Complete Sequence and Genomic Analysis of Rhesus Cytomegalovirus

Abstract

The complete DNA sequence of rhesus cytomegalovirus (RhCMV) strain 68-1 was determined with the whole-genome shotgun approach on virion DNA. The RhCMV genome is 221,459 bp in length and possesses a 49% G+C base composition. The genome contains 230 potential open reading frames (ORFs) of 100 or more codons that are arranged colinearly with counterparts of previously sequenced betaherpesviruses such as human cytomegalovirus (HCMV). Of the 230 RhCMV ORFs, 138 (60%) are homologous to known HCMV proteins. The conserved ORFs include the structural, replicative, and transcriptional regulatory proteins, immune evasion elements, G protein-coupled receptors, and immunoglobulin homologues. Interestingly, the RhCMV genome also contains sequences with homology to cyclooxygenase-2, an enzyme associated with inflammatory processes. Closer examination identified a series of candidate exons with the capacity to encode a full-length cyclooxygenase-2 protein. Counterparts of cyclooxygenase-2 have not been found in other sequenced herpesviruses. The availability of the complete RhCMV sequence along with the ability to grow RhCMV in vitro will facilitate the construction of recombinant viral strains for identifying viral determinants of CMV pathogenicity in the experimentally infected rhesus macaque and to the development of CMV as a vaccine vector.

Cytomegaloviruses are the prototypic members of the betaherpesvirus subgroup. Like all herpesviruses, they share several characteristics with other viruses of the family, including virion structure and the ability to establish persistent and latent infections. What makes the betaherpesvirus subgroup unique among the herpesviruses is their tropism for the salivary gland, their species specificity, and their slow growth in culture systems (25). Weller et al. coined the term cytomegalovirus to reflect the cytopathic effect caused by virus infection and the virus's role in genitally acquired cytomegalic inclusion disease (75).

Human cytomegalovirus (HCMV) infection is generally asymptomatic in immunocompetent hosts. Typically, HCMV is acquired subclinically during childhood and persists throughout the individual's life. Persistence is characterized by the presence of latent viral genomes that periodically reactivate to produce infectious virus that can be shed intermittently in saliva, urine, semen, cervical secretions, and breast milk (25). Data suggest that persistence is established not only by the virus's ability to infect various cell types within the host, but also by the virus's ability to differentially regulate gene expression and by virally encoded immunomodulators (23, 30, 32, 43, 44, 57, 60, 64). When primary HCMV infection is acquired during adulthood, the virus is capable of doing considerably more damage. In particular, primary infections during pregnancy can lead to congenital abnormalities in the fetus and morbidity and mortality in immunocompromised individuals, including transplant patients and persons with AIDS.

Thus, with advances in medical procedures such as organ allografting and immunosuppressive posttransplant therapies as well as with increasing numbers of people infected with the human immunodeficiency virus (HIV), HCMV is proving to be a significant pathogen. Currently, the only Food and Drug Administration-approved antiviral therapies include the drugs ganciclovir, foscarnet, and cidofovir, each of which targets the viral DNA polymerase, and Vitravene, a novel antisense compound used to treat CMV retinitis (38). Although helpful, these drugs have numerous drawbacks, which include limited efficacy and toxic side effects and frequent emergence of resistant viral isolates during long-term therapy (17, 19, 72).

Despite the growing information on HCMV gene products and pathogenesis, there is a limitation on the ability to develop effective treatments for primary HCMV infection or reactivation of latent infection. To address these limitations, four animal models are being evaluated for antiviral treatments against HCMV and for overall cytomegalovirus pathogenesis. These models are guinea pig cytomegalovirus, murine cytomegalovirus (MCMV), rat cytomegalovirus (RCMV), and more recently, rhesus cytomegalovirus (RhCMV). The guinea pig cytomegalovirus model is well suited for congenital cytomegalovirus infection because the virus can cross the placenta and produce infection in utero (31). Despite this advantage, the guinea pig cytomegalovirus model is not well suited to the analysis of drug therapy, as the virus is resistant to ganciclovir therapy (50). The genomes of MCMV and RCMV have been completely sequenced, and these models are providing valuable information on both cytomegalovirus pathogenicity and antiviral therapies (59, 73). However, MCMV and RCMV are evolutionarily far removed from HCMV, and the results obtained in these systems may not be applicable to treatment and prevention of HCMV infection (51, 61).

Over the past decade, researchers have been working to develop an RhCMV model to study HCMV pathogenesis and antiviral therapies (2, 3, 7, 8, 48, 49, 70, 74). RhCMV was first recognized 70 years ago as an incidental infection of rhesus macaques (15, 16). Several nonhuman primate cytomegaloviruses have been isolated since that time, each of which is different from HCMV in host range and DNA content (3, 18, 63, 69). RhCMV is ubiquitous in captive rhesus macaques, with infection rates of greater than 90% by the first year of life (39, 41, 74). Cytomegalovirus infections in macaques share several features with HCMV infections. The majority of RhCMV infections are subclinical, and healthy adults tend to shed virus in their urine, saliva, semen, cervical secretions, and breast milk for years. Experimental congenital central nervous system disease has been produced in rhesus macaques by inoculating fetuses in utero intracerebrally or intra-amniotically (49).

Like humans infected by HIV, rhesus macaques are susceptible to infection with simian immunodeficiency virus (SIV). Studies utilizing SIV-infected rhesus macaques have demonstrated that disseminated cytomegalovirus disease is quite similar to that observed in HIV-infected humans (8). These symptoms can include orchitis, encephalitis, and respiratory tract disease. In addition, cytomegalovirus infections have been seen in the bone marrow and lungs of rhesus macaques that have been immunosuppressed with antithymocyte globulin, cyclophosphamide, and cortisone and subsequently inoculated with varicella-zoster virus, which is similar to what has been observed in people who acquire HCMV infection after transplant surgery (53). Lastly, studies of the host immune response to RhCMV appear to parallel HCMV infection, as cytomegalovirus-specific cytotoxic T lymphocytes are essential in controlling infection (40).

Studies of a nonhuman primate cytomegalovirus such as RhCMV are relevant to the study of HCMV, as this virus is evolutionarily closer to humans than the rodent cytomegalovirus isolates (51). Until now, only a limited number of reports have demonstrated the similarity of HCMV and RhCMV at the genomic level. In this study. we completely sequenced the RhCMV strain 68-1 genome and confirmed that the virus is significantly homologous to HCMV. Data generated for the sequence analysis revealed that the genome is over 200 kb in size and is structurally similar to previously reported cytomegaloviruses. Like HCMV, RhCMV encodes a complement of immunomodulatory genes as well as structural proteins and enzymes required for replication. Unique to RhCMV is a cyclooxygenase-2 (prostaglandin E₂) homologue. Available sequence information on RhCMV will further the development of this virus as a model for HCMV pathogenesis.

MATERIALS AND METHODS

Preparation of viral DNA.

Primary rhesus fibroblasts were grown in two 850-cm² roller bottles and infected with the 68-1 strain of RhCMV (ATCC VR-677) at a multiplicity of infection of 0.1, and the virus was harvested from the culture supernatant at 12 to 14 days postinfection as previously described (70). Briefly, infected-cell supernatants were clarified to remove cellular debris by centrifugation at 1,000 × g for 10 min. The clarified supernatants were collected and stored, and 10 ml of culture medium was added to the pelleted debris. To remove intracellular virus particles, the cell debris pellet was sonicated until resuspended. Sonication was followed by centrifugation at 1,000 × g for 10 min, after which clarified supernatants were combined.

RhCMV was pelleted from supernatants by centrifugation at 12,500 × g for 1 h at 4°C. The virus pellet was resuspended in 1 ml of 1 mM Tris-HCl (pH 8.0)-1 mM EDTA (TE) and added to the top of a six-step sorbitol gradient ranging from 20 to 70%. The gradients were spun in a Beckman SW41 rotor for 2 h at 18,000 rpm at 4°C. The virus-containing band at the 50 to 60% interface was collected and diluted with 15 ml of cold 1 mM Tris-HCl and then pelleted by centrifugation in the SW41 rotor for 50 min at 18,000 rpm at 4°C. The washed virus pellet was resuspended in 9.2 ml of TE (pH 8.0). Virus particles were digested at 37°C overnight with 0.6 ml of 10% sodium dodecyl sulfate and 0.2 ml of proteinase K (10 mg/ml) to release viral DNA. Viral DNA was isolated by CsCl₂ gradient centrifugation in a Beckman Ti75 rotor at 38,400 rpm for 72 h. The viral DNA fraction, which routinely corresponds to refractive indices of 1.400 to 1.403, was collected and dialyzed in TE (pH 8.0) to remove CsCl₂.

