Fine analysis of humoral antibody response to envelope glycoprotein of SIV in infected and vaccinated macaques

Simian immunodeficiency virus infection evades vaccine-elicited antibody responses to V2 region

OBJECTIVES

An effective AIDS vaccine should elicit protective antibody responses against HIV/simian immunodeficiency virus (SIV) infection. We recently reported that mucosal priming with a replicating modified vaccinia Tiantan virus (MVTTgpe)-based vaccine regimen induces durable protection against pathogenic SIVmac239 infection in rhesus monkeys. Here, we aim to conduct a comprehensive analysis on antigenic determinants recognized by specific antibody responses generated by vaccination and SIVmac239 infection.

METHODS

A novel yeast surface displayed antigen library of entire SIVmac239 envelope (Env) glycoprotein was established and validated to map the major antigenic determinants (MAD) in monkey sera elicited by vaccination and infection. MAD-directed antibody responses were further analyzed for correlation of protection.

RESULTS AND CONCLUSIONS

The yeast surface displayed library allows the mapping of SIV-specific linear and conformational MAD. The MVTTgpe-based regimen induces antibodies targeting mainly to 6 antigenic domains covering the entire gp160. Critically, this regimen induced a uniquely predominant antibody response against a distinct MAD in variable region 2 (V2) as compared with the Ad5gpe-based vaccine and SIVmac239 infection. This MAD was associated with a higher titer of anti-V2 antibody responses, which was inversely correlated with peak and set-point viral loads. Unexpectedly, the pathogenic SIVmac239 challenge evaded the vaccine-elicited anti-V2 antibody response. Instead of recalling B-cell memory responses to the V2 MAD, viral infection directed anti-V1V2 antibodies primarily to V1 region. Moreover, the anti-V1V2 antibody responses diminished significantly in infected macaques after they enter the stage of simian AIDS. Our findings have critical implications to AIDS vaccine efforts with focus on V2 region.

Impact of glycosylation on antigenicity of simian immunodeficiency virus SIV239: induction of rapid V1/V2-specific non-neutralizing antibody and delayed neutralizing antibody following infection with an attenuated deglycosylated mutant

Infection of rhesus macaques with a deglycosylation mutant, Delta5G, derived from SIV239, a pathogenic clone of simian immunodeficiency virus (SIV), led to robust acute-phase viral replication followed by a chronic phase with undetectable viral load. This study examined whether humoral responses in Delta5G-infected animals played any role in the control of infection. Neutralizing antibodies (nAbs) were elicited more efficiently in Delta5G-infected animals than in SIV239-infected animals. However, functional nAb measured by 90% neutralization was prominent in only two of the five Delta5G-infected animals, and only at 8 weeks post-infection (p.i.), when viral loads were already below 10(4) copies ml(-1). These results suggest a minimal role for nAbs in the control of the primary infection. In contrast, whilst Ab responses to epitopes localized to the variable loops V1/V2 were detected in all Delta5G-infected animals at 3 weeks p.i., this response was associated with a concomitant reduction in Ab responses to epitopes in gp41 compared with those in SIV239-infected animals. These results suggest that the altered surface glycosylation and/or conformation of viral spikes induce a humoral response against SIV that is distinct from the response induced by SIV239. More interestingly, whereas V1/V2-specific Abs were induced in all animals, these Abs were associated with vigorous Delta5G-specific virion capture ability in only two Delta5G-infected animals that exhibited a functional nAb response. Thus, whereas the deglycosylation mutant infection elicited early virion capture and subsequent nAbs, the responses differed among animals, suggesting the existence of host factors that may influence the functional humoral responses against human immunodeficiency virus/SIV.

Comparative analysis of HIV-1 recombinant envelope glycoproteins from different culture systems

The productivity of stable Chinese hamster ovary cell lines secreting HIV-1 monomeric (IIIB gp120) and oligomeric (UG21 gp140) recombinant envelope glycoproteins was compared in serum-containing (S+), serum-free (S-) and protein-free (P-) culture media. UG21 gp140 expression was greatest in S+ medium, while IIIBgp120 production was lower than gp140 in all three media but highest in S-. UG21 gp140 production was highest in standard 850-cm2 roller bottle cultures in S+ media, peaking after 14 days of incubation, while expression levels in the three media were 0.5 (S+), 0.4 (S-) and 0.2 (P-) mg/l, from which 90, 80 and 12% of gp140, respectively, could be purified by immunoaffinity chromatography. Purified UG21 gp140 from S+ and S- media possessed biological functionality as evidenced by CD4 and monoclonal antibody (Mab) binding. In contrast, UG21 gp140 from P- medium appears to be misfolded and non-functional. Despite the possession of a different N-linked glycan profile, UG21 gp140 from S- media shows very similar CD4 and Mab binding characteristics to S+ UG21 gp140. The relevance of these findings to HIV vaccine development is discussed.

