Michael Palm Basic Science, Vaccines & Prevention Project Weblog

In Vitro Study of Adenovirus Vector Effects on HIV Replication in CD4 T Cells

Another abstract from the recent Keystone meeting (see below). The data suggests that, as some researchers had hypothesized, Ad5-specific memory CD4 T cells may be rendered particularly susceptible to HIV after Ad5 stimulation. Further research will be needed to try and ascertain if these observations are relevant to the evidence of enhanced susceptibility to HIV acquisition that emerged from the STEP trial. Notably, there was no indication that Ad5-specific immunity enhanced post-infection HIV viral load among STEP participants (people with pre-existing immunity to Ad5 actually had slightly lower viral load setpoints than Ad5 seronegative participants).

In light of the concerns about adenovirus vectors raised by the STEP data, it seems a little odd that a week after the NIAID HIV Vaccine Summit, GenVec announced receipt of a $600,000 NIAID grant to develop their Ad5-based malaria vaccine candidate and Crucell – another company that has invested heavily in the adenovirus vector vaccine platform – announced the initiation of a phase I trial of an Ad26-based HIV vaccine candidate. Ad26 is far less prevalent in nature than Ad5, so few people have antibody responses to the Ad26 hexon protein (the major target for antibody responses to adenoviruses). However, studies have clearly shown that Ad-specific memory CD4 T cell responses are highly cross-reactive against different serotypes, which may call into serious question the safety of any adenovirus-based vaccine candidate in individuals at risk for HIV infection (this may even hold true for vectors based on chimpanzee adenoviruses, see second abstract below). As the data stands currently, if GenVec successfully developed an Ad5-based malaria vaccine, it would be contraindicated in anyone at risk for HIV infection with pre-existing immunity to the vector; this would represent a substantial portion of the population that stands to benefit from immunization against malaria. And for the Crucell construct, it is also difficult to imagine a scenario where the vaccine could be ethically studied in individuals at risk for HIV infection, unless it can be proven that the risk of enhancement seen in STEP related entirely to immune responses specific to Ad5 and not immune responses cross-reactive with other adenovirus serotypes (or some other immune-modulating effect of adenoviruses generally). The comments of Crucell’s Jaap Goudsmit in the company’s press release seem not only premature, but bordering on delusional: “We are excited about the first in man study of this newly developed vector, that could provide a solution to the issues that raised from previous HIV vaccine trials.”

Keystone Symposia: HIV Vaccines: Progress and Prospects
Banff, Alberta, Canada, March 27-April 1, 2008

Abstract# 328

HIV-1 replication in CD4 T cells from Adenovirus experienced and Naïve Donors

A. Gregory Spies, Jason Stucky, John McNevin and Juliana McElrath.
Vaccine and Infectious Disease Research Institute, Fred Hutchinson Cancer Research Center, Seattle, Washington 98105.

Preliminary analysis of the Merck STEP Study suggests that there is a higher rate of HIV infection in vacinees who had positive human adenovirus serotype-5 titers on enrollment compared to those with undetectable Ad5 titers. Therefore, we have utilized a human in vitro T cell priming assay to determine whether prior infection by Ad5 affects replication of HIV-1 in human CD4 T cells following exposure to a replication-defective Ad5 vector. Immature dendritic cells were generated using GM-CSF/IL-4 treatment of CD14 positive cells. Immature DC were exposed to an empty Ad5 vector resulting in increased levels of HLA-Class I, HLA-DR, CCR7, CD40, CD80 and CD83. CD4 T cells were primed with mock or Ad5-stimulated DCs and subsequently infected with the CCR5-tropic HIV-1 strain, YU2. HIV-1 replication was monitored by p24 ELISA of supernatants from DC-primed CD4 T cell cultures. Exposure of DC to Ad5 resulted in increased HIV growth in CD4 T cells from both Ad5-naive and Ad-experienced donors. At day 6, following addition of T cells, the Ad5-experienced CD4 T cells were found to produce twice the p24 as the Ad5-naive CD4 T cells. However FACS analysis of these cells using CD38, Ki67, Bcl-2, CCR5, CCR7 and CD-27 showed no phenotypic difference.

Abstract #223

T-cell responses in healthy adults to replication deficient AdHu5 vector

Natalie Hutnick, Hildegund Ertl, Michael R. Betts.
Microbiology Dept, University of Pennsylvania, Philadelphia PA, USA, 19104

A current strategy for an HIV vaccine employs the use of a replication deficient AdHu5 vector. However, it has been shown that pre-existing antibodies against AdHu5 reduce vaccine efficacy. To avoid the problems associated with pre-existing AdHu5 humoral immunity, vectors based on 2 serologically distinct chimpanzee adenoviruses (AdC6 and AdC7) have been developed.. While this strategy may avoid issues with pre-existing humoral immunity, it remains possible that cross-reactive Ad-specific CD8+ T cells could similarly limit vaccine efficacy. Here, we examined the response to replication defective AdHu5, AdC6 and AdC7 vectors in 10 healthy subjects from North America not expected to have been exposed previously to chimp-derived Ad strains. Eight of the ten subjects exhibited a polyfunctional CD8 T cell response to AdHu5 as measured by IL-2, TNF-a, IFN-g and perforin expression. The responding cells were a mixed population composed of CD27- CD45RO+ effector memory and CD27+CD45RO+ central memory phenotype. The 8 subjects who responded to AdHu5 also exhibited a cross reactive response to AdC6, and three responded to AdC7, though at a lower frequency. We further examined Ad-specific T cell responses in one subject who participated in a phase I AdHu5 HIV vaccine safety trial. A substantial expansion of polyfunctional AdHu5-specific CD8+ (1.06%) and CD4+ (0.57%) T cells were noted approximately 4 months after the final inoculation. These cells also exhibited cross-reactivity with both AdC6 and AdC7. These data indicate that Adenovirus-specific T cell responses are common in humans, and that shared elements between disparate Ad strains can be serve as targets for Ad-specific T cells. Vaccination with replication defective AdHu5 vectors also appears to enhance the frequency of Ad specific T-cells.

HLA Class II Associations with Viral Load Control

Last week’s NIAID vaccine summit featured much handwringing pessimism regarding the potential for adaptive immunity to control HIV replication, leading Guiseppe Pantaleo to remind attendees about the strong and consistent association between the class I HLA allele B*57 and elite control/long term non progression. Class II HLA alleles, which influence CD4 as opposed to CD8 T cell responses, have been less well studied, although some associations with slowed disease progression and resistance to HIV infection have been reported. At the recent Keystone HIV pathogenesis conference in Banff, Rachel Owen from UCSF presented new data indicating that the class II HLA allele DRB1*13 is significantly overrepresented among individuals who maintain viral loads in the absence of therapy, suggesting that CD4 T cell responses contribute to the phenomenon. The findings echo a recently published report on class II alleles and elite control in SIV-infected macaques (see second abstract, below).

As a side note, the Keystone HIV pathogenesis conference occurs in parallel with an HIV vaccine meeting and has done so for many years. However, next year the titles of the two meetings will be changed to “Prevention of HIV/AIDS” and “HIV Immunobiology: From Infection to Immune Control.”

Keystone Symposia: HIV Pathogenesis (X8), March 27 - April 1, 2008

HLA Class II associations in HIV infection: Controllers versus Non-controllers

Rachel E. Owen1, 2, Elizabeth Sinclair3, C. Lorrie Epling3, Jeffrey N. Martin3, Steven G. Deeks3, Philip J. Norris1, 2,3.

1Blood Systems Research Institute, 270 Masonic Avenue, San Francisco, CA 94118, USA, 2Department of Laboratory Medicine and 3Department of Medicine, University of California San Francisco, San Francisco, CA 94143, USA.

Strong CD4+ and CD8+ T cell responses are detected in individuals who control viraemia, suggesting but not proving that these cells may be causally related to virus control. The consistent association between certain class I alleles (e.g., B5701) and virus control provides strong evidence that CD8 T cells are able to effectively exert control. However, the role of CD4+ T cells in controlling HIV infection is not fully understood, and few HLA class II gene associations have been made.

We investigated HLA class II gene associations in chronically infected HIV+ individuals from the UCSF SCOPE cohort and categorized individuals as: (1) “controllers”: defined as plasma HIV RNA levels <10,000 copies/ml (in absence of therapy, n= 38); and (2) “non-controllers”: defined as plasma HIV RNA levels >10,000 copies/ml (n=290).