The isolated viral DNA was tested for infectivity by transfection into primary rhesus fibroblasts with the calcium phosphate method without dimethyl sulfoxide shock, and the cells were observed for cytopathic effects. As a negative control, primary rhesus fibroblasts were exposed to transfection reagents and conditions in the absence of viral DNA.

Cloning and DNA sequencing.

To facilitate DNA sequencing of the viral genome, a shotgun subclone library was generated. Briefly, purified genomic DNA was hydrodynamically sheared in a high-pressure liquid chromatograph and then separated on a standard 1% agarose gel. A fraction corresponding to fragments 3,000 to 3,500 bp in length was excised from the gel and purified by the GeneClean procedure (Bio101, Inc., Carlsbad, Calif.). The purified DNA fragments were blunt-ended with T4 DNA polymerase. The treated DNA was then ligated to unique BstXI linker adapters. These linkers are complementary to the pGTC vector (Genome Therapeutics Corporation, Waltham, Mass.), while the overhang is not self-complementary. Therefore, the linkers will not concatamerize, nor will the cut vector religate easily. The linker-adapted inserts were separated from the unincorporated linkers on a 1% agarose gel and again purified with GeneClean. The linker-adapted inserts were ligated to BstXI-cut vector to construct a shotgun subclone library.

The library was transformed into Escherichia coli DH10B competent cells (Gibco/Invitrogen, San Diego, Calif.) and assessed on plates containing ampicillin. Bacterial transformants were used for plating of clones and picked for sequencing. The cultures were grown overnight at 37°C, and the DNA was purified by standard methods.

The purified DNA samples were sequenced with ABI (Applied Biosystems Inc., Foster City, Calif.) dye terminator chemistry. All subsequent steps were based on sequencing by automated DNA sequencing methods. The ABI dye terminator sequence reads were run on MegaBace 1000 (Amersham Biosciences, Piscataway, N.J.) machines, and the data were transferred to Unix machines. Base calls and quality scores were determined with the program Phred (21, 22). Reads were assembled with Phrap (29) with default program parameters and quality scores.

Computer-assisted analysis of the RhCMV ORFS and protein coding content.

Open reading frames (ORFs) were identified with MacVector version 6.5 (Accelrys, San Diego, Calif.). The target search criterion for an ORF was 100 codons on either the positive or negative strand. The ORFs that were identified in this manner were translated and saved as individual files for further analysis. Putative RhCMV proteins were then compared to GenBank files with the BlastP search engine to look for homology with other known proteins. Significant hits were compiled and are presented in Table 1. Homology with HCMV ORFs was analyzed further with the global alignment algorithm FastA (Table 2). Multiple sequence alignments were done with MacVector's ClustalW feature, with gap penalties of 10 (fixed) and 10 (floating).

TABLE 1.

