|Home | About | Journals | Submit | Contact Us | Français|
The immunogenicity and protective capacity of replication-defective herpes simplex virus (HSV) vector-based vaccines were examined in rhesus macaques. Three macaques were inoculated with recombinant HSV vectors expressing Gag, Env, and a Tat-Rev-Nef fusion protein of simian immunodeficiency virus (SIV). Three other macaques were primed with recombinant DNA vectors expressing Gag, Env, and a Pol-Tat-Nef-Vif fusion protein prior to boosting with the HSV vectors. Robust anti-Gag and anti-Env cellular responses were detected in all six macaques. Following intravenous challenge with wild-type, cloned SIV239, peak and 12 week plasma viremia levels were significantly lower in vaccinated compared to control macaques. Plasma SIV RNA in vaccinated macaques was inversely correlated with anti-Rev ELISPOT responses on the day of challenge (P-Value <0.05), anti-Tat ELISPOT responses at two weeks post challenge (P-Value <0.05), and peak neutralizing antibody titers pre-challenge (P-Value 0.06). These findings support continued study of recombinant herpesviruses as a vaccine approach for AIDS.
Development of an effective vaccine against HIV/AIDS faces a number of severe difficulties (Desrosiers, 2004; Garber and Feinberg, 2003). In the absence of blind luck in stumbling upon a successful vaccine, one important obstacle is the lack of knowledge regarding the characteristics of a protective immune response that we should be looking for from a vaccine. Once initial infection is established, HIV has an uncanny ability to replicate continuously in the face of readily-measurable immune responses to the virus. HIV has evolved a number of specific immune evasion strategies that allow this continuous, unrelenting viral replication (Desrosiers, 1999). These characteristics raise concern whether long-lived memory cells that may be present months or years after immunization can respond in time to provide sufficient blunting of viral replication during the initial days and weeks following exposure.
A defining feature of all herpesviruses is that they persist for the life of the infected individual. Replication-competent and replication-defective herpesviruses have been used experimentally as vectors for gene therapy (Burton, Fink, and Glorioso, 2005) and for vaccine purposes. Vaccine applications of herpesvirus vectors have included both AIDS (Murphy et al., 2000) and non AIDS (Hoshino et al., 2005; Morrison and Knipe, 1994; Spector et al., 1998) settings. Herpesvirus vectors may differ fundamentally from other viral vectors that have been tried experimentally, for example poxviruses (Amara et al., 2001; Dale et al., 2004; Horton et al., 2002; Pal et al., 2002), adenoviruses (Patterson et al., 2003; Santra et al., 2005; Shiver et al., 2002), and rhabdoviruses (Ramsburg et al., 2004; Rose et al., 2001), regarding the nature and persistence of the immune responses that are generated. Since herpesviruses may periodically or chronically reactivate, the numbers of memory cells and the numbers of actively-responding cells at any one time after the last exposure may be significantly higher than that which is observed with viruses that do not have the potential to persist. Consequently, we have been interested in the ability of recombinant herpesviruses to provide vaccine protection against AIDS in an experimental setting.
Here we describe the immune responses and protective capacity of replication-defective herpes simplex virus (HSV) vectors with and without prior priming by recombinant DNA.
To improve upon our earlier studies that utilized replication-competent and replication-defective HSV vectors (Murphy et al., 2000), we constructed new experimental HSV-SIV vaccines based on HSV-1 d106, a replication-defective vector with multiple HSV genes deleted (Samaniego, Neiderhiser, and DeLuca, 1998). The lack of expression of ICP4 and ICP27 prevents the shut off of host cell transcription and translation and allows prolonged expression of the transgene. The lack of host shut-off and early and late HSV protein synthesis prevents MHC downregulation and ICP47-associated blockage of the TAP peptide transporter. Separate HSV-1 d106 vectors were engineered to express gag, env and a rev-tat-nef fusion protein. High levels of SIV gag and env expression were achieved by use of codon-modified cassettes that were modified to achieve maximal levels of expression (Nasioulas et al., 1994; Schneider et al., 1997; Schwartz et al., 1992; Schwartz, Felber, and Pavlakis, 1992). The nef reading frame in the tat-rev-nef fusion protein was modified to render it incapable of MHC class I downregulation (Swigut et al., 2004). This vector expresses significantly higher levels of SIV protein for more prolonged periods than we have been able to achieve previously in the absence of HSV-induced cytopathic effect (Watanabe et al, manuscript in preparation).
We also included a group of monkeys in these studies that were primed by DNA immunization prior to recombinant HSV-1 d106 immunization. The recombinant DNA constructs that were used have been described recently (Rosati et al., 2005). Gag and Env expression plasmids were fused to MCP-3 to facilitate humoral immune responses. Gag, Env, Pol, Nef, Tat and Vif genes were fused with the β-catenin peptide in order to target the viral proteins to the proteasomal degradation pathway and enhance MHC class I antigen presentation. SIV239 sequences were used for all immunogen constructions.
