The Ribonucleic Acid of Tomato Spotted Wilt Virus - CiteSeerX

J. gen. Virol. 0977), 36, 81-91

81

Printed in Great Britain

The Ribonucleic Acid of Tomato Spotted Wilt Virus By J. V A N D E N H U R K ,

P. W. L. T A S AND D. P E T E R S

Laboratory of Virology, Agricultural University, Binnenhaven I I, Wageningen, The Netherlands (Accepted IX February I977) SUMMARY

TSWV nucleic acid, extracted from purified virus with phenol-SDS, was noninfectious. Following electrophoresis of the extracted nucleic acid in 2 % polyacrylamide gels containing o'5% agarose, 3 major and 2 minor bands were observed. The amount of material contained in the different bands varied with the season. Bands I, 2, 3 and 4 were sensitive to RNases and resistant to DNase. This was a strong indication of the presence of single-stranded R N A in these bands. Precipitation in 2 M-LiC1, effect of heating and mobility in gels containing an increasing percentage of acrylamide confirmed the single-stranded character. Band Ia was resistant to the action of RNases and stained with Coomassie brilliant blue. The mol. wt. of RNAs I, 2, 3 and 4 (corresponding to bands I to 4) were estimated under both non-denaturing and denaturing conditions and appeared to be 2. 5 to 2"7 × 1o 6, 1.9 × Io ~, 1"7 x io 6 and I"3 × lO6, respectively. Following dissociation of the virus with Nonidet P 4o, an infectious zone could be isolated after sedimentation in an 8 to 43 % glycerol gradient. INTRODUCTION

Very little information exists concerning the nucleic acid of tomato spotted wilt virus (TSWV). Van Kammen, Henstra & Ie 0966) concluded that the nucleic acid was RNA on the basis of positive orcinol and negative diphenylamine reactions. Best (I968) arrived at the same conclusion from data obtained by paper chromatography and estimation of the base composition of the nucleic acid (guanine 38 %, adenine 35 ~o, cytosine 9 % and uracil 18 %). In the cryptogram of TSWV (R/*:*/*:S/S:S/Th; Ie, I97o) data are missing on the strandedness and molecular weight of the nucleic acid, and its percentage content in the virus particles. This paper describes a study of the nucleic acid of TSWV, with respect to its type, segmentation, molecular weight and infectivity, in addition, the isolation of an infectious ribonucleoprotein complex obtained after dissociation of the virus with Nonidet P 4o is described. METHODS

Virus purification. Plant material infected with TSWV-S was ground in o.oI M-tris (pH 8.o) containing o.oI M-sodium sulphite and o.1% cysteine hydrochloride. The extract was centrifuged for Io min at 1oo0o g. The pellet was resuspended in buffer consisting of o.oI M-tris, O'0I M-glycine, O'Ot M-sodium sulphite and o . I /o/ o cysteine hydrochloride (pH 7"9). After clarification for Io min at IOOOOg, the suspension was spun for 3o rain at 25ooo rev/ min. The pellet was resuspended and incubated with antiserum against healthy plant 6-2