Summary of the location and features of the 230 ORFs of the RhCMV 68-1 genome^a

ORF	Strand	Position (nt)		Length (aa)	Size (kDa)	HCMV name	Comments
		From	To
Rh01	W	1040	2566	508	56.4	TRL1
rh02	W	1486	2037	184	21.1
rh03	W	2618	3082	155	17.1
rh04	C	2703	3401	233	26.9
Rh05	W	3528	4349	274	30.2	UL153	Membrane protein; RL11 family
rh06	W	4710	5198	163	18.6
rh07	W	5325	5906	194	22.2
rh08	W	5881	6396	172	19.6
rh09	C	6724	7029	102	11.4
rh10 Ex6	C	8521	9027	169	17.9		Prostaglandin-endoperoxide synthase; cyclooxygenase-2
rh10 Ex5	C	9270	9941	224	26.3		Prostaglandin-endoperoxide synthase; cyclooxygenase-2
rh11	W	9462	9902	154	17.7
rh10 Ex4	C	10031	10216	62	6.9		Prostaglandin-endoperoxide synthase; cyclooxygenase-2
rh10 Ex3	C	10289	10432	48	5.4		Prostaglandin-endoperoxide synthase; cyclooxygenase-2
rh10 Ex2	C	10527	10661	45	5.5		Prostaglandin-endoperoxide synthase; cyclooxygenase-2
rh10 Ex1	C	10762	10932	57	6.4		Prostaglandin-endoperoxide synthase; cyclooxygenase-2
rh12	W	11225	11920	232	26.0
rh13	C	11233	11547	105	11.9
rh14	W	12506	13087	194	21.5		Membrane protein; RL11 family
rh15	C	12556	12879	108	11.4
rh16	C	12937	13458	174	20.1
Rh17	W	13525	14625	367	41.4	UL11	RL11 family (early glycoprotein)
rh18	C	13557	13922	122	13.6
Rh19	W	14707	15642	312	35.3	UL07	Membrane protein; RL11 family
Rh20	W	15700	16293	198	21.9	UL06	Membrane protein; RL11 family
Rh21	W	16327	17004	226	26.2	UL11	Membrane protein; RL11 family
Rh22	W	17111	17818	236	27.3	UL11	Membrane protein; RL11 family
Rh23	W	17716	18411	232	25.9	UL11	Membrane protein; RL11 family
rh24	W	18428	18817	130	14.9		Membrane protein; RL11 family
Rh25	W	18896	19570	225	24.8	UL09	Membrane protein; RL11 family
Rh26	W	19554	20402	283	32.1	UL11	Membrane protein; RL11 family
rh27	W	20547	21152	202	22.9		Membrane protein; RL11 family
rh28	W	21154	21777	849	96.2
Rh29	W	21854	23239	462	50.4	UL11	Membrane protein; RL11 family
rh30	C	23315	23638	108	11.7
Rh31	W	23355	24662	436	50.3	UL13
rh32	C	23577	24029	151	16.9
Rh33	W	24942	25838	299	34.7	UL14
rh34	C	26718	27041	108	12.4
Rh35	W	26778	27101	108	12.6	UL17
Rh36	W	27826	29172	449	50.6	UL20
rh37	C	29275	29640	122	13.9
rh38	W	29898	30251	118	12.8
rh39	C	29916	30380	155	18.0
Rh40	C	30731	31669	313	35.9	UL23	US22 family
rh41	W	31416	31796	127	14.6
Rh42	C	31726	32652	309	35.2	UL24	US22 family
Rh43	W	32719	34476	586	66.7	UL25	UL25 family
Rh44	C	34537	35283	249	27.8	UL26
rh45	W	34897	35328	144	15.2
Rh46	C	35237	36976	580	65.9	UL27
Rh47	C	37064	38077	338	38.6	UL28	US22 family
rh48	W	37289	37786	166	18.3
Rh49	W	38190	38588	133	15.4
Rh50	C	38208	39218	337	38.8	UL29	US22 family; immediate-early protein
rh51	C	38351	38755	135	15.3
rh52	W	38925	39395	157	17.8
rh53	C	39548	39979	144	17.1
Rh54	W	39865	41487	541	60.9	UL31
Rh55	C	41498	43615	706	78.8	UL32	pp150, ppUL32
Rh56	W	43983	44972	330	37.4	UL33	GpUL33; G protein-coupled receptor; 7TM family
Rh57	W	45177	45977	267	30.8	UL34
rh58	W	45491	45877	129	13.9		US22 family; IRS1
Rh59	W	46098	47870	591	66.9	UL35	UL25 family
Rh60	C	47988	49151	388	44.7	UL36	US22 family;
Rh61	C	49115	49480	122	13.8	UL36	US22 family/PICK>
Rh62	C	49578	50396	273	31.2	UL37	Immediate-early glycoprotein
rh63	W	50598	50981	128	13.6
Rh64	C	50702	51583	294	33.4	UL38
rh65	W	50768	51133	122	13.7
Rh66	C	51625	51930	102	11.6	UL37	Immediate-early protein; IE glycoprotein
rh67	C	52255	52785	177	18.9
Rh68	C	53229	53621	131	14.4	UL42	Glycoprotein (gpUL42)
Rh69	C	53605	54606	334	38.5	UL43	US22 family
Rh70	C	54730	55902	391	43.9	UL44	ppUL44; DNA polymerase processivity factor; early nuclear antigen p41
rh71		55132	55608	159	18.1
Rh72	C	56143	58692	850	96.7	UL45	Ribonucleotide reductase-1
rh73	W	56404	56829	142	16.1
rh74	W	58181	58483	101	11.3
Rh75	C	58711	59583	291	33.1	UL46	Minor capsid binding protein (mCBP)
Rh76	W	59582	62458	959	110.6	UL47	Capsid assembly protein;; HMW-BP; ppUL47
rh77	C	62052	62357	102	11.5
Rh78	W	62479	69012	2178	246.9	UL48	Very large tegument protein; large tegument protein (U31)
rh79	W	69181	69846	222	23.8		F fragment DNA
Rh80	W	69295	70764	490	56.2	UL49	Virion protein (LF3); capsid protein; ORF 66, Epstein-Barr virus BFRF2 homologue
Rh81	C	70703	71629	309	33.8	UL50	Virion protein
Rh82	C	71655	71990	112	12.3	UL51
Rh83	W	72069	73724	552	62.5	UL52	Virion protein
rh84	C	72179	72484	102	10.8
Rh85	W	73717	74583	289	33.1	UL53	Virion protein
rh86	C	74128	74559	144	16.0		F fragment DNA
Rh87	C	74561	77668	1036	116.6	UL54	DNA polymerase
Rh88	W	77489	77866	126	14.1	UL53
Rh89	C	77687	80071	795	91.5	UL55	Glycoprotein B; envelope glycoprotein
rh90	W	79376	79822	149	16.9
Rh91	C	80217	82523	769	88.2	UL56	Transport protein; transport/capsid assembly protein; pUL56
Rh92	C	82670	86152	1164	128.9	UL57	ppUL57; ssDNA-binding protein, major DNA-binding protein
rh93	C	87380	87694	105	11.5
rh94	W	87758	88315	186	19.5
rh95	C	88049	88750	234	24.1
rh96	W	88116	89006	297	30.3		Ig heavy chain
Rh97	C	90525	92858	778	85.9	UL69	ppUL69
rh98	W	91522	91830	103	11.7		Ig heavy chain
rh99	W	92073	92855	261	28.5
Rh100	C	92792	95644	951	109.8	UL70	DNA helicase-primase component
Rh101	C	96316	97347	344	39.3	UL72	dUTPase
Rh102	W	97342	97656	105	11.9	UL73	Glycoprotein N
Rh103	C	97637	98806	390	45.6	UL74	Glycoprotein O
Rh104	C	99023	101185	721	81.7	UL75	Glycoprotein H
Rh105	W	101318	102202	295	32.8	UL76	Virion protein
Rh106	W	101871	103658	596	67.3	UL77	Virion protein
Rh107	W	103785	104924	380	44.2	UL78	GCR homologue; 7TM family
Rh108	C	105020	105820	267	30.5	UL79
Rh109	W	105819	107663	615	66.6	UL80	Capsid assembly protein
Rh110	C	107776	109422	549	61.6	UL82	Major late antigen; pp71; UM; ppUL82; UL82 family
Rh111	C	109552	111171	540	62.1	UL83	Phosphorylated matrix protein (pp65); LM; ppUL83; UL82 family
Rh112	C	111240	112868	543	61.7	UL83	Phosphorylated matrix protein (pp65); LM; ppUL83; UL82 family
rh113	W	112041	112487	149	16.3
Rh114	C	112990	114528	513	57.3	UL84	Early nonstructural protein
rh115	W	113051	113419	123	13.3
rh116	W	114270	114728	153	16.1
Rh117	C	114443	115369	309	34.7	UL85	Minor capsid protein; mCP
Rh118	C	115430	119461	1344	151.3	UL86	Major capsid protein; MCP
rh119	W	115995	116309	105	11.8
rh120	C	116922	117656	245	28.8
rh121	W	117836	118309	158	17.5
Rh122	W	119476	122022	849	96.2	UL87	Virion protein
Rh123	W	122035	123237	401	45.5	UL88	Virion protein
Rh124 Ex2	C	123234	124181	316	35.8	UL89	DNA packaging protein (conserved spliced gene)
rh125	W	123897	124469	191	21.0
Rh126	W	124501	124803	101	10.7	UL91
Rh127	W	124697	125413	239	26.6	UL92	Virion protein
Rh128	W	125367	126932	522	59.9	UL93	Virion protein
Rh129	W	127435	127848	138	14.8	UL94	Virion protein
Rh124 Ex1	C	127845	128768	308	35.8	UL89	Virion protein
Rh130	W	128767	130047	427	46.6	UL95	Virion protein
Rh131	W	130044	130433	130	14.8	UL96	Virion protein
Rh132	W	130491	132323	611	67.9	UL97	Phosphotransferase; HCMV UL97; phosphorylates ganciclovir; ppUL97
rh133	C	131142	131456	105	11.8
Rh134	W	132374	134044	557	63.4	UL98	Dnase; exonuclease
rh135	C	132886	133287	134	14.6
Rh136	C	133325	133918	198	22.3	UL99	Phosphoprotein; pp28
rh137	W	133981	134436	152	16.6
Rh138	C	134603	135673	357	41.1	UL100	Glycoprotein M
Rh139	W	135862	138036	725	80.4	UL102	Helicase-primase component
Rh140	C	138058	138813	252	28.9	UL103
Rh141	C	138740	140707	656	75.3	UL104	Structural protein; virion protein
Rh142	W	140544	143123	860	97.5	UL105	DNA helicase
Rh143	W	146491	146958	156	17.9	UL111	IL-10-like protein; cmvIL-10
Rh144 Ex1	W	147820	148620	267	28.4	UL112	Early phosphoprotein; pp34
Rh145	W	148719	149600	294	30.5	UL113
Rh144 Ex2	W	148719	149600	294	30.5	UL112	Early phosphoprotein, pp34
Rh146	C	149714	150457	248	26.3	UL114	Uracil N-glycosylase
Rh147	C	150420	151196	259	29.3	UL115	Glycoprotein L
Rh148	C	151207	152277	357	37.9	UL116	Serine-alanine-rich glycoprotein
rh149	C	151952	152416	155	19.2
Rh150	C	152259	153407	383	42.5	UL117
Rh151	C	153432	154031	200	23.5	UL118	Large splice transcript
Rh152	C	154111	154782	224	22.2	UL119
rh153	C	154260	154577	106	11.6		Major immediate-early protein
Rh154	C	154831	155427	199	22.7	UL120
Rh155	C	155429	155977	183	21.1	UL121	Serine-alanine-rich glycoprotein
Rh156 Ex5	C	156229	157689	487	52.7	UL122	Immediate-early protein 2; pp80
Rh156 Ex4	C	158214	159383	390	44.0	UL123	Immediate-early protein 1; pp72
Rh156 Ex3	C	161559	162053	165	18.6	UL122/3	Immediate-early exon
Rh156 Ex2	C	161621	161708	29	3.5	UL122/3	Immediate-early exon
Rh156 Ex1	C	161778	161902	41	4.4	UL122/3	Immediate-early exon
rh157	W	161947	162489	181	21.6
Rh158	W	163705	164166	154	17.7	UL147	VCXC-2
Rh159	W	164371	165351	327	36.8	UL148
Rh160	W	165417	166082	222	24.4	UL132
rh161	C	167130	167570	147	16.0
Rh162	C	167655	167960	102	11.2	UL145
Rh163	C	168378	168893	172	18.7	UL144	TNF receptor homologue; immune evasion viroceptor; ppUL144
Rh164	C	169096	170388	431	48.9	UL141
rh165	C	170857	171288	144	16.5
Rh166	C	171333	171857	175	19.3	UL133
rh167	C	171988	172491	168	18.2
rh168	C	172745	173401	219	24.5
rh169	C	173501	174064	188	20.9
rh170	C	174198	174764	189	21.1
Rh171	C	174768	175607	280	30.6	UL133	Glycoprotein cercopithecine herpesvirus 5
rh172	C	175806	176336	177	19.8		Glycoprotein cercopithecine herpesvirus 5
rh173	C	176388	177515	376	40.9		Glycoprotein cercopithecine herpesvirus 5
rh174	C	178695	179774	360	39.7		Glycoprotein cercopithecine herpesvirus 5
rh175	W	180470	180922	151	16.4
rh176	C	180619	181260	214	23.3
rh177	C	181226	181669	148	16.5
rh178	C	181320	182060	247	27.3
rh179	W	183231	183746	172	18.5
rh180	C	183347	183670	108	10.6
Rh181	C	183766	184272	169	19.4	US1
Rh182	C	184502	185092	197	23.1	US2	gpUS2; US6 family
Rh183	W	185590	185952	121	13.6	US5	gpUS5
Rh184	C	185617	186159	181	20.6	US3	gpUS3;immediate-early; US6 family
Rh185	C	187133	187645	171	18.9	US6	US6 family
rh186	C	187934	188638	235	28.2		US6 family
Rh187	C	188879	189559	227	25.5	US8	US6 family
rh188	C	189657	190031	125	14.6		GTP binding protein
Rh189	C	190318	191163	282	32.9	US11	US6 family
Rh190	C	191367	192149	261	30.0	US12	US12 family
rh191	C	191524	191856	111	11.8
Rh192	C	192207	192971	255	29.7	US13	US12 family
rh193	C	192977	193462	162	18.5
Rh194	C	193084	193917	278	31.5	US14	US12 family
Rh195	C	194048	194791	248	27.9	US14	US12 family
Rh196	C	194864	195622	253	29.4	US14	US12 family
Rh197	C	195728	196453	242	27.9	US14	US12 family
Rh198	C	196431	197255	275	30.3	US17	US12 family
Rh199	C	197361	198216	202	22.8	US18	US12 family; pUS18
Rh200	C	198336	199121	262	21.1	US19	US12 family; pUS19
Rh201	C	199182	199943	253	67.8	US20	US12 family; pUS20
Rh202	C	199991	200677	228	11.5	US21	US12 family
Rh203	C	200797	202521	575	65.9	US22	US22 family; ICP22; pUS22
Rh204	C	202680	204548	622	37.5	US23	US22 family
rh205	W	202743	203504	254	29.9
rh206	C	202979	203326	116	13.9
rh207	W	203175	203504	110	13.1
rh208	W	204333	204650	106	11.5
rh209	C	204572	206002	477	56.5		US22 family
rh210	W	205019	205594	192	21.1
rh211	C	206363	208156	598	67.8		US22 family
rh212	C	206818	207132	105	11.5
rh213	W	206909	207424	172	19.1
Rh214	W	208328	209314	328	37.5	US28	Chemokine receptor homologue; 7TM family
Rh215	W	209660	210673	338	38.8	US28	Chemokine receptor homologue; 7TM family
Rh216	W	210809	211809	333	35.8	US28	Chemokine receptor homologue; 7TM family
Rh217	C	211704	212006	101	11.4
Rh218	W	211882	212901	340	39.0	US28	Chemokine receptor homologue; 7TM family
rh219	C	212671	212976	102	11.5
Rh220	W	213046	214659	393	43.2	US28	Chemokine receptor homologue; 7TM family
Rh221	W	214653	215978	442	49.4	US29
rh222	W	215128	215451	108	12.6
Rh223	W	215896	216717	274	30.7	US30
rh224	C	216579	217196	206	22.5
Rh225	W	216793	217278	162	18.5	US31	US1 family
Rh226	W	217405	217965	187	22.2	US32	US1 family
rh227	C	217463	217879	139	14.7
rh228	W	218098	218403	102	10.9
rh229	C	219051	219506	152	16.5
Rh230	C	219127	221214	696	77.6	TRS1	TRS1; US22 family; pTRS1

