|Home | About | Journals | Submit | Contact Us | Français|
The recognition of naturally occurring rhadinoviruses in macaque monkeys has spurred interest in their use as models for human infection with Kaposi sarcoma-associated herpesvirus (human herpesvirus 8). Rhesus macaques (Macaca mulatta) and pig-tailed macaques (Macaca nemestrina) were inoculated intravenously with rhadinovirus isolates derived from these species (rhesus rhadinovirus [RRV] and pig-tailed rhadinovirus [PRV]). Nine rhadinovirus antibody-negative and two rhadinovirus antibody-positive monkeys were used for these experimental inoculations. Antibody-negative animals clearly became infected following virus inoculation since they developed persisting antibody responses to virus and virus was isolated from peripheral blood on repeated occasions following inoculation. Viral sequences were also detected by PCR in lymph node, oral mucosa, skin, and peripheral blood mononuclear cells following inoculation. Experimentally infected animals developed peripheral lymphadenopathy which resolved by 12 weeks following inoculation, and these animals have subsequently remained free of disease. No increased pathogenicity was apparent from cross-species infection, i.e., inoculation of rhesus macaques with PRV or of pig-tailed macaques with RRV, whether the animals were antibody positive or negative at the time of virus inoculation. Coinoculation of additional rhesus monkeys with simian immunodeficiency virus (SIV) isolate SIVmac251 and macaque-derived rhadinovirus resulted in an attenuated antibody response to both agents and shorter mean survival compared to SIVmac251-inoculated controls (155.5 days versus 560.1 days; P < 0.019). Coinfected and immunodeficient macaques died of a variety of opportunistic infections characteristic of simian AIDS. PCR analysis of sorted peripheral blood mononuclear cells indicated a preferential tropism of RRV for CD20+ B lymphocytes. Our results demonstrate persistent infection of macaque monkeys with RRV and PRV following experimental inoculation, but no specific disease was readily apparent from these infections even in the context of concurrent SIV infection.
The identification of Kaposi’s sarcoma (KS) and Pneumocystis carinii pneumonia in previously healthy homosexual men led to the first case definitions of AIDS (7, 8). In the years that followed, KS was recognized as the most common malignancy in human AIDS patients. Epidemiologic evidence suggested the presence of a second agent or additional cofactors which acted in conjunction with the human immunodeficiency virus (HIV) to produce the neoplasm (2, 33). In 1994, gammaherpesvirus DNA sequences were detected in KS tissues obtained from human AIDS patients by representational difference analysis (13). The virus that was subsequently identified has been named KS-associated herpesvirus (KSHV), or human herpesvirus 8 (HHV-8). Viral sequences have been detected in KS patients whether positive or negative for HIV (5, 14, 30), in body-cavity-based lymphomas (a rare form of non-Hodgkin’s lymphoma) (9, 10), and in multicentric Castleman’s disease (MCD) (22, 35).
We recently recognized a distinct new gammaherpesvirus in rhesus macaques (21). The presence of the virus was first suspected on the basis of immunoreactivity to herpesvirus saimiri antigens detected in serologic surveys of colony monkeys. The virus was isolated from peripheral blood mononuclear cells (PBMC) and could be grown lytically and to high titer in primary rhesus monkey fibroblast cells. Sequencing of the complete DNA polymerase and glycoprotein B genes revealed a closer relatedness to KSHV than to herpesvirus saimiri, Epstein-Barr virus, or any other herpesvirus (21). Furthermore, an interleukin-6 (IL-6) gene homolog, analogous to a virus-encoded IL-6 homolog found in KSHV but not in any other viruses, was identified following open reading frame 11. The recent completion of the DNA sequence of an independent isolate in Oregon revealed similarity in organization and sequence with KSHV across the full length of the genome (34). Collectively these findings demonstrated the presence of a distinct gammaherpesvirus, rhesus rhadinovirus (RRV), that is closely related to KSHV. Short stretches of KSHV-like sequences have also been amplified from tissues of macaque monkeys with retroperitoneal fibromatosis (3, 4, 32).
