- Email: [email protected]
Use of recombinant gp135 to study epitope-specific antibody responses to maedi visna virus
Journul of Virological Methods, 43 (1993) 221-232 m$’1993 Elsevier Science Publishers B.V. All rights
reserved
VIRMET
01506
/ 0166-0934/93/$06.00
Use of recombinant gpl35 to study epitope-specific antibody responses to maedi visna virus N. Carey,
D.J. Roy and R.G. Dalziel
Department of Veterinary Pathology. University of Edinburgh, Summerhall, Edinburgh EH9 IQH (UK) (Accepted
27 January
1993)
Summary The envelope glycoprotein gp135 of the ovine lentivirus maedi visna virus (MVV) is the main target for neutralising antibody in vivo, however little is known about the specific regions of gp135 which elicit this neutralising response. We have used the polymerase chain reaction (PCR) to generate overlapping fragments of the gp 135 gene which have been expressed as fusion proteins in the yeast Ty-VLP system. These fusion proteins have been used to analyse the antibody response to gp135 in MVV infected sheep and we are able to identify at least three distinct regions of gp135 to which antibodies are directed. The approach described in this paper provides a rapid and simple method of generating overlapping fusion proteins with which to carry out epitope mapping studies. Maedi visna virus; Glycoprotein
gpl35;
Epitope-specific
antibody
Introduction Maedi visna virus (MVV) is the prototype virus of the family lentiviridae other members of which include human immunodeficiency virus (HIV), feline immunodeficiency virus (FIV) and equine infectious anaemia virus (EIAV). MVV infection in sheep results in the development of chronic pneumonitis (Sigurdson, 1954) and/or a progressive demyelinating disease (Sigurdson et al., 1957) either of which may become manifest years after the initial infection. In Correspondence to: R.G. Dalziel, Department Summerhall, Edinburgh EH9 lQH, UK.
of Veterinary
Pathology,
University
of Edinburgh,
222
some cases, chronic mastitis (van der Molen et al., 1985) or arthritis (Oliver et al., 1981) may also be apparent. In contrast to HIV which infects predominately CD4’ lymphocytes and cells of the monocyte/macrophage lineage, MVV infection in vivo is limited to monocyte/macrophages with a possible involvement of cells in the CNS (Haase 1985). Following experimental infection of sheep a low level of MVV specific antibody (Ab) is produced one to two weeks post infection. However it is not until approximately six months post-infection that neutralising Ab can be demonstrated (De Boer, 1970). The component of the MVV virion to which neutraiising Abs are directed is the envelope glycoprotein. The envelope precursor protein is cleaved post-translationally to yield a membrane associated gp46 non-covalently linked to gp135, which is exposed on the virion surface. Despite the importance of the in vivo Ab response to gp135 the regions of the protein which elicit this immune response have not been studied. The recent publication of the nucleotide sequence of several strains of MVV (Sonigo et al., 1985; Braun et al., 1987; Querat et al., 1990; Sargan et al., 1991) has facilitated the production of recombinant envelope protein with which to carry out epitope mapping studies. We have made use of the yeast Ty-VLP expression system in which a truncated form of the TYA gene product (designated pl) derived from the yeast retrotransposon (Ty) is expressed in yeast from a multicopy plasmid. The expressed pl self assembles to generate Virus Like Particles (VLPs). Insertion of a gene of interest downstream of TY.4 and subsequent expression generates a Ty(pl):fusion protein which is also capable of self assembly into a hybrid TyVLP. The protein of interest is thought to be expressed on the surface of the hybrid VLPs (Adams et al., 1988) which are then readily purified by density gradient centrifugation. We have expressed the entire MVV gp135 as three overlapping fusion proteins. This is the first report of the production of recombinant MVV gpl35. The recombinant Ty:Env fusion proteins were used to investigate the immune response to gp135 in MVV infected sheep. We show that both naturally infected and experimentally infected animals produce Abs which recognise the Ty:Env fusion proteins. We further demonstrate that the Ab response to regions of gp135 differs in individual animals and that we are able to use the Ty:Env fusion proteins to partially epitope map the Ab response to gp135 in sheep. The combination of the polymerase chain reaction (PCR) using tailored primers to generate and clone gene sequences and subsequent expression of these sequences using the Ty-VLP system results in a rapid and simple method for the expression of protein sequences. The use of overlapping fusion proteins of the type described to map Ab responses is a tractable and cost effective alternative to conventional peptide binding approaches.
223
Materials and Methods Virus and cells: MVV strain EVI (Sargan et al., 1991) was propagated on the WSCP cell line (a sheep choroid plexus cell line obtained from Dr. M. Dawson, Central Veterinary Laboratories, Weybridge, UK) in Dulbecco’s Modification of Eagle’s Medium (DMEM) supplemented with 2% fetal calf serum (FCS).
sheep 848A and 754N are Finn/Blackface crosses which were Sheep. experimentally infected with MVV strain EVI. Sera used in the experiments described were taken from sheep 754N at 18 months post infection and from 848A at 12 months post infection. Sheep 2, 74, 68 and GO27 were naturally infected Texels which displayed classical maedi lesions at post-mortem. Sera from these sheep were taken at post-mortem and the duration of infection was unknown. Sera from all the sheep used in this study reacted by immunoblotting with gp135 from EVl infected cell lysates. PCR amplfication
of MVV
envelope gp135
The overlapping fragments of gp135 outlined in Fig. 1 were generated by PCR using the conditions of Ohara et al. (1989). Extra-chromosomal proviral DNA purified from EVl infected WSCP cells (Hirt, 1967) was used as template. The published sequence of EVI was used to derive primers whose location on the gp135 gene is shown in Fig. 1 and whose sequence is outlined in Table 1. Each primer contained a BamHI site at its 5’-end to facilitate subsequent cloning steps. Cloning of gp135 ,fragments
PCR-generated env DNA was digested with BamHI, gel purified and inserted into the BamHI site of the phagemid pTZl9R (Pharmacia) for sequence analysis. Sequencing was carried out by the method of Sanger et al. (1977) using a Sequenase Version 2.0 Kit (USB) and the manufacturer’s protocol. The PCR fragments were also inserted into the BamHI site of the yeast/E.coli shuttle vector pOGS40 (for Env B and C) or pOGS42 (Env A) (Gilmour et al., 1989) and transformed into E. coli JM83. The vectors were generously supplied by British Bio-Technology Ltd, Oxford. In these plasmids expression of the TYA gene product is driven from the galactose inducible hybrid PGK-GAL (PAL) promoter (Dobson et al., 1982, Kingsman et al., 1990). Transformants were screened by colony hybridisation (Buluwela et al., 1989) using appropriate gp135 specific PCR products as probes and those which contained env sequences were selected. These clones were further screened by diagnostic restriction enzyme digestion to determine the orientation of the inserted sequences with respect to the pl sequences. Clones containing single copies of the inserted DNA in the correct orientation were designated p:Env:A.l, p:Env.B. 1 or p:Env.C.l. etc.
