Molecular Aspects of the Antigenic Variation of ... - Semantic Scholar
J. gen. ViroL ( I 9 7 7 ) , 35, 2 9 9 - 3 1 5
299
Printed in Great Britain
Molecular Aspects o f the Antigenic Variation of Swine Vesicular Disease and Coxsackie B5 Viruses By T. J. R. H A R R I S ,
T. R. D O E L
AND F. B R O W N
Department of Biochemistry, Animal Virus Research Institute, Pirbright, Woking, Surrey, U.K. (Accepted 7 December 1976) SUMMARY
The antigenic variation of swine vesicular disease virus (SVDV) and Coxsackie B5 virus (CB5) has been examined at the molecular level by analysing the protein and nucleic acid of the virus particles. The tryptic peptides of carboxymethylated 35S-methionine labelled virus particles were very similar, although some minor differences were apparent. Competition hybridization experiments confirmed that there is variation in the RNA sequence of antigenically distinct SVD viruses and some limited homology of these RNAs to the CB5 virus RNAs. Competition hybridization using mixtures of two RNAs as competitors showed that the sequences shared by the CB5 virus RNAs were largely the same as those shared by the CB5 virus RNAs and the SVDV RNAs. Similar experiments with the SVDV RNAs established that the homologous regions in these RNAs were also shared. Thermal denaturation curves of SVDV RNA-RNA hybrids generally supported the competition hybridization results but the hybrids between SVDV RNA and CB5 virus RNA were shown to be mismatched. Ribonuclease Tx oligonucleotide maps of the virus RNAs were also compared to obtain another measure of relatedness. A number of long oligonucleotides were shared by the SVDV RNAs but few of these were found in either of the CB5 RNAs. Further ways of investigating the antigenic variation of SVDV at the molecular level are discussed.
INTRODUCTION
Although swine vesicular disease virus (SVDV) is antigenically related to Coxsackie B5 virus (CB5; Graves, i973) the viruses can be distinguished by cross neutralization and immunodiffusion tests (Brown, Talbot & Burrows, 1973; Brown et al. 1976). There is also antigenic variation within the two virus types which can be detected by immunodiffusion. For example, the United Kingdom isolate of SVDV (SVDV-UK) gives a line of partial identity with a Hong Kong isolate (SVDV-HK) and the prototype strain of CB5 (CB5Faulkner) gives a line of partial identity with a recent isolate CB5-8o68 (Brown et al. I976 ). Evidence for this variation has also been obtained from analysis of the virus structural polypeptides on SDS-polyacrylamide slab gels, since the polypeptides of anfigenically distinct virus isolates have slightly different mobilities (Harris & Brown, 1975). The virus RNAs which code for these polypeptides have also been studied. Using radioactively labelled double stranded RNA as a source of complementary RNA, Brown & Wild (I974) showed that about 5o ~ of the sequence of SVDV-UK RNA was shared by the RNAs of CB5Faulkner and CB5-8o68. In an extension of these studies we showed that SVDV-UK R N A 20-2
300
T.J.R.
H A R R I S , T. R. D O E L A N D F. B R O W N
was 80 ~ homologous to the RNA of SVDV-HK and 90 ~ homologous to the RNA of the SVDV (SVDV-It66) isolated in Italy in 1966 (Brown et al. I976 ). The RNAs of SVDVUK, CB5-Faulkner and CB5-8o68 have also been examined by specific nuclease digestion (Frisby et al. I976). Few, if any, long oligonucleotides produced by ribonuclease T1 (RNase TO digestion were shared. In view of the possible role of the human CB5 viruses in the etiology of swine vesicular disease, it was clearly important to characterize the antigenic variation in more detail by biochemical methods. This paper describes the analysis of the structural proteins of antigenically distinct SVDV and CB5 isolates by peptide mapping techniques and the examination of the sequence variation of the RNAs using more precise competition hybridization and oligonucleotide fingerprinting techniques.
METHODS
Viruses. The SVD and CB5 viruses were grown at 37 °C in Eagle's medium in monolayers of IBRS-2 cells in Roux bottles. The following SVDV isolates were used: United Kingdom 27[72 (SVDV-UK); Hong Kong 3617t (SVDV-HK) and Italy 1/66 (SVDV-It66). The CB5 isolates were the prototype strain CB5-Faulkner and a CB5 isolated in I973 (CB5-8o68). These viruses were passaged only twice in IBRS-z cells to retain as closely as possible the characteristics of the original isolates. Poliovirus (type ~, Mahoney) was grown in monolayers of HeLa cells at 37 °C. Preparation of radioactively labelled viruses. Monolayers of about 2 × IO7 IBRS-2 cells in Roux bottles were infected with virus at high multiplicity in 2o ml of Eagle's medium. After an adsorption period of 30 min, the monolayers were washed with the appropriate labelling medium, 20 ml of fresh medium added and the cells incubated at 37 °C. Radioactive label was added to the cultures I-5 to 2 h after infection. For 35S-methionine labelling Ioo #Ci of 35S-methionine (ioo Ci/mmol) in 2o ml of methionine-free Eagle's medium was used. Virus RNA was labelled to high specific activity with 3zp by adding 25o to 5oo #Ci/ml carrier free sZP-orthophosphate in 2o ml of phosphate-free Earle's saline. All isotopes were from the Radiochemical Centre, Amersham. Purification of virus. Radioactively labelled virus was purified from the supernatant medium at the end of the growth cycle when the cells had left the glass. The cellular debris was removed by centrifuging at 4ooog for I5 min at Io °C and the medium clarified further by centrifuging at ioooo g for 20 min at Io °C (Ioooo rev/min, MSE rotor 59595). Virus was then pelleted by centrifuging for 1-25 h at 8oooo g in the same rotor (3oooo rev/min) at Io °C. The pellet was resuspended in I"5 ml of o.I M-NaCI, 0"o5 M-tris-HCl, pH 7"6, for at least I6 h, SDS added to I ~ (w]v), and the virus suspension layered on to a sucrose gradient (I5 ~ to 45 ~ [w/v] in o.I M-NaC1, o'o5 M-tris-HC1, pH 7"6) which was centrifuged at 8oooo g for 2"25 h at Io °C (3oooo rev/min, MSE rotor 5959o). The gradient was collected in I ml fractions into bijou bottles containing 25/zg of bovine serum albumin, and the fractions containing labelled virus were pooled. The purity of the virus was judged by SDS polyacrylamide gel electrophoresis of the polypeptides. Unlabelled virus (3oo to 4oo ml) grown in 15 to 2o Roux bottles was first precipitated with 5o ~o (v/v) saturated ammonium sulphate buffered by o-o4 M-phosphate, pH 7"6, and the precipitate resuspended in 4o ml of o.o4 M-phosphate. After clarification at 1oooo rev/min, virus was pelleted from this suspension by centrifuging at 30ooo rev/min as described for radioactive virus. Sucrose gradients of unlabelled virus were monitored by measuring the extinction at 260 nm.
Molecular aspects of S VD V-Coxsackie B 5 variation
30 t
Trypsin digestion. Sucrose gradient fractions containing virus were diluted twofold and the virus particles precipitated with 2 vol. acetone overnight at - 2 o °C in the presence of I5O #g of bovine serum albumin. Virus was carboxymethylated in 8.o M-urea by the method of Crestfield, Moore & Stein (1963), modified by inclusion ofo'5 ~ (v/v) SDS and substitution of fl-mercaptoethanol with o.o6 M-dithiothreitol. The carboxymethylated protein was recovered after this procedure by acetone precipitation and washed by repeated acetone precipitation. The dried pellets were dissolved in z ml of o.I M-NH4HCO3, pH 8.o and digested with I ~ (w/w) TPCK trypsin (Worthington Biochemical Corporation, Freehold, New Jersey) at 37 °C for 24 h. A further I ~ (w[w) trypsin was added after 4 h. The digests were filtered through nitrocellulose membranes (25 m/t, Millipore) and evaporated to dryness in a stream of N 2 at 4o °C. The residues were dissolved in io to 25 #1 of o.z MNH4OH for tryptic peptide analysis. Analysis of tryptic peptides. Tryptic peptides were separated by electrophoresis and chromatography on 2o × zo cm silica gel thin layer sheets (Machery-Nagel, SIL G) as described by Sargent & Vadlamudi (1968) except that electrophoresis was continued at 2oo V for 6 h. This method has the advantage that two samples can be subjected to electrophoresis at the same time under identical conditions. For ascending chromatography the sheets were cut in half with scissors and both halves run for 2"5 h in the same rectangular tank containing H20: acetic acid:butan-I-ol (1 : I : 3). Autoradiographs were prepared by exposing Kodak RP Royal X-Omat X-ray film to the dried sheets for 5 to Io days. Extraction of RNA from virus particles. The sucrose gradient fractions containing virus were diluted with o'i5 M-NaC1, o'o5 M-tris-HC1, o'oo5 M-EDTA, pH 7"6 (TNE buffer), containing o.I ~ SDS, Ioo #g of E. coli tRNA added, extracted twice with a I ; I mixture of phenol: chloroform and the RNA precipitated with 2 vol. ethanol overnight at - 2o °C. The RNA was re-precipitated once with ethanol before use. Extraction of double stranded RNA from infected cells. Double stranded RNA (ds RNA) was extracted and purified from about 5 × Io8 virus-infected cells by phenol extraction and 2 M-LiC1 precipitation. The ds RNA in the LiC1 soluble fraction was purified further from low mol. wt. single stranded RNA (ss RNA) by centrifuging on I5 ml 5 to 25 ~ (w/v) sucrose gradients in o.i M-sodium acetate, o.I ~ SDS, pH 5"o, for 16 h at 8oooog (MSE rotor 591o8) at 2o °C. The ds RNA was recovered from the gradient by ethanol precipitation. These procedures have been described in detail previously (Harris & Brown, i977). The precipitates were dissolved in 5o ~ formamide, 5 × SSC (o'75 M-NaCI, o'75 M-Nacitrate) to give concentrations of ds RNA of approx. 5 to xo/zg/ml based on the extinction of the sucrose gradient fractions at 260 nm. Molecular hybridization. Saturation and competition hybridization experiments were done in duplicate in o.2 ml of 5o ~ formamide, 5 × SSC at 50 °C exactly as described by Harris & Brown (I977). The competitor RNAs were dissolved in 5o ~ formamide, 5 × SSC to give a concentration of 8 to Io/zg/ml. The results obtained with the hybridization procedures were handled as described in detail by Darby & Minson (I973). Briefly, the reciprocal plots of the competition hybridizations give intercepts on the I[f axis which allow determination of f, the fraction of RNA displaced by the competitor, at infinite competitor RNA concentration. The reciprocal plots for the saturation hybridizations give intercepts on the I[F axis which allow the determination of F, the fraction of labelled RNA homologous to the ds RNA, at infinite ds RNA concentration. Thermal denaturation profiles. These were obtained as described by Harris & Brown (I977). Briefly, labelled homologous and heterologous hybrids were prepared by annealing
302
T.J.R. HARRIS, T. R. DOEL AND F. BROWN
3ZP-virus RNAs to denatured ds RNA at 50 °C in o'o5 ml of 50 7o formamide, 5 x SSC. The samples were diluted to I ml with de-ionized H20, heated to various temperatures, cooled quickly and the fraction of label remaining in the hybrid determined from the residual ribonuclease resistance. Polyacrylamide gel electrophoresis of the RNase Ti-resistant oligonucleotides. Virus RNAs, labelled with 32p, were digested for I h at 37 °C with RNase T 1 (SANKYO, Japan) in IO/~l of o.oI M-tris-HCl, o.ooi M-EDTA, pH 7"4, at an enzyme: substrate ratio of 1:2o. The oligonucleotides produced were separated by two dimensional gel electrophoresis (De Wachter & Fiers, I972), using the buffer modification of Frisby et aL (1976) and the gel dimensions and running conditions of Harris & Brown (I977). Autoradiographs were prepared using Kodak RP-Royal X-Omat X ray film. Analysis of oligonucleotides .from polyacrylamide gels. The radioactive oligonucleotides were eluted from the gels as described by Porter, Carey & Fellner (I974), digested with ribonuclease A (RNase A, Worthington) in o-oi M-tris-HCl, o-ool M-EDTA at an enzyme: substrate ratio of I : 20 and the products fractionated by high voltage ionophoresis on DEAE paper (Whatman DE 81) in 0"5 7o pyridine, 5 ~ acetic acid, pH 3"5. The identity of the products was determined from their mobilities relative to markers. RESULTS
Tryptic peptide analysis of virus particles Although the polypeptides of antigenically distinct isolates of SVDV and CB5 can be distinguished by polyacrylamide gel electrophoresis (Harris & Brown, I975), it is not possible to decide from this observation to what extent the amino acid sequence has varied. To investigate this point sSS-methionine labelled virus particles of different SVDV and CB5 isolates were digested with trypsin and the tryptic peptides analysed by electrophoresis and chromatography on thin layer silica gel sheets (Sargent & Vadlamudi, I968). Fig. I (a to e) shows maps of the tryptic peptides of the isolates prepared by autoradiography of the thin layer plates. All the isolates have the same overall pattern of tryptic peptides and the maps of the SVDV isolates (Fig. I a to c) are more similar to each other than to the CB5 isolates (Fig. I d, e). There are, however, some distinct differences between the maps of the individual SVD viruses and the CB 5 viruses. As these maps simply represent the tryptic peptides of virus particles, it is impossible to say from which of the four polypeptides these differences arise. The tryptic peptide map of 35S-methionine labelled poliovirus prepared in the same way (Fig. If) bears little resemblance to any of the SVDV and CB5 maps. Tryptic peptide maps of 35S-methionine labelled SVD and CB 5 virus particles have also been prepared by high voltage paper electrophoresis and chromatography. The results confirmed the similarity of the maps of SVDV and CB5 determined by the thin layer method (T. R. Doel, unpublished results).