Serologic surveys of normal rhesus macaques housed at the New England Regional Primate Research Center (NERPRC) and elsewhere showed that greater than 90% of adult macaques are immunoreactive to RRV (21). Rhesus macaques have not previously been screened for the presence of this virus, and the vast majority of animals that have been used in biomedical research have been potentially infected with this agent. The effects of the virus on the pathogenesis of spontaneous disease and on previous experimental findings need to be critically evaluated. Attempts to correlate infection by RRV or closely related viruses with lymphoma and retroperitoneal fibromatosis (3, 32, 34) are complicated by the high rates of natural infection among normal animals. Controlled studies utilizing RRV-free animals and well-defined inocula should facilitate the investigation of disease causation. Furthermore, it is important to document the characteristics of primary infection and subsequent persistence as a potential model for KSHV infection of humans. Reported here are the results of experimental inoculation of normal and immunodeficient macaques with two distinct rhadinovirus isolates from nonhuman primates.
All macaques (Macaca mulatta and M. nemestrina) were housed at the NERPRC in accordance with standards of the American Association for Accreditation of Laboratory Animal Care and Harvard Medical School’s Internal Animal Care and Use Committee. Normal macaques were serologically negative for simian T-lymphotropic virus, type D retrovirus, herpesvirus type B, and simian immunodeficiency virus (SIV) (16, 19). Macaques experimentally inoculated with SIV and/or nonhuman primate-derived rhadinoviruses were individually housed in biolevel 2 and 3 containment facilities as previously described (28, 36a). Immunodeficient animals had previously been inoculated intravenously with one of two strains of SIV (SIVmac251 or SIVmac239), the history, preparation and in vivo and in vitro properties of which have been described extensively (17, 18, 20, 23, 26). These animals were included in a variety of infectivity and pathogenesis studies and received neither antiretroviral agents nor antimicrobial prophylaxis.
Five pig-tailed macaques (M. nemestrina) and 16 rhesus macaques (M. mulatta) were inoculated intravenously with 5 × 105 50% tissue culture infective doses of the H26-95 isolate of RRV and H19545 isolate of pig-tailed rhadinovirus (PRV) prepared as previously described (21). The serologic status to previous nonhuman primate rhadinovirus infection was established prior to inoculation. Same-species and cross-species transmissions of the respective rhadinoviruses were carried out in seronegative and seropositive animals. Species, animal number, and rhadinovirus and SIV serology status prior to inoculation for 21 animals are listed in Table Table1.1.
Macaque rhadinovirus was purified from productively infected rhesus monkey fibroblast cell cultures, lysed, coated onto plates, and used for the detection of reactive antibodies by enzyme-linked immunosorbent assay (ELISA) as previously described (21).
Following experimental rhadinovirus inoculation, all animals were examined daily and body temperature and clinical data were recorded with an implanted microchip and transponder (Bio Medic Data Systems, Maywood, N.J.). Oral mucosal, skin, and lymph node biopsies and PBMC were obtained prior to inoculation and at 2, 4, and 12 weeks postinoculation. Tissues were fixed in 10% neutral buffered formalin and snap frozen at −70°C. Blood for viral antibody response, viral load, and complete blood count were obtained at 0, 1, 2, 4, 8, and 12 weeks and monthly thereafter.
Viral isolation was performed as previously described (21). Viral load in infected macaques was estimated by quantitative cocultivation of PBMC with primary rhesus fibroblasts. This technique was adapted from that previously described to quantitate SIVmac viral load (20, 24). Briefly, PBMC were purified, counted in a hemocytometer, and subsequently cocultured in various numbers from 106 to 152 PBMC via serial dilutions with 105 rhesus fibroblasts. The presence of cytopathic effect was evaluated at 14 to 21 days, and the number of PBMC required to recover RRV was calculated. Evaluation of the levels of SIV RNA in plasma were also performed as previously described (20, 36).