One group of three Indian-origin rhesus macaques (Mm 136-90, 450-91, and 349-91) received the mixture of HSV/SIV recombinant d106 strains at weeks 0, 4, 12, and 20 (Fig. 1). Another group of three Indian-origin rhesus macaques (Mm 205-87, 297-91 and 240-91) were immunized with the mixture of DNAs at weeks 0 and 4 and the mixture of HSV/SIV recombinant d106 strains at weeks 12 and 20 (Fig. 1). Two macaques from each group were Mamu-A*01-positive (Mm 136-90, 450-91, 205-87 and 297-91). One macaque (Mm 450-91) was Mamu-B*17-positive. All of the rhesus monkeys were obtained from the herpes B-free, specific pathogen free breeding colony at the NEPRC (Desrosiers, 1997) and were negative for antibodies to HSV at the time of enrollment. All six vaccinated monkeys were challenged intravenously with WT SIV239 at week 26 (Fig. 1).
Antibodies to HSV were measured by ELISA using plates covered with whole lysed HSV virions. All six monkeys were negative for anti-HSV antibodies at the time of the first immunization (Fig. 2a). The three monkeys that received recombinant HSV-SIV d106 at week 0 responded with anti-HSV antibodies within weeks of this first immunization. Anti-HSV antibody levels appeared to be boosted in this group of animals upon each subsequent receipt of recombinant d106 inoculations (Fig. 2a). The initial inoculations of recombinant DNA, in contrast, did not result in anti-HSV reactivity in this assay, as expected. However, inoculation of this group of monkeys with the HSV-SIV recombinant d106 strains at week 12 did promptly result in readily-detectable levels of anti-HSV antibodies (Fig. 2a). These persisted in all six monkeys through challenge.
Antibodies to SIV assayed again by whole virus ELISA using methodologies that have been extensively described and validated previously (Daniel et al., 1988; Wyand et al., 1996) were not so readily detected (Fig. 2b). However, anti-SIV antibodies were measured in all six monkeys with this assay by two weeks after the final vaccine boost (Fig. 2b).
We next measured the ability of plasma to neutralize SIV infectivity. None of the plasma was able to neutralize infectivity of SIV239 detectably. This is not surprising because SIV239 is difficult to neutralize and even monkeys infected experimentally with SIV239 make neutralizing antibodies of low levels when measured against this strain (Johnson et al., 2003). Neutralizing activity was detected, however, when the laboratory-passaged, neutralization-sensitive SIV251 strain (SIV251-lab) was used (Fig. 3a and Table 1). Neutralizing activity against SIV251-lab was detected at multiple time points in five of the six monkeys prior to challenge (Table 1). The highest titers of neutralizing activity that were achieved against SIV251-lab prior to challenge were 1:160. This contrasted with titers of 1:-3200 to 1:5120 in pooled sera obtained from unvaccinated rhesus macaques infected with SIV239 and used as a positive control in the neutralization assay. Thus, the vaccine regimens did induce neutralizing activity but the titers were low and they did not detectably neutralize cloned SIV239 whose sequences were used in the immunogen constructions.
Cellular immune responses to SIV were measured by the IFN-γ ELISPOT and ICS assays using overlapping peptides of individual SIV239 proteins as immunogens. Additionally, Mamu-A*01/ Gag CM9 tetramer reactive CD8+ T lymphocytes were measured in the Mamu-A*01-positive rhesus macaques.
IFN-γ ELISPOT responses to the SIV proteins expressed by both DNA and HSV-1 d106 vectors (Gag, Env, Nef and Tat) were detected in all six vaccinated monkeys (Figure 4a). From the time of the HSV booster inoculation at week 12, the magnitude of ELISPOT responses to shared SIV immunogens remained significantly higher in the DNA-DNA-HSV-HSV immunized macaques compared to the HSV-HSV-HSV-HSV immunized macaques (Table 2). At its peak, Gag-specific responses as high as 1,442 spot forming cells (SFC)/106 PBMC and Env-specific responses as high as 3,208 SFC/106 PBMC were observed in the DNA-DNA-HSV-HSV group of rhesus macaques (Figure 4a). Such levels are comparable to or higher than those achieved after wild-type SIV infection of unvaccinated rhesus macaques. In contrast to Gag and Env, positive ELISPOT responses to the SIV regulatory proteins Nef, Tat and Rev were of considerably lower magnitude (Figure 4a).
Mamu-A*01/ Gag CM9 tetramer reactive CD8+ T lymphocytes were detected in all four Mamu-A*01-positive vaccinated rhesus macaques and reached a peak level of 2.3% of CD8+ T lymphocytes in the peripheral blood of Mm 205-87 (Figure 4b). There was a trend for increasing affinity of Gag CM9-specific CD8+ T lymphocytes with each HSV booster inoculation in the DNA primed macaques as evidenced by an eight-fold or greater decrease in the peptide concentration required for T cell activation (Figure 4c). One week after the last vaccine inoculation (week 21), tetramer-positive CD8+ T lymphocytes with both effector memory (CD28–, CD62L–, CCR7–) and central memory (CD28+, CCR7+, CD62L+) phenotype were observed (Figure 5 and Table 3). At this time-point, the gut-homing molecule α4β7 was expressed on 13–50% of tetramer-positive CD8+ T lymphocytes in the peripheral blood (Table 3). The frequency of Gag CM9 tetramer-reactive CD8+ T lymphocytes correlated positively with Gag ELISPOT responses suggesting that a significant portion of the Gag-specific T lymphocyte response in Mamu-A*01-positive rhesus macaques consisted of CD8+ T lymphocytes responding to the Gag CM9 peptide (Figure 6a). ELISPOT responses to the whole SIV239 Gag p55 antigen were also detected and correlated positively with Gag peptides (Figure 6b). Because whole proteins predominantly stimulate CD4+ T lymphocytes in vitro, the positive p55-specific ELISPOT response suggested that SIV-specific CD4+ T lymphocytes were also induced in vaccinated macaques. Induction of both CD4+ and CD8+ T lymphocyte responses to SIV was confirmed by the ICS assay (Figure 7 and Table 4). CD4+ T lymphocytes responding to SIV Gag and/or Env were detected in four vaccinated macaques (Table 4).
