Previous Article | Next Article ![]()
Journal of Virology, October 2003, p. 10339-10347, Vol. 77, No. 19
0022-538X/03/$08.00+0 DOI: 10.1128/JVI.77.19.10339-10347.2003
Plum Island Animal Disease Center, Agricultural Research Service, U.S. Department of Agriculture, Greenport, New York 11944,1 Area of Virology, School of Veterinary Sciences, University of Buenos Aires, 1427 Buenos Aires, Argentina,2 Veterinary Department, University of Viçosa, Viçosa, MG, Brazil 36571-000,3 Veterinary Department, University of Santa Maria, Santa Maria, RS, Brazil 97105-9004
Received 16 April 2003/ Accepted 2 July 2003
|
|
|---|
2 subgroup of herpesviruses in genomic organization and gene content. BHV-5 is very similar to BHV-1, the etiological agent of infectious bovine rhinotracheitis, as reflected by the high level of amino acid identity in their protein repertoires (average, 82%). The highest similarity to BHV-1 products (
95% amino acid identity) is found in proteins involved in viral DNA replication and processing (UL5, UL15, UL29, and UL39) and in virion proteins (UL14, UL19, UL48, and US6). Among the least conserved (
75%) are the homologues of immediate-early (IE) proteins BICP0, BICP4, and BICP22, the three proteins being longer in BHV-5 than in BHV-1. The structure of the BHV-5 latency-related (LR) region departs markedly from that of BHV-1 in both coding and transcriptional regulatory regions. Given the potential significance of IE genes and the LR region in virus-neuron interactions, it is likely these differences contribute to BHV-5 neuropathogenicity. |
|
|---|
Both BHV-5 and BHV-1 are neurotropic viruses, but only BHV-5 is capable of significant replication in the central nervous system (CNS) and induction of neurological disease (5, 10). Outbreaks of meningoencephalitis caused by BHV-5 have been reported in Australia (37), North and South America (8, 15, 42, 77), and Europe (9, 66). The course of the disease after experimental infection with BHV-5 depends on the virus isolate, the route of inoculation, and the immunological status and age of the animal. Calves up to 4 months of age are most susceptible. During the first week following intranasal inoculation of virus, the animals either present signs of mild rhinitis and conjunctivitis or they remain asymptomatic (5, 16, 66, 73). At this point, animals can recover from infection or, alternatively, progress to neurological disease and die. Regardless of the appearance of clinical neurological disease in infected calves, BHV-5 invades the CNS and causes various degrees of pathology (16, 65). The neural route followed by BHV-5 to invade the CNS in cattle has not been defined, although both the trigeminal and the olfactory pathways have been implicated (6, 65). In a rabbit model for BHV-5, the olfactory pathway is the main route for neural dissemination (19, 57).
At times when no infectious virus can be isolated from peripheral sites, surviving animals exhibit BHV-5 viral sequences in their trigeminal ganglia (TG), indicating that, as in BHV-1, BHV-5 remains latent in the TG (5, 16, 65, 91). Latent virus can be reactivated after treatment of latently infected animals with dexamethasone (10, 16). Although recrudescence of clinical disease has been observed after experimental virus reactivation, its occurrence in nature remains unknown (73). Previous vaccination of cattle against BHV-1 resulted in protection against BHV-5-induced neuropathology and clinical signs of disease (16). This is partially explained by the cross-reactivity of induced BHV-1 neutralizing antibodies (24). Vaccination with BHV-1, however, does not prevent establishment of latency by BHV-5 (16).
The marked neuroinvasiveness (and often neurovirulence) displayed by BHV-5 contrasts with the inability of BHV-1 to invade the CNS and cause neurological disease to any significant degree. During infection with other neurotropic herpesviruses (e.g., pseudorabies virus [PRV] and herpes simplex virus [HSV]), neuroinvasiveness and local virus spread in the CNS are properties which largely rely on certain viral envelope glycoproteins (46, 67). However, other viral functions could contribute to a successful infection of the CNS.