We found an increased frequency of the HLA-DRB1*13 allele in the controllers compared to the non-controllers (42.1% vs. 24.1%, p= 0.029), consistent with previous studies. Some DRB1*13 haplotypes were more common in controllers than non-controllers (p= <0.05 for each pair-wise comparison): DRB1*13 plus DQB1*03 (24% vs. 10%), DQB1*04 (8% vs. 0.7%), DQB1*06 (37% vs. 18%), DRB3*03 (13% vs. 2%) and DRB4*01 (21% vs. 8%). Controllers had higher CD4+ T cell IFN and IL-2 responses than non-controllers, measured by intracellular cytokine staining, following gag stimulation (mean IFN response 0.7% vs. 0.3%; mean IL-2 response 0.3% vs. 0.1%). No difference in the CD4+ IFN or IL-2 response following env or pol stimulation was measured. A small number of DRB1*13 expressing controllers had a trend towards higher CD4+ IFN and IL-2 responses following gag stimulation, when compared to controllers not expressing the DRB1*13 allele or non-controllers.

The epidemiological link between class II alleles and virus control is consistent with observations made with class I alleles, and argues that antigen-specific CD4+ T cells are playing a role in control of viraemia. These responses would be important to study in the pathogenesis of acute HIV infection and in HIV vaccine trials.

This work was supported by the National Institutes of Health to the UCSF-GIVI Center for AIDS Research, 5P30AI027763, and to the UCSF Clinical and Translational Sciences Institute, U54RR023566.

J Virol. 2008 Jan;82(2):859-70. Epub 2007 Nov 7.

The major histocompatibility complex class II alleles Mamu-DRB1*1003 and -DRB1*0306 are enriched in a cohort of simian immunodeficiency virus-infected rhesus macaque elite controllers.

Giraldo-Vela JP, Rudersdorf R, Chung C, Qi Y, Wallace LT, Bimber B, Borchardt GJ, Fisk DL, Glidden CE, Loffredo JT, Piaskowski SM, Furlott JR, Morales-Martinez JP, Wilson NA, Rehrauer WM, Lifson JD, Carrington M, Watkins DI.

Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, 555 Science Dr., Madison, WI 53711, USA.

The role of CD4(+) T cells in the control of human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) replication is not well understood. Even though strong HIV- and SIV-specific CD4(+) T-cell responses have been detected in individuals that control viral replication, major histocompatibility complex class II (MHC-II) molecules have not been definitively linked with slow disease progression. In a cohort of 196 SIVmac239-infected Indian rhesus macaques, a group of macaques controlled viral replication to less than 1,000 viral RNA copies/ml. These elite controllers (ECs) mounted a broad SIV-specific CD4(+) T-cell response. Here, we describe five macaque MHC-II alleles (Mamu-DRB*w606, -DRB*w2104, -DRB1*0306, -DRB1*1003, and -DPB1*06) that restricted six SIV-specific CD4(+) T-cell epitopes in ECs and report the first association between specific MHC-II alleles and elite control. Interestingly, the macaque MHC-II alleles, Mamu-DRB1*1003 and -DRB1*0306, were enriched in this EC group (P values of 0.02 and 0.05, respectively). Additionally, Mamu-B*17-positive SIV-infected rhesus macaques that also expressed these two MHC-II alleles had significantly lower viral loads than Mamu-B*17-positive animals that did not express Mamu-DRB1*1003 and -DRB1*0306 (P value of <0.0001). The study of MHC-II alleles in macaques that control viral replication could improve our understanding of the role of CD4(+) T cells in suppressing HIV/SIV replication and further our understanding of HIV vaccine design.

A Surreal, Deeply Troubling Week for AIDS Vaccine Research

At the prompting of several basic scientists, NIAID held a “vaccine summit” this past Tuesday. The event, which from my perspective had all the rationality and conviviality of a hastily-convened Soviet era show trial, was webcast for posterity (the archived version can be seen online). Essentially, frustration with the dismal, unacceptably low NIAID payline for investigator-initiated grants appears to have caused a number of basic researchers to see the failure of Merck’s HIV vaccine candidate as an appropriate latch on which to hang their argument that money should be directed away from human trials of other experimental HIV vaccine candidates and into basic research and discovery.

There is certainly logic to the argument, despite the clumsy and alienating manner in which it has been advanced. The STEP trial has at the least shown that T cell-mediated control of HIV replication will not easily be achieved, and there is a strong case for rigorously reviewing the current vaccine pipeline and trimming it down to only candidates that might be able to address the potential shortcomings of the Merck construct (particularly in terms of HIV-specific T cell breadth, proliferative capacity, functionality, CD4 help and - of course - vector safety). In the longer term, discovery will also be vitally important to develop neutralizing antibody-based candidates (or other novel approaches) with the potential to completely protect against HIV acquisition.

However, this type of discussion is not what primarily emerged from the summit. Instead, panelists such as virologist and summit co-chair Warner Greene used the opportunity as a bully pulpit to make outrageous and offensive claims about the state of the HIV vaccine field and the work of their colleagues. Even cellular immunologist Rafi Ahmed offered a bizarre revisionist history of his thinking on T cell-based candidates: he claimed he had “cringed” when hearing of Merck’s candidate - an odd assertion given that he is on the immunology advisory committee of the HVTN, which conducted the STEP study – and that vaccines that focus on one arm of the immune system are “doomed to failure.” When I interviewed him a little over a year ago he said he was "optimistic" about the potential for T cell-based approaches to reduce post-infection viral load: “I’m a big believer in memory T cells bringing viral load down because that’s what we keep seeing. That doesn’t mean it will hold up in HIV but at least in all the models we study if you have memory T cells - even without any antibody - the initial burst is less and it’s pulled down much faster. And then the HIV data suggests that if that happens the prognosis is better, so that’s where my optimism comes from." Ahmed also stated at the summit that “immunology 101” says antibodies are required for control of HIV replication, yet studies of elite controllers offer little evidence for this claim.

Predictably, the media coverage surrounding the event amplified these soundbites across the world. Researchers in places such as South Africa, where skepticism about AIDS research runs high and AIDS denial has been politically embraced, will be dealing with the fall-out for years to come. Even elsewhere and here in the US, there has been a growth of anti-research and anti-science movements which will feed off these false, widely-publicized statements that no progress has occurred in AIDS vaccine research.

The misinformation actually started on Friday March 21, when a David Brown article in the Washington Post offered a series of misrepresentations about the Merck vaccine in the guise of facts. The deck quote was from Bob Gallo who, in a truly woeful bout of self-serving hyperbole, compared the STEP trial results to the Space Shuttle Challenger Disaster. Brown went on to misleadingly suggest that the strong trend toward enhancement seen in a subset of the STEP study participants was somehow connected to the T cell-based vaccine strategy; he does so by saying that the vaccine “primed” people to be more susceptible, and then stating that immune activation – one of the factors that has been suggested as a potential explanation for the adverse events – has “never been” associated with increased susceptibility to HIV infection. This is simply false; studies have shown that several “immune–activating infections” (to use Brown’s words) such as CMV and HSV-2 are associated with an increased susceptibility to HIV acquisition. Peripheral blood cells from highly immune-activated individuals also show increased susceptibility to HIV infection. In addition, if the enhancement effect had been related to the HIV-specific T cell responses induced by the vaccine, the effect would have been worst for the participants with no pre-existing immunity to the Ad5 vector (who developed the highest magnitude HIV-specific responses), but the opposite was the case (at the vaccine summit, a CDC researcher noted that the multivariate analyses showed that these vaccinated individuals were actually at a slightly but non-significantly reduced risk of HIV infection compared to the other study participants, with a point estimate of 0.6-0.8). In the same piece, Brown also attempts to indict researchers for not testing the Merck vector in a low-dose SIV challenge model as if this was some kind of oversight, completely glossing over the many complexities and controversies that attend these models and not mentioning that David Watkins did in fact use the Merck vector in one of his repeated low dose SIV challenge experiments.

Brown’s follow-up article from the March 25 summit compounded and expanded upon the misinformation offered in the first, claiming that the event was "tantamount to an admission that almost no progress has been made in the search for an AIDS vaccine in the past 25 years." This must have come as an unwelcome surprise to the researchers and study participants who have been slogging away for years, slowly improving the scientific understanding of both the immunological and immunogenetic underpinnings of long-term non-progression and resistance to HIV infection in highly exposed seronegative individuals, not to mention those involved in trying to iteratively improve the immunogenicity of experimental vaccine constructs. There has certainly been waste, duplication, and a sunny over-simplification of the complexities and unknowns that still face the HIV vaccine field from some quarters but these problems are not best addressed by trashing the entire endeavor and everyone involved in it. The statements quoted in Brown’s piece from Warner Greene and James Hoxie are de facto accusations that researchers who followed the strong scientific evidence that T cells play a key role in controlling HIV replication and used that evidence to inform vaccine design have led the field in the wrong direction and brought us no closer to an AIDS vaccine; in addition to being horrifically hostile and unfair, these claims are simply untrue.

