaver, A.K. (1975). Intermediates in the biosynthesis of double-stranded ribonucleic acids of bacteriophage 06. Proceedings of the National Academy of Sciences, U.S.A, 72, 849-853. 243 Cowan, M.C. (1978). Lignification in wheat roots parasitised by Gaeumannomyces graminis and Phialophora radicicola. Annals of Applied Biology, 89, 101 (Abstract). Cross, R.K. and Fields, B.N. (1977). Genetics of reoviruses. In 'Comprehensive Virology', Ed. H. Fraenkel-Conrat and R.R. Wagner, vol. 9, pp. 291-340. New York: Plenum Press. Cunningham, P.C. (1975). Some consequences of cereal monoculture on Gaeumannomyces graminis (Sacc.) Arx and Olivier and the take-all disease. European and Mediterranean Plant Protection Organisation Bulletin, 5, 297-317. Davis, R.J. (1925). Studies on Ophiobolus graminis Sacc. and the take-all disease of wheat. Journal of Agricultural Chemistry, 31_, 801-825. Day, P.R. and Anagnostakis, S.L. (1973). The killer system in Ustilago maydis. Heterokaryon transfer and loss of determinants. Phytopathology, 63, 1017-1018. Day, P.R. and Dodds, J.A. (1979). Viruses of plant pathogenic fungi. In 'Viruses and Plasmids of Fungi', Ed. P.A. Lemke, pp. 201-238, New York: Marcel Dekker. Day, P.R., Dodds, J.A., Elliston, J.E., Jaynes, R.A. and Anagnostakis, S.L. (1977). Double-stranded RNA in Endothia parasitica. Phytopathology, 67_, 1393-1396. de la Pena, P., Barros, F., Gascon, S., Ramos, S. and P.S. Lazo. (1980). Primary effects of yeast killer toxin. Biochemical and Biophysical Research Communications, 96, 544-550. Deacon, J.W. (1973). Control of the take-all fungus by grass leys in intensive cereal croppings. Plant Pathology, 22, 88-94. Deacon, J.W. (1974). Interaction between varieties of Gaeumannomyces graminis and Phialophora radicicola on roots, stem bases and rhizomes of the Gramineae. Plant pathology, 23, 85-92. Deacon, J.W. (1976). Biological control of the take-all fungus, Gaeumannomyces graminis, by Phialophora radicicola and similar fungi. Soil Biology and Biochemistry, 8, 275-283. 244 Del Vecchio, V.G., Dixon, C. and Lemke, P.A. (1978). Immune electron microscopy of virus-like particles of Agaricus bisporus. Experimental Mycology, 2, 138-144. Dieleman-van Zaayen, A. (1979). Mushroom viruses. In 'Viruses and Plasmids of Fungi', Ed. P.A. Lemke, pp. 239-324. New York: Marcel Dekker. Dieleman-van Zaayen, A. and Temmink, J.H.M. (1968). A virus disease of cultivated mushrooms in the Netherlands. Netherlands Journal of Plant Pathology, 74, 48-51. Dobos, P., Hallett, R., Kells, D.T.C., Hill, B.J., Becht, H. and Teninges, D. (1978). Biochemical and biophysical studies of five animal viruses with bisegmented dsRNA genomes. Abstracts of the Fourth International Congress for Virology, p. 335; Wageningen: Centre for Agricultural Publishing and Documentation. Dodds, J.A. (1979). Double-stranded RNA and virus-like particles in Endothia parasitica. In American Chestnut Symposium Proceedings, Ed. W. McDonald. West Virginia University Agricultural Experiment Station and United States Dept. of Agriculture. Dodds, J.A. (1980). Association of type I viral-like dsRNA with club- shaped particles in hypovirulent strains of Endothia parasitica. Virology, 107, 1-12. Dodds, J.A. and Elliston, J.E. (1978). Association between double- stranded RNA and hypovirulence in an American strain of Endothia parasitica. Abstracts of the Third International Congress for Plant Pathology, p 57; Hamburg: Paul Parey. Dodds, J.A. and Hamilton, R.I. (1976). Structural interactions between viruses as a consequence of mixed infections. Advances in Virus Research, 20, 33-86. Douthart, R.J., Burnett, J.P., Beasley, F.W. and Frank, B.H. (1973). Binding of ethidium bromide to double-stranded ribonucleic acid. Biochemistry, J2, 214-220. 245 Ellis, L.F. and Kleinschmidt, W.J. (1967). Virus-like particles of a fraction of statolon, a mould product. Nature, London, 215, 649-650. Esser, K. and Blaich, R. (1973). Heterogenic incompatibility in plants and animals. Advances in Genetics, V7, 107-152. Fenner, F. (1974). In 'The Biology of Animal Viruses', 2nd. edition. London and New York: Academic Press. Fink, G.R. and Styles, C.A. (1972). Curing of a killer factor in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences, U.S.A., 69, 2846-2849. Frick, L.J. and Lister, R.M. (1978). Serotype variability in virus-like particles from Gaeumannomyces graminis. Virology, 85, 504-517. Fried, H.M. and Fink, G.R. (1978). Electron microscopic heteroduplex analysis of 'killer' double-stranded RNA species of yeast. Proceedings of the National Academy of Sciences, U.S.A., 75, 4224-4228. Garrett, S.D. (1948). Soil conditions and the take-all disease of wheat. IX. Interaction between host plant nutrition, disease escape, and disease resistance. Annals of Applied Biology, 35, 14-17. Geelen, J.L.M.C., Rezelman, G. and Van Kammen, A. (1973). The infectivity of the two electrophoretic forms of cowpea mosaic virus. Virology, 279-286. Ghabrial, S.A. (1980). Effects of fungal viruses on their hosts. Annual Review of Phytopathology, JH3, 441-461. Ghabrial, S.A., Sanderlin, R.S. and Calvert, L.A. (1979). Morphology and virus-like particle content of Helminthosporium victoriae colonies regenerated from protoplasts of normal and diseased isolates. Phyto- pathology, 69, 312-315. Gomori, G. (1955). Preparation of buffers for use in enzyme studies. In 'Methods in Enzymology', vol. 1. Eds. S.P. Colowick and N.O. Kaplan, pp. 138-146. New York: Academic Press. 246 Hankin, L. and Puhalla, J.E. (1971). Nature of a factor causing inter- strain lethality in Ustilago maydis. Phytopathology, 61_, 50-53. Hardy, K.G. (1975). Colicinogeny and related phenomena. Bacteriological Reviews, 39, 464-515. Hardy, D.B., Rosen, L. and Fields, B.N. (1979). Polymorphism of the migration of double-stranded RNA genome segments of reovirus isolates from humans, cattle and mice. Journal of Virology, 104-111. Haskins, R. (1950). Biochemistry of the Ustilaginales. 1. Preliminary cultural studies of Ustilago zeae. Canadian Journal of Research, 28, section C, 213-223. Hastie, N.D., Brennan, V. and Bruenn, J.A. (1978). No homology between double-stranded RNA and nuclear DNA of yeast. Journal of Virology, 28, 1002-1005. Herring, A.J. and Bevan, E.A. (1975). Double-stranded RNA containing particles from the yeast Saccharomyces cerevisiae and their relationship to the killer character. In 'Molecular Biology of Nucleocytoplasmic Relationships', Ed. S. Puiseux-Dao, pp. 149-154. Amsterdam: Elsevier Scientific Publishing Company. Herring, A.J. and Bevan, E.A. (1977). Yeast virus-like particles possess a capsid-associated single-stranded RNA polymerase. Nature, London, 268, 464-466. Holland, J.J., Grabau, E.A., Jones, C.L. and Semier, B.L. (1979). Evolution of multiple genome mutations during long-term persistent infection by Vesicular stomatitis Virus. Cell, J[6, 495-504. Hollings, M. (1962). Viruses associated with a die-back disease of cultivated mushroom. Nature, London, 196, 962-965. Hollings, M. (1978). Mycoviruses - viruses that infect fungi. Advances in Virus Research, 22, 3-53. Hollings, M. and Stone, O.M. (1971). Viruses that infect fungi. Annual Review of Phytopathology, 9, 93-118. 