Competition hybridization of the virus RNAs Saturation hybridization of SVDV and CB5 RNAs has shown that they have about 5o homology (Brown & Wild, 1974; Brown et al. I976). More precise estimates of homology can be obtained, however, by competition hybridization. In the method we have used, 3~P-ss RNA was prevented from annealing to unlabelled complementary RNA (denatured ds RNA) by the addition of increasing amounts of unlabelled ssRNAs from the heterologous viruses. The homology of the RNAs of SVDV-UK, SVDV-HK and SVDV-It66 were
Fig. t. Autoradiographs of the tryptic peptides of various virus particles separated in two dimensions by eleclrophoresis and chromatography on thin layer silica gel plates. (a) SVDV-UK, (b) SVDV-HK, (c) SVDV-It66, (d) CB5-Faulkner, (e) CB5-8o68, (f) poliovirus. The letter (o) marks the origin of the plates. The direction of chromatography is shown by the large arrow.
O
E"
304
T . J . R . H A R R I S , T. R. DOEL A N D F. B R O W N I
i
I
I
I
I
(a)
I
I
(b)
"•4"0 ~5 3-0 2.0
.~ 1"0 I
0.05
I
0.10
I
I
I
I
0-15 0.20 0.025 Competitor RNA (I/c)/zl
i
0-050
0.075
i
0.100
- t
Fig. 2. Competition hybridization of SVDV-UK RNA. Reciprocal plots of the annealing of 3~p. SVDV-UK ss RNA to denatured SVDV-UK ds RNA in the presence of increasing amounts of (a) SVDV-UK RNA (0--0); SVDV-HK RNA (©--©), SVDV-It66 RNA ( I - - I ) a n d (b) a 5o:5o mixture of both SVDV-HK and SVDV-It66 RNAs ( I - - I ) . Table i. RNA sequence homology of SVDV RNAs determined by
competition hybridization 3~P-ss RNA and unlabelled ds RNA r
Unlabelled competing RNAs SVDV-UK SVDV-HK SVDV-It66 SVDV-HK + SVDV-It66 SVDV-UK + SVDV-HK
*SVDV-UK
SVDV-HK
SVDV-It66
Ioo 86 91 91 --
85 Ioo 80
87 83 ioo
-
-
--
-
-
90
* These values are the means of two determinations, including that shown in Fig. z. first examined by this method. S V D V - U K RNA, labelled with 32p, was annealed to SVDVU K ds R N A in the presence of S V D V - H K and SVDV-It66 virus RNAs. The results, which also show the control homologous competition using SVDV-UK RNA, give I / f intercepts of 1.2o and I,I 3 for S V D V - H K and SVDV-It66 R N A (Fig. 2a), corresponding to a sequence homology with S V D V - U K R N A of 83 ~o and 89 ~ respectively. These values are in good agreement with those obtained previously using radioactive ds R N A (Brown et aL I976 ). A competition curve was also constructed using increasing amounts of a 5o: 5o mixture of S V D V - H K and SVDV-It66 R N A s as competitor. The intercept of I.I2 (Fig. 2b) corresponds to a homology of 89 to 90 ~ and is within the error of the value obtained with SVDV-It66 R N A alone as competitor (Fig. 2 a), demonstrating that the regions of homology are not additive. Thus, the sequences in S V D V - H K and SVDV-It66 R N A s that are homologous to S V D V - U K R N A are shared. Corresponding competition experiments have been done by annealing 32p-SVDV-HK R N A and 32p-SVDV-It66 to their complementary RNAs in the presence of unlabelled heterologous SVDV RNAs. The results of the reciprocal plots of these data (Table I) confirm those in Fig. 2 (a). To examine the relationship of SVDV R N A s to the RNAs of CB5-Faulkner and CB58o68, 32p-SVDV-UK R N A was annealed to denatured homologous ds R N ~ in the presence
Molecular aspects of SVD V-Coxsackie B5 variation I
I
I
I
8-0 - (a)
305
I
I
I
I
I
I
I
I
- (b)
6"0
o
•
4"0 "~
2-0
F. e~
I
e-
~16.0 O
I
I
I
(c)
_. (a) o
.~ 12.0 8.0 4"0 I
I
0"05
0-10
I
!
l
0"15 0.20 0"025 Competitor RNA (l/c)/11-1
t
0-050
I
0-075
I
0.100
Fig. 3. Competition hybridization of SVDV-UK and CB5-Faulkner RNA. Reciprocal plots of: 0 ) 3~P-SVDV-UK R N A annealed to denatured ds R N A in the presence of increasing amounts of (a) SVDV-UK R N A (O---O), CB5-Faulkner R N A ( O - - O ) , CB5-8o68 R_NA ( l l - - i ) ; ~,(b) a 50:50 mixture of CB5-Faulkner and CB5-8o68 RNA. (2) 3~P-CB5-Faulkner R N A annealed to denatured ds R N A in the presence of increasing amounts of (c) CB5-Faulkner R N A (O---O), SVDV-UK R N A ( O - - O ) , CB5-8o68 R N A ( i - - i ) ; (d)a 50:50 mixture of SVDV-UK R N A and CB5-8o68 RlqA.
Table z. RNA sequence homology of SVDV-RNA and CB5 RNA determined
by competition hybridization azP-ss R N A and unlabelled ds R N A t
Unlabelled competing ss RNAs SVDV-UK CB5-Faulkner CB5-8o68 CB5-Faulkner + CB5-8o68 SVDV-UK + CB5-8o68 SVDV-UK + CB5-Faulkner
*SVDV-UK
*CB5-Faulkner
CB5-8o68
! oo 45 53 54 ---
31 ioo 34 -39 --
35 33 ioo --43
* These values are taken from the intercepts shown in Fig. 3.
of increasing amounts of unlabelled SVDV-UK, CB5-Faulkner and CB5-8o68 virus RNAs. Fig. 3 (a) shows the reciprocal plot of these hybridization results. CB5 Faulkner RNA gives a I/f intercept of 2-20 and CB5-8o68 an intercept of 1.9o, corresponding to sequence homologies with SVDV-UK RNA of 45 ~ and 53 ~ respectively. These values are in agreement with those found by Brown & Wild (i974). The result of the mixed competition curve using a 50:50 mixture of both CB 5 RNAs as competitors (Fig. 3b) gives an intercept of ]'85 (corresponding to a sequence homology of 54 ~o)- The non-additive sequence homology
T. J. R. H A R R I S , T. R. D O E L A N D
306 I
I
I
I
I
"< 0-6 z 0.4 O
I
I
I
(b)
(a]
~0"8
I
F. B R O W N
16'0
/
12"0
o
8"0 4"0
•~ 0.2 I
10
I
I
I
20 30 40 ~1 dsRNA(d)
I
50
I
0-05
I
0-10 1/d
I
0-15
I
0.20
Fig. 4. Saturationhybridizationof SVDV and CB5 RNA. (a) Increasingamounts of denatured ds RNA from SVDV-UK infectedcells added to 3~P-SVDV-UK RNA (O--O), 3~P-CB5-Faulkner RNA (G--Q), 3~P-CB5-8o68RNA ( I - - I ) . (b) Reciprocalplots of the data. indicates that CB5-Faulkner RNA and CB5-8o68 RNA have similar sequences homologous to SVDV-UK RNA. CB5 Faulkner RNA labelled with 32p has also been annealed to denatured CB5-Faulkner ds RNA in the presence of unlabelled competitor RNAs. The results in Fig. 3 (c) show intercepts of 3.20 and 3"oo for SVDV-HK and CB5-8o68 RNA, corresponding to homologies of 31 ~ and 34 ~ respectively. In another experiment in which SVDV-HK and SVDV-It66 RNAs were used instead of SVDV-UK RNA, the results were similar to those with the SVDV-UK RNA. Fig. 3 (d) shows the result of the mixed competition with a 5o: 50 mixture of SVDV-UK and CB5-8o68 RNAs as competitor The intercept of 2.6o (39 homology) shows that the sequence homology of SVDV-UK, CB5-Faulkner and CB5-8o68 RNA is non-additive, again indicating that the homologous sequences in these three RNAs are shared. Table 2 shows the homology of SVDV RNA and CB 5 RNA taken from the intercepts in Fig. 3 and from the reciprocal plot of an additional confirmatory competition experiment using 32P-CB5-8o68 RNA and unlabelled CB5-8o68 ds RNA. Saturation hybridization and thermal denaturation of the RNA hybrids CB 5 RNAs, labelled with 32p, were added to increasing amounts of denatured SVDV-UK ds RNA and the level of annealing assessed from the residual ribonuclease resistance. Fig. 4 shows normal and reciprocal plots of the data. The intercepts of 5.zo and 4'8o in the reciprocal plots indicate a homology of I9 ~ and 2I ~ respectively for CB5-Faulkner and CB5-8o68 RNA. Similar results were obtained with CB5-Faulkner ds RNA. Thus, the estimate of sequence homology of SVDV-UK and CB5 RNAs measured by this method is even less than that measured by competition hybridization (Fig. 3, Table 2), or by saturation hybridization using denatured radioactive ds RNA (Brown et al. I976). In view of this discrepancy, RNA-RNA hybrids prepared by similar saturation hybridization experiments with denatured SVDV-UK ds RNA were subjected to thermal denaturation. The results in Fig. 5 (a) show that the CB5/SVDV RNA hybrids are much more mismatched than is the homologous SVDV-UK hybrid, since a Tm of 9o °C was obtained for the homologous hybrid compared with much lower values for SVDV-UK[CB5-Faulkner and SVDV-UK/ CB5-8o68 RNA hybrids. The melting curves of heterologous SVDV RNA hybrids were also examined. Denatured SVDV UK ds RNA was added to 32P-SVDV ss RNAs and the resulting hybrids subjected
Molecular aspects o f S VDV-Coxsackie B5 variation I
I
I
70
80
90
(b)
I
I
307 T
0.2
.E 0"4 < z =;
~ 0-6 0
0.8 ,
,
70 1O0 Temperature (°C)
80
!
I
90
100
Fig. 5. Thermal denaturation of hybrids prepared by annealing denatured SVDV-UK ds RNA to 32P-ss virus Rl'qAs. (a) SVDV-UK RNA (~--O), CB5-Faulkner RNA ( O - - O ) , CB5-8o68 RNA (D--D) and (b) SVDV-UK RNA (O-O), SVDV-HK RNA ( O - - O ) and SVDV-It66 RNA
(m--m).
to thermal denaturation. The saturation curves used to monitor the hybridizations are not shown, but in this case the results agreed closely with the competition hybridization data (Fig. 2, Table 0. Fig 5 (b) shows the thermal denaturation profile of the SVDV-UK RNA hybrids. The Tms of the heterologous hybrids (SVDV-HK RNA, 88"5 °C; SVDV-It66 RNA, 86 °C) indicate that these duplexes are less well matched than the homologous hybrid (Tm = 90 °C; Fig 5a and b) and that the SVDV-It66/SVDV-UK RNA duplex is more mismatched than the SVDV-HK/SVDV-UK RNA hybrid.