^aThe ORF finder of MacVector was used to identify all 230 putative ORFs. Putative ORFs are numbered by the order in which they appear in the genome. ORFs with homology to HCMV were given an Rh designation, and ORFs unique to RhCMV were given an rh designation. ORFs that read left to right are designated W, whereas ORFs that read right to left are designated C. aa, amino acids; 7TM, seven-transmembrane; ssDNA, single-stranded DNA; Ig, immunoglobulin; GCR, G protein-coupled receptor.

TABLE 2.

Similarity between RhCMV genes with amino acid sequence homology to HCMVa

RhCMV gene	Length (aa)	HCMV homologue	Homologue length (aa)	Identity (%)	Gaps (aa)	Score (except value)		RhCMV gene	Length (aa)	HCMV homologue	Homologue length (aa)	Identity (%)	Gaps (aa)	Score (except value)
Rh01	508	TRL1	311	37	35	304
Rh19	311	UL07	221	34	3	124
Rh20	197	UL06	284	26	31	228
Rh21	225	UL11	275	23	54	93
Rh22	235	UL11	275	25	10	110
Rh23	231	UL11	275	36	7	152
Rh25	224	UL09	228	23	14	125
Rh31	435	UL13	473	30	31	96
Rh33	298	UL14	343	31	17	343
Rh35	107	UL17	104	78	0	64
Rh36	448	UL20	340	30	17	291
Rh40	312	UL23	342	41	1	520
Rh42	308	UL24	358	53	1	869
Rh43	585	UL25	656	35	58	974
Rh44	248	UL26	188	45	5	414
Rh46	579	UL27	608	54	22	1559
Rh47	337	UL28	379	64	4	1207
Rh50	336	UL29	360	62	0	1107
Rh54	540	UL31	694	58	16	1515
Rh55	705	UL32	1,048	54	8	821
Rh56	329	UL33	390	60	7	953
Rh57	266	UL34	504	65	1	945
Rh59	590	UL35	640	39	60	1135
Rh61	121	UL36	476	63	52	233
Rh62	272	UL37	487	30	21	265
Rh64	293	UL38	331	51	3	754
Rh66	101	UL37	487	36	11	111
Rh68	130	UL42	158	44	39	148
Rh69	333	UL43	187	56	1	199
Rh70	390	UL44	433	67	49	1437
Rh72	849	UL45	906	61	16	2620
Rh75	290	UL46	290	72	0	11116
Rh76	958	UL47	982	41	32	1984
Rh78	2,177	UL48	2,241	37	181	3854
Rh80	489	UL49	570	71	46	2026
Rh81	308	UL50	397	76	3	788
Rh82	111	UL51	157	81	0	351
Rh83	551	UL52	668	54	77	1667
Rh85	288	UL53	376	80	0	1083
Rh87	1,035	UL54	1,242	59	219	3608
Rh89	794	UL55	906	57	28	2444
Rh91	768	UL56	850	70	80	2984
Rh92	1,163	UL57	1,235	68	41	4248
Rh97	777	UL69	744	57	14	1110
Rh100	950	UL70	1,062	46	75	3273
Rh101	343	UL72	388	57	24	1050
Rh102	104	UL73	138	56	41	203
Rh103	389	UL74	466	41	40	726
Rh104	720	UL75	743	48	11	1757
Rh105	294	UL76	325	52	9	663
Rh106	595	UL77	642	64	47	2117
Rh107	380	UL78	431	58	34	500
Rh108	266	UL79	295	38	25	797
Rh109	614	UL80	708	23	163	1006
Rh110	548	UL82	559	39	30	926
Rh111	539	UL83	561	32	30	781
Rh112	542	UL83	561	35	23	955
Rh114	512	UL84	586	47	76	1108
Rh117	308	UL85	306	67	5	1056
Rh118	1,343	UL86	1,370	75	27	5631
Rh122	848	UL87	941	65	60	2865
Rh123	400	UL88	429	47	30	972
Rh124	315	UL89	674	84	1	1425
Rh126	100	UL91	111	72	3	295
Rh127	234	UL92	234	90	188	946
Rh128	521	UL93	594	45	82	1166
Rh129	137	UL94	345	64	101	507
Rh124	307	UL89	674	86	2	1352
Rh130	426	UL95	531	65	72	1436
Rh131	129	UL96	115	57	1	319
Rh132	610	UL97	707	67	6	1649
Rh134	556	UL98	584	65	39	1890
Rh138	355	UL100	372	55	12	1112
Rh139	724	UL102	798	54	144	2059
Rh140	251	UL103	249	50	169	656
Rh141	655	UL104	697	64	39	2262
Rh142	859	UL105	956	69	97	3293
Rh144	266	UL112	864	48	2	333
Rh145	293	UL113	684	30	77	176
Rh146	247	UL114	250	69	1	933
Rh147	258	UL115	306	53	2	620
Rh148	356	UL116	344	27	27	144
Rh150	382	UL117	424	46	46	890
Rh151	199	UL118	209	32	18	266
Rh152	223	UL119	142	30	2	97
Rh154	198	UL120	201	43	12	322
Rh155	182	UL121	180	28	8	124
Rh156	486	UL122	411	46	30	850
Rh156	389	UL123	491	21	16	174
Rh160	221	UL132	270	52	5	221
Rh181	168	US01	212	43	16	323
Rh182	196	US02	199	23	4	148
Rh183	180	US05	186	23	7	86
Rh187	226	US08	227	22	11	76
Rh189	281	US11	215	21	6	242
Rh190	260	US12	281	33	6	353
Rh192	254	US13	261	24	10	268
Rh194	277	US14	310	23	11	242
Rh195	247	US14	310	23	17	132
Rh196	252	US14	310	29	9	238
Rh197	241	US14	310	23	22	93
Rh198	274	US17	293	36	6	383
Rh199	201	US18	274	32	2	320
Rh200	261	US19	240	29	19	178
Rh201	252	US20	357	40	149	540
Rh202	227	US21	239	58	1	646
Rh203	574	US22	593	51	6	1242
Rh204	621	US23	592	48	52	1615
Rh214	327	US28	323	27	7	255
Rh215	337	US28	323	24	4	261
Rh216	332	US28	323	34	0	109
Rh218	339	US28	323	25	7	325
Rh220	392	US28	323	37	7	463
Rh221	441	US29	462	22	14	565
Rh223	273	US30	349	24	41	144
Rh226	161	US31	197	44	1	226
Rh227	186	US32	183	38	3	274
Rh230	695	TRS1	788	34	96	819

Nomenclature and phylogenetic analysis.

