New insights into evaluating effective T-cell responses to HIV.

Abstract

Introduction CD8 cytotoxic T lymphocyte (CTL) and CD4 T-helper lymphocyte (HTL) responses play important roles in the containment of HIV and SIV. The appearance of CTL in the acute phase correlates with the initial control of primary HIV and SIV viremia [1-5], and depletion of these cells prior to SIV infection results in an inability to control SIV [6-8]. Similarly, it is well documented that patients who mount and maintain strong HIV-1-specific CD4 T-cell responses are better able to control virus replication [9,10]. However, despite numerous studies delineating the roles of CD8 and CD4 T-cell responses in controlling HIV and SIV, a comprehensive understanding of what particular attributes of T-cell responses are most effective at controlling HIV replication is still lacking. As such, the exact composition of CTL and HTL responses that should be engendered by a vaccine, or boosted during therapeutic treatment, remains largely undetermined. This review will highlight some of the contemporary techniques used to define T-cell responses to HIV and SIV, as well as some of the recent discoveries that are redefining what constitutes an effective T-cell response against HIV. Towards addressing questions related to HIV pathogenesis and vaccine design, we are fortunate to have the availability of the SIV-infected rhesus macaque model. The strengths of this model, which include a rapid rate of progression to AIDS, and knowledge of the dose, route, and strain of the infecting virus, complement studies in HIV-infected patients in which the reagents, host genetics, and access to samples are more extensive and better defined. Unfortunately, there is currently still too little known about the antiviral immune responses in either system to directly and accurately compare their similarities and differences, and to draw any definitive conclusions. Therefore, the data and views presented herein will simply reflect what has recently been discovered in both humans and non-human primate studies. Contemporary techniques for the evaluation of T-cell responses The advent of tetramer technology in 1996 revolutionized our ability to accurately measure T-cell responses to HIV and many other pathogens [11]. Tetramers now provide a level of sensitivity and ease with which to measure T-cell responses that was not previously possible using traditional methods such as 51Cr-release assays, limiting dilution assays and proliferation assays. Unfortunately, the use of these tetramers is limited by the requirement for the optimal peptide and restricting major histocompatibility complex (MHC) class I or class II molecule. The recent introduction of enzyme-linked immunospt (ELISPOT) [12-14] and intracellular cytokine staining (ICS) assays [15-17], however, has further improved our ability to measure cellular responses (Table 1). These techniques, in conjunction with overlapping peptide libraries to HIV and SIV, provide an unprecedented opportunity to critically define the strength, breadth and specificity of cellular immune responses against HIV and SIV. Similarly, the capability to sequence entire viral genomes with relative ease offers the ability to compare the selective pressures induced by various cellular immune responses and to monitor the evolution of the virus. These powerful techniques, combined with many of the traditional assays for measuring cellular immune responses, are rapidly yielding a greater understanding of the complexity of the immune systems response against HIV and SIV. These techniques are, however, also revealing the extent to which HIV and SIV are capable of thwarting many of these responses.Table 1: Methods available to measure T-cell responses.ELISPOT assays Perhaps the most powerful new tool for the analysis of T-cell responses is the ELISPOT assay. While less elegant than ICS, this method is far more applicable to the contemporary approach being used in screening patients for T-cell responses to the entire HIV genome. This is due largely to the relative simplicity of the assay that allows it to be adaptable to most laboratories worldwide, and the reduced cost of the assay compared with ICS. ELISPOT assays are now available for many cytokines, among them interferon-γ (IFN-γ), tumor necrosis factor-α, interleukin-2, and interleukin-4. This technique requires plating of cells, generally in a 96-well format, and exposing these cells to various antigens, often individual or pools of peptides. Following incubation periods of 6-24 h, the cells are removed and the local production of cytokines within each well is measured through a sandwich enzyme-linked immunosorbent assay-based assay. The magnitude of the response measured represents the total T-cell