Oligonucleotide fingerprints of the RNAs Few of the long oligonucleotides produced by RNase T z digestion of SVDV-UK RNA were found in similar digests of CB5-Faulkner and CB5-8o68 RNA (Frisby et al. T976). In view of the limited homology of these three virus RNAs (30 to 50 ~, Table 2) and the extensive mismatch of the hybrids (Fig. 5 a), this result is not unexpected. It was of considerable interest, however, to compare the characteristic RNase T1 oligonucleotides of the more closely related SVDV RNAs and to contrast them with the RNase T1 oligonucleotides of CB5-Faulkner and CB5-8o68 RNA. asP-labelled SVDV and CB 5 RNAs were extracted from purified virus particles and the RNase T1 oligonucleotides of each RNA separated by two dimensional gel electrophoresis (De Wachter & Fiers, I972). The fingerprints obtained by autoradiography of the gels are shown in Fig. 6(a to e), together with a key to the top half of the autoradiographs showing the long oligonucleotides (Fig. 6 f to j). The patterns for SVDV-UK RNA and the two CB5 RNAs (Fig. 6a, d, e), correspond to those obtained by Frisby et al. 0976) confirming the reproducibility of the method. There is a striking similarity in the position of a number of the long RNase Tz oligonucleotides of the SVDV RNAs (Fig. 6a to c, f t o h), but few, if any, similar long oligonucleotides were found in the CB5-Faulkner and CB5-8o68 fingerprints (Fig. 6d, e, i,/). Fig. 6 also shows that SVDV-UK RNA has more of its oligonucleotides in common with SVDV-HK RNA than with SVDV-It66 RNA.
308
T . J . R . HARRIS~ T. R.r'DOEL A N D F. B R O W N 1st 2nd
)
6 ,O
Fig. 6. (a to e) Autoradiographs of two dimensional polyacrylamide gels of the RNase T1 oligonucleotides of 32P-SVDV and CB5 RNAs. The large arrows beside (a) indicate the directions of electrophoresis.
Molecular aspects of SVD V-Coxsackie B5 variation UK
(f)
309
(g)
IlK
\ 1 O
1 2 OO
~
34 C~~5 ~ 61514
0
18 1716
22019
0 4) 5 10
8 7 .6 12 16 • ~B' 1 3 ~ 1~(~ T7 15
1~23
(/,)
It66
14
9
~'"
Faulkner
(0
1
1 0 3
0 2 o
2 0
4 ©
%% 22
Io0008 "/~pO 1615 12
g
4
3 C)
9 "~ j7
10
f- ,l
0
6 5 0
O
0
23
(/)
8068
0
©2
3
g
o
.o~ 1~3~
g g
Fig. 6. (fto]) Enlarged tracings of the top halves of the autoradiographs shown in (a to e), to provide a key for comparison. The numbers refer to those oligonucleotides analysed further by RNase A digestion (Tables 3 and 4). Oligonucleotides marked • are those shared by all 3 SVDV RNAs; O, shared by SVDV-UK and SVDV-HK RNAs; 0, shared by SVDV-UK and SVDV-It66 RNAs @, shared by SVDT-HK and SVDV-It66 RNAs (see Table 3). The shaded oligonucleotides in the keys to CBs-Faulkner and CB5-8o68 RNAs (i,]) are those possibly present in the SVDV RNAs (see Table 4). The streak is the heterogeneous poly(A) tract (Frisby et aL z976).
SVDV-UK RNA
Table 3.* Spot no. SVDV-HK RNA
IAAU, 2-3AU, IG, 2-3C, 3"4U
I5
I4
2-3U IAAAC, IAU, IAC, tG, IAAAU, 2-3U (IAAAU?), IAAAC, IAAC, 2AU, IAC, IG, 2C, 3U
2C, 3U IAAAC, 0AAU?)0AAC?) 12t IAAAU, IAU, IAC, I(3",2C, 3U I3 (IAAAU?) IAAAC, IAAC, 2AU, IAC, IG, 2C, 2U
:6
(IAAAU?.):AAAC(~AAU2) :AAC, 2AU, IAC, IG, I-2C, 2-3U
zU
14] I AAAUOAAAC?), 2AAU0 AAC ?) 2AU, IAC, IG, 2-3C, 3U 15 IAAU, IAAAC, IAU, IAC, IG, 2C,
I2 IAAAA(U), IAG, IAC, 2-3C , 3-4U 13 IAAU, IAAC, I-2AU, (tAC?), IG, 2C, 3U
9 IAAG, IAAU, z-3AAC, IAU, IAC, 2C, 2U IO IAAAC, IAAC, IAU, IAC, IG, I-2C, I-2U II IAAU, 2AU, 0AC?), :G, 2C, ~ 4U
1-2U
5 IAAU, IAAC, IAU, 2AC, IG, IC, 2-3U 6 I-2AAAC, IAAU, 2AAC, IAU, 2-3AC, IG, 2C, 2-3U 7 tAAG, IAAU, 2AAC, tAG, 2-3AU, 2-3AC, 3C, 3U 8 IAAAA(C), lAG, IAU, IAC, I-2C,
5 I-2AAU, I-2AAC, IAU, 2AC, IG, I-2C, 2-3U 6 t-2AAAC, IAAU, 2AAC, IAU, 2-3 AC, IG, I-2C, 2-3U 7 IAAG, IAAU, I-2AAC, IAG, 2AU, 2AC, 3C, 3-4U 8 IAAAA(C), IAG, IAU, 2AC, I-2C, I-2U
4
I-2AU, IAC, IG, 2-3C, 5U
3 IAAU, IAAC, tAG, IAU, IAC, 3-4C, 2-3U
4 zAU, I-2AC, tG, 2-3C, 3-4U
IO tAAAA(U), IAG, 2AC, 2-3C, 3-5U 9 tAAAA(U), IAG, 2-3AC, 2C, 4U II' IAAU, IAAC, IAU, IAC, tG, I-2C, 2-3U 12 (IAAU?)IAAC, IAU, IAC, IG, I-2C, I0 I-2AAC, I-2AU, IAC, IG, 2C, 2-3U 2-3U I3 1AAAU, I-2AAU, IAU, IAC, IG, 2C, I I ] IAAAU, I-2AAU, IAU, IAC, IG,
9
5 IAAU, tAAC, xAU, 2-3AC, IG, IC 2-3U 6 I-2AAAC, IAAU, 2AAC, IAU, 2-3AC, IG, IC, 2U 7 IAAG, IAAU, I-2AAC, IAG, 2AU, 3AC, 2-3C, 2U 8 1AAAA(C),IAG, IAU, 2-3AC, 2C, 2U
SVDV-It66 RNA
I IAAAC, IAAG, IAU, IAC, IC, 2U 2 IAAAC, IAAU, IAAC, 2AU, IAC, IG, 3-4C, 3U
Spot no.
Ribonuclease A analysis of the long RNase I"1 oligonucleotides of SVDV RNAs~
I IAAAC, IAAG, I-2AU, 2-3AC, IC, 2U I tAAAC, rAAG, IAU, 2AC, IC, IU z IAAAU, IAAAC, /AAU, IAAC, I-2AC, IG, 2-3C, 2-3U 3 IAAAC, IAAU, 2AU, IAC, IG, 2C, 2) IAAAU, I-2AAU, t-2AAC, IAG, I-zU 3J 2AU, I-2AC, IG, 2-3C, 3-4U 4 IAAU, IAAC, IAG, 2-3AU, IAC, 3-4C, zU
Spot no.~
Mixed oligonucleotides, but common to all 3 RNAs
Probably shared by UK and It66 RNAs Common Shared by UK and It66 RNAs Probably shared by UK and HK RNAs
Unique
Unique
Mixture of two oligonucleotides - common Common
Common
From gel position UK spot 4, HK spot 2 probably correspond. From analysis UK spot 4-It66 spot 3 correspond Shared by HK and It66 RNAs Common
Common Unique
Oligonucleotide comparison
Z
©
©
SVDV-UK RNA
IAAAA(C), IAAU, I-2AAC, IAG, IAU, zAC, 3C, 3U
IAAAC, IAG, IAC, 2C, IU
IAAG, IAAC, 2AC, 2C, I-2U
IAAG, IAAU, I-2AU, 2C, 3-4U
20
2I
22
23
SVDV-It66 RNA
20] IAAAA(U), tAG, tAU, IAC, 2C, IU 2 I ) IAAAC, IAAG, I-2AAC, IAU, 2AC, 2C, IU 22 tAAG, IAAU, IAU, 2-3C, 3-4U
)
I6 2AAU, 2AAC, IAC, IG, 3C, 2U 17 I-2AAAC, (IAAU?)(IAAC?)2-3AC, IG, 3C, 2U 18~ IAAAC, (IAAU?)IAAC, IAG, I-2AC, 3C, 2U 19 IAAAA(C), IAAU, tAAC, IAG, IAU, 2AC, 3-4C, 2U
I4
2z) IAAG, IAAC, IAU, 1AC, 2-3C
2I) IAAG, I-2AC, I-2C, 1U
20 I-2AAAU, IAAU, IAC, IG, 3-4C, 2-3U
I9
IAAAC, IAAU, IAAC, IAG, 2AU, 2AC, IG, 4C, 4U
Spot no.
15 IAAAU, IAAU, 2AAC, IAG, IAU, 2-3AC, IG, 4-5C, 3-4U
(cont.)
I-2AAU, IAAC, IAG, IAC, 2C, 2U
SVDV-HK RNA
Table 3
17 IAAAC, IAAU, IAU, zAC, IG, 2C, 2U I8 IAAU, IAAC, IAG, IAC, 2C, 3U
Spot no.
Shared by UK and HK RNAs
HK spots I8 and I9 mixed; spot I9 probably corresponds to UK spot 20. It66 spot 20 unique Unique Unique HK spots 20 and 2I mixed but unique
HK spot I5 and It66 spot I9 mixture of 2 oligonucleotides, one of which is shared. UK spot 18 - single oligonucleotide may correspond to it. Unique Unique
Unique
Common
Unique
Oligonucleotide comparison
* The table has been arranged so that oligonucleotides moving to about the same position in each two dimensional gel (Fig. 6a to c) are compared across the table. t The molarities of the products were estimated directly from the autoradiograph and the values should only be considered as approximate. :~ See Fig. 6(fto h).
IAAAU, IAAAC, IAAU, I-2AC, IG, 3-4C, 2-3U
I9
16 IAAU, tAAC, IAG, IAC, 2-3C, 2-3U I7 2AAU, IAG, IAU, 0AC?)3-4C, 2-3U 18 I-2AAC, IAC, IG, 2-3C, 2-3U
Spot no.:~
E"
r~
2AAU, 2AC, IG, IC, ~ 3U
15
)
15 16
13 14
Io ii 12
8 9"1
}
CB5-8o68 RNA
IAAU, I-zAAC, IAU, I-zAC, IG, z-3C, IU 2-3AU, IG, 3C, 4-5U IAAAU, 1AAU, IAG, 2AU, l-zAC, 1G, 2-3C, z-3U IAAAU, IAAU, 1-2AU, 1-zAC, 1G, 2-3C, 2-3U IAAAU, IAAU, I-zAU, I-zAC, IG, 3-4C, 2-3U IAAAG, IAAC, IAU, IAC, some C, some U (IAAU?)IAAC, IAU, IAC, IG, some C, some U IAAU, IAAC, IAC, IG, some C, some U IAAAU, IAAU, IAG, I-zAC, 3-4C, zU IAAU, 2AAC, lAG, z-3AU, I-zAC, IG, 3-5C, 2-3U IAAAC, IAAG, 2AU, 2AC, 2-3C , 2-3U IAAAU, IAAU, IAG (tAU?) 1AC, some C, some U IAAAU, IAG, ~-zAC, zC, 3-4U IAAU, 2-3AU, tG, 1-2C, 3-4U
some C, some U
IAAAC, IAAG, IAAU, IAAC, I-2AU, 2AC, I-2C, 3-4U IAAAA(C orU), IAAU, I-2AAC, IAC, IG,
Oligonucleotide comparison§
©
.~
.,--t
Unique Same analysis as SVDV UK spot 9, SVDVIt66 spot I I may correspond
Unique Unique
~Z
0
.,~
Unique CB5 8068 spot 12 two oligonucleotides; Unique :Z
Unique Unique; not SVDV HK/It66 spot 4 Mixed oligonucleotides. CBs-Faulkner spot 5 and CB-8o68 spot 6 may correspond to SVDV-UK spot I3: SVDV-HK spot It and SVDV-It66 spot 14. Unique Unique; CB5-8o68 spots 9 and IO mixed
Unique
Unique
* As for Table 3 see Fig. 6(d, e). t As for Table 3. See Fig. 6 (i, ]). § As far as possible these analyses have been compared to those in Table 3, using the position in the two dimensional gels as the basis for the comparison.
