A Recombinant Rhesus Monkey Rhadinovirus Deleted of Glycoprotein L Establishes Persistent Infection of Rhesus Macaques and Elicits Conventional T Cell Responses

Kaposi’s sarcoma-associated herpesvirus (KSHV) is associated with a substantial disease burden in sub-Saharan Africa, often in the context of human immunodeficiency virus (HIV) infection. The related rhesus monkey rhadinovirus (RRV) has shown potential as a vector to immunize monkeys with antigens from simian immunodeficiency virus (SIV), the macaque model for HIV. KSHV and RRV engage cellular receptors from the Eph family via the viral gH/gL glycoprotein complex. We have now generated a recombinant RRV that expresses the SIV Gag antigen and does not express gL. This recombinant RRV was infectious by the intravenous route, established persistent infection in the B cell compartment, and elicited strong immune responses to the SIV Gag antigen. These results argue against a role for gL and Eph family receptors in B cell infection by RRV in vivo and have implications for the development of a live-attenuated KSHV vaccine or vaccine vector.

ABSTRACT

A replication-competent, recombinant strain of rhesus monkey rhadinovirus (RRV) expressing the Gag protein of SIVmac239 was constructed in the context of a glycoprotein L (gL) deletion mutation. Deletion of gL detargets the virus from Eph family receptors. The ability of this gL-minus Gag recombinant RRV to infect, persist, and elicit immune responses was evaluated after intravenous inoculation of two Mamu-A*01⁺ RRV-naive rhesus monkeys. Both monkeys responded with an anti-RRV antibody response, and quantitation of RRV DNA in peripheral blood mononuclear cells (PBMC) by real-time PCR revealed levels similar to those in monkeys infected with recombinant gL⁺ RRV. Comparison of RRV DNA levels in sorted CD3⁺ versus CD20⁺ versus CD14⁺ PBMC subpopulations indicated infection of the CD20⁺ subpopulation by the gL-minus RRV. This contrasts with results obtained with transformed B cell lines in vitro, in which deletion of gL resulted in markedly reduced infectivity. Over a period of 20 weeks, Gag-specific CD8⁺ T cell responses were documented by major histocompatibility complex class I (MHC-I) tetramer staining. Vaccine-induced CD8⁺ T cell responses, which were predominantly directed against the Mamu-A*01-restricted Gag_181-189CM9 epitope, could be inhibited by blockade of MHC-I presentation. Our results indicate that gL and the interaction with Eph family receptors are dispensable for the colonization of the B cell compartment following high-dose infection by the intravenous route, which suggests the existence of alternative receptors. Further, gL-minus RRV elicits cellular immune responses that are predominantly canonical in nature.

IMPORTANCE Kaposi’s sarcoma-associated herpesvirus (KSHV) is associated with a substantial disease burden in sub-Saharan Africa, often in the context of human immunodeficiency virus (HIV) infection. The related rhesus monkey rhadinovirus (RRV) has shown potential as a vector to immunize monkeys with antigens from simian immunodeficiency virus (SIV), the macaque model for HIV. KSHV and RRV engage cellular receptors from the Eph family via the viral gH/gL glycoprotein complex. We have now generated a recombinant RRV that expresses the SIV Gag antigen and does not express gL. This recombinant RRV was infectious by the intravenous route, established persistent infection in the B cell compartment, and elicited strong immune responses to the SIV Gag antigen. These results argue against a role for gL and Eph family receptors in B cell infection by RRV in vivo and have implications for the development of a live-attenuated KSHV vaccine or vaccine vector.

INTRODUCTION

The rhesus monkey rhadinovirus (RRV) is a gamma-2 herpesvirus, also called rhadinovirus, that is found with high prevalence in both captive and wild rhesus macaques (1, 2). There are two major sequence groups or clades of RRV (3) for which two cloned isolates, RRV 26-95 (4) and RRV 17577 (5), exist. RRV is closely related to the only human rhadinovirus, Kaposi’s sarcoma-associated herpesvirus (KSHV), and serves as a model for selected aspects of KSHV’s biology (6). Experimental infection with RRV strain 17577 induced lymphoma in rhesus macaques that were coinfected with simian immunodeficiency virus (SIV) (7, 8), while RRV strain 26-95 did not cause immediate disease in immunocompetent or immunosuppressed monkeys (9). Nevertheless, RRV or its macaque homologs in Macaca nemestrina or Macaca fascicularis have been detected in various types of non-Hodgkin lymphoma in SIV-infected or simian retrovirus type 2-infected monkeys (10, 11). This is similar to the case with KSHV in humans, which is associated with primary effusion lymphoma (PEL) and multicentric Castleman’s disease in human immunodeficiency virus (HIV)-positive individuals (12). RRV, unlike KSHV with Kaposi’s sarcoma, is not consistently associated with solid malignancies, although RRV has been detected in solid malignancies as well (8). RRV establishes latency in the B cell compartment, which is another point of similarity to KSHV (13). With regard to cellular entry and receptor tropism, both KSHV and RRV can interact with Eph family receptor tyrosine kinases to infect target cells (14). While KSHV binds EphA2 with significantly higher avidity than other Ephs, RRV binds several A- and B-type Ephs with comparable avidities. RRV can also effectively utilize different A- and B-type Ephs to infect cells (14). Our recent results demonstrate that KSHV uses EphA7 to infect the BJAB B cell line (15), and the findings of other groups suggest that KSHV can use EphA4 and EphA5 as well in some settings (16, 17). The interaction with Eph family receptors is mediated by the gH/gL complex of KSHV or RRV, and the presence of gL in the gH/gL complex is essential for the receptor interaction to occur (18).

RRV has been used as a vaccine vector in the SIV macaque model of HIV infection (19–22). Recombinant RRV (rRRV) encoding SIV antigens alone or in combination with other vaccine strategies was able to elicit potent and durable T cell responses (19–21) and antibody responses to vectored SIV antigens (23). These vaccine-induced SIV-specific immune responses afforded significant reductions in viral replication in SIVmac239-infected vaccinees (19–21) and even protection against SIVmac239 acquisition in a recent study (24).