247 Hopper, J.E., Bostian, K.A., Rowe, L.B. and Tipper, D.J. (1977). Translation of the L-species dsRNA genome of the killer-associated virus-like particle of Saccharomyces cerevisiae. Journal of Biological Chemistry, 252, 9010-9017. Hornby, D. (1979). Take-all decline: a theorist's paradise. In 'Soil- borne Plant Pathogens', Eds. B. Schippers and W. Gams, pp 133-156. Academic Press, London and New York. Hornby, D., Slope, D.B., Gutteridge, R.J. and Sivanesan, A. (1977). Gaeumannomyces cylindrosporus, a new ascomycete from cereal roots. Transactions of the British Mycological Society, 69, 21-25. Huang, A.S. and Baltimore, D. (1977). Defective interfering animal viruses. In 'Comprehensive Virology', Eds. H. Fraenkel-Conrat and P.R. Wagner, vol. 10. pp. 73-116. New York: Plenum Press. Joklik, W.K. (1974). Reproduction of reoviridae. In 'Comprehensive Virology' Eds. H. Fraenkel-Conrat and R.R. Wagner, vol. 2, pp. 231-334. New York: Plenum Press. Kotani, H., Shinmyo, A and Enatsu, T. (1977). Killer toxin for sake yeast: Properties and effects of adenosine 5'-diphosphate and calcium ion on killing action. Journal of Bacteriology, 129, 640-650. Koltin, Y. (1977). Virus-like particles in Ustilago maydis: mutants with partial genomes. Genetics, 86, 527-534. Koltin, Y. and Day, P.R. (1975). Specificity of Ustilago maydis killer proteins. Applied Microbiology, 30, 694-696. Koltin, Y. and Day, P.R. (1976a). Inheritance of killer phenotypes and double-stranded RNA in Ustilago maydis. Proceedings of the National Academy of Sciences, U.S.A., 73, 594-598. Koltin, Y. and Day. P.R. (1976b). Suppression of the killer phenotype in Ustilago maydis. Genetics, 82, 629-637. Koltin, Y. and Kandel, J.S. (1978). Killer phenomenon in Ustilago maydis: the organisation of the viral genome. Genetics, 88, 267-276. 248 Lampson, G.P., Tytell, A.A., Field, A.K., Nemes, M.M. and Hilleman, M.R. (1967). Inducers of interferon and host resistance. I. Double-stranded RNA from extracts of Penicillium funiculosum. Proceedings of the National Academy of Sciences, U.S.A., 58, 782-789. Lapierre, H. (1973). Etude de l1influence des virus sur les champignons phytopathogenes du sol. In 'erspectives de lutte Biologique des Champignons Parasites des Plantes Cultivees et les Pourritures des Tissus Ligneux', pp. 62-64, Station Federale de Recherches Agronomiques de Lausanne, Switzerland. Lapierre, H., Lemaire, J.M., Jouan, B and Molin, G. (1970). Mise en evidence de particules virales associees a une perte de pathogenicite chez le Pietin-echaudage des cereales, Ophiobolus graminis Sacc. C.r hebd. Seanc. Acad. Sci. Paris, 271, 1833-1836. o Laskey, R.A. and Mills, A.D. (1975). Quantitative film detection of H 14 and C in polyacrylamide gels by fluorography. European Journal of Biochemistry, 56, 335-341. Lemaire, J.M., Doussinault, G., Jouan, B. and Tivoli, B. (1975). La luttte oontre les maladies des cereales par des moyens biologique. Bulletin Technique D1Information, Ministere Agriculture, Paris 297, 195-200. Lemaire, J.M., Jouan, B., Perraton, B. and Sailly, M. (1971). Perspectives de lutte biologique contre les parasites des cereales d'origine tellurique en particulier Ophiobolus graminis Sacc., Sciences Agronomiques Rennes, 1977, pp. 1-8. Lemaire, J.M., Lapierre, H., Jouan, B. and Bertrand, G. (1970). Decouverte de particules virales chez certains souches d'Ophiobolus graminis, agent du Pietin echaudage des cereales: consequences agroncmique previsibles. C. r. hebd. Seanc. Acad. Agric. Fr. 56, 1134-1138. Lemke. P.A. (1976). Viruses of eukaryotic microorganisms. Annual Review of Microbiology, 30, 105-145. Lemke, P.A. (1977). Fungal viruses in agriculture. In 'Virology in Agriculture. Eds. T.O. Diener and J.A. Romberger, pp. 159-175. Montclair, New Jersey: Allanheld Osman and Co. 249 Lester, E. and Shipton, P.J. (1967). A technique for studying inhibition of the parasitic activity of Ophiobolus graminis (Sacc.) in field soils. Plant Pathology, J6r 121-123. Levine, R., Koltin, Y. and Kandel, J. (1979). Nuclease activity associated with the Ustilago maydis virus induced killer proteins. Nucleic Acids Research, 6> 3717-3731. Lhoas, P. (1971). Infection of protoplasts from Penicillium stoloniferum with double-stranded RNA viruses. Journal of General Virology, 13, 365-367. Lilly, V.G. and Barnett, H.L. (1951). 'Physiology of the Fungi'. McGraw-Hill Co., New York. Lindberg, G.D. (1959). A transmissible disease of Helminthosporium victoriae. Phytopathology, 4£, 29-32. Loviny, T. and Szekely, M. (1973). Fingerprinting double-stranded non-radioactive RNA from a fungal virus. European Journal of Biochemistry, 35, 87-94. Marcus, A., Efron, D. and Weeks, D.P. (1974). The wheat embryo cell-free system. In 'Methods in Enzymology', vol. 30, part F, pp. 749-754. Eds. K. Moldave and L. Grossman. New York and London: Academic Press. McMaster, G.K. and Carmichael, G.C. (1977). Analysis of single and double-stranded nucleic acids on polyacrylamide gels using glyoxal and acridine orange. Proceedings of the National Academy of Sciences, U.S.A., 74, 4835-4838. Mitchell, D.J., Bevan, E.A. and Herring, A.J. (1973). The correlation between dsRNA in yeast and the "killer" character. Heredity, 133 (Abstract). Moffitt, E.M. and Lister, R.M. (1975). Application of a serological screening test for detecting double-stranded RNA mycoviruses. Phyto- pathology, 65, 851-859. 250 Mossop, D.W. and Franki, R.I.B. (1978). Survival of a satellite RNA in vivo and its dependence on cucumber mosaic virus for its replications. Virology, 86, 562-566. Nash, C.H., Douthart, R.J., Ellis, L.F., van Frank, R.M., Burnett, J.P. and Lemke, P.A. (1973). On the mycophage of Penicillium chrysogenum. Canadian Journal of Microbiology 97-103. Nilsson, H.E. (1969). Studies of root and foot rot diseases of cereals and grasses. 1. On resistance to Ophiobolus graminis Sacc. Annals of the Agricultural College of Sweden, ^5* 275-807. Oliver, S.G. McCready., S.J. Holm., C., Sutherland, P.A., McLaughlin, C.S. and Cox, B.S. (1977). Biochemical and physiological studies of the yeast virus-like particle. Journal of Bacteriology, 130, 1303-1309. Palfree, R.G.E. and Bussey, H. (1979). Yeast killer toxin: purification and characterisation Biochemistry, 93, 487-493. Pallett, I.H. (1976). Interactions between fungi and their viruses. In 'Microbial and Plant Protoplasts', Eds. J.F. Peberdy, A.H. Rose, H.J. Rogers and E.C. Cocking, pp. 107-124. London: Academic Press. Partridge, J.E., Vidaver, A.K. and van Etten, J.L. (1978). In vitro transcription of bacteriophage 06 double-stranded RNA. Abstracts of the Annual Meeting of the American Society for Microbiology, S53, 221. Passmore, E.L. and Frost, R.R. (1974). The detection of virus-like particles in mushrooms and mushroom spores. Phytopathologische Zeitschrift, 80, 85-87. Philliskirk, G. and Young, T.W. (1975). The occurrence of killer character in yeasts of various genera. Antonie van Leeuwenhoek, 41, 147-151. Polley, R.W. and Clarkson, J.D.S. (1980). Take-all severity and yield in winter wheat: relationship established using a single plant assessment method. Plant Pathology, 29, 110-116. Powell, H.M., Culbertson, C.G., McGuire, J.M., Hoehn, M.M. and Barker, L.A. (1952). A filtrate with chemoprophylactic and chemotherapeutic action against MM and Semiliki Forest virus in mice. Antibiotics and Chemotherapy, JJ_, 433-434. 251 Puhalla, J.E. (1968). Compatibility reactions on solid medium and interstrain inhibition in Ustilago maydis. Genetics, 60, 461-474. Ratti, G. and Buck, K.W. (1972). Virus particles in Aspergillus foetidus: a multicomponent system. Journal of General virology, 14, 165-175. Ratti, G. and Buck, K.W. (1975). RNA polymerase activity in double- stranded ribonucleic acid virus particles from Aspergillus foetidus. Biochemical and Biophysical Research Communications, 66, 706-711. Ratti, G. and Buck, K.W. (1978). Semi-conservative transcription in particles of a double-stranded RNA mycovirus. Nucleic Acids Research, 5, 3843-3854. Rawlinson, C.J., Hornby, D., Pearson, V. and Carpenter. J.M. (1973). Virus-like particles in the take-all fungus, Gaeumannomyces graminis. Annals of Applied Biology, 74, 197-209. Rawlinson, C.J. and Muthyalu, G. (1976). Report of the Rothamsted Experimental Station for 1975, p. 255. Rawlinson, C.J., Muthyalu, G. and Deacon, J.W. (1977). Natural transmission of viruses in Gaeumannomyces and Phialophora spp. Second International Mycological Congress, Tampa, Florida, 1977, p. 558. (Abstract). Rimon, A. and Haselkorn, R. (1978). Temperature-sensitive mutants of bacteriophage 06 defective in both transcription and replication. Virology, 89, 218-228. Rogers, D. and Bevan, E.A. (1978). Group classification of killer yeasts based on cross reactions between strains of different species and origin. Journal of General Microbiology, 105, 199-202. Romanos, M.A. (1981). The structure and function of some viruses of Gaeumannomyoces graminis var. tritici. Ph.D. Thesis. University of London. Romanos, M.A., Rawlinson, C.J., Almond, M.R. and Buck, K.W. (1980). Production of fungal growth inhibitors by isolates of Gaeumannomyces graminis var. tritici. Transactions of the British Mycological Society9 74, 79-88. 