I3 I4
IAAAC, IAAU, IAAC, IAC, IG, 1-2C, I-2U IAAAA(U), IAG, IAU, 2-3AC, 1-2C, I-2U IAAAU(IAAAC?), IAAU, IAAC, IAU(IAC?), IG, 2C, 2J IAAAA(C or U), IAU, IAC, IG, I-2C, I-2U IAAAC, IAAC, 3AU, IG, IC, N 3U
I0 II 12
(IAAAG?)IAAU, (IAU?)2-3C, 3U IAAG, I-2AU, IAC, some C, 2U 1AAAC, (IAAU?) IAAC, IAC, IG, I-2C, I-2U IAAC, 2AU, IAC, IG, I-2C, some U
3 4
IAAAA(C or U), IAU, 2AC, IG, 2C, 2U 2AAU, IAU, IAC, IG, IC, 3-4U IAAAU, (1AAU?)IAU, IAC, 1G, 1C, 2U
3 4
7 8 9
2
I
0AAAG?)IAAAU, IAAC, IAU, I-2C, I-2U
IAAG, 2AAC, 2AU, IAC, I-2C, I-2U
Spot no.
Ribonuclease A analysis of the long RNase T1 oligonucleotides of the CB5 virus RNAs~f
CB5-Faulkner RNA
T a b l e 4-*
2
Spot no.~
7~
t..o
Molecular aspects of S VD V-Coxsackie B5 variation
313
Since the two dimensional gels separate the oligonucleotides according to size and charge (De Wachter & Fiers, I972), it is likely that oligonucleotides moving to the same relative position in the gels will have the same composition. This has been confirmed recently in this laboratory in a study of the specific RNase T1 oligonucleotides of the RNA of a virulent and an attenuated foot-and-mouth disease virus (Harris & Brown, I977). Nevertheless to check that several oligonucleotides are shared by the SVDV RNAs and to determine those oligonucleotides that are different, we have eluted a number of the oligonucleotides from the gels and determined their composition by RNase A digestion and high voltage electrophoresis on DEAE paper at pH 3"5 (Table 3 and Table 4). The key (Fig. 6fto h) shows those oligonucleotides that are shared by the SVDV RNAs as determined by RNase A analysis (Table 3). The results support those determined visually; there are at least 12 oligonucleotides that are shared by all three SVDV RNAs, 3 oligonucleotides shared only by SVDV-UK and SVDV-HK RNAs, 2 oligonucleotides shared by SVDV-UK and SVDV-It66 and I oligonucleotide shared by SVDV-HK and SVDV-It66 RNAs. A similar analysis of the CB5 RNAs (Table 4) demonstrated that the majority of these oligonucleotides were unique. There were, however, two CB5-8o68 oligonucleotides that had a similar analysis to SVDV oligonucleotides (Table 4, Fig. 6 i, i). As these CB5-8o68 oligonucleotides moved to about the same position in the gels as the corresponding SVDV oligonucleotides, it is possible that they are shared. One of these oligonucleotides may also be present in CB5-Faulkner RNA (see Table 4)- This oligonucleotide may therefore be common to SVDV and CB5 RNAs. DISCUSSION
Considerable variation in the homology of SVDV and CB5 RNAs has been obtained by different hybridization methods. The initial saturation experiments of Brown & Wild (t 974) and Brown et aL (1976) using alkali-treated virus RNAs and radioactive ds RNA showed that SVDV-UK RNA was 5o ~ homologous to the RNA of CB5-FauJkner and CB5-8068. The competition hybridization experiments with a2P-SVDV-UK RNA done under the more stringent conditions used in the present experiments showed the same level of annealing (Fig. 3a). However, competition experiments with 32P-CB5 RNAs gave consistently lower values (30 to 35 ~, Fig. 3 c and Table 2). The saturation hybridization experiments done under the conditions used for constructing the competition curves (i.e. 50 ~o formamide, 5 × SSC, 50 °C for I6 h) also gave a lower estimate of the homology of SVDV-UK and the CB5 RNAs (Fig. 2). The results of the thermal denaturation experiments may help to explain the variability of the competition reactions since they demonstrate clearly that the heterologous SVDV/CB5 RNA hybrids are extensively mismatched. It is still not clear, however, why there is a discrepancy between the saturation curves (Fig. 4) and the competition curves (Fig. 3). Presumably some sequences which will not hybridize completely are still similar enough to sequences in the homologous 32P-RNA to prevent them from annealing to complementary RNA. The lack of variability in the hybridization of the more closely related SVDV RNAs, measured by the three different methods, supports the above conclusions. In addition, the thermal denaturation experiments (Fig. 5) show that the SVDV-UK/SVDV-It66 RNA hybrid is not as well matched as the SVDV-UK/SVDV-HK RNA hybrid, despite the fact that SVDV-UK and SVDV-It66 RNAs are apparently more homologous than SVDV-UK and SVDV-HK RNAs (9° ~ compared to 85 ~o, see Table I). This result is confirmed by the fingerprints (Fig. 6 a to c) which show that SVDV-UK RNA shares more oligonucleotides with SVDV-HK RNA than with SVDV-It66 RNA. In view of this it is interesting that 21
V I R 35
314
T . J . R . HARRIS~ T. R. DOEL AND F. BROWN
preliminary experiments using radioimmune assay to measure the antigenic variation of the SVD viruses have shown that SVDV-UK is more closely related to SVDV-HK than to SVDV-It66 (J. R. Crowther, personal communication). The results of the 'mixed competitions' using equal amounts of two RNAs as competitor showed that the homologous regions shared by the CB5 virus RNAs were largely the same as those shared by the CB5 virus RNAs and SVDV-UK (Table 2). The substantial sequence homology of the three SVDV RNAs was also shown to be shared (Table 0. These hybridization results are similar to those obtained by Young, Hoyer & Martin (I968) for the three serotypes of poliovirus. The results also imply that the three SVDV viruses are members of a closely related group whereas the CB5 viruses are not, since CB 5 Faulkner R N A and CB5-8o68~RNA are as different from each other as they are from SVDV-UK RNA. It has been suggested from measurements of the sizes of the heteroduplexes of the RNAs of the poliovirus serotypes that the common sequences are not continuous but are located at different positions along the molecule (Young, I973). The conclusions from the fingerprinting of the RNase T1 oligonucleotides of the virus RNAs studied here have some bearing on this observation. Our results (Fig. 6, Table 3 and Table 4) support the hybridization and Tm data given above. The sharing of several long oligonucleotides by the SVDV RNAs (Fig. 6 a to c) suggests that there are continuous regions of common sequence and the lack of long oligonucleotides common to both the CB5 and SVDV RNAs suggests that the limited homologous regions between these viruses are randomly distributed along the genome. The location of both the common and the different oligonucleotides in the RNAs of the antigenically distinct SVD viruses is currently being investigated by orientating the long oligonucleotides relative to the poly (A) tract at the 3'-end, as has been done recently for avian RNA tumour viruses (Wang et al. I975; Coffin & Billeter, I976 ) and foot-and-mouth disease virus R N A (Harris & Brown, I976). This approach, coupled with annealing reactions using the 5'-end of the R N A (the region coding for the structural polypeptides and hence the antigenicity) should enable us to locate those parts of the molecule responsible for the antigenic variation. Although the polypeptides of the SVD and CB5 virus isolates examined in this paper have different mobilities in discontinuous SDS polyacrylamide gels (Harris & Brown, I975) the tryptic peptide maps of the unfractionated virus polypeptides show that there is considerable similarity in the methionine containing tryptic peptides. If, as seems likely, the peptides that do not contain methionine are also similar then the results suggest that there are only limited differences in the primary structures of the polypeptides of SVDV and CB5. By examining the tryptic peptides of the individual polypeptides of antigenically distinct viruses it should be possible to ascertain the extent of the sequence conservation and to correlate for the SVD viruses changes in the individual polypeptides with the antigenic differences. Further detailed analysis of the RNAs and polypeptides of SVDV and CB5 should give some insight into the changes which occur during antigenic drift in this group of viruses.
REFERENCES
&BURROWS,R. (1973). Antigenic differencesbetween isolates of swine vesicular disease and their relationship to Coxsackie B5 virus. Nature, London 245, 3t5-316. BROWN,F. &WILD,F. (1974). Variation in the Coxsackie virus type B5 and its possible role in the etiology of swine vesicular disease. Intervirology 3, I25-I28. BROWN, F., WILD, T.F., ROWE, L.W., UNDERWOOD, B.O. & HARRIS, T..1. R. (i976). Comparison of swine vesicular disease virus and Coxsackie B5 virus by serological and RNA hybridization methods. Journal of General Virology 3x, 231-237. BROWN, F, TALBOT, P.
Molecular aspects of S VD V-Coxsackie B5 variation
315
cornN, J.M. & ~rL~.ETER, M.A. (1976). A physical map of the Rous sarcoma virus genome. Journal of Molecular Biology Too, 293-318. CRESTFIELD, A. M., MOORE, S. & STEIN, W. It. 0963)- The preparation and enzymatic hydrolysis of reduced and S-carboxymethylated proteins. Journal of Biological Chemistry 238, 622-627. DARaY, C. & MINSON,A. C. (1973). The structure of tobacco rattle virus ribonucleic acids: common nucleotide sequences in the R N A species. Journal of General Virology 2x, 285-295. DE WACHTER, R. & VIERS, W. (r972). Preparative two-dimensional poIyacrylamide gel electrophoresis of 32Plabelled RNA. Analytical Biochemistry 49, 184-197. FRISBY, D. P., NEWTON, C., CAREY, N. H., FELLNER, P., NEWMAN, J. F. E., HARRIS, T. J. R. & BROWN, F. (i976). Oligonueleotide mapping of picornavirus R N A s by two-dimensional electrophoresis. Virology 7x, 379388. GRAVES,J. H. (1973). Serological relationship of swine vesicular disease virus and Coxsackie B5 virus. Nature, London 245, 314-315. HARRIS, W. J. R. & BROWN, F. (1975). Correlation of polypeptide composition with antigenic variation in the swine vesicular disease and Coxsackie B5 viruses. Nature, London 258, 758-76o. HARRIS, T. S. R. & BROWN, F. (I976). The location of the poly(C) tract in the R N A of foot and mouth disease virus. Journal of General Virology 33, 493-5oi. HARRIS, X. J. R. & BROWN, F. (1977). Biochemical analysis of a virulent and an avirulent strain of foot and mouth disease virus. Journal of General Virology 34, 87-Io5. PORTER, a., CAREY, N. H. & FELLNER, P. (1974). Presence of a large poly(C) tract within the R N A of encephalomyocarditis virus. Nature, London 248, 675-678. SARGENT, J. a. & VADLAMUDI,B. P. (I968). Electrophoresis of peptides on thin layers of silica gel. Analytical Biochemistry 25, 583-587. WANG, L. H., DUESBERG, P , BEEMON, K. & VOGT, P. K. (1975). Mapping RNase Trresistant oligonucleotides of avian tumor virus RNAs: sarcoma specific oligonucleotides are near the poly (A) end and oligonucleotides common to sarcoma and transformation defective viruses are at the poly(A) end. Journal of Virology x6, lO5I-IO7O. YOUNO, N. A. (1973). Size of the gene sequences shared by polioviruses types I, 2 and 3. Virology56, 4oo-403. YOtrNo, N. A., HOYER, B. H. & MARTIN, M. A. (1968). Polynucleotide sequence homologies among polioviruses. Proceedings of the National Academy of Sciences of the United States of America 6x, 548-555.