The nomenclature used for the RhCMV genes is based on the same strategy used in designating both the MCMV and RCMV genomes. This system numbers the ORFs from left to right starting at the first open reading frame on the coding or complementary strand. RhCMV genes homologous to HCMV genes are indicated by uppercase prefixes (e.g., Rh88), whereas ORFs without sequence similarity to HCMV genes are indicated by lowercase prefixes (e.g., rh06).

Sequence data from ClustalW alignments were analyzed with ClustalX version 1.8 and the draw tree feature with bootstrapping. Tree data generated with this method were viewed with the phylogenetic tree-generating software TreeView version 1.6.1 (Logiciels & Duhem, Paris, France). The measure of divergence is presented as a scale at the bottom of each tree. Viral sequences for phylogenetic analysis were acquired from GenBank. The viruses and their accession numbers are as follows: herpes simplex virus type 1 (HSV-1), NC001806; Kaposi's sarcoma-associated herpesvirus (KSHV), NC003409; MCMV, U68299; RCMV, AF232689; chimpanzee cytomegalovirus (ChCMV), NC003521; and HCMV, NC001347. The proteins used for phylogenetic analysis were DNA polymerase, glycoprotein B (gB), helicase, major capsid protein, single-stranded-DNA-binding protein, and uracil N-glycosylase.

Nucleotide sequence accession number.

RhCMV 68-1 nucleotide sequence data have been deposited in the GenBank database and assigned accession number AY186194.

RESULTS

Features of nucleotide sequence of RhCMV.

The genome of RhCMV 68-1 is 221,459 bp in length, which is less than HCMV at 229,354 bp (52). The overall G+C content is 49% and is evenly distributed throughout the length of the genome. The only exception is a short stretch of sequence from nucleotides 15000 to 20000 (Fig. 1). . This region has a lower G+C content (20 to 40%) and represents ORFs Rh20 through Rh27 inclusive. These proteins have significant homology to the HCMV RL11 family members.

FIG. 1.

G+C content of RhCMV 68-1 genome. The distribution of G+C versus A+T content was analyzed over the entire genome. Results are presented as histograms.

Like all herpesviruses, the RhCMV genome is a long, double-stranded unique sequence. There are two PacI homologies (virion packaging sites) as defined by HCMV, one found at the start of the genome and the other between what would be the U_L and U_S regions of HCMV. Although our sequence data did not reveal large internal or terminal repeats like those present in the HCMV genome, Southern analysis with probes specific for the termini detected variable copies of a roughly 500-bp terminal direct repeat sequence that may be analogous to the a sequence of HCMV and other herpesviruses (13).

Restriction analysis of RhCMV genomic DNA.

To verify the genomic structure of RhCMV DNA, purified viral DNA was digested with the restriction enzymes BamHI and HindIII (Fig. 2). . Both digests produced a large number of distinguishable fragments. The resultant bands correspond to predicted restriction digest patterns (Fig. 2A and B and Table 3). The combination of sequence and restriction digest data, along with data from Chang and Barry (13), supports the hypothesis that the organization presented in this paper is the correct structure for RhCMV and that the virus does not isomerize, unlike HCMV.

FIG. 2.

Restriction analysis of RhCMV genome. (A) Restriction digests of the RhCMV 68-1 genome. Purified genomic DNA (10 μg) was digested overnight with 25 U of either BamHI (lane B) or HindIII (lane H) at 37°C. Digests were supplemented with an additional 25 U of enzyme the following day for 4 h prior to separation on a 0.7% agarose gel. Bands were resolved at 20 V overnight. Lanes MW, size markers. (B) Deduced BamHI and HindIII restriction maps. Digested products were numbered based on the sizes shown in Table 3.

TABLE 3.

Predicted fragment sizes

Endonuclease	Fragment no.	Fragment size (bp)
BamHI	1	25,743
	2	19,539
	3	11,138
	4	10,445
	5	9,746
	6	9,496
	7	8,965
	8	8,468
	9	8,355
	10	7,738
	11	7,180
	12	6,665
	13	6,304
	14	6,243
	15	6,240
	16	6,181
	17	5,092
	18	4,994
	19	4,136
	20	3,885
	21	3,484
	22	3,457
	23	3,167
	24	3,114
	25	2,819
	26	2,661
	27	2,415
	28	2,415
	29	2,305
	30	2,161
	31	1,955
	32	1,856
	33	1,616
	34	1,551
	35	1,352
	36	1,226
	37	1,184
	38	1,004
	39	919
	40	841
	41	801
	42	686
	43	600
	44	528
	45	293
	46	165
	47	132
	48	78
	49	66
	50	30
	51	25
HindIII	1	18,969
	2	17,516
	3	17,139
	4	15,818
	5	15,135
	6	13,257
	7	13,144
	8	11,186
	9	9,370
	10	9,169
	11	7,978
	12	7,469
	13	6,589
	14	5,772
	15	5,281
	16	4,677
	17	4,567
	18	4,368
	19	3,512
	20	3,413
	21	3,040
	22	2,762
	23	2,509
	24	2,498
	25	2,218
	26	2,217
	27	2,204
	28	2,105
	29	2,004
	30	1,916
	31	1,401
	32	1,347
	33	487
	34	264
	35	135
	36	32

Protein coding content.

The arrangement of the RhCMV genes is shown in Fig. 3. The genome contains 230 ORFs that are arranged colinearly with those of previously sequenced betaherpesviruses, including HCMV, MCMV, and RCMV. All of the conserved herpesvirus gene blocks are retained in RhCMV in both position and orientation. Of the 230 ORFs, 138 (60%) encode products that are homologous to HCMV proteins (Table 2). These homologues are dispersed throughout the genome. Less than half of the ORFs are predicted to encode genes unique to RhCMV; these ORFs are possibly host range specific. Unlike MCMV, RhCMV encodes almost all classified HCMV gene families, including RL11, UL25, UL82, seven transmembrane, US1, US2/6, US12, and US22 families.

FIG. 3.

Map of RhCMV genome. Map of the RhCMV 68-1 genome showing the ORFs on the coding strand (dark arrows) and the noncoding strand (light arrows). Putative RhCMV genes are numbered as shown in Table 1 from Rh01 to Rh236. A selected number of HCMV homologues are listed: pp150, large structural phosphoprotein (UL32); DNA pol, DNA-dependent DNA polymerase (UL54); gB, glycoprotein B (UL55); MDBP, major DNA-binding protein (UL57); pp71, phosphoprotein 71 (UL82); pp65, phosphoprotein 65 (UL83); MCP, major capsid protein (UL86).

Genome and individual ORFs. (i) Proteins essential for origin-dependent replication.

RhCMV encodes six of the seven proteins necessary for HSV-1 replication. These proteins are the DNA polymerase (Rh87), the polymerase accessory protein (Rh70), the single-stranded-DNA-binding protein (Rh92), and the helicase-primase complex (Rh100, Rh139, and Rh142). The DNA helicase (Rh142) contains the six motifs required for function in congruent positions described for HSV and HCMV (77). RhCMV lacks the origin-binding protein (UL9) of HSV-1 that is required for HSV replication. The only betaherpesvirus known to encode an HSV-1 UL9-like protein is human herpesvirus 6 (28). RhCMV does encode all of the remaining trans-activating factors required for transient complementation of HCMV origin of lytic replication-dependent DNA replication (54). These include UL36 to 38 (Rh61, Rh62, Rh64, and Rh66), TRS1 (Rh230), and IE1/IE2 (Rh156 exons 4 and 5, respectively), which are known regulatory proteins, and ORFs UL84 (Rh114) and UL112/UL113 (Rh144/Rh145), which encode early temporal class nucleus-associated proteins. HCMV UL84 and IE2 have been shown to interact and are thought to be responsible for transcriptional control (65). The DNase gene (Rh134) shown to be required for HSV-1 replication in vitro is 65% conserved with HCMV UL98 (Table 2) (51).