As we have recently demonstrated, a gL null mutant of RRV is unable to interact with Ephs (18). We found the resulting reduction in infectivity to be more pronounced on endothelial cells than on primary rhesus monkey fibroblasts, the cells that are usually used to grow RRV to high titers. While a clear reduction in infectivity was observed upon ablation of gL expression, gH was still incorporated into the virion particle to wild-type (wt) levels, and the virus remained replication competent in vitro (18), which is similar to results with murine gammaherpesvirus 68 (MHV-68) and pseudorabies virus (25, 26) but clearly different from results obtained with herpes simplex virus (27). These findings prompted us to ask several questions: (i) whether a gL null mutant of RRV would still be able to establish persistent infection in macaques, (ii) whether it could still serve as a potent vaccine vector, and (iii) whether any attenuation with regard to viral loads would result from the ablation of gL expression. Since recombinant forms of the 68-1 strain of rhesus cytomegalovirus (RhCMV), which is fibroblast adapted and lacks genes associated with formation of the pentameric gH/gL/UL128/UL130/131 complex, have been shown to induce noncanonical CD8⁺ T cell responses of unusual major histocompatibility complex (MHC) restriction (28), we also examined the MHC restriction of the SIV-specific CD8⁺ T cell response induced by the rRRVgagΔgL vector described herein.

RESULTS

We constructed an rRRVgagΔgL virus, similar to the RRV yellow fluorescence protein (YFP) ΔgL virus used in our previous study (18). Instead of the YFP expression cassette, an expression-optimized gag (gag c.o.) expression cassette (29) was inserted into the RRV genome (Fig. 1A). Before inoculating the animals, the virus stock was tested for the deletion in the orf47 locus that codes for gL (Fig. 1B), and expression of the gag transgene in infected primary rhesus monkey fibroblasts was verified by Western blotting (Fig. 1C). Two adult, RRV-negative, Mamu-A*01-positive rhesus macaques (Table 1) were infected with rRRVgagΔgL by the intravenous route. Sera were collected and tested for reactivity to RRV antigens by enzyme-linked immunosorbent assay (ELISA) at weeks 0, 2, 4, 12, and 16 to assess the infection status of the two monkeys. Both animals seroconverted following the rRRVgagΔgL inoculation. Anti-RRV antibody responses were readily detected and increased over the 16 weeks of follow-up (Fig. 2A, red lines). The levels of anti-RRV antibodies appeared similar to levels observed in previous experiments using wild-type recombinant RRV expressing the gag, env, or a nef-tat-ref fusion gene of SIV (Tri-mix) (Fig. 2A, black lines). We also analyzed the serological responses to gH and gL (Fig. 2B). To this end, we tested Western blot membrane strips that were displaying recombinant gH and gL for reactivity with sera from the two rRRVgagΔgL-infected monkeys (Fig. 2B, 3rd and 4th lanes from the left) and with 6 control sera from the Tri-mix-infected monkeys (Fig. 2B, 5th to 10th lanes from the left). While the preinfection sera (Fig. 2B, first and second lanes from the left) were not reactive toward either protein, the rRRVgagΔgL-infected monkeys mounted a strong antibody response to gH, but not to gL, whereas the monkeys from the Tri-mix control group mounted vigorous antibody responses to both gH and gL.

FIG 1.

Recombinant rRRVgagΔgL. (A) Position of the gag transgene cassette in rRRV. The expression cassette for codon-optimized SIVmac239 Gag was inserted between the left terminal repeats (TR) and the first open reading frame (ORF), R1, of RRV. The expression cassette consists of the cytomegalovirus promoter (pCMV), the SIVmac239 gag transgene, the bovine growth hormone (BGH) poly(A) signal, and a kanamycin resistance gene cassette under the control of a prokaryotic promoter. In addition, 128 bp of the RRV gL-encoding gene (orf47) was deleted. (B) Confirmation of gL deletion. Viral DNA from either wild-type RRV or rRRVgagΔgL was isolated from infected rhesus fibroblast (RF) culture supernatants and a PCR was performed using primers flanking orf47 of RRV, yielding amplicons of 647 bp for the wt sequence and 519 bp for the gL deletion. (C) Confirmation of expression of SIVmac239 Gag. RF lysates were prepared. Proteins were separated by SDS-polyacrylamide gel electrophoresis, and the separated proteins were transferred to a PVDF membrane and subsequently tested with an antibody specific for SIVmac239 Gag (KK64). “Ctrl.” represents cell lysate from wt RRV-infected RF, and “rRRVgagΔgL” represents cell lysate from RF infected with rRRVgagΔgL.

TABLE 1.

Characteristics of research animals

Monkey	Sex	Age (yrs)	Wt (kg)	Inoculum	Mamu-A*01	RRV status
						DOIa	PIb
r13018	Female	3.9	5.13	rRRVgagΔgL	+	−	+
r13041	Female	3.8	6.07	rRRVgagΔgL	+	−	+

^aDOI, day of inoculation.

^bPI, postinoculation.

FIG 2.

Serological responses to infection with rRRVgagΔgL. (A) Seroconversion as established by RRV ELISA. Sera of two rRRVgagΔgL-inoculated rhesus macaques and of six monkeys that had received a mixture of three recombinant RRVs (including gag, env, or nef-tat-ref; Tri-mix) were tested for the presence of anti-RRV responses by using ELISA plates coated with purified RRV lysate. Samples were collected at the indicated time points. (B) Antibody response to gH and gL. Sera from experimentally RRV-infected macaques (with either rRRVgagΔgL or Tri-mix) were collected 16 weeks postinfection and tested for reactivity toward recombinant RRV gH or gL protein by Western blot analysis. For the two rRRVgagΔgL-infected monkeys, samples from the time point of infection were included as a negative control (first two lanes).

In order to further establish whether the virus was infectious for rhesus monkeys and could establish persistent infection, peripheral blood mononuclear cells (PBMC) were isolated and RRV DNA copy number was determined by quantitative PCR (qPCR) at early and late time points postinoculation (Table 2). Both animals were negative at the 2-week time point. The animals had detectable RRV genomic DNA in PBMC by weeks 6 (r13018) and 16 (r13041; no result available for week 6), respectively (Table 2). RRV genomic DNA copy number in PBMC from the two rRRVgagΔgL-infected monkeys at week 20 postinfection was, on average, not lower than in PBMC from Tri-mix-infected animals (Fig. 3A). After 64 weeks (69 weeks for the CD14⁺ cells), the animals were analyzed again, and PBMC were sorted into CD3⁺, CD14⁺, and CD20⁺ populations. These populations were analyzed individually, as were PBMC subpopulations from naturally infected or RRV-naive monkeys as controls (Fig. 3B). Animal r13041 displayed an approximately 10-fold-higher viral load in the CD20⁺ population than did animal r13018. Overall, viral loads were in the same range as in two naturally infected macaques with detectable viral loads (r04170 and r10043), while two naturally infected (seropositive) macaques had undetectable viral loads in the CD20⁺ population (r07054 and r10008). These results were unexpected, as we had previously found infection of immortalized B cell lines to strongly depend on interaction with receptors from the Eph family (14), which is disrupted by deletion of gL in rRRVgagΔgL (18). To confirm that ablation of gL expression would lead to reduced infectivity on different B cell lines, we performed in vitro infection experiments with RRV-YFP ΔgL virus, which expresses a YFP reporter gene. Indeed, this virus infected immortalized B cell lines with considerably lower efficiency than RRV-YFP wt gL virus (Fig. 4). While approximately 5-fold more RRV-YFP ΔgL input virus (normalized to genome copy number) was needed to achieve RRV-YFP gL wt infection levels on rhesus monkey fibroblasts, this defect was considerably more pronounced on the B cell lines that were tested, ranging from >25-fold-reduced infectivity on MMB1845 cells to >125-fold-reduced infectivity of RRV-YFP ΔgL on Raji and BJAB cells compared to that with RRV-YFP wt gL.