252 Sanderlin, R.S. and Ghabrial, S.A. (1978). Physioochemical properties of two distinct types of virus-like particles from Helminthosporium victoriae. Virology, 87, 142-151. Schumaker, V.N. and Schachman, H.K. (1957). Ultracentrifugal analysis of dilute solutions. Biochimica et Biqphysica Acta, 23, 628-639. Scott, P.R. (1970). Phialophora radicicola, an avirulent parasite of wheat and grass roots. Transactions of the British Mycological Society, 55, 163-167. Shatkin, A.J. and Sipe, J.D. (1968). FNA polymerase activity in purified reoviruses. Proceedings of the National Academy of Sciences, U.S.A., 61, 1462-1469. Shope, R.E. (1948). The therapeutic activity of a substance from Penicillium funiculosum Thorn against swine influenza virus infection of mice. American Journal of Botany, 35/ 803 (Abstract). Silverstein, S.C., Christman, J.K. and Acs, G. (1976). Ihe reovirus replicative cycle. Annual Review of Biochemistry, 45, 375-408. Sinden, J.W. and Hauser, E. (1950). Report on two new mushroom diseases. Mushroom Science, 96-100. Sivasithamparam, K. (1975). Phialophora and Phialophora-1 ike fungi occurring in the root region of wheat. Australian Journal of Botany, 23, 193-212. Sivasithamparam, K., Stukely, M. and Parker, C.A. (1975). A volatile factor inducing transmissible lysis in Gaeumannomyces graminis (Sacc.) Arx and Olivier var. tritici Walker. Canadian Journal of Microbiology, 21, 293-300. Skipper, N. and Bussey, H. (1977). Mode of action of yeast toxins: energy requirement for Saccharomyces cerevisiae killer toxin. Journal of Bacteriology, 129, 669-677. Slope, D.B., Prew, R.D., Gutteridge, R.J. and Etheridge, J. (1979). Take-all, Gaeumannomyces graminis var. tritici and yield of wheat grown after ley and arable rotations in relation to the occurrence of Phialophora radicicola var. graminicola. Journal of Agricultural Science, (Cambridge), 93, 377-389. 253 Slope, D.B., Salt, G.A., Broom, E.W. and Gutteridge, R.J. (1978). Occurence of Phialophora radicicola var. graminicola and Gaeumannomyces graminis var. tritici on roots of wheat in field crops. Annals of Applied Biology, 88, 239-246. Smith, E.L., Austen, D.M., Blumenthal, K.M. and Nyc, J.F. (1975). Glutamate dehydrogenases. In 'The Enzymes', vol. 11. pp. 293-367. Ed. P.D. Boyer, London: Academic Press. Speakman, J.B. (1977). Interactions between Gaeumannomyces graminis and Phialophora radicicola in relation to take-all disease of wheat. Ph.D. Thesis, University of East Anglia, Norwich. Speakman, J.B. and Lewis, B.G. (1978). Limitation of Gaeumannomyces graminis by wheat root responses to Phialophora radicicola. New Phytolog^, 80, 373-380. Speakman, J.B. and Lewis, B.G. (1980). Vitamin requirements of Phialophora radicicola var. graminicola compared with Gaeumannomyces graminis var. tritici. Transactions of the British Mycological Society, 74, 410-413. Squire, P.G., Moser, P. and O'Kinski, C.T. (1968). The hydronamic properties of bovine serum albumin and dimer. Biochemistry, 1_, 4261- 4272. Sugiura, M. and Miura, K. (1977). Transcription of double-stranded RNA by Escherichia coli DNA-dependent RNA polymerase. European Journal of Biochemistry, 73, 179-184. Szybalski, W. (1968). In 'Methods in Enzymology', Eds. S.P. Colowick and N.O. Kaplan. Vol. 12, Part B, pp. 330-360. London and New York: Academic Press. Tivoli, B., Lemaire, J.M. and Jouan, B. (1974). Premunition du ble contre Ophiobolus graminis Sacc. par des souches peu aggressive du meme parasite. Annales de Phytopathologie. j5, 395-406.. Toh-E, A., Guerry, P. and Wickner, R.B. (1978). Chromosomal superkiller mutants of Saccharomyces cerevisiae. Journal of Bacteriology, 136, 1002-1007. 254 van Alfen, N.K., Jaynes, R.A., Anagnostakis, S.L. and Day, P.R. (1975). Chestnut blight: biological control by transmissible hypovirulence in Endothia parasitica. Science, 189, 890-891. van Etten, J.L., Burbank, D.E., Cuppels, D.A., Lane L.C. and Vidaver, A.K. (1980). Semi-conservative synthesis of single-stranded RNA by bacteriophage 06 polymerase. Journal of Virology, 33, 769-777. van Etten, J.L., Lane, L.C., Gonzalez, C., Partridge, J. and Vidaver, A.K. (1976). Comparative properties of bacteriophage 06 and 06 nucleocapsid. Journal of Virology, J8/ 652-658. van Etten, J.L., Vidaver, A.K., Koski, R.K. and Semancik, J.S. (1973). RNA polymerase activity associated with bacteriophage 06. Journal of Virology, 12, 464-471. Vandewalle, M.J. and Siegel, A. (1976). A study of nucleotide sequence homology between strains of tobacco mosaic virus. Virology, 73, 413-418. Vodkin, M. (1977). Homology between double-stranded RNA and nuclear DNA of yeast. Journal of Virology, 2J_, 516-521. Walker, J. (1981). Taxonomy of take-all fungi and related genera and species. In 'Biology and Control of Take-all'. Eds. M.J.C. Asher, and P.J. Shipton, P.J. New York and London: Academic Press. Weber, K., Pringle, J.R. and Osborn, M. (1972). Measurement of molecular weights by electrophoresis on SDS-acrylamide gel. In 'Methods in Enzymology' vol. 26, part C, pp. 3-27. Eds. C.H.W. Hirs and S.N. Tamashieff. New York and London: Academic Press. Weste, G. and Thrower, L.B. (1963). Production of perithecia and microconidia in culture by Ophiobolus graminis. Phytopathology, 53, 354. Wickner, R.B. (1974). Chromosomal and non-chromosomal mutations affecting the 'killer character' of Saccharomyces cerevisiae. Journal of Bacteriology, 117, 681-686. Wickner, R.B. (1976). Killer of Saccharomyces cerevisiae: a double- stranded ribonucleic acid plasmid. Bacteriological Reviews, 40, 757-773. 255 Wickner, R.B. (1978). Twenty-six chromosomal genes needed to maintain the killer double-stranded RNA plasmid of Saccharaomyces cerevisiae. Genetics, 88, 419-425. Wickner, R.B. and Leibowitz, M.J. (1976). Two chromosomal genes required for killing expression in killer strains of Saccharomyces cerevisiae. Genetics, 82, 429-442. Wong, P.T.W. (1975). Cross-protection against the wheat and oat take-all fungi by Gaeumannomyces graminis. var. graminis. Soil Biology and Biochemistry, 7, 189-194. Wong, P.T.W. (1981). Biological control of take-all by cross-protection. In 'Biology and Control of Take-all', Eds. M.J.C. Asher and P.J. Shipton. London and New York: Academic Press. Young, T.W. and Yagiu, M. (1978). A comparison of the killer character in different yeasts and its classification. Antonie van Leeuwenhoek, 44, 59-77. Zweerink, H.J., McDowell, M.J. and Joklik, W.K. (1971). Essential and nonessential noncapsid reovirus proteins. Virology, £5, 716-723. Vol.98, No. 2,1981 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS January 30, 1981 Pages 501-506 TRANSCRIPTASE ACTIVITY ASSOCIATED WITH A TYPE 2 DOUBLE-STRANDED RNA MYCOVIRUS R.M. McGinty , K.V. Buck and C.J. Raulxnson ^Department of Biochemistry Imperial College of Science and Technology London SW7 2AZ UK ^Plant Pathology Department Rothamsted Experimental Station Harpenden Herts AL5 2JQ UK Received December 2,1980 SUMMARY It is shown that the virion-associated RNA polymerase of Phialophora virus A, a type 2 double-stranded RNA mycovirus with a genome