Analysis of the bovine viral diarrhea virus genome for possible cellular insertions

VIROLOGY

189, 285-292

Analysis

(1992)

of the Bovine Viral Diarrhea

FENGXIA QI,* JULIA F. RIDPATH,t *Department

of Veterinary

Virus Genome for Possible

Cellular Insertions

TERRY LEWIS,“,’ STEVE R. BOLIN,t AND EUGENE S. BERRY*,*

and Microbiological Sciences, North Dakota State University, P. 0. Box 5406, Fargo, North Dakota 58 105; and tNational Animal Disease Center, USDA-ARS, P.O. Box 70. Ames, Iowa 500 10 Received December

2, 199 1; accepted April 7, 1992

Mucosal disease is the most severe disease resulting from bovine viral diarrhea virus (BVDV) infection in cattle. Two biotypes of BVDV may be isolated from animals with mucosal disease: cytopathic (cp) and noncytopathic (ncp). These “pairs” of cp/ncp viruses are often closely related and it has been suggested that the cp virus arises from a ncp virus by insertion of cellular RNA in the ~125 region of the BVDV genome. We have used four pairs of cp/ncp BVDV isolated from cattle with mucosal disease, to examine the genomic sequence of the region of the genome coding for the nonstructural protein ~125 (processed to ~541~80 in cp viruses) by PCR analysis and sequencing. We did not detect any cellular gene insertions in any of the four ncp viruses; however, we found a large duplication of the p80 gene and a ubiquitin gene insertion in three of the four cp isolates. Our results suggest that cellular RNA insertions in the ~125 region may contribute significantly to the cytopathogenicity of BVDV. However, this does not appear to be the only mechanism of cytopathogenicity as we did not detect any insertions or duplications in one of the cp viruses. Comparison of the DNA sequence in the p80 region revealed greater homology within the “pairs” than to NADL, which lend further support to the hypothesis that a cp virus is originated from a ncp virus 0 1992 Academic Press, Inc.

INTRODUCTION

that a cp virus arises from a ncp virus by some mutational event(s) (Corapi eta/., 1988). Meyers eta/. (1990, 1991) have proposed that the mutational event that induces a switch in biotype is an insertion of cellular RNA within the ~125 gene. The basis of this proposal was the identification of cellular RNA insertions in the two sequenced BVDV strains, NADL and Osloss (Collett et a/., 1988a; Meyers et a/., 1990). NADL has a 270-b insertion of an unknown cellular gene in the 3’ region of the p54 gene, and Osloss has a complete copy of the bovine ubiquitin gene inserted at the junction of the p54 and p80 genes (Meyers et al., 1990, 199 1). Both insertions are within the proposed protein processing sites (Collett et al., 198813; Wiskerchen et al., 1991). Further support of this hypothesis comes from the findings of Meyers et al. (1991), who partially sequenced a cp-ncp pair of BVDV isolates and found that the cp virus not only had a ubiquitin gene insertion, but also a duplication of the entire p80 gene. In contrast, the ncp counterpart had no insertion or duplication. However, DeMoerlooze et al. (1990) examined eight BVDV cp and ncp isolates for cellular RNA insertions in the ~125 region by polymerase chain reaction and DNA sequencing and found that none of the isolates had cellular RNA insertions. In the present study, we have examined four cp-ncp “pairs” of BVDV, by PCR and DNA sequencing, for possible cellular gene insertions or duplications of viral genome. Although cellular gene insertions or duplications were absent in the ncp viruses, we found ubiquitin gene insertions and p80 gene duplications in three

Bovine viral diarrhea virus (BVDV), hog cholera virus (HCV), and border disease virus (BDV) belong to the genus Pestivirus recently reclassified to the Flaviviridea family (Franchi et al., 1991). BVDV has a positivestrand RNA genome of approximately 12.5 kb (Collett et al., 1988a; Meyers et al., 1989). Two biotypes of BVDV, cytopathic (cp) and noncytopathic (ncp), are distinguished microscopically by their effects in cell culture. Biochemical differences between biotypes occur in the nonstructural protein, ~125, which is processed to proteins p54 and p80 by cp viruses (Lee and Gillespie, 1957; Gillespie et al., 1960, 1962; Howard et a/., 1987; Pocock et al., 1987; Donis and Dubovi, 1987; Corapi et al., 1988). While BVDV is associated with numerous disease problems, mucosal disease is the most severe form of BVDV infection. The development of mucosal disease requires congenital persistent infection with ncp BVDV and postnatal superinfection with specific cp BVDV (Brownlie et al., 1984; Bolin et al., 1985). In natural outbreaks of mucosal disease, the cp and ncp BVDV involved are often, but not always, antigenically similar (Corapi et al., 1988; Howard et al., 1987; Ridpath et a/., 1991). The finding of antigenically similar viruses of different biotypes in several cattle led to the hypothesis

1 Present address. Palo Alto VA Medical Center, GI Division, Palo Alto, CA 94304. ’ To whom reprint requests should be addressed. 285

0042.6822/92

$5 00

Copyright 0 1992 by Academic Press. Inc All rights of reproducllon I” any form reserved

286

QI ET AL.

of the four cp viruses. We present the amino acid sequence and the deduced genomic structure of portions of the ~125 region of these viruses, compare the homology within the pairs in the ~125 region and between the viruses and NADL, and discuss the possible mechanisms of RNA recombination and cytopathogenicity.

