Michael Palm Basic Science, Vaccines & Prevention Project Weblog

Transmission of CD8 T Cell Escape Mutants is Associated with Lower Viral Loads in Newly Infected Individuals

A new study by Paul Goepfert and colleagues offers compelling evidence that CD8 T cell responses can pressure HIV into mutating in ways that compromise viral fitness. CD8 T cells target tiny slices of viral proteins called epitopes, which are displayed by infected cells as a sort of alarm signal; CD8 T cells that recognize a particular epitope (via a docking bay-type structure called a T cell receptor or TCR) can mediate destruction of an infected cell by releasing cell-destroying substances such as perforin and granzyme B. HIV mutations that impact the structure of an epitope can abrogate CD8 T cell recognition and this phenomenon is called immune escape (it is loosely analogous to the way mutations can allow the virus to resist the effects of antiretroviral drugs). If a particular epitope-specific CD8 T cell response is effective, viruses with escape mutations in that epitope are at a selective advantage because they can persist despite the presence of the immune response. However, certain parts of HIV can tolerate mutations more easily than others, and studies have shown that the Gag protein is so vital to replication that mutations affecting epitopes in Gag can impair the ability of the virus to replicate in a lab dish (in vitro).

To assess whether these observations are relevant in people, Goepfert et al analyzed data from 114 epidemiologically linked transmission pairs in Zambia (the individuals had been participants in a larger cohort of “discordant” couples in which one partner was HIV-infected; despite counseling and increased condom use in the cohort, transmission still occurred at a rate of approximately 8% per year). The researchers looked for evidence of CD8 T cell escape mutations in Gag and Nef and then analyzed whether the transmission of HIV containing mutations impacted viral load in the newly infected individual. The timepoint for the analysis was 6 months after infection, because 579 of the 610 mutations documented in the transmitting partners were still present in recipients at this time point. The results showed that higher numbers of CD8 T cell escape mutations in the Gag protein of transmitted viruses were significantly associated with lower viral loads in the newly infected individuals (no such effect was seen for Nef). Further analyses revealed that the effect was most consistent for Gag epitopes targeted by HLA-B-restricted CD8 T cells (HLA genes manufacture the CD8 T cell TCR and thus govern the epitope structures that a CD8 T cell can recognize). The researchers hypothesized that the impact on HIV replication of mutations in epitopes targeted by HLA-B-restricted CD8 T cells would be most prominent in individuals lacking the same HLA-B genes themselves, and indeed this turned out to be the case: when the analysis was restricted to only these study participants the associations between more mutations in Gag and lower viral load became much stronger (p=0.0003).

In discussing their results, the study authors note that there was a ~10-fold difference in viral loads when individuals infected with viruses containing less than two escape mutations in Gag were compared to those with more than six such mutations, suggesting that infection with these multiple escape mutants may slow disease progression (although longer term follow up of a subset of participants indicates that viral load is, as is typical, increasing over time). They also state that: “these data imply that for CTL-based HIV vaccines to effectively control viral load, they must simultaneously target multiple Gag epitopes, thereby ensuring that fitness constraints prevent the virus from easily mutating.” In an accompanying news brief, JEM editor Hema Bashyam speculates that the study “might explain why a T cell vaccine that induces immune responses against two Gag epitopes failed in a recent trial.” In fact, recipients of the Merck HIV vaccine developed responses to an average of just one Gag epitope.

The Journal of Experimental Medicine
Published online 21 April 2008
doi:10.1084/jem.20072457

BRIEF DEFINITIVE REPORT

Transmission of HIV-1 Gag immune escape mutations is associated with reduced viral load in linked recipients

Paul A. Goepfert1,2, Wendy Lumm4, Paul Farmer4, Philippa Matthews5, Andrew Prendergast5, Jonathan M. Carlson6,7, Cynthia A. Derdeyn4,8, Jianming Tang1,2, Richard A. Kaslow3, Anju Bansal1, Karina Yusim10, David Heckerman6, Joseph Mulenga11, Susan Allen9, Philip J.R. Goulder5,12,13, and Eric Hunter4,8