224
Plasmid DNA purified by alkaline lysis and chromatography using ~523 columns (5-Prime ~ 3 Prime Inc) was used for yeast transformation. Yeast transformation Yeast were cultured as described by Kingsman et al. (1990) and transformed using the method of Hinnen et al. (1978). The protease deficient Saccharomyces cerevisiae strain BJ2 168 (a, leu2 _, trpl ~ , ura3-52, prbl-1122, pep4-3, pcrl-407, ga12) (Jones 1991) was simultaneously transformed to leucine and uracil independence with the relevant recombinant pl :Env yeast expression plasmid and PUG 41 S, a plasmid which expresses Gal4 and allows upregulation of expression from the PAL promoter of the pOGS plasmids upon galactose induction (Burns et al., 1991). Following galactose induction whole cell extracts of transformed yeast (Burns et al., 199 1) were analysed for the presence of a protein species corresponding to the Ty:Env fusion protein by reducing SDSPAGE (Laemmli, 1970) and immunoblotting. All primary transformants derived using the same pl .Env plasmid expressed approximately equal amounts of fusion protein when analysed by immunoblotting. One transformant expressing each fusion protein was chosen for further analysis and designated T:EnvA, T:EnvB or T:EnvC. Fusion proteins obtained were termed Ty:EnvA, Ty:EnvB or Ty:EnvC. Purification
of Ty:Env fusion proteins
Ty:Env fusion proteins were purified from 4L yeast cultures 24 hours after galactose induction essentially as described by Reyburn et al. (1992). Briefly, cells were resuspended in 100 mM Tris-HCl pH 7.4; 2 mM EDTA; 140 mM NaCl (TEN buffer), disrupted by agitation with glass beads and the lysate centrifuged at 100 000 x g for 1 h at 4°C onto a 60% sucrose (in TEN) cushion. The material which banded at the interface was dialysed overnight against TEN buffer to remove sucrose and the Ty:Env particles purified by centrifugation (53 000 x g 3 h at 4°C) through 1545% linear sucrose gradients overlaid on a 60% sucrose cushion. The gradient was fractionated by positive displacement and fractions analysed by SDS-PAGE. Fractions containing hybrid Ty: EnvVLPs were stored at - 70°C. Immunoblotting Ty:Env proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes by electrophoresis at 0.4 mA cmy2 for 1 h in 25 mM Tris, 20% methanol using a semi-dry blotter. Unfilled sites on the nitrocellulose membrane were blocked by incubation in 5% milk powder/sterile phosphate buffered saline pH 7.0 (Milk powder/SPBS) for 30 min at room temperature. Sera were diluted in milk powder/SPBS and membrane strips incubated with this sera for 16 h at 4°C. After extensive washing bound
225
antibody was visualised using alkaline-phosphatase conjugated second antibody and NBT/BCIP (Nitro-blue Tetrazolium/bromo-chloro-indoyl phenol toluidine salt) as the detection system. Preparation
of yeast for electron microscop?.
Yeast were prepared for transmission electron microscopy following the method of Byers and Goetsch (1975). 60 nm sections were cut, post stained with uranyl acetate and lead citrate and examined using a Phillips EM400 transmission microscope.
Results Generation of Ty:Env fusion proteins Since little information on the domain structure of gp135 is available we chose primers (Table 1) to generate gp135 fragments with predicted sizes of 37.1 kDa (EnvA), 34.2 kDa (EnvB) and 20.1 kDa (EnvC) which overlap by approximately 60 amino acids (Fig. 1). The sizes of the gp135 fragments are significantly smaller than 43 kDa which is the maximum size of protein so far reported to be expressed in the Ty-VLP system (Burns et al., 1992). Sequencing of the PCR derived clones revealed that individual clones displayed variation from the published EVI sequence (Sargan et al. 1991) of between 2 and 7%. None of the clones sequenced contained a stop codon in the envelope reading frame. The observed variation was not introduced as a result of the PCR procedure as sequencing of subsequent PCR material generated using this cloned material as template did not display variation from the original clone (data not shown). Immuno-blot analysis of transformed cell lysates using a rabbit anti-Ty-VLP antiserum which detects the pl component of the fusion protein revealed that the pl protein, which has a predicted molecular weight of 43 kDa, migrates in TABLE I Nucleotide sequence of PCR primers. Primer
no.
1
5’ TCCCGGATCCTTTGTTTAGGATGGCAAGCAC 5’ TCCCGGGATCCGAGGAAACAGGACATGCAC 5’ TCCCGGGATCCGGATGTAATTGCTCAAGGTCAGG 5’ ACACCGGGATCCAGGACTCTTCCCTCGTCCTG3’ 5’ ACACCCGGGATCCCCCTATTCTGTCTAACGTGC 5’ACACCCGGGATCCTCACCAACCCTATGCCCCTC3’
2 3 4 5 6
Nucleotide no. 3’ 3’ 3’ 3’
59846005 67356753 74647485 6922.-6903 7652-7633 8000-798 I
Primers are numbered as in Fig. 1. The position on the MVV genome spanned by the primer is shown on the right. Numbering is as in Sargan et al. (1991). Bold letters denote nucleotides which are not part of the MVV sequence, with the RamHI site underlined.
226
670
490 -) 3
EnvC
* 6
Fig. I. Regions of gpl35 expressed in this study. The regions of gp135 referred to in the text as EnvA. EnvB and EnvC are shown diagrammatically. Numbers above the lines indicate the position of the amino acid residue which bounds the fragment. Numbering is as in Sargan et al. (1991). Arrows and numbers below the line refer to the primers described in Table 1 used to generate the PCR fragments.
