Enzyme-assisted immune detection of plant virus proteins electroblotted onto nitrocellulose paper

Journal of Virological Methods, 5 (1982) 267-

267

278

Elsevier Biomedical Press

ENZYME-ASSISTEDIMMUNEDETECTIONOFPLANTVIRUSPROTEINS ELECTROBLOTTEDONTONITROCELLULOSEPAPER

E.P. RYBICKI and M. BARBARA VON WECHMAR Department

of Microbiology,

University of Cape Town, Private Bag, Rondebosch,

7700, South Africa

(Accepted 16 August 1982)

A technique for the detection of plant virus coat proteins in plant sap is described. The method entails the electroblotting of sodium dodecyl sulphate-polyacrylamide gel electrophoresis-fractionated plant extracts onto nitrocellulose paper, probing the paper with virus-specific rabbit antisera, and in-

direct detection of virus proteins with horseradish peroxidaseconjugated goat anti-rabbit globulins. The sensitivity and specificity of the technique were tested using brome mosaic and barley stripe mosaic viruses. As little as 1 ng per track of virus protein was detectable, either as pure virus or when mixed with plant sap. Distant serological relationships were detected amongst tobamoviruses, and amongst the bromoviruses, with single antisera. The uses of the technique in probing capsid configuration in a presumed aphid picornavirus, and in routine diagnostic practice, are described. enzyme immune virus detection

electroblotting

bromovirus

tobamovirus

picornavirus

INTRODUCTION

An increasingly

important

problem in plant virology is the detection

of viruses which either occur at very low concentrations

and identification

in plant tissues, or which are

extremely labile and/or hard to extract from plant tissues. Sensitive serological techniques such as immunosorbent

electron

microscopy

(ISEM) and enzyme-linked

immunosorbent

assays (ELBA) have helped enormously in recent years to alleviate the problem (van Regenmortel, 1981); however, these techniques may not always be the ideal solution. An attractive alternative technique for virus detection, which combines the fractionating power of sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and antigen specificity of a radioimmunoassay (RIA), has recently

with the sensitivity

been described (O’Donnell

et al., 1982). The technique

entails the electrophoretic

trans-

fer of SDS-PAGE-fractionated proteins onto activated paper, and subsequent specific indirect immune detection of viral coat proteins by rabbit antibodies and ‘251-labelled protein A. We have investigated the potential of a similar immuno-electroblotting technique, which involves the electrophoretic transfer of proteins from SDS-polyacrylamide gels to nitrocellulose paper and subsequent indirect detection of virus proteins by use of 0166-0934/82/0000-0000/$02.75

@ 1982 Elsevier Biomedical Press

268

rabbit

antisera

(Towbin

and goat

anti-rabbit

et al., 1979). The sensitivity

of capsid disruption

on the antigenicity

horseradish

(CAR-HRP)

of the technique,

of a putative

Our aim in this work was first to investigate tool for the less well-equipped

peroxidase

and specificity

aphid picornavirus

the value of the technique

plant virus laboratory;

conjugate

and the effect were studied. as an analytical

and second, to develop a tech-

nique for our own use in the study of viruses affecting small grains in South Africa. The applicability of our results to plant virology, and possible future applications of electroblotting techniques in general, are discussed. METHODS

Viruses: propagation and purification All viruses were obtained from the stock collection University

of the Microbiology

of Cape Town. Plants used for propagation

Department,

were grown in fully controlled

plant growth rooms with a day/night cycle of 16 h/8 h, at an average temperature of 24°C. Brome mosaic virus (BMV) and broad bean mottle virus (BBMV) were propagated and purified as described elsewhere (Rybicki and von We&mar, 1981). Barley stripe mosaic virus (BSMV) was propagated on barley (Hordeurn vulgare L.) and purified by the method described by Atabekov and Novikov (1971). Tobacco mosaic virus common strain (TMV-vulgare), TMV nitrous acid mutant Ni 109 and cucumber

green mottle mosaic virus (CGMMV) were propagated

cribed by van Regenmortel(l975)

and purified as des-

and von Wechmar and van Regenmortel(l970).