MATERIALS AND METHODS Viruses The BVDV pairs used in this study were VM/l90, NEB/21 10, TGAN/TGAC, and III-NC/III-C (ncp/cp, respectively, in all pairs). The cp viruses were purified by plaque isolation, and the ncp viruses were purified by limiting dilutions. The history of these viruses has been reported (Ridpath et a/., 1991).

RNA preparation Subconfluent MBDK cells were infected with either cp or ncp isolates at 10 m.o.i. Cells were harvested at 24 hr postinfection for cp isolates and at 48 hr for ncp isolates by freezing the monolayer at -70”. Ten milliliters of GIT buffer (4 M guanidine isothiocyanate, 0.4 M mercaptoethanol, 0.01 M EDTA, 0.05 M Tris, pH 8.0) was added to each 150-ml flask and the cells were broken by vigorous pipeting. A 1O-ml aliquot of the cell lysate was layered over a 2-ml cushion of 5.7 M CsCl and centrifuged at 36,000 rpm in a Beckman centrifuge for 16 hr at 15”. The RNA pellet was rinsed once with 70% ethanol, resuspended in 200 ~1 TE (10 mM Tris, pH 7.6, 1 mM EDTA), and precipitated with 70% ethanol in 0.3 M NaAc (pH 5.4) at -70” overnight. The RNA pellet was then washed with 70% ethanol, dried, resuspended in 100 ~1 TE buffer, and quantitated.

Oligonucleotide

primers

The sequence of the primers was chosen according to the conserved sequence between the published NADL and Osloss genomes. The sequences of the one cDNA primer and two PCR primers are listed below; nucleotide numbers are based on the NADL genome. 11: 5’-CTGTTGT-TGCTTTGGCAA-3’ (cDNA primer) (5703-5686) 10: 5’-GGACllTATGTACTAC-3’ (PCR primer) (4546-4564) 14: 5’-TCCCAATGATAACAGACATA-3’ (PCR primer) (7545-7564)

The PCR strategy, shown in Fig. 1, was based on the reports of Meyers et al. (1990, 1991) that gave locations of cellular mRNA insertions in BVDV genomes. PCR amplification with primers 11 and 14 only occurred when there was a duplication of the p80 (~125) region.

First strand cDNA synthesis, polymerase chain reaction (PCR), and DNA sequencing For the first-strand cDNA synthesis, 5 pg of total RNA was used. The reaction was performed according to the procedure provided with the Riboclone cDNA synthesis system (Promega, Madison, WI). Completed first-strand reaction mixture (25 ~1)was added to a PCR cocktail containing 25 pm of the PCR primer, 1X Taq buffer, 100 pm dNTPs, and 2 U of Taq DNA polymerase (Promega) in a total volume of 100 ~1. The reaction was run in an automatic tempcycler (Coy, Inc., Ann Arbor, MI) at the following profile: 94”/5’-50”/1’-72”/ l’, 1 cycle; 94”/1’-50”/1’-72”/1’, 30 cycles. The PCR product was isolated from a 1% agarose gel, purified with GeneClean (BiolOl, La Jolla, CA), blunt ended with T4 DNA polymerase, and cloned into either M 13mpl8 or M 13mpl9 at the Smal site. The procedure provided with the DNA Sequenase kit (United States Biochemical Co., Cleveland, OH) was followed for DNA sequencing.

