Duplicated genes within the variable right end of the ... - CiteSeerX

Journal of General Virology (1993), 74, 21255130. Printedin Great Britain

2125

Duplicated genes within the variable right end of the genome of a pathogenic isolate of African swine fever virus S o o p a y a h V y d e l i n g u m , 1 S a l l y A. Baylis,2~" Chrissy Bristow, 1 G e o f f r e y L . S m i t h 2 and Linda K. D i x o n ~* 1 A F R C Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Woking, Surrey GU24 ONF and 2 Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE, U.K.

The right variable region of the genome of a pathogenic strain of African swine fever virus (ASFV), Malawi LIL20/1, has been sequenced and 15 open reading frames (ORFs) identified by computer analysis. Eight of these ORFs were found to be similar to previously described ASFV ORFs and three of these belong to two previously described multiple gene families (MGF), 360 and 110. Four of the remaining five ORFs belong to a novel MGF, designated MGF 100, and the last ORF encodes a protein that is similar to the virus structural protein, p22. Copies of MGF 110 and the gene coding

for p22 have previously been characterized only at the left end of the ASFV genome. The organization of these genes suggests evolution by duplications, deletions and sequence transposition from one end of the genome to the other. Sequence comparisons of members of MGF 360 suggest that the Malawi LIL20/1 genome has undergone separate DNA rearrangements compared to the Ba71V genome. Lastly, one ORF was found to be similar to the myeloid differentiation primary response protein, M y D l l 6 and to the herpes simplex virus neurovirulence-associated factor ICP34"5.

Introduction

Gonz~ilez et al., 1986). The central genome region of 125 kb is conserved in length but regions near both the left and the right termini are variable, the latter varying in size from 11 to 16 kb and the former from 38 to 47 kb (Blasco et al., 1989). In addition, a variable region located between 93 and 97 kb from the left end has been reported (Sumption et aI., 1990). Two families of homologous genes have been identified at the variable ends of the genome (Gonz~ilez et aL, 1990; Almendral et al., 1990). Copies of multiple gene family (MGF) 360, so designated because the genes belonging to that family code for proteins of approximately 360 amino acids, are present at both ends of the genome. Copies of a second MGF, MGF 110, have so far been found only near the left end of the genome (Almendral et al., 1990). The similarities between the genomes of ASFV and poxviruses and the presence of several genes encoding virulence factors at the right end of the poxvirus genome (Goebel et al., 1990) led us to sequence an 11.3 kb fragment from the right end of the genome of a virulent ASFV isolate from Malawi (LIL20/1) which has a different restriction site map (Dixon, 1988) compared to those isolates (Ba71V, LIS57) from which sequences of MGF 360 and M G F 110 have previously been determined. Here the presence, near the right end of the LIL20/1 genome, of additional members of MGF 360 and MGF 110 is reported with an unrelated and novel MGF (MGF 100). One ORF with similarity to a virus

African swine fever virus (ASFV) is a large, icosahedral, cytoplasmic DNA virus which is responsible for an economically important disease of swine. African swine fever is currently enzootic only within sub-Saharan Africa, the Iberian peninsula and Sardinia (Wilkinson, 1984). The disease varies from hyperacute to acute and chronic to inapparent depending on the virus isolate involved (Hess, 1981). ASFV infects wild and domestic pigs as well as warthogs, bushpigs and ticks of the family Ornithodoros. The virus infects cells of the reticuloendothelial system in vivo and primarily monocytes and macrophages in vitro (Casal et al., 1984). The genome structure and replication strategy of ASFV are similar to those of poxviruses although the virus particle is structurally different (for reviews, see Vifiuela, 1985 and Costa, 1990). The genome consists of a linear double-stranded DNA molecule of 170 to 190 kbp with terminal inverted repetitions and terminal hairpin loops (Sogo et al., 1984;

t Presentaddress: Departmentof Cell Biology,WetlcomeResearch Laboratories, LangleyCourt, South Eden Road, Beckenham, Kent BR3 3BS, U.K. The nucleotidesequencedata discussedin this paper will appear in the EMBLdatabaseunder the accessionno. X71982. 0001-1748 © 1993 SGM

2126

S. V y d e l i n g u m and others

structural protein encoded near the left end of the Ba71V genome is also described. Another O R F has similarities over a carboxy-terminal domain with both a myeloid differentiation primary response protein ( M y D l l 6 ) (Lord et aL, 1990) and a herpesvirus-encoded neurovirulence-associated protein (McGeoch & Barnett, 1991).