All six vaccinated macaques were challenged intravenously with 10 animal infectious doses of SIV239 at week 26, six weeks after the last vaccine dose (Figure 1). All vaccinated monkeys became infected with SIV (Figure 8). In the vaccinated monkeys, the median viral load at peak two weeks after SIV challenge was 2.4x106 (Range 4.0x105 to 2.7x107) plasma SIV RNA copies/ml; at 12 weeks it was 8.7x103 copies/ml (Range 2.1x102 to 1.3x106 copies/ml); and at 24 weeks it was 1.6x103 copies/ml (Range 7.0x101 to 4.5x106 copies/ml). One HSV-HSV-HSV-HSV vaccinated macaque (Mm 136-90) died suddenly three weeks after SIV inoculation and was likely a rapid progressor. The set-point plasma SIV RNA at 24 weeks post SIV challenge in three of the five surviving vaccinated macaques was <5000 copies/ml (Figure 8).
To examine whether there was a vaccine effect on control of viremia, we compared plasma viral load in the vaccinated macaques with historical controls consisting of 31 unvaccinated SIV-infected Indian rhesus macaques, 14 from NEPRC and 17 from the Wisconsin Primate Research Center. All had been inoculated with the same stock of cloned SIV239 used for SIV challenge in the vaccinated macaques. While there is animal to animal variation, there have not been changes in the behavior of this virus stock in groups of unvaccinated control Indian-origin rhesus macaques over the course of several years. The median plasma SIV RNA level at peak (week 2) was 0.9 log lower, and at 12 weeks it was 1.7 log lower in the vaccinated as compared to the historical control rhesus macaques, and these differences were statistically significant by the nonparametric Mann Whitney-U test (Figure 9). At later time-points (20–24 weeks), the difference in plasma viral load between vaccinated and historical control macaques failed to reach statistical significance (Figure 9).
Four of six vaccinated macaques (67%) and 23 of 31 controls (74%) were Mamu-A*01-positive (Table 5). Because the presence of Mamu-A*01 may be an independent predictor of SIV viral control (Mothe et al., 2003; Muhl et al., 2002; Pal et al., 2002; Zhang et al., 2002), we compared viral loads in vaccinated and control macaques stratified on the basis on their Mamu-A*01 status (Table 5). Vaccinated Mamu-A*01-positive macaques continued to demonstrate a significant reduction in peak and 10–14 wk time-point viral loads as compared to control Mamu-A*01-positive macaques (Table 5). Although there was a trend for lower viral loads in vaccinated Mamu-A*01-positive macaques at 17–28 wks, this difference did not reach statistical significance (Table 5). It is to be noted that one vaccinated macaque (Mm 450-91) contained two MHC Class I alleles, Mamu-A*01 and Mamu-B*17, a combination that is particularly associated with low SIV viral loads (O'Connor et al., 2003). On exclusion of the Mamu-B*17 positive monkeys (one vaccinated and five controls), the difference between vaccinees and controls continued to be statistically significant (P-value <0.05) for peak plasma viremia, but did not reach statistical significance (P-value 0.07) for the 10–14 week timepoint. Given the small number of vaccinated macaques, we cannot definitively evaluate vaccine efficacy independent of the attenuating effect of MHC Class I alleles in the current study.
Neutralizing activity in plasma against SIV251-lab, presumably neutralizing antibodies, increased abruptly in all six animals following challenge, consistent with the viral load measurements demonstrating that all six vaccinated monkeys had become infected with SIV (Fig. 4b and Table 1). Neutralizing activity was dramatically increased in five of the six vaccinated monkeys by 2 weeks post challenge consistent with an anamnestic antibody response. Neutralizing activity was not detected against SIV239 even by 8 weeks post SIV infection. Dramatic increases in Mamu-A*01 / Gag CM9 tetramer reactive CD8+ T lymphocytes and ELISPOT responses to Gag, Env, and Nef were observed in all vaccinated macaques in the first two to four weeks after SIV challenge (Figure 10a and data not shown).
Since three of the six vaccinated macaques had set-point plasma SIV RNA levels of <5000 copies/ml, we investigated whether there was a correlation between vaccination-induced anti-SIV immune responses and the level of SIV viremia post challenge. Peak, day of challenge and two weeks post challenge anti-SIV immune responses were examined for inverse correlation with the area under curve (AUC) plasma SIV RNA values between 0 and 24 weeks of SIV infection (Figure 10b–c and data not shown). Although Gag- and Env-specific cellular immune responses dominated the anti-SIV IFN-γ ELISPOT response prior to challenge, their magnitude at peak response, or at day of challenge, or at two weeks post challenge, did not correlate with control of viral load (Figure 10b–c and data not shown). Instead, cellular immune responses targeting accessory SIV proteins appeared to be important. The magnitude of the Rev-specific IFN-γ ELISPOT response on the day of challenge (Figure 10b), as well as the magnitude of the Tat-specific IFN-γ ELISPOT response two weeks post challenge (Figure 10c) showed a significant inverse correlation with viral load (Rho –0.86 and –0.87 respectively, P-Values <0.05, Spearman Rank Correlation test). Neutralizing antibody titers on the day of challenge and two weeks post challenge did not correlate significantly with viral load (Figure 10b–c). However, the maximal titer of neutralizing antibodies detected during the vaccine phase showed a strong trend towards an inverse correlation with AUC plasma SIV RNA (Rho –0.83, P-Value 0.06, Spearman Rank Correlation test; data not shown).