82

J. VAN DEN HURK, P. W. L. TAS AND D. PETERS

material for I h. After removal of the precipitate the virus was subjected to a cycle of rate and equilibrium centrifugation in sucrose gradients (Tas, Boerjan & Peters, I977a). Purification of RNA. Sterile solutions and glassware were used throughout all operations. Pellets of purified virus were dissolved in o.oi M-tris-HCl, o.I M-NaC1, o.ooi M-Na2 EDTA, pH 7"6, containing I or 2 0/0 SDS. This mixture was stirred vigorously for IO rain at room temperature, cooled to 4°C, and an equal volume of phenol, saturated with buffer containing 8-hydroxyquinoline at a final concentration of o.I %, was added. After stirring for IO rain at 4°C the aqueous phase was separated from the phenol phase by centrifugation at 3ooo rev/min for 15 min at 4 °C. The phenol phase was re-extracted with one third volume of buffer. The aqueous phases were pooled and re-extracted three times with one third volume of phenol saturated with buffer. The nucleic acid was precipitated from the final aqueous phase using 2. 5 volumes of ethanol and a few drops of 3 M-sodium acetate, p H 5"5. Precipitation was completed by storage overnight at - 2 o °C. The nucleic acid precipitate was pelleted by centrifugation in a table centrifuge at 3ooo rev/min for 15 min and was washed three times with 7o~o ethanol containing o-I M-sodium acetate. The nucleic acid was stored at - 2 o °C either in ethanol or in o'I5 M-NaC1 and o'oi5 M-Na-citrate, p H 7"2 (I × SSC). Electrophoresis of RNA. Polyacrylamide gel electrophoresis under non-denaturing conditions was~performed by a modification of the method of Loening & Ingle (I967); electrophoresis under denaturing conditions was by the method of Reijnders et aL (I973). Acrylamide and bisacrylamide were re-crystallized according to Shepherd & Gurley (i966). Gels containing 2 % acrylamide and 0"5 0/0 agarose were polymerized in glass tubes (65 × 6 ram) at 37 °C for 3o rain. Gels were left at room temperature overnight and usually pre-swollen in electrophoresis buffer for I z to 16 h before use. Upper and lower parts of the gel were removed with a sharp razor blade. The length of the gel was adjusted to 5"5 cm. For the analysis of nucleic acid under non-denaturing conditions the gels were pre-run at 4mA/gel in o-o36M-tris-H3PQ, 0"03 M-NaH2PO~.2H~O, o.ooi M-EDTA, p H 7"7, containing o:2 % SDS for I h at 4 °C. Following this run the electrophoresis buffer was renewed. The nucleic acid, dissolved in electrophoresis buffer supplemented with o-2% SDS and 5 % glycerol, was layered on top of the gels through the electrophoresis buffer. Electrophoresis of R N A samples was performed at 4 °C for 3 to 4 h at 4 mA/gel. After electrophoresis the gels were soaked in water for 2 to 3 h and scanned at 260 nm in a Beckman D U spectrophotometer equipped with a Gilford 24Io Linear Transport, an ERA attachment and a linear/log recorder. Gels were also stained with o.oi ~o toluidine blue O in 4o Uo methanol, and destained in water or 1% acetic acid and stored in the dark. Incubation with RNase and DNase. After completion of electrophoresis of the RNA, the gels were washed in 2 x SSC for 16 h and incubated with 25 units/ml RNase T1 grade III (Sigma), from Aspergillus oryzae, and 25 #g/ml RNase A type t-A, 5 x crystallized, from bovine pancreas, (Sigma), at room temperature in 2 × SSC (15 ml/gel). The distribution of the extinction at 26o nm over the gel was measured before and after incubation. To determine the sensitivity of the nucleic acid to DNase the gels were washed in 2 × SSC containing 5 mi-MgC12 for I6 h and incubated with 5 #g/ml DNase I in the same medium (I5 ml/gel). Precipitation in 2 M-LiC1. Nucleic acid extracted from TSWV was mixed with an equal volume of 4 M-LiC1. This suspension was stored overnight at - 2 o °C. The resulting precipitate was removed by centrifugation at low speed, and R N A in the supernatant was concentrated by precipitation with 2 vol. cold ethanol. The low speed pellet and ethanol precipitate were analysed by electrophoresis in polyacrylamide gels.

R N A o f tomato spotted wilt virus

83

Denaturation of the nucleic acid by heating. Nucleic acid from TSWV, suspended in o.oi M-EDTA, pH 7"o, was heated at IOO °C for 3 to IO min and then cooled rapidly in ice. The treated extracts were analysed by electrophoresis in polyacrylamide gels. Estimation of molecular weights of the different RNA segments in polyaerylamide gels. The single-stranded RNA of cowpea mosaic virus middle component (CPMV RNA~) and bottom component (CPMV RNA~) were used as standards. The tool. wt. values were taken as I"37 x I06 and 2.02 × IO 6, respectively (Reijnders et al. 1974). Double stranded R N A molecules consisting of the respective replicative forms of the RNAs of CPMV were kindly supplied by A. M. Assink (Assink, Swaans & Van Kammen, 1973). Isolation of infectious ribonucleoprotein. Purified virus was treated with 0"o5 % Nonidet P 40 (Shell) for 2 rain, layered on an 8 to 43 % glycerol gradient in I x SSC, and centrifuged for I h at 4oooo rev/min in an SW 6o Ti rotor of a MSE 65 centrifuge. The gradient was fractionated and the extinction of each fraction measured at 257 nm. The infectivity of the fractions was determined by inoculation on detached leaves of Petunia hybrida. RESULTS