0
z
w i; _i_
c”
-.I
94 67 -
. l
$ .
a4
43
30
Fig. 2. Immunoblot of puritied hybrid Ty:Env VLPs detected using VLP specific rabbit antisera pl VLPs were purified and immunoblotted as descrtbed in the text. Dots to the right of each track indicate the position of the fusion proteins. pl routinely migrates as a doublet and is indicated by two dots. M, (kDa) of molecular-weight standards is given.
221
SDS-PAGE with an apparent molecular weight (M,) of 55-60 kDa (Fig. 2). Based on the observed M, of pl, Ty:EnvA and Ty:EnvB have predicted molecular weights of approximately 92-97 kDa and 89-94 kDa respectively and Ty:EnvC a predicted molecular weight of approximately 75580 kDa. The observed M, values of the pl.fusion proteins were Ty:EnvA, 94-98 kDa, Ty:EnvB, 85 kDa and Ty:EnvC, 76 kDa (Fig. 2) which are consistent with those predicted. Purified hybrid VLPs were also analysed by SDS-PAGE and protein visualised by silver staining. The pl and Ty:EnvA protein bands were clearly visible but no bands corresponding to Ty:EnvB or C were distinguishable (data not shown), despite detection by immunoblotting. When analysed by SDS-PAGE the purified Ty:Env preparations all contain a number of contaminating bands which are present despite sucrose gradient purification. It was not therefore possible to determine accurately the amount of fusion protein using a standard protein assay system. Examination of the structure of the intracellular hybrid-VLPs by transmission electron microscopy revealed that VLPs containing pl alone are visible as circular structures approximately 15 to 35 nm in diameter with an electron dense outer region (Fig. 3b). similar structures are present in T:EnvB and T:EnvC cells although at much lower levels compared to control transformants expressing Ty-VLPs, a finding which is consistent with the low yield of fusion protein (data not shown). Fig. 3a shows that the VLPs formed after induction
Fig. 3. Electron micrographs of intracellular VLPs. Electron micrographs of thin sections of transformed yeast showmg size and morphology of native and hybrid VLPs. a) T:EnvA cells, b) Control yeast which produce native Ty-VLPs. VLPs are indicated by arrows scale bar = 2 /cm.
228
94 67
B
43
f 30
20
Ty:EnvA
TY
94 67
67
43
30
Ty:EnvB
Ty:EnvC
229
of T:EnvA cells are more numerous, larger (approximately 70 to 150 nm in diameter) and exhibit a heterogenous morphology. Possible reasons for the aberrant nature of the Ty-VLPs formed in T:EnvA cells are discussed later. Recognition
of Tv:Env jikon
proteins by serajiom
MVV infected sheep
The relative amount of pl protein in native Ty, Ty:EnvA, Ty:EnvB and Ty:EnvC preparations was determined by immunoblotting using a rabbit TyVLP specific antiserum (data not shown). Equal amounts of Ty:EnvA, Ty:EnvB and Ty:EnvC were loaded with native Ty loaded in slight excess. Sera from four naturally infected sheep and two which had been experimentally infected with the EVl strain of MVV were assayed by immunoblotting for reactivity with the Ty:Env fusion proteins. Sera from sheep G027,754N and 74 reacted with all three fusion proteins, with 754N reacting more strongly than GO27 or 74 (Fig. 4). Sera from sheep 848A reacts weakly with Ty:EnvC but not with Ty:EnvA or Ty:EnvB. None of these sera reacted with pl alone. Sera from sheep 2 and 68 show a faint reactivity against pl alone. Serum from sheep 68 reacts more intensely with all three Ty:Env fusion proteins suggesting strongly that this serum is also reacting specifically with gpl35 sequences. Similarily serum from sheep 2 is reactive with Ty:EnvB and Ty:EnvC but not Ty:EnvA. Normal sheep sera reacted very weakly against pl alone but not at all against any of the fusion proteins, this is probably a consequence of pl being in excess in the control samples.
We have used PCR in combination with the yeast Ty-VLP expression system to express MVV gp135 as three overlapping fusion proteins (Ty:EnvA, Ty:EnvB and Ty:EnvC) each of which overlaps the adjacent fragment(s) by 50 to 60 amino acids. This is the first report of expression of any component of the MVV envelope glycoprotein. These fusion proteins have been used to epitope map the Ab response to gp135 in MVV infected sheep. The particles produced intracellularly by Ty:EnvA differ markedly from those of Ty:EnvB, Ty:EnvC and pl alone. This is unlikely to be a simple consequence of the size of EnvA as EnvB is approximately equal in size. gp135 of MVV contains a putative signal sequence between amino acids 78-100, (Sonigo et al., 1985) a region which is present in Ty:EnvA. As the signal sequence is proposed to function under normal conditions at an internal site + Fig. 4. Reactivity of sheep sera with fusion proteins. Fusion proteins were separated by SDS-PAGE, blotted and reacted with sheep sera as described. Numbers above each lane refer to the sheep from which serum was obtained. NSS is normal sheep sera. Arrows to the right of each panel indicate the position of the fusion protein. M, (kDa) of molecular-weight standards is given.
230
rather than at the amino terminus of the protein it may be capable of acting within the fusion protein to direct the protein to a cellular compartment where it is differentially processed resulting in the unusual particle formation observed. Sera from naturally and experimentally MVV infected sheep recognise the Ty:Env fusion proteins suggesting that the fusion proteins contain epitopes which are widely conserved between MVV strains. All sera tested reacted with at least one of the expressed regions of gpl35. 6/6 sera reacted with Ty:EnvC, 5/ 6 reacted with Ty:EnvB and 4/6 reacted with Ty:EnvA. Sera from sheep 2 and 754N both exhibit comparable strong reactions with Ty:EnvC, however, whilst sheep 754N also recognises Ty:EnvA and B to similar levels sheep 2 fails to recognise these fusion proteins. We take this as evidence that the differential reactivity of the sera by immunoblotting is due to differing immune responses to the three regions of gpl35 represented by the Ty:Env fusion proteins. These results allow the preliminary mapping of Ab responses to specific regions of MVV gp135. All sheep recognise Ty:EnvC which contains sequences spanning amino acids 490-670 and which overlaps with Ty:EnvB between amino acids 490-553. Since only 5/6 sera recognise Ty:EnvB it is logical to assume that there is at least one distinct epitope lying between residues 553-670 which is recognised by at least one sheep, that is sheep 848A which recognises Ty:EnvC but not Ty:EnvB. Similarly Ty:EnvB spans residues 247 to 553 and overlaps with Ty:EnvA between residues 247 and 31 I. As sheep 2 recognises Ty:EnvB and Ty:EnvC but not Ty:EnvA it must recognise epitopes within the region of Ty:EnvB from approximately 311 to 553. As all sheep which recognise Ty:EnvA also recognise Ty:EnvB but all sheep which recognise Ty:EnvB do not recognise Ty:EnvA it is probable that there is a third class of epitope which lies between residues l-247. We are therefore able to identify at least three distinct regions of MVV gp135 to which sheep mount an immune response. The approach outlined in this paper is directly applicable to a wide range of viral and non-viral proteins and represents a rapid and simple method of investigating antibody specificities.