Formalin-fixed paraffin-embedded and snap-frozen tissues were used in immunohistochemical procedures to define the immunophenotype of cells within lymphoid tissue as previously described (38). Briefly, tissue sections were fixed in 2% paraformaldehyde and immunostained with an avidin-biotin-horseradish peroxidase complex technique with diaminobenzidine chromogen. The primary antibodies used were Snv71.1 (from C. Collignon and C. Thiriart, SmithKline Beecham Biologicals, Rixensart, Belgium) for SIV gp120, B1 for CD20-positive B lymphocytes, EBM11 (Dako Corp., Carpinteria, Calif.) for CD68-positive macrophages, DK25 for CD8-positive T lymphocytes, Nu-Th/1 (from M. Yokoyama and Y. Matsuo, Nicheri Research Institute, Fukuoka, Japan) for CD4-positive T lymphocytes, CR3/43 for HLA-DR, and MIB-1 (Immunotech, Westbrook, Maine) for Ki67.
To examine biopsy and necropsy tissues for the presence of RRV, DNA was extracted from fresh frozen tissues, using a QIAmp tissue kit (Qiagen, Valencia, Calif.) according to the manufacturer’s instructions. DNA was eluted in 30 to 50 μl of sterile water treated with diethylpyrocarbonate, and PCR was performed as described below. To analyze cell types harboring RRV, PBMC were separated from whole blood from a rhesus macaque (270-97) 13 days after inoculation with RRV, using standard Ficoll isolation techniques as instructed by the manufacturer (Organon Teknika, Malvern, Pa.). Isolated PBMC were washed in 1× phosphate-buffered saline with 1% heat-inactivated fetal calf serum, resuspended in 2 ml of RPMI 1640 medium containing 5% fetal calf serum, and counted with a hemacytometer. Cells were pelleted at 500 × g, resuspended at 10 × 106 cells/ml, aliquoted into Falcon tubes with 5 × 106 to 10 × 106 cells/tube, and repelleted. Supernatant was decanted, and pelleted cells were resuspended and labeled with 100 μl of CD20-fluorescein isothiocyanate (Becton Dickinson [BD]) CD4-phycoerythrin (PE) (Ortho), CD8 peridinin chlorophyll protein (BD), CD14-PE (BD), or CD3-PE (PharMingen). Cells were incubated on ice in the dark for 30 min, then washed in 1× phosphate-buffered saline pelleted, resuspended in 200 μl 2% paraformaldehyde, and stored at 4°C until analyzed. Cells were sorted in a Vantage flow cytometer (BD). Sorted cells were then pelleted, decanted, and stored frozen at −80°C.
Genomic DNA from cells was isolated by using a Promega Wizard kit with a modified protocol. Briefly, cells were lysed in 300 μl nuclei lysis solution containing proteinase K and incubated overnight at 37°C. Protein precipitation solution (100 μl) was added to the nuclear lysate, which was then vortexed and centrifuged at 14,000 × g. The supernatant was transferred to a new tube containing 300 μl of isopropanol to precipitate DNA. DNA was resuspended in sterile diethylcarbonate-treated H2O at a concentration of 500, 1,000, or 2,000 cells/μl.
Nested PCR was performed on DNA isolated from PBMC and tissue as previously described. Outer primers (RRV8758 [5′-GCC AAA CCG TCT CTC ATT CT] and RRV9767 [5′-CGA CCC CCA TCC CCA CAT AG]) were used to produce a 1,029-bp product. Briefly, PCR was performed with 0.2 μM each primer, 100 μM deoxynucleoside triphosphates, 1.5 mM MgCl2, 2.5 IU of taq DNA polymerase, and buffer (50 mM KCl, 10 mM Tris [pH 9.0]) in a 50-μl reaction volume. Amplification was performed at 94°C for 30 s, 56°C for 60 s, and 72°C for 45 s for a total of 30 cycles in a Perkin-Elmer 9600 thermocycler; 1 μl of this reaction was used as template for a second reaction utilizing internal primers (RRV8832 [5′ CCC TCG CCA CAC AAA ACC AG] and RRV9009 [5′ GGC GCG GAG TCT AAT GAA AA]). Precautions against PCR contamination including physical separation of areas used for DNA isolation, PCR, and post-PCR manipulations and use of appropriate negative controls were utilized as previously described (25). PCR products were resolved in a 2% ethidium bromide-stained agarose gel.