Amidst all the sturm and drang at the summit, few people seemed to pick up on the comments of immunologist Mark Connors. Over the years, Connors has studied HIV-specific immune responses and authored many skeptical articles questioning some of the correlations with control of HIV replication that have been reported in the literature. In 2002, however, he identified HIV-specific CD8 T cell proliferation as a potentially important correlate, a finding several other research groups have now confirmed. At the summit, he cited additional assays his lab is working on, and expressed confidence that robust and broadly applicable correlates of immunological control are within striking distance. Furthermore, he also cited the fact that several Merck vaccine recipients who became infected in the STEP trial and carry the favorable immune response gene HLA B*57 are controlling their viral loads to undetectable levels. Initially, this was assumed to be an HLA B*57 effect but Connors argued that the frequency of the occurrence is far higher than would be anticipated in the absence of immunization, suggesting that it reflects an interaction between the vaccine and the favorable HLA allele akin to that seen in Merck’s now notorious SIV challenge studies, in which only immunized macaques bearing the favorable Mamu A*01 allele showed significant viral load reductions. The implication is that while the Merck vaccine was far from optimal, it may have been able to enhance the HLA B*57 effect. If this finding actually holds up, it may offer additional validation of the SIV/macaque model and it would also strengthen the argument for developing improved immunogens with the potential to achieve this outcome in people lacking HLA B*57.

The AIDS Vaccine Advocacy Coalition and the National Association of People with AIDS both issued concerned statements about the summit and the media coverage it engendered. Exactly where things go from here remains to be seen.

CORRECTION: The first version of this posting incorrectly stated that the class I Mamu A*01 allele in macaques is the equivalent to HLA B*57 in humans; thanks to David Watkins for alerting me to the fact that this is not the case, and that the likely analog is Mamu B*17. Evidence also suggests that Mamu B*08 is equivalent to HLA B*27, another HLA allele associated with control of HIV replication in humans.

Understanding Control of HIV Replication in the Absence of Therapy

Individuals who maintain control of HIV replication in the absence of any therapy may offer important clues to aid the development of both vaccines and novel immune-based therapies. Many research groups are studying such individuals, with the largest and most ambitious research project being the elite controller study helmed by Bruce Walker from Partners AIDS Research Center in Boston. In a recent issue of the Journal of Infectious Diseases, lead investigator Florence Pereyra and colleagues report preliminary data from 126 individuals who have enrolled into the study. Participants are divided into two categories: elite controllers, defined as those with viral load consistently below 50 copies/mL (the limit of detection for the viral load test) and viremic controllers, whose viral loads are consistently below 2000 copies/mL. The JID paper contains data on 66 individuals in the former category and 60 in the latter; 30 individuals with chronic, progressive infection recruited over the same time period are also included for comparative purposes.

The major finding is that controllers (both elite and viremic) are a heterogeneous group with regard to many of the factors that previous studies have associated with viral load control. In terms of host genetics, HLA B*57 alleles were significantly overrepresented among controllers, but the researchers emphasize that the frequency of HLA B*57 was considerably lower than reported previously in smaller cohorts. When the analysis was expanded to any HLA allele that has been associated with a favorable prognosis in HIV infection, 68% of elite controllers and 60% of viremic controllers were found to carry at least one such allele, compared with 37% of the chronic progressors, “leaving almost one-third of HIV controllers without any known relatively protective HLA alleles.” CCR5 and CCR2 polymorphisms that have been associated with slowed disease progression were not overrepresented among controllers. Absolute CD4 counts were different between elite and viremic controllers (884 vs. 602 cells), in line with a recent paper showing that CD4 declines can still occur – albeit very slowly - despite low or even undetectable viral loads.

Randomly selected subsets from each group of study participants were further evaluated for HIV-specific immune responses. In terms of CD8 T cell responses (as measured by interferon gamma ELISpot), elite controllers displayed the lowest breadth and magnitude of HIV-specific CD8 T cells. Median breadth was 15, 19 and 27 epitopes among elite controllers, viremic controllers and chronic progressors respectively while average magnitude was 5428, 6253 and 8300 spot-forming cells (SFC) per million PBMC (peripheral blood mononuclear cells). The simplistic interpretation of these data is that the CD8 T cell response cannot be important, but the broader, larger responses associated with higher viral loads are to be expected due to the persistent recruitment from the naïve CD8 T cell pool that can occur in the setting of chronic infection, and because interferon gamma production is the last function to be lost by exhausted cells (so a large, broad response does not necessarily equate to an effective response). When the researchers looked at IL-2 and interferon gamma production together, responses were significantly higher in elite controllers than viremic controllers or progressors (both for HIV-specific CD4 T cells and CD8 T cells), but there was still heterogeneity in that these responses were not detectable in some individuals. Interestingly, elite controllers had the highest ratio of IL-2+interferon-gamma-producing HIV-specific CD4 T cells to HIV-specific CD8 T cells while chronic progressors had the lowest, suggesting – as have many prior studies - that CD4 T cell help plays an important, perhaps undervalued, role in the immune response to HIV. Adding to the heterogeneity, elite controllers also displayed the widest range of CD8 T cell interferon gamma ELISpot responses in terms of both breadth and magnitude, with responses ranging from 2-101 epitopes in breadth and from less than 500 SFC to over 25,000 SFC in magnitude. Despite the range of responses, the vast majority of both elite and viremic controllers displayed responses to many more CD8 T cell epitopes than have been reported to date in recipients of T cell-based HIV vaccine candidates (the average response in the recent Merck trial was to three epitopes).

Consistent with previously published studies, the HIV Gag protein was preferentially targeted among controllers. In progressors, Gag, Pol and Nef were targeted equally and more targeting of Env was observed than in controllers. The researchers also looked at neutralizing antibody responses. Plasma samples from both viremic controllers and progressors showed neutralizing activity against viruses sampled from the same groups as well as against laboratory HIV strains. There were no differences in neutralizing antibody titers between these two groups. Elite controllers had significantly lower neutralizing antibody responses overall but, again, heterogeneity was seen: some individuals had broad neutralizing antibody responses whereas others had very low or undetectable responses.

In concluding, the authors state that: “although elite and viremic controllers share some immunologic features and are distinct from persons with chronic progressive infection, the elite controllers are a distinct subgroup. However, even within these groups there is substantial heterogeneity in all of the parameters studied, which suggests that there are as-yet-undefined viral or host factors or combinations of factors that contribute to this remarkable phenotype.” They also cite the elite controller study linked to above and their hope that this large collaborative effort will better define the pathways that lead to durable control of HIV replication.

Two other recent studies offer additional perspectives on the same topic. In the first, Shiv Ghandi and colleagues show that the antiretroviral activity of APOBEC proteins does not appear to explain elite control of HIV replication. In the second, recently presented at CROI, researchers from Barbara Shacklett’s lab at UC Davis demonstrate that controllers have significantly more polyfunctional HIV-specific CD4 and CD8 T cells in their rectal mucosa, a novel finding indicating that analyses of additional compartments beyond the blood may add to the understanding of the elite controller phenomenon.

The Journal of Infectious Diseases 2008;197:563–571
DOI: 10.1086/526786
MAJOR ARTICLE

Genetic and Immunologic Heterogeneity among Persons Who Control HIV Infection in the Absence of Therapy

Florencia Pereyra,1,2 Marylyn M. Addo,1 Daniel E. Kaufmann,1 Yang Liu,5 Toshiyuki Miura,1 Almas Rathod,1 Brett Baker,1 Alicja Trocha,1,4 Rachel Rosenberg,1 Elizabeth Mackey,1 Peggy Ueda,1 Zhigang Lu,1 Daniel Cohen,3 Terri Wrin,5 Christos J. Petropoulos,5 Eric S. Rosenberg,1 and Bruce D. Walker1,4

1Partners AIDS Research Center, Massachusetts General Hospital and Division of AIDS, Harvard Medical School, 2Brigham and Women’s Hospital, Division of Infectious Diseases, and 3Fenway Community Health Care Center, Boston, Massachusetts; 4Howard Hughes Medical Institute, Chevy Chase, Maryland; and 5Monogram Biosciences, South San Francisco, California

Background: Spontaneous control of human immunodeficiency virus (HIV) infection has been documented in a minority of HIV-infected individuals. The mechanisms behind this outcome remain largely unknown, and a better understanding of them will likely influence future vaccine strategies. Methods: HIV-specific T cell and antibody responses as well as host genetics were examined in untreated HIV-infected patients who maintain comparatively low plasma HIV RNA levels (hereafter, controllers), including those with levels of <50 RNA copies/mL (elite controllers, n=64), those with levels of 50–2000 copies/mL (viremic controllers, n=60); we also examined HIV-specific T cell and antibody responses as well as host genetics for patients with levels of >10,000 copies/mL (chronic progressors, n=30). Results: CD8+ T cells from both controller groups preferentially target Gag over other proteins in the context of diverse HLA class I alleles, whereas responses are more broadly distributed in persons with progressive infection. Elite controllers represent a distinct group of individuals who have significantly more CD4 and CD8 T cells that secrete interferon-γ and interleukin-2 and lower levels of HIV-neutralizing antibodies. Individual responses were quite heterogeneous, and none of the parameters evaluated was uniquely associated with the ability to control viremia. Conclusions: Elite controllers are a distinct group, even when compared to persons with low level viremia, but they exhibit marked genetic and immunologic heterogeneity. Even low-level viremia among HIV controllers was associated with measurable T cell dysfunction, which has implications for current prophylactic vaccine strategies.