A small isometric virus infecting Rhopalosiphon padi L. and Diuraphis noxia aphids was purified from barley plants (Hordeurn vulgare L. cv. Clipper) which had been exposed 21 days previously

to infected

R. padi (Rybicki

and von Wechmar,

1982). A

similar virus has been described in the U.S.A. by D’Arcy et al. (1981 a,b), and tentatively named R. padi virus (RhPV). Various naturally

infected

plants were maintained

as described by Von Wechmar and

Rybicki (1981). Antiserum production Antisera to intact virions of BMV, BBMV, BSMV, TMV strains and RhPV, and other pure and semi-pure virus extracts, were raised in rabbits by 3 intramuscular injections at weekly intervals of 1 ml of emulsified 1: 1 mixtures of virus suspension and Freund’s incomplete adjuvant (Rybicki and von Wechmar, 1981), followed by a booster 6 wk later. Bleedings used for assays were those obtained 3 mth or longer after the initial immunisation. All antisera to purified viruses had Ouchterlony double-diffusion gel titres of l/256 or higher. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) This was performed by the method of Laemmli (1970), in a vertical slab gel apparatus

269 (SE-600, Hoefer Scientific

Instruments,

San Francisco).

Slab gels were 1.5 mm thick and

consisted of a 12.5% resolving gel 13 cm long, and a 4.5% stacking gel 3 cm long. Bisacrylamide concentration

was 2.6% of the total. Ten sample wells of width 8 mm were cast

per gel. Gels were cooled at 4°C during electrophoresis. maintained

until bromophenol

A current

of 35 mA/gel was

blue tracker dye reached the bottom edge of the gels.

Sample preparation Fresh plant tissue was thoroughly crushed in a mortar and pestle with 1 ml/g of disruption buffer (125 mM Tris-HCl pH 6.8/10% SDS/lO%P-mercaptoethanol/ 15% glycerol), heated at 95’C for 10 mm, and clarified by centrifugation in a bench-top centrifuge. Liquid samples were mixed 1: 1 with disruption buffer and heated similarly. All disrupted samples were stored sealed at -20°C until needed. 20-40 ~1 of sample were loaded per gel slot, depending on the degree of purification. Marker proteins were obtained from Pharmacia (Sweden). Gels were stained for total protein by overnight soaking in 0.2% Coomassie brilliant blue (BDH, Poole, U.K.) in 45% methanol/lo% acetic acid, and destained in 25% methanol/lO% 3MM filter paper.

acetic acid before being vacuum-dried

Electrophoretic transfer (electroblotting) Electroblotting was performed essentially

by the method

onto Whatman

of Towbin

et al. (1979).

Resolving gels were laid upon wetted nitrocellulose sheets (0.45 pm pore, Schleicher and Schuell BA 85, NH, U.S.A.), then sandwiched between wetted filter paper sheets (Whatman 3MM). The gel sandwiches were laid upon 15 X 20 X 0.8 cm Scotch-Brite scouring pads: 5 pads plus gels could be accommodated in a single transfer, with the topmost gel overlaid with another scouring pad. Large carbon electrodes (20 X 15 X 1 cm) were secured on either side of the scouring pad assembly by elastic bands strung around the outside. The electrode assembly was placed vertically upright in a narrow 3.5 1 capacity tank containing transfer buffer (25 mM Tris/192 mM glycine/20% (v/v) methanol, pH 8.3) and connected to a Shandon Southern 50V/lA destainer powerpack with the anode nearest the nitrocellulose. temperature

A current of 0.6-l

A was applied for 4-10

to 45°C on long runs did not noticeably

h. Increase in tank

affect transfer of bands from the

gels; indeed, the longer times were found to make transfer of even high MW proteins essentially quantitative, as assessed by Coomassie staining of blotted gels. Electroblots were not stained with Coomassie blue, as in our hands this proved a less effective (albeit faster) means of visualising protein bands than staining of duplicate gels, due to uneven shrinkage and fading of dried, stained blots. Enzyme-assisted indirect immunoassay Electroblots were soaked for 3-4 h at 37’C, or overnight at 22”C, in a 1% (w/v) suspension of bovine serum albumin (BSA) in 10 mM Tris-HCl/saline pH 7.4 (Tris-salineBSA buffer) to saturate free protein binding sites. Rabbit antisera were diluted 1/251/100 in Tris-saline-BSA,