Acknowledgements This work was supported by AFRC research grant No. AG 15/504 to RGD. DJR is supported by a Wellcome Trust Programme Grant. For a portion of this study NC was in receipt of a MRC studentship. We also thank Drs. S. Adams and N. Burns of British Bio-technology Ltd for useful discussions and for providing the Yeast Ty-VLP system.
231
References Adams, S.E., Senior, J.M., Kingsman, S.M. and Kingsman, A.J. (1988). Induction of HIV antibodies by Ty-HIV hybrid virus-like particles. In: Technological Advances in Vaccine Development, pub. Alan R. Liss, Inc. pp 1 I7 -126. Braun, M.J., Clements, J.E. and Gonda. M.A. (1987). The visna virus genome: evidence for a hypervariable site in the cn): gene and sequence homology among lentivirus envelope proteins. J. Viral. 61, 404664054. Buluwela, L., Forster, A., Boehm, T. and Rabbits, T.H. (1989) A rapid procedure for colony screening using nylon filters. Nucl. Acid. Res. 17, 452. Burns, N.R.. Gilmour, J.E.M., Kingsman, S.M., Kingsman, A.J. and Adams, S.E. (1991) Production and purification of hybrid Ty-VLPs. In: M. Collins‘(Ed), Methods in Molecular Biology, Vol. 8 (Viral Vectors and Transfection Techniques). Humana Press Inc. Clifton NJ. Chapter 25. Burns, N.R., Saibil, H.R., White, N.S., Pardon. J.F., Timmins. P.A., Richardson, S.M.H.. Richards, B.M., Adams, S.E.: Kingsman, S.M. and Kingsman, A.J. (1992) Symmetry, flexibility and permeability in the structure of yeast retrotransposan virus like particles. Embo J. I I, I l551164. Byers, B. and Goetsch, C. (1975) Behaviour of spindles and spindle plaques in the cell cycle and conjugation of Succharom_tws cerevisiar. J. Bact. 124, 5 I I 523. De Boer, G.F. (1970) Antibody formation in zwoegerziekte, a slow infection in sheep. J. Immunol. 104, 414 422. Dobson, M.J., Tuite, M.F., Roberts, N.A.. King, R.M., Kingsman, A.J.and Kingsman, SM. (1982) Conservation of high efficiency promoter sequences in Sacchurom_tws cerevisiae. Nucl. Acids. Res. IO, 2625-2637. Gilmour, J.E.M., Senior, J.M.. Burns, N.R., Esnouf, M.P., Gull, K., Kingsman, S.M., Kingsman, A.J. and Adams, S.E. (1989) A novel method for the purification of HIV-I p24 protein from hybrid Ty virus like particles (Ty-VLPs). AIDS 3. 717-723. Haase, A. (1985) Pathogenesis of lentivirus infections. Nature 322, l30- 136. Hinnen. A., Hicks, J.B. and Fink, G.R. (197X) Transformation of yeast. Proc. Natl. Acad. Sci. USA. 75, 1929-1933. Hirt, B. (1967) Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26, 365-m369. Jones, E.W. (1991) Tackling the protease problem in Saccharomyce.s cerevisine. Methods Enzymol. 194, 428453. Kingsman, SM., Cousens, D., Stanway, CA., Chambers, A., Wilson, M. and Kingsman, A.J. (1990) High efficiency yeast expression vectors based on the promoter of the phosphoglycerate kinase,gene. Methods Enzymol. 185, 329~-341. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Qhard, 0.. Dorit, R.L. and Gilbert, W. (1989) One-sided polymerase chain reaction: the amplification of cDNA. Proc. Natl. Acad. Sci. USA. 86, 567335677. Oliver, R.E., Gorham, J.R., Parish, SF., Hadlow, W.J. and Narayan, 0. (1981) Ovine progressive pneumonia: pathologic and virologic studies in the naturally occuring disease. Am. J. Vet. Res. 42. 1554-l 559. Querat, G., Audoly, G., Sonigo, P. and Vigne, R. (1990) Nucleotide sequence analysis of SAOMVV, a visna-related ovine lentivirus: phylogenetic history of lentiviruses. Virology 175. 434447. Reyburn, H.T., Roy, D.J., Blacklaws, B.A., Sargan, D.R. McConnell, L (1992) Expression of maedi-visna virus major core protein, ~25: development of a sensitive ~25 antigen detection assay. J. Virol. Methods 37, 3055320. Sanger, F., Nicklem, S. and Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA. 74, 5463 -5467.
232 Sargan, D.R., Bennet, I.D., Cousens, C., Roy, D.J., Blacklaws, B.A., Dalzie!, R.G., Watt, N.J. and McConnell, I. (1991) Nucleotide sequence of EVI, a British isolate of maedi-visna virus. J. Gen. Virol. 72, 189331903. Sigurdson, B. (1954) Maedi, a slow progressive pneumonia of sheep: an epizoological and pathological study. Brit. Vet. J. I IO, 2555270. Sigurdson, B., Palsson, P.A. and Crimson, H. (1957) Visna, a demyelinating transmissible disease of sheep. J. Neuropathol. Exp. Neurol. 16, 389403. Sonigo, P., Alizon, M., Staskus, K., Klatzmann, D., Cole, D., Danos, O., Retzel, E., Tiollais, P., Haase, A. and Wain-Hobson. S. (1985) Nucleotide sequence of the visna lentivirus: relationship to the AIDS virus. Cell 42, 369 382. van der Molen, E.J., Vecht. U. and Houwers, D.J. (1985) A chronic indurative mastitis in sheep associated with maedijvisna infection. Vet. Q. 7, 112-l 19.