Journal of Virology, March 2008, p. 3125-3130, Vol. 82, No. 6
doi:10.1128/JVI.01533-07

Role of APOBEC3G/F-Mediated Hypermutation in the Control of Human Immunodeficiency Virus Type 1 in Elite Suppressors

Shiv K. Gandhi,1 Janet D. Siliciano,1 Justin R. Bailey,1 Robert F. Siliciano,1,2 and Joel N. Blankson1

Department of Medicine, Johns Hopkins University School of Medicine,1 Howard Hughes Medical Institute, Baltimore, Maryland 212052

While many studies show that the APOBEC3 family of cytidine deaminases can inhibit human immunodeficiency virus type 1 (HIV-1) replication, the clinical significance of this host defense mechanism is unclear. Elite suppressors are HIV-1-infected individuals who maintain viral loads below 50 copies/ml without antiretroviral therapy. To determine the role of APOBEC3G/F proteins in the control of viremia in these patients, we used a novel assay to measure the frequency of hypermutated proviral genomes. In most elite suppressors, the frequency was not significantly different than that observed in patients on highly active antiretroviral therapy. Thus, enhanced APOBEC3 activity alone cannot explain the ability of elite suppressors to control viremia.

15th Conference on Retroviruses & Opportunistic Infections
Abstract #355

Polyfunctional HIV-specific T Cells in Rectal Mucosa of HIV Controllers

April Ferre*1, P Hunt2, D Young1, J Garcia1, H Yee2, R Pollard1, S Deeks2, and B Shacklett1

1Univ of California, Davis, US and 2Univ of California, San Francisco, US

Background: Among the HIV-infected population exists a unique group of individuals who achieve control over HIV replication in the absence of ART. The study of such individuals is crucial to understanding how the immune system may effectively control viral replication and limit progression to AIDS. Methods: In this study we examined cell-mediated immune responses in 26 HIV controllers (viral load <2000 copies/mL), 14 non-controllers (viral load >10,000 copies/mL), and 10 HAART-suppressed individuals (viral load <50 copies/mL) in peripheral blood mononuclear cells (PBMC) and rectal mucosa. Of the controllers, 14 were classified as elite controllers, individuals who maintain plasma viral load <75 copies/mL in the absence of ART. This group is believed to encompass <1% of all HIV-infected individuals. In intracellular cytokine assays, we measured the production of 3 cytokines (interferon-γ, interleukin-2, tumor necrosis factor-α), 1 chemokine (MIP-1β), and the cytolytic granule marker CD107 in response to stimulation by HIV-1 Gag peptides. We hypothesized that “polyfunctional” T cells, capable of producing multiple antiviral factors, are critical in limiting viral replication and disease progression. Results: Mucosal CD8 T cell responses in controllers were significantly stronger and more complex than those in HAART-suppressed individuals (p = 0.0006). Differences between controllers and non-controllers were more subtle, but included a higher frequency of 4-function HIV-specific CD8 T cells in rectal mucosa of controllers than in non-controllers (p = 0.002). CD4 T cell responses were less complex and of lower magnitude than CD8 responses, but several controllers had unusually strong, polyfunctional mucosal CD4 responses. Conclusions: These findings demonstrate that many controllers mount strong and complex HIV-specific T cell responses in mucosal tissues. These polyfunctional cells may play an important role in immune surveillance of gut mucosa, as suggested by their relative enrichment among individuals who appear to be controlling HIV replication in absence of therapy.

Souping Up Therapeutic Vaccines

Two new papers in the advance online section of the Journal of Experimental Medicine report that blocking immunosuppressive signals can give therapeutic immunization a boost. The studies follow on from work covered on the blog previously in which the IL-10/IL-10 receptor and PD-1/PD-L1 signaling pathways were identified as key contributors to viral persistence in the mouse model of chronic infection with lymphocytic choriomeningitis virus (LCMV).

In the first paper, David Brooks and colleagues from Mike Oldstone’s laboratory at Scripps show that combining DNA vaccination with an antibody blocking the IL-10 receptor leads to improved LCMV-specific CD4 and CD8 T cell function and accelerated clearance of the virus from chronically infected mice. In the second paper, Sang-Jun Ha and a team from Rafi Ahmed’s lab at Emory University report that immunization with a vaccinia vector encoding a single CD8 T cell epitope from LCMV also accelerates viral clearance when combined with an antibody blocking PD-L1 (the ligand for the molecule PD-1). The vaccine had no effect when given alone. In addition, these researchers demonstrate an effect of the vaccine plus PD-L1 blockade in mice transiently depleted of CD4 T cells, suggesting that the approach might have an impact in settings of CD4 T cell deficiency.

Both groups conclude by citing each other’s work and noting that, taken together, the results indicate that immunosuppressive signaling in chronic infection may explain the disappointing results of most therapeutic vaccination studies conducted to date. The authors also argue that their data point toward new strategies for boosting the impact of therapeutic immunization. In the hopeful closing words of Brooks et al: “a similar dual immunotherapeutic adjuvant/vaccination strategy that alleviates immunosuppressive signals while boosting immunity may stimulate effective, long-term, immune-mediated control of persistent viral infections in humans, and may also be an adjuvant for primary vaccination.”

The Journal of Experimental Medicine
Published online 10 March 2008
doi:10.1084/jem.20071948

BRIEF DEFINITIVE REPORT

IL-10 blockade facilitates DNA vaccine-induced T cell responses and enhances clearance of persistent virus infection

David G. Brooks1, Andrew M. Lee1, Heidi Elsaesser1, Dorian B. McGavern1, and Michael B.A. Oldstone1,2

1 Viral-Immunobiology Laboratory, Department of Molecular and Integrative Neuroscience, and 2 Department of Infectology, The Scripps Research Institute, La Jolla, CA 92037

Therapeutic vaccination is a potentially powerful strategy to establish immune control and eradicate persistent viral infections. Large and multifunctional antiviral T cell responses are associated with control of viral persistence; however, for reasons that were mostly unclear, current therapeutic vaccination approaches to restore T cell immunity and control viral infection have been ineffective. Herein, we confirmed that neutralization of the immunosuppressive factor interleukin (IL)-10 stimulated T cell responses and improved control of established persistent lymphocytic choriomeningitis virus (LCMV) infection. Importantly, blockade of IL-10 also allowed an otherwise ineffective therapeutic DNA vaccine to further stimulate antiviral immunity, thereby increasing T cell responses and enhancing clearance of persistent LCMV replication. We therefore propose that a reason that current therapeutic vaccination strategies fail to resurrect/sustain T cell responses is because they do not alleviate the immunosuppressive environment. Consequently, blocking key suppressive factors could render ineffective vaccines more efficient at improving T cell immunity, and thereby allow immune-mediated control of persistent viral infection.

The Journal of Experimental Medicine
Published online 10 March 2008
doi:10.1084/jem.20071949

ARTICLE

Enhancing therapeutic vaccination by blocking PD-1–mediated inhibitory signals during chronic infection

Sang-Jun Ha1, Scott N. Mueller1, E. John Wherry1, Daniel L. Barber1, Rachael D. Aubert1, Arlene H. Sharpe2, Gordon J. Freeman3, and Rafi Ahmed1

1 Emory Vaccine Center and Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322
2 Department of Pathology, Harvard Medical School and Brigham and Women's Hospital, 3 Department of Medical Oncology, Dana-Farber Cancer Institute, Department of Medicine, Harvard Medical School, Boston, MA 02115

Therapeutic vaccination is a potentially promising strategy to enhance T cell immunity and viral control in chronically infected individuals. However, therapeutic vaccination approaches have fallen short of expectations, and effective boosting of antiviral T cell responses has not always been observed. One of the principal reasons for the limited success of therapeutic vaccination is that virus-specific T cells become functionally exhausted during chronic infections. We now provide a novel strategy for enhancing the efficacy of therapeutic vaccines. In this study, we show that blocking programmed death (PD)-1/PD-L1 inhibitory signals on exhausted CD8+ T cells, in combination with therapeutic vaccination, synergistically enhances functional CD8+ T cell responses and improves viral control in mice chronically infected with lymphocytic choriomeningitis virus. This combinatorial therapeutic vaccination was effective even in the absence of CD4+ T cell help. Thus, our study defines a potent new approach to augment the efficacy of therapeutic vaccination by blocking negative signals. Such an approach may have broad applications in developing treatment strategies for chronic infections in general, and perhaps also for tumors.