response (CD4 + CD8). The power of this technique lies in its ability to sensitively analyze hundreds of responses in a single simple assay. Computerized imagers are also now available that greatly reduce the subjectivity and time required for the post-assay analysis of the samples. Careful handling of these cells also allows for recycling of cells following the assay to generate CTL or T-helper lines, an important consideration when samples are limited. Furthermore, in utilizing peptide pools and a matrix approach to the screening of these peptide pools, the entire T-cell response to HIV in a patient can be preliminarily assessed using a single ELISPOT plate requiring as few as 10 million peripheral blood mononuclear cells [18]. Together, these attributes will undoubtedly make the ELISPOT assay the 'workhorse' for the evaluation of current and future vaccine trials assessing cellular immune responses. Intracellular cytokine staining ICS represents the state-of-the-art methodology for analyzing T-cell responses. Like ELISPOT, ICS similarly measures antigen-specific IFN-γ and tumor necrosis factor-α production from individual cells. This method represents an extremely sensitive assay for accessing the ability of various cell subsets to respond to particular antigens. It requires a brief incubation period during which cells are exposed to individual antigens, peptides, or pools of peptides. Cytokines produced by antigen-specific cells accumulate within the cells during this time. Surface labeling of cells for CD8 or CD4 cells followed by intracellular staining for the specific cytokines of interest then allows for identification of antigen-specific cells. ICS offers the unique ability to measure both CD4 and CD8 cell responses in a single tube with incredible sensitivity. Furthermore, it is also adaptable to a 96-well format, enabling the measurement of hundreds of responses in a single assay. Unfortunately, this method is limited to fewer laboratories by its requirement for access to flow cytometers and the expense of the antibodies used. Furthermore, this assay generally requires many more cells than a comparable ELISPOT assay, an important consideration to many investigators. Because of these limitations, it is unlikely that ICS will prove useful as a major screening assay for vaccine trials. However, in conjunction with tetramers, this assay has begun to prove invaluable in the analysis of T-cell function and dysfunction [19,20]. Viral genome sequencing One of the newer approaches to identifying and characterizing cellular immune responses to HIV and SIV is through whole viral genome sequencing [21,22] to identify viral escape mutants. While the approach of viral sequencing to define the degree to which the virus has evolved away from a particular CD8 cell response is not novel, the concept that this approach can be routinely used to define unique CD8 cell responses that exert selective pressure represents a potentially powerful technique. For example, direct bulk sequencing of the open reading frames of SIV at 4 weeks post-infection from rhesus macaques infected with cloned SIVmac239 revealed a number of unique mutations [22]. Not only did this approach reveal some previously unidentified CD8 cell epitopes, it also revealed a unique set of CD8 cell responses capable of exerting selective pressure very early after infection (D.H. O'Connor, manuscript submitted). Unlike conventional approaches to epitope identification that rely on monitoring the levels of immune responses in the host, this approach allows the virus to highlight those regions that are under selective pressure and, therefore, those immune responses most capable of exerting such selective pressure. Whole viral genome sequencing will also be important for accurately defining the sequence of the particular HIV strain infecting any given patient. Since current HIV peptide libraries are based on HIV clade consensus sequences, many T-cell responses in individuals remain undetectable due to viral mutations in the infecting strain that vary from the consensus peptide library sequence. Therefore, employing whole HIV genome sequencing will also be important in defining T-cell responses in HIV-infected subjects. Overview Traditionally, whole antigen or vaccinia virus vectors expressing particular viral proteins have been used to stimulate or measure particular T-cell responses. Overlapping peptide libraries, often based on clade consensus sequences, now represent the state-of-the-art for both stimulation and measurement. Virtually all assays yield significantly higher responses to peptides (or pools of peptides) compared with proteins or whole antigens, and the use of overlapping peptides dramatically simplifies the definitive identification of epitopes. Costs