Methods ASFV DNA was isolated from clone LMw23, which is part of a bacteriophage 2 library containing inserts from the genome of the ASFV Malawi LIL20/1 isolate (Dixon, 1988). The isolated ASFV DNA was sonicated, end-repaired and cloned into M 13 vectors. Singlestranded DNA was sequenced by the chain termination method (Sanger et al., 1977), using [~S]dATP. Sequences were read using a sonic digitizer and assembledby the computer programs DBUTIL and DBAUTO (Staden, 1982). ORFs were identified using the program ORFFILE (kindly provided by M. E. G. Boursnell, Cantab Pharmaceuticals, Cambridge, U.K.). Database searches were carried out using FASTA (Pearson & Lipman, 1988). Protein sequence comparisons, secondary structure and potential signal peptide predictions were carried out using the University of Wisconsin Genetics Computer Group package of programs (Devereux et al., 1984).

mapping indicated that a b o u t 2.5 k b p from the right end of the genome terminus, consisting of repeated sequences present within the terminal inverted repeats (K. J. Sumption, L. K. Dixon & P. J. Wilkinson, unpublished), is not contained within the insert in clone LMw23 (Dixon, 1988) and has therefore not been sequenced. O R F s coding for 65 amino acids or more were identified by computer analysis and the deduced sequences compared with each other and with protein sequence databases. O R F s were designated 1 to indicate the SalI fragment from which they are derived and numbered sequentially from the left. Those read towards the left end of the genome are designated L and those to the fight R. This analysis identified duplicated genes and O R F s encoding proteins similar to those in the databases and these are described here. A sequence of seven thymidines has been defined as the transcription termination signal for some early and late A S F V genes (Almaz~n et al., 1992, 1993). This sequence has been found downstream within 140 bp of five ORFs. Other ASFV transcription signals have yet to be defined. Protein p 2 2

Results Fig. 1 shows the position of the 11.3 kbp portion from the SalI 1 fragment contained in clone LMw23 (Dixon, 1988), which has been sequenced. Restriction enzyme site

a I

b I

c J

de I

f

(I

g hi f

j

k

II i

---~13~

1

q.~,.....i l i

l12R 15L ]'6LT[L7 ll0L

115 l14L

18L Fig. 1. Region of Malawi LIL20/1 genome sequenced in this paper. The upper panel shows the SalI restriction map of the Malawi LIL20/1 genome. An enlarged map of the ORFs encoded within the SalI 1 fragment is shown underneath. Arrows indicate the position and orientation of ORFs. ORFs described in the text are named.

IlOL . . . . . . . . . M p22 Ba71V M F N I K M T I S T

ll0L lL'21~ E 1 ~

YYP LLI

V L

V~IG IDItV I

DNFNTE VFLYYK

The O R F coding for protein p22, an external virion protein which is also expressed on the cell surface, was identified at the left end of the genome of the attenuated Spanish isolate Ba71V (Camacho & Vifiuela, 1991). The protein predicted to be encoded by O R F l l 0 L of the Malawi isolate has 58% amino acid similarity and 40.8 % identity with p22 o f Ba71V. An alignment of these sequences is shown in Fig. 2. The Malawi l l 0 L O R F codes for a protein of 167 amino acids, 10 amino acids shorter than the p22 of Ba71V which has nine extra amino acids at the N terminus. Despite this the hydrophobic character of the N-terminal sequence is conserved with stretches of 24 or 15 amino acids in BA71V p22 and Malawi t l 0 L respectively. In contrast to the p22 of Ba71V, the Malawi-encoded protein does not contain a predicted cleavage site downstream of the Nterminal hydrophobic region. Protein p22 of the Ba71V isolate has 10 cysteine residues, all of which are retained within the ll0L protein described here. Furthermore, five of the cysteines retain

FLSS KQQP

S

E

L

59

QB I

p22 Ba7IV ~ m D ~ M h l K

R N L ~ K Zm~l S K

IIOL $ I D H ~ Q p22 Ba71V P Q E V m E K

WGTD

S Lr~c Y[] S

T

S

F ~ N ~ R

Y . . VGDEK

ITPSPSES

PVLKYGKD

IAL

119

HA

177

Fig. 2. Sequence alignment of LIL20/1 ll0L and Ba71V p22 proteins. Sequence alignments were carried out using the program PILEUP and output was generated using PRETTYBOX (Devereux et al., 1984). Identical amino acids are boxed.