Evanescing Th17 CD4 T Cells & HIV Pathogenesis

A slew of new studies presented at CROI revealed that a recently discovered subset of CD4 T cells – dubbed T-helper type 17 or Th17 cells for short – appear preferentially depleted in HIV and pathogenic SIV infections. Th17 cells were identified as a separate lineage of CD4 T cells in 2005; they are characterized by the production of the cytokines IL-17 and IL-22 and have been shown to be involved in both pro-inflammatory autoimmune diseases, and in protection against certain pathogens such as Klebsiella pneumonia and Salmonella typherium. The cytokines IL-17 and IL-22 also play a role in protecting mucosal sites from pathogens and maintaining the integrity of the mucosal barrier. Prior studies conducted outside of the HIV field indicate that Th17 cells make up less than 1% of circulating CD4 T cells in healthy people with slightly higher levels (~1-2%) found in the gut. The cytokine IL-23 appears involved in the development of Th17 CD4 T cells, which arise – like other Th subsets – from the activation of naïve CD4 T cells. The development of Th17 cells is also suppressed under certain conditions, particularly by the presence of type I & II interferons (e.g. alpha interferon and interferon gamma).

At CROI, Mirka Paiardini presented data showing that Th17 cell levels were similar or slightly increased in the blood of HIV-infected individuals compared to uninfected controls. But in the gut, the percentage of CD4 T cells producing IL-17 after in vitro stimulation (with the pan-T cell stimulating agents PMA and ionomycin or anti-CD3) was significantly lower in people with HIV and antiretroviral therapy did not fully restore the proportion of Th17 cells to those seen in uninfected individuals. Paiardini also showed that there was a significant inverse correlation between immune activation levels in both gut and peripheral blood (as assessed by the proportion of CD4 T cells expressing the proliferation marker Ki67) and the proportion of Th17 cells in the gut.

It has recently been reported that non-pathogenic SIV infection of sooty mangabeys is associated with gut CD4 T cell loss, so Paiardini measured Th17 levels in the gut of mangabeys to ascertain if they were also depleted, as they appear to be in HIV-infected humans. He reported that the proportion of gut Th17 cells remained stable in sooty mangabeys despite the overall reduction in the percentage of gut CD4 T cells. Paiardini hypothesized that preservation of gut Th17 cells in mangabeys preserves the integrity of the mucosal barrier and thus prevents translocation of microbial products and systemic immune activation, supporting this hypothesis with slides demonstrating the health of the epithelial barrier in SIV-infected animals. However it is perhaps worth noting that – as is so often the case with research results – cause and effect are not necessarily easy to tease apart. Immune activation in HIV infection is associated with elevated levels of type I interferons, so if Th17 cells are generated dynamically in the gut, immune activation would be expected to inhibit their differentiation and thus reduce the proportion of Th17 cells. In sooty mangabeys, where SIV infection does not result in persistent immune activation, Mark Feinberg has shown that type I interferons are not upregulated, suggesting that preservation of Th17 cells could be a consequence rather than a cause of the observed lack of activation.

Valentina Cecchinato from Genoveffa Franchini’s lab at the National Cancer Institute sounded a similar theme to Paiardini, showing that in a study of SIV-infected rhesus macaques, the percentage of Th17 cells was reduced in the gut while the proportion of Th1 (interferon gamma producing) CD4 T cells was unaffected. In contrast, both populations were depleted from the peripheral blood. Further analyses revealed an inverse correlation between SIV viral load and percentages of Th17 cells in the colon and rectum. But when Cecchinato looked for evidence of microbial translocation by measuring blood levels of lipopolysaccharides (LPS), there was no correlation between LPS levels and the proportion of Th17 cells in the gut (or anywhere else). This finding may be explained by the technical challenges associated with LPS measurement, but it could also be an indication that Th17 loss is an effect rather than a cause of virus-driven immune activation in these animals.

In the last of three talks on the topic, David Favre from UCSF showed that the proportion of Th17 cells also declines at mucosal sites in SIV-infected pigtailed macaques. Favre focused on the balance between T regulatory (Treg) cells (another CD4 T cell subset with the ability to suppress immune responses) and Th17 cells, noting that the balance is maintained in African green monkeys, a species that – like sooty mangabeys – does not develop immunodeficiency as a result of SIV infection. In contrast, the loss of Th17 cells dysregulated the Th17/Treg balance in pigtailed macaques.

At the close of CROI, conference chair Mario Stevenson highlighted these presentations as representing an important advance in the understanding of HIV pathogenesis. However, until the cause and effect relationship between immune activation and Th17 cell decreases is better understood, the data should probably be interpreted with caution.

Webcasts for the presentations are available on the CROI website (the fourth oral abstract session on Wednesday).

Study abstracts:

15th Conference on Retroviruses & Opportunistic Infections
Session 36 Oral Abstracts
New Insights into Mechanisms of Viral Pathogenecity

Abstract #115

Preferential Loss of Th17 CD4 T Cells in the Gastrointestinal Tract of HIV-infected Individuals but Not SIV-infected Sooty Mangabeys

Barbara Cervasi*1, J Brenchley2, M Paiardini1, S Gordon1, A Asher2, I Frank1, J Else3, D Douek2, and G Silvestri1
1Univ of Pennsylvania, Philadelphia, US; 2Vaccine Res Ctr, NIAID, NIH, Bethesda, MD, US; and 3Yerkes Natl Primate Res Ctr, Emory Univ, Atlanta, GA, US

Background: A new subset of CD4 T helper cells, designated Th17, has been recently identified, leading to a reappraisal of the classically proposed dichotomy in which memory CD4 T cells belong exclusively to either a Th1 or Th2 lineage. While Th17 cells may be involved in the pathogenesis of autoimmune diseases, emerging data suggest that they are important in mucosal immunity to extra-cellular bacteria infection. Here we examined the frequency, antigen specificity, functionality, phenotype, and infection frequency of Th17 cells in the peripheral blood and gastrointestinal tract of HIV-infected and uninfected humans and simian immunodeficiency virus (SIV) -infected and uninfected sooty mangabeys, a natural, non-progressing host species of SIV infection.

Methods: Peripheral blood- and gastrointestinal tract-derived cells were obtained from 30 HIV-infected individuals and 15 healthy controls; peripheral blood, gastrointestinal, and lymph node cells were also obtained from 12 naturally SIV-infected and 10 uninfected sooty mangabeys. Phenotypical and functional characterization of Th17 cells was performed by 10-color flow cytometry.

Results: Interleukin (IL) -17+, interferon (IFN) -g–, IL-4–, IL-2+, IL23R+ memory CD4+ Th17 cells can be identified in all the examined tissues of humans and sooty mangabeys, with the majority of Th17 cells being CCR5– in peripheral blood, but CCR5+ in the gastrointestinal tract. The main results of this study were: There is significant depletion of Th17 cells in the gastrointestinal tract of HIV-infected individuals than in uninfected individuals; SIV-infected sooty mangabeys maintain healthy frequencies of Th17 cells in the the gastrointestinal tract; Th17 cells are infected by HIV in vivo, but not preferentially so compared to Th1 cells; and human Th17 cells are specific for bacterial and fungal antigens, but not common viral antigens.

Conclusions: The preferential loss of Th17 in the the gastrointestinal tract of HIV-infected individuals represents a new mechanism of mucosal immune dysfunction during HIV infection. These data thus further elucidate the immunodeficiency of HIV disease and may provide a mechanistic basis for the enteropathy and associated microbial translocation that characterize HIV infection. Conversely, maintenance of gastrointestinal Th17 cells may account for the non-progressive nature of non-pathogenic SIV infection in sooty mangabeys.

Abstract #116

Preferential Loss of Th17 T Cells at Mucosal Sites Predicts AIDS Progression in Simian Immunodeficiency Virus-infected Macaques

Valentina Cecchinato*1, C Trindade1, J M Heraud1, A Laurence2, J Brenchley3, E Tryniszewska1,4, D Venzon1, D Douek3, J O'Shea2, and G Franchini1
1NCI, NIH, Bethesda, MD, US; 2Natl Inst of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, MD, US; 3NIAID, NIH, Bethesda, MD, US; and 4Med Univ of Bialystok, Poland

Background: Depletion of CD4+ T cells in the gut is necessary, but not sufficient to cause AIDS in animal models, raising the possibility that differential depletion of CD4+ T cell subtypes may be important. Th17 is a recently identified lineage of effector CD4+ T-helper interleukin-17 (IL-17) -secreting cells, important in mucosal immunity to extracellular bacteria. Here we investigated whether the frequency of Th17 lymphocytes is modified during simian immunodeficiency virus (SIV) infection, particularly at mucosal sites.