TABLE 2.

rRRVgagΔgL copies in total PBMC

Monkey	rRRVgagΔgL copiesa /million PBMC at wk postinoculation
	0	2	6	16	20
r13018	0	0	987	984	693
r13041	0	0	NAb	2,032	2,242

^arRRVgagΔgL copies were measured via quantitative real-time PCR.

^bNA, not available.

FIG 3.

RRV genome copy number in PBMC and RRV presence in PBMC subsets of experimentally infected rhesus macaques. (A) RRV genome copy numbers in PBMC of experimentally infected rhesus macaques. PBMC of the two rRRVgagΔgL-infected rhesus macaques and of five monkeys that had received a mixture of three rRRVs including an RRVgag recombinant were obtained at week 20 postinoculation. DNA was isolated and subsequently tested for the presence of recombinant RRV genomes via quantitative real-time PCR. Error bars represent the SDs of triplicate reactions (for animal r11004, only duplicate values were available). Values marked with a star are potentially below the linear range of the PCR. (B) Presence of RRV DNA in PBMC subpopulations of experimentally or naturally infected rhesus macaques. PBMC of two rRRVgagΔgL-inoculated rhesus macaques were obtained at week 64 (for CD3⁺ and CD20⁺ subpopulations) and at week 69 (for the CD14⁺ subpopulation) postinoculation. In addition, PBMC of two RRV-naive and four animals naturally infected with RRV were isolated. PBMC were sorted into CD3⁺, CD14⁺, and CD20⁺ subsets, and DNA was isolated and subsequently tested for the presence of RRV genomes by quantitative real-time PCR.

FIG 4.

In vitro infection of primary rhesus monkey fibroblasts and three CD20-positive suspension cell lines with RRV-YFP wt gL and RRV-YFP ΔgL. Target cells of human (BJAB and Raji) or rhesus macaque (RF and MMB1845) origin were infected with RRV-YFP wt gL or RRV-YFP ΔgL at the indicated virus concentrations. YFP expression as an indicator of infection was measured by flow cytometry.

Next, we characterized the Gag-specific CD8⁺ T cell response induced by the rRRVgagΔgL vector. The fact that both r13018 and r13041 were Mamu-A*01⁺ enabled us to track their vaccine-induced Gag-specific CD8⁺ T cell responses in PBMC by Mamu-A*01 tetramer staining. We selected two Mamu-A*01-restricted SIV Gag epitopes for this analysis: Gag_181-189CM9 (immunodominant) (30) and Gag_254-262QI9 (subdominant) (31). Neither r13018 nor r13041 developed detectable CD8⁺ T cell responses against Gag_254-262QI9 (Fig. 5A), consistent with the subdominant rank of this epitope. Conversely, both animals mounted Gag_181-189CM9-specific CD8⁺ T cells, which peaked at week 4 postvaccination at frequencies of 11.4% (r13018) and 5.5% (r13041) of the entire peripheral CD8⁺ T cell compartment (Fig. 5B). These responses contracted in the ensuing weeks and eventually plateaued at frequencies of 1.02% (r13018) and 2.44% (r13041) at week 20 postvaccination (Fig. 5B and C). At this time, we carried out a memory phenotype analysis and found that the fraction of Mamu-A*01/Gag_181-189CM9 tetramer⁺ CD8⁺ T cells displaying a fully differentiated effector memory T cell (T_EM2) signature (CD28⁻ CCR7⁻) ranged from 89.5% in r13018 to 94.1% in r13041 (Fig. 5C). Moreover, 68 to 96% of tetramer⁺ CD8⁺ T cells expressed the cytotoxicity-associated molecule granzyme B (Fig. 5C). The durability of these vaccine-induced CD8⁺ T cell responses and their T_EM-biased phenotype are consistent with the ability of the rRRVgagΔgL vector to persist in vivo.

FIG 5.

Kinetics and memory phenotype of rRRVgagΔgL-induced CD8⁺ T cell responses in peripheral blood. Fluorescently labeled Mamu-A*01 tetramers folded with peptides corresponding to the Gag_254-262QI9 (A) or Gag_181-189CM9 (B) epitope were used to monitor the ontogeny of vaccine-induced CD8⁺ T cell responses in two Mamu-A*01⁺ rRRVgagΔgL-inoculated monkeys (r13018 and r13041). (C) Memory phenotype of vaccine-induced Gag_181-189CM9-specific CD8⁺ T cells. The frequencies of Mamu-A*01/Gag_181-189CM9 tetramer⁺ CD8⁺ T cells at week 20 post-rRRVgagΔgL inoculation are shown in the left plots for r13018 (top row) and r13041 (bottom row). The middle plots show the delineation of memory subsets within the tetramer⁺ gate based on the differential expression of CD28 and CCR7. Three subsets were identified: central memory (T_CM; CD28⁺ CCR7⁺), transitional memory (T_EM1; CD28⁺ CCR7^–), and effector memory (T_EM2; CD28^– CCR7^–). The histograms in the right plots show the levels of granzyme B expressed by tetramer⁺ CD8⁺ T cells (black lines). The gray lines correspond to tetramer⁺ CD8⁺ T cells stained with an isotype-matched control MAb.