consisting of three RNA species, is a transcriptase. Synthesis of single stranded RNA in vitro continues for a least 24 hours and after this time two full length transcripts are produced, on average, per dsRNA molecule, i.e.re-initiation of transcription occurs in this in vitro system. Analysis of the products by polyacrylamide gel electrophoresis indicated that the efficiences of transcription of all three double-stranded RNA species were similar. The transcriptase activity of Phialophora virus A differs from the replicase activity of Penicillium stoloniferum virus S, the only other type 2 double- stranded RNA mycovirus whose RNA polymerase has been characterised. INTRODUCTION Isometric viruses with genomes of dsRNA occur commonly in fungi (1). There are two basic types; type 1 viruses have an undivided genome of polycist- ronic dsRNA, whereas type 2 viruses have a genome which is divided into two or more segments of monocistronic dsRNA, Viruses of either type may carry, in some fungal strains, additional segments of satellite or defective dsRNAs which are not essential for virus replication (2), Virion-associa- ted RNA polymerases have been found in viruses of both types (3), although Abbreviations: dsRNA - double-stranded RNA; ssRNA - single-stranded RNA; TCA - trichloroacetic acid. 0006-291X/81/020501-06$01.00/0 Copyright © 1981 by Academic Press, Inc. 501 All rights of reproduction in any form reserved. Vol. 98. No. 2.1981 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS only a few of these have been adequately characterised. In Saccharomyces cerevisiae L virus , a type 1 virus^, the RNA polymerase in mature virions is a transcriptase which catalyses the synthesis and release of full-length ssRNA copies of one of the strands of the dsRNA genome (4, 5). Transcrip- tase activity iaaa also b^en detected in virions of two nrher mycoviruses with polycistTonic ds"RNAs, Aspergillus foetidrrs vims S <-6) "and Alltmyces arbuscula virus (7). In contrast the virion-associated RNA polymerase of Penicillium stoloniferum virus S, a type 2 virus, is a replicase, which catalyses the synthesis in vitro of a new molecule of dsRNA, which remains within the particles, giving rise to diploid virions (8). Since the Penicillium virus replicase was the only RNA polymerase of a type 2 dsRNA mycovirus to have been characterised, it was of interest to know if replicase activity is characteristic of other type 2 dsRNA mycoviruses. Two serologically unrelated type 2 dsRNA mycoviruses (A and B) have been obtained from an isolate of a Phialophora sp. parasitic on barley roots (9) We now report that both of these viruses have virion-associated RNA poly- merase activity and that the virus A enzyme is a transcriptase, which catalyses the synthesis of ssRNA copies of each of the three virus dsRNA components. This activity differs from that of Penicillium stoloniferum virus S (8) and this is the first report of transcriptase activity associated with particles of a type 2 dsRNA mycovirus. METHODS Preparation of viruses. Viruses A and B from Phialophora sp.(lobed hyphopodia) isolate 2-2 were isolated, purified and separated by the methods described previously (10). Virus preparations were finally dialysed against TNE buffer (0.05M-tris-HCl + 0.15M - NaCl +. O.lmM- EDTA, pH 7.9) and stored at -20°C after addition of an equal volume of glycerol. RNA polymerase assay. Standard reaction mixtures contained 0.15mM-ATP; 0.15mM - GTP; 0.15mM - CTP; 0.15mM - [3H]UTP (sp. act. 17 to 68mCi/mmol) MgCl2 (2.5mM,virus A; 5.0mM, virus B); actinomycin D (lOOyg/ml); bentonite 25 to 5 TNE (800pg/ml); virus (A26o °- )5 buffer. Incubation was at'30°C. Incorporation of [^H]UMP into acid insoluble material was determined by adding 3ml of 10% (w/v) TCA to 10 to lOOyl of chilled reaction mixture 502 Vol. 98. No. 2.1981 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS and allowing precipitation to occur at 0°C for 30 minutes. The precipitate was collected on Whatman GF/F filters and washed extensively with 21 TCA and then with ethanol. Filters were dried and radioactivity was determined by liquid scintillation counting in a toluene-based scintillation fluid. Preparation of RNA. Solutions were made 1Z in sodium dodecyl sulphate and then extracted with phenol. The ethanol precipitated RNA was xe&uspended in th£ required .buffer and then dialysed against the same buffer, "For separating ssHNA and dsKHA selective precipitation of ssRNA in 2M-LiCl vas used (11) . Analysis of RNA by electrophoresis. Electrophoresis of RNA was carried out in 4Z polyacrylamide slab gels in a tris-acetate-EDTA buffer containing 8M-urea (12) at 5 volts/cm for 15h. Gels were either stained with 0.01Z aqueous toluidine blue or examined by fluorography (13). Denaturation and glyoxalation of dsRNA prior to electrophoresis, when required, was carried out as described by McMaster & Carmichael (14), except that 60Z (v/v), rather than 50Z (v/v), dimethyl sulphoxide was employed in the reaction mixture. RNA hybridization assays. These were carried out by the Method B described by Ratti and Buck (6). RESULTS AND DISCUSSION Both viruses A and B catalysed the incorporation of [3H]UMP into TCA- insoluble material in RNA polymerase reaction mixtures. Both reactions were dependent on magnesium ions with optima of 2.5mM and 5mM respectively and no pretreatment of the virions with a proteinase or by heat shock treatment was necessary to activate the polymerase as is the case with certain other dsRNA viruses e.g. reovirus (15). Omission of one of the four nucleoside triphosphates almost completely abolished the reaction in both cases. In the case of virus A [3H]UMP incorporation was linear for at least 30 h whereas the virus B reaction slowed down after a short period of time (Fig.l), probably due to the greater instability of this virus under the reaction conditions. The products of an 18h RNA polymerase reaction with virus A were isolated by phenol/sodium dodecyl sulphate extraction and examined by electro- phoresis in polyacrylamide gels containing 8M-urea. After staining of gels with toluidine blue six clear bands were revealed, three bands corresponding to virus template dsRNAs, mol. wt. 1.29 x 10^, 1.22 x 10& and 1.03 x 10^ and three bands of newly synthesised RNA of slower mobility. 503 Vol. 98. No. 2.1981 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 10 20 Tim«,h Fig. 1: RNA synthesis catalysed by Phialophora viruses A and B RNA polymerases. Reaction mixtures containing [-*H]UTP (specific activity 17.5 mCi/mmol) and virus A or B (2 A2go units) were incubated at 30°C as described in the Methods. At the indicated times, lOyl samples were withdrawn and the radioactive material insoluble in cold TCA was determined. • virus A; # virus B. When the RNA products were incubated with ribonuclease A (0.25iig/ml) for 2h in hybridisation buffer (6), prior to electrophoresis, the three slower moving bands could no longer be observed, but the three bands of virus dsRNA Were unchanged. The three newly synthesised RNA components are therefore single-stranded. In order to prove that the ssRNA. products arose by transcription of the o virus template dsRNAs, the [ H] ssRNA products of an 18h polymerase react- ion were isolated by precipitation with 2M-LiCl and annealed with increas- ing amounts of denatured unlabelled virus dsRNA. The extent of hybridisa- tion which occurred was measured by the amount of label which remained TCA-insoluble after treatment with ribonuclease A in hybridisation buffer (6). The proportion of products capable of hybridising to the virus template RNA increased with the amount of virus dsRNA up to at least 85% (Fig.2). No self-annealing of the product, in the absence of added denatured virus dsRNA, could be detected. Co-electrophoresis experiments 504 Vol. 98. No. 2.1981 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Waight ratio d»RNA:«sRNA Fig.2: Hybridisation of virus A RNA polymerase product to virus template dsRNA. 