Comparative genomics has proven useful in identifying genes involved in virulence. The complete BHV-1 genomic sequence, a composite including sequences of five different virus strains, is available (GenBank accession number AJ004801). However, less than 15% of the BHV-5 genome (mostly representing envelope glycoproteins) has been sequenced (1, 18, 20, 22, 31, 40, 64, 80, 86). Here we present the complete sequence of a neurovirulent strain of BHV-5, with analysis and comparison to BHV-1.
|
|
|---|
To obtain viral DNA, BHV-5-infected MDBK cells were pelleted, washed in NTE buffer (10 mM Tris [pH 8.0], 150 mM NaCl, 5 mM EDTA), resuspended in KTE buffer (10 mM Tris [pH 8.0], 10 mM KCl, 5 mM EDTA), and incubated in KTE buffer containing 0.25% ß-mercaptoethanol and 10% Triton X-100 for 10 min on ice. After centrifugation at 3,000 rpm (Sorvall), 5 min, 4°C, virus in the supernatant was pelleted onto a 30% sucrose cushion (13,000 rpm, 1 h, 4°C), resuspended in Tris-EDTA (TE; pH 8.0), incubated in the presence of sodium dodecyl sulfate (0.5% [wt/vol]) and proteinase K (10 mg/ml) for 2 h at 37°C, and phenol extracted. Viral DNA was precipitated with ethanol and resuspended in TE buffer (pH 8.0). Random 1- to 8-kbp DNA fragments were obtained by incomplete enzymatic digestion with AciI endonuclease (New England Biolabs, Beverly, Mass.). DNA fragments of 1.5 to 3 kbp were isolated after separation in size exclusion columns (Clontech), cloned into the dephosphorylated AciI site of pUC19 plasmids, and grown in Escherichia coli DH10B cells (Gibco BRL, Gaithersburg, Md.). Plasmids were purified by alkaline lysis as instructed by the manufacturer (Eppendorf 5 Prime; Boulder, Colo.). DNA templates were sequenced from both ends with M13 forward and reverse primers and from selected plasmids with transposon insertion (EZ::TN<KAN-2> transposon insertion kit; Epicentre Tech., Madison, Wis.), using dideoxy-chain terminator sequencing chemistries (82) and an Applied Biosystems PRISM 3700 automated DNA sequencer (PE Biosystems, Foster City, Calif.). Bases were called from chromatogram traces with Phred (35), which also produced a quality file containing a predicted probability error at each base position.
DNA sequence analysis.
DNA sequences were assembled with Phrap (34), using the quality files and default settings to produce a consensus sequence which was manually edited with Consed (41). An identical sequence was assembled using the Cap3 assembler with quality files and clone length constraints (45). The final DNA consensus sequence represented an average eightfold redundancy at each base position. Gap closure was achieved by primer walking of gap-spanning clones and sequencing of PCR products. A total of 7,835 usable traces were assembled into a 127,162-bp contig by bidirectional sequencing of random clones and 61 PCR products, including those that crossed the terminal junction sequences from concatemeric replicative intermediates. The assembled contig had an estimated error rate of <0.03/10 kbp and showed no evidence of polymorphism using Polyphred analysis (34), including within the single, complete unique short (US) repeat sequence which assembled with double average redundancy at each base position, consistent with a bimolar representation of repeat sequences derived from two identical copies, as found in other alphaherpesviruses. Thus, the assembled contig contained 570 bp of the terminal repeat (TR) at the left terminus, all of the unique long (UL), internal repeat (IR), and US sequences, and 738 bp of the TR at the right terminus. IR and TR sequences were also assembled separately with clones containing the unique-repeat junctions and overlapping clones, using length constraints and position as provided by the computer assembly programs. These assemblies were manually joined at the unique-repeat boundaries, thus providing the complete genome. For descriptive purposes, we have presented BHV-5 in a linearized fashion as described by Dolan et al. (29). Genome DNA composition, structure, repeats, and restriction enzyme patterns were analyzed as previously described (2). Open reading frames (ORFs) encoding proteins of
60 amino acids with a methionine start codon (88) were evaluated for coding potential using the Hexamer (ftp.sanger.ac.uk/pub/rd) and Glimmer (81) computer programs. Other criteria included similarity to other herpesvirus and compact gene arrangements with little gene overlap. Homology searches were conducted using BLAST (3), PSIBLAST (4), FASTA (72), BLIMPS (92), and HMMER (87) programs with the Prosite, Pfam, Prodom, Sbase, Blocks, Domo, and GenBank databases (14). GCG (26), MEMSAT (54), and SAPS (12) programs were used for gene analysis. The coding potential and splicing patterns in the latency-related (LR) region of BHV-5 and BHV-1 were analyzed with Glimmer, Hexamer, Splice (a neural network program for eukaryotic splice site prediction; ftp://genome.lbl.gov/pub/reese/SPLICE), NNPP (eukaryotic promoter prediction; ftp://genome.lbl.gov/pub/reese/NNPP), TIGR GeneSplicer (74), HMM gene trained on human and Drosophila gene sets (55), and Eponine for transcription start site detection (30). We also compared BHV-1 mapping data to the BHV-5 LR region.
