Journal o f General Virology (1992), 73, 695-700. Printed in Great Britain 695
Epitope mapping on fragments of beet necrotic yellow vein virus coat protein
U. Commandeur, ~* R. Koenig, ~ D.-E. Lesemann, ~ L. Torrance, 2 W. Burgermeister, ~ Y. Liu, 3 A. Schots, 4 M. Alric 5 and G. Grassi 6
1Biologische Bundesanstalt f~r Land- und Forstwirtschaft, Institut fftr Biochemie und Pflanzenvirologie, Messeweg 11, D-3300 Braunschweig, Germany, 2 Scottish Crop Research Institute, Invergowrie, Dundee D D 2 5 D A, U.K., 3 Beijing Agricultural University, Beijing, China, 4Laboratory for Monoclonal Antibodies, NL-6700 G W Wageningen, The Netherlands, 5 Laboratoire de Biologic Cellulaire et Molkculaire, F-63170 Aubiere, France and 6 Istituto Sperimentale per le Colture Industriale, S.O.P. de Rovigo, 1-45000 Rovigo, Italy
The location of five SDS-stable epitopes on the coat protein (CP) of beet necrotic yellow vein virus was determined by reacting Escherichia col#expressed free CP, as well as fusion proteins (FP) containing fragments of the CP, with polyclonal and monocional antibodies on Western blots. Epitope 1, which has previously been found to be exposed on only one extremity of the virus particle, was located in the region between amino acids (aa) 1 and 7, i.e. on the N terminus of the CP. It was blocked when the N terminus of the CP was linked to a portion of the /3-galactosidase sequence in an FP. Epitope 3, which has previously been found to be exposed on the opposite extremity of the
particle, was located in the region between aa 37 and 59. Epitope 4, which is exposed along the entire length of the particle, occurs on the C terminus o f C P (aa 183 to 188). Two previously unknown epitopes were identified in the regions between aa 115 and 125 and 125 and 140, respectively. The former was located on the same extremity of the particle as epitope 3, the latter became accessible only after denaturation of the particle. Nothing is known about the probably non- adjacent aa sequences that participate in the formation of the two SDS-labile epitopes (epitopes 2 and 5) which are found on one extremity and along the entire length of the particle, respectively.
Introduction
In previous studies (Lesemann et al., 1990; Koenig et al., 1990a) we have demonstrated the presence of five different epitopes on particles of beet necrotic yellow vein virus (BNYVV). Epitope 1 (the term 'epitope' is used in the singular throughout the text, but preliminary results of scanning experiments with synthetic peptides to overlapping regions of the virus CP suggest that at least some of the antigenic regions described here consist of a number of closely spaced, overlapping epitopes) is exposed only on one extremity of the particle, whereas epitopes 2 and 3 are exposed on the opposite one. In contrast, epitopes 4 and 5 are found along the entire length of the particle. Epitopes 2 and 5 are destroyed by treatment with SDS, whereas epitopes 3 and 4, and to some extent epitope 1, survive this treatment and thus can be studied by Western blotting. Trypsin degradation experiments had suggested that epitope 4 is located on the C terminus of the coat protein (CP) amino acid sequence, amino acids (aa) 183 to 188. Using Eseherichia coil-expressed free CP and a series of fusion proteins
(FPs) containing fragments of the CP we confirmed the presence of epitope 4 on the C terminus, and located epitope 1 in the region of the N terminus (aa 1 to 7) and epitope 3 in the region between aa 37 and 59 of the CP. Also, two new SDS-stable epitopes were found in the regions between aa 115 and 125 (epitope 6) and 125 and 140 (epitope 7), respectively. Epitope 6 was found on the same extremity of the particles as epitope 3, whereas epitope 7 became exposed only after denaturation of the particle. A preliminary account of part of this work has been given (Commandeur et al., 1990).