1 Department of Medicine, 2 Department of Microbiology, and 3 Department of Epidemiology, University of Alabama at Birmingham, Birmingham, AL 35294
4 Emory Vaccine Center at Yerkes National Primate Research Center, Atlanta, GA 30322
5 Department of Pediatrics, The Peter Medawar Building for Pathogen Research, University of Oxford, Oxford OX1 3SY, England, UK
6 Microsoft Research, Redmond, WA 98052
7 Department of Computer Science and Engineering, University of Washington, Seattle, WA 98195
8 Department of Pathology and Laboratory Medicine and 9 Department of Global Health, Emory University, Atlanta, GA 30322
10 Los Alamos National Laboratory, Los Alamos, NM 87545
11 Zambia-Emory HIV Research Group, Lusaka, Zambia
12 HIV Pathogenesis Program, Doris Duke Medical Research Institute, University of KwaZulu-Natal, Durban 4013, South Africa
13 Partners AIDS Research Center, Massachusetts General Hospital, Charlestown, MA 02129

CORRESPONDENCE Paul A. Goepfert: paulg@uab.edu

In a study of 114 epidemiologically linked Zambian transmission pairs, we evaluated the impact of human leukocyte antigen class I (HLA-I)–associated amino acid polymorphisms, presumed to reflect cytotoxic T lymphocyte (CTL) escape in Gag and Nef of the virus transmitted from the chronically infected donor, on the plasma viral load (VL) in matched recipients 6 mo after infection. CTL escape mutations in Gag and Nef were seen in the donors, which were subsequently transmitted to recipients, largely unchanged soon after infection. We observed a significant correlation between the number of Gag escape mutations targeted by specific HLA-B allele–restricted CTLs and reduced VLs in the recipients. This negative correlation was most evident in newly infected individuals, whose HLA alleles were unable to effectively target Gag and select for CTL escape mutations in this gene. Nef mutations in the donor had no impact on VL in the recipient. Thus, broad Gag-specific CTL responses capable of driving virus escape in the donor may be of clinical benefit to both the donor and recipient. In addition to their direct implications for HIV-1 vaccine design, these data suggest that CTL-induced viral polymorphisms and their associated in vivo viral fitness costs could have a significant impact on HIV-1 pathogenesis.

HLA Class II Associations with Resistance & Susceptibility to HIV Infection

Following a similar theme to a recently posted study indicating that certain class II HLA genes – which impact antigen presentation to CD4 T cells – are associated with elite control of HIV replication in infected individuals, a new paper just published in the journal AIDS adds to the literature showing significant associations between class II HLA genes and resistance/susceptibility to HIV infection. The data comes from a large cohort study of sex workers in the Pumwani district of Nairobi, approximately 10% of whom persistently resist HIV infection despite an estimated ~60 unprotected exposures per year.

AIDS. 22(7):807-816, April 23, 2008.

BASIC SCIENCE

Human leukocyte antigen-DQ alleles and haplotypes and their associations with resistance and susceptibility to HIV-1 infection.

Hardie, Rae-Anne; Luo, Ma; Bruneau, Brigitte; Knight, Erin; Nagelkerke, Nico JD; Kimani, Joshua; Wachihi, Charles; Ngugi, Elizabeth N; Plummer, Francis A

Abstract:

Objectives: To determine the association of DQ antigens with resistance and susceptibility to HIV-1.

Design: Despite repeated exposure to HIV-1, a subset of women in the Pumwani Sex Worker cohort established in Nairobi, Kenya in 1985 have remained HIV-1 negative for at least 3 years and are classified as resistant. Differential susceptibility to HIV-1 infection is associated with HIV-1 specific CD4+ and CD8+ T cell responses. As human leukocyte antigen-DQ antigens present viral peptides to CD4+ cells, we genotyped human leukocyte antigen -DQ alleles for 978 women enrolled in the cohort and performed cross-sectional and longitudinal analyses to identify associations of human leukocyte antigen -DQ with resistance/susceptibility to HIV-1.