of these peptide libraries, however, are prohibitively expensive. Regardless of the source of the antigen, one limitation is that for any patient studied these peptides or proteins do not accurately reflect the exact sequence of the virus against which any given individual is mounting immune responses. Therefore, to begin addressing the degree to which an individual's global T-cell response is being underestimated, autologous peptide libraries will need to be synthesized once the exact sequence of the virus in an individual is defined through whole genome viral sequencing. Again, while this represents a prohibitively expensive undertaking that will probably be limited in its scope, the data generated from such studies may prove invaluable towards our understanding of HIV-specific immune responses. Comparisons of CD8 T-cell responses using the various assays available indicate that tetramer, ELISPOT, and ICS assays are equally sensitive, with limiting dilution assays generally underestimating levels 10-fold to 100-fold [23]. However, despite the numerous assays available to measure both CD8 and CD4 T-cell responses, and a growing understanding of their role in controlling both HIV and SIV infections, there exists no clear approach or method of choice to most effectively evaluate these responses. The combination of various assays has, however, begun to reveal a degree of T-cell dysfunction in HIV and SIV infections. This has now been described for both the lytic [19] and cytokine [20,24] responses of CD8 T cells. Similarly, deficiencies in CD4 T-cell proliferation and cytokine production have been described [25,26]. What is missing, however, is a more clear understanding of the importance of such deficiencies to HIV's ability to evade the immune system. It will only be through continued analysis of HIV-specific and SIV-specific immune responses with combinations of these approaches that we might begin to reveal more precise correlates of immunity and begin to redefine the appropriateness of these assays. Furthermore, the ability to more routinely define regions of viral escape throughout the HIV genome, and the rate and frequency with which these escapes are occurring in numerous patients, will better enable us to define correlates of 'protective' CD8 and CD4 T-cell responses and the impact of the associated T-cell dysfunction on the control of HIV infections. CD8 T-cell correlates of immunity The role of CD8 T cells in controlling various viral infections is well documented. Some specificities of the CD8 T-cell response that are probably important in the control of HIV include breadth (number of epitopes recognized), specificity (viral protein target or MHC restriction), effector function (i.e., lytic versus cytokine), potency (i.e., selective pressure), and compartmentalization (i.e., mucosal). While examination of these attributes of the CD8 T-cell responses to HIV have revealed some remarkable findings, no clear correlates of protective immunity have been identified to better enable us to define an effective HIV vaccine. However, with the more powerful techniques of ELISPOT, ICS and whole genome viral sequencing available to accurately assess the characteristics of successful and failed immune responses, we may be better able to define the types of immune responses that should be engendered in a vaccine. Breadth The importance of broad CD8 cell responses in the control of various pathogens including HIV [27], hepatitis B virus (HBV) [28,29], Epstein-Barr virus (EBV) [27], malaria [30] and influenza [31] is well defined. Breadth may be especially important to the control of HIV since this virus is capable of evading many CD8 cell responses through rapid mutation. Fortunately, there now exists an extensive list of identified HIV and SIV/SHIV CD8 cell epitopes [32,33]. We are only now, however, beginning to fully comprehend the true breadth of an immune response to HIV or SIV. For example, in the SIV-infected rhesus macaque model, a single MHC class I molecule (Mamu-A*01) has been shown to present at least 14 CD8 T-cell epitopes from a variety of SIV proteins [34]. Similarly, as many as 29 individual CD8 cell responses have been identified in a single HIV-infected individual who is homozygous at all three MHC class I loci (M. Altfeld, personal communication, 2001). These studies reinforce the immense breadth of CD8 cell responses we might expect to see from an immune response to a pathogen but, more importantly, indicates that we are undoubtedly still unaware of many of the CD8 cell responses elicited against HIV and SIV. Without complete knowledge of all of the CD8 cell epitopes an individual will mount responses against, defining the precise CD8 cell responses involved in controlling an HIV infection remains