Methods
Polyclonal (PAbs) and monoclonal antibodies (MAbs) were the same as those used in previous studies (Lesemann et al., 1990; Koenig et al., 1990a, b) except MAb Ch 1 prepared by Liu et al. (1990) and MAb SCR 84 prepared by L. Torrance (unpublished results). A cDNA clone of the BNYVV CP gene (pCG700) (Ehlers et aL, 1991) served as the starting material for the construction of deletion clones using the modified expression vector pEX2 (Kocken et al., 1988) (Table 1). The deletion clones were obtained by restriction endonuclease subfragment exci- sion, using the natural sites available in the cDNA of the CP gene
T a b l e 1. Expression vector constructions based on the modified pEX2 vector used for the expression o f fragments o f B N Y V V CP as [3-gal FPs*
BNYVV CP sequence
Name of expressed No. of plasmid* Origin (aa) FP
pEV-N Cleavage of the BNYVV CP gene cDNA clone pGC700 t with ScaI and SmaI, insertion 1-103 9 of the fragment containing the 5' part of the CP gene into EcoRV/SmaI linearized pEX2
pEV-NAAva Complete digestion of pEV-N with AvaI and Sail, blunt-end repair and religation 1-59 2 pEV-NASty Complete digestion of pEV-N with StyI and SalI, blunt-end repair and religation 1-43 1 pEC-Sph Complete digestion of pEV-NAAva with ClaI, blunt-end repair, subsequent digestion 37-59 11
with SphI, in-frame connection of the fl-gal and CP genes to a polylinker fragment obtained from pGEM-3Zf(+) (Promega) by digestion with AvaI, blunt-end repair and digestion with SphI
pEC-Sty Complete digestion of pEV-NAAva with StyI, blunt-end repair, subsequent digestion 43-59 10 with ClaI, in-frame connection of the fl-gal and CP genes to a polylinker fragment obtained from pGEM-3Zf(+) (Promega) by digestion with EcoRI, blunt-end repair and digestion with AccI
pEV-C Cleavage of plasmid pGC700 with ScaI, insertion of the fragment containing the 3" part 104-188 9 of the CP gene into EcoRV/SmaI-linearized vector pEX2
pEV-CANsi Complete digestion of pEV-C with NsiI and PstI, and religation 104-114 4 pEV-CAEcR Complete digestion of pEV-C with EcoRI and Sail, blunt-end repair and religation 104-140 5 pEV-CAAva Complete digestion of pEV-C with AvaI and Sail, blunt-end repair and religation 104-165 6 pEV-CANco Complete digestion of pEV-C with NcoI and Sail, blunt-end repair and religation 104-175 7 pEV-CAEcN Complete digestion of pEV-C with EcoNI and SalI, blunt-end repair and religation 104-179 8 pEC-Nco Complete digestion of pEV-C with ClaI and NcoI, blunt-end repair and religation 176-188 13 pEC-BsE Complete digestion of pEV-CAEcR with BspEl and ClaI, blunt-end repair and religation 125-140 12
* Plasmids designated pEV encode FPs with a 45K N-terminal portion of fl-gal, those designated pEC encode FPs with a portion of fl-gal.
t pGC700 is described by Ehlers et al. (1991).
37K N-terminal
sequence of our BNYVV isolate (U. Commandeur, unpublished results) and in the pEX vector (Stanley & Luzio, 1984). By using these deletion clones, defined fragments of BNYVV CP were expressed as FPs with either a 37K or a 45K portion of the fl-galactosidase (fl-gal) sequence at the N terminus (Table 1). Full-length BNYVV CP was expressed from plasmid pJLAcp. For its construction, the cDNA of the BNYVV CP gene was excised from pGC700 by digestion with AflIII and BamHI, and ligated into the NcoI/BamHI-linearized vector pJLA502 (Schauder et al., 1987). Recombinants were screened on the basis of the size of the inserts and of the expressed FPs, and were checked by restriction analysis. The transformed bacteria were grown in 2 ml LB medium containing 100 ~tg/ml ampicillin and the expression of the FPs was induced by incubation for 6 h at 42 °C. The cells were harvested by centrifugation at 12000g for 10 min and resuspended in 100 pl 1 x extraction buffer (Laemmli & Favre, 1973). The mixture was kept in a boiling water bath for 10 min and 5 ~tl aliquots were analysed on 10 ~ or 15 ~ polyacrylamide gels (Laemmli & Favre, 1973). Western blotting (Burgermeister & Koenig, 1984) and immunogold electron microscopy (Lesemann et aL, 1990) were done as described previously.