Methods: DQA1 and DQB1 were genotyped using taxonomy-based sequence analysis. SPSS 13.0 was used to determine associations of DQ alleles/haplotypes with HIV-1 resistance, susceptibility, and seroconversion rates.

Results: Several DQB1 alleles and DQ haplotypes were associated with resistance to HIV-1 infection. These included DQB1*050301 (P = 0.055, Odds Ratio = 12.77, 95% Confidence Interval = 1.44-112), DQB1*0603 and DQB1*0609 (P = 0.037, Odds Ratio = 3.25, 95% Confidence Interval = 1.12-9.47), and DQA1*010201-DQB1*0603 (P = 0.044, Odds Ratio = 17.33, 95% Confidence Interval = 1.79-168). Conversely, DQB1*0602 (P = 0.048, Odds Ratio = 0.68, 95% Confidence Interval = 0.44-1.05) and DQA1*010201-DQB1*0602 (P = 0.039, Odds Ratio = 0.64, 95% Confidence Interval = 0.41-1.03) were overrepresented in the HIV-1 infected population. DQA1*0504-DQB1*0201, DQA1*010201-DQB1*0201, DQA1*0402-DQB1*0402 and DQA1*0402-DQB1*030101 genotypes were only found in HIV-1 positive subjects (Odds Ratio = 0.30-0.31, 95% Confidence Interval = 0.03-3.70), and these women seroconverted rapidly. The associations of these DQ alleles and haplotypes with resistance and susceptibility to HIV-1 were independent of the previously reported human leukocyte antigen-DRB*01, human leukocyte antigen A2/6802, and human leukocyte antigen-A*2301.

Conclusion: The associations of DQ alleles and haplotypes with resistance and susceptibility to HIV-1 emphasize the importance of human leukocyte antigen-DQ and CD4 in anti-HIV-1 immunity.

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.

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.

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.

Two New Papers on Polyfunctional T cell Responses

Two new papers add to the literature demonstrating that the presence of polyfunctional HIV-specific T cell responses correlates with low viral load in HIV infection. In one paper, in the European Journal of Immunology, Rick Koup and colleagues report that virus-specific CD4 and CD8 T cells capable of multiple functions (involving various combinations of IL-2, TNF alpha, interferon gamma and MIP-1 beta production and expression of CD107a, a marker indicating cell-killing ability) are far more frequently detectable in HIV-2 than HIV-1 infection. The study also confirms prior reports that such polyfunctional T cells make 15-20 times more interferon gamma and TNF alpha on a per-cell basis than monofunctional T cells. The authors note that the data do not prove a causative role of the T cell responses in controlling viral load and preventing or slowing disease progression in HIV-2-infected individuals, but cite evidence from other recent studies - including Robert Seder's murine Leishmania vaccine work - as suggesting that polyfunctional T cells do contribute to control of pathogens and are thus unlikely to have arisen simply as a consequence of low viral load in HIV-2 infection.

In a second paper - just published online in J. Virology by Marybeth Daucher from Daniel Douek's laboratory at NIAID - polyfunctional CD8 T cell responses targeting HIV-1 are reported to be associated with maintenance of low viral loads subsequent to antiretroviral therapy interruption. The study only involves six individuals so the authors are appropriately cautious in interpreting the data, but they emphasize that other parameters such as the magnitude and breadth of the HIV-specific CD8 T cell response showed no correlations with post-interruption outcomes.  In summing up their work, the researchers state: "Despite the small number of patients and the inherent issues associated with the interpretation of observational studies in humans, these data suggest that functional attributes of the HIV-specific CD8 T-cell response might be important correlates of virological outcome after exposure to SIT regimens and could represent useful biological parameters to measure in the clinical context."