A rhadinovirus was isolated from PBMC of a pig-tailed macaque (animal 19545). This virus was called PRV19545 for PRV isolate 19545, to distinguish it from RRV isolates. Virus stocks of PRV19545 were prepared as described previously for RRV (21). Sequencing of the R1 gene (15) and gB gene (1) revealed PRV19545 to be closely related to but distinct from RRV isolate 26-95 (21). Sequencing of full-length gB genes from nine rhadinovirus isolates from three species of macaques has recently revealed all to be rather closely related, with up to 7.2% divergence at the amino acid level (1). However, phylogenetic analysis demonstrated that the rhadinovirus isolates grouped according to species of origin, not primate facility of origin (1). These analyses suggested that RRV and PRV are closely related but distinct viruses. RRV26-95 and PRV19545 were inoculated intravenously into a total of 11 macaque monkeys in the absence of concurrent SIV infection (Table (Table1).1). Both rhesus and pig-tailed macaques were used for these experimental inoculations in order to examine the effects of same- versus cross-species infection on pathogenic potential (Table (Table1).1). Two of the rhesus monkeys that were inoculated with PRV were antibody positive to RRV at the time of inoculation. The other nine recipient monkeys were antibody negative. Blood samples were obtained at periodic intervals after inoculation and used for measurement of antibody responses and virus recovery.
All monkeys inoculated with RRV seroconverted to positive anti-RRV antibody status within the early weeks after inoculation (Fig. (Fig.1A).1A). The anti-RRV antibodies have persisted at high levels for as long as we have followed the animals. The three antibody-negative macaques that were inoculated with PRV seroconverted to positive anti-PRV antibody status within the early weeks after inoculation (Fig. (Fig.1B).1B). The anti-PRV antibodies have also persisted at high levels for as long as we have followed the animals. Animals infected with RRV or PRV made antibodies that reacted strongly with both RRV and PRV antigens. Serologic cross-reactivity between RRV and PRV was extensive, with a tendency toward slightly increased reactivity to the homologous virus (data not shown). These results demonstrate consistent, persistent infection of naive macaques by RRV and PRV. The two RRV-antibody-positive rhesus monkeys (190-96 and 195-96) that were inoculated with PRV exhibited an increase in antibody reactivity to PRV following inoculation (Fig. (Fig.1C),1C), suggesting a possible take of PRV in rhesus monkeys already infected with RRV.
Experimental infection of macaque monkeys with PRV and RRV was also demonstrated by virus recovery from PBMC using rhesus monkey fibroblast cultures. RRV or PRV was recovered from the majority of inoculated animals at two or more time points. Attempts to recover rhadinovirus from RRV-seronegative animals in control experiments have repeatedly failed. RRV or PRV was recovered at one or more time points from 11 of the 11 monkeys used in these studies. We also attempted to roughly quantitate the numbers of infectious cells in PBMC by performing RRV and PRV recoveries with serial threefold dilutions of cells starting at 106 PBMC in duplicate. Representative results from one set of monkeys inoculated at the same time with RRV are shown in Fig. Fig.2.2. RRV and PRV loads in this assay appeared to peak 1 to 4 weeks following the inoculation. The highest loads reached a numerical score of 5, which corresponds to virus recovery with 12,345 PBMC (Fig. (Fig.2).2). gB sequences from virus recovered from 190-96 and 195-96 that were already RRV positive when PRV19545 was inoculated did not correspond to the PRV19545 gB sequence (data not shown).