Results
The reactivity of FPs containing different parts of the BNYVV CP amino acid sequence with PAbs and MAbs was checked by means of Western blotting (Fig. 1 and 2). PAbs did not react with the FP containing aa 1 to 43 of BNYVV CP; however, they did react strongly with the FP containing aa 1 to 59. This suggested that all or part of
the region between aa 44 and 59 is involved in the formation of an SDS-stable epitope (Fig. 1). No increase in reactivity was observed when the BNYVV CP sequence in an FP was increased in length to aa 103, indicating that there are apparently no strong SDS-stable epitopes in the region between aa 60 and 103. The FP containing aa 104 to 114 did not react with PAbs, suggesting that this region does not contain SDS-stable epitopes. However, the FP containing aa 104 to 140 of BNYVV CP strongly reacted with PAbs, suggesting that all or part of the region between aa 115 and 140 is involved in the formation of SDS-stable epitopes. Further increasing the length of the BNYVV CP sequence in FPs to aa 179 did not cause a pronounced increase in reactivity with PAbs. Thus, there are apparently no strong SDS-stable epitopes in the region between aa 141 and 179. However, a strong increase in reactivity was observed with the FP containing the entire C terminus of BNYVV CP (aa 180 to 188). This suggests that all or part of the sequence between aa 180 and 188 is involved in the formation of SDS-stable epitopes (Fig. 1).
Testing the same FPs with MAbs showed that each of the three antigenic regions described was recognized specifically by some MAbs, e.g. the region between aa 44 and 59 by MAbs 41 and 47, the region between aa 115 and 140 by MAbs Chl and SCR 84, and the region between aa 180 and 188 by MAbs MAFF 6, MAFF 7,
Epitopes on B N Y V V coat protein 697
(a)
(b)
(e)
No. B N Y V V of FP CP aa
1 1~43 o ~ L - - - ~ _ - ~ 1 2 I 59 o - - U ~ ' Z - ~ _ ~ _ _ . ~ 3 1 103 o . - - I ] 4 104 114 o E ~ 5 104 140 0 [ £ _ _ ~ 6 104-165 o E - - I
7 104 175 0 E - - ]
8 t04-179 o [ 1~ 9 104-188 o
I | 37-59 o - - - - L - - - 3 12 125-140 o 13 176-188 o ~ - - - { ~
MAbs 41 and 47
O I_ . . . . 3 1 l 43
2 1-59 0 [ 1 10 43-59 O - - ~
11 37-59 0-------4 : I
M A b Chl /
4 104-114 o - - - - - ~ 5 104-140 o L . . . . . . . l
12 125 140 o . . . . .
MAb SCR 84 /
4 104-114 C ~ - - - ~ 5 ~04-140 o L - 7
12 125 140 0 - - - ~ _ ~
MAbs M A F F 6 and 7, 3H12, 4 F l l , 8B6
8 104-179 o - - - q Z ] ,[ 9 104-188 O - - - - ~ _=
13 176-188 o [ ]
MAbs M A F F 8, 9 and 10
14 1-188 o - - - - 4 -
8 188 1[
1 188
1 I I 1 CP regi . . . . ith ~ - ~ - ~ _ _ ~ [ . ~ ~1 antigenic reactivity
Fig. 1. Schematic representation of the reactivity of PAbs (a) and MAbs (b) with CP and with fragments of the CP expressed as FPs in E. coil (c) Comparison of the location of identified epitopes and antigenic sites predicted by computer analysis. The data marked 1" were extimated from experiments with trypsin-treated particles (Koenig et al., 1990a). O , p-gal; tzz~, FP does not react, ~ , FP reacts only very weakly; ~ , FP reacts well; I , FP reacts much more strongly than the preceding FP with a somewhat shorter CP sequence. Arrows indicate epitopes. AI, Antigenic index.
(a)
(b)
(c)
(d)
(e)
1 2 3 4 5 6 7 8 9
8B6, 3H12 and 4 F l l (Fig. 1). The epitope for which MAbs 41 and 47 are specific had formerly been named epitope 3, the one for which MAbs MAFF 6, MAFF 7, 8B6, 3H12 and 4 F l l are specific had been named epitope 4 (Koenig et al., 1990a).