An RhCMV origin of replication (ori) is situated between homologues of HCMV UL57 and UL69 (between nucleotides 88161 and 90525). The ori shares several structural features with HCMV oriLyt, most notably the Y-R element (78), and mediates RhCMV-dependent replication in transient assays (David Anders, unpublished results).

(ii) Other genes involved in DNA and nucleotide metabolism.

RhCMV encodes all of the enzymes found in HCMV that are required for nucleotide metabolism, replication, and repair. These are uracil-DNA glycosylase (Rh146), ribonucleotide reductase (Rh72), and dUTPase (Rh101). The enzymes all have very high homologies to their HCMV counterparts, UL114 (69%), UL45 (61%), and UL72 (57%), respectively (Table 2). The ribonucleotide reductase homologue is structurally identical in size (97 kDa) to both the HCMV and MCMV proteins. This is in contrast to HSV-1, which encodes a functional protein of 124 kDa. The RhCMV protein lacks a homologue of the small subunit of ribonucleotide reductase and thus may not be functional. Also, like HCMV and MCMV, the RhCMV genome does not have an ORF that encodes a thymidylate synthetase or a thymidine kinase.

RhCMV ORF Rh132 is the HCMV UL97 homologue. The UL97 protein of HCMV is a phosphotransferase and has been shown to phosphorylate ganciclovir (47). Unlike the alphaherpesviruses, RhCMV does not encode any additional protein kinases. However, RhCMV does encode a homologue to the pyruvoyl decarboxylase gene (Rh106) found in HCMV. Consistent with HCMV, the putative protein homologue contains a single pyruvoyl decarboxylase enzyme prosthetic group represented by the sequence ALGSSLFN (single-amino-acid code) from amino acids 566 to 574.

(iii) Regulatory genes.

Examination of RhCMV ORFs for viral regulatory genes involved in the control of viral gene expression reveals that the virus encodes homologues of the major immediate-early region (Rh156), the UL36 to UL38 region (Rh60 to Rh66), one of the TRS1 transactivator genes (Rh230), the UL69 gene (Rh97), and the US3 transcription unit (Rh184). The major immediate-early region is colinear with that of HCMV and has matching splicing patterns (7). Transcription for the HCMV major immediate-early genes is controlled by a strong enhancer or promoter sequence at positions −733 to +625 (relative to the start site of transcription at +1) (2). In HCMV, TRS1 and IRS transactivate UL44 as well as other early genes (9, 37, 66).

(iv) Structural proteins.

Capsid assembly proteins are very similar among the betaherpesviruses. RhCMV is no exception, encoding homologues to the major capsid protein (Rh118), the large tegument protein (Rh78), and a minor capsid protein (Rh75). Also present in RhCMV is a UL80 (Rh109) homologue that is thought to be involved in assembly and packaging of DNA into the virion. RhCMV also encodes other structural proteins common to the betaherpesvirus group, including the upper matrix protein pp71 (Rh110), the lower matrix protein pp65 (Rh111 and Rh112), the large phosphoprotein pp150 (Rh55), and the small phosphoprotein pp28 (Rh136). Rh91 is homologous to the HCMV UL56 gene product. However, the predicted size of Rh91 is 88.2 kDa, compared with 130 kDa for HCMV. Although gaps are found throughout the protein, accounting for the size difference, the protein is still highly conserved at 70% identity. The HCMV UL56 gene product has been found associated with the nucleocapsid tegument fraction and is thought to be involved in capsid maturation (10).

(v) Glycoproteins.

A total of 21 glycoproteins were found within the RhCMV genome. These predictions were based on known homologies within GenBank. RhCMV encodes homologues to the gB (Rh89), gH (Rh104), gL (Rh147), gM (Rh138), gN (Rh102), and gO (Rh103) glycoproteins. In addition, homologues to UL37 (Rh62), UL42 (Rh68), UL116 (Rh148), UL121 (Rh155), US02 (Rh182), US03 (Rh184), and US11 (Rh189) are present in RhCMV. No homologues to HCMV glycoproteins gpUL04, gpUL16, or gpUS10 were found. Glycoprotein B (gB) is a prominent component of the viral envelope and is involved in virus attachment, penetration, and cellular spread (12). Many gB homologues carry the motif RXK/RR, which is the specific recognition motif for cleavage by the cellular protease furin. The gB homologues of varizella-zoster virus, bovine herpesvirus type 1, pseudorabies virus, equine herpesvirus, Marek's disease virus, and all known betaherpesviruses, including HCMV and human herpesvirus 6, contain this cleavage site (reviewed in reference 68). The RhCMV gB sequence contains several potential furin sites; however, none of these sites is an exact match to the sequence RXK/RR.

In HCMV, glycoprotein O (gO) is encoded by the UL74 gene (36). gO has homologues only among the betaherpesviruses (36). Similar to HCMV, the RhCMV gO homologue has numerous putative N- and O-linked glycosylation sites. gO has been shown to be dispensable for growth of the virus; however, a gO knockout virus is severely attenuated in growth and spread in culture (33). A tripartite complex of glycoproteins H, L, and O (gH/gL/gO) forms through disulfide bonds and has been implicated in viral entry (via membrane fusion), cell-to-cell spread, and virion maturation (36, 55, 71). Recent studies have found that gO represents a hypervariable region of the genome, with some sequences varying from each other by as much as 45% (55, 58). Currently, we do not know if RhCMV gO is a hypervariable region. Sequencing of primary isolates will indicate if RhCMV is similar to HCMV in this respect.

HCMV gH remains membrane bound and is the anchor for the gH/gL/gO fusion complex. RhCMV gL has two putative N-linked glycosylation sites defined by the motif NXT/S, where X is any amino acid (46). HCMV gL has one glycosylation site, whereas MCMV is predicted to have five. The RhCMV gM homologue (Rh138) is predicted to have eight putative transmembrane domains, based on the hydrophobicity plot. A similar profile is seen with both HCMV and MCMV gM proteins. The HCMV gM protein has five putative glycosylation sites, only one of which is present in RhCMV gM. However, the characteristic HCMV gM acidic C-terminal tail is present in Rh138. Unlike MCMV, RhCMV encodes a gN protein. Therefore, the gM/gN structural complex is preserved in RhCMV. In HCMV, gN has been designated a hypervariable region, with at least four distinct subtypes identified (reviewed in reference 56).

(vi) Immunomodulators.

Like most DNA viruses, RhCMV encodes a total of 13 putative immunomodulators. Six of these show significant homology to known HCMV immunomodulatory proteins, including UL111 (Rh143), UL144 (Rh163), US02 (Rh182), US03 (Rh184), and US11 (Rh189). The UL36 gene of HCMV encodes a cell death suppressor, denoted vICA (62). Immediately downstream of the HCMV UL36 gene locus is a mitochondrion-localized inhibitor of apoptosis (vMIA), a product of the UL37 gene (27). RhCMV ORFs Rh62 and Rh66 show homology to UL37. Recently, the HCMV UL118/119 spliced transcripts have been found to encode viral Fcγ receptor homologues (4). RhCMV potentially encodes these receptors through the Rh151 and Rh152 ORFs. RhCMV encodes two genes that show homology to the HCMV UL36 genes, Rh60 and Rh61. Rh60 and Rh61 may be spliced to form a functional RNA transcript. The putative ORFs rh96 and rh98 show homology with the immunoglobulin heavy chain. While rh186 is structurally similar to the US6 family members, the protein does not have an HCMV homologue by Blast analysis. Rh143 is an interleukin-10 (IL-10) homologue that has previously been shown to be expressed and may act as a cytokine synthesis inhibitory factor in vivo (48). Rh163 has homology to a tumor necrosis factor (TNF) receptor, which may play a role in preventing apoptosis. The US2, US3, US8, and US11 homologues have been shown to downregulate major histocompatibility complex (MHC) class I (24).

(vii) Unique to RhCMV.