Infectivity The nucleic acid isolated by the phenol-SDS method was non-infectious. Purified virus disrupted with o-I % SDS and inoculated on to Petunia hybrida was also non-infectious. Several changes in the extraction conditions (buffer composition, presence or absence of bentonite and RNase inhibitors) failed to yield infective preparations.

Polyacrylamide gel electrophoresis The nucleic acid was analysed by electrophoresis on 2 ~o polyacrylamide gels containing 0"5 % agarose (Fig. I). In the electrophoretic diagrams obtained by scanning of the gels at 260 rim, 4 to 5 peaks were detected; they corresponded with the bands found by staining. The relative amount of material contained in the different bands varied with the season. The reason for this inconsistancy is not understood. All further experiments to be described were performed with nucleic acid extracted from virus preparations purified in the period June-July. The virus could be purified from infected leaves in high yield and with a high degree of purity (Tas et al. I977a). The extinction profiles obtained at 260 nm a/ways revealed bands I, 3 and 4 clearly, while band 2 was seen only when a high resolution was obtained. In most cases component u appeared as a shoulder of component 3 in the electrophoretic patterns, although bands 3 and u were always clearly resolved upon visual observation of stained gels. Band Ia was not always visible; this may be attributable to the relatively small amount of this component. The observed band patterns cannot be due to the effects of storage of purified virus at - 2 o °C prior to extraction. The same species were found when nucleic acid was extracted from virus used directly after purification.

Incubation with RNase and DNase The nature of the nucleic acids was investigated by incubating the gels after electrophoresis with RNase and DNase. Fig. 2 clearly shows that the bands which form peaks I, a, 3 and 4 in the profile and the single stranded CPMV RNAs are equally sensitive to RNases. Double-stranded CPMV R N A was resistant under the incubation conditions employed. Incubation without RNase did not change the observed band pattern (Fig. 2).

84

J. VAN DEN HURK, P. W. L. TAS AND D. PETERS I

:

(a)

I

I

B M

]

(b)

I

I

B M

I

J

(c)

I

I

B

M

I

I

I

-~

3

4

1

0

I

i

I

I

1

2

3

4

0 1 2 3 4 5 0 1 2 3 4 Distance migrated (cm)~

Fig. I. Electrophoretic patterns of RNA extracted from TSWV purified in (a) February, (b) April and (c) June-July on 2 ~o polyacrylamide gels containing o-5 ~ agarose. The peaks in the extinction patterns are numbered I to 4 from high to low mol. wt. respectively; CPMV RNA~ (= M) and CPMV RNAB (= B) were co-electrophoresed. Electrophoresis was from left to right. It is interesting to note that band Ia appeared to be rather resistant to the action of the enzymes. Following incubation of the gels with RNase and staining with toluidine blue O, band Ia was still visible, whereas bands I, 2, 3 and 4 bad disappeared. Fig. 3 shows the band pattern observed after incubation of the gels with DNase. A test for the presence of RNase activity in the DNase preparation using singIe stranded CPMV R N A proved negative. From the almost complete disappearance of the calf thymus D N A peak following incubation with DNase for 12 h, it was concluded that the DNase preparation was active. Under these conditions no breakdown of TSWV nucleic acid was observed. From the results obtained by the incubation of the gels with RNases and DNase it was concluded that bands I, 2, 3 and 4 consisted of R N A of a single-stranded nature. The material in band I a appeared not to be sensitive to either RNase or DNase.