(Received 25 October I 9 7 6 )
21-2
299
Printed in Great Britain
Molecular Aspects o f the Antigenic Variation of Swine Vesicular Disease and Coxsackie B5 Viruses By T. J. R. H A R R I S ,
T. R. D O E L
AND F. B R O W N
Department of Biochemistry, Animal Virus Research Institute, Pirbright, Woking, Surrey, U.K. (Accepted 7 December 1976) SUMMARY
The antigenic variation of swine vesicular disease virus (SVDV) and Coxsackie B5 virus (CB5) has been examined at the molecular level by analysing the protein and nucleic acid of the virus particles. The tryptic peptides of carboxymethylated 35S-methionine labelled virus particles were very similar, although some minor differences were apparent. Competition hybridization experiments confirmed that there is variation in the RNA sequence of antigenically distinct SVD viruses and some limited homology of these RNAs to the CB5 virus RNAs. Competition hybridization using mixtures of two RNAs as competitors showed that the sequences shared by the CB5 virus RNAs were largely the same as those shared by the CB5 virus RNAs and the SVDV RNAs. Similar experiments with the SVDV RNAs established that the homologous regions in these RNAs were also shared. Thermal denaturation curves of SVDV RNA-RNA hybrids generally supported the competition hybridization results but the hybrids between SVDV RNA and CB5 virus RNA were shown to be mismatched. Ribonuclease Tx oligonucleotide maps of the virus RNAs were also compared to obtain another measure of relatedness. A number of long oligonucleotides were shared by the SVDV RNAs but few of these were found in either of the CB5 RNAs. Further ways of investigating the antigenic variation of SVDV at the molecular level are discussed.
INTRODUCTION
Although swine vesicular disease virus (SVDV) is antigenically related to Coxsackie B5 virus (CB5; Graves, i973) the viruses can be distinguished by cross neutralization and immunodiffusion tests (Brown, Talbot & Burrows, 1973; Brown et al. 1976). There is also antigenic variation within the two virus types which can be detected by immunodiffusion. For example, the United Kingdom isolate of SVDV (SVDV-UK) gives a line of partial identity with a Hong Kong isolate (SVDV-HK) and the prototype strain of CB5 (CB5Faulkner) gives a line of partial identity with a recent isolate CB5-8o68 (Brown et al. I976 ). Evidence for this variation has also been obtained from analysis of the virus structural polypeptides on SDS-polyacrylamide slab gels, since the polypeptides of anfigenically distinct virus isolates have slightly different mobilities (Harris & Brown, 1975). The virus RNAs which code for these polypeptides have also been studied. Using radioactively labelled double stranded RNA as a source of complementary RNA, Brown & Wild (I974) showed that about 5o ~ of the sequence of SVDV-UK RNA was shared by the RNAs of CB5Faulkner and CB5-8o68. In an extension of these studies we showed that SVDV-UK R N A 20-2
300
T.J.R.
H A R R I S , T. R. D O E L A N D F. B R O W N
was 80 ~ homologous to the RNA of SVDV-HK and 90 ~ homologous to the RNA of the SVDV (SVDV-It66) isolated in Italy in 1966 (Brown et al. I976 ). The RNAs of SVDVUK, CB5-Faulkner and CB5-8o68 have also been examined by specific nuclease digestion (Frisby et al. I976). Few, if any, long oligonucleotides produced by ribonuclease T1 (RNase TO digestion were shared. In view of the possible role of the human CB5 viruses in the etiology of swine vesicular disease, it was clearly important to characterize the antigenic variation in more detail by biochemical methods. This paper describes the analysis of the structural proteins of antigenically distinct SVDV and CB5 isolates by peptide mapping techniques and the examination of the sequence variation of the RNAs using more precise competition hybridization and oligonucleotide fingerprinting techniques.
METHODS
Viruses. The SVD and CB5 viruses were grown at 37 °C in Eagle's medium in monolayers of IBRS-2 cells in Roux bottles. The following SVDV isolates were used: United Kingdom 27[72 (SVDV-UK); Hong Kong 3617t (SVDV-HK) and Italy 1/66 (SVDV-It66). The CB5 isolates were the prototype strain CB5-Faulkner and a CB5 isolated in I973 (CB5-8o68). These viruses were passaged only twice in IBRS-z cells to retain as closely as possible the characteristics of the original isolates. Poliovirus (type ~, Mahoney) was grown in monolayers of HeLa cells at 37 °C. Preparation of radioactively labelled viruses. Monolayers of about 2 × IO7 IBRS-2 cells in Roux bottles were infected with virus at high multiplicity in 2o ml of Eagle's medium. After an adsorption period of 30 min, the monolayers were washed with the appropriate labelling medium, 20 ml of fresh medium added and the cells incubated at 37 °C. Radioactive label was added to the cultures I-5 to 2 h after infection. For 35S-methionine labelling Ioo #Ci of 35S-methionine (ioo Ci/mmol) in 2o ml of methionine-free Eagle's medium was used. Virus RNA was labelled to high specific activity with 3zp by adding 25o to 5oo #Ci/ml carrier free sZP-orthophosphate in 2o ml of phosphate-free Earle's saline. All isotopes were from the Radiochemical Centre, Amersham. Purification of virus. Radioactively labelled virus was purified from the supernatant medium at the end of the growth cycle when the cells had left the glass. The cellular debris was removed by centrifuging at 4ooog for I5 min at Io °C and the medium clarified further by centrifuging at ioooo g for 20 min at Io °C (Ioooo rev/min, MSE rotor 59595). Virus was then pelleted by centrifuging for 1-25 h at 8oooo g in the same rotor (3oooo rev/min) at Io °C. The pellet was resuspended in I"5 ml of o.I M-NaCI, 0"o5 M-tris-HCl, pH 7"6, for at least I6 h, SDS added to I ~ (w]v), and the virus suspension layered on to a sucrose gradient (I5 ~ to 45 ~ [w/v] in o.I M-NaC1, o'o5 M-tris-HC1, pH 7"6) which was centrifuged at 8oooo g for 2"25 h at Io °C (3oooo rev/min, MSE rotor 5959o). The gradient was collected in I ml fractions into bijou bottles containing 25/zg of bovine serum albumin, and the fractions containing labelled virus were pooled. The purity of the virus was judged by SDS polyacrylamide gel electrophoresis of the polypeptides. Unlabelled virus (3oo to 4oo ml) grown in 15 to 2o Roux bottles was first precipitated with 5o ~o (v/v) saturated ammonium sulphate buffered by o-o4 M-phosphate, pH 7"6, and the precipitate resuspended in 4o ml of o.o4 M-phosphate. After clarification at 1oooo rev/min, virus was pelleted from this suspension by centrifuging at 30ooo rev/min as described for radioactive virus. Sucrose gradients of unlabelled virus were monitored by measuring the extinction at 260 nm.
Molecular aspects of S VD V-Coxsackie B 5 variation
30 t
Trypsin digestion. Sucrose gradient fractions containing virus were diluted twofold and the virus particles precipitated with 2 vol. acetone overnight at - 2 o °C in the presence of I5O #g of bovine serum albumin. Virus was carboxymethylated in 8.o M-urea by the method of Crestfield, Moore & Stein (1963), modified by inclusion ofo'5 ~ (v/v) SDS and substitution of fl-mercaptoethanol with o.o6 M-dithiothreitol. The carboxymethylated protein was recovered after this procedure by acetone precipitation and washed by repeated acetone precipitation. The dried pellets were dissolved in z ml of o.I M-NH4HCO3, pH 8.o and digested with I ~ (w/w) TPCK trypsin (Worthington Biochemical Corporation, Freehold, New Jersey) at 37 °C for 24 h. A further I ~ (w[w) trypsin was added after 4 h. The digests were filtered through nitrocellulose membranes (25 m/t, Millipore) and evaporated to dryness in a stream of N 2 at 4o °C. The residues were dissolved in io to 25 #1 of o.z MNH4OH for tryptic peptide analysis. Analysis of tryptic peptides. Tryptic peptides were separated by electrophoresis and chromatography on 2o × zo cm silica gel thin layer sheets (Machery-Nagel, SIL G) as described by Sargent & Vadlamudi (1968) except that electrophoresis was continued at 2oo V for 6 h. This method has the advantage that two samples can be subjected to electrophoresis at the same time under identical conditions. For ascending chromatography the sheets were cut in half with scissors and both halves run for 2"5 h in the same rectangular tank containing H20: acetic acid:butan-I-ol (1 : I : 3). Autoradiographs were prepared by exposing Kodak RP Royal X-Omat X-ray film to the dried sheets for 5 to Io days. Extraction of RNA from virus particles. The sucrose gradient fractions containing virus were diluted with o'i5 M-NaC1, o'o5 M-tris-HC1, o'oo5 M-EDTA, pH 7"6 (TNE buffer), containing o.I ~ SDS, Ioo #g of E. coli tRNA added, extracted twice with a I ; I mixture of phenol: chloroform and the RNA precipitated with 2 vol. ethanol overnight at - 2o °C. The RNA was re-precipitated once with ethanol before use. Extraction of double stranded RNA from infected cells. Double stranded RNA (ds RNA) was extracted and purified from about 5 × Io8 virus-infected cells by phenol extraction and 2 M-LiC1 precipitation. The ds RNA in the LiC1 soluble fraction was purified further from low mol. wt. single stranded RNA (ss RNA) by centrifuging on I5 ml 5 to 25 ~ (w/v) sucrose gradients in o.i M-sodium acetate, o.I ~ SDS, pH 5"o, for 16 h at 8oooog (MSE rotor 591o8) at 2o °C. The ds RNA was recovered from the gradient by ethanol precipitation. These procedures have been described in detail previously (Harris & Brown, i977). The precipitates were dissolved in 5o ~ formamide, 5 × SSC (o'75 M-NaCI, o'75 M-Nacitrate) to give concentrations of ds RNA of approx. 5 to xo/zg/ml based on the extinction of the sucrose gradient fractions at 260 nm. Molecular hybridization. Saturation and competition hybridization experiments were done in duplicate in o.2 ml of 5o ~ formamide, 5 × SSC at 50 °C exactly as described by Harris & Brown (I977). The competitor RNAs were dissolved in 5o ~ formamide, 5 × SSC to give a concentration of 8 to Io/zg/ml. The results obtained with the hybridization procedures were handled as described in detail by Darby & Minson (I973). Briefly, the reciprocal plots of the competition hybridizations give intercepts on the I[f axis which allow determination of f, the fraction of RNA displaced by the competitor, at infinite competitor RNA concentration. The reciprocal plots for the saturation hybridizations give intercepts on the I[F axis which allow the determination of F, the fraction of labelled RNA homologous to the ds RNA, at infinite ds RNA concentration. Thermal denaturation profiles. These were obtained as described by Harris & Brown (I977). Briefly, labelled homologous and heterologous hybrids were prepared by annealing
302
T.J.R. HARRIS, T. R. DOEL AND F. BROWN
3ZP-virus RNAs to denatured ds RNA at 50 °C in o'o5 ml of 50 7o formamide, 5 x SSC. The samples were diluted to I ml with de-ionized H20, heated to various temperatures, cooled quickly and the fraction of label remaining in the hybrid determined from the residual ribonuclease resistance. Polyacrylamide gel electrophoresis of the RNase Ti-resistant oligonucleotides. Virus RNAs, labelled with 32p, were digested for I h at 37 °C with RNase T 1 (SANKYO, Japan) in IO/~l of o.oI M-tris-HCl, o.ooi M-EDTA, pH 7"4, at an enzyme: substrate ratio of 1:2o. The oligonucleotides produced were separated by two dimensional gel electrophoresis (De Wachter & Fiers, I972), using the buffer modification of Frisby et aL (1976) and the gel dimensions and running conditions of Harris & Brown (I977). Autoradiographs were prepared using Kodak RP-Royal X-Omat X ray film. Analysis of oligonucleotides .from polyacrylamide gels. The radioactive oligonucleotides were eluted from the gels as described by Porter, Carey & Fellner (I974), digested with ribonuclease A (RNase A, Worthington) in o-oi M-tris-HCl, o-ool M-EDTA at an enzyme: substrate ratio of I : 20 and the products fractionated by high voltage ionophoresis on DEAE paper (Whatman DE 81) in 0"5 7o pyridine, 5 ~ acetic acid, pH 3"5. The identity of the products was determined from their mobilities relative to markers. RESULTS
Tryptic peptide analysis of virus particles Although the polypeptides of antigenically distinct isolates of SVDV and CB5 can be distinguished by polyacrylamide gel electrophoresis (Harris & Brown, I975), it is not possible to decide from this observation to what extent the amino acid sequence has varied. To investigate this point sSS-methionine labelled virus particles of different SVDV and CB5 isolates were digested with trypsin and the tryptic peptides analysed by electrophoresis and chromatography on thin layer silica gel sheets (Sargent & Vadlamudi, I968). Fig. I (a to e) shows maps of the tryptic peptides of the isolates prepared by autoradiography of the thin layer plates. All the isolates have the same overall pattern of tryptic peptides and the maps of the SVDV isolates (Fig. I a to c) are more similar to each other than to the CB5 isolates (Fig. I d, e). There are, however, some distinct differences between the maps of the individual SVD viruses and the CB 5 viruses. As these maps simply represent the tryptic peptides of virus particles, it is impossible to say from which of the four polypeptides these differences arise. The tryptic peptide map of 35S-methionine labelled poliovirus prepared in the same way (Fig. If) bears little resemblance to any of the SVDV and CB5 maps. Tryptic peptide maps of 35S-methionine labelled SVD and CB 5 virus particles have also been prepared by high voltage paper electrophoresis and chromatography. The results confirmed the similarity of the maps of SVDV and CB5 determined by the thin layer method (T. R. Doel, unpublished results).