One vaccinated macaque, Mm 136-90, died 23 days after intravenous challenge with SIV239 with no premonitory signs of illness. Gross and microscopic abnormalities were confined to the brain. Cut sections of the formalin-fixed brain showed extensive tan discoloration with multiple red-brown foci confined to the white matter of the left occipital lobe (Figure 11A). On light microscopic examination, multiple and randomly distributed ring hemorrhages and inflammatory cell infiltrates dotted the otherwise pale white matter; extension into the inner layers of the gray matter was infrequent (Figure 11B). On higher magnification, the white matter was rarefied with angiocentric hemorrhage and/or inflammatory cell infiltrates (Figure 11C). At high magnification, vascular walls were necrotic and surrounded by degenerate inflammatory cells primarily composed of neutrophils (Figure 11D). Sections of the left occipital lobe immunostained with anti-myelin basic protein showed perivascular loss of myelin (Figure 11E). Rare glial nodules were observed within the white matter of the cerebellum and the corpus striatum (Figure 11F). The histopathologic findings are most consistent with a necrotizing and suppurative vasculitis culminating in secondary ischemic necrosis of the white matter with perivascular ring hemorrhages and angiocentric loss of myelin in the foci of inflammation. Light microsopic examination of multiple sections of the left occipital lobe stained with H&E and several special stains did not detect viral inclusion bodies, protozoa, bacteria, acid fast bacilli, or fungal organisms. A battery of immunohistochemical markers for herpes simplex virus, cytomegalovirus, adenovirus, SV40, and SIV did not establish a viral etiology. Furthermore, electron microscopic examination of several sites from the brain lesion did not identify an infectious etiology, including herpes simplex. Co-cultivation of trigeminal and sacral ganglia from Mm 136-90 onto Vero cells and the E-11 complementing cell line did not recover any virus.
The ability of herpesviruses to persist and induce durable immune responses in their infected hosts makes them an attractive viral vector for delivery of an AIDS vaccine. We have previously reported on the efficacy of recombinant replication-deficient and replication-competent herpes simplex virus (HSV) vaccine vectors expressing SIV Env and Nef to protect rhesus macaques against a mucosal challenge with pathogenic SIV239 (Murphy et al., 2000). Using a new generation of HSV-SIV recombinants based on the replication-deficient HSV vector, HSV-1 d106 (Samaniego, Neiderhiser, and DeLuca, 1998), we now demonstrate considerable increase in vaccine immunogenicity as compared to the previous HSV vaccine vectors (Murphy et al., 2000).
The previous HSV recombinants induced weak humoral and cellular immune responses against SIV. Binding antibodies were not readily detected by ELISA-based assays and neutralizing activity against lab-adapted SIV251 was not detected (Murphy et al., 2000). Using the traditional CTL assay, weak CTL activity was detected at a single time-point in only two of seven immunized monkeys (Murphy et al., 2000). In contrast, the HSV-1 d106 recombinants used in the current study induced readily detectable anti-SIV antibodies, and a robust cellular immune response in all six vaccinated monkeys. A single inoculation of HSV-1 d106 SIV recombinants expressing Gag, Env, and a Rev-Tat-Nef fusion protein was sufficient to induce anti-SIV T cell responses that were detectable ex vivo by IFNγ ELISPOT assay and by tetramer staining. Vaccination with recombinant HSV-SIV in the absence of DNA priming gave rise to peak anti-SIV T cell responses to the vaccine immunogens ranging between 528 and 818 SFC/106 PBMC, with more than half of this response being directed towards the Gag and Env proteins. These levels are higher or comparable in magnitude to those induced by several other recombinant viral vector SIV vaccines in rhesus macaques including attenuated poxviruses, vaccinia virus, vesicular stomatitis virus, semliki forest virus, and adeno-associated virus, given alone or in combination with DNA priming or as heterologous virus boost regimens (Dale et al., 2004; Doria-Rose et al., 2003; Johnson et al., 2005; Koopman et al., 2004; Pal et al., 2002; Ramsburg et al., 2004; Santra et al., 2002). The magnitude of SIV-specific cellular immune responses in the DNA primed macaques was even higher and in the range of other highly immunogenic DNA prime / viral vector boost SIV vaccination regimens in rhesus macaques consisting of Modified Vaccinia Ankara (MVA) or replication-deficient Adenovirus Ad5 (Amara et al., 2001; Casimiro et al., 2003; Horton et al., 2002; Shiver et al., 2002), as well as with heterologous prime-boost immunization regimen using MVA and replication-defective Adenovirus vectors (Casimiro et al., 2004), and replication-competent Adenovirus based SIV vaccines (Patterson et al., 2003). Both CD4+ and CD8+ T lymphocyte responses to SIV were elicited in the vaccinated macaques. Furthermore, up to 50% of the vaccine-induced tetramer-positive CD8+ T lymphocytes in peripheral blood expressed the gut-homing molecule, α4β7, suggesting that systemic administration of recombinant HSV may be able to elicit mucosal immunity. These data are consistent with the finding of a recombinant NYVAC-SIV vaccine being able to induce anti-SIV CD8+ T cells in the rectal and vaginal mucosa even when administered via the intramuscular route (Stevceva et al., 2002). Since mucosal priming can further accentuate mucosal immunity (Bertley et al., 2004; Egan et al., 2004; Evans et al., 2003), the ability to deliver recombinant HSV vectors via mucosal routes may be an added advantage of this vaccine approach. In all, our finding of robust vaccine-induced cellular immune responses, combined with the potential of HSV vectors to elicit durable immune responses and induce mucosal immunity makes recombinant HSV-based AIDS vaccine worthy of further study.