CD8+ lymphocytes do not mediate protection against acute superinfection 20 days after vaccination with a live attenuated simian immunodeficiency virus

In order to test the hypothesis that CD8+ cytotoxic T lymphocytes mediate protection against acute superinfection, we depleted >99% of CD8+ lymphocytes in live attenuated simian immunodeficiency virus macC8 (SIVmacC8) vaccinees from the onset of vaccination, maintained that depletion for 20 days, and then challenged with pathogenic, wild-type SIVmacJ5. Vaccinees received 5 mg per kg of humanized anti-CD8 monoclonal antibody (MAb) 1 h before inoculation, followed by the same dose again on days 3, 7, 10, 13, and 17. On day 13, peripheral CD8+ T lymphocytes were >99% depleted in three out of four anti-CD8 MAb-treated vaccinees. At this time attenuated SIVmacC8 viral RNA loads in anti-CD8 MAb-treated vaccinees were significantly higher than control vaccinees treated contemporaneously with nonspecific human immunoglobulin. Lymphoid tissue CD8+ T lymphocyte depletion was >99% in three out of four anti-CD8 MAb-treated vaccinees on the day of wild-type SIVmacJ5 challenge. All four control vaccinees and three out of four anti-CD8 MAb-treated vaccinees were protected against detectable superinfection with wild-type SIVmacJ5. Although superinfection with wild-type SIVmacJ5 was detected at postmortem in a single anti-CD8 MAb-treated vaccinee, this did not correlate with the degree of preceding CD8+ T lymphocyte depletion. Clearance of attenuated SIVmacC8 viremia coincided with recovery of normal CD8+ T lymphocyte counts between days 48 and 76. These results support the view that cytotoxic T lymphocytes are important for host-mediated control of SIV primary viremia but do not indicate a central role in protection against acute superinfection conferred by inoculation with live attenuated SIV.

Vaccination with live attenuated simian immunodeficiency virus for 21 days protects against superinfection

The identification of mechanisms that prevent infection with human immunodeficiency virus (HIV) or simian immunodeficiency virus (SIV) would facilitate the development of an effective AIDS vaccine. In time-course experiments, protection against detectable superinfection with homologous wild-type SIV was achieved within 21 days of inoculation with live attenuated SIV, prior to the development of detectable anti-SIV humoral immunity. Partial protection against superinfection was achieved within 10 days of inoculation with live attenuated SIV, prior to the development of detectable anti-SIV humoral and cellular immunity. Furthermore, co-inoculation of live attenuated SIV with wild-type SIV resulted in a significant reduction in peak virus loads compared to controls that received wild-type SIV alone. These findings imply that innate immunity or non-immune mechanisms are a significant component of early protection against superinfection conferred by inoculation with live attenuated SIV.

Humoral Immunity and the Evolution of HIV-2

Human immunodeficiency virus type 2 (HIV-2) evolved from the zoonotic transmission of simian immunodeficiency virus (SIV) that naturally infects sooty mangabeys found in West Africa. Using sera from HIV-2-infected humans, we discovered that an hypervariable region (the V4 loop) of HIV-2 induces antibody responses only weakly reactive against itself but strongly reactive against analogous sequences from the V4 loop of strains of SIV. Available sequence data indicates that all strains of HIV-2 have large deletions in the V4 region that truncate an immunodominant neutralizing B cell epitope among strains of SIV. Infection of a macaque with a sequenced clone of HIV-2 similarly elicited antibodies that poorly recognized the V4 loop of HIV-2 but readily bound to analogous SIV sequences. Our data are consistent with a scenario whereby a disparate antibody response directed against the V4 loop may have influenced the selective expansion and survival of HIV-2 in humans.