RESULTS Sequence of the ~125 region As described in the legend for Fig. 1, the PCR product of primers 10 and 1 1 should be an 890-bpfragment encompassing nucleotides 4546-5703 of the NADL sequence if the viral gene lacks any cellular RNA insertions, but a larger product is expected if the viral genome has cellular RNA insertions. PCR with primers 1 1 and 14 should not generate any product unless the p80 region is duplicated. PCR with primers 10 and 1 1 produced a fragment of about 900 bp in all four viral pairs and a fragment of about 1200 bp in NADL (data not shown). PCR with primers 11 and 14 produced fragments of different sizes in three of the cp viruses (approximately 800 bp for 190, 900 bp for TGAC, and 1600 bp for Ill-C), but not in their ncp counterparts, nor in the viral pair NEB/21 10 (data not shown), suggesting that the ncp viruses do not have cellular RNA insertions and that three of the cp viruses have a duplication of the p80 gene. Both the 1O-l 1 and 1 l-14 fragments were cloned into M 13 in both orientations and sequenced. Because

GENOMIC

STRUCTURE

A.

e

x10

x11

x10

+

I P54

wthout Insert

#ll

f

Ndubq

With Insert p80

B. XII

#14

rc

WIthout Duplication P54

pso T

#14 + I

-w

With Duplication w

Ubq

w

FIG. 1. PCR Strategy for ~125 arrangement analysis. (A) Detectron of a possible cellular insertion at the junction of ~54 and ~80. Primer 1 1 (5686-5703) is 286 bp downstream of the ubiquitin insertion site in Osloss and is complementary to the viral genome. Primer 10 (454664564) is 450 bp upstream of the cellular insertion site in NADL. By synthesizrng cDNA from the vrral genome with primer 11 and PCR with primer 10, a 890.bp fragment should be generated if the virus does not have an Insert and a 1. l- to 1 Zkb fragment should be obtained if the virus has an insertion of cellular RNA. (B) Detection of possible duplications in the ~125 region. Primer 14 (7545-7564) is 450 bp upstream of the proposed 3’end of ~80. PCR using the first-strand BVDV cDNA as template and primers 11 and 14 should not generate any product since the two primers go in opposite directions. A product can be obtained only when the p80 gene is duplicated.

of the low fidelity of the Taq DNA polymerase, three clones from each virus were sequenced. Variations in one of the clones were found; however, each sequence presented here was from two clones and both strands. Therefore, we are confident that the sequences represent the authentic RNA sequences of the respective viruses. For convenience, only the deduced amino acid sequences are presented (Fig. 2). As shown in Fig. 2a, the sequences of the 10-l 1 fragments from all four pairs of BVDV did not show any insertions in the region reported for NADL and Osloss (Meyers et a/., 199 1); the sequences of the 1 l-l 4 fragments from two of the three cp isolates (190 and TGAC) showed a duplication of the p80 region as well as a ubiquitin insertion between the C-terminal region of ~125 gene (at aa 2524 for 190; at 2464 for TGAC) and the N-terminal region of the second copy of p80 (at aa 1680) (Fig. 2b). The sequence of Ill-C showed that, in addition to a ubiquitin insertion and a p80 duplication, the C-terminal half of the p54 gene was also duplicated (aa 1384 to 1679). If the first copy of the p80 gene is considered duplication, then the different lengths of the 1 l-l 4 fragment among the three cp viruses reflect the different

287

OF BVDV ~125

lengths of the duplication. Our sequence firmed our results obtained by PCR. Genomic organization of BVDV

data con-

of the mucosal disease pairs

Based on the strategy by which we designed our PCR primers (Fig. 1) and on the subsequent sequence data (Fig. 2), we deduced the ~125 genomic structures of three cp BVDV isolates (190, TGAC, and III-C) and their ncp counterparts (Fig. 3). The ncp viruses have no insertions or duplications in the ~125 region, while in each of the three cp viruses there is a p80 gene duplication as well as a cellular ubiquitin RNA insertion. Several features in Fig. 3 are notable. (1) The second copy of p80 started at the same site from its N-terminus in all three cp isolates (1680 aa/ nt, numerization in the text is based on the NADL sequence numbers) and was preceded by a complete ubiquitin monomer. (2) The first copy of p80 was truncated at different points from its C-terminus (aa 2524/nt 8418 for 190, aa 2464/nt 7754 for TGAC, and aa 2479/nt 7819 for Ill-C). (3) The ubiquitin insert in 190 and TGAC has two copies of ubiquitin monomer, one complete copy and a truncated copy from the N-terminus (aa 60 in 190 and 62 in TGAC; the total length of an ubiquitin monomer is 76 amino acids). (4) In 190 and TGAC, the N-terminus of the truncated ubiquitin insert was fused to the C-terminus of the first copy of ~80, while in III-C, the N-terminus of the single ubiquitin gene was fused to the C-terminus of a second copy of p54 (truncated from its N-terminus at aa 1384), and then the N-terminus of the second p54 was fused to the C-terminus of the first copy of ~80. Also of interest in Ill-C was that the ubiquitin insert had a direct repeat of four amino acids, KQLE, in the middle of the gene. DNA sequences

flanking

the ubiquitin

insert

In an attempt to understand the mechanism of RNA recombination, the sequences flanking the ubiquitin insert were compared. As shown in Fig. 4, certain patterns were obvious. Figure 4A shows sequences flanking the 3’ end of the ubiquitin insert. The sequences of the ubiquitin and BVDV p80 gene flanking the junction site are highly conserved among all isolates sequenced so far; the concensus sequence for BVDV is GGG/ACCTGCCGTGTG. In contrast, the sequence flanking the 5’ end of the ubiquitin insert varies from virus to virus (Fig. 4B and 4C), and no conser-vations are found except for Ill-C and Osloss. In Ill-C and Osloss, a single copy of the ubiquitin gene was inserted between the p54 and ~80. Although the ~125 region with the ubiquitin insert was duplicated in Ill-C, the se-

QI ET AL.