and incubated

with the blots in individual

closed containers

210 on a shaker at 22°C for l-2

h. Clarified healthy plant sap extract - made by crushing

leaves 1: 1 w/v with Tris-saline, as a l/3 dilution

and clarified by centrifugation

with Tris-saline-BSA

for dilution

was incubated

for 1 h at 37°C and centrifuged

has previously

been used in sandwich

- was occasionally

used

of antisera, in which case the mixture before use. This absorption

procedure

ELISA tests for cereal viruses (Rybicki

and von

Wechmar, 1982). Blots were washed for 10 min on a shaker in at least four changes of saline. Goat anti-rabbit horseradish peroxidase conjugate (GAR/HRP, Miles Laboratories, Cape Town) was diluted l/500 in Tris-saline-BSA buffer, and incubated with the blots on a shaker for l-2

h at 22°C. After further washing in saline, the enzyme substrate solution

(25 pg/ml o-dianisidine (Sigma)/O.Ol% Hz02/10 blots were left at 22°C for 30 min. The colour water. Goat anti-rabbit fluorescein isothiocyanate in early experiments as described by Towbin et

mM Tris-HCl pH 7.4) was added, and the reaction was stopped by washing in tap (GAR-FITC) conjugates were also used al. (1979): however, detection efficien-

cy proved less than with GAR-HRP conjugates, and in addition, intrinsic fluorescence of plant proteins and pigments interfered seriously with photographic recording of results. RESULTS

Sensitivity Serial 4-fold dilutions of BMV and BSMV were electrophoresed on duplicate gels, one of each of which was then directly Coomassie-stained, and the other subjected to immuno-electroblotting.

Both virus proteins were detectable

to barely visible end-points

of

16 pug/ml (0.31 pg total) on stained gels, while BMV and BSMV proteins on electroblots were detectable

to end-points

of 0.06 pg/ml (1 ng) and 0.24 pg/ml (5 ng), respectively

(Fig. 1). Sample size in all cases was 20 pi/slot. that the technique

The enzyme immunoassay

results mean

is at least 64 times - and up to 256 times - more sensitive than

Coomassie staining. Detection BSMV diluted in to obscure

Fig. 1.

Illustration of the sensitivity of the IEB technique.

5 four-fold marker blot. l/40.

of virus in sap antisera could be used without absorption to specifically detect virus protein plant extracts; however, BMV antisera reacted sufficiently with plant proteins virus protein reaction. Incubation of these sera with clarified plant sap extract,

serial

proteins Tracks

dilutions

of BMV,

total

protein

of MW 30, 20.1 and 14.4 kilodaltons

1-8:

serial four-fold

(c) Peroxidase-stained

20 fig. BSMV antiserum

dilutions

electroblot.

was diluted

l/40.

(a) Coomassie-stained

in track

1 = 20 pg. Track

(top to bottom).

of BMV, starting

Tracks

l-8:

Arrows

indicate

gel. Tracks 6: Pharmacia

(b) Peroxidase-stained

at 20 pg. BMV antiserum

serial four-fold end-point

dilutions on originals.

l-5: LMW electro-

was diluted

of BSMV, starting

at

271

12

3

4

12345678

b

12345678

C

56

272

however, rendered the reactions

far more specific, and less subject to intense background

staining (Fig. 2). Viruses in sap were generally detectable

to the same level of efficiency

as that for pure virus. The presence of minor bands for either pure BMV protein or virus/ sap mixtures

run on gels has been noticed

disappearance

of these bands on dilution

previously (E.P. Rybicki, unpubl.

sap, indicates

that they are virus-associated,

results): the

of either pure virus, or virus diluted in plant and probably

band corresponds in MW to coat protein dimers; products of in situ proteolysis of capsid protein.

virus-derived.