We also produced FPs containing only the BNYVV CP sequences between aa 43 and 59, 125 and 140 or 176 and 188, respectively. The latter FP reacted as strongly with the MAbs as the FP containing aa 104 to 188, suggesting that epitope 4 resides entirely within the sequence between aa 176 and 188 (Fig. 1). However, the FP containing aa 43 to 59 showed weaker reactivity with MAbs 41 and 47 than the FP containing aa 1 to 59. This suggested that some amino acids preceding the sequence
Fig. 2. Examples of the results of Western blotting, on which the scheme in Fig. 1 is based. Reactivity of FPs 1 to 9 (lanes 1 to 9) with PAbs to fl-gal (a) and BNYVV particles (b), and MAbs 41 (c), Chl (d) and 3H12 (e).
between aa 43 and 59 may also be involved in the formation of epitope 3, although the FP containing aa 1 to 43 showed no reactivity. Indeed, we found that an FP containing aa 37 to 59 reacted just as strongly with MAbs 41 and 47 as the FP containing aa 1 to 59 (Fig. 1).
MAb Chl reacted equally well with FPs containing either aa 104 to 140 or 125 to 140. However, MAb SCR 84 and several PAbs reacted only with the FP containing
698 U. Commandeur and others
1 2 3 a free protein in E. coli. In Western blots such preparations yielded several closely spaced bands which all reacted with PAbs (Fig. 3). However, only the most slowly moving band, which moved at the same rate as intact CP from virus particles, reacted with MAbs MAFF 8 or MAFF 9. The nucleotide sequence of the CP gene is known to contain several in-frame AUGs in the 5' region (Bouzoubaa et al., 1986) which may provide internal translation starts in bacteria. The more rapidly moving bands in Western blots apparently represent truncated forms of the CP which start at aa 8 or 13. Since MAbs MAFF 8 and MAFF 9 did not react with these truncated forms of the CP, we conclude that they are specific for an epitope which is predominantly or entirely formed by N-terminal aa 1 to 7 of BNYVV CP.
Fig. 3. Reactivity of E. coli-expressed CP starting at aa 1 (upper band), aa 8 (middle band) or aa 13 (lower band) with PAbs (lane 1) MAbs to epitope 1 (lane 2), or of CP from virus particles with PAbs (lane 3). Arrowheads denote bands of interest.
aa 104 to 140, and not with that containing aa 125 to 140. Neither of the MAbs reacted with the FP containing aa 104 to 114 (Fig. 1). The difference in the reactivity of the two MAbs suggests that the region between aa 115 and 140 encompasses at least two (groups of) epitopes; one, for which MAb Chl is specific (named epitope 7), resides within the sequence between aa 125 and 140, whereas the other, for which MAb SCR 84 is specific (named epitope 6), is located in the sequence between aa 115 and 125 (Fig. 1). MAb SCR 84 reacted specifically with an FP containing aa 104 to 125 (result not shown). The conclusion that MAbs Chl and SCR 84 are specific for different epitopes was confirmed by the electron microscopic immunogold technique (Lesemann et al., 1990). The binding of MAb SCR 84, specific for epitope 6, was readily detected on one extremity of the particle; experiments with mixtures of MAbs suggested that it is the same extremity on which epitope 3 is exposed (results not shown). However, MAb Chl was not visibly bound to the particles. Therefore epitope 7 is apparently a cryptotope which becomes exposed only after denatur- ation of the particle.
With FPs we were not able to detect SDS-stable epitopes on the N terminus of the CP amino acid sequence (Fig. 1). However, from previous Western blotting studies it was known that MAbs MAFF 8 and MAFF 9 detect SDS-stable epitopes on the CP isolated from virus particles (Torrance et al., 1988). In our experiments with FPs these MAbs did not react with any of the four SDS-stable epitopes described. To identify the reaction site of these MAbs we expressed BNYVV CP as
Discussion
The results of this paper, summarized in Fig. 1, indicate that at least five different regions of BNYVV CP are involved in the formation of SDS-stable epitopes, i.e. the N terminus (aa 1 to 7), the C terminus (aa 180 to 188) and the regions between aa 37 and 59, 115 and 125, and 125 and 140. Previous studies with untreated and trypsin- treated virus particles had already suggested that C- terminal aa 183 to 188 participate in the formation of an epitope, designated epitope 4 (Koenig et al., 1990a). With E. coli-expressed fragments of BNYVV CP we have located this epitope to a similar region, i.e. aa 180 to 188. The position of epitope 3, which was formerly located in the region between aa 1 and 103 (Koenig et al., 1990a), has now been mapped more precisely in the region between aa 37 and 59. Since the FP containing aa 37 to 59 reacted more strongly with MAbs 41 and 47 than the FP containing aa 43 to 59 and the FP containing aa 1 to 43 showed no reactivity, jt seems likely that it is mainly the N-terminal part of the sequence between aa 43 and 59 which contributes to the formation of epitope 3, but that, in addition, there is also some involvement of the C- terminal part of the sequence between aa 37 and 43.