We also carried out intracellular cytokine staining (ICS) assays at weeks 3 to 6 postvaccination to determine the breadth and MHC restriction of vaccine-induced Gag-specific CD8⁺ T cells in r13018 and r13041. PBMC from both animals produced gamma interferon (IFN-γ) and tumor necrosis factor alpha (TNF-α) in response to stimulation with a pool of peptides spanning the first half of the Gag polyprotein (Gag 1; amino acids [aa] 1 to 263) but not the second (Gag 2; aa 248 to 510) (Fig. 6A). These CD8⁺ T cell responses against Gag 1 were then deconvoluted using a set of 7 peptide pools covering 50-aa stretches of Gag 1 (Fig. 6A). This analysis revealed that vaccine-induced Gag-specific CD8⁺ T cells in both animals were narrowly focused on Gag E (aa 161 to 211), which contains the sequence corresponding to the Gag_181-189CM9 epitope (Fig. 6A and B).

FIG 6.

rRRVgagΔgL vaccination elicits narrow Gag-specific CD8⁺ T cell responses that are sensitive to MHC-I blockade. (A) ICS gating strategy for deconvoluting Gag-specific CD8⁺ T cell responses induced by rRRVgagΔgL vaccination. Representative results from r13041 are shown. PBMC were left unstimulated (top left plot) or stimulated with pools of 15-mer peptides overlapping by 11 aa covering amino acids 1 to 263 (Gag 1, middle plot, top row) or aa 248 to 510 (Gag 2, right plot, top row). Reactivity to the Gag 1 pool was deconvoluted using seven pools of peptides spanning 50-aa segments of Gag 1 (Gag A to G) and found to be concentrated on Gag E (aa 161 to 211). Plots are gated on live CD3⁺ CD8⁺ lymphocytes. (B) Summary of the ICS results in panel A. The graph shows the background-subtracted percentages of CD8⁺ T cells from r13018 and r13041 producing both IFN-γ and TNF-α against each of the Gag A to G peptide pools. (C and E) In order to identify the MHC-restricting element for the CD8⁺ T cell reactivity against the Gag E peptide pool in r13018 (C and D) and r13041 (E and F), the ICS assays were set up in the presence of 10 μg/ml of the MHC-I-blocking MAb W6/32 or 1.0 μg/ml of the MHC-II-blocking MAb G46-6. Because addition of these MAbs resulted in high levels of nonspecific TNF-α production (data not shown), the proportion of background-subtracted CD69⁺ IFN-γ⁺ CD8⁺ T cells was used as the readout for these assays. Response frequencies in parentheses were less than 2-fold higher than the corresponding value in the “No stimulus” test and therefore were considered negative. (D and F) Summary of MHC blockade ICS results for r13018 (D) and r13041 (F).

Hansen et al. have reported that rhesus macaques vaccinated with RhCMV 68-1 vectors expressing SIV antigens develop unusually broad SIV-specific CD8⁺ T cell responses (28). These unconventional CD8⁺ T cell responses were restricted by MHC class II (MHC-II) molecules and the nonclassical MHC-Ib molecule MHC-E and did not conform to normal immunodominance hierarchies (28, 32). Indeed, RhCMV 68-1 vaccinees did not mount CD8⁺ T cell responses against canonical SIV epitopes restricted by MHC-Ia molecules, like the Mamu-A*01-restricted Gag_181-189CM9 epitope (28). RhCMV 68-1 lacks genes whose encoded proteins associate with gH/gL to form the pentameric CMV virion surface complex (33) and, similar to our rRRVgagΔgL vector, displays altered cell tropism (28). Therefore, we also evaluated whether the Gag-specific CD8⁺ T cell response elicited by the rRRVgagΔgL vector displayed unusual MHC restriction. To do that, we set up ICS assays in the presence of the same anti-MHC-I monoclonal antibody (MAb) (W6/32) or anti-MHC-II MAb (G46-6) used by Hansen and colleagues in their studies (28). In our assays, addition of these MAbs increased background production of TNF-α and, to a lesser extent, IFN-γ. This effect was more pronounced for W6/32 than G46-6. Because of this nonspecific cytokine production, we used the frequency of recently activated (CD69⁺) IFN-γ⁺ CD8⁺ T cells as the readout for the MHC blockade ICS assays (Fig. 6C to F). Since Gag E was the only pool recognized by both vaccinees, we focused our detailed analysis on this antigen. We found that in the presence of the MHC-I-blocking MAb W6/32, CD8⁺ T cell recognition of Gag E was completely (r13018 [Fig. 6C and D]) or nearly (r13041 [Fig. 6E and F]) reduced to background levels, indicating that MHC-I molecules were the main restricting element for the epitopes present in the Gag E pool. Of note, since W6/32 also blocks MHC-E molecules, this assay did not assess whether Gag E peptides other than Gag_181-189CM9 were being presented by Mamu-E. Surprisingly, incubation with the MHC-II-blocking MAb G46-6 also decreased CD8⁺ T cell recognition of Gag E, although to a lesser extent than that observed in the presence of W6/32 (Fig. 6C to F). However, since CD8⁺ T cell recognition of Gag_181-189CM9 was reduced by similar degrees (Fig. 6D and F), this effect is unlikely due to MHC-II presentation of Gag E peptides. We conclude from these assays that rRRVgagΔgL vaccination induced predominantly canonical, MHC-I-restricted Gag-specific CD8⁺ T cell responses.

DISCUSSION

Here we have described the experimental infection of two rhesus monkeys with recombinant rRRVgagΔgL by the intravenous route. Persistent infection was established as confirmed by persistent antibody responses, persistence of cellular immune responses, and detection of viral genomic DNA over a period of 1 year. RRV genome copy numbers in PBMC from rRRVgagΔgL-infected animals were similar to those in other experimentally infected animals that received RRV wt-based vaccine vectors, and CD20⁺ PBMC subpopulations from both rRRVgagΔgL-infected animals were positive for RRV DNA more than 1 year postinoculation. Therefore, the first important finding of our study is that gL, and by extension also the interaction with receptors from the Eph family, is dispensable for establishment of persistent infection by RRV following high-dose intravenous administration. The phenotype of rRRVgagΔgL infection in monkeys resembles that of an MHV-68 gL deletion mutant in mice (25), and it is interesting that a KSHV gH null mutant was reported to still infect B cells in vitro (34). It remains to be determined whether rRRVgagΔgL can enter cells by a similar gH-independent mechanism. It should be noted, however, that the high-dose intravenous inoculation of 10⁹ viral genomes, which bypassed the mucosal barrier, does not reflect natural infection. Whether gL and/or interaction with receptors from the Eph family plays a major role during mucosal transmission will need to be evaluated in separate studies.