Duplicated genes within A S F V genome Table 1. Amino acid identities* between members of MGF 360 from two different A S F V isolates and ORFs 13R and l15R Ba71V K'362

L356

J319

D'311

D'363

33 74

38 61

38 53

37 47

89 37

36 52

13R l15R

115R ORF, which is the final ORF sequenced at the right end of the LIL20/1 genome, encodes an MGF 360 product that is more similar to that of the MGF 360 furthest to the left of the Ba71V genome (74 % identity within the Ba71V K'360 product) than to products of MGF 360 copies at the right end of the Ba71V genome. All the amino acids that are identical within all MGF 360 products previously described are present in the products of ORFs 13R and l15R.

LIS57

K'360

LIS LIS 375 382 31 30

2127

30 29

* The percentage identity was determined by pairwise alignment of amino acid sequences using GAP. Sequences of MGF 360 copies, i.e. K'360, K'362, L356, J319, D'311, D'363 from the Ba71V isolate and LIS375, LIS382 from the LIS57 isolate, were compared with the 13R and l15R products.

Multiple gene family 110 The amino acid identity between ORF l12R and members of MGF 110 described earlier (Almendral et al., 1990) varies from 42 to 22.6%, and the degree of similarity varies from 55% to 47% (Table 2). Eight cysteine residues were shown to be conserved within proteins of MGF 110 in Ba71V but only three of these are conserved in the product of ORF l12R (Fig. 3). Copies of MGF 110 have not been identified near the right termini of genomes of other ASFV isolates, but so far have been found near the left end of all isolates except one virus clone (de la Vega et al., 1990; Agiiero et al., 1990).

the motif CXsCXsCX4CX4C described before for the p22 molecule encoded by Ba71V. There are three potential Asn glycosylation sites present within the ll0L product of the Malawi isolate genome whereas none was identified within the p22 of Ba71V. The absence of a cleavage site downstream of the N-terminal hydrophobic region within the ll0L-encoded molecule suggests that it is expressed on the surface of infected cells rather than secreted. The ll0L protein may also be incorporated into the external envelope of the virion.

Multiple gene family 100 Multiple gene family 360

Four other ORFs (15L, 16L, 17L, 18L) (Fig. l), are similar to each other and therefore form a new MGF, designated MGF 100. The proteins encoded by these ORFs are aligned in Fig. 4 and range from 77 amino acids (ORF 16L) to 141 amino acids (ORF 15L) with ORFs 17L and 18L coding for proteins of 102 and 103 amino acids respectively. The amino acid identity between the products of the four ORFs varies from 52 % to as low as 18 % (Table 3). However, there are many conservative amino acid changes giving rise to a correspondingly higher percentage similarity (Table 3). These ORFs are contiguous and all read in the same direction towards the left end of the virus genome supporting a model for their evolution by gene duplication. Although the products of

Two ORFs (13R and l15R) encode proteins similar to those belonging to the MGF 360 described earlier (de la Vega et aL, 1990; Gonz~ilez et al., 1990). These code for proteins of 352 (13R) and 310 (115R) amino acids. Table 1 shows the percentage of identical amino acids when the products of ORFs 13R and 115R are compared with each of the MGF 360 copies on the Ba71V and LIS57 genomes. The product o f O R F 13R has between 33 % and 89% identity compared to different products of MGF 360 encoded by Ba71V. The most similarity is observed between the 13R product and the Ba71V D'311 product and these two ORFs are located at similar positions on their respective genomes. In contrast, the

Table 2. Amino acid similarity* and identity between members of MGF llO from two different isolates and ORF ll2R of Malawi LIL 20/1 Ba71V V104 Vl18 Identity

(%)

Similarity

(%)

LIS57

V270 V124

LIS 119-1

LIS LIS 119-2 121-1

LIS LIS 121-2 124-1

LIS 124-2

LIS 137

LIS 290

22.6

27.5

42

27

23

26

25

27

28

28

26

23

50

55

52

47

48

46

53

50

49

46

43

46

* Percentage amino acid identity and similarity was determined by pairwise alignment of sequences using GAP (Devereux et al., 1984).