Relative to the magnitude of SIV-specific cellular immune responses, anti-SIV antibody responses elicited by recombinant HSV were modest. Binding anti-SIV antibodies were elicited in all vaccinated macaques and neutralizing antibodies to the neutralization-sensitive lab-adapted strain of SIVmac251 were detected at several time-points prior to challenge in five of the six vaccinated macaques. A similar range of neutralizing antibody titers to HxB2, a lab isolate of HIV-1, were recently reported in rhesus macaques immunized with DNA and recombinant Adenovirus Ad5 vectors encoding HIV Env proteins (Mascola et al., 2005). The neutralizing antibody titers against lab-adapted SIVmac251 generated in the immunized macaques in our study are 100 to 1000-fold lower than following wild-type SIV infection and at least 10-fold lower than that induced by live attenuated SIV 3 vaccination (Wyand et al., 1999). The low levels of antibody responses are perhaps not surprising. In SIV vaccination approaches used thus far in rhesus macaques, high titer neutralizing antibodies during the vaccination phase have generally only been seen with regimens that incorporate a booster inoculation of envelope protein (Patterson et al., 2004; Peng et al., 2005; Zhao et al., 2003), or AT2-inactivated SIV particles (Willey et al., 2003), and even in these instances neutralizing antibodies to primary SIV isolates are not detected.
It is of interest that the enhanced immunogenicity of the current HSV-1 d106 SIV recombinants as compared to the earlier generation of HSV vectors (Murphy et al., 2000) did not translate to better protection against SIV239 challenge. In the previous study, two of seven vaccinated macaques were solidly protected while one macaque had reduced viral load (Murphy et al., 2000). In this study, three of six vaccinated macaques had reduced viral loads. However both studies differ in important aspects, particularly with regards to the route of SIV239 challenge (intrarectal versus intravenous), and the timing between immunization and challenge (22 weeks versus six weeks after the last vaccination), making direct comparison between the two studies difficult. A discrepancy between vaccine immunogenicity and protective efficacy was apparent on analysis of the six vaccinated macaques in this study. Thus, the magnitude of vaccine-induced SIV-specific cellular immunity to immunodominant antigens (Gag and Env), as measured by the frequency of IFNγ-secreting or tetramer-positive CD8+ T lymphocytes in peripheral blood, did not correlate with post challenge viral load. On the contrary, Mm 205-87, the rhesus macaque with the highest responses during the vaccine phase showed no evidence of viral control post challenge. Interestingly, immune parameters that significantly correlated with lower viral loads post challenge included the presence of Rev-specific T cell responses on the day of challenge, the rapidity of emergence of the anamnestic Tat-specific T cell response post challenge, and the magnitude of neutralizing antibody activity prior to challenge. These findings suggest that a successful AIDS vaccine may need to prime for T cell responses to accessory HIV proteins in addition to inducing neutralizing antibodies.
The vaccinated monkey that died 3 weeks after SIV challenge, likely a rapid progressor, had a unilateral lesion in the left occipital lobe. The microscopic findings were characterized by a necrotizing and suppurative vasculitis with secondary necrosis of the white matter. In addition, rare glial nodules were present in the cerebral white matter and brainstem but were unrelated to the occipital lobe lesion. Glial nodules are nonspecific lesions that usually result from viral infection (An et al., 2002; Prineas, MacDonald, and Franklin, 2002); in this case the nodules were negative by IHC for viruses, including SIV, and of unknown significance. Differential diagnoses for the occipital lobe lesion included a septic lesion of hematogenous origin (Kennedy, 2004; Moore, 2000), HSV encephalitis (Kennedy, 2004; Kennedy, 2005), and acute hemorrhagic leukoencephalitis (AHL). An extensive histopathologic workup failed to identify an infectious etiology, including HSV, as the cause of the vasculitis. Encephalitis related to reversion of the replication-deficient HSV vaccine to replication-competent HSV was considered unlikely for several reasons. In vivo replication of HSV-1 d106 vectors could hypothetically occur in two scenarios; one, in the presence of host cellular proteins able to compensate for the deleted immediate early HSV genes; and two, by recombination with herpes B virus naturally infecting rhesus macaques. Rhesus macaques in this study were seronegative for herpes B. Although a seronegative status for herpes B does not preclude herpes B infection in macaques (Ward and Hilliard, 1994), the absence of herpes virions by electron microscopy makes the diagnosis of HSV encephalitis highly unlikely. In the absence of an infectious etiology, the morphologic changes of perivascular hemorrhages, perivascular myelinolysis and neutrophilic infiltration in the white matter are consistent with a diagnosis of AHL as the cause of the encephalitis (An et al., 2002; Bennetto and Scolding, 2004; Hart and Earle, 1975; Prineas, MacDonald, and Franklin, 2002). AHL is a rare, fulminant, and often fatal disorder of unknown etiology that is thought to be immune-mediated or a sequela of the acute perivascular inflammation (Bennetto and Scolding, 2004; Hart and Earle, 1975; Prineas, MacDonald, and Franklin, 2002). A febrile or nonspecific respiratory infection often precedes onset of both AHL and postinfectious encephalitis with a temporal relationship of 1 to 4 weeks between the precipitating cause and onset of neurologic disease. While there was a temporal relationship (3 weeks) between experimental SIV infection and death, it cannot be ascertained that SIV was the precipitating cause. Finally, the morphology of the lesion and the long interval (8 weeks) between the last vaccination and death argues against a diagnosis of postvaccinal encephalomyelitis.