Methods: We enrolled in the study 31 rhesus macaques, either naive or SIV-infected. Blood and lymphoid tissues were collected 2 weeks after infection or at the time of necropsy. The frequency of CD4+IL-17+ cells in tissues was determined by intracellular cytokines staining following stimulation with phorbol 12-myristate 13-acetate (PMA) and ionomycin. Flow cytometric analysis was performed to determine Th17 phenotype. Quantification of SIV Gag DNA in sorted memory CD4 T cells was performed by quantitative polymerase chain reaction (PCR). Human CD4+ T cells were infected in vitro with SIVmac251 and analyzed for IL-17 and p27 expression. Differences between groups were assessed by t-test. The Spearman rank correlation method was used to determine the correlation between the frequency of Th17 in tissues and viral load in plasma.

Results: Our data demonstrate that Th17 cells express CCR5 and can be infected in vitro and in vivo. We found that healthy macaques have a significantly higher number of Th17 T cells in the gut than in systemic tissues. Importantly, in the SIVmac251 model, Th17 cells are preferentially depleted within the first weeks of infection and their frequency is not restored to normal levels over time. Furthermore, we found that SIVmac251-infected macaques that could effectively control viral replication maintain normal levels of Th17 T cells in all tissues, and that there is a negative correlation between Th17 cell frequency at mucosal sites and plasma virus level, suggesting their importance in HIV/SIV pathogenesis.

Conclusions: Because Th17 cells play a central role in innate and adaptive immune response to extracellular bacteria, these data provide a rational explanation for the chronic enteropathy in HIV infection. Thus, therapeutic approaches to reconstitute Th17 numbers and/or function may be of benefit in HIV infection.

Abstract #117LB

Primary SIV Infection Causes Rapid Loss of the Balance between TH17 and T Regulatory Cell Populations in Pathogenic Infection of Non-human Primates

David Favre*1, S Lederer2, B Kanwar1, Z M Ma3, S Proll2, Z Kasakow1, C Miller3, M Katze2, and J McCune1
1Univ of California, San Francisco, US; 2Univ of Washington, Seattle, US; and 3Univ of California, Davis, US

Background: Progression to AIDS is observed after simian immunodeficiency virus (SIV) infection of some but not all non-human primates. To better define the basis for this dichotomy, we performed a comparative analysis of non-pathogenic infection of the African green monkey and pathogenic infection of the Asian pigtailed macaque. We hypothesized that a critical distinction between pathogenic and nonpathogenic SIV infections may lie in a shift in the equilibrium between pro- and anti-inflammatory host immune responses during acute infection.

Methods: To address this hypothesis, we investigated early virologic, immunologic, and gene expression events that occurred in the peripheral blood and in lymphoid organs, including the colonic mucosa, after acute infection of pigtailed macaques and African green monkeys with the same SIVagm.sab92018 isolate. In particular, we explored the absolute number and relative frequency of TH17 cells, which produce the pro-inflammatory cytokine interleukin (IL) -17, and FoxP3+ T regulatory cells (Treg), the function of which is instead immunosuppressive.

Results: After infection with SIVagm.sab92018, both African green monkeys and pigtailed macaques developed high peak and set point viral loads, but only pigtailed macaque showed generalized CD4+ T cell depletion and sustained immune activation and inflammation. In addition, the following novel features were observed. First, signs of chronic T cell activation and inflammation were detected in the colonic mucosa of both species, while systemic immune activation (e.g., in the blood and lymph node) was only sustained in pigtailed macaques. Second, an increase in the frequency of FoxP3+ Treg was detected early in the blood and in lymph nodes after infection of African green monkeys, but only at later time points in the colon of both African green monkeys and pigtailed macaques. Finally, and unexpectedly, the TH17 population was maintained in infected African green monkeys, but rapidly lost during the course of progressive SIV disease in pigtailed macaques.

Conclusions: These results suggest that balanced representation of both Treg and TH17 cells plays a unique and possibly determinative role in the setting of acute lentiviral infection. Because IL-17 acts to preserve the integrity of the mucosal barrier, thus enhancing host defenses against microbial agents, maintenance of robust TH17 function in mucosal tissue during SIV infection may decrease systemic inflammation and disease progression by preventing the immune activation that would otherwise occur after microbial translocation and spread.

Functional Impairment of MTB-Specific Memory Cells in HIV Infection

Following on from yesterday’s posting about central memory T cells, a new paper in the Journal of Infectious Diseases offers what may be a clearer view of the impact of the loss of this cell population in HIV infection. Cheryl Day and collaborators (from the University of KwaZulu Natal in Durban and Partners AIDS Research in Boston) investigated memory CD4 and CD8 T cell responses to Mycobacterium tuberculosis (MTB) in a cohort of individuals with HIV infection and latent (asymptomatic) TB, finding that cells producing IL-2 – a signature cytokine of central memory T cells - are depleted compared to cells producing TNF-alpha and interferon gamma or interferon gamma alone. Furthermore, they report an inverse correlation between HIV viral load and the proportion of IL-2-producing MTB-specific CD4 T cells, stating that the data “suggests that MTB-specific CD4 T cells in HIV-1-positive subjects may gradually lose IL-2 secretion capacity as HIV-1 disease progresses.” The authors note that further studies are required to identify the mechanisms contributing to the impairment of MTB-specific T cell responses in people with HIV, and the data also suggest that more studies of the impact of antiretroviral therapy on the functional profile of MTB-specific T cell responses are needed.

The Journal of Infectious Diseases 2008;197:000–000
DOI: 10.1086/529048

MAJOR ARTICLE

Detection of Polyfunctional Mycobacterium tuberculosis–Specific T Cells and Association with Viral Load in HIV-1–Infected Persons

Cheryl L. Day,1,3,4,a Nompumelelo Mkhwanazi,1 Sharon Reddy,1 Zenele Mncube,1 Mary van der Stok,1 Paul Klenerman,2 and Bruce D. Walker1,3,4,5

1HIV Pathogenesis Programme, Doris Duke Medical Research Institute, University of KwaZulu Natal, Durban, South Africa; 2Nuffield Department of Medicine, The Peter Medawar Building for Pathogen Research, Oxford University, Oxford, United Kingdom; 3Partners AIDS Research Center, Massachusetts General Hospital, and 4Division of AIDS, Harvard Medical School, Boston, Massachusetts; and 5Howard Hughes Medical Institute, Chevy Chase, Maryland

Background. The human immunodeficiency virus type 1 (HIV-1) epidemic is associated with a significant increase in the incidence of tuberculosis (TB); however, little is known about the quality of Mycobacterium tuberculosis (MTB)–specific cellular immune responses in coinfected individuals.

Methods. A total of 137 HIV-1–positive individuals in Durban, South Africa, were screened with the use of overlapping peptides spanning Ag85A, culture filtrate protein 10 (CFP-10), early secretory antigen target 6 (ESAT-6), and TB10.4, in an interferon (IFN)–γ enzyme-linked immunospot (ELISPOT) assay. Intracellular cytokine staining for MTB-specific production of IFN-γ, tumor necrosis factor (TNF)–α, and interleukin (IL)–2 was performed, as was ex vivo phenotyping of memory markers on MTB-specific T cells.

Results. A total of 41% of subjects responded to ESAT-6 and/or CFP-10, indicating the presence of latent MTB infection. The proportion of MTB-specific IFN-γ+/TNF-α+ CD4+ cells was significantly higher than the proportion of IFN-γ+/IL-2+ CD4+ cells (p=.0220), and the proportion of MTB-specific IL-2–secreting CD4 cells was inversely correlated with the HIV-1 load (p=.0098). MTB-specific CD8 T cells were predominately IFN-γ+/TNF-α+/IL-2−. Ex vivo memory phenotyping of MTB-specific CD4 and CD8 T cells indicated an early to intermediate differentiated phenotype for the population of effector memory cells.

Conclusions. Polyfunctional MTB-specific CD4 and CD8 T cell responses are maintained in the peripheral blood of HIV-1–positive individuals, in the absence of active disease, and the functional capacity of these responses is affected by HIV-1 disease status.

The Journal of Infectious Diseases 2008;197:000–000
DOI: 10.1086/529049

EDITORIAL COMMENTARY

T Cells and Tuberculosis: Beyond Interferon-γ

Ajit Lalvani and Kerry A. Millington

Tuberculosis Immunology Group, Department of Respiratory Medicine, National Heart and Lung Institute, Imperial College London, London, United Kingdom
Received 18 December 2007; accepted 19 December 2007; electronically published 4 March 2008.