difficult. ELISPOT and ICS methods capable of measuring the global response against the virus should, however, better allow us to assess the breadth of CD8 cell responses mounted against HIV and SIV. What further complicates the issue of breadth is that the majority of CD8 T-cell responses have been identified and measured during the chronic stages of HIV infection. Results from a recent study now illustrate that CD8 cell responses against a well-defined HIV CTL epitope do not emerge until late in infection, well after resolution of acute viremia [35]. This HLA-A*0201-SL9-specific CD8 cell response was previously associated with effective control of HIV replication [36]. This suggests that this epitope, perhaps along with many other CD8 cell responses that may only arise during the chronic phase of HIV infection, is unlikely to play a role in the primary control of HIV. While these studies highlight the importance for careful characterization of CD8 cell responses, they also reveal the true complexity of discerning a correlation between CD8 cell responses and control of viremia. This finding has significant implications since the vast majority of CD8 cell responses studied to date have been identified in the chronic phase of infection. Careful re-evaluation of the CD8 cell responses to HIV at all stages of infection will therefore be critical. Specificity Although the majority of identified HIV and SIV CD8 cell epitopes are derived from structural proteins [32,33], CD8 cell responses against regulatory gene products such as Tat and Rev have been associated with a slower progression to AIDS [37]. More recently, vaccine strategies targeting these regulatory genes in primates have also shown some promise [38-40], although this is still a matter of some debate [41]. The identification of acute viral escape in SIV from a CD8 cell response to Tat [22] has further heightened interest in characterizing CD8 cell responses to these small, but potentially important, viral proteins. To this end, many CD8cell responses to HIV-encoded regulatory proteins including Tat, Rev, Vif and Vpr have now been identified [42,43]. Due to the large size of the structural proteins (Env, Gag, and Pol), and numerous regions within them that are relatively well conserved, a preponderance of CD8 cell responses directed against these proteins have been defined in the chronic phase [32,33]. As such, significant focus has traditionally been placed on T-cell responses to these structural proteins. However, it is possible that immune responses directed against regulatory and accessory proteins, produced very early after infection, will represent effective targets for HIV vaccine development [22]. Interestingly, there is enormous variability in these smaller proteins [44]. One might argue that the reason for such variability is because of the repeated impact of immune system selective pressure on these regions by effective CD8 cell responses. Unfortunately, such variability will make comprehensive analysis of responses against these non-structural HIV proteins difficult. Regardless, CD8 T-cell responses against these smaller regulatory and accessory proteins are currently under investigation, which will hopefully clarify their role in the control of HIV infections. However, only through careful analysis of responses against the autologous virus of an individual will the true extent of cellular responses to these proteins be revealed. Specific MHC class I alleles have also been associated with different rates of HIV and SIV progression [45-47]. For example, HLA-B*5701 is associated with containment of HIV replication in some long-term non-progressors (LTNP) [48]. The beneficial impact of a particular MHC class I allele may be associated with a molecule's ability to bind an unusually large number of CD8 cell epitopes because of a unique peptide binding motif [49-52]. This hypothesis would fit with observations that individuals who are homozygous at their MHC class I loci are less able to effectively control HIV infections [53]. Alternatively, this association could be related to a molecule's ability to bind particular CD8 cell epitopes that are unusually effective at controlling viral replication. The observation that particular MHC class I molecules may be associated with acute CD8 cell escape in SIV-infected rhesus macaques [22] (D.H. O'Connor, manuscript submitted) may support this latter hypothesis. Continued identification of CTL responses directed against HIV and SIV, and their restricting MHC class I molecules, will be necessary to provide a foundation to understand which CTL responses are better able to control an HIV infection. Effector function CD8 T cells control viral infections through target cell lysis using a perforin/granzyme pathway or