3H-labelled ssRNA product (17 000 ct/min) was hybridised with increasing amounts of unlabelled denatured virus dsRNA. Hybridisation is expressed as the percentage of input ssRNA label which becomes resistant to ribonuclease A in high salt buffer. showed that the three ssRNA products had mobilities identical to those of the three virus dsRNAs, after denaturation and glyoxalation. It was concluded that these products are full length ssRNA transcripts of one of the strands of each of the three virus dsRNA components. The amount of ssRNA product synthesised after a 24h RNA polymerase reaction was calculated (a) from absorbance at 260nm, assuming A26O 1 is equivalent to AOyg/ml ssRNA and 50yg/ml dsRNA and (b) from [JH]UMP incorporation, assuming a UMP content of 25% for ssRNA. Both methods indicated that a weight of ssRNA approximately equal to that of the template dsRNA had been synthesised, i.e. on average, two rounds of transcription per dsRNA molecule had occurred. This result shows that re-initiation of transcription can occur in the _in vitro system. The proportions of individual transcripts, as judged from intensities of bands in toluidine blue stained gels or after fluorography of gels, following electrophoresis of product RNA, were similar to those of the template dsRNA molecules, indicating approximately equivalent efficiencies of transcrip- tion of each of the three dsRNA components. The results have shown that the virion-associated RNA polymerase of the 505 Vol. 98. No. 2.1981 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Phialophora virus A is a transcriptase which catalyses the synthesis of full-length ssRNA copies of one of t^he strands of each of the three virus dsRNA template molecules. This polymerase differes fundamentally from that of Penicillium stoloniferum virus S which is a replicase. Another difference is that transcription can "be initiated in Vitro by the Phialophora vim enzyme, whereas the Penicillium virus replicase is unable to initiate the reaction and RNA synthesis takes place only in particles which contain ssRNA primers in addition to template dsRNA(8). These results, which show for the first time transcriptase activity in virions of a type 2 dsRNA mycovirus, suggest sub-division of the type 2 viruses into type 2a, those with viripn-associated replicase activity and type 2b, those with virion-associated transcriptase activity. ACKNOWLEDGEMENT. We thank the Science Research Council for the award of a studentship (to R. M. McGinty). REFERENCES 1. Ghabrial, S. (1980). Annu. Rev. Phytopathol. 18, 441-461. 2. Buck, K.W. (1980) in The Eukaryotic Microbial Cell (Gooday, G.W., Lloyd, D. 4 Trinci, A.P.J., eds.), 30th Symp. Soc. Gen. Microbiol., pp.329-375, Cambridge University Press, London. 3. Buck, K.W. (1979) in Fungal Viruses (Molitoris, H.P., Wood, H.A. 4 Hollings, M.), pp.62-67, Springer Verlag, Berlin, Heidelberg 4 New York. 4. Herring, A.J. 4 Beavan, E.A. (1977). Nature (London), 268, 464-466. 5. Bruenn, J., Bobek, L., Brennan, V. 4 Held, W. (1980). Nucleic Acids Res. 8, 2985-2997. 6. Ratti, G. 4 Buck, K.W. (1979). J. Gen. Virol. 42, 59-72. 7. Khandjian, E.W. 4 Turian, G. (1977). FEMS Microbiology Lett.2, 121-124. 8. Buck, K.W. (1975). Nucleic Acids Res. 2, 1889-1902. 9. Rawlinson, C.J. 4 Buck, K.W. (1981) in Biology and Control of Take-All (Shipton, P. 4 Asher, M.J.C., eds.), Academic Press, London 4 New York, in press. 10. Buck, K.W., Almond, M.R., McFadden, J.J.P., Romanos, M.A. 4 Rawlinson, C.J. (1981). J. Gen. Virol., in press. 11. Balitmore, D. (1966). J. Mol. Biol. 18, 421-428. 12. Buck, K.W. 4 Ratti, G. (1977). J. Gen. Virol. 37, 215-219. 13. Laskey, R.A. 4 Mills, A.D. (1975). Eur. J. Biochem. 56, 335-341. 14. McMaster, G.K. 4 Carmichael, G.C. (1977). Proc. Nat. Acad. Sci. U.S.A., 74, 4835-4838. 15. Shatkin, A.J. 4 Sipe, J.D. (1968). Proc. Nat. Acad. Sci., U.S.A. 61, 1426-1469. 506 fpROoTi J. gen. Virol. (1981). 55.000-000. Printed in Great Britain 00 1 Key words: Phialophora sp./virus isolates/physiochemicalproperties Two Serologically Unrelated Viruses Isolated from a Phialophora sp. {Accepted3 March 1981) SUMMARY Two isometric viruses, with similar diameters (34 to 36 nm). obtained from an isolate of a Phialophora sp. with lobed hyphopodia were separated by salt gradient elution from DEAE-celiulose and from SP-Sephadex. Virus A had a sedimentation coefficient of 119S, three double-stranded RNA components of mol. wt. 1 -29 x 106. 1-22 x 106 and 1-03 x 106 and one major capsid polypeptide, mol. wt. 60000. Virus B had a sedimentation coefficient of 122S, three double-stranded RNA components of mol. wt. 1-32 xlO6, 1-25 x 106 and 1-03 x 106 and one major capsid polypeptide, mol. wt. 66000. The two viruses were unrelated serologically to each other and to 13 viruses obtained from a related fungus Gaeumannomyces graminis var. tritici. Although isometric virus-like particles have been reported to occur in more than 100 species of fungi (Hollings. 1978; Buck, 1980). very few have been adequately characterized and many more data are required if progress is to be made in their taxonomy (Hollings. 1979). As part of our programme investigating possible biological effects of viruses on the wheat take-all fungus, Gaeumannomyces graminis var. tritici (hereafter called G. graminis) and related fungi we have detected isometric panicles in several isolates of a Phialophora sp. with lobed hyphopodia. an avirulent cereal root parasite related to G. graminis (Walker. 1931). In the present paper we report on the separation and properties of two viruses obtained from one of these isolates. A Phialophora sp. with lobed hyphopodia, designated as isolate 2-2, was grown in a 60 1 fermenter for 3 days in a glucose-corn steep liquor medium and the filtered mycelium was suspended in three times its wet wt. of 0-03 M-sodium phosphate buffer pH 7-6 (P buffer) and disrupted by passage through a Manton-Gaulin homogenizer (The A.P.V. Co. Ltd.. Crawley. Sussex, U.K.) at 8000 lb/in2. Virus was partially purified from the homogenate by polythylene glycol precipitation and differential centrifugation using the method of Buck et al. (1981) and was further purified by sucrose density-gradient sedimentation. Partially purified virus (1500>4260 units) was loaded on to a 20 to 50% (w/w) linear sucrose gradient (500 ml in P buffer) in an MSE B XIV zonal rotor (MSE Scientific Instruments. Crawley. Sussex. U.K.) and centrifuged at 47000 rev/min for 3 h. After centrifugation the gradient was passed through an ISCO Model 222 u.v. analyser and 5 ml fractions were collected. The material in a u.v.-absorbing peak, about halfway down the gradient, was shown by electron microscopy to comprise isometric particles of diam. 34 to 36 nm. Nucleic acid, isolated from the purified virus particles by phenol/SDS extraction, was shown to be double-stranded (ds)RNA by its differential susceptibility to ribonuclease A in buffers containing high or low salt concentrations (Bellamy ei al.. 1967). Analysis and mol. wt. determination of viral dsRNA by polyacrylamide gel electrophoresis, using the methods described by Buck & Ratti (1977). revealed five components with mol. wt. 1-32 x 106, 1-29 x 106. 1-25 x 106. 