FOXO3a and the Survival of Central Memory CD4 T cells in HIV Infection

Recent research on the pathogenesis of HIV and SIV infections has identified the depletion of “central memory” (TCM) CD4 T cells as being a crucial – perhaps the crucial – factor precipitating the development of clinical immunodeficiency and AIDS. As covered previously on the blog, TCM are a subset of memory T cells endowed with the capacity for self-renewal and capable of prodigious proliferation if they encounter a pathogen that they recognize. The research group of Rafik-Pierre Sekaly at the University of Montreal recently identified the transcription factor forkhead box O3a (FOXO3a) as a key mediator of TCM survival. FOXO3a controls several genes involved in the cell cycle (the process of cell division) and apoptosis (cell death) such as bim, but when FOXO3a is phosphorylated, it is excluded from the cell nucleus and thus cannot promote expression of these genes. In their paper published in the Journal of Experimental Medicine in late 2006 (see abstract, below), Catherine Riou and colleagues showed that TCM expressed higher levels of the transcriptionally inactive phosphorylated forms of FOXO3a and, consequently, lower levels of the cell-death promoting protein Bim. They also reported that FOXO3a phosphorylation can be promoted synergistically by signaling through the T cell receptor (TCR) and cytokines known to enhance memory T cell survival such as IL-7.

Now, in paper just published online by Nature Medicine, the same group demonstrate that HIV-infected elite controllers (individuals controlling viral load to less than 50 copies in the absence of any treatment) have higher levels of phosphorylated FOXO3a in both TCM and effector memory CD4 T cells (TEM) compared to individuals on successful antiretroviral therapy (ART). Conversely, levels of the pro-apoptotic protein Bim were higher in TCM and TEM from people on ART compared to elite controllers. Consistent with these findings, TCM and TEM from elite controllers survived repeated rounds of in vitro stimulation far better than those from individuals on ART. The researchers also used various techniques to inactivate FOXO3a in TCM and TEM cells from ART-treated study subjects and found that these interventions made their in vitro lifespan comparable with TCM and TEM from the elite controllers.

In discussing the results, the authors note that it is possible signaling pathways are perturbed in the cells of individuals on ART such that the integration of TCR and cytokine signals does not cause FOXO3a phosphorylation in the normal way. They also suggest that specific genetic polymorphisms affecting cell signaling pathways could be present in elite controllers; ongoing whole-genome analyses (in a study led by Partners AIDS Research Center in Boston) will be able to address this possibility. The authors close by stating that their work confirms the importance of maintaining TCM integrity in preventing HIV disease progression, and that approaches with the potential to enhance TCM survival may deserve study as therapies for HIV infection.

Nat Med. 2008 Mar 2 [Epub ahead of print]

Transcription factor FOXO3a controls the persistence of memory CD4(+) T cells during HIV infection.

van Grevenynghe J, Procopio FA, He Z, Chomont N, Riou C, Zhang Y, Gimmig S, Boucher G, Wilkinson P, Shi Y, Yassine-Diab B, Said EA, Trautmann L, Far ME, Balderas RS, Boulassel MR, Routy JP, Haddad EK, Sekaly RP.

[1] Laboratoire d'Immunologie, Centre de Recherche, Hôpital Saint-Luc, Centre Hospitalier de l'Université de Montréal, 264 Boulevard Rene-Levesque Est, Montréal, Québec H2X 1P1, Canada. [2] Laboratoire d'Immunologie, Département de Microbiologie et d'Immunologie, Université de Montréal, 264 Boulevard Rene-Levesque Est, Montréal, Québec H3T 1J4, Canada. [3] Institut national de la Santé et de la Recherche médicale U743, Centre de Recherche, Centre Hospitalier de l'Universite de Montréal, 264 Boulevard Rene-Levesque Est, Montreal, Québec H2X 1P1, Canada.

The persistence of central memory CD4(+) T cells (T(CM) cells) is a major correlate of immunological protection in HIV/AIDS, as the rate of T(CM) cell decline predicts HIV disease progression. In this study, we show that T(CM) cells and effector memory CD4(+) T cells (T(EM) cells) from HIV(+) elite controller (EC) subjects are less susceptible to Fas-mediated apoptosis and persist longer after multiple rounds of T cell receptor triggering when compared to T(CM) and T(EM) cells from aviremic successfully treated (ST) subjects or from HIV(-) donors. We show that persistence of T(CM) cells from EC subjects is a direct consequence of inactivation of the FOXO3a pathway. Silencing the transcriptionally active form of FOXO3a by small interfering RNA or by introducing a FOXO3a dominant-negative form (FOXO3a Nt) extended the long-term survival of T(CM) cells from ST subjects to a length of time similar to that of T(CM) cells from EC subjects. The crucial role of FOXO3a in the survival of memory cells will help shed light on the underlying immunological mechanisms that control viral replication in EC subjects.

J Exp Med. 2007 Jan 22;204(1):79-91. Epub 2006 Dec 26.

Convergence of TCR and cytokine signaling leads to FOXO3a phosphorylation and drives the survival of CD4+ central memory T cells.

Riou C, Yassine-Diab B, Van grevenynghe J, Somogyi R, Greller LD, Gagnon D, Gimmig S, Wilkinson P, Shi Y, Cameron MJ, Campos-Gonzalez R, Balderas RS, Kelvin D, Sekaly RP, Haddad EK.

Laboratoire d'Immunologie, Centre de Recherche, Hôpital Saint-Luc du Centre Hospitalier de l'Université de Montréal (CHUM), Montréal, Québec H2X 1P1, Canada.

The molecular events involved in the establishment and maintenance of CD4+ central memory and effector memory T cells (TCM and TEM, respectively) are poorly understood. In this study, we demonstrate that ex vivo isolated TCM are more resistant to both spontaneous and Fas-induced apoptosis than TEM and have an increased capacity to proliferate and persist in vitro. Using global gene expression profiling, single cell proteomics, and functional assays, we show that the survival of CD4+ TCM depends, at least in part, on the activation and phosphorylation of signal transducer and activator of transcription 5a (STAT5a) and forkhead box O3a (FOXO3a). TCM showed a significant increase in the levels of phosphorylation of STAT5a compared with TEM in response to both IL-2 (P<0.04) and IL-7 (P<0.002); the latter is well known for its capacity to enhance T cell survival. Moreover, ex vivo TCM express higher levels of the transcriptionally inactive phosphorylated forms of FOXO3a and concomitantly lower levels of the proapoptotic FOXO3a target, Bim. Experiments aimed at blocking FOXO3a phosphorylation confirmed the role of this phosphoprotein in protecting TCM from apoptosis. Our results provide, for the first time in humans, an insight into molecular mechanisms that could be responsible for the longevity and persistence of CD4+ TCM.

Updates on the STEP and Phambili Vaccine Trials

As reported yesterday by AVAC's Advocates Network, NIAID released a statement this past Monday reporting that a total of 11 infections occurred in the South African study of Merck's HIV vaccine, dubbed the Phambili trial or HVTN 503. Immunizations in the Phambili were stopped before it was fully enrolled due to the results from the preceding international STEP/HVTN 502 trial, which showed no efficacy of the vaccine and evidence of enhanced susceptibility in a subset of immunized volunteers. Of the 11 infections in Phambili, seven occurred in vaccine recipients (although most had not received all three immunizations) and four in placebo recipients. Among the vaccine recipients, six infections occurred in people with titers of anti-Ad5 antibodies over 1:18 compared to three infections among people with similar titers in the placebo group. In participants with no detectable antibodies to Ad5 at enrollment, there was one case of HIV infection among vaccine recipients and one case of HIV infection among placebo recipients. Given the small numbers, the three additional infections in the vaccine group cannot be interpreted as evidence that the vaccine enhanced susceptibility in this population, although it is of course possible that such evidence would have emerged, had the trial not been stopped. The apparent difference in infections based on anti-Ad5 antibody titer (6 vs. 3 infections compared to 1 vs. 1) also involves far too few events for any inferences to be drawn, although it may still prompt speculation that a similar scenario to the STEP trial was developing. Notably, all but one of the cases of HIV infection in the Phambili trial were among women, who made up 360 of the 801 participants at the time immunizations in the study were halted.

NIAID have also released more materials regarding the multivariate analyses of the STEP data that have recently been conducted. The analyses find that receipt of the vaccine was associated with enhanced susceptibility but that this effect was strongest in uncircumcised men with antibody titers against the Ad5 vector of over 1:200. Circumcised men with undetectable Ad5 antibodies did not show any evidence of vaccine-related enhanced susceptibility. However, the picture is less clear for men with either risk factor (uncircumcised or Ad5 antibody titer over 1:200), as both groups showed some evidence of vaccine-related enhanced susceptibility in the multivariate analyses.