RhCMV encodes a putative cyclooxygenase-2 or prostaglandin E₂ homologue. A total of six exons make the putative cyclooxygenase-2 transcript. RhCMV has several gene duplications, one of which is a duplication of the pp65 homologue. In addition, the UL11 gene appears to be present in three copies (Rh21 to Rh23), the US14 homologue has four copies (Rh194 to Rh197), and the US28 gene homologue has six copies (Rh214 to Rh216, Rh218, and Rh220).

RhCMV gene families.

RhCMV encodes eight different gene families that are also present in HCMV (Table 4).

TABLE 4.

RhCMV gene families homologous with HCMV genes

Index ORF of gene family	Members of gene family	Family characteristicsa	Homologous HCMV family	Reference(s)
Rh22	Rh05, Rh17, Rh21, Rh22, Rh23, Rh24, Rh26	Genome location, structure, glycosylation sites	RL11
Rh43	Rh43, Rh59	Genome location, structure	UL25	58
Rh56	Rh56, Rh107, Rh214, Rh215, Rh216, Rh218, Rh220	GPCR motif, 7-transmembrane domains, S-T-rich carboxy terminus	7TM	59-62
Rh110	Rh110, Rh111, Rh112	Genome location, structure, phosphoprotein	UL82
Rh181	Rh181, Rh225, Rh226	Genome location, structure	US01
Rh185	Rh182, Rh184, Rh185, Rh187, Rh189	Structure, function	US06
Rh190	Rh190, Rh192, Rh194, Rh195, Rh196, Rh197, Rh198, Rh199, Rh200, Rh201, Rh202	Genome location, structure	US12
Rh203	Rh40, Rh42, Rh47, Rh50, Rh60, Rh61, Rh69, Rh203, Rh204, Rh230	OCCXD/EX1-40XxoG, two stretches of hydrophobicity, carboxy-terminal acidic residues	US22	65, 66

^aGPCR, G protein-coupled receptor; 7TM, seven-transmembrane.

(i) RL11 family.

The RL11 family is represented by seven putative ORFs. The RL11 family of membrane glycoproteins is represented by RhCMV ORFs Rh05 (UL153), Rh17 (UL11), Rh20 (UL06), Rh21 (UL11), Rh22 (UL11), Rh23 (UL11), and Rh25 (UL09). The function of these proteins in HCMV has not been completely elucidated. However, the RL11 sequences are fairly conserved within the long repeat sequence of herpesviruses.

(ii) UL25.

The UL25 family of proteins is represented by two RhCMV ORFs that are also conserved in MCMV and RCMV. These are RhCMV ORFs Rh43 (UL25) and Rh59 (UL35). The HCMV UL25 family members may be virion associated, although the data are controversial (6, 79).

(iii) Seven-transmembrane family.

Proteins within this family have seven transmembrane domains; some family members have been shown to act as G protein-coupled receptor homologues (67). Family members include UL33 (Rh56), UL78 (Rh107), and US28 (Rh214, Rh215, Rh216, Rh218, and Rh220). There is no apparent RhCMV homologue to HCMV US27, also a seven-transmembrane family member.

(iv) UL82 family.

HCMV UL82 family members comprise the upper and lower matrix proteins pp71 and pp65. RhCMV homologues to these proteins are Rh110 (pp71) and Rh111/Rh112 (pp65). The RhCMV pp71 protein has 39% identity with HCMV pp71. The two pp65 copies of RhCMV have 32% (Rh111) and 35% (Rh112) identity with the HCMV protein and 39% identity between themselves. Immunologically, the pp65 homologue encoded by Rh112 ORF elicits a greater T-cell response than that generated against Rh111 (L. Picker, personal communication).

(v) US01 family.

The US01 family of proteins has three members in HCMV and includes the US01, US31, and US32 proteins. RhCMV encodes homologues to all three of these proteins (Rh181, Rh225 and Rh226, respectively) that are at colinear positions within the genome.

(vi) US06 family.

The HCMV US06 family of proteins show both structural and functional similarities with one another. The majority of these family members are glycoproteins, although US06 proteins have been shown to function as downregulators of MHC class I on the cell surface. RhCMV encodes five potential US06 family members capable of immunomodulation: US02 (Rh182), US03 (Rh184), US06 (Rh185), US08 (Rh187), and US11 (Rh189). Historically, US02 and US03 were considered US02 family members, but we have chosen to classify the RhCMV proteins as US06 family proteins due to structural and functional features in common with US06.

(vii) US12 family.

There are 11 US12 family members within the U_S region of RhCMV; four of these (Rh194, Rh195, Rh196, and Rh197) are duplications of US14. The remaining seven ORFs include Rh190 (US12), Rh192 (US13), Rh198 (US17), Rh199 (US18), Rh200 (US19), Rh201 (US20), and Rh202 (US21). Family classification is based on structural similarities among the proteins. The function of these proteins has yet to be elucidated.

(viii) US22 family.

US22 family members are found throughout the genomes of both RhCMV and HCMV. RhCMV encodes 10 homologues to HCMV ORFs. These include Rh40, Rh42, Rh47, Rh50, Rh60, Rh61, Rh69, Rh203, Rh204, and Rh230. Proteins of the US22 family have two stretches of hydrophobic sequences and predominantly acidic carboxy termini; the RhCMV proteins meet both of these criteria.

Phylogenetic analysis.

A phylogenetic analysis of the relationship of RhCMV to other herpesviruses is shown in Fig. 4. Six ORFs found in HSV-1, KSHV, MCMV, RCMV, and ChCMV were examined by bootstrap analysis with the maximum parsimony method. Alignments were performed with ClustalW and analyzed to ensure that sequences were properly aligned. ClustalW data were then used to generate a phylogenetic tree for DNA polymerase, glycoprotein B, helicase, major capsid protein, single-stranded-DNA-binding protein, and uracil N-glycosylase. The HCMV proteins were used as the root for the analysis, and thus the figure shows divergence from these proteins. The tree indicates that the virus most closely related to HCMV is ChCMV, followed by RhCMV, MCMV, RCMV, KSHV, and finally HSV-1.

FIG. 4.

Phylogenetic analysis. Phylogenetic trees for alpha-, beta-, and gammaherpesviruses. Six proteins were selected for phylogenetic analysis with the ClustalX program. Proteins were aligned by ClustalW. The alignments were adjusted to remove any unnecessary gaps, and then tree data was calculated. Trees were drawn with the TreeView program. The measure of divergence is presented as a scale at the bottom of each tree. Proteins used for the analysis were DNA polymerase, glycoprotein B (gB), helicase, major capsid protein, single-stranded-DNA-binding protein, and uracil N-glycosylase.

DISCUSSION

The RhCMV genome has now been completely sequenced. The RhCMV strain 68-1 genome is 221,459 bp in length and potentially encodes over 200 proteins. Like the other sequenced betaherpesviruses, RhCMV contains ORFs that are homologous to genes from both the unique long (U_L) and unique short (U_S) regions, similar to other viruses within the herpesvirus family. However, unlike human and chimpanzee CMVs, RhCMV does not have repeats between the putative U_L and U_S regions. Restriction analysis of the RhCMV genome further indicated that the genome does not isomerize, which is not uncommon within the betaherpesvirus family. Consistent with this finding, probes specific for each terminus failed to detect the altered terminal fragments that would be predicted by isomerization on Southern blots of RhCMV genomic DNA (13). Moreover, the closely related simian CMV strain Colburn was likewise found by restriction analysis not to isomerize (45). Thus, based on the organization of the RhCMV genome, the genome is most similar to the tupaia herpesvirus, which was recently sequenced and is the prototypic member of the F group of herpesvirus genomes (5).

The complete DNA sequence of RhCMV was compared to all of the published herpesviruses genomes within GenBank. From this analysis, we found that RhCMV and the other herpesvirus genomes are very similar at the genetic and nucleotide levels (34). For HCMV, the G+C content is about 57%, whereas for RhCMV the G+C content is around 49%; however, slight differences are seen at the termini, possibly due to the lack of repeats. Even so, homologous proteins are found within terminal regions, suggesting that host determinants may account for these differences at the nucleotide level.