reserved
VIRMET
01506
/ 0166-0934/93/$06.00
Use of recombinant gpl35 to study epitope-specific antibody responses to maedi visna virus N. Carey,
D.J. Roy and R.G. Dalziel
Department of Veterinary Pathology. University of Edinburgh, Summerhall, Edinburgh EH9 IQH (UK) (Accepted
27 January
1993)
Summary The envelope glycoprotein gp135 of the ovine lentivirus maedi visna virus (MVV) is the main target for neutralising antibody in vivo, however little is known about the specific regions of gp135 which elicit this neutralising response. We have used the polymerase chain reaction (PCR) to generate overlapping fragments of the gp 135 gene which have been expressed as fusion proteins in the yeast Ty-VLP system. These fusion proteins have been used to analyse the antibody response to gp135 in MVV infected sheep and we are able to identify at least three distinct regions of gp135 to which antibodies are directed. The approach described in this paper provides a rapid and simple method of generating overlapping fusion proteins with which to carry out epitope mapping studies. Maedi visna virus; Glycoprotein
gpl35;
Epitope-specific
antibody
Introduction Maedi visna virus (MVV) is the prototype virus of the family lentiviridae other members of which include human immunodeficiency virus (HIV), feline immunodeficiency virus (FIV) and equine infectious anaemia virus (EIAV). MVV infection in sheep results in the development of chronic pneumonitis (Sigurdson, 1954) and/or a progressive demyelinating disease (Sigurdson et al., 1957) either of which may become manifest years after the initial infection. In Correspondence to: R.G. Dalziel, Department Summerhall, Edinburgh EH9 lQH, UK.
of Veterinary
Pathology,
University
of Edinburgh,
222
some cases, chronic mastitis (van der Molen et al., 1985) or arthritis (Oliver et al., 1981) may also be apparent. In contrast to HIV which infects predominately CD4’ lymphocytes and cells of the monocyte/macrophage lineage, MVV infection in vivo is limited to monocyte/macrophages with a possible involvement of cells in the CNS (Haase 1985). Following experimental infection of sheep a low level of MVV specific antibody (Ab) is produced one to two weeks post infection. However it is not until approximately six months post-infection that neutralising Ab can be demonstrated (De Boer, 1970). The component of the MVV virion to which neutraiising Abs are directed is the envelope glycoprotein. The envelope precursor protein is cleaved post-translationally to yield a membrane associated gp46 non-covalently linked to gp135, which is exposed on the virion surface. Despite the importance of the in vivo Ab response to gp135 the regions of the protein which elicit this immune response have not been studied. The recent publication of the nucleotide sequence of several strains of MVV (Sonigo et al., 1985; Braun et al., 1987; Querat et al., 1990; Sargan et al., 1991) has facilitated the production of recombinant envelope protein with which to carry out epitope mapping studies. We have made use of the yeast Ty-VLP expression system in which a truncated form of the TYA gene product (designated pl) derived from the yeast retrotransposon (Ty) is expressed in yeast from a multicopy plasmid. The expressed pl self assembles to generate Virus Like Particles (VLPs). Insertion of a gene of interest downstream of TY.4 and subsequent expression generates a Ty(pl):fusion protein which is also capable of self assembly into a hybrid TyVLP. The protein of interest is thought to be expressed on the surface of the hybrid VLPs (Adams et al., 1988) which are then readily purified by density gradient centrifugation. We have expressed the entire MVV gp135 as three overlapping fusion proteins. This is the first report of the production of recombinant MVV gpl35. The recombinant Ty:Env fusion proteins were used to investigate the immune response to gp135 in MVV infected sheep. We show that both naturally infected and experimentally infected animals produce Abs which recognise the Ty:Env fusion proteins. We further demonstrate that the Ab response to regions of gp135 differs in individual animals and that we are able to use the Ty:Env fusion proteins to partially epitope map the Ab response to gp135 in sheep. The combination of the polymerase chain reaction (PCR) using tailored primers to generate and clone gene sequences and subsequent expression of these sequences using the Ty-VLP system results in a rapid and simple method for the expression of protein sequences. The use of overlapping fusion proteins of the type described to map Ab responses is a tractable and cost effective alternative to conventional peptide binding approaches.
223
Materials and Methods Virus and cells: MVV strain EVI (Sargan et al., 1991) was propagated on the WSCP cell line (a sheep choroid plexus cell line obtained from Dr. M. Dawson, Central Veterinary Laboratories, Weybridge, UK) in Dulbecco’s Modification of Eagle’s Medium (DMEM) supplemented with 2% fetal calf serum (FCS).
sheep 848A and 754N are Finn/Blackface crosses which were Sheep. experimentally infected with MVV strain EVI. Sera used in the experiments described were taken from sheep 754N at 18 months post infection and from 848A at 12 months post infection. Sheep 2, 74, 68 and GO27 were naturally infected Texels which displayed classical maedi lesions at post-mortem. Sera from these sheep were taken at post-mortem and the duration of infection was unknown. Sera from all the sheep used in this study reacted by immunoblotting with gp135 from EVl infected cell lysates. PCR amplfication
of MVV
envelope gp135
The overlapping fragments of gp135 outlined in Fig. 1 were generated by PCR using the conditions of Ohara et al. (1989). Extra-chromosomal proviral DNA purified from EVl infected WSCP cells (Hirt, 1967) was used as template. The published sequence of EVI was used to derive primers whose location on the gp135 gene is shown in Fig. 1 and whose sequence is outlined in Table 1. Each primer contained a BamHI site at its 5’-end to facilitate subsequent cloning steps. Cloning of gp135 ,fragments
PCR-generated env DNA was digested with BamHI, gel purified and inserted into the BamHI site of the phagemid pTZl9R (Pharmacia) for sequence analysis. Sequencing was carried out by the method of Sanger et al. (1977) using a Sequenase Version 2.0 Kit (USB) and the manufacturer’s protocol. The PCR fragments were also inserted into the BamHI site of the yeast/E.coli shuttle vector pOGS40 (for Env B and C) or pOGS42 (Env A) (Gilmour et al., 1989) and transformed into E. coli JM83. The vectors were generously supplied by British Bio-Technology Ltd, Oxford. In these plasmids expression of the TYA gene product is driven from the galactose inducible hybrid PGK-GAL (PAL) promoter (Dobson et al., 1982, Kingsman et al., 1990). Transformants were screened by colony hybridisation (Buluwela et al., 1989) using appropriate gp135 specific PCR products as probes and those which contained env sequences were selected. These clones were further screened by diagnostic restriction enzyme digestion to determine the orientation of the inserted sequences with respect to the pl sequences. Clones containing single copies of the inserted DNA in the correct orientation were designated p:Env:A.l, p:Env.B. 1 or p:Env.C.l. etc.