All animals that were inoculated with RRV wt-based strains mounted vigorous antibody responses toward gH and gL, whereas the rRRVgagΔgL-infected animals mounted an antibody response to gH but not to gL (Fig. 2B). This demonstrates seroconversion established by ELISA measurements and confirms, as expected, the absence of gL presentation following infection with our rRRVgagΔgL. Antibody responses to truncated gL variants, which could theoretically arise through mutations that restore the remaining open reading frame (ORF), should also be detected by our Western blot assay with recombinant full-length gL (Fig. 2B). The strong antibody responses to gH following infection by rRRVgagΔgL are consistent with our previous data showing incorporation of gH into RRV-YFP ΔgL virions to wt levels (18).

For KSHV, orf47, orf46, and orf45 are found on a tricistronic transcript (35); orf46 and orf47 are also on bi- and monocistronic transcripts (36, 37). The existence of additional transcripts that originate from the KSHV orf47 promoter region has been described (37), namely, two transcripts coding for an ORF47/ORF45 fusion protein, which has been reported to be important for KSHV reactivation. Whether analogous transcripts and proteins are expressed during RRV infection was not the focus of our study, but our results suggest that if they exist, they are not essential for infection in vitro and in vivo, as their expression should be abrogated by the deletion in rRRVgagΔgL. Differences between RRV and KSHV are likely, as the RRV orf47 transcript exhibits extremely suboptimal codon usage and requires ORF57 or codon optimization for expression of gL (38), which is not the case for KSHV. Expression of ORF46 and ORF45 was not analyzed in our study.

The lack of an attenuated phenotype with regard to PBMC viral genome copy number (Fig. 3A) and in particular with regard to infection of the CD20⁺ subpopulation (Fig. 3B) is surprising in that we observed a strong reduction in infectivity of RRV-YFP ΔgL on CD20⁺ B cell lines that are widely used as models for B cells (39) (Fig. 4). We also had previously shown that infection of three immortalized B cell lines was strongly dependent on the interaction with receptors from the Eph family (14). While our high-dose intravenous inoculation regimen may have enabled the virus to directly establish persistence in, e.g., a B cell precursor compartment, the discrepancy between in vitro and in vivo observations is nevertheless striking. Three main possible causes for these discrepancies, or a combination thereof, come to mind: (i) receptor expression may change after immortalization, (ii) immortalization by, e.g., lymphocryptovirus infection may select for certain cells that are preferentially infected by lymphocryptovirus, and, finally, (iii) certain developmental stages or B cell subsets may be more prone to give rise to lymphoma, the source of B cell lines such as BJAB and Raji (40). Our results demonstrate that results with cell lines in vitro have only limited predictive value for the situation in vivo, clearly making a case for the need for animal studies to confirm in vitro findings.

We observed CD8⁺ T cell responses against SIV Gag antigen in rRRVgagΔgL-infected animals (Fig. 5A and B). In fact, these CD8⁺ T cell responses were similar in magnitude to those from a previous vaccine study with wt RRV-vectored Gag antigen (19). In contrast to the unusually broad and unconventionally MHC-restricted CD8⁺ T cell responses observed in RhCMV 68-1-vaccinated rhesus macaques (28), Gag-specific CD8⁺ T cells induced by the rRRVgagΔgL vector were narrowly focused on the canonical, Mamu-A*01-restricted Gag_181-189CM9 epitope. In line with this finding, CD8⁺ T cell recognition of Gag peptides was largely inhibited by MHC-I blockade, further supporting the conclusion that deletion of gL alone is insufficient to render RRV capable of inducing the unconventional CD8⁺ T cell responses seen in RhCMV 68-1 vaccinees. In support of this view, a recent analysis of anti-CMV CD8⁺ T cell responses elicited by two fibroblast-adapted human CMV vaccines found no evidence for MHC-II or HLA-E restriction (41). Rather, these vaccine-induced CMV-specific CD8⁺ T cell responses were focused on a few immunodominant epitopes restricted by classical HLA-I molecules. Unexpectedly, incubation with the MHC-II-blocking MAb G46-6 partially reduced CD8⁺ T cell recognition of Gag E in our experiments (Fig. 6C to F). Whether this really indicates some degree of MHC-II presentation of Gag E peptides or represents an experimental artifact will have to be determined in future experiments.

Overall, our study demonstrates that gL and the interaction with Eph receptors are not essential for the establishment of persistent RRV infection after high-dose inoculation by the intravenous route. This clearly indicates the existence of additional receptors for RRV, which remain to be characterized and may provide important clues also for the biology of KSHV and potentially also the more distantly related Epstein-Barr virus. Further, our results demonstrate that an RRV gL-null mutant can still elicit potent immune responses to a vectored antigen. In the light of recent findings that show an association between variants in the gene coding for the human EphA2 KSHV receptor and KSHV infection as well as development of Kaposi’s sarcoma (42), our findings with an Eph-detargeted rhadinovirus mutant successfully vectoring an SIV antigen may have implications for the development of a live-attenuated KSHV vaccine or the use of live-attenuated KSHV as a vaccine vector.

MATERIALS AND METHODS

Cells.

293T, BJAB, and Raji cell lines were obtained from the Leibniz-Institut DSMZ-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH. MMB1845, a clonal cell line established from rhesus macaque PBMC that were immortalized by infection with herpesvirus papio, were a kind gift from Ulrike Sauermann. All suspension cells were cultured in RPMI 1640 medium (Thermo Fisher Scientific) supplemented with 10% fetal calf serum (FCS) and 50 μg/ml of gentamicin (R10 medium). Primary rhesus monkey fibroblasts were a kind gift of Rüdiger Behr or from the New England Primate Research Center (NEPRC). Rhesus fibroblasts (RF; NEPRC) were cultured and maintained in Dulbecco’s modified Eagle medium (DMEM; Life Technologies) supplemented with 20% fetal bovine serum (FBS; Life Technologies) and Primocin (InvivoGen) or in DMEM with 10% FBS and 50 μg/ml of gentamicin (those RF used for infection experiments in Fig. 4).

Recombinant viruses.

rRRVgagΔgL was constructed based on the RRV BAC17 (43). The deletion in the gL locus was introduced into RRV BAC17 as described previously (18) using the method of Tischer et al. (44). A total of 128 bp was deleted, which introduced a frameshift after amino acid 26 and resulted in a stop codon after amino acid 37, leaving only six amino acids of the original gL sequence after the putative signal peptide cleavage site.