S. Vydelingum and others

2128

LISI21-1 Ba71V 118 Ba71V 124 Ba71V 104 I12R

•

s ~

s~

v~qm

vu~lu

SL

~

LISI21-1 Ba71V 118 B a T l ¥ 124 Ba71V 104 l12R

Q

~ ~

~

mv~mA~z

~

z

il]x~

LIS121-1 Ba71V 118 Ba71V 124 Ba71V 104 lI2R

I

R

FKN

RF~QI~Q~,K

N~

40

Q

.

ili~i

~s

v~v~ A L q L ~ L : D S K

. H

WS KY F R ~ E W K K ~ L vP RL G R ~ I Y E K ~ L ATHLINA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Q

KN ~ ¥-IgM ~ER

LFIAK~LNTD

.... .... . . . .

EGN H'~ RDG F K * ..............

s B ~ m Q

H[qM~H~... R

~

~ ~ s ~

RIP . . . . . . . . . P. . . . . . . . . . FH . . . . . . . . . YAYNKIHYLQ

104 97 104

118 124 104 104

Fig. 3. Sequence alignment of LIL20/1 ll2R, Ba71V 118, 124, 104 and LIS57 121-1 (Lis121-1) encoded proteins. Sequence alignments were carried out using the program PILEUP and output was generated using PRETTYBOX (Devereux et al., 1984). Identical amino acids present in two or more of the sequences are boxed in black and conservative changes of amino acids among sequences are shown in shaded boxes.

17L 18L 15L LISIgO 16L

::: iiii:i

K P . . . . . . . . . . . . . . . R . . . . . . . . . . . . . . . . . . . . . . .

17L 18L 15L LIS100 16L

17L 18L 15L LISI00 16L

E

L

V

LK E . . . . . . . . . . . . . . . . . . . . . . . . V F ~ I ~ I R KLFYNKWLKY QEKKLEKRLR P I ~ I ~ Y R ..... RSRKI QEKKLRKSLK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SFLKKKY~V~

.

zAmGms-'aFm~ m N Z a E B F T M S D] ~ ~ N ~M

VI

F L~N

L

HD.I,~ Y R ~ [ ] L ~ v~l I C T

K T~E

~VVL'JHr'~NLIy~

K S R K E

* * Y N

.

.

.

.

.

.

.

.

.

.

.

.

.

.

S F S CF~TRT~I

VLBH[]E~SA AQJI~D LEarnS

L

.

.

.

.

.

.

.

.

.

.

M H L :~:~:~NmC Y D

|i!!

. . LLSF KLNF . . . . . . .

Q • Q E

zFT~ne~Is

emM]L~x~D

N IAL~E

A m ~ K N B F

FK[ME

21 59 39 0

81

119

53

.... 103 .... 141 Y N I * 124 K * . . 77

Fig. 4. Sequence alignment of members of MGF 100. Amino acid sequences of proteins encoded by ORFs 15L, 16L, 17L and 18L of the Malawi LIL20/1 isolate and of LIS100, translated from the sequence at the left end of LIS57, were aligned using PILEUP and output was generated using PRETTYBOX (Devereux et al., 1984). Identical amino acids present in three or more sequences are boxed in black and if the other sequences contain similar amino acids, they are shaded.

T a b l e 3. Amino acid identities and similarities* between

members o f M G F 100 15L 15L 16L lVL llSL

16L

17L

18L

LIS 100t

58 (34)

46 (19) 41 (18)

48 (22) 49 (20) 70 (51)

64 (44) 55 (29) 43 (19) 50 (24)

* Percentage identities and similarities were calculated by pairwise alignment of sequences using GAP (Devereux et al., 1984). The first number indicates percentage similarity, the number in parenthesis is percentage identity. ]" LIS100 is the newly identified member of MGF 100 translated from previously published sequence at the left end of the LIS57 isolate (de la Vega et al., 1990).

these O R F s are n o t r e l a t e d to o t h e r A S F V - e n c o d e d O R F s t h a t h a v e b e e n described, we h a v e identified an O R F within a p u b l i s h e d sequence f r o m the left e n d o f the LIS57 isolate g e n o m e (de la Vega et al., 1990) w h i c h is

similar to p r o d u c t s o f M G F 100. This LIS57 O R F ( L I S 100) is l o c a t e d b e t w e e n M G F 110 copies 137 a n d 290 n e a r the left end o f the g e n o m e a n d encodes a p r o d u c t o f 124 a m i n o acids• T h e L I S 100 p r o d u c t has the highest h o m o l o g y (44 % identity) a n d is m o s t similar in length to the 15L p r o d u c t . A l i g n m e n t s between e a c h o f these M G F 100 m e m b e r s were f o u n d to be statistically significant using F A S T A C H E C K , w h e r e a s w h e n the sequences were r a n d o m l y shuffled using S H U F F L E the a l i g n m e n t s were n o t statistically significant.