Differential precipitation by salt Information about the nature of a nucleic acid can also be deduced from its behaviour in solutions of increasing salt concentration (Erikson & Franklin, t966 ). Whereas singlestranded R N A is precipitated by 2 M-LiCI, double-stranded RNA, double-stranded D N A and transfer R N A are not precipitated under these conditions (Segal & Sreevalsan, I974). The procedure of Segal & Sreevalsan was applied to the nucleic acid o f TSWV (Fig. 4)The material forming bands I, 3 and 4 on the gels was almost completely precipitated by 2 M-LiC1 suggesting that they are single-stranded. Band 2 was also found but was not well

R N A o f tomato spotted wilt virus I

1

(a)

(b)'

I

3

I

85

t

I

I

(c)

t

1

I

(d) B

M M

n

la

0

2

1

2

ii

I

!

( I

I

I

I

3

4

1

2

I

I

J I

]../

I

0 1 2 Distance migrated (cm)-+ 3

4

• .

I

|

I

I

I

I

3

4

1

2

3

4

Fig. z. Electrophoretic pattern of (a) TSWV RNA control incubated in 2 x SSC alone for o h ( ), 5 h ( - - - - ) or iz h ( ...... ), and of (b) TSWV RNA, (c) single-stranded CPMV RNA and (d) double-stranded CPMV RNA incubated for o h ( ), 5 h ( - - - - ) and I2 h ( ...... ) with 25 #g RNase A/ml and 25 units RNase T1/ml in z x SSC. resolved in the electrophoretic pattern (Fig. 4) suggesting that the material present in this band exhibited the same behaviour as that forming bands I, 3 and 4. The concentrated supernatant contained material forming band ra and a relatively small amount of material forming band 3- The occurrence of the latter material in the supernatant suggested that the single stranded R N A was not completely precipitated by 2 M-LiCI. The material forming band ~a behaved differently from that forming bands I, 2, 3 and 4.

Effects of heating at ~oo °C Information about the nature of the nucleic acid was gained from an investigation of its sensitivity to denaturation at I oo °C. The components of double-stranded nucleic acid will be separated from each other, resulting in a change in the mobility. The positions of bands I, 2, 3 and 4 did not change after heating (Fig. 5), suggesting that these bands consisted of single-stranded RNA. However, band Ia disappeared following this treatment. The material contained in the supernatant and pellet fraction obtained after precipitation with 2 M-LiC1 was also heated at Ioo°C and analysed by electrophoresis in polyacrylamide gels. The material in the low speed pellet fraction appeared to be resistant to denaturation, since no change in the electrophoretic pattern occurred. Analysis of the supernatant fraction after heating revealed the presence of some peak 3 material while peak ~a could not be detected. This experiment furnished additional proof that band Ia consisted of material differing in properties to that contained in bands I to 4.

86

J. V A N D E N H U R K , I

I

i

P. W. L. T A S A N D D. P E T E R S I

l

(a)

I

I

I

2

3

4

(b)

~2

1

...,

...........\ -...

i

0

l

2

3

4

0

1

Distance migrated (cm)-+ Fig. 3. Electrophoretic patterns of (a) calf thymus D N A and (b) TSWV RNA, scanned before and after incubation of the gels with DNase. The gels were incubated in 2 x SSC containing Mg-ions and 5 #g DNase I/ml for o h ( ), 5 h ( - - - - ) and [z h ( ...... ) respectively.

Properties of band [a Determination of the extinction of the bands in the gels at 260 nm and 280 nm revealed that the 260/280 nm ratio of band Ia was lower (~'3 to ['4) than that of the bands [, 2, 3 and 4 (2.o to 2"2). This pointed to possible contamination of band Ia with protein. Staining of a gel with Coomassie brilliant blue did not reveal protein in bands I, 2, 3 and 4; however, band [a reacted positively suggesting the presence of protein.