It is interesting that epitope 1, which is apparently at the N terminus of the CP, was detected only on the free CP and not on FPs; apparently it is blocked when part of the /3-gal sequence is attached to the N terminus of BNYVV CP. Two previously unknown SDS-stable epitopes, epitopes 6 and 7, were mapped in the regions between aa 115 and 125 and 125 and 140, respectively.
The location of the five SDS-stable epitopes described here was compared with the predictions of an antigeni- city plot (Jameson & Wolf, 1988) for BNYVV CP (Fig. 1). The predictions of the computer program (Devereux et al., 1984) are based on calculations for hydrophilicity, surface probability and flexibility along the protein
Epitopes on B N Y V V coat protein 699
. . - . , ~ ~ ~ @ (aal-7)
---j (aa 183 188) " ~ / ' @
• "%
! I
I /
m m m
I I
/ Q
(aa 37-59)
/ J -)_
W
@ (aa 115 125)
(~) (aa 125-140) (not accessible)
Fig. 4. Schematic representation of the location of the SDS-stable epitopes (encircled, approximate position in amino acid sequence in parentheses) on particles of BNYVV.
chain. The epitopes on the N and C termini and the newly described epitope 7 between aa 125 and 140, lie in regions of high predicted antigenicity, but epitope 3 (aa 37 to 59) and 6 (aa 125 to 140) do not. Predicted antigenic areas in other parts of the CP aa sequence were recognized by neither PAbs nor MAbs.
Each of the five SDS-stable epitopes on BNYVV CP reacted specifically with some MAbs for which we know the binding sites on the virus particles from either previous (Lesemann et al., 1990; Koenig et al., 1990a) or the present studies. This enabled us to map the position of these five epitopes on the virus particle (Fig. 4). The C terminus of BNYVV CP, which contains epitope 4, is exposed along the entire length of the particle and is readily removed by treatment with trypsin or plant proteases. The N terminus, which contains epitope 1 is exposed only on one extremity of the particles because MAbs MAFF 8 and MAFF 9, which are specific for this epitope, bind only in this area (Lesemann et al., 1990). Epitope 3, in the region between aa 37 and 59, and epitope 6, in the region between aa 115 and 125, are also exposed on only one extremity of the particle, apparently the opposite one to that which carries epitope 1 because particles treated with a mixture of MAbs specific for epitopes 1 and 3 or for epitopes 1 and 6 showed gold labelling on both extremities, whereas particles treated
with a mixture of MAbs specific for epitopes 3 and 6 were labelled on only one extremity. The newly recognized epitope 7 in the region between aa 125 and 140 is apparently a cryptotope because we were unable to detect binding of the respective MAb to the intact virus particle. It became exposed only after denaturation of the particle.
In addition to the five groups of SDS-stable epitopes described here, there are at least two groups of SDS- labile epitopes, i.e. epitope 5 along the entire length and epitope 2 on one extremity of the particle (Koenig et al., 1990a). We do not know which parts of the CP amino acid sequence participate in their formation.
We are greatly indebted to Mrs Petra Liiddecke, Mrs Jutta Burghardt and Mrs Marianne Koerbler for expert technical assistance, and to the Deutsche Forschungsgemeinschaft (grant Ko 518/11-1) for financial support.
R e f e r e n c e s
BOUZOUBAA, S., ZIEGLER, V., BECK, D., GUILLEY, H., RICHARDS, K. & JONARD, G. (1986). N ucleotide sequence of beet necrotic yellow vein virus RNA-2. Journal of General Virology 67, 1689-1700.