through secretory cytokine production (i.e., IFN-γ). While most CD8 T cells possess measurable lytic function, little work has directly addressed whether differences in lytic function (i.e., the rate at which a target cell is lysed) exist between different CTL responses. Testing of CTL lines specific for different epitopes or viral proteins against recently infected CD4T-cell lines could begin to address this issue [54,55]. Alternatively, differences in the types of cytokines produced by a CD8 cell response have been observed [19]. Investigating the relationship between such cytokine expression profiles and a CD8 cell response's ability to control HIV needs to more formally addressed. Understanding the antiviral mechanism of particular CTL responses, combined with knowledge of when these CTL responses arise and to what degree they are able to induce viral escape, will prove invaluable towards the selection of CD8 cell responses that should be engendered by a vaccine. A more recent area of investigation, due in large part to the availability of tetramer and ICS technologies, is that of CD8 T-cell dysfunction. It was recently observed that, due to a lower level of perforin expression, HIV-specific CD8 T cells exhibit some degree of impaired cytolytic activity, as compared with cytomegalovirus-specific T cells from HIV chronically infected individuals [19]. In contrast, the majority of CD8 T cells specific for HIV and cytomegalovirus appear to be functionally intact in their ability to produce antiviral cytokines. This phenomenon would not have been elucidated through measurements of IFN-γ production alone. Similarly, in the setting of both HIV and SIV infections, a significant reduction in IFN-γ cytokine production was observed to occur during the chronic stage of infection [20,24]. Unfortunately, the cause for this eventual dysfunction of CD8 T cells is poorly understood but may be related to the gradual decline in CD4 T-cell responses during chronic infection. In contrast, however, a similar study did not find any evidence for lytic or cytokine production dysfunction when tetramer, ELISPOT, ICS and limiting dilution assays were used to measure HIV responses [23]. Therefore, not only is careful monitoring of the various CD8 responses important for identifying possible correlates of protection, but similar monitoring of the effector functions of such CD8 responses will be critical towards providing a more complete understanding of successful CD8 T-cell responses. Potency While numerous methods now exist to measure the breadth, frequency, and effector function of various CD8 T-cell responses, our ability to define the potential of a particular CD8 T-cell response to control HIV is still lacking. Perhaps the easiest and most direct approach to measure the impact of a CD8 cell response on HIV is through viral sequencing. The propensity for HIV and SIV to escape from CD8 cell responses is well documented [45,56-62]. A recent study examined CD8 cell escape during the acute phase of a SIV infection [23], a task more easily accomplished in non-human primates than in human subjects where early post-infection samples are difficult to obtain. Surprisingly, viral escape was detectable by 4 weeks post-infection in macaques that mounted an early CD8 cell response against an epitope in Tat. Interestingly, despite a similarly strong CD8 cell response against a well-characterized epitope in Gag, escape was only observed in the Tat epitope. These findings illustrated the dichotomy that exists in the ability of only particular CD8 cell responses to induce rapid viral escape. More extensive analysis of the viral genomes in 20 rhesus macaques early after SIV infection has now revealed that acute escape from CTL is a hallmark of SIV infection, and therefore acute escape is probably also a hallmark of acute HIV infection (D.H. O'Connor, manuscript submitted). Moreover, those epitopes observed to rapidly escape do so repeatedly in several animals sharing the same MHC class I molecule, thus demonstrating the reproducibility of the phenomenon. Furthermore, additional data suggests that acute escape responses share a common feature in demonstrating high functional avidity as measured by the ability of whole peripheral blood mononuclear cells to respond to low concentrations of peptide (D.H. O'Connor, manuscript submitted). In other viral systems, these high-avidity CTL have been shown to be more effective than low-avidity CTL in resolving infections [63,64]. More importantly, the identification of these acute escape epitopes, and the dichotomy that exists between them and those responses unable to induce rapid escape, provides an opportunity to begin to measure and compare other attributes of these responses. These include comparing viral