In this study, historical controls were used for assessment of vaccine efficacy. Although not ideal, the use of historical controls allowed us to include a large number of animals that would not otherwise have been possible with concurrent controls. The historical controls consisted of 31 SIV naïve rhesus macaques that had been infected with the same challenge stock of SIV239 that was used for infecting the vaccinated macaques in the present study. Peak height and set-point viral loads with this SIV239 are reasonably consistent from animal-to-animal and study-to-study (Johnson et al., 2003). Furthermore, we were able to compare the vaccinated and control groups as a whole as well as stratified based on their Mamu-A*01 status. A significant vaccine effect on viral load persisted even when only the Mamu-A*01-positive macaques were examined. Although these results are encouraging, the small number of vaccinated macaques in the current study precludes definitive conclusion of vaccine efficacy independent of other attenuating MHC Class I alleles such as Mamu-B*17 and Mamu-A*1303 (Muhl et al., 2002; O'Connor et al., 2003). Furthermore, relative efficacy of the two vaccine approaches cannot be assessed with the small number of vaccinated monkeys in this study.
It is somewhat disappointing that stronger levels of protection were not observed in these experiments, particularly given the breadth and nature of the immune responses that were generated by the vaccines. The disappointing levels of protection occurred despite attempts to use unrealistic vaccine/challenge conditions that favor the possibility of protection. The challenge virus, SIV239, is not only a homogeneous clone, it is exactly matched in sequence to the sequences present in the vaccine. Furthermore, challenge occurred six weeks after the fourth vaccine administration. The challenge virus, like primary isolates of HIV-1, principally uses CCR5 as its second receptor, is difficult to neutralize, and induces a chronic, progressive disease course following long-term persistent infection. It could be argued that IV inoculation of 10 monkey infectious doses may be uncharacteristically higher than the 1–2 infectious doses that typify natural infection. However, in the absence of sterilizing immunity, it is still disappointing that the cellular responses could not do a more effective job in controlling the sequence-matched challenge virus. It seems likely that a vaccine will need to be able to provide much better levels of protection against homologous SIV239 challenge to have any hope of working in human field trials. Among factors that could improve protective efficacy are efficient induction of mucosal immunity, induction of long-lived memory T cells with good proliferative capacity, i.e., central memory T cells (Vaccari et al., 2005; Wherry et al., 2003), and induction of higher levels of potently-neutralizing antibodies. In DNA/viral vector vaccine approaches, priming with viral vector via the mucosal route followed by systemic administration of DNA can induce excellent mucosal and systemic T cell and humoral immune responses (Eo et al., 2001). Whether priming with recombinant HSV via the mucosal route will similarly improve induction of mucosal immunity and improve protective efficacy against the AIDS virus remains to be determined.
The plasmids used for DNA vaccination contain the CMV promoter without an intron, the bovine growth hormone polyadenylation site, and the kanamycin resistant gene. The RNA (codon) optimized SIV genes were generated introducing multiple silent point mutations not affecting the sequence of the encoded proteins, as previously described for HIV-1 gag and env (Nasioulas et al., 1994; Schneider et al., 1997; Schwartz et al., 1992; Schwartz, Felber, and Pavlakis, 1992) using synthetic DNAs. The secreted and intracellularly degraded variants of the SIV antigens were generated by fusion of either IP10-MCP3 or of a beta-catenin (CATE)-derived peptide, respectively, at the N-terminus of Gag and Env, as previously described (Rosati et al., 2005) or by fusion of the CATE-derived peptide to polNTV, encoding a fusion protein comprising of an inactive pol, nef, tat and vif (von Gegerfelt et al, in preparation).
HSV recombinant vectors were constructed in the HSV-1 d106 mutant strains (Samaniego, Neiderhiser, and DeLuca, 1998). HSV-1 d106 contains mutations that inactivate expression of four immediate early genes, ICP4, ICP27, ICP22, and ICP47 and thus was grown in the E-11 cell line containing the ICP4 and ICP27 genes. The vaccine vector properties of d106 and the isolation of d106 based recombinant vectors expressing SIV env, SIV gag and a rev-tat-nef fusion protein will be described elsewhere (Watanabe et al, in preparation). Expression cassettes for SIV env and SIV gag utilized RNA-optimized ORFs described above, and the rev-tat-nef cassette was constructed from sequences in the p239SpE3’/nef-open plasmid (Regier and Desrosiers, 1990) and the nef plasmid (Kestler et al., 1990). The expression cassettes were inserted into transfer plasmids and recombined into the d106 genome at the UL54 (ICP27) gene locus by homologous recombination. Stocks for the immunization studies were prepared by growth on the E-11 complementing cell line.