Nucleotide sequence accession number. The BHV-5 genome sequence has been deposited in GenBank under accession no. AY261359.
|
|
|---|
The BHV-5 origins of DNA replication (ORI) are located in the repeat regions from nucleotide positions 113206 to 113418 and 129595 to 129807. ORI sequences consist of two imperfect AT-rich direct repeats (ORIa and ORIb) which contain herpesvirus consensus sites for the origin-binding protein (84). As in BHV-1, an additional truncated repeat (ORIc) is located 130 bp downstream from ORIb.
Gene characterization.
BHV-5 contains 72 genes (Table 1), of which 68 are present as single copies within the unique regions and 2 initiate and are completely located within the repeat regions (BICP4 and BICP22). BHV-5 proteins are most similar to homologues from BHV-1, averaging 82% amino acid identity. All BHV-5 ORFs are present in BHV-1; however, BHV-5 lacks a homologue of UL0.5. Among nonbovine herpesviruses, ORFs of equine herpesviruses (EHV) 1 and 4 are the most similar to those of BHV-5 (28 to 69% amino acid identity). The similarity in gene arrangement and the high percentage of amino acid identity between BHV-5 and previously sequenced alphaherpesviruses support the inclusion of BHV-5 in the
2 subgroup of herpesviruses (62), as was previously suggested by envelope glycoprotein B sequence analysis (80).
|
View this table: [in a new window] |
TABLE 1. Characterization of BHV-5 genes
|
95% amino acid identity) involving viral DNA replication and processing (UL5, UL15, UL29, and UL39), tegument (UL14 and UL48), and capsid (UL19). Compared with nonbovine herpesviruses, all BHV-5 capsid proteins and six of eight proteins involved in viral DNA replication or processing are the most conserved (
60%).
BHV-5 UL49, UL44, UL24, UL11, UL3.5, UL3, UL0.7, LR, and BICP0 are the least conserved UL genes (
75%) relative to BHV-1 and are discussed below. Homologues of UL49 and UL3 in HSV-1 and of UL49 in BHV-1 are not essential for virus growth in cultured cells, suggesting a role for these genes in viral pathogenesis and host range (7, 59, 75).
BHV-5 UL44 encodes glycoprotein C (gC), which is not essential for neurovirulence in strain TX89; however, it affects neurotropism and is important for high levels of virus replication and full expression of virulence in the rabbit CNS (21). As alphaherpesvirus gC mediates primary attachment of virus to target cells via binding to surface glycans, variability in the heparin binding sites of BHV-5 gC (gC5) and BHV-1 gC (gC1) likely account for differences in their heparin-binding phenotypes (60). Although gC5 and gC1 are 75% identical, the amino-terminal third of the proteins (amino acids 1 to 102 and 1 to 123, respectively) differ significantly. Notably, there is a 35-amino-acid deletion in gC5 which removes two potential N-linked glycosylation sites present in gC1 (amino acids 93 and 111) (36) and a gC1-specific epitope (amino acids 103 to 122) (18). Comparison of gC sequences between strain SV507/99 (this report) and strain TX89 (18) reveals substantial differences at the amino-terminal third. However, a more detailed analysis shows that the proteins are 97% identical and that the discrepancy likely results from a missing base at position 480 in the published TX89 sequence.