protein targets, MHC:peptide binding affinities, T-cell receptor diversity, and T-cell effector activity. While these acute escape CD8 cell responses may represent potent immune responses, the shortcomings in their ability to control HIV and SIV infections may be due, in part, to what we now understand about the early failures of antiviral drug therapies against HIV. Drawing from lessons learned from monotherapy and bitherapy drug treatments, which were only transiently effective due to eventual drug resistance, an effective CD8 cell response may require the induction of CTL against multiple epitopes capable of inducing significant selective pressure early after infection. While this approach has in effect already been tested using whole protein vaccinations, focusing of the immune response against a few CD8 cell epitopes capable of exerting significant selective pressure may be a more appropriate test of the hypothesis. These studies could be properly addressed in the SIV-infected rhesus macaque model where CD8 cell responses restricted by high-frequency alleles have been identified and the strain (sequence) of the infecting virus can be appropriately selected. These studies suggest that sequencing of acute phase virus from HIV-infected individuals may be an effective means of identifying unique CD8 cell responses capable of exerting substantial early selective pressure [21]. While the precise impact of such CD8 cell responses on HIV and SIV is still unclear, more extensive analysis of the degree of viral escape in rapid progressors and LTNP may help address this issue. If CD8 cell responses that exert such pressure on the virus provide greater protection, then careful characterization of these acute escape CD8 cell epitopes could be important to HIV vaccine design. Two other methods are available that may also begin to address the potency of individual CD8 cell responses. The first, co-culture assays, require incubation of a CD8 cell line or clone with infected CD4 T cells [54,55]. Such an assay should delineate the abilities of particular CD8 cell responses to prevent or inhibit viral replication. These in vitro assays are expected to model, to some degree, what might be occurring naturally during HIV infection. Second, viral fitness assays assess the inherent replicative capacity of a particular viral clone to replicate [65]. Such measurements can be used to directly assess the impact that a particular CD8 T-cell-selected viral mutation has on virus replication. Unfortunately, the assays required to measure viral fitness are extremely labor intensive. Mucosal responses The intestinal gut-associated lymphoid tissue represents a major site of viral replication and CD4T-cell depletion due to the propensity of HIV and SIV to infect activated memory CD4 T cells [66]. Therefore, it is critical to understand the role that CD8 cell responses in the mucosal compartments play in controlling these infections. The importance of mucosal-homing CTL in clearing an intestinal virus has previously been demonstrated with a murine rotavirus [67]. To this end, an examination of HIV-infected patients has revealed that mucosal-specific CTL are detectable in the gut-associated lymphoid tissue of HIV-infected patients using a mucosal lymphocyte integrin marker [68]. Furthermore, in non-human primates, it has been illustrated that pre-existing mucosal SIV-specific CTL correlate with control of viral replication against an intracolonically delivered SIV challenge [69,70]. More recently, it was demonstrated that rhesus macaques receiving an intrarectal administration of a peptide-based vaccine prior to intrarectal SHIV infection eliminated virus more completely than controls in both the blood and the intestine (I.M. Belyakov, Z. Hel, B. Kelsall, et al., manuscript submitted). Together, these studies strongly support the role of mucosal CD8 cell responses in controlling an HIV infection. Various DNA prime boost vaccine regimens have demonstrated the ability to induce very strong peripheral CD8 cell responses in rhesus macaques [71,72]. Unfortunately, these levels of CD8 cell responses have been unable to adequately control against pathogenic SIV challenges [73] (T.M. Allen, manuscript submitted). However, it was recently observed that there is effective induction of mucosal homing CTL in animals vaccinated with live attenuated SIV [73], which are subsequently protected against pathogenic SIV challenges [74]. This is in sharp contrast to CD8 cell responses in those DNA/MVA-vaccinated macaques that were apparently unable to home the mucosa [73]. Finally, even more provocative evidence comes from the analysis of cervical samples from a group of highly HIV-exposed but persistently seronegative female sex workers in Nairobi [75]. Elevated