1-22 x 106 and 1-03 x 106. Purified virus preparations, denatured by boiling with 1 % SDS and 1 % 2-mercapto- ethanol. were subjected to electrophoresis in polyacrvlamide-SDS gels, using the methods 002M 317/81/0000-4541 S02.00 £ 1981 SGM 002 Shori communications Fig. 1. Gel immunodiffusion tcsis with Phialophora viruses. AS. Antiserum to viruses A and B; VM. mixture of viruses A and B: VA. virus A: VB. virus B. The gel was stained with Coomassie Blue. and standards described by Buck & Kempson-Jones (1974); Coomassie Blue staining revealed two polypeptide components of mol. wt. 66000 and 60000. When undenatured virus preparations were subjected to electrophoresis in 0-5% agarose tube gels, as described by Ratti &. Buck (1972), two components with mobilities of about 1-2 x 10~5 and 3-8 x 10~5 cmVs/V towards the anode were detected by staining gels with either Coomassie Blue or with toluidine blue. The fact that virus particles took up both stains indicated their nucleoprotein nature. After electrophoresis and staining the agarose gel was sliced and the gel slices corresponding to each of the two nucleoprotein components were separately boiled with 1 % SDS and 1% 2-mercaptoethanol and the products were loaded on 8% polyacrylamide-SDS tube gels. Electrophoresis and staining showed that the nucleoprotein with the slower electrophoretic mobility in agarose gel electrophoresis contained the polypeptide of mol. wt. 60000, whereas the component with the faster mobility contained the mol. wt. 66000 polypeptide. These results suggested either that isolate 2-2 contained two distinct viruses with similar diameters, as has been found in several fungi (Hollings, 1978) or that the virus particles may have been partially degraded by a protease giving rise to two electrophoretic forms, as has been shown with cowpea mosaic virus (Geelen et aL 1973). To help to distinguish between these two possibilities a rabbit antiserum to the purified virus particles was prepared using the method of Hollings (1962). When purified virus preparations were allowed to diffuse against this antiserum in gel immunodiffusion tests, two clear precipitin lines were formed (Fig. 1), consistent with the presence of two distinct viruses. The two viruses were separated by chromatography on a column of an anion exchanger by an adaptation of the method of Buck & Kempson-Jones (1973). Purified virus preparation (4 ml. A 260 = 5) in P buffer was run on to a column of DEAE-cellulose (Whatman DE 52, 10 ml) (anion exchanger), previously equilibrated with P buffer. The column was eluted first with P buffer (10 ml) and then with a linear gradient of 0-6 M-NaCl (50 ml) in P buffer at a flow rate of 20 ml/h; 0-5 ml fractions were collected and analysed by agarose gel electrophoresis as above. The virus with the slower elecirophoretic mobility in agarose (virus A) was eluted in P buffer and was homogeneous. The virus with the faster electrophoretic mobility (virus B) was eluted with 0-4 M-NaCl. but contained a trace of virus A. For further purification virus B was dialysed against 10 mM-sodium acetate buffer pH 4-5 and adsorbed on to a column of SP-Sephadex (10 ml) (cation exchanger) previously equilibrated with the same buffer. The 3Shor icommunications (a) c oUl yrt rt uo 5. (B) < Migration Fig. 2. Electrophoresis patterns of dsRNA isolated from Phialophora viruses: (a) virus B: (A) virus A. column was eluted first with the acetate buffer (10 ml) and then with a linear gradient of 0 to 1 M-KCI in the same buffer (50 ml). From this column, virus B was eluted first with 0-07 M-KCI, whereas virus A eluted with 0-3 M-KCI. Virus A consisted of isometric particles of diam. 34 to 36 nm and s20 = 119S (in P buffer). It contained one major polypeptide of mol. wt. 60000 and three dsRNA components of mol. wt. 1-29 x 106, 1-22 x 106 and 1-03 x 106. Virus B consisted of isometric particles also of diam. 34 to 36 nm and s22 = 122S (in P buffer). It contained one major polypeptide of mol. wt. 66000 and three dsRNA components of mol. wt. 1-32 x 106, 1-25 x 10* and 1-03 x 106. In gel immunodiffusion tests using the antiserum described above, each virus gave rise to a single precipitin line. When the two viruses were placed in adjacent wells, crossing precipitin lines were obtained (Fig. 1), indicating that the two viruses were serologically unrelated. When the antiserum was titrated in twofold serial dilutions against each virus, the antiserum titre was found to be 1:2048 against virus A and 1:512 against virus B. The proportions of the dsRNA components in each virus, as measured from the area of the peaks in toluidine blue-stained polyacrylamide gels after electrophoresis (Fig. 2). were not equal, suggesting that the three dsRNA components were not enclosed in a single particle, but were probably encapsidated separately as has been found for several other dsRNA mycoviruses (Hollings, 1978). Further evidence to support this view was obtained by polyacrylamide gel electrophoretic analysis of dsRNA prepared from virus fractions obtained from the sucrose density gradients used for virus purification. Some fractions on the slower sedimenting side of the peak contained only the 1-03 x 106 mol. wt. RNA. whereas fractions on the faster sedimenting side were enriched in the higher mol. wt. RNAs. Both viruses contained a dsRNA of mol. wt. 1-03 x 106. The small amount of RNA of this mol. wt. found in virus B preparations was not due to contamination with virus A. Traces of virus A which remained in virus B after separation on the anion exchanger DEAE-cellulose 004 Shori communications were removed by separation on the cation exchanger SP-Sephadex from which virus B was eluted first. After this second column stage, no virus A could be detected in virus B preparations by immunodiffusion analysis, by agarose gel electrophoresis of intact virions or by polvacrvlamide gel electrophoresis of virus polypeptides, and neither of the other two dsRNAs of virus A could be detected in virus B. The occurrence of a dsRNA of the same mol. wt. in the two viruses could be due to genomic masking (encapsidation of the RNA of one virus by the capsid of another) or. to the necessity of having this gene for say replication in Phialophora sp. However, identity of mol. wt. does not necessarily indicate identity of nucleotide sequence and the two RNAs could be unrelated. The dsRNA mycoviruses are transmitted only intracellularly, during fungal growth, in spores or via hyphal anastomosis (cell fusion). They, therefore, may be expected to have a limited natural host range and the occurrence of serologically related viruses has, in general, been confined to related fungi (Hollings, J978). Recently, 13 isometric viruses from G. graminis. a fungus related to Phialophora sp. with lobed hyphopodia (Walker, 1981), have been characterized and classified into three groups on the basis of their physicochemical and serological properties (Buck el al.. 1981). On the basis of its physicochemical properties virus A from Phialophora sp. with lobed hyphopodia isolate 2-2 could be classified with the G. graminis group I viruses. Virus B. on the other hand, has a sedimentation coefficient and dsRNA mol. wt. similar to the group I viruses, but its polypeptide mol. wt. is closer to that of the group II viruses. In gel immunodiffusion tests none of the 13 G. graminis viruses, 11 of which were in groups I and II, reacted with the antiserum to the Phialophora viruses A and B described here, nor did viruses A and B react with any of eight antisera containing antibodies to a total of 11 of the G graminis viruses. However, the absence of serological cross-reactions does not preclude other relationships between these viruses. The antigenic determinants of a virus are encoded by a comparatively small portion of its genome and a fairly small number of mutations may lead to marked changes in its serological properties. Studies of nucleotide sequence homology of the viral dsRNAs, preferably carried out by direct comparison of sequences, may provide a greater insight into possible relationships between the viruses of these two fungal species. We thank Mr R. D. Woods for electron microscopy and the Science Research Council for the award of a C.A.S.E. studentship (to R.M.M.). Department of Biochemistry K. W. BUCK* Imperial College of Science and Technology R. M. MCGINTY+ London SW7 2AZ, U.K. 1 Plant Pathology Department C. J. RAWLINSON1 Rothamsted Experimental Station Harpenden, Herts AL5 2JQ, UJC. + Present address: Department of Medical Oncology. Southampton General Hospital. Southampton. U.K. REFERENCES BELLAMY, A. r.. SHAPIRO, u. AUGUST. J. T. & JOKUK. w. K. (1967). Studies on reovirus RNA. 1. Characterisation of reovirus genome RNA. Journal of Molecular Biolog\• 29. 