Electrophoretie behaviour of the RNA in gels with different percentages of acrylamide Polyacrylamide gel electrophoresis in gels with differing amounts of acrylamide can be used to differentiate between single-stranded, double-stranded linear and double-stranded circular R N A (Harley, White & Rees, I973). The mobility of bands I, 2, 3 and 4 was measured by electrophoresis in gels containing different concentrations of acrylamide (Fig. 6). Single- and double-stranded CPMV R N A s were used as standards. The curves obtained for bands [, 2, 3 and 4 of TSWV nucleic acid were similar in shape to each other and to the curves of single-stranded CPMV R N A (RNA~ + RNA~), stressing again their single-stranded character.

R N A o f tomato spotted wilt virus [ (a) l

L

l

1

I

I

I

[

(b)

I

87 I

!

!

2

3

4

(c) 3

,2

0

1

2

3

4

5 ! 2 3 4 5 0 Distance migrated (cm)~

1

5

Fig. 4- Electrophoretic patterns of (a) TSWV RNA, (b) the pellet and (c) the supernatant fractions obtained after precipitation by 2 M-LiCI. The ethanol precipitate of the supernatant was resuspended in one tenth the original volume.

Molecular weight determination The mol. wt. of the R N A segments corresponding to bands r, 2, 3 and 4 were determined under both non-denaturing and denaturing conditions (Table r). It was assumed that the R N A o f TSWV was totally denatured when analysed in 8 M-urea and low salt concentrations at 6o°C (Reijnders et al. I973, I974). The values obtained by the two methods were in good agreement (Table I). The ratio of the number of R N A segments present in bands I, 3 and 4 (Fig. I) was determined by dividing the peak area by the tool. wt. of the corresponding RNA. The ratio of the number of molecules of R N A I, R N A 3 and R N A 4 was I : to: I for TSWV extracted in the months June-July, 2 : 8 : 3 for R N A extracted in February and 6: I : r2 for R N A extracted in April.

Isolation of an infectious ribonueleoprotein complex from purified T S W V Because the R N A extracted from TSWV appeared to be non-infectious, experiments were conducted to try to isolate an infectious ribonucleoprotein from purified virus. The non-ionic detergent Nonidet P 4o, at a final concentration of o'o5 %, was used to dissociate TSWV. TSWV preparations treated in this manner exhibited a very low infectivity. N o complete virus particles were seen when virus treated with Nonidet P 4o was observed by electron microscopy (Tas et al. I977b). To exclude the possibility that the residual infectivity was due to the presence of nondisrupted virus particles, two virus samples, one treated with Nonidet P 4o and the other

88

I. VAN J

DEN

HURK~ ]

I

P.

W.

L. T A S

AND

!

{

I

D.

PETERS

I

J

(b)

(a)

~S

1

4 4

I

l

I

I

1

2

3

4

I

I

1

2

N 3

4

5

Distance migrated (cm)~ Fig. 5. Electrophoretic patterns of (a) TSWV RNA and (b) TSWV RNA heated for 5 rain at Ioo °C. Table I. Molecular weights of the RNA segments I - 4 under both

non-denaturing and denaturing conditions Peak

Non-denaturing conditions (Loening & Ingle, I967)

Denaturing conditions (Reijnders et al. I973)

I 2

2" 7 X I O e + I O ~ o 1"9× IOn±5~oo

2"5XIO n I ' 9 × IO n

3 4

['7x EO6+5% I'3 × IO~---+5~o

~'7x

to 6

1'3 X I 0 ~

untreated, were layered on separate gradients and centrifuged. The extinction profiles and the distribution of infectivity of the gradients are shown in Fig. 7. The extinction profile exhibited by detergent-treated virus consisted of 3 peaks (I to 3). Peak r was due to Nonidet P 4 o remaining at the top of the gradient. The nature of the material contained in peak z was unclear. The infectivity was contained in fractions in the region of peak 3. These peaks were not observed in the extinction profile of untreated virus and the infectivity was restricted to the pellet. These observations suggested that the infectivity of virus treated with Nonidet P 40 could not be ascribed to complete virus particles, but to an infectious component sedimenting more slowly than the complete virus particles.