The kanamycin resistance-coding KanS cassette from pEPKanS (44), which also harbors an I-SceI site, was inserted into an expression-optimized Gag expression cassette between the Gag coding sequence and the bovine growth hormone (BGH) poly(A) site of the Gag expression cassette, flanked by a 51-bp repeat to allow removal of the cassette by recombination. To this end, the Gag expression vector was amplified in two fragments using two primer pairs (Gag co reverse [CTACTGGTCTCCTCCAAAGAGAG] and AmpQas [TGAGGCACCTATCTCAGCGATCTG]; Gag co STOP-50 forward [TGACAGAGGATTTGCTGCACC] and AmpQs [CAGATCGCTGAGATAGGTGCCTCA]), and the KanS cassette from pEPKanS was amplified using primers KanS + gag STOP-25 forward (TTCTCTCTTTGGAGGAGACCAGTAGGGATGACGACGATAAGTAGG) and KanS + gagSTOP-50to-25 reverse (TTGAGGTGCAGCAAATCCTCTGTCACAACCAATTAACCAATTCTGATTAG). These fragments were joined by Gibson assembly to yield pCMV_c.o._gag_KanS.

This construct was inserted into RRV BAC17 ΔgL using the protocol by Tischer et al. (44) and the following primers: Gag co in forward (GTTACAAGCACCTGCTTAACTTGCTTTGGCTCTGTGCGGTTTTGTTGCTAGGTACTTAGTTATTAATAGTAATCAATTACGG) and Gag co in reverse (TCGGATGCTGTGGAGCGTATGCAGGCGCCATGCGATGAAATACTATGTGTTACTACTCTCATTACTTCCTCACATGTTGGAG). Despite several attempts, we were not able to obtain clones with intact genome after the second recombination to remove the prokaryotic kanamycin resistance cassette located in the 3′ untranslated region (UTR) of the gag gene by means of recombination between the two 51-bp repeats that were generated in pCMV_c.o._gag_KanS. Therefore, we decided to go forward with a clone that still had the cassette in the 3′ UTR, where it should not affect transgene expression. After transfection of BAC DNA into rhesus fibroblasts, virus was recovered, and infection with this virus yielded strong expression of the Gag protein. This initial stock was then sequenced by Illumina-based sequencing before one more passage on RF to generate the inoculation stock as previously described (19).

Subsequently, recombinant RRV-infected RF culture supernatants were harvested and spun down twice at 2,000 × g for 5 min to remove any cell debris, and resulting virus titers in the culture supernatants of infected rhesus fibroblasts were measured via quantitative real-time PCR using an RRV latency-associated nuclear antigen (LANA)-specific primer set. Quantification was performed using the TaqMan fast virus 1-step master mix (Life Technologies) in a real-time PCR thermocycler (Applied Biosystems 7500 fast and 7500 real-time PCR system). The following primers and reporter were used: forward primer, ACCGCCTGTTGCGTGTTA; reverse primer, CAATCGCCAACGCCTCAA; and reporter, 6-carboxyfluorescein (FAM)-CAGGCCCCATCCCC.

The RRV Tri-Mix viruses are described elsewhere (19, 23).

In vitro expression of SIVmac239 Gag.

A total of 250,000 RF were seeded into each well of a six-well plate. The following day, cells were infected with 50 μl of a stock of rRRVgagΔgL (1 × 10⁹ genome copies/ml) or wt RRV (4). Cells were kept in culture 6 days before harvesting of the cells or cell culture supernatants when cells were exhibiting advanced cytopathic effect. For Western blot analysis of gag expression, cells were harvested, resuspended, and lysed with an NP-40-based lysis buffer containing protease inhibitors (Roche). After centrifugation, supernatants were transferred into new tubes and protein levels were evaluated using a bicinchoninic acid protein assay kit (Pierce) which allowed for protein normalization in the samples. The lysates were mixed with equal volumes of 2× SDS Laemmli sample buffer (Sigma-Aldrich). Subsequently, samples were incubated at 97°C for 5 min and separated by SDS-PAGE under reducing conditions. The proteins from the gels were transferred to a polyvinylidene difluoride (PVDF) membrane and subjected to immunoblotting. Membranes were blocked with 1× phosphate-buffered saline (PBS) containing 5% skim milk and incubated with a 1:5,000 dilution of a mouse-derived anti-SIVmac239-gag antibody (KK64; NIH AIDS Reagent Program) overnight. Three wash steps with PBS-T (phosphate-buffered saline containing 0.05% Tween 20) were followed by incubation with goat anti-mouse-horseradish peroxidase (HRP)-conjugated secondary antibody (SouthernBiotech) in 1× PBS containing 5% skim milk. Three final wash steps were performed, and specific signals were detected with a LAS4000 minisystem (GE Healthcare Systems) using SuperSignal West Pico chemiluminescent substrate (Pierce).

Confirmation of gL deletion in rRRVgagΔgL by PCR.

Viral DNA from either wild-type RRV or rRRVgagΔgL was isolated from infected RF culture supernatants via a NucleoSpin tissue kit (Macherey-Nagel) according to the manufacturer’s protocol. Subsequently, a PCR was performed using primers flanking orf47 of RRV, which codes for gL: forward primer, AAATTTAAGCCATGAGTCGCTAA, and reverse primer, ACAATATACATGCAGTTTACAACGTA. The resulting amplicons were analyzed via agarose (1%) electrophoresis and visualized with ethidium bromide.

Recombinant RRV inoculation.

rRRVgagΔgL stock was diluted to a total of 1 × 10⁹ genome copies/ml in PBS. Each monkey received 1 ml of the diluted virus stock intravenously. The three recombinant RRVs of the Tri-mix were inoculated at 3 × 10⁸ genome copies/ml each in PBS.

RRV ELISA.

Antibodies reactive to RRV were detected by ELISA as described previously (1). Sera were tested at 1:200 dilution.

Western blotting-based serology with recombinant gH/gL.