The high MW

the lower MW bands appear to be

Detection of serological relationships in virus groups The serological relationship between BMV and BBMV - two members of the bromovirus group (Lane, 1981) - was tested with antisera to all three characterised members of the group. Antisera to BMV, BBMV and the related cowpea chlorotic mottle virus (CCMV) all reacted with both BMV and BBMV proteins. The reactions of the two viruses with antisera to BMV and BBMV are shown in Fig. 3. Although end-point titrations were not performed, heterospecific detection of virus proteins persisted down to 5 ng/band (not shown). This is the first confirmation of the relationship between BMV and BBMV first demonstrated by indirect ELISA (Rybicki and Von Wechmar, 1981). TMV vulgare and Nil09 proteins reacted strongly, and CGMMV protein only weakly, with a high-titre TMV vulgare-specific antiserum (Fig. 4). This is in agreement with the antigenic

relationships

2

1

between

tobamoviruses

36

2

1

Fig. 2. Detection

of BMV in barley

in a 1:l (w/v) mixture

and 20 ~1 applied BMV antiserum

per track. diluted

adsorbed

with

HC1/0.15

M NaCl,

Track

elsewhere

3

(van Regenmortel,

1

2

3

0

b

a rupted

described

l/3

fresh

macerated

Tracks

1-3:

1:l

was diluted

Samples

serially

(w/v)

electroblot, with

BMV dilutions,

ten-fold

were then heated

gel. (b) Peroxidase-stained

(c) Peroxidase-stained

barley

pH 7.4).

buffer.

(a) Coomassie-stained

l/50.

B: 10 pg of pure BMV.

sap. BMV at 2.5 mg/ml

with disruption

amount

electroblot,

BMV antiserum

incubation

buffer

in track

in barley

dis-

at 95°C for 10 min, unadsorbed

diluted

(1% BSA/O.Ol

l/50

and

M Tris-

1 = 5 ng of pure BMV.

273

a

c

b

Fig. 3. Reaction of BMV and BBMV with antisera to BMV and BBMV. (a) Coomassie-stained gel. (b), (c) Peroxidase-stained electroblots. Track 1: 5 fig of BMV. Track 2: 5 pg of BBMV. Electroblot (b) was reacted with a l/50-diluted BMV antiserum, electroblot (c) with a l/50-diluted BBMV antiserum.

1234

123

Fig. 4. Reaction of three tobamoviruses with a TMV antiserum. (a) Coomassie-stained gel. (b) Peroxidase-stained electroblot. Track 1: 20 pg of TMV-vulgare. Track 2: 20 gg of TMV-Nil09. Track 3: 20 pg of CGMMV. Track 4: Pharmacia LMW markers (molecular weights indicated, x 10-s). Electroblot in (b) was reacted with a l/50-diluted antiserum to intact TMV. The blot in (b) is offset with regard to the gel in (a), to show apparent virus protein-derived peptides visible only after immunoperoxidase staining.

274

1978; van Regenmortel that the TMV Nil09 both

reacting

mixture

and Burckard,

1980). An interesting

result was the demonstration

isolate appeared to have two distinct coat proteins of different MW,

strongly

with the vulgare antiserum.

Thus the ‘mutant’

was probably

a

of strains, both closely related to TMV vulgare.

Antisera to turnip yellow mosaic virus did not react with any of the bromo- or tobamovirus proteins used in these experiments. Alteration ofantigenicity by capsid disruption The reaction of the three distinct capsid proteins RhPV (Rybicki

and von Wechmar,

for double-antibody

of sucrose gradient-fractionated

1982) was tested using antibody

sandwich ELISA detection

globulins

prepared

of the virus (von Wechmar and Rybicki,

1981). Molecular weights of the subunits were calculated with reference to the MW markers by a linear regression programme (Statistician, Compucorp, U.S.A.). Although Coomassie staining clearly detected all three proteins present in roughly equimolar proportions, electroblots indicated that the 31 kilodalton protein reacted more weakly than the 30 and 28 kilodalton proteins. In addition, impurities not visible on stained gels gave stronger reactions than the virus proteins (Fig. 5). Preliminary tests using GARFITC detecting antibodies (E.P. Rybicki and C. Roberts, only the impurities and not the virus proteins.