Six Indian rhesus macaques housed in the herpes B-free specific pathogen free colony of the New England Primate Research Center (NEPRC) were enrolled in the study. Animals were maintained in accordance with federal and institutional guidelines mandated by the Animal Care and Use Committee of Harvard Medical School. Three rhesus macaques received a DNA prime, recombinant HSV boost vaccine regimen (Group DNA-DNA-HSV-HSV) while three rhesus macaques were only vaccinated with recombinant HSV (Group HSV-HSV-HSV-HSV).
The vaccine regimen consisted of four inoculations administered at 0, 4, 12 and 20 weeks (Figure 1). The DNA inoculum included five plasmid DNA constructs; MCP3p39 gag, MCP3p39 env, CATEDX gag, CATE239 env, and CATEpolNTV. One mg of each plasmid DNA (total 5 mg DNA) was injected at two to three sites via the intramuscular route. The HSV d106 vaccine inoculum included three recombinant HSV constructs, d106env, d106gag and d106rev-tat-nef, each administered at a dose of 1.25 to 3x109 pfu per dose. Two-thirds of the HSV vaccine dose was inoculated via the intramuscular route while one-third of the dose was administered via the subcutaneous route. All six vaccinated animals were challenged at 26 weeks with 10 animal infectious doses of cloned homologous pathogenic SIV239 via the intravenous route.
ELISA plates for HSV were obtained commercially (Viral Antigens, Memphis TN); SIV ELISA plates were constructed as previously described (Daniel et al., 1988; Wyand et al., 1996). Serum was assayed at a dilution of 1:20 or 1:200 in phosphate buffered saline (PBS)-Tween20 0.05% supplemented with 5% inactivated goat serum or 5% inactivated fetal bovine serum for the SIV and HSV ELISA, respectively. Plates were incubated at room temperature for 1 hour, washed three times with PBS-Tween20 0.05%, and incubated with 100 μl diluted goat anti-human IgG-conjugated alkaline phosphatase (Kirkegaard and Perry Laboratories, Gaithersburg, MD). Following a 1 hour incubation at room temperature, plates were washed three times and color was developed at room temperature with 200 μl p-nitrophenyl substrate solution (KPL) per well. After 30 minutes, the color reaction was stopped by the addition of 50 μl 3N NaOH per well, and absorbance was determined at 410nm on a microplate reader (Dynex Technologies, Chantilly, VA).
Neutralizing antibody titers to SIV239 and lab-passaged SIV251-lab were determined by the ability to block infection of CEMx174SIV-SEAP cells harboring a Tat-inducible SEAP reporter construct as described previously (Means, Greenough, and Desrosiers, 1997). Serial two-fold dilutions of plasma were incubated with virus at 100 μl per well in 96-well plates for 1 hour prior to addition of 3x104 CEMx174SIV-SEAP cells (100 μl) for a final volume of 200 μl per well. After 72 hours, 18 μl of supernatant was removed and assayed for SEAP activity using the Phospha-Light SEAP detection kit (Applied Biosystems, Foster City, CA) according to manufacturers instructions, with analysis on a Wallac multilabel microplate reader (Perkin Elmer, Boston, MA). Control wells for mock-infected cells and virus treated with pooled SIV-negative plasma were used to determine background and maximal SEAP production, respectively. The percent neutralization was calculated from the mean SEAP counts at each plasma dilution divided by the maximal SEAP counts after subtracting background activity. Fifty percent neutralization titers were calculated as the reciprocal of the plasma dilution at which infectivity was reduced by 50%.
The following anti-human monoclonal antibodies (mAbs) were obtained from BD Biosciences Pharmingen (San Jose, CA) unless otherwise indicated: anti-CD3 (clone SP34) fluorescein isothiocyanate (FITC) or phycoerythrin (PE), anti-CD4 (clone L200) allophycocyanin (APC), anti-CD8 (clone SK1) peridinin chlorophyll protein (PerCP), anti-CD11a (clone H111) FITC, anti-CD28 (clone CD28.2) FITC or PE, anti-CD62L-selection (clone SK11) PE, anti-β7 integrin (clone FIB 504) PE, anti-CCR7 (clone 150503, R&D System, Inc., Minneapolis, MN) PE and α4β7 (clone ACT-1; a gift from Leukosite, Inc.) PE. Anti-IFN-γ (clone B27) APC or FITC, anti-TNF-α (clone Mab11) APC or FITC, anti-CD69 (clone FN50) PE, anti-Granzyme B (clone GB12) PE and anti-Ki67 (clone B56) FITC were used for intracellular staining. Purified anti-CD28 (clone 28.2, 10 μg/ml) and anti-CD49d (clone 9F10, 10 μg/ml) were used for costimulation in cytokine staining assays (ICS). Affinity-purified F(ab′)2 fragments of goat anti-mouse IgG (H+L) (GAM, Kirkegaard and Perry laboratories, Gaithersburg, MD) was used for cross-linking the costimulatory antibodies in ICS assays.