For detection of antibodies reactive to gH and gL by Western blotting, recombinant soluble RRV gH/gL complex was expressed in 293T cells by transfection of 2.5 μg of gH and 7.5 μg of gL expression plasmid per 10-cm dish using the polyethylenimine (PEI) method (45). The recombinant protein was purified from the cell culture supernatant using StrepTactinXT beads (IBA) as per the manufacturer’s protocol. Expression constructs were based on codon-optimized versions of gH and gL (3). The soluble ectodomain of gH (aa 21 to 697) expressed behind a heterologous murine IgG kappa signal peptide was amplified by PCR (forward primer pAX22 BamHI for, AAGGATCCTACGAATATAATGAGGAGAAGGTG; reverse primer AX47 NotI StrepSTOP ApaI rev, ATGGGCCCTACTTCTCGAACTGGGGGTGGCTCCAGCGGCCGCCCAGGGCATGCCTC) using the previously described gH-Fc expression plasmid (14) as a template. Thereby, a C-terminal StrepTag (amino acid sequence WSHPQFEK) followed by a stop codon was added and the resulting open reading frame was reinserted into the pSecTag2/HygroA backbone via BamHI and ApaI. The gL construct with a C-terminal Flag tag (amino acid sequence DYKDDDK) was described previously (14). A total of 0.3 μg of purified recombinant soluble gH/gL complex was denatured in reducing Laemmli sample buffer (Sigma-Aldrich) and separated on 8 to 16% polyacrylamide gradient 1-mm Tris-glycine minigels with Tris-glycine SDS running buffer (Thermo Fisher Scientific) and then transferred to 0.22-μm PVDF membranes (Bio-Rad) by semidry blotting in a Trans-Blot (Bio-Rad) apparatus using 2× Tris-glycine transfer buffer (Thermo Fisher Scientific) with 10% (vol/vol) ethanol and 150 mA per minigel for 1 h. The membranes were blocked in membrane blocking solution (Thermo Fisher Scientific) and cut into strips. Individual strips were then incubated with the respective sera diluted 1:100 in membrane blocking solution for 5 h at room temperature. After three washes with PBS-T, the membrane strips were incubated with anti-rhesus IgG HRP-conjugated secondary antibody (SouthernBiotech) diluted 1:2,000 in membrane blocking solution for 2 h at room temperature. After three washes in PBS-T, the strips were imaged using SuperSignal West Pico-ECL (Thermo Fisher Scientific) and an ImageQuant LAS 4000 miniluminescent image analyzer.

RRV genome measurement in cells.

DNA from PBMC was isolated with a NucleoSpin tissue kit (Macherey-Nagel) in accordance with the manufacturer’s protocol. Between 4 and 8 million PBMC were used for isolation of PBMC DNA, between 0.2 and 1.9 million cells were used for isolation from sorted PBMC subsets, and samples were eluted into 100 μl. A total of 10 μl was used per PCR. The isolated DNAs were analyzed via a quantitative real-time PCR using a probe specific for the ORF 59/60 junctional region of RRV (46) and RRV cosmid DNA (4) as standards. Quantification was performed using the TaqMan fast virus 1-step master mix (Life Technologies) in a real-time PCR thermocycler (Applied Biosystems 7500 fast and 7500 real-time PCR system). The following primers and reporter were used: forward primer, TCTGAATATGTCACATCCGTTCATA; reverse primer, GGCCCGGAAAATGAGTAACA; and reporter, FAM-TGATCTGTAGTCCCCATGTGTCC. Cycling conditions were 1 min at 95°C for initial denaturation and then 45 cycles of 95°C for 15 s and 60°C for 60 s. The limit of detection was below 21 copies per reaction, and the linear range of the PCR ended between 210 and 21 copies per reaction.

Cell sorting and quantitative real-time PCR.

Using a BD FACSJazz cell sorter (BD Biosciences), Ficoll-isolated PBMC were separated into CD3⁺ and CD20⁺ populations at week 64 postinfection (anti-CD3 Alexa Fluor 488 [clone SP34-2] and anti-CD20 allophycocyanin [APC]-H7 [clone L27], both from BD Biosciences). Additionally, Ficoll-isolated PBMC were fixed with 2% paraformaldehyde (PFA; diluted from 32% stock in PBS; Electron Microscopy Systems) and separated into CD3⁺, CD20⁺, and CD14⁺ cells (anti-CD14 peridinin chlorophyll protein [PerCP]-Cy5.5, clone MΦP9; BD Biosciences) at week 69 postinfection to allow for enrichment of CD14⁺ cells, which otherwise was technically not feasible. Subsequently, DNA from the isolated populations was isolated using a NucleoSpin tissue kit (Macherey-Nagel) according to the manufacturer’s protocol. The isolated DNAs were analyzed via a quantitative real-time PCR using the same probe, instrument, and parameters as described for the RRV genome measurements in PBMC.

Infection of suspension cells with RRV-YFP reporter virus.

RRV production and quantitative real-time-PCR-based viral genome copy number analysis were performed as described before (18). For infection assays, rhesus fibroblasts (plated 1 day prior to infection at 25,000/cm²) or suspension cells at 200,000 cells/ml (BJAB, Raji, and MMB1845) were infected with the desired amounts of virus, normalized to genome copies. At 24 h (RF) or 48 h (BJAB, Raji, and MMB1845) postinfection, cells were harvested, washed with PBS, and fixed in 2% formaldehyde (Carl Roth) in PBS. A minimum of 10,000 cells was analyzed per sample for YFP expression on an LSRII flow cytometer (BD Biosciences). Data were analyzed using Flowing software (version 2.5).

MHC-I tetramer staining of T cells.

Phycoerythrin-labeled Mamu-A*01/Gag_181-189CM9 (MBL) and allophycocyanin-labeled Mamu-A*01/Gag_254-262QI9 (NIH Tetramer Core Facility) tetramers were used to quantify Gag-specific CD8⁺ T cells in PMBC according to a recently published protocol (47). For the time course analysis of vaccine-induced SIV epitope-specific CD8⁺ T cell responses (Fig. 5A and B), approximately 800,000 PBMC were incubated in R10 medium with titrated amounts of each tetramer at room temperature for 45 min. The cells were then stained with fluorochrome-labeled MAbs directed against the surface molecules CD3 (clone SP34-2), CD8α (clone RPA-T8), CD14 (clone M5E2), CD16 (clone 3G8), and CD20 (clone 2H7) for 25 min at room temperature. This surface-staining MAb cocktail also included an amine-reactive dye (ARD; LIVE/DEAD fixable aqua dead cell stain; Life Technologies). The cells were then washed with wash buffer (Dulbecco’s PBS with 0.1% bovine serum albumin [BSA] and 0.45 g/liter of NaN₃) and fixed with 1× BD FACS lysing solution (BD Biosciences) for 10 min at room temperature. The cells were washed one more time before they were acquired on a BD LSR II cytometer equipped with a 50-mW 405-nm violet laser, a 100-mW 488-nm blue laser, and a 30-mW 635-nm red laser using FACSDIVA (version 6) software.