The homologue of UL24 in HSV-1 is required for efficient replication in TG of mice (50). The homologue of UL11 in HSV-1 encodes a tegument protein with roles in virion envelopment and egress (7). BHV-5 and BHV-1 UL3.5 and UL0.7 are not found in other herpesviruses, and their functions remain unknown. BHV-5 lacks a homologue of BHV-1 UL0.5, which is predicted to encode an 87-amino-acid protein of unknown function.
The latency-related (LR) gene is postulated to contribute via alternative splicing to the LR protein(s) (27, 44), whereas BICP0 encodes a homologue of BHV-1 BICP0, an immediate-early (IE) and early (E) transactivator (95). Given the potential roles of LR and BICP0 genes in virulence and host range, these genes are treated in more detail below (see "IE genes" and "The LR region," below).
US region. The US region, extending from positions 116733 to 126280, contains eight genes (US1.67, 2 to 4, and 6 to 9), four of which have been previously sequenced (1, 20, 22, 31). BHV-5 US genes exhibit 69 to 98% amino acid identity (average, 79%) and overall less conservation to BHV-1 homologues than those within the UL region (84% average amino acid identity) (58). BHV-5 US2 to US9 are syntenic with homologues in HSV-1, with the exception of the US5 homologue, which is lacking in BHV-5.
The two BHV-5 genes located at the ends of the US are likely significant for virus-host interactions. The gene at the US-IR boundary, US1.67, contains the 75 carboxy-terminal amino acids within the IR and is homologous to ORFs found in other members of the
2 herpesvirus subgroup. Notably, the homologue of US1.67 in EHV-1 is a virulence determinant and is involved in egress of viral nucleocapsids (70, 71, 89). The gene at the US-TR boundary, US9, is essential for neurovirulence in the TX89 strain. Following intranasal inoculation of rabbits, a TX89 strain US9 deletion mutant failed to invade the CNS mainly due to an inability of the virus to spread to the olfactory bulb via anterograde transport (22).
BHV-5 US2, US4, and US8 are the least-conserved US ORFs (
75%) in comparison with BHV-1. The US2 homologue in HSV-1 and HSV-2 seems dispensable in tissue culture and is not involved in HSV-2 neuropathogenesis in mice (52, 61). BHV-5 US4 encodes a glycosaminoglycan-associated protein (31) and is 72% identical to BHV-1 gG, which is involved in cell-cell virus transmission in vitro (69) and in prevention of apoptosis in certain cell lines (68). BHV-5 US8 encodes gE, which is important for neurovirulence in rabbits. Deletion of BHV-5 US8 or its substitution by BHV-1 gE resulted in viruses that replicated and spread much less efficiently in the rabbit brain than revertant or wild-type viruses (20).
Repeats.
The IR and TR regions, located at nucleotide positions 104624 to 116732 and 126281 to 138389, respectively, are 12,109 bp in length. Each repeat contains two genes, BICP4 and BICP22, which are 75 and 68% identical to BHV-1 BICP4 and BICP22, respectively (see "IE genes," below). Among nonbovine herpesviruses, genes of EHV-1 are the most similar to BICP4 (42% amino acid identity) and BICP22 (40% amino acid identity). Interestingly, a homologue of the
1 34.5 gene, a neurovirulence determinant of HSV-1, is not present in BHV-5 (79).
A 4,867-bp noncoding region separates BICP4 from BICP22 start codons. Overall, this region is 56% identical to the homologous BHV-1 intergenic region and contains transcriptional regulatory elements for the flanking IE genes and the origins of DNA replication.
IE genes. Herpesvirus IE genes are critical regulators of viral gene expression. BHV-5 BICP0, BICP4, and BICP22, the homologues of BHV-1 IE genes BICP0, BICP4, and BICP22, are relatively less conserved than other classes of viral genes (Table 1).