288

a

1479

1419

NWSMEEEESKGLKKFYLLSGRLRNLIIKHKVRNETVASWYGEEEX'YGI4PKIMTIIKASTL R .... D...RR ................ ................................. G......VFI................Q................V....R.C .. ....... G ..... ..FI................Q................F....R.C .. ....... ...... .G ..... ..FI................Q..........C.....F....R.C .. ..... ..G.......LI................Q................V....R.C .. ............... FI................Q................V....R.C .. ............. ..FI................Q................W...R.C .. ... ..G.G......VFI................Q................W...R.C ..

NADL 190 VM 111-c Ill-NC 2110 NE TGAC TGAN

1539

1460

NADL 190 VM 111-c Ill-NC 2110 NE TGAC TGAN

SKSRHCIICTVCEGREWKGGTCPKCGRHGKPITCGMSLADYKRIFIREGNFEGMC N.NK ......... ..K ........................................ N.NK ....... ..A.K....N...............T.....................T ................... N.NK.........A.K....N ............... N.NK.........A.K....N.L....D........T.....................T ................... N.NK.........A.K....N ............... T ................... N.NK ......... AKK....N ............... T ................... N.NK ......... AKK....N ............... N.NK.........A.K.NLFN...............T.....................-

NADL 190 VM 111-c Ill-NC 2110 NE TGAC TGAN

-_-_-----_____----__---------------------------------

---..- ..- ..- ..- -

------1659

1600

TEWAGCQRVGISPDTHRVPCHISFGSRMPFRQEYNGFVQYTAQLFLRNLPVLATKVKM --------------------------QW.........I..............I..A .... ____-----------------------..I ....... ...................... ----------------------------........................I ....... ____-----------------------..I ....... ...................... ----------------------------........................I ....... ----------------------------......S.............KD..I ....... ----------------------------.....HS.................I....L .. ____----------~~~~~~~~~~~~~~ I ....... ........................

NADL 190 VM 111-c Ill-NC 2110 NE TGAC TGAN

1719

1660

NADL 190 VM 111-c Ill-NC 2110 NE TGAC TGAN

LMVGNLGEEIGNLEHLGWILPAVCKKITEHEKCHINILDKLTAFFGI~RGTTP~PV ..v ........... V ......... ..D ........................ ......... ..V.D..Y.....................V...........V ........... ....... V ........... G ... ..v ........... ......... ..D .................. ..D..Y.....................V...........V ........... ......... ..v ........... V ......... ..D ........................ ......... V ......... ..v ........... ..D ........................ ......... ..V.D........................V............T .......... ....... ..V.D........................V...........V ........... ....... 1720

NADL 190 VM 111-c Ill-NC 2110 NE TGAC TGAC

1753

RFPTSLLKVRRGLETGWAYTHQGGISSVDHVTAG .... A ............................. .... A ............................. .... A ............................. .... A ............................. .... . ............................. .... A ............................. M .............. ................... .... A .............................

FIG. 2. (a) Sequence alignment of the 1O-l 1 fragments of BVDV isolates in the ~125 region. For convenrence. only the deduced amino acid sequence is shown, The numbers correspond to the published NADL sequence; the differences between NADL and the mucosal disease isolates are noted. (b) The deduced amino acid sequence of the 1 l-l 4 PCR fragments from the ~125 region of the cp isolates (19OC, TGAC. ILLC) surrounding the ubiquitin (Ubq) insertion. The (*) denotes the corresponding amino acid position in the NADL sequence and the positions in the ubiquitin monomer. The C-terminal portion of the ~125 region is represented by aa 2388-2524 for 190, aa 2391-2464 for TGAC, and aa 2388-2479 for III-C. The N-terminal part of the p80 is represented by aa 1680-l 757 for 190, aa 1680-l 769 for TGAC, and aa 1680-l 768 for Ill-C. The C-terminal part of the p54 of Ill-C is represented by aa 1384-l 679. The Ubq sequence is in bold and underlined, and the first amino acid of the Ubq monomer is double underlined.