Competition hybridization of the virus RNAs Saturation hybridization of SVDV and CB5 RNAs has shown that they have about 5o homology (Brown & Wild, 1974; Brown et al. I976). More precise estimates of homology can be obtained, however, by competition hybridization. In the method we have used, 3~P-ss RNA was prevented from annealing to unlabelled complementary RNA (denatured ds RNA) by the addition of increasing amounts of unlabelled ssRNAs from the heterologous viruses. The homology of the RNAs of SVDV-UK, SVDV-HK and SVDV-It66 were
Fig. t. Autoradiographs of the tryptic peptides of various virus particles separated in two dimensions by eleclrophoresis and chromatography on thin layer silica gel plates. (a) SVDV-UK, (b) SVDV-HK, (c) SVDV-It66, (d) CB5-Faulkner, (e) CB5-8o68, (f) poliovirus. The letter (o) marks the origin of the plates. The direction of chromatography is shown by the large arrow.
O
E"
304
T . J . R . H A R R I S , T. R. DOEL A N D F. B R O W N I
i
I
I
I
I
(a)
I
I
(b)
"•4"0 ~5 3-0 2.0
.~ 1"0 I
0.05
I
0.10
I
I
I
I
0-15 0.20 0.025 Competitor RNA (I/c)/zl
i
0-050
0.075
i
0.100
- t
Fig. 2. Competition hybridization of SVDV-UK RNA. Reciprocal plots of the annealing of 3~p. SVDV-UK ss RNA to denatured SVDV-UK ds RNA in the presence of increasing amounts of (a) SVDV-UK RNA (0--0); SVDV-HK RNA (©--©), SVDV-It66 RNA ( I - - I ) a n d (b) a 5o:5o mixture of both SVDV-HK and SVDV-It66 RNAs ( I - - I ) . Table i. RNA sequence homology of SVDV RNAs determined by
competition hybridization 3~P-ss RNA and unlabelled ds RNA r
Unlabelled competing RNAs SVDV-UK SVDV-HK SVDV-It66 SVDV-HK + SVDV-It66 SVDV-UK + SVDV-HK
*SVDV-UK
SVDV-HK
SVDV-It66
Ioo 86 91 91 --
85 Ioo 80
87 83 ioo
-
-
--
-
-
90
* These values are the means of two determinations, including that shown in Fig. z. first examined by this method. S V D V - U K RNA, labelled with 32p, was annealed to SVDVU K ds R N A in the presence of S V D V - H K and SVDV-It66 virus RNAs. The results, which also show the control homologous competition using SVDV-UK RNA, give I / f intercepts of 1.2o and I,I 3 for S V D V - H K and SVDV-It66 R N A (Fig. 2a), corresponding to a sequence homology with S V D V - U K R N A of 83 ~o and 89 ~ respectively. These values are in good agreement with those obtained previously using radioactive ds R N A (Brown et aL I976 ). A competition curve was also constructed using increasing amounts of a 5o: 5o mixture of S V D V - H K and SVDV-It66 R N A s as competitor. The intercept of I.I2 (Fig. 2b) corresponds to a homology of 89 to 90 ~ and is within the error of the value obtained with SVDV-It66 R N A alone as competitor (Fig. 2 a), demonstrating that the regions of homology are not additive. Thus, the sequences in S V D V - H K and SVDV-It66 R N A s that are homologous to S V D V - U K R N A are shared. Corresponding competition experiments have been done by annealing 32p-SVDV-HK R N A and 32p-SVDV-It66 to their complementary RNAs in the presence of unlabelled heterologous SVDV RNAs. The results of the reciprocal plots of these data (Table I) confirm those in Fig. 2 (a). To examine the relationship of SVDV R N A s to the RNAs of CB5-Faulkner and CB58o68, 32p-SVDV-UK R N A was annealed to denatured homologous ds R N ~ in the presence
Molecular aspects of SVD V-Coxsackie B5 variation I
I
I
I
8-0 - (a)
305
I
I
I
I
I
I
I
I
- (b)
6"0
o
•
4"0 "~
2-0
F. e~
I
e-
~16.0 O
I
I
I
(c)
_. (a) o
.~ 12.0 8.0 4"0 I
I
0"05
0-10
I
!
l
0"15 0.20 0"025 Competitor RNA (l/c)/11-1
t
0-050
I
0-075
I
0.100
Fig. 3. Competition hybridization of SVDV-UK and CB5-Faulkner RNA. Reciprocal plots of: 0 ) 3~P-SVDV-UK R N A annealed to denatured ds R N A in the presence of increasing amounts of (a) SVDV-UK R N A (O---O), CB5-Faulkner R N A ( O - - O ) , CB5-8o68 R_NA ( l l - - i ) ; ~,(b) a 50:50 mixture of CB5-Faulkner and CB5-8o68 RNA. (2) 3~P-CB5-Faulkner R N A annealed to denatured ds R N A in the presence of increasing amounts of (c) CB5-Faulkner R N A (O---O), SVDV-UK R N A ( O - - O ) , CB5-8o68 R N A ( i - - i ) ; (d)a 50:50 mixture of SVDV-UK R N A and CB5-8o68 RlqA.
Table z. RNA sequence homology of SVDV-RNA and CB5 RNA determined
by competition hybridization azP-ss R N A and unlabelled ds R N A t
Unlabelled competing ss RNAs SVDV-UK CB5-Faulkner CB5-8o68 CB5-Faulkner + CB5-8o68 SVDV-UK + CB5-8o68 SVDV-UK + CB5-Faulkner
*SVDV-UK
*CB5-Faulkner
CB5-8o68
! oo 45 53 54 ---
31 ioo 34 -39 --
35 33 ioo --43
* These values are taken from the intercepts shown in Fig. 3.
of increasing amounts of unlabelled SVDV-UK, CB5-Faulkner and CB5-8o68 virus RNAs. Fig. 3 (a) shows the reciprocal plot of these hybridization results. CB5 Faulkner RNA gives a I/f intercept of 2-20 and CB5-8o68 an intercept of 1.9o, corresponding to sequence homologies with SVDV-UK RNA of 45 ~ and 53 ~ respectively. These values are in agreement with those found by Brown & Wild (i974). The result of the mixed competition curve using a 50:50 mixture of both CB 5 RNAs as competitors (Fig. 3b) gives an intercept of ]'85 (corresponding to a sequence homology of 54 ~o)- The non-additive sequence homology
T. J. R. H A R R I S , T. R. D O E L A N D
306 I
I
I
I
I
"< 0-6 z 0.4 O
I
I
I
(b)
(a]
~0"8
I
F. B R O W N
16'0
/
12"0
o
8"0 4"0
•~ 0.2 I
10
I
I
I
20 30 40 ~1 dsRNA(d)
I
50
I
0-05
I
0-10 1/d
I
0-15
I
0.20
Fig. 4. Saturationhybridizationof SVDV and CB5 RNA. (a) Increasingamounts of denatured ds RNA from SVDV-UK infectedcells added to 3~P-SVDV-UK RNA (O--O), 3~P-CB5-Faulkner RNA (G--Q), 3~P-CB5-8o68RNA ( I - - I ) . (b) Reciprocalplots of the data. indicates that CB5-Faulkner RNA and CB5-8o68 RNA have similar sequences homologous to SVDV-UK RNA. CB5 Faulkner RNA labelled with 32p has also been annealed to denatured CB5-Faulkner ds RNA in the presence of unlabelled competitor RNAs. The results in Fig. 3 (c) show intercepts of 3.20 and 3"oo for SVDV-HK and CB5-8o68 RNA, corresponding to homologies of 31 ~ and 34 ~ respectively. In another experiment in which SVDV-HK and SVDV-It66 RNAs were used instead of SVDV-UK RNA, the results were similar to those with the SVDV-UK RNA. Fig. 3 (d) shows the result of the mixed competition with a 5o: 50 mixture of SVDV-UK and CB5-8o68 RNAs as competitor The intercept of 2.6o (39 homology) shows that the sequence homology of SVDV-UK, CB5-Faulkner and CB5-8o68 RNA is non-additive, again indicating that the homologous sequences in these three RNAs are shared. Table 2 shows the homology of SVDV RNA and CB 5 RNA taken from the intercepts in Fig. 3 and from the reciprocal plot of an additional confirmatory competition experiment using 32P-CB5-8o68 RNA and unlabelled CB5-8o68 ds RNA. Saturation hybridization and thermal denaturation of the RNA hybrids CB 5 RNAs, labelled with 32p, were added to increasing amounts of denatured SVDV-UK ds RNA and the level of annealing assessed from the residual ribonuclease resistance. Fig. 4 shows normal and reciprocal plots of the data. The intercepts of 5.zo and 4'8o in the reciprocal plots indicate a homology of I9 ~ and 2I ~ respectively for CB5-Faulkner and CB5-8o68 RNA. Similar results were obtained with CB5-Faulkner ds RNA. Thus, the estimate of sequence homology of SVDV-UK and CB5 RNAs measured by this method is even less than that measured by competition hybridization (Fig. 3, Table 2), or by saturation hybridization using denatured radioactive ds RNA (Brown et al. I976). In view of this discrepancy, RNA-RNA hybrids prepared by similar saturation hybridization experiments with denatured SVDV-UK ds RNA were subjected to thermal denaturation. The results in Fig. 5 (a) show that the CB5/SVDV RNA hybrids are much more mismatched than is the homologous SVDV-UK hybrid, since a Tm of 9o °C was obtained for the homologous hybrid compared with much lower values for SVDV-UK[CB5-Faulkner and SVDV-UK/ CB5-8o68 RNA hybrids. The melting curves of heterologous SVDV RNA hybrids were also examined. Denatured SVDV UK ds RNA was added to 32P-SVDV ss RNAs and the resulting hybrids subjected
Molecular aspects o f S VDV-Coxsackie B5 variation I
I
I
70
80
90
(b)
I
I
307 T
0.2
.E 0"4 < z =;
~ 0-6 0
0.8 ,
,
70 1O0 Temperature (°C)
80
!
I
90
100
Fig. 5. Thermal denaturation of hybrids prepared by annealing denatured SVDV-UK ds RNA to 32P-ss virus Rl'qAs. (a) SVDV-UK RNA (~--O), CB5-Faulkner RNA ( O - - O ) , CB5-8o68 RNA (D--D) and (b) SVDV-UK RNA (O-O), SVDV-HK RNA ( O - - O ) and SVDV-It66 RNA
(m--m).
to thermal denaturation. The saturation curves used to monitor the hybridizations are not shown, but in this case the results agreed closely with the competition hybridization data (Fig. 2, Table 0. Fig 5 (b) shows the thermal denaturation profile of the SVDV-UK RNA hybrids. The Tms of the heterologous hybrids (SVDV-HK RNA, 88"5 °C; SVDV-It66 RNA, 86 °C) indicate that these duplexes are less well matched than the homologous hybrid (Tm = 90 °C; Fig 5a and b) and that the SVDV-It66/SVDV-UK RNA duplex is more mismatched than the SVDV-HK/SVDV-UK RNA hybrid.