1-17. BUCK. K. w. (1980). Viruses and killer factors of fungi. In The Eukaryotic Microbial Cell. Society for Genera! Microbiology Symposium, vol. 30. pp. 329-3~5. Edited by G. W. Gooday. D. Lloyd and A. P. J. Trinci. Cambridge: Cambridge University Press. BUCK. K. W. & KEMPSON-JONES. G. F. (1973). Biophysical properties of Penicillium stoloniferum virus S. Journal of General I'irologv 18. 223-235. 5 Shori communications BLCK. K. \V. A KEMPSONJONES. G. F. (1974). Capsid polypeptides of two viruses isolated from Penicillium stoloniferum. Journal of General Virology 22. 441 —445. BICK. K. W. & RATTI. G. (1971). Molecular weight of double-stranded RNA: a re-examination of Aspergillus foetidus virus S RNA components .Journal of General Virolog\' 37, 215-219. BUCK. K. W., ALMOND, M. R., MiFADDEN, J. J. P., ROMANOS, M. A. & RAWLINSON, c. J. (1981). Properties of thirteen viruses and virus variants obtained from eight isolates of the wheat take-all fungus. Gaeumannomyces graminis var. tritici. Journal of General Virology 53. 235-245. GEELEN. J. L. M. C.. REZELMAN, G. A VAN KAMMEN, A. (1973). The infectivity of the two electrophoretic forms of cowpca mosaic virus. Virology 51, 279-286. HOLLINGS. M. (1962). Viruses associated with a die-back disease of cultivated mushroom. Sature, London 196. 962-965. HOLLINGS. M. (1978). Mycoviruses - viruses that infect fungi. Advances in Virus Research 22. 3-53. HOLLFNGS. M. (1979). Taxonomy of Fungal Viruses. In Fungal Viruses, pp. 165-175. Edited by H. P. Molitoris. M. Hollings and H. A. Wood. Berlin. Heidelberg and New York: Springer-Verlag. RATTI. G. A BUCK, K. W. (1972). Virus particles in Aspergillus foetidus: a multicomponent system. Journal of General Virology 14. 165-175. WALKER, J. (1981). Taxonomy of take-all fungi and related genera and species. In Biology and Control of Take-all. Edited by M. J. C. Asher and P. J. Shipton. New York and London: Academic Press (in press). (Received 12 December 1980) Phytopathologie Zeitschrift R.M.McGinty,K.W.Buck,C.J.Rawlinson (Accepted 2 February, 1981) VIRUS PARTICLES AND DOUBLE-STRANDED RNA IN ISOLATES OF PHIALOPHORA SP. WITH LOBED HYPHOPODIA, PHIALOPHORA GRAMINICOLA AND GAEUMANNOMYCES GRAMINIS VAR. GRAMINIS ·BY;- R.M. McGINTY *t , K.W. BUCK* and C.J. RAWLINSON t * Department of Biochemistry, Imperial College of Science and Technology, London, SW7 2AZ. t Plant Pathology Department, Rothamsted Experimental Station, Harpenden, Hert AL5 2JQ. Isometric virus particles containing double-stranded RNA (dsRNA) are of common occurrence in fungi (Hollings, 1978; Buck 1980). These viruses are unusual in that they do not ly$e their hosts and are transmitted only intracellularly, during growth, in sexual and asexual spores and via heterokaryosis. Although many fungi are able to tolerate these particles without adverse effect and few virus diseases of fungi have been reported (Hollings, 1962;· Ghabrial, Sanderlin & Calvert, 1979; Castanho, Butler and Shepherd, 1978), it is becoming increasingly clear that dsRNA can affect a fungus phenotype in other ways. For example, protein toxins secreted by killer strains of Saccharomyces_ cerevisiae Hansen and of Ustilago maydis (DC) Corda are encoded by specific dsRNA molecules (Bostian et al. 1980;· Koltin and Kandel, 1978), and in Endothia parasitica (Murr.) And. hypovirulence, t Present address: Department of Medical Oncology, Southampton General Hospita Southampton. • 2 • "tJithout loss of saprophytic ability, is caused by a cytoplasmic factor, .probably dsRNA (Day et al. 1977). Virus particles have been found in many isolates of fungi in the Gaeumannomyces/Phialophora complex, including the wheat take-all fungus, Gaeumannomyces graminis (Sacc.) Arx & Olivier var. tritici Walker (GGT) (Lapierre et al. 1970; Rawlinson et al. 1973; Rawlinson & Muthyalu, 1976). Most fungi in this complex, including hypovirulent GGT, are able to restrict th growth of virulent GGT on cereal roots (Deacon, 1974; Wong, 1975; Sivasithamparam, 1975), but the role of virus particles in such cross-protectio is uncertain. Lapierre et al. (1970) aud Lemaire et al. (1970) correlated hypovirulence and ability to cross-protect with virus particles in GGT, but Rawlinson et al. (1973) in a study of more than 150 isolates of GGT found no consistent association of virus particles with weak pathogenicity. In a more recent study Frick & Lister (1978) found serotype variability in virus particle from twenty isolates of G. graminis and considered that this might reflect similar virUs biotype variability and hence offer an explanation for the apparent inconsistencies of their putative association with hypovirulence. Almond, Buck & Rawlinson (1977, 1978) and Almond (1979) found not only serotype variation, but also considerable variation in the number and molecular weights of dsRNA components obtained from GGT isolates collected from six different countries. In the present study we have extracted virus particles and dsRNA from isolates of Phialophora sp. with lobed hyphopodia, Phialophora graminicola (Deacon) Walker and Gaeumannomyces graminis (Sacc.) Arx & Olivier var. graminis (Table 1) and have found that dsRNA variability occurs even in isolates taken from the same field. • 3 . Isolates were first grown on 3.9% potato dextrose agar at 25 0 C in darkness for one week. A 6mm disk of PDA culture was added to 100 ml of 2% malt extract or a basal corn steep liquor medium (Banks, Buck & Fleming, 1971) and shake cultured for 7 to 14 days. Virus particles were then prepared from the fungal mycelium either as described by Rawlinson et al. (1973) or as follows. Filtered mycelium was disrupted by passage three times through a Pascali triple roll mill and the homogenate was suspended in ten times its weight of 0.03 ~- ~odium phosphate buffer, pH 7.6. After removal of debris by centrifugation (23 OOOg, 30 min), sodium chloride and polyethylene glycol 5000 were added to supernatant to give final concentrations of 1~ and 10% respectively. After storing for 24 to 48h the precipitate was collected by centrifugation and ~esuspended in one tenth of the original volume of phosphate buffer. After low speed centrifugation to remove insoluble debris, virus particles were sedimented from the supernatant fluid by centrifugation for 2h at 100 OOOg. The virus pellets were resuspended in one fiftieth of the original volume of phosphate buffer. Final preparations were centrifuged at low speed to remove residual insoluble material. All operations were carried out at 0 to 4°C. Virus preparations were examined in a Siemens Elmiskop la electron - microscope after mixing with neutral 2% sodium phosphotungstate. All preparations made from isolates of Phialophora sp. with lobed hyphopodia and Gaeumannomyces graminis var. graminis were found to contain isometric particles, 35nm in diameter (Table 2). Preparations from isolates 2-2 and 12-2 of Phialophora sp. with lobed hyphopodia contained in addition particles of 40nm diameter. Virus particles obtained from Phialophora graminicola measured 30nm in diameter, and showed a marked tendency to form . 