GENOMIC

b

STRUCTURE

289

OF BVDV ~125

190 *2380 PMITDIYTTEDQRLEDTTHLQYAPNAIRTEGKENELKELAVGDLDKIMGSISDYASGGLDWRSQAEKMRSAPVFKE 2524**60 NVEAAKGYVQKFIDSLIENKETIIRYGLWGTHTALYKSIAARLGHETAFATLVIKWLAFGVQ~S~~ QIFvKTLTGKTITLKvEP

*1

76"+1680 SDTIENVXAlCIODKEGIPPDQQRLIFAGKQLEDGR!l'LSDYWIQKSSTLRLVLRLRGGGP

1757* AVCKKITEHEKCHVNILDKLTAFFGVMPRGTTPRAPVRFPTALLKVRRGLETGWAYTH~ISSMHVTAGKDLL

TGAC 2464+*62 l 2391 TDIYTIEDQRLEDTTHLQYAPNAIKTEGAETELKELASGDVGKIMGAISDYAAGGLEFVKSQAERIEHLGWIL~STIlI l 1 LvLRLRGGxQIFvKTLTGKTI-PSrn1ENvKA.K

IQDKEGIPPDQQRLTFAGKQLEDGRTLSDYNIOKBSTLRLVLR

76**1680 LROOGPAVCKKITEHERCHTNILDKLTAFFGIMPRGTTPRL l 1769 LVCDNMGRTRWC

111-c l 2380 PMITDIYTIEDQRLEDTTHLQYAPNTIRTEGKETELKELAVGDLDKIMGSISDYAFGGLDF~SQ~KIRSAPAF~~ 2479* AAKGYVQKFIDYLTLDFMYYMHRKVIEEISGGTNVMSRVIAALIELNWSHEEEGSKGLKKFFILSGRLRNLIIKHKVRNQ l 1304 TVASWYGEDEVYGMPKVMTIIRACTLNKNKHCIICTVCEARKWKGGNCPKCGRGKPITYGMTLADFEERHYKG 1679**1 NFEGPFRQEYNGFVQHT~VQLFLRNFPILATKVKMLMVGNLGEEIGDLEHLGWILRlKLINICTrMI[TITLGVILPSIITI

l 1768 AFFGVMPRGTTPRAPVRFPTALLKVRRGLETGWAYTHQGGISSMHVTAGKDLLVCDSMGRTRVG FIG. 2-Continued

quences flanking the ubiquitin insert are almost identical for the two viruses (Fig. 4C). A unique feature for III-C is that in addition to the two junctions at the 5’ and 3’ termini of the ubiquitin insert, there is a third junction between the truncated first copy of p80 and a truncated second copy of ~54. The sequence at the third junction is shown in Fig. 4D. Comparison of the 3’ p80 sequence with that of 190, TGAC, and CPl (Fig. 4B) did not reveal any homology either. The implications of the differences in the flanking sequences in regard to the mechanism of RNA recombination will be discussed later.

Homology of the ~125 region within between the viruses and NADL

viral pairs and

The hypothesis for the origin of cp BVDV implies that each pair of cp and ncp isolates has a very close relationship. To test this idea, the DNA sequences from the ~125 region within each pair of BVDV isolates were compared to each other and to NADL. As shown in Table 1, the nucleotide homology within each viral pair ranges from 90-98%, and the nucleotide homology of each virus to NADL varies from 82-860/o. The homology at the amino acid level ranges from 92-98’70 within

290

QI ET AL.

NCP

-

2404

-

I

w

I p54

Ukl

w

ELubq

Iwo

TGAC pso

0sloss pea

FIG. 3. Deduced genomic structure of the ~125 region. The genomic structures of the 3 cp mucosal disease isolates, 190, TGAC, and III-C, are shown with the structure of a ncp BVDV (without insert) and the cytopathic Osloss and CPl (Meyers et al., 1991); ncp BVDV does not have any insertion or duplication. Osloss has a Ubq insertion, but no duplication. TGAC, 190, III-C, and CPl have Ubq insertions as well as duplications. The different lengths of the first p80 region in 190, TGAC, I-K, and CPl represent the different points of C-terminal truncation. The different lengths of the first Ubq insertion indicate the different points of N-terminal truncation. The genomic positions of the amino acids surrounding the Ubq insert are indicated. The amino acid positions indicated are based on those of NADL.

the pairs and 89-93% between each virus and NADL. These results suggest that the viruses within each pair have a closer relationship to each other than to NADL. Since there are two copies of the p80 gene in three of the cp viruses, we compared the homology between the two copies within each virus and the homology between the second copy of p80 and the p80 of their ncp counterpart. As shown in Table 1, the homology between the two copies of p80 for 190, TGAC, and Ill-C is 100, 89, and 96%, respectively; the homology between the second copy of p80 and the ncp counterpart is 89.8, 79, and 99%, respectively. The high degree of homology in 190 and Ill-C suggests a common origin for the two copies of p80 gene. Possible reasons for the low homology in TGAC will be discussed below. DISCUSSION