European Journal of Immunology
Volume 38, Issue 2 , Pages 350 - 363
Published Online: 17 Jan 2008

Polyfunctional T cell responses are a hallmark of HIV-2 infection (free access to full text)

Melody G. Duvall 1 2, Melissa L. Precopio 2, David A. Ambrozak 2, Assan Jaye 4, Andrew J. McMichael 1, Hilton C. Whittle 4, Mario Roederer 3, Sarah L. Rowland-Jones 1 4, Richard A. Koup, Dr. 2

1MRC Human Immunology Unit, Weatherall Institute of Molecular Medicine, University of Oxford, Oxford, UK
2Immunology Laboratory, Vaccine Research Center, NIAID, National Institutes of Health, Bethesda, MD, USA
3Immuno Technology Section, Vaccine Research Center, NIAID, National Institutes of Health, Bethesda, MD, USA
4MRC Laboratories Fajara, Banjul, The Gambia, West Africa

Funded by:
National Institutes of Health, USA
Medical Research Council, UK

Abstract

HIV-2 is distinguished clinically and immunologically from HIV-1 infection by delayed disease progression and maintenance of HIV-specific CD4+ T cell help in most infected subjects. Thus, HIV-2 provides a unique natural human model in which to investigate correlates of immune protection against HIV disease progression. Here, we report a detailed assessment of the HIV-2-specific CD4+ and CD8+ T cell response compared to HIV-1, using polychromatic flow cytometry to assess the quality of the HIV-specific T cell response by measuring IFN-, IL-2, TNF-, MIP-1, and CD107a mobilization (degranulation) simultaneously following Gag peptide stimulation. We find that HIV-2-specific CD4+ and CD8+ T cells are more polyfunctional that those specific for HIV-1 and that polyfunctional HIV-2-specific T cells produce more IFN- and TNF- on a per-cell basis than monofunctional T cells. Polyfunctional HIV-2-specific CD4+ T cells were generally more differentiated and expressed CD57, while there was no association between function and phenotype in the CD8+ T cell fraction. Polyfunctional HIV-specific T cell responses are a hallmark of non-progressive HIV-2 infection and may be related to good clinical outcome in this setting.

JVI Accepts, published online ahead of print on 30 January 2008

J. Virol. doi:10.1128/JVI.02212-07

Virological Outcome After Structured Interruption of Antiretroviral Therapy For HIV Infection is Associated with the Functional Profile of Virus-Specific CD8+ T-Cells

Marybeth Daucher*, David A Price, Jason M Brenchley, Laurie Lamoreaux, Julia A Metcalf, Catherine Rehm, Elizabeth Nies-Kraske, Elizabeth Urban, Christian Yoder, Diane Rock, Julie Gumkowski, Michael R Betts, Mark R Dybul, and Daniel C Douek

Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892; Graduate Genetics Program, George Washington University, Washington, DC 20522; Human Immunology Section, and Immunology Laboratory, Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892; Department of Medical Biochemistry and Immunology, University of Cardiff, Heath Park, Cardiff CF14 4XN, UK; Department of Microbiology, University of Pennsylvania, Philadelphia, PA 19104; Office of the U.S. Global AIDS Coordinator, U.S. Department of State, Washington, DC 20522

A clear understanding of the antiviral effects of CD8+ T-cells in the context of chronic HIV infection is critical for the development of prophylactic vaccines and therapeutics designed to support T-cell-mediated immunity. However, defining the potential correlates of effective CD8+ T-cell immunity has proven difficult; notably, comprehensive analyses have demonstrated that the size and shape of the CD8+ T-cell response is not necessarily indicative of efficacy determined by measures of plasma viral load. Here, we conducted a detailed quantitative and qualitative analysis of CD8+ T-cell responses to autologous virus in a cohort of six HIV-infected individuals with a history of structured interruption of antiretroviral therapy (SIT). The magnitude and breadth of the HIV-specific response did not, by itself, explain the changes observed in plasma virus levels after cessation of ART. Furthermore, mutational escape from targeted epitopes could not account for the differential virological outcomes in this cohort. However, the functionality of HIV-specific CD8+ T-cell populations upon antigen encounter, determined by the simultaneous and independent measurement of five CD8+ T-cell functions (degranulation, IFN{gamma}, MIP1{beta}, TNF{alpha} and IL2) reflected the emergent level of plasma virus with multiple functions being elicited in those individuals with lower viremia after SIT. These data show that the quality of the HIV-specific CD8+ T-cell response, rather than the quantity, is associated with the dynamics of viral replication in the absence of ART and suggest that the effects of SIT can be assessed by measuring the functional profile of HIV-specific CD8+ T-cells.