Five of 11 animals had a febrile reaction, as defined by an increase from baseline measurements of >2.0°C or absolute temperature >104.0°C at any time point. Fever was recognized as early as 48 h following inoculation and persisted for up to 14 days. Of the two rhesus macaques seropositive to RRV prior to inoculation (Mm 190-96 and 195-96), neither became febrile. Of the animals that became febrile, two were rhesus macaques inoculated with RRV (same species), two were rhesus macaques inoculated with PRV (cross species), and one was a pig-tailed macaque inoculated with RRV (cross species).
The complete blood counts for each animal included leukocyte count, absolute and relative counts of lymphocytes, neutrophils, eosinophils, basophils, and monocytes, hematocrit, platelet count, erythrocyte count, hemoglobin, mean corpuscular hemoglobin, mean corpuscular volume, and mean corpuscular hemoglobin concentration. Values remained within normal limits in all animals postinoculation. While there were minor fluctuations, there were no consistent changes in any of these values postinoculation. This was also true when the group was examined as a whole and when subgroups (RRV versus PRV, seropositive versus seronegative or M. mulatta M. nemestrina) were examined individually.
Skin and oral biopsies showed no unusual features. There were no clinically apparent cutaneous manifestations, and these animals remained healthy and free of disease throughout the 437 days of follow-up.
Clinically evident lymphadenopathy was detected in 8 of the 11 animals as soon as 2 weeks after rhadinovirus inoculation. Microscopic morphologic features were similar regardless of the rhadinovirus inoculum. Histologically, the lymphadenopathy was characterized at 2 weeks by marked paracortical lymphocytic hyperplasia which effaced normal architecture (Fig. (Fig.3A).3A). This paracortical expansion was accompanied by an abundance of small arborizing vessels lined by hypertrophied and hyperplastic endothelium (Fig. (Fig.3B).3B). The expanded paracortex contained increased numbers of immunoblasts, mitotic figures, and histiocytes and moderate numbers of small lymphocytes. Immunostaining revealed an increase in both CD20-positive B lymphocytes and CD3-positive T lymphocytes within the expanded paracortical zone. Dispersed within this diffuse paracortical expansion were occasional regions of follicular hyperplasia. Mantle zones were variably developed, and occasional follicles were confluent. An increased number of small blood vessels were found between the developing follicles.
Lymphadenopathy was absent in the two rhesus macaques that were seropositive at the time of PRV inoculation, 190-96 and 195-96. Histologically these animals lacked the marked vascular changes present in most of the other animals. Mild to moderate paracortical expansion was present in animal 195-96.
The paracortical expansion in animals with lymphadenopathy was less pronounced by 4 weeks postinoculation and had been replaced by extensive follicular hyperplasia. The most severe follicular hyperplasia was seen in rhesus macaques 380-96, 282-96, and 266-97, all of which received PRV. In two animals (380-96 and 266-97), follicular hyperplasia had completely effaced the normal architecture of the medulla and cortex at 4 weeks after infection (Fig. (Fig.3D).3D). Regions of vascular hyperplasia were still evident in the paracortex and surrounding follicles. Rhesus macaques 190-96 and 195-96 lacked any follicular hyperplasia.
By 12 weeks postinoculation, clinically recognized lymphadenopathy had resolved in all of the eight monkeys in which lymphadenopathy was seen. In all animals, there were increased numbers of involuted follicles characterized by small germinal centers and the presence of periodic acid-Schiff stain-positive material. These hyalinized follicles were occasionally penetrated by a single small blood vessel and less commonly surrounded by layers of loosely concentric lymphocytes (Fig. (Fig.3E3E and F). These follicular changes, although nonspecific, are unusual in normal animals and share features with the hyaline-vascular variant of Castleman’s disease in humans. In animals in which enlargement of the peripheral lymph nodes was still present, there was a combination of follicular hyperplasia and continued paracortical expansion with vascular hyperplasia.
Six rhesus monkeys that had been previously infected with SIV were also inoculated with PRV or RRV (Table (Table1).1). Three of these six were already antibody positive to RRV at the time of PRV inoculation (Table (Table1).1). Four additional rhesus monkeys that were RRV negative and SIV negative were coinoculated with RRV plus SIV or PRV plus SIV (Table (Table11).