We have cloned and sequenced part of the ~125 gene region of four pairs of BVDV that were isolated from mucosal disease animals and found that three of the four cp viruses have a ubiquitin gene insertion and a duplication of the p80 gene in the ~125 region. Our results extend the findings of Meyers et al. (1990, 1991) and provide further support to the hypothesis

that cytopathogenicity of BVDV correlates with cellular gene insertions or duplications in the ~125 region. However, our results also make the mechanism of cytopathogenicity of BVDV seemingly more complicated than before. For example, one of the cp isolates (21 10) did not show any cellular gene insertion or viral gene duplication in the region studied; also, PCR of the 21 10 cDNA with a ubiquitin-specific primer did not generate any product (data not shown), suggesting that mechanisms in addition to cellular gene insertion may also contribute to the cytopathic effect. Since we have not sequenced the entire ~125 region, cellular gene insertions in areas beyond our detection are also possible. Both cp and ncp biotypes may be isolated from cattle with mucosal disease. The only difference between the two viruses at the molecular level is the synthesis of p80 protein in cp virus-infected cells. It has been proposed that the presence of p80 is closely associated with the development of cytopathic effect. Wiskerchen eTal. (1991) recently verified that the p80 is a serine proteinase that can act autocatalytically or in

A. 3'-Ubiquitin TGAC 190 111-c CPl 0aloss

5’-BMV

CGTCTGAGGGGTGGT CGTCTGAGGGGTGGC CGCCTCAGAGGTGGG CGTCTGAGGGGTGGC CGTCTGAGGGGTAGT

BVDV consensus

sequence

pa0

GGACCTGCCGTGTGCAAGAAAATAACT GGGCCTGCCGTGTGTAAGAAGATTACA GGACCTGCCGTGTGCAAGAAAATMCT GGGCCTGCCGTGTGTAAGAAGATCACA GGGCCTGCCGTGTG CAAAAAGATTACT GGM!CTGCCGTGTG~!$tA G TA

B.

3'-BVDV 190 TGAC CPl

p80

ACCTTGGGTGGATCCTAAGG AAAAGTGGCTGGCCTTCGGG AGCTGCTAAGGGGTATGCTT

5'-Ubiquitin AAAGAGACGACCCTGCACCT GTCCAGAAAGAGTCGACCCT ACCAGGATAAGGAAGGCATT

C. 3'-BVDV 111-c 0s10ss

p54

CACCTAGGATGGATCTTMGG CACCTAGGATGGATCT

5'-Ubiquitin ATGCAGATCTTCGTGMGA ATGCAGATCTTCGTGAAAA

D. 3'-BVDV 111-c

p80

TACGTCCAAAAATTTATTGAT

5'-BVDV

p54

TACCTMCCCTGGACTTTA

FIG. 4. DNA sequences flanking the ubiquitin insertion. (A) The 5’-BVDV sequence flanking the ubiquitin insertion in TGAC, 190, IIIC, CPl , and Osloss. Note that this flanking sequence (the 5’ end of ~80) is highly conserved among all CP isolates sequences. (B) The BVDV sequence flanking the 5’-end of the ubiquitin insertion in 190, TGAC, and CPl. No conservation was observed in either the BVDV sequence (3’ end of ~125) or the ubiquitin insert. (C) The 5’-BVDV flanking sequence in Ill-C and Osloss. There is only a 2-bp difference at the 3’ end of p54 and no difference at the 5’end of ubiquitin insert. (D) The sequence at the junction of the first copy of p80 and the second copy of p54 in III-C. No conservation was observed compared to (B).

GENOMIC

STRUCTURE

OF BVDV ~125

291

TABLE 1 PERCENTHOMOLOGYOF THE ~125 REGIONSBETWEENNADL AND THE INDIVIDUALVIRUSES,WITHINTHE CP-NCP PAIRS, BETWEENTHE Two COPIES OF ~80 WITHINA CP VIRUS, AND BETWEENTHE SECOND COPY OF ~80 AND ITS NCP COUNTERPART Percent homology

NADL Nucleotide Amino acid Within pairs Nucleotide Amino acid 1st copy vs 2nd copy Nucleotide 2nd copy vs ncp Nucleotlde