CD4 T cells & Dendritic Cells on Film

A new paper in the "online ahead of issue" section of the journal Immunity captures interactions between CD4 T cells and dendritic cells (DCs) on film using intravital imaging. Although access to the full text of the article requires a subscription to the journal, the Quicktime videos can be viewed free of charge online. DCs are stained green, CD4 T cells are stained red. Some similar footage has been published previously, but this study goes a step further by using a system that allows the first encounter between DCs and antigen-specific CD4 T cells to be captured in real time.

Immunity
10.1016/j.immuni.2007.08.018

Real-Time Manipulation of T Cell-Dendritic Cell Interactions In Vivo Reveals the Importance of Prolonged Contacts for CD4+ T Cell Activation

Susanna Celli,1,2 Fabrice Lemaître,1,2 and Philippe Bousso1,2,

1 Institut Pasteur, G5 Dynamiques des Réponses Immunes, Paris, F-75015, France
2 Inserm, Equipe Avenir, U668, Paris, F-75015, France

T cells interact with dendritic cells (DCs) for periods lasting from minutes to hours. However, a causal link between the duration of this interaction and the efficiency of T cell activation has not been established in vivo. Employing intravital two-photon imaging, we manipulated T cell-DC interactions in real time and found that the first T cell-DC encounter often resulted in a long-lived interaction. Moreover, the cessation of T cell receptor-major histocompatibility complex signals promoted cellular dissociation, suggesting that antigen availability on DCs regulates contact duration. Finally, at least 6 hr of in vivo T cell-DC interaction were required for naive CD4+ T cells to undergo clonal expansion. These results establish the importance of prolonged T cell-DC interactions for efficient CD4+ T cell activation in vivo.

Hijacking Dendritic Cells: HIV-1 vs. HIV-2

A new paper in the Journal of Virology reports that dendritic cells (DCs) are more susceptible to infection by HIV-1 than its far less pathogenic retroviral sibling, HIV-2. The work is the product of collaboration between researchers at the NIH’s Vaccine Research Center and Oxford University in the UK. DCs act as immune system sentinels that capture pathogens as they first enter the body and transport them to lymph nodes in order to initiate immune responses by T cells and B cells. DCs are referred to as antigen-presenting cells for this reason.

The study authors point out that previous studies have shown that DCs can be susceptible to HIV-1 infection, but that infection of DC’s also leads to presentation of HIV antigens to HIV-specific CD4 T cells. They write: “this suggests that the infection of DCs by HIV-1 and subsequent processing and presentation of HIV-1 antigens on the surface of DCs is a double-edged sword. In order to induce an HIV-specific immune response, it is necessary for DCs to deliver appropriate signals to memory CD4+ T cells. However, in doing so, the intimate contact of DCs and T cells leaves HIV-1-specific CD4+ T cells susceptible to becoming infected by these HIV-1-carrying DCs. This could help to explain mechanistically how HIV-1-specific CD4+ T cells become preferentially infected in vivo.” However, although HIV-2 was extremely inefficient at infecting DCs, the researchers found that HIV-2-specific CD4 T cells were still more likely to be infected by HIV-2 compared to CD4 T cells targeting other pathogens, suggesting that transfer from DCs may not be the only means by which the virus preferentially infects HIV-2-specific CD4 T cells.