M y e l o i d differentiation primary response protein A n o t h e r O R F ( O R F l14L) has s o m e similarity w i t h a m y e l o i d differentiation p r i m a r y r e s p o n s e a n t i g e n ( M y D l l 6 ) ( L o r d et at., 1990; S u s s m a n et al., 1992)• M y D l l 6 is expressed in m y e l o i d cells only• O R F l14L e n c o d e s a p o l y p e p t i d e o f 184 a m i n o acids w i t h a n M r o f 21.3K which has 5 1 % i d e n t i t y a n d 6 1 % similarity o v e r a 40 a m i n o acid c a r b o x y - t e r m i n a l d o m a i n w i t h M y D 1 1 6

Duplicated genes within A S F V g e n o m e

2129

I14L

MSRRNKRSRR

RRKKPLNTIQ

PGPSKPSAQD

EPIKSVSHHS

SKIGTNPMLA

FILGGNEDLS

60

l14L

DDSDWDEDFS

LENTLMPLNE

VSLKGKHDSK

HFNKGFDNNT

ALHEVNTKWE

AFYSSVKIRQ

120

MyD116 114L

RDVKVYFATD

DILIKVREAD

DI

A

180

I14L

MEHE*Z84

LRKKLNP

Fig. 5. Sequencealignmentof a conservedcarboxy-terminalregionbetweenthe proteinencodedby ORF l14Land the mousemyeloid primary responsedifferentiationantigen, MyD116.An alignment betweenthe amino acid residuesencodedby ORF l14L and those between residues 567 and 607 of the 622 amino acid long MyD116product was producedusing PILEUP and output was generated using PRETTYBOX(Devereuxet al., 1984).Identical amino acids are boxedin black. (Fig. 5). This was found to be statistically significant using the PROSRCH program (Collins & Coulson, 1987), which shows that the observed alignment was one million times more likely than that predicted from a random match. Similarity between the MyD 116 domain and the herpes simplex virus neurovirulence-associated factor ICP34.5 has also been reported (McGeoch & Barnett, 1991).

Discussion A striking feature of the ASFV genome is the large number of duplicated genes encoded. Variation in the number of members of two MGFs, M G F 110 and M G F 360, accounts for at least some of the large length variations which occur close to the left end of the virus genome (de la Vega et al., 1990; Agtiero et al., 1990). Copies of M G F 360 have previously been described at both ends of the Ba71V genome (Gonzilez et al., 1990). We have now identified three other types of ORFs at the right end of the Malawi LIL20/1 genome that have previously been found only at the left end of ASFV genomes. These include one ORF which is similar to the gene for p22, a virus structural protein, one ORF which is similar to copies of M G F 110 and four ORFs which constitute a novel MGF, M G F 100. Gonzilez et al. (1990) proposed a model to explain the presence of members of M G F 360 at both ends of the virus genome. The basic tenets of the model are that gene duplications and sequence divergence of duplicated members within one end of the genome were followed by translocation of sequences from one genome end to the other. Disparity in the number of members of the M G F at opposite ends may be explained either by deletion of genes or further duplications following translocation. This model explains the presence of related genes at opposite ends of the genome and the observation that individual members of M G F 360 at opposite genome ends are more similar than individual members within one end of the genome. Such sequence transpositions from one end of the genome to the other also occur in