Oligonucleotide fingerprints of the RNAs Few of the long oligonucleotides produced by RNase T z digestion of SVDV-UK RNA were found in similar digests of CB5-Faulkner and CB5-8o68 RNA (Frisby et al. T976). In view of the limited homology of these three virus RNAs (30 to 50 ~, Table 2) and the extensive mismatch of the hybrids (Fig. 5 a), this result is not unexpected. It was of considerable interest, however, to compare the characteristic RNase T1 oligonucleotides of the more closely related SVDV RNAs and to contrast them with the RNase T1 oligonucleotides of CB5-Faulkner and CB5-8o68 RNA. asP-labelled SVDV and CB 5 RNAs were extracted from purified virus particles and the RNase T1 oligonucleotides of each RNA separated by two dimensional gel electrophoresis (De Wachter & Fiers, I972). The fingerprints obtained by autoradiography of the gels are shown in Fig. 6(a to e), together with a key to the top half of the autoradiographs showing the long oligonucleotides (Fig. 6 f to j). The patterns for SVDV-UK RNA and the two CB5 RNAs (Fig. 6a, d, e), correspond to those obtained by Frisby et al. 0976) confirming the reproducibility of the method. There is a striking similarity in the position of a number of the long RNase Tz oligonucleotides of the SVDV RNAs (Fig. 6a to c, f t o h), but few, if any, similar long oligonucleotides were found in the CB5-Faulkner and CB5-8o68 fingerprints (Fig. 6d, e, i,/). Fig. 6 also shows that SVDV-UK RNA has more of its oligonucleotides in common with SVDV-HK RNA than with SVDV-It66 RNA.
308
T . J . R . HARRIS~ T. R.r'DOEL A N D F. B R O W N 1st 2nd
)
6 ,O
Fig. 6. (a to e) Autoradiographs of two dimensional polyacrylamide gels of the RNase T1 oligonucleotides of 32P-SVDV and CB5 RNAs. The large arrows beside (a) indicate the directions of electrophoresis.
Molecular aspects of SVD V-Coxsackie B5 variation UK
(f)
309
(g)
IlK
\ 1 O
1 2 OO
~
34 C~~5 ~ 61514
0
18 1716
22019
0 4) 5 10
8 7 .6 12 16 • ~B' 1 3 ~ 1~(~ T7 15
1~23
(/,)
It66
14
9
~'"
Faulkner
(0
1
1 0 3
0 2 o
2 0
4 ©
%% 22
Io0008 "/~pO 1615 12
g
4
3 C)
9 "~ j7
10
f- ,l
0
6 5 0
O
0
23
(/)
8068
0
©2
3
g
o
.o~ 1~3~
g g
Fig. 6. (fto]) Enlarged tracings of the top halves of the autoradiographs shown in (a to e), to provide a key for comparison. The numbers refer to those oligonucleotides analysed further by RNase A digestion (Tables 3 and 4). Oligonucleotides marked • are those shared by all 3 SVDV RNAs; O, shared by SVDV-UK and SVDV-HK RNAs; 0, shared by SVDV-UK and SVDV-It66 RNAs @, shared by SVDT-HK and SVDV-It66 RNAs (see Table 3). The shaded oligonucleotides in the keys to CBs-Faulkner and CB5-8o68 RNAs (i,]) are those possibly present in the SVDV RNAs (see Table 4). The streak is the heterogeneous poly(A) tract (Frisby et aL z976).
SVDV-UK RNA
Table 3.* Spot no. SVDV-HK RNA
IAAU, 2-3AU, IG, 2-3C, 3"4U
I5
I4
2-3U IAAAC, IAU, IAC, tG, IAAAU, 2-3U (IAAAU?), IAAAC, IAAC, 2AU, IAC, IG, 2C, 3U
2C, 3U IAAAC, 0AAU?)0AAC?) 12t IAAAU, IAU, IAC, I(3",2C, 3U I3 (IAAAU?) IAAAC, IAAC, 2AU, IAC, IG, 2C, 2U
:6
(IAAAU?.):AAAC(~AAU2) :AAC, 2AU, IAC, IG, I-2C, 2-3U
zU
14] I AAAUOAAAC?), 2AAU0 AAC ?) 2AU, IAC, IG, 2-3C, 3U 15 IAAU, IAAAC, IAU, IAC, IG, 2C,
I2 IAAAA(U), IAG, IAC, 2-3C , 3-4U 13 IAAU, IAAC, I-2AU, (tAC?), IG, 2C, 3U
9 IAAG, IAAU, z-3AAC, IAU, IAC, 2C, 2U IO IAAAC, IAAC, IAU, IAC, IG, I-2C, I-2U II IAAU, 2AU, 0AC?), :G, 2C, ~ 4U
1-2U
5 IAAU, IAAC, IAU, 2AC, IG, IC, 2-3U 6 I-2AAAC, IAAU, 2AAC, IAU, 2-3AC, IG, 2C, 2-3U 7 tAAG, IAAU, 2AAC, tAG, 2-3AU, 2-3AC, 3C, 3U 8 IAAAA(C), lAG, IAU, IAC, I-2C,
5 I-2AAU, I-2AAC, IAU, 2AC, IG, I-2C, 2-3U 6 t-2AAAC, IAAU, 2AAC, IAU, 2-3 AC, IG, I-2C, 2-3U 7 IAAG, IAAU, I-2AAC, IAG, 2AU, 2AC, 3C, 3-4U 8 IAAAA(C), IAG, IAU, 2AC, I-2C, I-2U
4
I-2AU, IAC, IG, 2-3C, 5U
3 IAAU, IAAC, tAG, IAU, IAC, 3-4C, 2-3U
4 zAU, I-2AC, tG, 2-3C, 3-4U
IO tAAAA(U), IAG, 2AC, 2-3C, 3-5U 9 tAAAA(U), IAG, 2-3AC, 2C, 4U II' IAAU, IAAC, IAU, IAC, tG, I-2C, 2-3U 12 (IAAU?)IAAC, IAU, IAC, IG, I-2C, I0 I-2AAC, I-2AU, IAC, IG, 2C, 2-3U 2-3U I3 1AAAU, I-2AAU, IAU, IAC, IG, 2C, I I ] IAAAU, I-2AAU, IAU, IAC, IG,
9
5 IAAU, tAAC, xAU, 2-3AC, IG, IC 2-3U 6 I-2AAAC, IAAU, 2AAC, IAU, 2-3AC, IG, IC, 2U 7 IAAG, IAAU, I-2AAC, IAG, 2AU, 3AC, 2-3C, 2U 8 1AAAA(C),IAG, IAU, 2-3AC, 2C, 2U
SVDV-It66 RNA
I IAAAC, IAAG, IAU, IAC, IC, 2U 2 IAAAC, IAAU, IAAC, 2AU, IAC, IG, 3-4C, 3U
Spot no.
Ribonuclease A analysis of the long RNase I"1 oligonucleotides of SVDV RNAs~
I IAAAC, IAAG, I-2AU, 2-3AC, IC, 2U I tAAAC, rAAG, IAU, 2AC, IC, IU z IAAAU, IAAAC, /AAU, IAAC, I-2AC, IG, 2-3C, 2-3U 3 IAAAC, IAAU, 2AU, IAC, IG, 2C, 2) IAAAU, I-2AAU, t-2AAC, IAG, I-zU 3J 2AU, I-2AC, IG, 2-3C, 3-4U 4 IAAU, IAAC, IAG, 2-3AU, IAC, 3-4C, zU
Spot no.~
Mixed oligonucleotides, but common to all 3 RNAs
Probably shared by UK and It66 RNAs Common Shared by UK and It66 RNAs Probably shared by UK and HK RNAs
Unique
Unique
Mixture of two oligonucleotides - common Common
Common
From gel position UK spot 4, HK spot 2 probably correspond. From analysis UK spot 4-It66 spot 3 correspond Shared by HK and It66 RNAs Common
Common Unique
Oligonucleotide comparison
Z
©
©
SVDV-UK RNA
IAAAA(C), IAAU, I-2AAC, IAG, IAU, zAC, 3C, 3U
IAAAC, IAG, IAC, 2C, IU
IAAG, IAAC, 2AC, 2C, I-2U
IAAG, IAAU, I-2AU, 2C, 3-4U
20
2I
22
23
SVDV-It66 RNA
20] IAAAA(U), tAG, tAU, IAC, 2C, IU 2 I ) IAAAC, IAAG, I-2AAC, IAU, 2AC, 2C, IU 22 tAAG, IAAU, IAU, 2-3C, 3-4U
)
I6 2AAU, 2AAC, IAC, IG, 3C, 2U 17 I-2AAAC, (IAAU?)(IAAC?)2-3AC, IG, 3C, 2U 18~ IAAAC, (IAAU?)IAAC, IAG, I-2AC, 3C, 2U 19 IAAAA(C), IAAU, tAAC, IAG, IAU, 2AC, 3-4C, 2U
I4
2z) IAAG, IAAC, IAU, 1AC, 2-3C
2I) IAAG, I-2AC, I-2C, 1U
20 I-2AAAU, IAAU, IAC, IG, 3-4C, 2-3U
I9
IAAAC, IAAU, IAAC, IAG, 2AU, 2AC, IG, 4C, 4U
Spot no.
15 IAAAU, IAAU, 2AAC, IAG, IAU, 2-3AC, IG, 4-5C, 3-4U
(cont.)
I-2AAU, IAAC, IAG, IAC, 2C, 2U
SVDV-HK RNA
Table 3
17 IAAAC, IAAU, IAU, zAC, IG, 2C, 2U I8 IAAU, IAAC, IAG, IAC, 2C, 3U
Spot no.
Shared by UK and HK RNAs
HK spots I8 and I9 mixed; spot I9 probably corresponds to UK spot 20. It66 spot 20 unique Unique Unique HK spots 20 and 2I mixed but unique
HK spot I5 and It66 spot I9 mixture of 2 oligonucleotides, one of which is shared. UK spot 18 - single oligonucleotide may correspond to it. Unique Unique
Unique
Common
Unique
Oligonucleotide comparison
* The table has been arranged so that oligonucleotides moving to about the same position in each two dimensional gel (Fig. 6a to c) are compared across the table. t The molarities of the products were estimated directly from the autoradiograph and the values should only be considered as approximate. :~ See Fig. 6(fto h).
IAAAU, IAAAC, IAAU, I-2AC, IG, 3-4C, 2-3U
I9
16 IAAU, tAAC, IAG, IAC, 2-3C, 2-3U I7 2AAU, IAG, IAU, 0AC?)3-4C, 2-3U 18 I-2AAC, IAC, IG, 2-3C, 2-3U
Spot no.:~
E"
r~
2AAU, 2AC, IG, IC, ~ 3U
15
)
15 16
13 14
Io ii 12
8 9"1
}
CB5-8o68 RNA
IAAU, I-zAAC, IAU, I-zAC, IG, z-3C, IU 2-3AU, IG, 3C, 4-5U IAAAU, 1AAU, IAG, 2AU, l-zAC, 1G, 2-3C, z-3U IAAAU, IAAU, 1-2AU, 1-zAC, 1G, 2-3C, 2-3U IAAAU, IAAU, I-zAU, I-zAC, IG, 3-4C, 2-3U IAAAG, IAAC, IAU, IAC, some C, some U (IAAU?)IAAC, IAU, IAC, IG, some C, some U IAAU, IAAC, IAC, IG, some C, some U IAAAU, IAAU, IAG, I-zAC, 3-4C, zU IAAU, 2AAC, lAG, z-3AU, I-zAC, IG, 3-5C, 2-3U IAAAC, IAAG, 2AU, 2AC, 2-3C , 2-3U IAAAU, IAAU, IAG (tAU?) 1AC, some C, some U IAAAU, IAG, ~-zAC, zC, 3-4U IAAU, 2-3AU, tG, 1-2C, 3-4U
some C, some U
IAAAC, IAAG, IAAU, IAAC, I-2AU, 2AC, I-2C, 3-4U IAAAA(C orU), IAAU, I-2AAC, IAC, IG,
Oligonucleotide comparison§
©
.~
.,--t
Unique Same analysis as SVDV UK spot 9, SVDVIt66 spot I I may correspond
Unique Unique
~Z
0
.,~
Unique CB5 8068 spot 12 two oligonucleotides; Unique :Z
Unique Unique; not SVDV HK/It66 spot 4 Mixed oligonucleotides. CBs-Faulkner spot 5 and CB-8o68 spot 6 may correspond to SVDV-UK spot I3: SVDV-HK spot It and SVDV-It66 spot 14. Unique Unique; CB5-8o68 spots 9 and IO mixed
Unique
Unique
* As for Table 3 see Fig. 6(d, e). t As for Table 3. See Fig. 6 (i, ]). § As far as possible these analyses have been compared to those in Table 3, using the position in the two dimensional gels as the basis for the comparison.