1

2

3

4

1

2

massie-stained Track

1: sucrose

of R. padi virus capsid

20 pg of BSMV. Track 30.20.1

proteins

gel. (b) Peroxidasestained density

results) could detect

3

b

a Fig. 5. Reaction

unpubl.

gradient

and 14.4 kilodaltons.

electroblot.

fractionated

4: Pharmacia

with an antiserum RhPV

RhPV. Track

LMW markers.

specific

antiserum

2: unfractionated

Molecular

weights

for whole

virus. (a) COO-

was used as a l/25 (top

RhPV extract. to bottom):

dilution. Track

3:

94, 67, 43,

275

Routine virus testing with the electroblotting technique The results demonstrated

in Fig. 6a and b represent the routine application

assisted immuno-electroblotting clarified glycol

by low-speed precipitation,

in our laboratory.

centrifugation, and tested

of enzyme-

Five wheat and barley extracts were

concentrated

by centrifugation

by the Ouchterlony

or polyethylene

double-diffusion

gel precipitin

technique for the presence of BMV, BSMV and sugarcane mosaic virus (SCMV), using antisera specific for local isolates of these viruses. This preliminary testing gave apparently contradictory results (M.B. von Wechmar, unpubl. results), so the extracts were tested by immuno-,electroblotting assay using antisera specific for highly-purified BMV and BSMV. These antisera were not host-absorbed. In Fig. 6a, BSMV is readily identifiable the BMV antiserum

in lane 1, but not in lanes 2-5;

reacted with all five samples at the characteristic

in Figure 6b

position

of BMV

protein (control not shown), but most strongly with samples in lanes 1 and 4. As mentioned earlier, the BMV antiserum presumably presence

resulting

also showed up peptides

from in situ capsid proteolysis.

of relatively high concentrations

smaller than BMV coat protein, The results thus demonstrated

the

of BSMV and BMV in sample 1, of high con-

centration of BMV in sample 4, and low concentrations in samples 2,3 and 5. However, both antisera also reacted strongly with a protein of Mr approximately 43,000 daltons,

12345

12345

43 kd 25 kd 30 kd

Fig. 6. Semi-purified plant extracts tested with BMV and BSMV antisera. (a) Peroxidase-stained electroblot, reacted with BSMV antiserum diluted l/40. (b) Peroxidase-stained electroblot reacted with BMV antiserum diluted l/40. Track 1: barley cv. ‘Loerie’ infected with BSMV derived from field infection. Track 2: wheat cv. ‘Betta’, grown in absence of virus. Track 3: wheat cv. ‘Helena’, extract from field-collected plants. Track 4: barley cv. ‘Heine’, grown as healthy. Track 5: wheat cv. ‘Palala’, grown as healthy. Molecular weights derived from Coomassie-stained gels are indicated.

276 though not identically

(see Fig. 6). This band could not be correlated

any of the major plant proteins

in SDS-PAGE Coomassie-stained

immediately

with

gels; it was interesting

that a supposedly host-absorbed SCMV antiserum recognised several proteins in samples l-4 at the same relative position (results not shown). However, the apparent M, was too high for most known amentous

local SCMV strains (R.C. Chauhan,

unpubl.

results),

and no fil-

particles were visible by electron microscopy.

DISCUSSION Our results

demonstrate

the usability

and general

applications

of this technique.

Enzyme-assisted immuno-electroblotting is both a sensitive and specific means for the detection of viral coat proteins in plant extracts (see Figs. 1 and 2); it may be used successfully for the investigation

of serological relationships

within virus groups (Figs. 3 and 4);

it is capable of being used for structural investigations of virus capsids (discussed later); and yields valuable information on the specificities of supposedly virus-specific antisera (Fig. 6). That the technique may be of great value to the smaller, less well-equipped plant virus laboratory appears obvious: it is a powerful analytical and diagnostic tool which does not require much sophisticated or expensive equipment; nitrocellulose paper and o-dianisidine

are both readily available commercially,

and have long shelf lives; GAR

and HRP may be conjugated relatively easily and cheaply (Barbara and Clark, 1982) in the laboratory, and frozen for long-term storage (E.P. Rybicki and A. Kaufman, pers. obs.). The specific