Prior infection with SIV appeared to result in weaker and/or delayed antibody responses to RRV or PRV (Fig. (Fig.11 and and4).4). Rhesus monkeys 196-94, 181-90, and 229-91 were all SIV infected and RRV negative at the time of RRV or PRV inoculation, and all had weaker or delayed antibody responses to RRV or PRV (Table (Table1,1, Fig. Fig.1,1, and Fig. Fig.4).4). Coinoculation with RRV and SIV or PRV and SIV also appeared to weaken or delay the antibody response to PRV or RRV in most animals (Fig. (Fig.5).5). Despite the marginal or absent anti-PRV antibody response in coinoculated monkeys 84-96 and 91-96, PRV was recovered from PBMC on repeated occasions following the PRV-plus-SIV inoculation, and thus these animals were clearly infected by the PRV inoculation.
Interpretation of the lymph node changes in immunodeficient macaques was confounded by preexisting pathology in the lymph nodes, including follicular hyperplasia and dysplasia, findings common in SIV-infected macaques. Clinically there was enlargement of lymph nodes (lymphadenopathy) relative to the preinoculation state in all animals following rhadinovirus inoculation, and this enlargement was most severe in animals rhesus monkeys 229-91 and 168-94. In immunodeficient RRV-negative rhesus macaques (229-91, 181-90, and 196-94) inoculated with rhadinovirus, histologic changes included paracortical expansion and vascular hyperplasia accompanied by florid follicular hyperplasia. In seropositive animals, paracortical expansion was seen to a lesser degree and there was exacerbation of follicular hyperplasia. The specificity of these changes is unknown.
Of the six rhesus macaques that were previously infected with SIV and subsequently inoculated with RRV or PRV, four have died. Three of these four PRV/RRV-inoculated rhesus macaques that were previously infected with SIV died with thrombosis, vascular hypertrophy, and nonsuppurative vasculitis of pulmonary vessels. Aseptic proliferative endocarditis was present in two cases (Fig. (Fig.3G).3G). These changes are characteristic of SIV arteriopathy, a condition of unknown etiology recognized commonly in SIV-infected macaques (12). Pulmonary lymphocytic infiltrates were present in the three animals with arteriopathy and accompanied by renal lymphoid infiltrates in two cases. These changes, indicative of a lymphoproliferative disorder recognized in SIV-infected macaques (11), are of unknown etiology and a common morphologic feature found at necropsy.
Animals coinoculated with SIVmac251 and PRV or RRV developed a clinical lymphadenopathy by 2 weeks postinoculation. Histologic features in these animals were similar to those seen in 8 of the 11 immunologically normal macaques described above, including extensive vascular hyperplasia and paracortical expansion. The degree of vascular hyperplasia within lymph nodes was more florid than that seen in immunologically normal animals and persisted in three animals (rhesus macaques 121-96, 91-96, and 84-96) at 12 weeks postinoculation.
Monkeys coinoculated with SIV and RRV or SIV and PRV also appeared to have an attenuated antibody response to SIV (Fig. (Fig.6).6). Coinfected animals not only had weaker antibody responses but also had shorter mean survival times than 40 SIVmac251-infected controls studied previously (155 days versus 560 days; P < 0.019). Coinoculated monkeys developed SIV RNA loads in plasma that were about 10-fold higher than those found in historical controls inoculated with the same stock of SIVmac251 only (Fig. (Fig.7).7). Differences at weeks 1, 2, 4, and 12 were statistically significant by the Mann-Whitney rank sum test (P = 0.020 to 0.048). Animals that were coinfected with SIV and RRV or PRV at the same time died with a variety of opportunistic infections characteristic of simian AIDS, including those caused by Cryptosporidium parvum, P. carinii, and enteropathogenic Escherichia coli (Table (Table2).2). Pulmonary lymphocytic infiltrates were present in two of four animals. In contrast to the animals previously infected with SIV and then inoculated with rhadinovirus, pulmonary vascular lesions were not identified in any of these four animals (Table (Table2).2).