The findings may also suggest that HIV-1’s superior ability to target DCs contributes to the different courses of disease associated with HIV-1 and HIV-2 infection. Recent studies by the same researchers have shown that HIV-2-specific CD4 T cell responses are relatively preserved compared to the loss and dysfunction of HIV-1-specific CD4 T cells that occurs in progressive HIV-1 infection. Further investigations of the role of DCs in HIV-1 vs. HIV-2 infection may help shed light on the reasons for these differences. This area of research may also have implications for HIV-1 vaccines; although T cell-based candidates like the Merck Ad5 vaccine reliably induce HIV-specific memory CD4 T cells, it could be the case that HIV-1 infection of DCs (or other antigen-presenting cells such as macrophages) somehow prevents these CD4 T cell responses from being activated efficiently.

JVI Accepts, published online ahead of print on 3 October 2007
J. Virol. doi:10.1128/JVI.00976-07

Dendritic Cells Are Less Susceptible to HIV-2 Than HIV-1 Infection

Melody G. Duvall, Karin Loré, Hetty Blaak, David A. Ambrozak, William C. Adams, Kathlyn Santos, Christof Geldmacher, John R. Mascola, Andrew J. McMichael, Hilton C. Whittle, Sarah L. Rowland-Jones, and Richard A. Koup

HIV-1 infection of dendritic cells (DCs) has been documented in vivo, and may be an important contributor to HIV-1 transmission and pathogenesis. HIV-1-specific CD4+ T cells respond to HIV antigens presented by HIV-1-infected DCs and in this process become infected, thereby providing a mechanism through which HIV-1-specific CD4+ T cells could become preferentially infected in vivo. HIV-2 disease is attenuated with respect to HIV-1 and host immune responses are thought to be contributory. Here, we investigated the susceptibility of primary myeloid DCs (mDCs) and plasmacytoid DCs (pDCs) to infection by HIV-2. We found that neither CCR5-tropic primary HIV-2 isolates nor a lab-adapted CXCR4-tropic HIV-2 could efficiently infect mDCs or pDCs, though these viruses could infect primary CD4+ T cells in vitro. HIV-2-exposed mDCs were also incapable of transferring virus to autologous CD4+ T cells. Despite this, we found that HIV-2-specific CD4+ T cells contained more viral DNA than memory CD4+ T cells of other specificities in vivo. These data suggest that either infection of DCs is not an important contributor to infection of HIV-2-specific CD4+ T cells in vivo, or that infection of DCs by HIV-2 occurs at a level undetectable in vitro. The frequent carriage of HIV-2 DNA within HIV-2-specific CD4+ T cells, however, does not appear to be incompatible with preserved numbers and functionality of HIV-2-specific CD4+ T cells in vivo, suggesting that additional mechanisms contribute to maintenance of HIV-2-specific CD4+ T cell help in vivo.

CD4 T cell Responses to HIV: Walking a Tightrope between Infection and Protection

Continuing the recent theme of covering issues raised by the Merck HIV vaccine trial results, one longstanding concern with T cell-based vaccines is that they might create more CD4 T cell targets for HIV infection. HIV has a well-described proclivity for activated, dividing CD4 T cells, as is evinced by studies showing that CD4 T cell activation by licensed vaccines and co-infections can often lead to a transient increase in viral load in HIV-infected individuals.

Immunization of an HIV-negative person with a T cell-based HIV vaccine creates a pool of “memory” CD4 and CD8 T cells that specifically recognize the HIV components included in the vaccine. If that individual is subsequently exposed to HIV, the HIV-specific CD4 T cells will respond by becoming activated. On one hand, this is a good thing in that these CD4 T cells help initiate an immune response by assisting HIV-specific CD8 T cells in their efforts to kill virus-infected cells and by signaling to B cells to make antibodies. On the other hand, these activated HIV-specific CD4 T cells also represent ideal targets for the virus and some proportion will inevitably become infected. One grim possibility is that infection of even a small proportion of the HIV-specific memory CD4 T cell response could be enough to compromise the overall immune response to HIV, rendering vaccination useless.