190

VM

III-C

III-NC

2110

NE

TGAC

TGAN

86.0 93.0

82.7 91.0

81.6 91.8

82.1 90.6

82.9 92.2

83.6 91.0

83.6 91.2

81.6 89.3

89.8 92.1 100 89.8

98.7 97.9

94.8 97.7

89.6 94.1

96.2

89.6

99.2

79.0

trans on the downstream protein sequences and further characterized the ~125 protein as having a highly hydrophobic region at its N-terminus. Based on these data, it was proposed that the N-terminal hydrophobic region of ~125 may restrict it to the membrane of the endoplasmic reticulum, while the release of ~80, which lacks this hydrophobic region, may make it more accessible to some critical cellular proteins. The degradation of these cellular proteins then may cause cell damage and death. Previous studies of Ridpath and Bolin (1990) showed that superinfection of ncp-infected cells with a cp virus did not result in cytopathic effect although p80 was produced. This result suggests that p80 may be necessary, but not sufficient, to cause cytopathic effect, although other explanations are also possible. For example, the ncp virus may have some inhibitory effect on the replication of the superinfecting cp virus so that the p80 protein cannot sufficiently accumulate to cause cytopathic effect. It is also possible that ~125, which is synthesized by ncp virus, acts as a repressor for the function of ~80. This speculation is supported by the finding that the ~125 region of 4 cp isolates (TGAC, Ill-C, 190, and CPl (Meyers eta/., 1991)) are truncated from their C-terminus; TGAC has the largest deletion of 77 amino acids; Ill-C, CPl , and 190 have deletions of 61, 56, and 10 amino acids, respectively. The truncation of ~125 in the cp viruses may inactivate the repressor so that p80 can function. However, the recently revised genomic map of BVDV shifted the C-terminal border of ~125 to about 200 amino acids upstream (Collett eta/., 1991) thus leaving the ~125 of the cp viruses intact. Therefore, the repressor hypothesis seems to be unjustified. The possible correlation between the production of p80 and the generation of the cytopathic effect requires more experimental data.

Homology studies of the ~125 region within the pairs and between the individual virus and NADL revealed higher homology within the pairs than to NADL, suggesting a closer relationship within a pair of viruses. However, the degree of homology is not consistent among the four pairs; III-NC/III-C and NEB/2110 have higher homology (98.7 and 94.8%, respectively) than VM/l90 and TGAN/TGAC (89.8 and 89.69/o, respectively). Since the sequence we compared represented only 6.4% of the viral genome, it is too early to draw any conclusion as to whether all cp viruses originate from their ncp counterpart. To explore the origin of duplication, we also compared the homology between the two copies of p80 and between the second copy of p80 and the ncp counterpart. The homology of the two p80 copies in 190 is 100% although the homology within the pair is only 89.8%. This result suggests that the cp virus diverted from the ncp virus early during the course of viral evolution and that the duplication of the p80 is a more recent event. The opposite result is seen for the Ill-C/IIINC pair. Here, the second p80 has a higher homology to its ncp counterpart than the first copy (99% vs 969/o). This result suggests that the duplication of p80 occurred at the time when the cp virus was derived from its ncp counterpart. In contrast to the other three pairs, both homologies within the pair and between the two copies of p80 are lower for TGAUTGAN. If the same mutation rate is at work during viral replication, and the same selective pressure is exerted on every virus, then the low homology suggests that this cp virus may have persisted in the animal for a long period of time. In explaining the mechanism of cellular ubiquitin gene insertion and p80 duplication, Meyers et a/. (199 1) suggested that a template switch occurred during the negative-strand RNA synthesis, in which the

292

QI ET AL.

RNA polymerase jumped from its BVDV template to the ubiquitin mRNA (first template switch). After copying more than one copy of the ubiquitin gene, the polymerase jumped back to the BVDV template (second template switch), resulting in either an insertion or an insertion plus a duplication depending on the resumption site. Comparison of the sequence flanking the 3’end of the ubiquitin insert reveals a high conservation among all cp viruses, which suggests that the first template switch is probably guided by homologous recombination. In contrast to the 3’ end of the insert, the sequence flanking the 5’ end of the insert is completely different from virus to virus. This could mean that the second template switch may happen as a rare event of heterologous recombination. However, the sequence flanking the 5’ border of the ubiquitin insert is highly conserved between Ill-C and Osloss, suggesting that the second template switch may also be based on homologous recombination, at least in these two viruses.

ACKNOWLEDGMENTS The nucleotide sequence data reported in this paper have been submitted to the GenBank nucleotide sequence database and have been assigned the accession numbers M87805-M87815. The authors thank D.L. Berryhill and T.L. Glass for critical review of the manuscript. This research was supported In part by USDA Grant 91-37204-6763 (E.S.B.). Published with the approval of the director of the North Dakota Agricultural Experiment Station as Journal Artcle Number 2012.