Fifteen amino acid (aa) long peptides with 11aa overlaps spanning the entire SIV239 proteome were obtained through the NIH AIDS Research and Reference Reagent program and synthesized at the Massachusetts General Hospital peptide core facility (Charlestown, MA) by F-moc chemistry using an automated peptide synthesizer (MBS 396; Advanced Chemtech, Inc., Louisville, KY). Peptide pools for individual SIV proteins were made up and used for stimulation in ELISPOT and ICS assays. The individual peptide concentration used for stimulation ranged between 1 to 2 μg/ml. SIVmac251 P55 Gag antigen derived from the supernatant of baculovirus was obtained from Quality Biological, Inc (QBI, Gaithersburg, MD) through the Vaccine Research and Development Branch, Division of AIDS, NIAID, NIH and used at 5 μg/ml for stimulation.
Three complementary assays, the interferon-γ (INF-γ) ELISPOT assay, intracellular cytokine staining (ICS) assay and MHC tetramer analysis were used for quantitation and phenotypic characterization of SIV-specific CD4+ and CD8+ T lymphocytes.
DNA was extracted from PBMC or B-lymphoblastoid cell lines using the QIAamp DNA blood kit (Qiagen, Valencia, CA) and subjected to PCR with rhesus macaque MHC Class I sequence specific primers for Mamu-A*01 and Mamu- B*17 as described (Knapp et al., 1997; Muhl et al., 2002).
Blood collected in tubes containing EDTA were spun at 2000 rpm for 10 min within 3 h of drawing blood and stored at –800C for quantitation of SIV RNA. SIV RNA was quantified using a real time RT PCR assay, as previously described (Cline et al., 2005). As used in the present study the assay has a threshold sensitivity of 10 SIV gag RNA Copy Eq/mL of plasma.
Trigeminal and sacral ganglia obtained at necropsy were minced and co-cultured onto Vero cells and the E-11 complementing cell line. Co-cultures were visually monitored twice daily for one week for evidence of cytopathic effects.
At necropsy, five-μ sections of formalin-fixed, paraffin-embedded tissues were stained with hematoxylin and eosin. Special stains included: Brown and Brenn Gram, modified Steiner’s silver, Ziehl-Neelsen acid fast, Grocott’s methenamine silver and periodic acid-Schiff. Monoclonal antibodies used for immunohistochemistry included: anti-SIVnef (clone KK75, Centralised Facility for AIDS Reagents, National Institute for Biological Standards and Control, Potters Bar, U.K.), anti-HIV1-p24/SIVp27 (clone 183-H12-5C, NIH AIDS Research & Reference Reagent Program, Germantown, MD), anti-SV40 T Ag Ab-2 (clone PAb416, Oncogene, Boston, MA), anti-adenovirus (clone 20/11, Chemicon, Temecula, CA); polyclonal antibodies included: anti-rhesus cytomegalovirus IE1 (gift from Dr. Peter Barry, University of California, Davis), anti-human myelin basic protein (Dako, Carpinteria, CA), and anti-herpes simplex virus 1 (Dako), which also immunolabels herpes B. Five-μ sections of paraffin-embedded sections of the left occipital lobe were deparaffinized and rehydrated, and hydrogen peroxide was applied for 5 minutes to quench endogenous peroxidase. Antigen retrieval was achieved by microwaving for 20 minutes in citrate buffer with the exceptions of anti-adenovirus, which was treated with proteinase K for 5 minutes, and anti-herpes simplex, which required no antigen retrieval. Dako Protein Block was applied for 10 minutes, and duplicate sections were incubated with either the primary antibody or the appropriate irrelevant antibody for negative controls (irrelevant IgG1 for monoclonal antibodies, and anti-rabbit immunoglobulin fraction for polyclonal antibodies). Next, biotin block was applied for 20 minutes, followed by incubation with the secondary antibody (biotinylated horse anti-mouse for monoclonal antibodies, and biotinylated goat anti-rabbit for polyclonal antibodies) for 30 minutes. Vectastain avidin-biotin complex (Vector, Burlingame, CA) was applied for 30 minutes and the reaction was visualized with DAB chromogen (Dako) and counterstained with Mayer’s hematoxylin
Formalin-fixed tissue from the left occipital lobe was processed for electron microscopy. Specimens were washed in 0.1 M phosphate buffer and postfixed in 1% osmium tetroxide. The specimens were dehydrated, embedded in Epon araldite, and stained with uranyl acetate and Sato’s lead stain for examination using a Jeol 1010 Electron Microscope (Jeol, Tokyo, Japan).S
Statistical analysis was performed using Statview software (Abacus Concepts, Inc., Berkeley, CA). The nonparametric Mann-Whitney U test was used for comparison of unpaired groups. Correlations between viral load and immune parameters were performed by the Spearman Rank Correlation test. Area under curve (AUC) for plasma SIV RNA was calculated using the GraphPad Prism software (GraphPad software Inc., San Diego, CA).
This study was supported by HIV Vaccine Research and Development grant AI46006 from the NIAID, PHS grant RR00168 and with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. NO1-CO-12400. The authors thank the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH for providing SIV239 p55 antigen and the 183-H12-5C mAb to HIV-1 p24, Dr K Kent and Ms C. Arnold and the NIBSC Centralised Facility for AIDS Reagents supported by EU programme EVA contract and the UK Medical Research Council for the KK75 mAb to SIV Nef, P. Roth for DNA preparation, M. Piatak Jr. and K. Oswald for assistance with plasma viral load measurements, Keith Mansfield, Angela Carville and the staff of the Division of Primate Medicine at NEPRC for assistance with animal studies, David Watkins for providing viral load data on control SIV239-infected rhesus macaques and Neal DeLuca for providing the HSV d106 vector. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organization imply endorsement by the U.S. Government.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.