RNA of tomato spotted wilt virus

15

89

i

I X

7 t

>

10

,.o 0

O e~

o

5

2.0

2.2 2.4 acrylamide

2.6

Fig. 6. Electrophoretic mobility of TSWV RNA I (A--A), 2 ( m - - m ) and 4 (O--O), singlestranded CPMV RNA~ (Lk--A), single-stranded CPMV RNA~ (D--V1), double-stranded CPMV RNA~ (O--©) and double-stranded CPMV RNAB ( x - x ) in gels containing different percentages of acrylamide. DISCUSSION

The material in bands I, 2, 3 and 4 was sensitive to RNases and resistant to DNase. This result pointed to the presence of single-stranded R N A segments in these bands. Precipitation in 2 M-LiC1, resistance to denaturation at IOO °C and mobility in gels with increasing percentage of acrylamide confirmed the single-stranded character of these segments. Band Ia, which was not always visible in the gel, appeared resistant to RNases. The material forming this peak could not be precipitated with 2 M-LiC1 and appeared to be sensitive to heat. This indicated that the material in band Ia was not single-stranded R N A . The presence of protein in band [a was suggested by staining with Coomassie brilliant blue. There is no direct evidence for the presence of nucleic acid in band I a because toluidine blue O also stains protein. It is, however, not plausible that band Ia consists solely of protein because such a protein would have a very high mol. wt. An association of protein with nucleic acid can account better for the occurrence of band Ia in the nucleic acid extract and can also explain the behaviour in 2 M-LiC1 and the disappearance of band Ia from the profiles as a result of heating at Ioo°C. The fact that the R N A extracted from TSWV was non-infectious could possibly be

J. V A N D E N H U R K , P. W. L. TAS A N D D. P E T E R S

9o

I l l l l l l l l

I

f

(a)

I I I I

I

I

I

I

(b)

0-4

800

0.4

0-3

600

0"3

0-2

400

0.1

200

~2

I

1

o.24 A

2

0.1

/•

i

t

t

~

t

z

|

i

2 3 4 5 6 7 8 910 Fractions

I

I

|

123456

60 O

40 ~

t

I

t

i

i

t

80

-i

L.

i ,

I

~' ~i !i !i

w

!

i

20 ~,. i

i

7 8 9 101112 Fractions

Fig. 7. Extinction pattern (E2~7) and infectivity distribution of an 8 to 43 ~ glycerol gradient for (a) untreated virus and (b) virus treated with Nonidet P 40, Centrifugation was for I h at 4oooo rev/min in the SW 60 Ti rotor of a MSE 65 centrifuge. Sedimentation is from left to right. The gradient was collected in fractions and the infectivity of the fractions determined on P. hybrida. , E257; O - - . - - O , lesions/leaf.

ascribed to the production of specific fractures in the R N A during virus purification or extraction of the RNA. However, the observed R N A patterns in polyacrylamide gels did not suggest a simple fragmentation of a large R N A molecule. If one assumes that infectious TSWV particles contain the equivalent of one of each of the 4 R N A segments the genome would have a mol. wt. of 7"4 x io 6. RNA segment T with a mol. wt. of 2"5 x Io 6 can code, theoretically, for a maximum of about 26oooo daltons of protein (Paucha, Seehafer & Colter, I974) and since the sum of the tool. wt. of proteins I to 5 (Tas et al. I977b) is more than 26oooo it seems reasonable to assume that more than one R N A molecule is needed to code for the proteins of TSWV. The presence of a segmented genome predicts a high frequency of recombination, a prediction supported by the conclusions of Best & Gallus (I955). However, no recombinants have been characterized so far. In the properties of its nucleic acid, TSWV shows some resemblance to influenza virus, namely single-stranded RNA that occurs in segments and is non-infectious. Heterogeneity in the R N A pattern has also been observed for influenza virus (Etchison et al. I971 ; Palese & Schulman, 1976). It would be interesting to discover if other properties characteristic of influenza virus, such as genetic recombination, multiplicity reactivation and production of incomplete virus particles also occur with TSWV. On the basis of the results obtained the cryptogram of TSWV may be written as: R/I: Y,7"4/*:S/S:S/Th. The classification of TSWV in a monotypic group (Harrison et al. I97I) is in agreement with the data presented here concerning the R N A and with data presented elsewhere (Tas et al. I977b ) concerning the protein composition of TSWV. The authors thank Professor Dr Ir. J. P. H. van der Want for his stimulating support and discussions, and Dr A. Ziemiecki for his help in preparing the manuscript.