REFERENCES BOLIN, S. R., MCCLURKIN, A. W., CUTLIP, R. C., and CORIA, M. F. (1985). Severe clinical disease induced in cattle persistently infected with noncytopathic bovine viral diarrhea virus by superinfection with cytopathic bovine viral diarrhea virus. Am. /. Vet. Rest 46, 573-576. BROWNLIE,J., CLARK, M. C., and HOWARD, C. J. (1984). Experimental production of fatal mucosal disease in cattle. Vet. Rec. 114, 535536. COLLE~, M. S., WISKERCHEN,M. A., WELNIAK, E., and BELZER, S. K. (1991). Bovine viral diarrhea virus genomic organization. Arch. Viral. S3, 19-27. COLLECT, M. S., LARSON, R., GOLD, C., STRINCK, D., ANDERSON, D. K.. and PURCHIO,A. F. (1988a). Molecular cloning and nucleotide sequencing of the pestivirus bovine viral diarrhea virus. virology 165, 191-199. COLLETT, M. S., LARSON, R., BELZER, S. K., and RETZEL, E. (1988b).

Proteins encoded by bovine viral diarrhea virus: The genomic organization of a pestivirus. Virology 165, 200-208. CORAPI, W. V., DONIS, R. O., and DUBOVI, E. F. (1988). Monoclonal antibody analysis of cytopathic and noncytopathic viruses from fatal bovine viral diarrhea virus infections. /. Viral. 62, 2823-2827. DE MOERLOOZE,L., DESPORT,M., RENARD,A., LECOMTE,C., BROWNLIE, J., and MARTIAL, J. A. (1990). The coding region for the 54.KDa protein of several pestiviruses lacks host insertions but reveals a “zinc finger-like” domain. Virology 177, 812-815. DONIS, R. O., and DUBOVI, E. J. (1987). Differences in virus-induced polypeptides in cells infected by cytopathic and noncytopathic biotypes of bovine virus diarrhea-mucosal disease virus. Virology 158, 168-173. FRANCHI, R. I. B., FRANQUET, C. M., KMUSON, D. L., and BROWN, F. (Eds.) (1991). Classification and nomenclature of viruses. Arch. Virol. S2, 228. GILLESPIE,J. H., BAKER,J. A., and MCENTEE, K. (1960). A cytopathogenie strain of virus diarrhea virus. Cornell Vet. 50, 73-79. GILLESPIE,J. H., MADIN, S. H., and DARBY, N. B. (1962). Cellular resistence in tissue culture, induced by noncytopathogenic strains to a cytopathogenic strain of virus diarrhea virus of cattle. Proc. Sot. Exp. Biol. Med. 110, 248-250. HOWARD, C. J., BROWNLIE,J., and CLARKE, M. C. (1987). Comparison by the neutralization assay of pairs of noncytopathogenic and cytopathogenic strains of bovine diarrhea virus isolated from cases of.mucoial disease. Vet. Microbial. 13, 361-369. LEE, K. M., and GILLESPIE,J. H. (1957). Propagation of virus diarrhea virus of cattle in tissue culture. Am. /. Vet. Res. 18, 953. MEYERS, G., RUMENAPT,T., and THIEL, H. J. (1989). Molecular cloning and nucleotide sequencing of the genome of hog cholera virus. Virology 171 I 555-567. MEYERS, G., RUMENAPF,T., and THIEL, H. J. (1990). Insertion of ubiquitin-coding sequence identified in the RNA genome of a togavirus. In “New Aspects of Positive-Strand RNA Viruses” (M. A. Brinton and F. X. Heinz, Eds.), pp. 25-30. American Society for Microbiology, Washington, DC. MEYERS, G., TAUTZ, N., DUBOVI, G. J., and THIEL, H. J. (1991). Viral cytopathogenicity correlated with integration of ubiquitin coding sequences. Virology 180, 602-616. POCOCK, D. H., HOWARD, C. 1.. CLARKE, M. C., and BROWNLIE, J. (1987). Variation in the intracellular polypeptide profiles from different isolates of bovine virus diarrhea virus. Arch. Viral. 94, 43-53. RIDPATH. J. F., and BOLIN, S. R. (1990). Viral protein products in homogenous mixed infection of cytopathic and noncytopathic bovine viral diarrhea virus. Arch. Viral. 111, 247-256. RIDPATH, J. F., LEWIS, T. L., BOLIN, S. R., and BERRY, E. S. (1991). Antigenic and genomic comparison between noncytopathic and cytopathic bovine viral diarrhea viruses isolated from cattle that had spontaneous mucosal disease. /. Gen. Viro/ogy72,725-729. WISKERCHEN,M.. BELZER.S. K.. and COLLET, M. S. (1991). Pestivirus gene expression: Protein ~80 of bovine viral diarrhea virus is a proteinase involved in polyprotein processing. Virology 184, 341350.