levels of HIV-specific CD8 T cells were detected in the cervix of these women compared with samples from HIV-infected women. Therefore, it will be important to continue to investigate the role of mucosal CD8 responses in the control of HIV and SIV infections. CD4 T-cell correlates of immunity While many reports support the importance of maintaining CD4 T-cell responses, what exactly comprises an effective CD4 T-helper response with respect to strength, specificity, and durability remains difficult to determine. This is due, in part, to limitations in the number of characterized HIV and SIV MHC class II epitopes, identification of infected subjects with sustained CD4 T-helper responses, and the difficulty in measuring the small number of antigen-specific CD4 lymphocytes compared with the stronger CD8 cell responses. While it seems probable that strong, broad, and durable antigen-specific CD4 T-cell responses will be important in the control of viral replication, continued investigation of these responses is still necessary to begin to better define whether particular CD4 cell responses are more beneficial than others. Strength, maintenance and function The role of CD4 HTL in controlling various viral infections is one of maintenance and facilitation of viral-specific CD8 CTL responses during both acute and chronic stages of viral infections [25,76-79]. CD4 T-helper responses are, however, characteristically undetectable in chronically HIV-1-infected patients [10,80-82]. This is due to the ability of HIV to eliminate CD4T cells [83,84] and affect the normal function of these cells by limiting their ability to proliferate [25,26,85,86]. Strong HIV-specific CD4 T-cell proliferative responses have been detected primarily in infected individuals that have controlled viremia [9,10]. These studies have supported the concept that maintenance of strong CD4 cell responses is critical in the setting of an HIV infection. The protection of macaques vaccinated with live attenuated SIV against pathogenic challenge [74,87] may be due, in part, to the induction and maintenance of strong T helper 1 (Th1)-type CD4 T-helper responses by these attenuated viruses [88,89]. Fortunately, the recent testing of structured treatment interruption (STI) highly active antiretroviral therapy (HAART) therapy is now showing substantial success in enabling acutely-treated HIV-infected patients to control the infection after cessation of therapy [9]. This control of viremia is associated with substantial increases in antigen-specific CD4 T-helper responses [90], again supporting the role of strong T-helper responses in the control of viremia [10,91]. The recent measurement of a MHC class II response to a SHIV epitope restricted by Mamu-DR*w201 indicated that, unlike most MHC class I responses, the number of HTL responding to this epitope were too low to be detected with tetramers in the peripheral blood without restimulation [92]. It is possible, however, that the responses against this epitope may be lower than most typical antigen-specific HTL responses. In support of this, recent use of ICS for IFN-γ revealed the existence of significant levels of antigen-specific CD4 T-cell responses to various proteins in HIV-infected patients [16]. These studies suggest that tetramers should be able to detect responses to many other HIV-specific MHC class II-restricted responses and, therefore, remain important tools for dissecting the frequency and function of HIV-specific CD4 cell responses. Measuring CD4 cell responses to viral antigens has been routinely accomplished using proliferation assays. Unlike ELISPOT and ICS, this method measures a unique characteristic of CD4 T-cell responses, which is the marked ability to proliferate on recognition of the cognate antigen. The importance of carefully measuring both cytokine production and proliferation of CD4 T cells was recently revealed. Populations of HIV-specific CD4 T cells were identified that were unable to proliferate yet were detectable through antigen-specific IFN-γ production [25,26]. This suggests that a combination of approaches will be required to more comprehensively measure CD4 T-cell responses to HIV. Interestingly, the more sensitive ICS techniques are also now revealing that, while CD4 cell responses are indeed present in most acutely-infected patients, they subsequently decline over the course of the infection [16]. This loss of function rather than elimination of the CD4 cells opens up the possibility of restoring the function of these cells during chronic infection. Recent studies have also illustrated the impact of this loss of CD4 T-cell function on CTL function during HIV and SIV infections. While HIV-specific CTL can persist in the absence of HIV-specific CD4 T-cell help [93], the CTL's