R N A of tomato spotted wilt virus

91

REFERENCES

ASSINK, A. M., SWAANS, H. & VAN KAMMEN, A. (I973). T h e localization o f virus specific double stranded R N A o f cowpea m o s a i c virus in subcellular fractions o f infected Vigna leaves. Virology 53, 384-391, BEST, R. J. (1968). T o m a t o spotted wilt virus. Advances in Virus Research x3, 65-I46. BEST, R. J. & GALLUS, H. P. C. (I955). F u r t h e r evidence for the transfer of character-determinants (recombi-

nants) between strains of t o m a t o spotted wilt virus. Enzymologia I7, 2o7-221. ERIKSON, R. L. & FRANKLIN, R. M. (1966). S y m p o s i u m on replication of viral nucleic acids I. F o r m a t i o n a n d properties of a replicative intermediate in the biosynthesis of viral ribonucleic acid. Bacteriological Reviews 3o, 267-278. ETCHISON, J., DOYLE, M., PENHOET, E. & HOLLAND, J. (1971). Synthesis a n d cleavage of influenza virus proteins. Journal of Virology 7, I55-I67. HARLEY, E. H., WHITE, J. S. & REES, K. R. (I973). T h e identification o f different structural classes o f nucleic acids by electrophoresis in polyacrylamide gels o f different concentration. Bioehimica et Biophysiea Acta 299, 253-263. HARRISON, B. D., FINCH, J. T., GIBBS, A. J., HOLLINGS, M., SHEPHERD, A. 1., VALENTA, V. & WETTER, C. (I971). Sixteen groups o f p l a n t viruses. Virology 4 5 , 356-363. IE, T. S. (1970). T o m a t o spotted wilt virus. CMllAAB Descriptions of Plant Viruses N o . 39. C o m m o n w e a l t h Mycological Institute, Kew. LOENING, U. E. & INGLE, J. (I967). Diversity o f R N A c o m p o n e n t s in green p l a n t tissues. Nature, London 215, 363-367. PALESE, P. & SCHULMAN,J. L. (I976). Differences in R N A patterns o f influenza A virus. Journalof Virology XT, 876-884. PAUCHA, E., SEEHAFER, J. & COLTER, J. S. (1974). Synthesis o f viral specific polypeptides in M e n g o virusinfected L cells: evidence for a s y m m e t r i c translation of the viral genome. Virology 6 i , 315-326. REI.[NDERS, L., SLOOF, e., SIVAL, J. & BORST, P. (1973)- Gel electrophoresis o f R N A u n d e r d e n a t u r i n g conditions. Biochimica et Biophysica Acta 324, 32o-333. REIJNDERS, L., AALBERS,A. M. J., VAN KAMMEN,A. & THURING, R. W. J. (1974). Molecular weights o f plant viral R N A s determined by gel electrophoresis u n d e r denaturing conditions. Virology 6o, 515-52I. SEGAL, S. & SREEVALSAN,T. (1974). Sindbis virus replication intermediates: purification a n d characterization. Virology 59, 428-442. SHEPHERD, G. R. & GURLEY, L.R. (1966). High-resolution disc electrophoresis of histones. I. A n i m p r o v e d m e t h o d . Analytical Biochemistry x4, 356-363. TAS, v. W. L., BOERJAN, M. L. & PETERS, D. (I 977 a). Purification a n d serological analysis o f t o m a t o spotted wilt virus. Netherlands Journal of Plant Pathology 83, 61-72. TAS, P. W. L., BOERJAN, M. L. & PETERS, D. (I977b). T h e structural proteins o f t o m a t o spotted wilt virus. Journal of General Virology 36 (in the press). VAN KAMMEN, A., HENSTRA, S. & IE, T. S. (I966). M o r p h o l o g y o f t o m a t o spotted wilt virus. Virology 30, 574577.

(Received 19 October I976)