Lymph node, oral mucosa, skin, and PBMC were obtained at 2, 4, and 12 weeks postinoculation and tested for the presence of RRV by PCR (Table (Table3).3). Viral DNA was amplified from all three tissues and from PBMC. Samples from RRV-inoculated, SIV-negative animals were PCR positive for RRV sequences by 2 weeks postinoculation in lymph node (four of four), oral mucosa (two of four), skin (one of four), and PBMC (one of three). Virus persisted in the tissues to at least 12 weeks postinoculation, as demonstrated by PCR in lymph node (three of four), skin (two of four), and PBMC (three of four). Fewer tissues were positive for RRV by PCR in two animals that were coinoculated with RRV and SIVmac251.
To determine the cell type infected in peripheral blood, we performed PCR using sorted cells from PBMC. PBMC were harvested from pig-tailed macaque 270-97 at 2 weeks after inoculation with RRV. Cells were stained for CD20, CD8, CD4, and CD14 and were sorted by flow cytometry. DNA was isolated from the sorted, pelleted cells, and PCR was performed to determine the primary cell type infected with RRV. CD20+ cells were positive at 15,000 cells per reaction and remained positive down to a dilution of 3,200 cells per reaction (Fig. (Fig.8).8). CD8+ cells were positive at 15,000 cells per reaction and were negative at dilutions of 7,500 cells or less. CD4+ and CD14+ cells were PCR negative for RRV. This dilution analysis indicated that the CD20+ B lymphocyte is the primary cell type infected with RRV in M. nemestrina.
Experimental inoculation of normal, seronegative, juvenile macaques with RRV or PRV produced a clinical syndrome characterized by a mild febrile reaction and moderate lymphadenopathy. The lymphadenopathy was initially characterized by marked paracortical expansion accompanied by vascular hypertrophy and hyperplasia which was replaced by follicular hyperplasia in 4 to 8 weeks and which resolved by 12 weeks postinoculation in most animals. These histologic features, while nonspecific, have been recognized in colony-housed animals with clinical lymphadenopathy and spontaneous seroconversion to RRV (data not shown). While qualitatively similar to changes observed in lymph nodes following experimental inoculation of macaques with rhesus lymphocryptovirus (rhesus Epstein-Barr virus) (29), the rhadinovirus-associated lymphadenopathy was characterized by a greater degree of vascular hyperplasia and hypertrophy and a lesser degree of paracortical lymphocytic hyperplasia. Similar changes have been described during the acute phase of HHV-8 infection of humans (27).
All rhadinovirus-seronegative animals developed a strong antibody response that persisted throughout the experimental period. Virus was recovered from the peripheral blood, and viral DNA could be detected in lymph node, skin, oral mucosa and PBMC by PCR as early as 2 weeks postinoculation. The high rate of seropositivity noted in the NERPRC colony indicates that the virus is readily transmitted among macaques, and our ability to demonstrate viral sequences by PCR in oral mucosa for up to 12 weeks postinoculation suggests that oral secretions may play a role in viral transmission. Cross-species transmission of the viruses was demonstrated and resulted in no observable difference in the clinical manifestations of infection. However, the most severe follicular hyperplasia was observed in monkeys inoculated with PRV.
The absence of a febrile reaction, clinical lymphodenopathy, and follicular hyperplasia in animals 190-96 and 195-96 suggests that prior exposure to RRV may have attenuated any infection following PRV inoculation into these RRV-positive animals. The spike in anti-PRV antibody levels following inoculation suggests that there may have been a take of the PRV. The failure to detect PRV 19545 sequences in the virus recovered during the weeks following inoculation indicates only that the newly inoculated virus did not overcome the indigenous virus to become the predominant species; this result is not surprising. More sensitive tests using tagged viruses will be needed to investigate the issue of superinfection more rigorously.