BURGERMEISTER, W. & KOENIG, R. (1984). Electro-blot-immunoassay- a means for studying serological relationships among plant viruses ? Phytopathologische Zeitschrift 111, 15-25.
COMMANDEUR, U., KOENIG, R., TORRANCE, L., BURGERMEISTER, W., GRASSI, G., AERIe, M. & Scnors, A. (1990). Epitope mapping on E. coli expressed fragments of beet necrotic yellow vein virus coat protein. Schriftenreihe der Deutschen Phytomedizinischen Gesellschafi, Bd. 1. Proceedings of the First Symposium of the International Working Group on Plant Viruses with Fungal Vectors. Braunschweig, Germany, August 21-24, 1990, pp. 61-64. Stuttgart: Ulmer-Verlag.
DEVEREUX, J., HAEBERLI, P. & SMITHIES, O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Research 12, 387-395.
EHLERS, U., COMMANDEUR, U., FRANK, R., LANDSMANN, J., KOENIG, R. & BURGERMEISTER, W. (1991). Cloning of the coat protein gene from beet necrotic yellow vein virus and its expression in sugar beet hairy roots. Theoretical and Applied Genetics 81, 777-782.
JAMESON, B. A. & WOLF, H. (1988). The antigenic index: a novel algorithm for predicting antigenic determinants. CABIOS 4, 181- 186.
KOCKEN, C. H. M., SCHEER, J. M. J., WELLING, G. W. & WELLING ° WESTER, S. (1988). Purification of fusion proteins expressed by pEX3 and a truncated pEX3 derivative. FEBS Letters 236, 132-134.
KOENIG, R., COMMANDEUR, U., LESEMANN, D.-E., BURGERMEISTER, W., TORRANCE, L., GRASSI, G., ALR1C, M., KALLERHOFF, J. & ScHo~, A. (1990a). Antigenic analysis of the coat protein of beet necrotic yellow vein virus by means of monoclonal antibodies. Journal of General Virology 71, 2229 2232.
KOENIG, R., COMMANDEUR, U., LESEMANN, D.-E., TORRANCE, L., BURGERMEISTER, W., GRASSI, G., ALRIC, M., KALLERHOFF, J. & Scno~, A. (1990b). Antigenic analysis of beet necrotic yellow vein virus by means of monoclonal antibodies. Sehrifienreihe der Deutschen Phytomedizinischen Gesellschafi, Bd. 1. Proceedings of the First Symposium of the International Working Group on Plant Viruses with Fungal Vectors. Braunschweig, Germany, August 21-24, 1990, pp. 57q/0. Stuttgart: Ulmer-Verlag.
LAEMMLI, U. K. & FAVRE, M. (1973). Maturation of the head of bacteriophage T4. I. DNA packaging events. Journal of Molecular Biology 80, 575-599.
700 U. Commandeur and others
LESEMANN, D.-E., KOENIG, R., TORRANCE, L., BUXTON, G., BOONE- KAMP, P. M., PETERS, D. & SCHOTS, A. (1990). Electron microscopi- cal demonstration of different binding sites for monoclonal antibodies on particles of beet necrotic yellow vein virus. Journal of General Virology 71, 731-733.
LIU, J. S., TSAr, Z. N. & LIu, Y. (1990). Preparation of antisera and McAB against BNYVV with some isolates. First Symposium of the International Working Group on Plant Viruses with Fungal Vectors. Braunschweig, Germany, August 21-24, 1990. Abstracts of Papers, p. 14.
SCHAUDER, B., BLOCKER, H., FRAme, R. & McCARTHY, J. E. G. (1987).
STANLEY, K. K. & LUZIO, J. P. (1984). Construction of a new family of high efficiency bacterial expression vectors: identification of cDNA clones coding for human liver proteins. EMBO Journal 3, 1429-1434.
TORRANCE, L., PEAD, M. T. & BUXTON, G. (1988). Production and some characteristics of monoclonal antibodies against beet necrotic yellow vein virus. Annals of Applied Biology 113, 519-530.
(Received 1t September 1991; Accepted 19 November 1991)