In terms of the association between anti-Ad5 antibody titers and risk of HIV infection seen in the placebo arm, the NIAID document states that: "Although the study was stratified on Ad5 titers, the Ad5 seropositive population differed substantially from the Ad5 seronegative population. When we looked in the multivariate analyses for association of Ad5 serostatus with infection risk independent of vaccination, Ad5 serostatus was not independently associated with HIV infection. We did find, though, that being North American was associated with an increased risk of acquisition. It may be that the reason for the higher rates of infection in the placebo arm among those who were Ad5 seronegative was because that subgroup was primarily in the North American cohorts, which reported riskier sexual practices and substance use. We are exploring this in other observational cohort studies."

So in this formulation, people in the placebo group with anti-Ad5 antibody titers below 1:18 were more likely to be North American, people with antibody titers between 1:18 and 1:1000 were a bit less likely to be North American and people with antibody titers over 1,000 were much less likely to be North American. Individuals with antibody titers over 1:200 were also less likely to be circumcised, so while circumcision appeared to have a protective effect in vaccine recipients, no such effect was seen in the placebo arm. People in the vaccine arm with higher anti-Ad5 antibody titers should presumably also have been at less risk because they were less likely to be North American, but the vaccine increased their risk somehow. To provide some context, this is the overall incidence table from STEP, with the confidence intervals included to give a sense of the uncertainty associated with the findings.

Anti-Ad5 antibody titer	HIV incidence vaccine (95% CI)	HIV incidence placebo (95% CI)	Relative incidence V:P
less than 1:18	4% (2.5-6.3)	4% (2.5-6.2)	1.0
1:18-1:200	4.4% (1.9-8.8)	2.2% (0.6-5.5)	2.0
1:200-1:1000	6.1% (3.3-10.2)	3% (1.2-6.2)	2.0
over 1:1000	4.4% (1.2-6.2)	1.2% (0.2-4.5)	3.5

The NIH is holding a summit on March 25 to discuss the implications of the STEP results for the future of HIV vaccine research, meeting and registration information are available online.

Update 2/29/08: Thanks to Sarah Alexander from the HVTN for pointing out several significant errors in the original posting:

1) the sample size for Phambili was slated to be 3,000, the same as for STEP, so it was not intended to be a smaller trial
2) women did not make up the majority of the Phambili enrollees (of the 801 participants, 360 were women and 441 men)
3) follow-up is continuing in both the STEP and Phambili trials so while immunizations and enrollment have ceased, the trials continue.

Apologies for the mistakes, which have also been corrected in the body of the post

Genetic Associations with Control of HIV Replication

At the recent CROI meeting, David Goldstein from Duke University gave a plenary presentation on genetic determinants of viral load set point in HIV-infected individuals. Goldstein’s talk was an update on work published recently with CHAVI (Center for AIDS Vaccine Immunology) collaborator Amalio Telenti from the University of Lausanne. Their approach involved analyzing a staggering 500,000 different single nucleotide polymorphisms (SNPs) present in the human genome to see if they were associated with lower viral load set points (although Goldstein noted that this initial work is restricted to SNPs documented in >1% of the population and thus excludes rare polymorphisms). As reported in the Science paper, three sets of significant associations emerged. The first was a SNP in a gene called HCP5. This is linked to an immune response gene called HLA B*5701 which is well known to be over-represented among HIV-infected long-term non-progressors (HLA B*5701 makes a receptor on CD8 T cells which appears particularly good at recognizing HIV epitopes). However, HCP5 is also an endogenous retroviral element (a part of the human genome derived from an ancient retroviral infection which gained access the human germ line by infecting an egg or sperm cell) and so Goldstein initially speculated that the SNP in HCP5 might somehow have a direct anti-HIV effect. At CROI, he reported that studies in which the SNP-containing version of HCP5 was overexpressed in cells showed no inhibition of HIV replication, suggesting that the SNP is mediating its effect via other means (exactly how is under investigation).

The second association uncovered by Goldstein’s work was with a SNP in the gene for HLA-C. HLA-C molecules are part of a family called class I HLA molecules that are involved in the recognition of pathogens by CD8 T cells. Thousands of HLA molecules are displayed on the surface of every cell in the body (except red blood cells) and their job is to constantly lift cellular debris from inside the cell and display it to passing CD8 T cells. If an HLA molecule displays a protein fragment (epitope) from a pathogen that has infected the cell, this can trigger the CD8 T cell to release toxic enzymes, which cause the cell to die. The majority of HLA molecules on cells belong to classes HLA-A and HLA-B, while HLA-C molecules are less frequent and, as a result, less studied. But because HIV’s Nef protein is known to cause a reduction in HLA-B molecules on infected cells (as a means of escaping the immune response), Goldstein’s hypothesis is that the SNP he has identified causes more HLA-C molecules to be expressed, thereby making it easier for CD8 T cells to identify HIV-infected cells. He is currently collaborating with Andrew McMichael at Oxford University to measure the effect of the SNP on HLA-C expression.

The third association described by Goldstein involves a set of seven SNPs in or near two genes, ZNRD1 and RNF39. In the Science paper, this association was only seen with disease progression (as assessed by time to a CD4 T cell count of less than 350) but at CROI Goldstein reported that in an expanded analysis involving 1,000 more individuals (in addition to the 446 studied initially), an association with viral load set point was also documented. ZNRD1 encodes a protein involved in RNA transcription and, interestingly, was also identified in the recent, widely publicized study of host proteins needed by HIV to replicate. The function of RNF39 remains to be determined. Goldstein stressed that the discovery of all these associations is a first step, and efforts are underway to uncover the mechanisms by which the identified SNPs mediate their effects.

To give a sense of how a combination of genetic factors can have a profound impact on HIV disease progression, Goldstein showed an analysis that included the SNPs in the HCP5, HLA-C, and ZNRD1/RNF39 genes and two other known favorable genetic polymorphisms in CCR5 and CCR2 genes (CCR5Δ32 and CCR2 V64I). HIV-infected individuals with no favorable mutations in any of these genes showed an average time of less than two years from infection to a CD4 T cell count of less than 350. In contrast, people with one or two favorable mutations in at least four of these genes did not experience a CD4 T cell decline to this level for an average period of more than eight years.

Goldstein’s group is now embarking on an effort to uncover genetic associations with the magnitude of antibody responses generated against a vaccine, using data from the North American efficacy trial of AIDSVAX. This is a potentially important area of study because the ability of an individual to generate a high titer antibody response was correlated with reduced susceptibility to HIV infection in the trial cohort. Some researchers have suggested an analogy with the association between the magnitude of anti-Ad5 antibody responses and susceptibility to HIV infection seen in the placebo group of the recent Merck vaccine trial, so Goldstein’s work may have the potential shed light on that mystery also. Another important area of ongoing study mentioned by Goldstein is an analysis of genetic associations with viral load set point restricted to African American individuals.

The webcast of David Goldstein’s talk is available on the CROI website, it is the second presentation on Monday.

Shortly after CROI, a study of untreated HIV-infected individuals with viral loads under 50 copies – a group now dubbed “elite controllers” – highlighted the complexities and limitations of genetic associations. In a research letter published in the journal AIDS, Yefei Han and colleagues from Bob Siliciano’s laboratory at Johns Hopkins report that, of 16 elite controllers analyzed, only four possessed the SNP in HLA-C identified in the CHAVI work and none had the SNP in HCP5 (although two had HLA B*5702 and HLA B*5703 genes, which are closely related to HLA B*5701). The researchers did find the HCP5 SNP in two other infected individuals with HLA B*5701, one of whom had a low but detectable viral load and the other with progressive disease. The authors also note that the frequency of the HLA-C SNP in the elite controllers was not significantly different from what would be expected in a larger population of people from the same ethnic background. Citing these data and an example of an individual in their cohort who developed an increasing viral load after developing a CD8 T cell escape mutation, they argue that adaptive immune responses are likely an important contributor to the control of viral replication in elite controllers.

AIDS. 22(4):541-544, February 19, 2008.

The role of protective HCP5 and HLA-C associated polymorphisms in the control of HIV-1 replication in a subset of elite suppressors.

Research Letters

Han, Yefei; Lai, Jun; Barditch-Crovo, Patricia; Gallant, Joel E; Williams, Thomas M; Siliciano, Robert F; Blankson, Joel N

Abstract:

Elite suppressors (ES) are untreated HIV-1-infected patients who maintain undetectable viral loads. A recent whole-genome analysis identified two independent polymorphisms associated with low viral loads in untreated HIV-1 infection. We screened 16 ES; none were positive for the protective HLA complex 5 gene polymorphism, and only four were positive for the protective polymorphism associated with the HLA-C gene. These results suggest that some ES control viremia by mechanisms independent of the newly-identified genetic factors.