All experimentally infected, immunologically normal animals remained healthy and have survived for greater than 437 days. These findings are consistent with our observations of NERPRC colony animals that indicate common spontaneous infection of animals greater than 1 year of age and absence of obvious disease sequelae. Retroperitoneal fibromatosis, a neoplastic proliferation of spindle cells resembling KS of humans, has previously been associated by PCR analysis with a macaque gammaherpesvirus related to RRV (4, 32). We found no evidence for the ability of RRV or PRV isolates H26-95 and H19545 to induce this condition. Whether other closely related viruses or cofactors may play an etiologic role in the development of this or other neoplasms remains to be determined.
Coinoculation of animals with SIVmac251 and macaque-derived rhadinovirus resulted in an attenuated antibody response to both agents and shorter mean survival compared to SIVmac251-inoculated controls (155 days versus 560 days; P < 0.019). SIV infection of CD4+ T lymphocytes in conjunction with RRV infection may thus result in an impaired humoral immune response to both viruses and adverse clinical outcome. The mechanism responsible for producing this effect is unknown but may relate to direct infection of CD20-positive B cells by the macaque rhadinovirus or to immune activation resulting from the rhadinovirus infection.
Experimental infection of macaques was associated with a clinical lymphadenopathy characterized initially by paracortical hyperplasia and vascular hypertrophy/hyperplasia that subsequently was replaced by marked follicular hyperplasia. In the most severe cases, this follicular hyperplasia obliterated medullary sinuses and completely effaced the normal lymph node architecture. Similar changes have been found in HIV-negative human patients with histologic features of angioimmunoblastic lymphadenopathy and reactive lymphadenopathy in which HHV-8 sequences could be detected by PCR (27). Our findings suggest that following infection these morphologic alterations may represent a continuum that occurs temporally and to differing degrees. B-cell proliferation is a feature common to MCD and angioimmunoblastic lymphadenopathy and may result from expression of the virus-encoded cytokine IL-6 (vIL-6). vIL-6 has been shown to have proliferative activity on myeloma cells, and its expression has been demonstrated in HIV-negative MCD (6, 31). The presence of a vIL-6 homolog in RRV suggests a similar mechanism may operate during histogenesis of the rhesus rhadinovirus-associated lymphadenopathy.
An arteriopathy has been found in 19 of 85 monkeys examined retrospectively following death from SIV-induced immunodeficiency (12). This arteriopathy has never been seen outside the setting of SIV-induced immunodeficiency in our macaque monkey colony, and a viral etiology has been suspected. A possible etiological role for RRV in this lesion is intriguing because of fundamental similarity of the vascular endothelial proliferation in this lesion and in HHV-8-associated KS. The lesion is also similar to a large vessel arteritis induced in mice by the murine gammaherpesvirus 68 (37). The occurrence of this arteriopathy in three of four monkeys sequentially infected with SIV and RRV is consistent with a possible etiologic role. More detailed studies, including comparison of the occurrence of lesions in RRV-positive versus RRV-negative animals that die with SIV and the demonstration of RRV sequences in the lesions by in situ hybridization, will be needed to draw meaningful associations of RRV with this vascular proliferative syndrome.
Our results have not demonstrated any clear evidence for prolonged or terminal disease induced by experimental infection with RRV or PRV despite variation of a number of investigational parameters. These include same- versus cross-species infection, prior antibody status, and SIV coinfection. While pathologic conditions such as arteriopathy and lymphoproliferative disease could still be caused, at least in part, by RRV, they do not appear to be consistently present even in SIV-plus-RRV/PRV coinfected animals. Nonetheless, the experimental system described here still should prove valuable for modeling KSHV infection. Since RRV and PRV can be grown lytically in cell culture, it should be possible to construct a variety of gene knockouts, gene substitutions, and point mutations within the viral genome. These can be studied not only for their effects on viral replication and B-cell persistence in cell culture but also for their effects on the acute replication phase in monkeys, ability to persist, levels of persistence, and sites of localization.
We thank Jeff Lifson for the SIV plasma RNA analysis, Daniel Silva, Allan McPhee, and Dong-Ling Xia for technical support, and Joanne Newton for manuscript preparation.
This work was supported by PHS grants AI 38131, RR07000, AI 42845, and RR00168.