To determine the memory phenotype of vaccine-induced Gag-specific CD8⁺ T cells (Fig. 5C), 2.4 × 10⁶ PBMC were incubated with the Mamu-A*01/Gag_181-189CM9 tetramer at room temperature for 45 min. The cells were then stained with the same MAbs specific for CD8α, CD14, CD16, and CD20 as described above, plus MAbs against CD28 (clone 28.2) and CCR7 (clone 150503). ARD aqua was also included in this surface-staining MAb cocktail. After a 25-min incubation at room temperature, cells were treated with 1× BD FACS lysing solution (BD Biosciences) for 10 min and subsequently washed with wash buffer (Dulbecco’s PBS with 0.1% BSA and 0.45 g/liter of NaN₃). Cells were then permeabilized by treatment with permeabilization buffer (1× BD FACS lysing solution 2 [Becton Dickinson] and 0.05% Tween 20 [Sigma-Aldrich]) for 10 min, washed, and stained with MAbs against CD3 (clone SP34-2) and granzyme B (clone GB12). After a 30-min incubation in the dark at room temperature, cells were washed and stored at 4°C until acquisition. Samples were acquired in the same flow cytometer as described above.

FlowJo versions 9.9 and 10.5.3 (FlowJo, LLC) were used to analyze the data. Doublets were excluded by gating out events with disproportionally high width first in a plot of forward-scatter (FS) area versus FS width and then in a plot of side-scatter (SC) area versus SC width. Next, a time gate was created that included only those events that were recorded within the 15th and 85th percentiles of acquisition time. The resulting cells were then gated on “dump channel” (CD14, CD16, CD20, and ARD)-negative, CD3⁺ cells. Because MHC-I tetramer binding to the T cell receptor (TCR) can lead to TCR internalization, the CD3 gate at this stage included cells that expressed intermediate levels of CD3. Next, the lymphocyte population was delineated based on its FS and SS properties and subsequent analyses were conducted within CD8⁺ cells. After outlining MHC-I tetramer⁺ cells, the memory phenotyping analysis was performed within this gate. Rhesus macaque memory T cells can be classified into three subsets based on the differential expression of CD28 and CCR7: central memory (T_CM; CD28⁺ CCR7⁺), transitional memory (T_EM1; CD28⁺ CCR7⁻), and fully differentiated effector memory (T_EM2; CD28⁻ CCR7⁻). Granzyme B expression was also evaluated within MHC-I tetramer⁺ cells. Cells stained with fluorochrome-labeled MAbs of the same isotypes as the anti-granzyme B, anti-CD28, and anti-CCR7 MAbs guided the identification of the memory subsets and granzyme B-expressing cells within the MHC-I tetramer⁺ population. Based on this gating strategy, all tetramer frequencies mentioned in this report correspond to percentages of live CD14⁻ CD16⁻ CD20⁻ CD3⁺ CD8⁺ tetramer⁺ lymphocytes.

ICS assay.

The antigen stimuli for the intracellular cytokine staining (ICS) assays consisted of pools of 15-mer peptides overlapping by 11 amino acids spanning various regions of the SIVmac239 Gag polyprotein or a peptide corresponding to the Mamu-A*01-restricted Gag_181-189CM9 epitope, as indicated in Fig. 6. The final assay concentration of each peptide was 1.0 μM. PBMC obtained from the research animals were stimulated with the appropriate peptides in R10 medium (RPMI 1640 supplemented with GlutaMax [Life Technologies], 10% fetal bovine serum [VWR], and 1% antibiotic-antimycotic [VWR]) containing costimulatory MAbs against CD28 and CD49d for 9 h at 37°C in an incubator with a 5.0% CO₂ concentration. Brefeldin A (Biolegend, Inc.) was added to all tubes 1 h into the incubation period to inhibit protein transport. Surface molecules of cells were stained with MAbs against the following molecules: CD14 (clone M5E2), CD16 (clone 3G8), CD20 (clone 2H7), CD4 (clone OKT4; Biolegend, Inc.), and CD8 (clone RPA-T8; Biolegend, Inc.). Amine-reactive dye (LIVE/DEAD fixable aqua dead cell stain; Life Technologies) was also added to this MAb cocktail. After a 25-min incubation in the presence of this surface-staining MAb master mix, the cells were washed with wash buffer (Dulbecco’s PBS with 0.1% BSA and 0.45 g/liter of NaN₃), fixed with a 2% paraformaldehyde solution for 20 min at 4°C, and then washed again with wash buffer. At this point, the cells were permeabilized by resuspension in permeabilization buffer (1× BD FACS lysing solution 2 [Becton Dickinson] and 0.05% Tween 20 [Sigma-Aldrich]) for 10 min and subsequently washed with wash buffer. Cells were then incubated with MAbs against CD3 (clone SP34-2), IFN-γ (clone 4S.B3), TNF-α (clone Mab11), and CD69 (clone FN50) for 1 h in the dark at room temperature. After this incubation was completed, the cells were washed and subsequently stored at 4°C until acquisition. The data were analyzed by gating first on live CD14^– CD16^– CD20^– CD3⁺ lymphocytes and then on cells expressing either CD4 or CD8 but not both markers. Functional analyses were conducted within these two compartments based on the expression of IFN-γ and/or TNF-α (Fig. 6A and B) or the coexpression of CD69 and IFN-γ (Fig. 6C to F). Leukocyte activation cocktail (BD Pharmingen)-stimulated cells stained with fluorochrome-labeled MAbs of the same isotypes as those against IFN-γ, TNF-α, and CD69 guided the identification of positive populations. We used two criteria to determine if responses were positive. First, the frequency of gated events had to be at least 2-fold higher than their corresponding values in background-subtracted negative-control tests. Second, the gates for each response had to contain ≥10 events.

For the MHC blockade ICS assays, we used the same protocol as described above, except that PBMC were incubated at room temperature with the anti-MHC-I MAb W6/32 (10 μg/ml) or anti-HLA-DR (MHC-II) MAb G46-6 (1.0 μg/ml) for 1 h prior to the addition of the SIV peptides.

Research animals.

Two rhesus macaques of Indian origin (Macaca mulatta) were used in this study. The research animals were housed at the Wisconsin National Primate Research Center (WNPRC) and cared for in accordance with the guidelines of the Weatherall Report (48) under a protocol approved by the University of Wisconsin Graduate School Animal Care and Use Committee. The animals were under anesthesia during recombinant RRV inoculations and subsequent blood draws.

Accession number(s).

The sequences obtained in this study have been deposited in GenBank under accession numbers MN488837 (rRRVgagΔgL), MN488838 (RRV-YFP ΔgL), and MN488839 (RRV-YFP).

ADDENDUM IN PROOF

After this article was accepted for publication, we noticed that RRV-YFP∆gL (used for the in vitro experiments in Fig. 4), which is negative for gL expression in virus particles, harbors an additional single nucleotide deletion 13 bp before the intended 128-bp deletion (GenBank accession number MN488838, as mentioned above).