In BHV-1, the BICP0 gene is transcribed at IE and E times postinfection from two separate promoters, leading to accumulation of 2.9- and 2.6-kb RNAs, respectively (94, 95). The IE promoter also controls BICP4 expression at IE times postinfection. A BICP0-specific transcript of similar size has been detected in BHV-5-infected cells at IE times (96). Several sequence elements involved in IE transcription and RNA processing of BHV-1 BICP0, including the TATA box (BHV-5 position 110799), major transcription start site (position 110766), donor and acceptor splice sites (positions 110417 and 104413, respectively), and polyadenylation signal (position 102150), are conserved in BHV-5. Further, potential TATA and SP1 sites are located at -39 and -71 relative to the first BICP0 codon, respectively, suggesting a similar promoter may direct E expression of BHV-5 BICP0 as in BHV-1 (95). Taken together, these features suggest that common mechanisms control BICP0 transcription in the two viruses.
BHV-5 BICP0 shares 70% amino acid identity with BHV-1 BICP0. Most amino acid differences are found at the carboxy half of the protein, a region which plays a role in subcellular localization of ICP0-like proteins (33, 47). BHV-5 BICP0 is 44 amino acids longer that its BHV-1 counterpart, including several small amino acid insertions and a 30-amino-acid extension at the carboxy terminus. This difference is likely responsible for the slightly larger size of IE transcripts originated from BHV-5 BICP0 relative to those from BHV-1 (96). Both BHV-5 and BHV-1 BICP0s are most similar in the amino half, which contains a conserved acidic cluster (amino acids 273 to 323 in BHV-5) and a C3HC4 zinc ring finger (amino acids 22 to 59). However, the region between amino acids 119 and 145 (110 to 155 in BHV-1) is poorly conserved and includes a 20-amino-acid deletion in BHV-5. In BHV-1, the BICP0 ring finger has been implicated in promoting transactivation, stimulation of productive infection, and cytotoxicity (47). It is not known if BHV-1 BICP0 contributes to virus pathogenesis; however, a point mutation in HSV-1 ICP0 significantly reduced neuroinvasiveness of an HSV-1 neurovirulent strain after peripheral inoculation of mice (90).
BHV-5 BICP4 encodes the homologue of BHV-1 BICP4, an IE transactivator and transrepressor (84). Based on sequence homologies, alphaherpesvirus ICP4 orthologs are divided into regions I to V from the amino to the carboxy terminus (13, 17, 43, 84). BHV-5 and BHV-1 ICP4 are most similar (90 to 100% amino acid identity) within regions II and IV. When compared with HSV-1, BHV-5 BICP4 regions II and IV are also the most similar (>50% amino acid identity). HSV-1 ICP4 region II is involved in homodimerization and DNA binding, whereas region IV is required for efficient transactivation (see reference 13 and references therein). BHV-5 BICP4 is 63 amino acids longer than BHV-1 BICP4, and this may be responsible for the slightly larger size of IE transcripts originated from BHV-5 BICP4 relative to their BHV-1 counterparts (96). The difference in length is partially due to a 47-amino-acid insertion in BHV-5 BICP4 region III, which includes a unique stretch of the dipeptide DG between two glutamic acid-rich clusters unique to bovine alphaherpesvirus ICP4s. BHV-5 BICP22 encodes the homologue of the BHV-1 IE and late transrepressor BICP22 (85). Most differences are observed in the carboxy halves of the proteins (48% amino acid identity versus 82% in the amino termini). This region contains 30 serines in BHV-5 and 16 in BHV-1 in a tract of 35 residues.
The LR region. The BHV-1 LR region is the only transcriptionally active region in latently infected neurons and is hypothesized to play a role in latent infections (44, 76). In vitro, the BHV-1 LR gene inhibits cell cycle progression (83), BICP0-dependent activation of productive infection (39), and apoptosis (23). Within the LR gene there exist two major ORFs designated ORF1 and ORF2, where ORF1 is completely overlapped by BICP0 (56). Using an antibody against the amino terminus of ORF2, an LR protein was detected in lytically infected cells and in latently infected TG neurons (44, 51). Infection of cattle with mutant BHV-1 containing three stop codons that should prevent LR protein expression leads to a reduced ability of virus to replicate in the eye during acute infection (48), lower levels of viral DNA in latently infected TG, and a severe defect in virus reactivation after dexamethasone treatment of latently infected animals (49).