224
Plasmid DNA purified by alkaline lysis and chromatography using ~523 columns (5-Prime ~ 3 Prime Inc) was used for yeast transformation. Yeast transformation Yeast were cultured as described by Kingsman et al. (1990) and transformed using the method of Hinnen et al. (1978). The protease deficient Saccharomyces cerevisiae strain BJ2 168 (a, leu2 _, trpl ~ , ura3-52, prbl-1122, pep4-3, pcrl-407, ga12) (Jones 1991) was simultaneously transformed to leucine and uracil independence with the relevant recombinant pl :Env yeast expression plasmid and PUG 41 S, a plasmid which expresses Gal4 and allows upregulation of expression from the PAL promoter of the pOGS plasmids upon galactose induction (Burns et al., 1991). Following galactose induction whole cell extracts of transformed yeast (Burns et al., 199 1) were analysed for the presence of a protein species corresponding to the Ty:Env fusion protein by reducing SDSPAGE (Laemmli, 1970) and immunoblotting. All primary transformants derived using the same pl .Env plasmid expressed approximately equal amounts of fusion protein when analysed by immunoblotting. One transformant expressing each fusion protein was chosen for further analysis and designated T:EnvA, T:EnvB or T:EnvC. Fusion proteins obtained were termed Ty:EnvA, Ty:EnvB or Ty:EnvC. Purification
of Ty:Env fusion proteins
Ty:Env fusion proteins were purified from 4L yeast cultures 24 hours after galactose induction essentially as described by Reyburn et al. (1992). Briefly, cells were resuspended in 100 mM Tris-HCl pH 7.4; 2 mM EDTA; 140 mM NaCl (TEN buffer), disrupted by agitation with glass beads and the lysate centrifuged at 100 000 x g for 1 h at 4°C onto a 60% sucrose (in TEN) cushion. The material which banded at the interface was dialysed overnight against TEN buffer to remove sucrose and the Ty:Env particles purified by centrifugation (53 000 x g 3 h at 4°C) through 1545% linear sucrose gradients overlaid on a 60% sucrose cushion. The gradient was fractionated by positive displacement and fractions analysed by SDS-PAGE. Fractions containing hybrid Ty: EnvVLPs were stored at - 70°C. Immunoblotting Ty:Env proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes by electrophoresis at 0.4 mA cmy2 for 1 h in 25 mM Tris, 20% methanol using a semi-dry blotter. Unfilled sites on the nitrocellulose membrane were blocked by incubation in 5% milk powder/sterile phosphate buffered saline pH 7.0 (Milk powder/SPBS) for 30 min at room temperature. Sera were diluted in milk powder/SPBS and membrane strips incubated with this sera for 16 h at 4°C. After extensive washing bound
225
antibody was visualised using alkaline-phosphatase conjugated second antibody and NBT/BCIP (Nitro-blue Tetrazolium/bromo-chloro-indoyl phenol toluidine salt) as the detection system. Preparation
of yeast for electron microscop?.
Yeast were prepared for transmission electron microscopy following the method of Byers and Goetsch (1975). 60 nm sections were cut, post stained with uranyl acetate and lead citrate and examined using a Phillips EM400 transmission microscope.
Results Generation of Ty:Env fusion proteins Since little information on the domain structure of gp135 is available we chose primers (Table 1) to generate gp135 fragments with predicted sizes of 37.1 kDa (EnvA), 34.2 kDa (EnvB) and 20.1 kDa (EnvC) which overlap by approximately 60 amino acids (Fig. 1). The sizes of the gp135 fragments are significantly smaller than 43 kDa which is the maximum size of protein so far reported to be expressed in the Ty-VLP system (Burns et al., 1992). Sequencing of the PCR derived clones revealed that individual clones displayed variation from the published EVI sequence (Sargan et al. 1991) of between 2 and 7%. None of the clones sequenced contained a stop codon in the envelope reading frame. The observed variation was not introduced as a result of the PCR procedure as sequencing of subsequent PCR material generated using this cloned material as template did not display variation from the original clone (data not shown). Immuno-blot analysis of transformed cell lysates using a rabbit anti-Ty-VLP antiserum which detects the pl component of the fusion protein revealed that the pl protein, which has a predicted molecular weight of 43 kDa, migrates in TABLE I Nucleotide sequence of PCR primers. Primer
no.
1
5’ TCCCGGATCCTTTGTTTAGGATGGCAAGCAC 5’ TCCCGGGATCCGAGGAAACAGGACATGCAC 5’ TCCCGGGATCCGGATGTAATTGCTCAAGGTCAGG 5’ ACACCGGGATCCAGGACTCTTCCCTCGTCCTG3’ 5’ ACACCCGGGATCCCCCTATTCTGTCTAACGTGC 5’ACACCCGGGATCCTCACCAACCCTATGCCCCTC3’
2 3 4 5 6
Nucleotide no. 3’ 3’ 3’ 3’
59846005 67356753 74647485 6922.-6903 7652-7633 8000-798 I
Primers are numbered as in Fig. 1. The position on the MVV genome spanned by the primer is shown on the right. Numbering is as in Sargan et al. (1991). Bold letters denote nucleotides which are not part of the MVV sequence, with the RamHI site underlined.
226
670
490 -) 3
EnvC
* 6
Fig. I. Regions of gpl35 expressed in this study. The regions of gp135 referred to in the text as EnvA. EnvB and EnvC are shown diagrammatically. Numbers above the lines indicate the position of the amino acid residue which bounds the fragment. Numbering is as in Sargan et al. (1991). Arrows and numbers below the line refer to the primers described in Table 1 used to generate the PCR fragments.