Mechanisms of protection induced by attenuated simian immunodeficiency virus

To determine whether attenuated simian immunodeficiency virus (SIV) vaccines confer protection against superinfection via secondary cellular immune responses, we searched for markers of immune activation following rechallenge. Productive infection with either attenuated SIVmacC8 or wild-type SIVmacJ5 resulted in a transient increase in T-lymphocyte CD25 and Mafa-DR expression. A pronounced increase in the frequency of FAS+ CD8+ lymphocytes was observed following SIVmacJ5 infection only. A transient increase in lymphocytes positive for intracellular IFN-gamma and IL-4 was observed following primary infection with either virus. In contrast, lymphocytes positive for intracellular IL-2 were reduced. Following SIVmacJ5 challenge of SIVmacC8-infected vaccinees, no evidence of detectable superinfection was obtained. Rechallenge of vaccinees did not alter the frequency of activated peripheral T-lymphocytes, perturb cytokine profiles, or generate an anamnestic antibody response. These data do not support the hypothesis that protection conferred by live attenuated SIV is mediated by the induction of vigorous T-cell responses upon rechallenge.

The quest for an AIDS vaccine: is the CD8+ T-cell approach feasible?

The rationale for developing anti-HIV vaccines that stimulate cytotoxic T-lymphocyte responses is given. We argue that such vaccines will work, provided that attention is paid to the development of memory T-cell responses that are strong and preferably activated. Furthermore, the vaccine should match the prevailing virus clade as closely as possible. Vaccines will have to stimulate a wide range of responses, but it is not clear how this can be achieved.

Characterization of neutralization epitopes of simian immunodeficiency virus (SIV) recognized by rhesus monoclonal antibodies derived from monkeys infected with an attenuated SIV strain

A major limitation in the simian immunodeficiency virus (SIV) system has been the lack of reagents with which to identify the antigenic determinants that are responsible for eliciting neutralizing antibody responses in macaques infected with attenuated SIV. Most of our information on SIV neutralization determinants has come from studies with murine monoclonal antibodies (MAbs) produced in response to purified or recombinant SIV envelope proteins or intact SIV-infected cells for relatively short periods of time. While these studies provide some basic information on the potential immunogenic determinants of SIV envelope proteins, it is unclear whether these murine MAbs identify epitopes relevant to antibody responses elicited in monkeys during infection with either wild-type or attenuated SIV strains. To accomplish maximum biological relevance, we developed a reliable method for the production of rhesus monoclonal antibodies. In the present study, we report on the production and characterization of a unique panel of monoclonal antibodies derived from four individual monkeys inoculated with SIV/17E-CL as an attenuated virus strain at a time when protective immunity from pathogenic challenge was evident. Results from these studies identified at least nine binding domains on the surface envelope glycoprotein; these included linear determinants in the V1, V2, cysteine loop (analogous to the V3 loop in human immunodeficiency virus type 1), and C5 regions, as well as conformational epitopes represented by antibodies that bind the C-terminal half of gp120 and those sensitive to defined mutations in the V4 region. More importantly, three groups of antibodies that recognize closely related, conformational epitopes exhibited potent neutralizing activity against the vaccine strain. Identification of the epitopes recognized by these neutralizing antibodies will provide insight into the antigenic determinants responsible for eliciting neutralizing antibodies in vivo that can be used in the design of effective vaccine strategies.

Specific proliferative T cell responses and antibodies elicited by vaccination with simian immunodeficiency virus Nef do not confer protection against virus challenge

The efficacy of immunizing with a combination of simian immunodeficiency virus (SIV) Nef vaccines was evaluated. Four vaccinates received three intradermal immunizations with recombinant vaccinia virus that expressed SIV Nef, followed by three intramuscular immunizations with rDNA also expressing SIV Nef. Finally, the four vaccinates received two subcutaneous boosts with recombinant SIV Nef protein. This immunization protocol elicited anti-Nef antibodies in all of the vaccinates as well as specific proliferative responses. However, specific cytotoxic T cell responses were not detected before virus challenge. All vaccinates were challenged intravenously with 10 MID(50) of SIVmacJ5 along with four controls. All eight subjects became infected after SIV challenge and there were no group-specific differences in virus load as measured by virus titration and vRNA analysis. The results of this study support indirectly the report from Gallimore and colleagues (Nat Med 1995;1:1667) suggesting that CD8(+) T lymphocyte responses are required for Nef-based vaccines to restrict SIV infection. If Nef-based vaccines are to be beneficial in controlling infection with immunodeficiency viruses, then it will be necessary to develop more effective immunization protocols that elicit potent CD8(+) cell responses reproducibly.