The sequence analysis revealed that RhCMV encodes homologues to four regions of HCMV that are transcribed at immediate-early times: the major immediate-early transcripts (IE-1 and IE-2), which have been described previously (7), UL36 to UL38, US3, and TRS1. In HCMV infection, these proteins are expressed early during infection and are involved in the regulation of gene expression throughout the virus's life cycle. MCMV does not contain a US3 homologue; however, the US3 locus is dispensable for growth in tissue culture (42). In addition, MCMV does not appear to encode the second transcript of the UL37 gene, which may not be functional. RhCMV encodes both transcripts of the UL37 gene. RhCMV UL37 therefore may function like its HCMV counterpart (1, 14). The TRS1 homologue in RhCMV shows significant homology to the HCMV protein (35% identity). TRS1 in HCMV has been shown to be expressed at immediate-early times and to act as a transactivator of several early promoters (46, 66).

At least two additional RhCMV gene homologues may have regulatory functions. The UL121 homologue Rh155 is located colinearly with the HCMV protein and has homology to the ICP0 protein of HSV-1. ICP0 has been implicated in viral switching between latent and lytic infection and has been shown to transactivate other herpesvirus promoters (20). However, the Rh155 protein does not contain a zinc finger motif and may not function in the same manner as ICP0. Rh155 does contain a highly charged carboxy terminus, which is a feature of some trans-acting transcriptional activators. Rh97 encodes a homologue of the UL69 transactivator of HCMV and is also highly conserved at 59% identity. Taken together, regulatory proteins and transactivators are highly conserved between HCMV and RhCMV.

Examination of the enzyme-coding capacity of RhCMV indicates that the virus encodes the same enzymes as HCMV and, like HCMV, lacks an identifiable thymidine kinase gene. The RhCMV DNA polymerase has been reported previously and has a sequence identical to the one described here (70). Present within the RhCMV genome is a homologue of the UL97 gene of HCMV. The UL97 product has protein kinase activity and has been shown to phosphorylate the nucleoside analog ganciclovir (47). Previous studies have shown that RhCMV is sensitive to ganciclovir treatment (70). Homologues to the HCMV DNA helicase-primase and the DNA repair enzymes dUTPase and DNase were present in RhCMV. The RhCMV enzymes are highly conserved with the HCMV proteins (Table 2). [Unique to the betaherpesviruses are homologues to the UL77 protein.] UL77 is a pyruvoyl decarboxylase homologue that can be found in HCMV virions. Pyruvoyl decarboxylase enzymes are involved in the synthesis of cellular polyamines, which is increased during HCMV infection in vitro (35). When polyamines are decreased, so are the number of HCMV particles released from infected cells (26).

Another group of proteins that are highly conserved between HCMV and RhCMV are the structural proteins. As expected, RhCMV encodes homologues to the major capsid protein, the minor capsid protein, and the capsid assembly protein. In addition to the capsid proteins, there are homologues to most of the known tegument proteins, including UL32 (pp150), UL48 (large tegument protein), UL82 (pp71), UL83 (pp65), and UL99 (pp28). RhCMV does not encode a copy of the UL65 protein, which is also absent in MCMV. Interestingly, RhCMV encodes an additional copy of the UL83 ORF (pp65). Current data are lacking on whether both proteins are present in virions, but immunological data indicate that there is a T-cell response to both proteins, suggesting that each is expressed (L. Picker, personal communication). The identity between the two RhCMV copies is 39%, while Rh111 has 32% identity and Rh112 has 35% identity to HCMV UL83. Such a low identity score suggests that either a gene duplication event occurred during the evolution of the virus or that these two genes are distinct and have separate functions.

The sequenced betaherpesviruses all contain gene families, although this feature is not common to all herpesviruses. Sequence analysis determined that RhCMV encodes gene homologues to all the HCMV families. Within the HCMV genome, nine gene families have been described; we have described eight for RhCMV. In HCMV, two families that are similar in structure, US2 and US6, have both been shown to downregulate MHC class I expression on the cell surface. We have chosen to classify US02-like family members as US06 family members due to structural similarities.

Immunomodulation is responsible for a large part of a virus's survival and propagation within a host. Herpesviruses encode numerous immunomodulatory proteins that are thought to be involved in immune evasion. The genomic analysis of RhCMV demonstrates that the virus encodes numerous immunomodulators. On average, the identity of these proteins with their HCMV counterparts is 21%. This may reflect host adaptation and not imply a lack of function. A recent publication has shown that RhCMV encodes and expresses an IL-10 homologue (48). The function of the RhCMV protein has yet to be elucidated, but it may function as a cytokine synthesis inhibitory factor. Like HCMV, RhCMV encodes a TNF receptor homologue. A TNF receptor homologue may be used as a mechanism to prevent the cells' progression towards apoptosis. However, there is insufficient evidence to determine if the RhCMV TNF receptor homologues are expressed and if the protein binds cytokines.

Viral infection may induce apoptosis within the cells they infect. In addition to the TNF receptor, HCMV encodes the apoptosis inhibitors vICA (UL36) and vMIA (UL37). The product of the vICA ORF inhibits apoptosis by binding to the prodomain of caspase-8, preventing activation (62). RhCMV has two potential ORFs that show homology to the UL36 gene. However, it is not known if these genes are spliced to yield one message or if they function independently of one another. HCMV's vMIA protein acts by inhibiting Fas-mediated apoptosis at a point downstream of caspase-8 activation and Bid cleavage but upstream of cytochrome c release (27). Rh62 and Rh66 are homologous to vMIA; however, the function of these proteins is not known.

RhCMV potentially encodes spliced transcripts homologous to the Fcγ receptor glycoproteins (4). The Fcγ receptors have recently been identified in HCMV. The HCMV proteins belong to the immunoglobulin gene superfamily and represent a diverse group of Fcγ receptor structures. As mentioned above, RhCMV encodes several homologues to US06 family members. US06 proteins are present in HCMV, in other herpesviruses, and in poxviruses. Many of these have been characterized and shown to downregulate MHC class I on the cell surface. MHC class I presentation of antigen is a significant host defense directed towards the clearance of pathogens. RhCMV also encodes two ORFs that show low sequence homology to the immunoglobulin heavy chain (rh96 and rh98). These RhCMV proteins may prove to be additional herpesvirus immunomodulators or the equivalent of the HCMV UL16 protein.

The viral cyclooxygenase-2 homologue may also have an immunomodulatory function. Cellular cyclooxygenase-2 is produced during an inflammatory response when immunoglobulin molecules are cross-linked. This triggers a rapid permeabilization of the lipid membrane through the production of arachidonic acid and releases inflammatory mediators into the surrounding area. Many anti-inflammatory drugs such as aspirin are inhibitors of arachidonic acid metabolism. Recently, Zhu et al. demonstrated that cyclooxygenase-2 is induced during HCMV infection and performs an important function supporting viral replication (76). Increased cyclooxygenase-2 activity early in infection promotes accumulation of the viral transcriptional regulatory protein IE-2, which is in turn required for subsequent viral transcription and DNA replication. RhCMV-encoded cyclooxygenase-2 could play an analogous role. RhCMV cyclooxygenase-2 could also possibly function as a transcriptional control molecule or be involved in latency by slowing viral replication. It remains to be established whether the RhCMV homologue is active and, if so, how the gene's expression is regulated.

The seven-transmembrane family member proteins are G protein-coupled receptors that are present on the cell surface and couple to G proteins to activate extracellular ligands (11). Once activated, the G proteins transduce signals from the cell surface to an effector molecule in order to modulate intracellular functions. RhCMV encodes all but one of the G protein-coupled receptors found in HCMV. However, RhCMV does encode five potential copies of the US28 protein, which has been shown to function as a chemokine receptor and to elicit cell migration in vitro. (68). Structurally, the G protein-coupled receptor homologues are very similar to their HCMV counterparts, although identity between the proteins is about 36%. All of the G protein-coupled receptors in HCMV are transcribed. It will be an interesting point of study in RhCMV to determine whether the seven-transmembrane family members are transcribed and whether they function.

Phylogenetic analysis of HCMV and comparison with ChCMV, RhCMV, MCMV, RCMV, KSHV, and HSV-1 for several different conserved genes revealed that the primate cytomegaloviruses are most closely related to HCMV. The analysis also demonstrates that both MCMV and RCMV are very distant from HCMV on the evolutionary tree and therefore may not be the best model with which to study HCMV pathogenesis and host immune response to the virus. Although chimpanzees are most closely related to humans evolutionarily, limited numbers of chimpanzees in the wild make ChCMV infection of chimps an unsuitable study model. With the availability of the genomic sequence, RhCMV will offer significant advantages over the other cytomegalovirus model systems. RhCMV is highly homologous to HCMV, encodes many of the same proteins, and has a similar genomic structure, and there is a relevant animal model with which to study both cytomegalovirus and cytomegalovirus/SIV infections in the natural host.