0
z
w i; _i_
c”
-.I
94 67 -
. l
$ .
a4
43
30
Fig. 2. Immunoblot of puritied hybrid Ty:Env VLPs detected using VLP specific rabbit antisera pl VLPs were purified and immunoblotted as descrtbed in the text. Dots to the right of each track indicate the position of the fusion proteins. pl routinely migrates as a doublet and is indicated by two dots. M, (kDa) of molecular-weight standards is given.
221
SDS-PAGE with an apparent molecular weight (M,) of 55-60 kDa (Fig. 2). Based on the observed M, of pl, Ty:EnvA and Ty:EnvB have predicted molecular weights of approximately 92-97 kDa and 89-94 kDa respectively and Ty:EnvC a predicted molecular weight of approximately 75580 kDa. The observed M, values of the pl.fusion proteins were Ty:EnvA, 94-98 kDa, Ty:EnvB, 85 kDa and Ty:EnvC, 76 kDa (Fig. 2) which are consistent with those predicted. Purified hybrid VLPs were also analysed by SDS-PAGE and protein visualised by silver staining. The pl and Ty:EnvA protein bands were clearly visible but no bands corresponding to Ty:EnvB or C were distinguishable (data not shown), despite detection by immunoblotting. When analysed by SDS-PAGE the purified Ty:Env preparations all contain a number of contaminating bands which are present despite sucrose gradient purification. It was not therefore possible to determine accurately the amount of fusion protein using a standard protein assay system. Examination of the structure of the intracellular hybrid-VLPs by transmission electron microscopy revealed that VLPs containing pl alone are visible as circular structures approximately 15 to 35 nm in diameter with an electron dense outer region (Fig. 3b). similar structures are present in T:EnvB and T:EnvC cells although at much lower levels compared to control transformants expressing Ty-VLPs, a finding which is consistent with the low yield of fusion protein (data not shown). Fig. 3a shows that the VLPs formed after induction
Fig. 3. Electron micrographs of intracellular VLPs. Electron micrographs of thin sections of transformed yeast showmg size and morphology of native and hybrid VLPs. a) T:EnvA cells, b) Control yeast which produce native Ty-VLPs. VLPs are indicated by arrows scale bar = 2 /cm.
228
94 67
B
43
f 30
20
Ty:EnvA
TY
94 67
67
43
30
Ty:EnvB
Ty:EnvC
229
of T:EnvA cells are more numerous, larger (approximately 70 to 150 nm in diameter) and exhibit a heterogenous morphology. Possible reasons for the aberrant nature of the Ty-VLPs formed in T:EnvA cells are discussed later. Recognition
of Tv:Env jikon
proteins by serajiom
MVV infected sheep
The relative amount of pl protein in native Ty, Ty:EnvA, Ty:EnvB and Ty:EnvC preparations was determined by immunoblotting using a rabbit TyVLP specific antiserum (data not shown). Equal amounts of Ty:EnvA, Ty:EnvB and Ty:EnvC were loaded with native Ty loaded in slight excess. Sera from four naturally infected sheep and two which had been experimentally infected with the EVl strain of MVV were assayed by immunoblotting for reactivity with the Ty:Env fusion proteins. Sera from sheep G027,754N and 74 reacted with all three fusion proteins, with 754N reacting more strongly than GO27 or 74 (Fig. 4). Sera from sheep 848A reacts weakly with Ty:EnvC but not with Ty:EnvA or Ty:EnvB. None of these sera reacted with pl alone. Sera from sheep 2 and 68 show a faint reactivity against pl alone. Serum from sheep 68 reacts more intensely with all three Ty:Env fusion proteins suggesting strongly that this serum is also reacting specifically with gpl35 sequences. Similarily serum from sheep 2 is reactive with Ty:EnvB and Ty:EnvC but not Ty:EnvA. Normal sheep sera reacted very weakly against pl alone but not at all against any of the fusion proteins, this is probably a consequence of pl being in excess in the control samples.
We have used PCR in combination with the yeast Ty-VLP expression system to express MVV gp135 as three overlapping fusion proteins (Ty:EnvA, Ty:EnvB and Ty:EnvC) each of which overlaps the adjacent fragment(s) by 50 to 60 amino acids. This is the first report of expression of any component of the MVV envelope glycoprotein. These fusion proteins have been used to epitope map the Ab response to gp135 in MVV infected sheep. The particles produced intracellularly by Ty:EnvA differ markedly from those of Ty:EnvB, Ty:EnvC and pl alone. This is unlikely to be a simple consequence of the size of EnvA as EnvB is approximately equal in size. gp135 of MVV contains a putative signal sequence between amino acids 78-100, (Sonigo et al., 1985) a region which is present in Ty:EnvA. As the signal sequence is proposed to function under normal conditions at an internal site + Fig. 4. Reactivity of sheep sera with fusion proteins. Fusion proteins were separated by SDS-PAGE, blotted and reacted with sheep sera as described. Numbers above each lane refer to the sheep from which serum was obtained. NSS is normal sheep sera. Arrows to the right of each panel indicate the position of the fusion protein. M, (kDa) of molecular-weight standards is given.
230
rather than at the amino terminus of the protein it may be capable of acting within the fusion protein to direct the protein to a cellular compartment where it is differentially processed resulting in the unusual particle formation observed. Sera from naturally and experimentally MVV infected sheep recognise the Ty:Env fusion proteins suggesting that the fusion proteins contain epitopes which are widely conserved between MVV strains. All sera tested reacted with at least one of the expressed regions of gpl35. 6/6 sera reacted with Ty:EnvC, 5/ 6 reacted with Ty:EnvB and 4/6 reacted with Ty:EnvA. Sera from sheep 2 and 754N both exhibit comparable strong reactions with Ty:EnvC, however, whilst sheep 754N also recognises Ty:EnvA and B to similar levels sheep 2 fails to recognise these fusion proteins. We take this as evidence that the differential reactivity of the sera by immunoblotting is due to differing immune responses to the three regions of gpl35 represented by the Ty:Env fusion proteins. These results allow the preliminary mapping of Ab responses to specific regions of MVV gp135. All sheep recognise Ty:EnvC which contains sequences spanning amino acids 490-670 and which overlaps with Ty:EnvB between amino acids 490-553. Since only 5/6 sera recognise Ty:EnvB it is logical to assume that there is at least one distinct epitope lying between residues 553-670 which is recognised by at least one sheep, that is sheep 848A which recognises Ty:EnvC but not Ty:EnvB. Similarly Ty:EnvB spans residues 247 to 553 and overlaps with Ty:EnvA between residues 247 and 31 I. As sheep 2 recognises Ty:EnvB and Ty:EnvC but not Ty:EnvA it must recognise epitopes within the region of Ty:EnvB from approximately 311 to 553. As all sheep which recognise Ty:EnvA also recognise Ty:EnvB but all sheep which recognise Ty:EnvB do not recognise Ty:EnvA it is probable that there is a third class of epitope which lies between residues l-247. We are therefore able to identify at least three distinct regions of MVV gp135 to which sheep mount an immune response. The approach outlined in this paper is directly applicable to a wide range of viral and non-viral proteins and represents a rapid and simple method of investigating antibody specificities.