A complex pattern of alternative splicing in the LR transcript (LRT) has been described (27). LRT splicing varied according to the source of LR RNA (cell lines or TG), the time postinfection, and the type of RNA [poly(A)+ or poly(A)-] used for the experiments. The splicing pattern allows for different combinations between methionine-initiated ORFs and between methionine-initiated and non-methionine-initiated ORFs which may result in different protein isoforms.
Figure 1 summarizes major differences between BHV-5 and BHV-1 in the LR region. Whereas both viruses contain ORF1 in frame 1 (66% amino acid identity), the structure of BHV-5 frame 2 departs from that of BHV-1. In BHV-1, frame 2 contains an ORF (LRORF2) which could encode a 181-amino-acid protein and a downstream reading frame without an initiating methionine. In BHV-5, frame 2 is interrupted four times, resulting in reading frames a to d (Fig. 1), with reading frame a being the only one to be methionine initiated. Reading frame a, which is truncated at 51 amino acids, and reading frame b are 82 and 75% identical to the corresponding regions in BHV-1 ORF2, respectively, while reading frames c and d are frameshifted. Interestingly, all splicing donor sites mapped in the BHV-1 LR transcript fall downstream of the carboxy terminus of frame b, suggesting that if BHV-5 proteins were made from this region they would differ substantially from BHV-1 LR products. A detailed comparison of LRORF2 and LRORF1 amino acid sequences between the two viruses is shown in Fig. 2A and B.
![]() View larger version (12K): [in a new window] |
FIG. 1. Comparison of the LR regions of BHV-5 and BHV-1. The LR nucleotide positions are given above (BHV-5) and below (BHV-1) the central stippled box, and the relative nucleotide positions are in bold (1 to 3,000 bp). Restriction sites are XhoI (X), XbaI (Xb), and SphI (S). The dots above the BHV-1 LRT represent the positions of splice donor signals (27). Methionine- and non-methionine-initiated ORFs are represented by black and open boxes, respectively. Asterisks indicate the positions of in-frame stop codons. BHV-1 neuron-specific transcriptional regulatory regions NSB and NSTA are represented by boxes between XhoI and SphI sites. The hatched box in the BHV-5 regulatory region represents deleted sequences. The structure of BHV-5 frames 1 and 2 is discussed in the text and shown in detail below in Fig. 2A and B.
|
![]() ![]() View larger version (86K): [in a new window] |
FIG. 2. (A) Alignment of BHV-1 LRORF2 with translated BHV-5 genomic sequence using tfastx, indicating the homologous BHV-5 ORF in frame 2 (boxed), stop codons (*), frameshift to BHV-5 frame 3 (/), conserved substitutions (:), identities (|), and remaining open sequence in BHV-5 frame 3 (![]() ![]() ). Total aligned sequence (182 amino acids) is 69% identical with BHV-1 LRORF2. (B) Alignment of BHV-1 LRORF1 with translated BHV-5 sequence using tfastx. Identities and substitutions are labeled as for panel A. Total aligned sequence (367 amino acids) is 66% identical with BHV-1 LRORF1. (C) Alignment of BHV-1 and BHV-5 LR regulatory sequences represented by the hatched box in Fig. 1, using FASTA. In BHV-1, the underlined sequence 1 was specifically protected from exonuclease digestion by ganglionic nuclear factors (25); underlined sequence 2 was protected from exonuclease digestion by neuroblastoma nuclear factors (11). Potential TATA and CAT boxes are shown in bold.
|
In summary, the BHV-5 LR region differs substantially from the BHV-1 LR region in both coding and transcriptionally regulatory regions. Given the potential significance of this region in viral latency and/or host range and the differential pathogenesis of these closely related viruses, it is likely that these differences are of biological significance for aspects of virus-neuron-host interactions.
Conclusions. Although BHV-5 and BHV-1 are closely related viruses, only BHV-5 causes CNS disease. The molecular basis for the differential pathogenesis remains unknown but probably arises from multiple genetic contributions. The complete BHV-5 genome provides a framework from which comparisons between BHV-5 and BHV-1 pathogenesis may be made. Strategies based on the construction of chimeric viruses will certainly contribute to our overall understanding of pathogen-host interactions and the evolution of herpesvirus virulence.
|
|
|---|
1 34.5 gene. J. Neurovirol. 3(Suppl. 1):S1-S2.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»