I3 I4
IAAAC, IAAU, IAAC, IAC, IG, 1-2C, I-2U IAAAA(U), IAG, IAU, 2-3AC, 1-2C, I-2U IAAAU(IAAAC?), IAAU, IAAC, IAU(IAC?), IG, 2C, 2J IAAAA(C or U), IAU, IAC, IG, I-2C, I-2U IAAAC, IAAC, 3AU, IG, IC, N 3U
I0 II 12
(IAAAG?)IAAU, (IAU?)2-3C, 3U IAAG, I-2AU, IAC, some C, 2U 1AAAC, (IAAU?) IAAC, IAC, IG, I-2C, I-2U IAAC, 2AU, IAC, IG, I-2C, some U
3 4
IAAAA(C or U), IAU, 2AC, IG, 2C, 2U 2AAU, IAU, IAC, IG, IC, 3-4U IAAAU, (1AAU?)IAU, IAC, 1G, 1C, 2U
3 4
7 8 9
2
I
0AAAG?)IAAAU, IAAC, IAU, I-2C, I-2U
IAAG, 2AAC, 2AU, IAC, I-2C, I-2U
Spot no.
Ribonuclease A analysis of the long RNase T1 oligonucleotides of the CB5 virus RNAs~f
CB5-Faulkner RNA
T a b l e 4-*
2
Spot no.~
7~
t..o
Molecular aspects of S VD V-Coxsackie B5 variation
313
Since the two dimensional gels separate the oligonucleotides according to size and charge (De Wachter & Fiers, I972), it is likely that oligonucleotides moving to the same relative position in the gels will have the same composition. This has been confirmed recently in this laboratory in a study of the specific RNase T1 oligonucleotides of the RNA of a virulent and an attenuated foot-and-mouth disease virus (Harris & Brown, I977). Nevertheless to check that several oligonucleotides are shared by the SVDV RNAs and to determine those oligonucleotides that are different, we have eluted a number of the oligonucleotides from the gels and determined their composition by RNase A digestion and high voltage electrophoresis on DEAE paper at pH 3"5 (Table 3 and Table 4). The key (Fig. 6fto h) shows those oligonucleotides that are shared by the SVDV RNAs as determined by RNase A analysis (Table 3). The results support those determined visually; there are at least 12 oligonucleotides that are shared by all three SVDV RNAs, 3 oligonucleotides shared only by SVDV-UK and SVDV-HK RNAs, 2 oligonucleotides shared by SVDV-UK and SVDV-It66 and I oligonucleotide shared by SVDV-HK and SVDV-It66 RNAs. A similar analysis of the CB5 RNAs (Table 4) demonstrated that the majority of these oligonucleotides were unique. There were, however, two CB5-8o68 oligonucleotides that had a similar analysis to SVDV oligonucleotides (Table 4, Fig. 6 i, i). As these CB5-8o68 oligonucleotides moved to about the same position in the gels as the corresponding SVDV oligonucleotides, it is possible that they are shared. One of these oligonucleotides may also be present in CB5-Faulkner RNA (see Table 4)- This oligonucleotide may therefore be common to SVDV and CB5 RNAs. DISCUSSION
Considerable variation in the homology of SVDV and CB5 RNAs has been obtained by different hybridization methods. The initial saturation experiments of Brown & Wild (t 974) and Brown et aL (1976) using alkali-treated virus RNAs and radioactive ds RNA showed that SVDV-UK RNA was 5o ~ homologous to the RNA of CB5-FauJkner and CB5-8068. The competition hybridization experiments with a2P-SVDV-UK RNA done under the more stringent conditions used in the present experiments showed the same level of annealing (Fig. 3a). However, competition experiments with 32P-CB5 RNAs gave consistently lower values (30 to 35 ~, Fig. 3 c and Table 2). The saturation hybridization experiments done under the conditions used for constructing the competition curves (i.e. 50 ~o formamide, 5 × SSC, 50 °C for I6 h) also gave a lower estimate of the homology of SVDV-UK and the CB5 RNAs (Fig. 2). The results of the thermal denaturation experiments may help to explain the variability of the competition reactions since they demonstrate clearly that the heterologous SVDV/CB5 RNA hybrids are extensively mismatched. It is still not clear, however, why there is a discrepancy between the saturation curves (Fig. 4) and the competition curves (Fig. 3). Presumably some sequences which will not hybridize completely are still similar enough to sequences in the homologous 32P-RNA to prevent them from annealing to complementary RNA. The lack of variability in the hybridization of the more closely related SVDV RNAs, measured by the three different methods, supports the above conclusions. In addition, the thermal denaturation experiments (Fig. 5) show that the SVDV-UK/SVDV-It66 RNA hybrid is not as well matched as the SVDV-UK/SVDV-HK RNA hybrid, despite the fact that SVDV-UK and SVDV-It66 RNAs are apparently more homologous than SVDV-UK and SVDV-HK RNAs (9° ~ compared to 85 ~o, see Table I). This result is confirmed by the fingerprints (Fig. 6 a to c) which show that SVDV-UK RNA shares more oligonucleotides with SVDV-HK RNA than with SVDV-It66 RNA. In view of this it is interesting that 21
V I R 35
314
T . J . R . HARRIS~ T. R. DOEL AND F. BROWN
preliminary experiments using radioimmune assay to measure the antigenic variation of the SVD viruses have shown that SVDV-UK is more closely related to SVDV-HK than to SVDV-It66 (J. R. Crowther, personal communication). The results of the 'mixed competitions' using equal amounts of two RNAs as competitor showed that the homologous regions shared by the CB5 virus RNAs were largely the same as those shared by the CB5 virus RNAs and SVDV-UK (Table 2). The substantial sequence homology of the three SVDV RNAs was also shown to be shared (Table 0. These hybridization results are similar to those obtained by Young, Hoyer & Martin (I968) for the three serotypes of poliovirus. The results also imply that the three SVDV viruses are members of a closely related group whereas the CB5 viruses are not, since CB 5 Faulkner R N A and CB5-8o68~RNA are as different from each other as they are from SVDV-UK RNA. It has been suggested from measurements of the sizes of the heteroduplexes of the RNAs of the poliovirus serotypes that the common sequences are not continuous but are located at different positions along the molecule (Young, I973). The conclusions from the fingerprinting of the RNase T1 oligonucleotides of the virus RNAs studied here have some bearing on this observation. Our results (Fig. 6, Table 3 and Table 4) support the hybridization and Tm data given above. The sharing of several long oligonucleotides by the SVDV RNAs (Fig. 6 a to c) suggests that there are continuous regions of common sequence and the lack of long oligonucleotides common to both the CB5 and SVDV RNAs suggests that the limited homologous regions between these viruses are randomly distributed along the genome. The location of both the common and the different oligonucleotides in the RNAs of the antigenically distinct SVD viruses is currently being investigated by orientating the long oligonucleotides relative to the poly (A) tract at the 3'-end, as has been done recently for avian RNA tumour viruses (Wang et al. I975; Coffin & Billeter, I976 ) and foot-and-mouth disease virus R N A (Harris & Brown, I976). This approach, coupled with annealing reactions using the 5'-end of the R N A (the region coding for the structural polypeptides and hence the antigenicity) should enable us to locate those parts of the molecule responsible for the antigenic variation. Although the polypeptides of the SVD and CB5 virus isolates examined in this paper have different mobilities in discontinuous SDS polyacrylamide gels (Harris & Brown, I975) the tryptic peptide maps of the unfractionated virus polypeptides show that there is considerable similarity in the methionine containing tryptic peptides. If, as seems likely, the peptides that do not contain methionine are also similar then the results suggest that there are only limited differences in the primary structures of the polypeptides of SVDV and CB5. By examining the tryptic peptides of the individual polypeptides of antigenically distinct viruses it should be possible to ascertain the extent of the sequence conservation and to correlate for the SVD viruses changes in the individual polypeptides with the antigenic differences. Further detailed analysis of the RNAs and polypeptides of SVDV and CB5 should give some insight into the changes which occur during antigenic drift in this group of viruses.
REFERENCES
&BURROWS,R. (1973). Antigenic differencesbetween isolates of swine vesicular disease and their relationship to Coxsackie B5 virus. Nature, London 245, 3t5-316. BROWN,F. &WILD,F. (1974). Variation in the Coxsackie virus type B5 and its possible role in the etiology of swine vesicular disease. Intervirology 3, I25-I28. BROWN, F., WILD, T.F., ROWE, L.W., UNDERWOOD, B.O. & HARRIS, T..1. R. (i976). Comparison of swine vesicular disease virus and Coxsackie B5 virus by serological and RNA hybridization methods. Journal of General Virology 3x, 231-237. BROWN, F, TALBOT, P.
Molecular aspects of S VD V-Coxsackie B5 variation
315
cornN, J.M. & ~rL~.ETER, M.A. (1976). A physical map of the Rous sarcoma virus genome. Journal of Molecular Biology Too, 293-318. CRESTFIELD, A. M., MOORE, S. & STEIN, W. It. 0963)- The preparation and enzymatic hydrolysis of reduced and S-carboxymethylated proteins. Journal of Biological Chemistry 238, 622-627. DARaY, C. & MINSON,A. C. (1973). The structure of tobacco rattle virus ribonucleic acids: common nucleotide sequences in the R N A species. Journal of General Virology 2x, 285-295. DE WACHTER, R. & VIERS, W. (r972). Preparative two-dimensional poIyacrylamide gel electrophoresis of 32Plabelled RNA. Analytical Biochemistry 49, 184-197. FRISBY, D. P., NEWTON, C., CAREY, N. H., FELLNER, P., NEWMAN, J. F. E., HARRIS, T. J. R. & BROWN, F. (i976). Oligonueleotide mapping of picornavirus R N A s by two-dimensional electrophoresis. Virology 7x, 379388. GRAVES,J. H. (1973). Serological relationship of swine vesicular disease virus and Coxsackie B5 virus. Nature, London 245, 314-315. HARRIS, W. J. R. & BROWN, F. (1975). Correlation of polypeptide composition with antigenic variation in the swine vesicular disease and Coxsackie B5 viruses. Nature, London 258, 758-76o. HARRIS, T. S. R. & BROWN, F. (I976). The location of the poly(C) tract in the R N A of foot and mouth disease virus. Journal of General Virology 33, 493-5oi. HARRIS, X. J. R. & BROWN, F. (1977). Biochemical analysis of a virulent and an avirulent strain of foot and mouth disease virus. Journal of General Virology 34, 87-Io5. PORTER, a., CAREY, N. H. & FELLNER, P. (1974). Presence of a large poly(C) tract within the R N A of encephalomyocarditis virus. Nature, London 248, 675-678. SARGENT, J. a. & VADLAMUDI,B. P. (I968). Electrophoresis of peptides on thin layers of silica gel. Analytical Biochemistry 25, 583-587. WANG, L. H., DUESBERG, P , BEEMON, K. & VOGT, P. K. (1975). Mapping RNase Trresistant oligonucleotides of avian tumor virus RNAs: sarcoma specific oligonucleotides are near the poly (A) end and oligonucleotides common to sarcoma and transformation defective viruses are at the poly(A) end. Journal of Virology x6, lO5I-IO7O. YOUNO, N. A. (1973). Size of the gene sequences shared by polioviruses types I, 2 and 3. Virology56, 4oo-403. YOtrNo, N. A., HOYER, B. H. & MARTIN, M. A. (1968). Polynucleotide sequence homologies among polioviruses. Proceedings of the National Academy of Sciences of the United States of America 6x, 548-555.
(Received 25 October I 9 7 6 )
21-2