Characterization and epitope mapping of neutralizing monoclonal antibodies produced by immunization with oligomeric simian immunodeficiency virus envelope protein

In an attempt to generate broadly cross-reactive, neutralizing monoclonal antibodies (MAbs) to simian immunodeficiency virus (SIV), we compared two immunization protocols using different preparations of oligomeric SIV envelope (Env) glycoproteins. In the first protocol, mice were immunized with soluble gp140 (sgp140) from CP-MAC, a laboratory-adapted variant of SIVmacBK28. Hybridomas were screened by enzyme-linked immunosorbent assay, and a panel of 65 MAbs that recognized epitopes throughout the Env protein was generated. In general, these MAbs detected Env by Western blotting, were at least weakly positive in fluorescence-activated cell sorting (FACS) analysis of Env-expressing cells, and preferentially recognized monomeric Env protein. A subset of these antibodies directed toward the V1/V2 loop, the V3 loop, or nonlinear epitopes were capable of neutralizing CP-MAC, a closely related isolate (SIVmac1A11), and/or two more divergent strains (SIVsmDeltaB670 CL3 and SIVsm543-3E). In the second protocol, mice were immunized with unfixed CP-MAC-infected cells and MAbs were screened for the ability to inhibit cell-cell fusion. In contrast to MAbs generated against sgp140, the seven MAbs produced using this protocol did not react with Env by Western blotting and were strongly positive by FACS analysis, and several reacted preferentially with oligomeric Env. All seven MAbs potently neutralized SIVmac1A11, and several neutralized SIVsmDeltaB670 CL3 and/or SIVsm543-3E. MAbs that inhibited gp120 binding to CD4, CCR5, or both were identified in both groups. MAbs to the V3 loop and one MAb reactive with the V1/V2 loop interfered with CCR5 binding, indicating that these regions of Env play similar roles for SIV and human immunodeficiency virus. Remarkably, several of the MAbs generated against infected cells blocked CCR5 binding in a V3-independent manner, suggesting that they may recognize a region analogous to the conserved coreceptor binding site in gp120. Finally, all neutralizing MAbs blocked infection through the alternate coreceptor STRL33 much more efficiently than infection through CCR5, a finding that has important implications for SIV neutralization assays using CCR5-negative human T-cell lines.

Mechanisms of protection induced by attenuated simian immunodeficiency virus. II. Lymphocyte depletion does not abrogate protection

To determine the role that cellular immune responses play in the protection conferred by vaccination with attenuated SIVmac32H (pC8), we have attempted to deplete macaques of their CD8+ cells prior to challenge with wild-type SIVmac32H (pJ5). In two of four pC8-infected macaques, N109 and N112, a transient partial depletion of CD8+ cells by antibody treatment was achieved. On the day of challenge peripheral CD2+CD4-CD8+ cell counts were reduced by 92 and 95%, respectively, in animals N109 and N112 and their lymph nodes revealed a 46 and 58% reduction, respectively, in CD2+CD4-CD8+ cells. Two other pC8-immunized macaques, N110 and N111, treated in the same way, did not show significant depletion of CD8+ cells. None of these four pC8-immunized animals became infected when challenged with 50 MID50 of pJ5. Treatment of a further four pC8-infected and protected macaques and two naive control animals with Campath-1H antibody successfully depleted peripheral CD3+ cell counts by >99% in all treated animals. Campath-1H depletion resulted in enhanced, longer lasting lymphoid depletion. Yet subsequent challenge with 20 MID50 of pJ5 still failed to infect the pC8-immunized animals. All eight of the naive controls, including two Campath-1H-treated animals, became infected following challenge. In summary, partial depletion of circulating CD8+ cells or total lymphocytes prior to challenge failed to abrogate the protection conferred by vaccination with pC8.

Simian immunodeficiency virus (SIV) envelope-specific Fabs with high-level homologous neutralizing activity: recovery from a long-term-nonprogressor SIV-infected macaque

An antibody phage display library was constructed from RNA extracted from lymph node cells of a simian immunodeficiency virus (SIV)-infected long-term-nonprogressor macaque. Seven gp120-reactive Fabs were obtained by selection of the library against SIV monomeric gp120. Although each of the Fabs was unique in sequence, there were two distinct groups based on epitope recognition, neutralizing activity in vitro, and molecular analysis. Group 1 Fabs did not neutralize SIV and bound to a linear epitope in the V3 loop of the SIV envelope. In contrast, two of the group 2 Fabs neutralized homologous, neutralization-sensitive SIVsm isolates with high efficiency but failed to neutralize heterologous SIVmac isolates. Based on competition enzyme-linked immunosorbent assays with mouse monoclonal antibodies of known specificity, these Fabs reacted with a conformational epitope that includes domains V3 and V4 of the SIV envelope. These neutralizing and nonneutralizing Fabs provide valuable standardized and renewable reagents for studying the role of antibody in preventing or modifying SIV infection in vivo.