investigations

in this work have both

confirmed

earlier

studies,

and

yielded valuable new information. The generation from BMV protein of antigenically reactive peptides by apparent in situ capsid proteolysis has been noticed previously (E.P. Rybicki, unpubl. results; M.H.V. van Regenmortel, pers. comm.). However, these are far more easily and specifically detectable by the immune technique than by general protein staining with Coomassie blue. The antigenic relationship between TMV vulgare and CGMMV (CV, J) has been described (van Regenmortel, 1978), as has the relationship between BMV and BBMV (Rybicki and von We&mar, 1981); this report, however, constitutes the first we know of to re-affirm the latter relationship by a technique other than indirect ELISA. The putative aphid picornavirus RhPV (D’Arcy et al., 1981a,b; Rybicki and von Wechmar, 1982) is known to have three capsid proteins, and to be stable on storage (E.P. Rybicki, pers. obs.). Immuno-electroblotting (IEB) results indicate that the capsid proteins are far less antigenic when disrupted than in the intact virion; as the antiserum used for detection of the proteins - a high titre (l/128, Ouchterlony test) late bleeding also used for sandwich ELISA detection of the virus (Rybicki and von Wechmar, 1982) - detected contaminant proteins not visible in Coomassie-stained gels at higher staining intensity than the Coomassie-detectable virion proteins (see Fig. 5). The differential staining of the three virion proteins, as compared with Coomassie-stained bands, indicates that the 3 1,000 MW protein was less antigenic than the other two, lower MW proteins. This could be due to (1) greater relative configurational changes of the

211

protein,

resulting in greater ‘distortion’

of antigenic

determinants,

location

of the protein in the intact capsid. Such effects are well known in picornavirus

serology (Putnak and Philips, 1981) and will be investigated The routine

application

of the technique

or (2) a more internal

further.

in our work can be seen in Fig. 6. It is note-

worthy that only sample 1 in lane 1 (Fig. 6, a and b) was deliberately infected, and that was with supposedly pure BSMV derived from a field infection. Samples 2,4 and 5 were grown from commercial seed without inoculation, and sample 3 was field-collected. The presence of BMV in all five samples, especially 1 and 4, is both an indication of its endemic occurrence in South African small grams (von Wechmar and Rybicki, 1981) and of its apparent seed-transmission. This is being investigated further (von We&mar and Rybicki,

in prep.).

The reaction

of both

BMV and BSMV antisera with 43,000 MW

proteins in plant sap explains earlier anomalous results in Ouchterlony tests (M.B. von We&mar, unpubl. results), which underlines the usefulness of the IEB technique. The relative merits of electroblotting methodology with regard to ELISA and ISEM techniques appear obvious. IEB tests do not require: (1) non-specific adsorption of antigen to surfaces in the presence of competing contaminants as in indirect ELISA (Rybicki and von Wechmar, 1981); (2) pre-coating of surfaces with antibodies or other proteins (Derrick,

1973; van Regenmortel

and Clark, 1982); (3) virus-specific and Burckard,

and Burckard,

antibodies

1980); (4) strain-specific

1980; Torrance,

1981; Barbara

from two animal species (van Regenmortel

antibody-enzyme

conjugates

as for sandwich

ELISA (Clark and Adams, 1977; Koenig, 1978; Rochow and Carmichael, 1979); or (5) pure virus and/or virus-specific antibodies for standardisation purposes. A strong advantage over EM ‘trapping’ and ‘decoration’ techniques, moreover, is that viruses are detected in IEB as coat protein subunits, localised by MW and serological reactivity and excess free coat protein subunits can severely inhibit ISEM detection of virus (R.G. Milne, pers. comm.) while being effectively their small size. One potentially

serious disadvantage

invisible

in decoration

of the IEB techniques

experiments is the possibility

antisera to intact capsids will not recognise dissociated subunits (van Regenmortel, Indications

due to that 1981).

of this, however, were noticed with only one out of 17 viruses tested by IEB

(this paper, and O’Donnell

et al., 1982). The problem

could perhaps be avoided alto-

gether by use of antisera specific for SDS-disrupted capsids (Purcifull et al., 1981). A possible advantage of this approach in virus group relationship studies lies in the observation that virus strain subunits are often more closely serologically related than the intact virions (Shepard et al., 1974; Rybicki and von Wechmar, 1981). IEB could thus become the method of choice for studying intra-group relationships, rather than ISEM or ELISA. The enzyme-assisted IEB procedure described here is probably less sensitive, and less amenable to quantitation, than the radioimmunoassay (RIA)variant described by O’Donnell et al. (1982). However, it is probably both cheaper and more convenient, as it does not require radiochemicals, photographic film, a scintillation counter, or special handling techniques. An advantage of the RIA technique is the option for re-probing of proteins