poxviruses (Moyer & Graves, 1981; Pickup et al., 1984), a group of viruses with many similarities with ASFV. The direction of transcription of copies of M G F 360, M G F 110, M G F 100 and p22 either towards or away from the genome terminus is the same at the right end of the Malawi LIL20/1 genome as at the left end of the Ba71V or LIS57 genomes. This is as expected from models proposed to explain poxvirus terminal sequence transpositions. Although one of the two (13R) M G F 360 genes present near the right end of the Malawi LIL20/1 genome is most closely related to the Ba71V M G F 360 gene, which has a similar genome location, the second (l15R) is most closely related to the Ba71V M G F 360 gene that is furthest to the left on this genome and is not very closely related to M G F 360 members located near the right terminus of the Ba71V genome. This suggests that the right terminal regions of the Ba71V and Malawi LIL20/1 genomes may have undergone deletions at different locations within the genome, leading to loss of different genes and thus the direct homologue of l15R and possibly other genes may have been deleted from the Ba71V genome. Based on sequence comparisons between M G F 360 members on the Ba71V genome, Gonz~ilez et al. (1990) also suggested that the two right terminal M G F 360 members have been deleted from the Ba71V genome. Our results support their prediction. ORF l14L, which encodes a product that has homology to MyD 116, is adjacent to l15R on the LIL20/1 genome and thus is one ORF which may also have been deleted from the Ba71V genome. The 114L product may be involved in the interaction of ASFV with its host. The protein might function by manipulating the differentiation process of myeloid cells, which are targets for ASFV replication, since the myeloid differentiation primary response genes are believed to be involved in the conversion of undifferentiated myeloid precursors into nonproliferating macrophages and granulocytes (Lord et al., 1990). A similar gene in herpes simplex virus types 1 and 2 has been shown to be a neurovirulence factor (McGeoch & Barnett, 1991; Chou & Roizman, 1990) that may interfere with the regulation of programmed

2130

S. Vydelingum and others

cell death in neuroblastoma cells (Chou & Roizman, 1992). The presence of MGFs within the three virus isolates for which sequence information is available clearly points to some important functions. It has earlier been shown that MGF l l0 is not involved in either virulence or pathogenicity in domestic pigs (Aguero et al., 1990) and it has been proposed that the products of this MGF may be part of the virus adaptation for growth in ticks (Dixon & Wilkinson, 1988). If that is true and given the genetic diversity within a given genus of ticks, such as Ornithodoros moubata (Walton, 1964), it is not surprising that the genes are present in multiple copies and are slightly different from each other. We would like to acknowledge financial support from M A F F and AFRC.

References AGUERO, M., BLASCO, R., WILKINSON, P. & VINUELA, E. (1990). Analysis of naturally occurring deletion variants of African swine fever virus: multigene family 110 is not essential for infectivity or virulence in pigs. Virology 176, 195-204. ALMAZA.N,F., RODRfGUEZ,J. M., ANDRI3S, G., P~REZ, R., VINUELA, E. & RODRIGUEZ, J. F. (1992). Transcriptional analysis of mnltigene family 110 of African swine fever virus. Journal of Virology 66, 6655 6667. ALMAZ.~N, F., RODRIGUEZ, J.M., ANGULO, A., VII~UELA, E. & RODRIGUEZ, J. F. (1993). Transcriptional mapping of a late gene coding for the p12 attachment protein of African swine fever virus. Journal ~?fVirology 67, 553-556. ALMENDRAL, J. M., ALMAZA.N,F., BLASCO, R. & VII~UELA, E. (1990). Multigene families in African swine fever virus: family 110. Journal of Virology 64, 2064-2072. BLASCO, R., DE Lg VEGA, R., ALMAZAN,F., AG/.)ERO, H. & VII~IUELA,E. (1989). Genetic variations of African swine fever virus: variable regions near the ends of the viral DNA. Virology 173, 251-257. CAMACHO,A. & V1NUELA,E. (1991). Protein p22 of African swine fever virus: an early structural protein that is incorporated into the membrane of infected cells. Virology 181,251-25% CASAL, I., ENJUANES, L. & VII~UELA, E. (1984). Porcine leukocyte cellular subsets sensitive to African swine fever virus in vitro. Journal of Virology 52, 37-46. CHOC, J. & ROIZMAN, B. (1990). The herpes simplex virus 1 gene for ICP34.5, which maps in inverted repeats, is conserved in several limited-passage isolates but not in strain 17 syn +. Journal of Virology 64, 1014-1020. CHOU, J. & ROIZMAN, B. (1992). The 71 34.5 gene of herpes simplex virus 1 precludes neuroblastoma cells from triggering total shutoff of protein synthesis characteristics of programmed cell death in neuronal cells. Proceedings of the National Academy of Sciences, U.S.A. 89, 3266-3270. COLLINS, J.F. & COULSON, A. F.W. (1987). Molecular sequence comparison and alignment. In Nucleic Acid and Protein Sequence Analysis." A PracticalApproach, pp. 323-358. Edited by M. Bishop & C. Rawlings. Oxford: IRL Press. COSTA, J. V. (1990). African swine fever virus. In Molecular Biology of Iridoviruses, pp. 247-270. Edited by G. Darai. Dordrecht: Kluwer Academic Publishers.