Identification of the V1 region as a linear neutralizing epitope of the simian immunodeficiency virus SIVmac envelope glycoprotein

The sequence variability of viral structure polypeptides has been associated with immune escape mechanisms. The V1 region of simian immunodeficiency virus (SIV) is a highly variable region of the SIVmac env gene. Here, we describe the V1 region as a linear neutralizing epitope. V1 region-specific neutralizing antibodies (NAb) were first demonstrated in a rabbit infected with a recombinant vaccinia virus carrying the env gene of human immunodeficiency virus type 2 strain ben (HIV-2ben). Since we detected in this animal V1 region-specific NAb that were able to neutralize not only human immunodeficiency virus type 2 but also SIVmac32H, we investigated whether a similar immune response is evoked in macaques (Macaca mulatta) either infected with SIVmac or immunized with the external glycoprotein (gp130) of the same virus. Distinctly lower NAb titers were found in the SIVmac-infected animals than in the gp130-immunized macaques. Since the NAb titers in both groups were high enough for competition experiments, we used five overlapping peptides encompassing the whole V1 region for a detailed identification of the epitope. In each of the 12 macaques investigated, we detected a high level of NAb reacting with at least one peptide located in the central part of the V1 region. The relatively high degree of divergence, especially within the central part of the V1 region, which characterized the evolution of the retroviral sequences from the original inoculum in the infected macaques suggests the development of escape mutants. Furthermore, 3 of 12 animals developed NAb directed against the amino-terminal end of the V1 region epitope. Sequence analysis, however, revealed relatively low levels of genetic drift and genetic variability within this part of the V1 region. The induction of V1 env-specific NAb not only in gp130-immunized macaques but also in SIVmac-infected animals in combination with the increased genetic variability of this region in vivo indicates a marked biological significance of this epitope for the virus.

Efficient purification and rigorous characterisation of a recombinant gp120 for HIV vaccine studies

A recombinant HIV-1 gp120 (rgp120) was expressed in a permanent Chinese Hamster Ovary (CHO) cell line (L761h) that constitutively secretes the product of clone p4 derived from the env gene of HIV-1 isolate GB8. The rgp120 was isolated from cell culture supernatants by a simple, rapid, non-denaturing and efficient purification procedure based on a novel combination of lectin affinity and FPLC ion-exchange chromatography. The purity of the isolated glycoprotein was rigorously confirmed by SDS-PAGE, capillary electrophoresis, laser desorption mass spectrometry, total amino acid analysis and N-terminal amino acid sequencing. The retention of biological activity by the purified rgp120 was assessed by determining the dissociation constant of rgp120 binding to sCD4. After formulation of this highly purified and biologically active rgp120 with "conventional" adjuvants, including types already used in clinical trials of candidate gp120-based HIV vaccines, antibody responses in immunised rabbits were analysed using panels of overlapping synthetic peptides. The consequences of using currently available adjuvants to deliver highly specialised and perhaps conformation-dependent molecules, like HIV gp120, are presented and discussed in the context of HIV vaccine development.

The simian immunodeficiency virus transmembrane protein is poorly immunogenic in inactivated virus vaccine

The transmembrane proteins (TMP) of immunodeficiency lentiviruses are primary candidates for inclusion in AIDS vaccines, the design and testing of which is facilitated by the SIV-macaque infection model. Antibody responses to linear determinants in the SIVmac TMP were investigated in rhesus macaques either infected with the SIVmac J5 molecular clone or vaccinated with partially purified, formalin-inactivated SIVmac. Infected animals were shown to recognise predominantly four regions in the external domain and three regions in the internal domain of the TMP defined by a series of nominally 20mer overlapping peptides. In contrast SIV vaccinates had extremely restricted and weak antibody responses to the TMP, indicating a selective loss of immunogenicity of this component in the vaccine.

Neutralising epitopes of simian immunodeficiency virus envelope glycoprotein

Monoclonal and polyclonal antibodies with weak SIV neutralising activity bind to the V2 and V4 regions of gp120 or bind to the amino acids DWNND in gp41. Antibodies with the most potent neutralising activity recognise conformation-dependent epitopes involving the V3 and V4 regions of gp120. Monoclonal antibodies that map to the V3 region of SIVmac failed to neutralise. However, one antibody to SIV AGM neutralised but only in the presence of soluble CD4.