4 . regular aggregates. Representative electron micrographs are given in Figs 1 to 4. Nucleic acids were prepared from disrupted mycelium and from isolated .virus preparations by phenol extraction and analysed by polyacrylamide gel electrophoresis, essentially as described by Moffitt & Lister (1975). Nucleic acid components were shown to be dsRNA and their molecular weights were determined by standard methods (Bellamy et al. 1967; Berry & Bevan, 1972; Buck Ratti, 1977). Numbers and molecular weights of dsRNA components (Table 2) were the same whether dsRNA was obtained directly from fungal mycelium or from isolated virus preparations. It is clear that there is considerable variation in the dsRNA components obtained from different fungal isolates. Out of the eight isolates studied only two (55-4 and 17-3) had the same pattern of dsRNA components. Moreover five different dsRNA patterns were obtained from isolates of Phialophora sp. with lobed hyphopodia collected from a single field at Rothamsted. The observed variation could be due both to different viruses and 6 to virus variants. It is noteworthy that dsRNA components of mol. wt.>6 X 10 were found only in the two isolates (2-2 and 12-2) which contained the particle of 40 nm diameter. The results reported here have two ma1n implications. Firstly becaus of their intracellular mode of transmission dsRNA mycoviruses become a fairly permanent cytoplasmic feature of particular fungal strains, rather like DNA plasmids in bacteria. Hence the five different dsRNA patterns obtained from si isolates of Phialophora sp. with lobed hyphopodia indicate a rather heterogeneous population of this fungus colonising barley roots in a single field. Secondly it is clear that dsRNA variability and hence possible biotype variability must be taken into account in any studies of the effect of virus infection on the biology of these fungi. • 5 . We thank Mr. R.D. Woods for electron microscopy, D.B. Slope and P.T.W. Wong for fungal isolates and the Science Research Council for the award of a studentship (to R.M. McGinty). . I 6 REFERENCES ALMOND, M.R. (1979). Biochemical characterisation of viruses isolated from the 1 phytopathogenic fungus Gaeumannomyces graminis var. tritici. Ph.D Thesis,University of London. ALMOND, M.R., BUCK, K.W. & RAWLINSON, C.J. (1977). Viruses of Gaeumannomyces graminis var. tritici. Second International Mycological Congress, Tampa, Florida. 1977. Abstract, p.l3. ALMOND, M.R., BUCK, K.W. & RAWLINSON, C.J. (1978). The virus complex of Gaeumannomyces graminis var. tritici. Bulletin of the British Mycological Society~' 115. BANKS, G.T., BUCK, K.W. & FLEMING, A. (1971). Isolation of viruses and viral ribonucleic acid from filamentous fungi on a pilot plant scale. The Chemical Engineer 251, 259-261. BELLAMY, A.R., SHAPIRO, L., AUGUST, J.T. & JOKLIK, W.K. (1967).Studies on reovirus RNA. I. Characterisation of reovirus genome RNA. Journal of Molecular Biology~' 1-17. BERRY, E.A. & BEVAN, E.A. (1972). A new species of double-stranded RNA from yeast. Nature, London 239, 279-28a BOSTIAN, K.A., HOPPER, J.E., ROGERS, D.T. & TIPPER, D.J. (1980). Translational analysis of the killer - associated virus-like particle dsRNA-genome-' of Saccharomyces cerevisiae: M ds RNA encodes toxin. Cell ~' 403-414 BUCK, K.W. (1980). Viruses and killer factors of fungi. In The Eukaryotic Microbial Cell. Society for General Microbiology Symposium 30, pp. 329-375. Edited by G.W. Gooday, D. Lloyd & A.P.J. Trinci. Cambridge: Cambridge University Press. BUCK, K.W. & RATTI, G. (1977). Molecular weight of double-stranded RNA: a re-examination of Aspergillus foetidus virus S RNA components. 7 Journal of General Virology 37, 215-219. CASTANHO, B., BUTLER, E.E. & SHEPHERD, R.J. (1978). The association of double-stranded RNA with Rhizoctonia decline. Phytopathology 68, 1515-1518. DAY, P.R.-, DODDS, J.A., ELLISTON, J.E., JAYNES, R.A. & ANAGNOSTAKIS, S.L. (197~ Double-stranded RNA in Endothia parasitica. Phytopathology 67, 1393• 1396. DEACON, J.W. (1974). Further studies on Phialophora radicicola and Gaeumannomyces graminis on roots and stem bases of grasses and cerea~ Transactions of the British Mycological Society~, 307-327. FRICK, L.J. & LISTER, R.M. (1978). Serotype variability in virus-like particle from Gaeumannomyces graminis. Virology 85, 504-517. GHABRIAL, S.A., SANDERLIN, R.S. & CALVERT, L.A. (1979). Morphology and virus- like particle content of Helminthosporium victoriae colonies regenerated from protoplasts of normal and diseased isolates. Phytopathology 69, 312~315. ·HoLLINGS, M. (1962). Viruses associated with a die-back disease of cultivated mushroom.- Nature, .:London, 196., 962-965. -!HOLLINGS, M. (1978). Mycoviruses - viruses that infect fungi. Advances in Virus Research 22, 3-53 • .KOLTIN, ·~. & KANDEL, J.S. (1978). Killer phenomenon in Ustilago maydis : the organisation of the viral genome. Genetics 88, 267-276. lAPIERRE, H., LEMAIRE, J:M., JOUAN, B & MOLIN, G. (1970). Mise en ~vidence de I particules virales associees ~ une perte de pathogenicite chez le pietin-echaudage des cereals, Ophiobolus graminis Sacc. Compte rendu hebdomadaire des Seances de 1' Academie des Sciences,~' Paris 271, 1833-1836. 8 , LEMAIRE, J~M., LAPIERRE, H., JOUAN, B & BERTRAND, G (1970). Decouverte de particules virales chez certaines souche d' Ophiobolus graminis agent du pi~tin-~chaudage des c~r~ales. Cons~quences agronomique pr~visibles. Compte rendu hebdomadaire des S~ances de 1' Acad~ie d' Agriculture de France 56, 1134-1137. MOFFITT, E.M. & LISTER, R.M. (1975). Application of a serological test for detecting double-stranded RNA mycoviruses. Phytopathology ~' 851-85 RAWLINSON, C.J., HORNBY, D., PEARSON, V. & CARPENTER, J.M. (1973). Virus-like particles in the take-all fungus, Gaeumannomyces graminis. Annals of Applied Biology 74, 197-209. RAWLINSON, C.J. & MUTHYALU, G. (1976). Report of the Rothamsted Experimental Station for 1975, p.255. SIVASITHAMPARAM, K. (1975). Phialophora and Phialophora- like fungi occurring in the root region of wheat. Australian Journal of Botany ~' 193-21 WALKER, J. (1981). Taxonomy of take-all fungi and related genera and species. In Biology and Control of Take-all, ed. M.J.C. Asher and P.J. Shipton New York & London : Academic Press, in press. WONG, P.T.W. (1975). Cross protection against the wheat and oat take-all fungi by Gaeumannomyces graminis var. graminis. Soil Biology and Biochemistr~ I, 189-194. Table 1 Source of Fungi Isolate Year of · Source isolation Phialophora sp. with lobed Continuous barley, Hoosfiel hyphopodl& 2-2, 12-2, 17-3, 24-4, 55-4 1973 74/1007-2 1974 Rothamsted, U.K.(B.D.Slope Gaeumannomyces graminis (Sacc.) Arx & Olivier var.graminis Gl 1973 Prairie grass, Rydalmere, Australia (P.T.W.Wong) Phialophora graminicola (Deacon) wa·l ker First wheat after ley, 1348-2 1974 Summerdells, Rothamsted, U.K. (D. B. S1 ope) . Table 2. Molecular weights and yields of dsRNA components and diameters of virus particles obtained from isolates of Gaeumannomyces and Phialophora spp. Fungal Diameter Mol.wt. of Yield of dsRNA Isolate of virus dsRNA components (pg/g particles compogents wet weight of (nm) mycelium) Gaeumannomyces graminis var. gramims £1 35 1.65 1.54 1.31 1.19 Phialophora sp, (lobed hyphopodia) 2-2 35,- 40 6.1 4 1.31 4 1.28 4 1.23 4 1.04 4 12-2 35, 40 6.1 0.2 1.57 0.8 17-3 35 1.19 0.1 1.14 0,1 24-4 35 1.69 0.8 1.62 0.3 55-4 35 1.19 0.1 1.14 0.1 74/1007-2 35 1.35 4 1.28 4 1.23 4 1.04 4 Phialophora graminicola 1348-2 30 1.32 1.23 1.17 1.14 11 Legends to figures Figs 1 to 4. Electron micrographs of virus particles negatively stained with phosphotungstate. (1) 35nm particles from Phialophora sp. with lobed hyphopodia isolate 2-2; (2) a group of 40nm particles from Phialophora sp. with lobed hyphopodia isolate 12-2; (3) 35mn particles from Gaeumannomyces graminis var, graminis isolate Gl; (4) aggregated 30nm particles from Phialophora graminicola isolate 1348-2. The bar represents lOOnm. 12 0 i. ,i . . i . P.sp.(lh)2-2 pale grey, floccose colonies with entire PDA margins