ability to produce IFN-γ can be significantly compromised [20,24]. Therefore, perhaps the greatest hurdle to HIV prevention/treatment is how to best preserve the function of these critical CD4 lymphocytes. Specificity Despite CD8 cell responses directed against virtually every HIV and SIV protein, the proliferative responses against HIV and SIV studied thus far appear to be limited to Gag, Env and Tat [10,40,88,89]. The strongest CD4 cell responses to HIV have routinely been detected against the Gag and Env structural proteins, although only responses to Gag have been found to correlate with control of viremia [10]. More recently, vaccine-induced proliferative responses to Tat were also associated with control of viremia in non-human primates [40]. However, with the introduction of ELISPOT and ICS methodologies, it is possible that additional CD4 T-helper responses will be identified in many other structural and non-structural proteins. More extensive identification of HIV-specific CD4 cell responses is still desperately needed. This might lead to a better understanding of the correlates of 'protection'. Finally, careful consideration will also need to be given to the effects of imbalances in the Th1-T helper 2 (Th2) cytokine profile that occur during HIV and SIV infections. The Th1-Th2 balance can significantly influence which arms of an immune response successfully develop after infection. Experimental data from vaccine studies in primates suggests that induction of balanced Th1-Th2 CD4 cell responses will be important to drive both the humoral and cellular arms of the immune system [94]. Therefore, careful monitoring of the impacts of cytokine shifts at various stages of disease progression will be important in determining the proper balance of cytokines required to maintain a successful immune response. Conclusions Few studies last year inspired more hope in the field of AIDS research than those illustrating the dramatic capabilities of STI HAART therapy [9,95]. These studies clearly indicate the potential for long-term control of an HIV infection in acutely infected subjects without the need for lifelong adherence to HAART therapy. More importantly, they suggest that the immune system can effectively control HIV replication. Unfortunately, very similar studies on the effects of STI in post-acute and chronically infected individuals have shown less promise [96-98]. However, detailed analysis of the immune responses in those patients who successfully control HIV infections after acute STI HAART therapy provides an excellent opportunity to characterize a successful immune response. Similarly, in most patients undergoing STI, a number of viral breakthroughs occur after removal of therapy, requiring re-initiation of HAART. Through careful analysis of the immune responses, and perhaps viral escape, in these patients, we may be able to more accurately define what changes in the immune response or virus are associated with loss of control. Such studies should begin to help us define the importance of breadth, specificity, T-cell dysfunction, and viral escape in the control of HIV. While it is important to understand the impact of the various arms of the immune system to control HIV infections, it is equally important to continue to enhance our ability to induce these responses through vaccines. Various vaccine regimens recently demonstrated control against SHIV 89.6P infections in the rhesus macaque [99,100]. While SHIV 89.6P is more sensitive to antibody-mediated neutralization than most primary isolates of HIV, these studies nonetheless illustrate the potential for vaccines to control HIV and SIV infections. Therefore, it will be necessary to continue research into the design and efficacy of vaccines capable of inducing stronger, broader, and tissue specific immune responses. Despite these recent advances in our understanding of the immune system's complex response to HIV, the potential of STI of HAART, and promising vaccine regimens [99,100], we still lack an adequate understanding of what is required to control HIV. The past year has highlighted the importance of combining various techniques to analyze both the quantitative and qualitative aspects of T-cell responses to HIV and SIV. The combination of information, with respect to the multitude of T-cell epitopes already identified and new more powerful tools for analyzing these responses, presents a unique opportunity to decipher these responses. More intensive use of ELISPOT, ICS, and whole viral genome sequencing in critical areas of investigation such as acute cellular immune responses should allow us to begin to identify those T-cell responses that might correlate with effective containment of the virus. Acknowledgment D.I.W. is an Elizabeth Glaser Scientist.