Acknowledgements This work was supported by AFRC research grant No. AG 15/504 to RGD. DJR is supported by a Wellcome Trust Programme Grant. For a portion of this study NC was in receipt of a MRC studentship. We also thank Drs. S. Adams and N. Burns of British Bio-technology Ltd for useful discussions and for providing the Yeast Ty-VLP system.
231
References Adams, S.E., Senior, J.M., Kingsman, S.M. and Kingsman, A.J. (1988). Induction of HIV antibodies by Ty-HIV hybrid virus-like particles. In: Technological Advances in Vaccine Development, pub. Alan R. Liss, Inc. pp 1 I7 -126. Braun, M.J., Clements, J.E. and Gonda. M.A. (1987). The visna virus genome: evidence for a hypervariable site in the cn): gene and sequence homology among lentivirus envelope proteins. J. Viral. 61, 404664054. Buluwela, L., Forster, A., Boehm, T. and Rabbits, T.H. (1989) A rapid procedure for colony screening using nylon filters. Nucl. Acid. Res. 17, 452. Burns, N.R.. Gilmour, J.E.M., Kingsman, S.M., Kingsman, A.J. and Adams, S.E. (1991) Production and purification of hybrid Ty-VLPs. In: M. Collins‘(Ed), Methods in Molecular Biology, Vol. 8 (Viral Vectors and Transfection Techniques). Humana Press Inc. Clifton NJ. Chapter 25. Burns, N.R., Saibil, H.R., White, N.S., Pardon. J.F., Timmins. P.A., Richardson, S.M.H.. Richards, B.M., Adams, S.E.: Kingsman, S.M. and Kingsman, A.J. (1992) Symmetry, flexibility and permeability in the structure of yeast retrotransposan virus like particles. Embo J. I I, I l551164. Byers, B. and Goetsch, C. (1975) Behaviour of spindles and spindle plaques in the cell cycle and conjugation of Succharom_tws cerevisiar. J. Bact. 124, 5 I I 523. De Boer, G.F. (1970) Antibody formation in zwoegerziekte, a slow infection in sheep. J. Immunol. 104, 414 422. Dobson, M.J., Tuite, M.F., Roberts, N.A.. King, R.M., Kingsman, A.J.and Kingsman, SM. (1982) Conservation of high efficiency promoter sequences in Sacchurom_tws cerevisiae. Nucl. Acids. Res. IO, 2625-2637. Gilmour, J.E.M., Senior, J.M.. Burns, N.R., Esnouf, M.P., Gull, K., Kingsman, S.M., Kingsman, A.J. and Adams, S.E. (1989) A novel method for the purification of HIV-I p24 protein from hybrid Ty virus like particles (Ty-VLPs). AIDS 3. 717-723. Haase, A. (1985) Pathogenesis of lentivirus infections. Nature 322, l30- 136. Hinnen. A., Hicks, J.B. and Fink, G.R. (197X) Transformation of yeast. Proc. Natl. Acad. Sci. USA. 75, 1929-1933. Hirt, B. (1967) Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26, 365-m369. Jones, E.W. (1991) Tackling the protease problem in Saccharomyce.s cerevisine. Methods Enzymol. 194, 428453. Kingsman, SM., Cousens, D., Stanway, CA., Chambers, A., Wilson, M. and Kingsman, A.J. (1990) High efficiency yeast expression vectors based on the promoter of the phosphoglycerate kinase,gene. Methods Enzymol. 185, 329~-341. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Qhard, 0.. Dorit, R.L. and Gilbert, W. (1989) One-sided polymerase chain reaction: the amplification of cDNA. Proc. Natl. Acad. Sci. USA. 86, 567335677. Oliver, R.E., Gorham, J.R., Parish, SF., Hadlow, W.J. and Narayan, 0. (1981) Ovine progressive pneumonia: pathologic and virologic studies in the naturally occuring disease. Am. J. Vet. Res. 42. 1554-l 559. Querat, G., Audoly, G., Sonigo, P. and Vigne, R. (1990) Nucleotide sequence analysis of SAOMVV, a visna-related ovine lentivirus: phylogenetic history of lentiviruses. Virology 175. 434447. Reyburn, H.T., Roy, D.J., Blacklaws, B.A., Sargan, D.R. McConnell, L (1992) Expression of maedi-visna virus major core protein, ~25: development of a sensitive ~25 antigen detection assay. J. Virol. Methods 37, 3055320. Sanger, F., Nicklem, S. and Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA. 74, 5463 -5467.
232 Sargan, D.R., Bennet, I.D., Cousens, C., Roy, D.J., Blacklaws, B.A., Dalzie!, R.G., Watt, N.J. and McConnell, I. (1991) Nucleotide sequence of EVI, a British isolate of maedi-visna virus. J. Gen. Virol. 72, 189331903. Sigurdson, B. (1954) Maedi, a slow progressive pneumonia of sheep: an epizoological and pathological study. Brit. Vet. J. I IO, 2555270. Sigurdson, B., Palsson, P.A. and Crimson, H. (1957) Visna, a demyelinating transmissible disease of sheep. J. Neuropathol. Exp. Neurol. 16, 389403. Sonigo, P., Alizon, M., Staskus, K., Klatzmann, D., Cole, D., Danos, O., Retzel, E., Tiollais, P., Haase, A. and Wain-Hobson. S. (1985) Nucleotide sequence of the visna lentivirus: relationship to the AIDS virus. Cell 42, 369 382. van der Molen, E.J., Vecht. U. and Houwers, D.J. (1985) A chronic indurative mastitis in sheep associated with maedijvisna infection. Vet. Q. 7, 112-l 19.