DE LAVEGA,I., VI]~UELA,E. & BLASCO,R. (1990). Genetic variation and multigene families in African swine fever virus. Virology 179, 234-246. DEVEREUX,J., HAEBERLI,P. & SMITHIES,O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Research 12, 387-395. DIXON, L.K. (1988). Molecular cloning and restriction enzyme mapping of an African swine fever virus isolate from Malawi. Journal of General Virology 69, 1683-1694. DIXON, L. K. & WILKINSON, P. J. (1988). Genetic diversity of African swine fever virus isolates from soft ticks (Ornithodoros moubata) inhabiting warthog burrows in Zambia. Journal of General Virology 69, 2981--2993. GOEBEL, S. J., JOHNSON, G. P., PERKUS, M. E., DAVIS, S. W., WINSLOW, J. P. & PAOLETTI,E. (1990). The complete sequence of vaccinia virus. Virology 179, 247-266. GONZALEZ, A., TALAVERA,A., ALMENDRAL,J.M. & VINUELA, E. (1986). Hairpin loop structure of African swine fever virus DNA. Nucleic Acids" Research 14, 683~6844. GONZALEZ,A., CALVO,V., ALMAZAN,F., ALMENDRAL,J. M., RAMIREZ, J. C., DE LA VEGA, I., BLASCO, R. & V1NUELA, E. (1990). Multigene families in African swine fever virus: family 360. Journal of Virology 64, 2073-2081. HEss, W. R. (1981). African swine fever: a reassessment. Advances in Veterinary Science and Comparative Medicine 25, 39-69. LORD, K. A., HOFFMAN-LIEBERMANN,B. • LIEBERMANN,D. A. (1990). Sequence of M y D l l 6 c D N A : a novel myeloid differentiation primary response gene induced by IL6. Nucleic Acids Research 18, 2823. McGEocH, D.J. & BARNETT, B.C. (1991). Neurovirulence factor. Nature, London 353, 609. MOVER, R. W. & GRAVES,R. L. (1981). The mechanism of cytoplasmic orthopoxvirus DNA replication. Cell 27, 391-40i. PEARSON, W. R. & LIPMAN, D. J. (1988). Improved tools for biological sequence comparison. Proceedings of the National Academy of Sciences, U.S.A. 85, 2444-2448. PICKUP, D. J., INK, B. S., PARSONS, B.L., HU, W. & JOKLIK, W. K. (1984). Spontaneous deletions and duplications of sequences in the genome of cow20ox virus. Proceedings of the National Academy of Sciences, U.S.A. 81, 6817-6821. SANGER, F , NICKLEN, S. & COULSON, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences, U.S.A. 74, 5463 5467. SOGO, J. M,, ALMENDRAL,J. M., TALAVERA,A. & VII~UELA,F,. (1984). Terminal and internal inverted repetitions in African swine fever virus DNA. Virology 133, 271-275. STADEN, R. (1982). Automation of computer handling of gel reading data produced by the shotgun method of DNA sequencing. Nucleic Acids Research 10, 4731-4751. SUMPTION,K. J., HUTCHINGS, G. H., WILKINSON,P. J. & DIXON, L. K. (1990). Variable regions on the genome of Malawi isolates of African swine fever virus. Journal of General Virology 71, 2331-2340. SUSSMAN,M. n., Lu, Z., KUTISH, G., ALFONSO, C. L., ROBERTS, P. & ROCK, D, C. (1992). Identification of an African swine fever virus gene with similarity to a myeloid differentiation primary response gene and neurovirulence associated gene of herpes simplex virus. Journal of Virology 66, 5586-5589. VINUELA, E. (1985). African swine fever virus. Current Topics in Microbiology and Immunology 116, 151-170. WALTON, G. A. (1964). The Ornithodoros "moubata" group of ticks in Africa. Control problems and implications. Journal of Medical Entomology 1, 53 64. WILKINSON, P.J. (1984). The persistence of African swine fever in Africa and the Mediterranean. Preventive Veterinary Medicine 2, 71-82.

(Received 7 April 1993; Accepted 1 June 1993)