218

blotted

onto diazophenylthioether

techniques protein

paper. An important

in general could include

future

application

their use - in conjunction

of the IEB

with sensitive

general-

silver staining (Ochs et al., 198 1) - as a new criterion of purity of viruses, and of

the specificity detection

of antisera used to study them. Other conceivable

of viral coat protein

antigens in cell-free translation

applications products,

are: (1) the

in the absence

of radio-labelled amino acids; and (2) the antigenic mapping of virus protein peptides after one- or two-dimensional peptide mapping (Cleveland et al., 1977; Koenig et al., 1981). That the latter proposition is feasible is shown by the detection of both BMV and TMV peptides in this study (Figs. 1,2 and 4). ACKNOWLEDGEMENTS

We wish to thank the University of Cape Town and the Wheat Board for their financial support, and Mr. P. Smith, Miss C. Roberts and Mrs. A. Kaufman for their valuable technical assistance. REFERENCES Atabekov,

J.G. and V.K. Novikow,

Barbara,

D.J. and M.F. Clark,

1971, CMI/AAB

Clark, M.F. and A.N. Adams,

1977, J. Gen. Virol.

Cleveland,

M.W. Kirschner

D.W., S.G. Fischer,

D’Arcy,

C.J., P.A. Burnett,

D’Arcy,

C.J., P.A. Burnett

Derrick,

KS.,

A.D. Hewings

R., 1978, J. Gen. Virol. 40, 309. R., H. Francksen 1981,

34,473.

and R.M. Goodman, 198lb,

and H. Stegemann,

U.K., 1970, Nature

L.C.,

(London)

in: Handbook

(Elsevier/North-Holland

1977, J. Biol. Chem.

198la,

Virology

Virology

1981, Phytopathology

2 100,347.

227,680.

of Plant

Biomedical

Virus

Infections.

Press, Amsterdam)

Comparative

and D.W. Sammons, and K.H. Gough,

1982, J. Virol. Methods

and J.A.A.

1981, Phytopathology

I. J., D.D. Shukla D.E., S.R. Christie

J.R. and B.A. Phillips,

Lima,

Rochow,

W.F. and L.E. Carmichael,

Rybicki,

E.P. and M.B. von Wechmar,

Shepard, Torrance,

J.F., G.A. Secor and D.E. Purcifull, L., 1980, J. Gen. Virol. 5 1, 229.

Towbin,

H., T. Staehaelin

1981, Electrophoresis

1981, Microbial.

Van Regenmortel,

M.H.V.,

1978, Ann. Rev. Phytopathol.

viruses.

in: Comprehensive

Publishing

Corpcration,

Van Regenmortel.

M.H.V.

58,464.

1979, Proc. Natl. Acad. Sci. U.S.A.

1975, Virology 1981,

71, 1221.

Z. 103, 306.

1974, Virology

M.H.V., M.H.V.,

2, 304. 4, 19.

95,415.

1982, Phytopath.

Van Regenmortel, Van Regenmortel,

76,435O.

64,415.

Serological

Virology,

vol.

methods 17, eds.

16,57. in the identification H. Frankel-Conrat

and

1980, Virology

Von Wechmar,

M.B. and M.H.V. van Regenmortel,

Von Wechmar,

M.B. and E.P. Rybicki,

106, 327.

1970, S. Afr. Med. J. 44, 15 1.

1981, S. Afr. J. Sci. 77, 488.

characterisation

and R. Wagner

New York) p. 183. and J. Burckard,

ed. E. Kurstak

Rev. 45,287.

1979, Virology

and J. Gordon,

Diagnosis.

p. 333.

Ochs, D.C., E.H. McConkey

Putnak,

252, 1102.

112, 346.

114,268.

O’Donnell, Purcifull,

No. 68.

65,652.

Koenig,

Lane,

of Plant Viruses,

and U.K. Laemmli,

and A.D. Hewings,

1973, Virology

Koenig, Laemmli,

Descriptions

1982, J. Gen. Virol. 58, 315.

of

(Plenum