However, in the absence of memory CD4 T cells targeting HIV, the first-responders in a person exposed to HIV are naïve (inexperienced) CD4 T cells that recognize the virus. Naïve CD4 T cells have to become activated and undergo multiple divisions in order to develop into memory CD4 T cells, and the few studies that have looked at this issue have found that activated naïve CD4 T cells are even more susceptible to HIV infection than activated memory CD4 T cells. Furthermore, a recent study found that developing memory CD4 T cells are preferentially infected with HIV during acute infection. These findings, although very limited, suggest that having vaccine-induced memory CD4 T cells on board prior to exposure might confer an advantage. Additionally, particular types of memory CD4 T cell may be less susceptible to infection than others; recently presented data has shown that CMV-specific memory CD4 T cells that make the chemokine MIP-1alpha resist HIV infection significantly better than memory CD4 T cells that do not.

The only study to look closely at this issue in the SIV model was published last year in the Journal of Experimental Medicine. Joseph Mattapallil and colleagues studied macaques that were immunized with a DNA prime/adenovirus vector boost regimen encoding SIVmac239 Env and a Gag-Pol fusion protein. Animals were challenged with SIVmac251(an uncloned virus with similar virulence to SIVmac239) a month after their last booster immunization. Post-infection viral loads were significantly lower in vaccinated animals (by approximately 1 log) compared to controls and vaccination was also associated with a relative preservation of gut CD4 T cells (controls lost 50-80% of this CD4 T cell population, vaccine recipients 5-20%). Mattapallil and colleagues specifically evaluated the extent to which SIV infected SIV-specific CD4 T cells; in both vaccine recipients and controls, SIV-specific CD4 T cells were preferentially infected to a similar extent (containing ~2-fold more SIV DNA than CD4 T cells of other specificities). The only difference was that the preferential infection of SIV-specific CD4 T cells was detectable a few days earlier in vaccinated macaques. The researchers conclude that the infection of SIV-specific CD4 T cells did not abrogate the virological and immunological benefits of vaccination in this study.

At worst, the Merck results could indicate that these findings in the macaque/SIV model do not translate to HIV-infected people, and that the susceptibility of CD4 T cells to HIV infection is an Achilles heel for all vaccination approaches. Among the less pessimistic possibilities, it could be the case that the vaccine-induced memory CD4 T cells were overly susceptible to HIV infection either because of their activation state at the time of exposure (see prior post on resting vs, activated memory T cells) or because they lacked the ability to make potentially protective substances like MIP-1alpha. The breadth of the memory CD4 T cell response could also come into play; Merck's vaccine induces responses only to Gag, Pol and Nef, meaning naive CD4 T cells would be responsible for responding to the HIV proteins not included in the vaccine (Env, Tat, Rev, Vif, Vpu and Vpr). This issue of HIV infection of virus-specific CD4 T cell responses is among the many that will need to be explored as the Merck results are investigated in detail.

Mechanisms of T cell Dysfunction in HIV Infection

Daniel Kaufmann’s study regarding CTLA-4 expression on HIV-specific CD4 T cells, which was presented at the Keystone meeting back in March (see this prior posting for a detailed description of the study findings), has now been published in the advance online section of Nature Immunology (UPDATE 10/20/07: the Nature Immunology website has made the full text of the paper available free of charge). Mass General Hospital also issued a press release highlighting the publication of the paper. Expression of CTLA-4 can be associated with T cell dysfunction and anti-CTLA-4 strategies are being explored in people with cancer to ascertain if tumor-specific T cell responses can be beneficially enhanced. Kaufmann and colleagues suggest that perhaps these strategies should also be studied in the context of anti-HIV immunotherapy.

There are some broad similarities between CTLA-4 and PD-1, another molecule upregulated in the context of T cell dysfunction (specifically, a type of dysfunction called exhaustion), but PD-1 levels are typically elevated on HIV-specific CD4 and CD8 T cells while Kaufmann’s study shows that elevated CTLA-4 expression is restricted to HIV-specific CD4 T cells.

Another recent paper on PD-1 expression in HIV spotlights some of the potential confounding issues involved in exploring the role of these molecules in T cell dysfunction. The study shows that PD-1 can be transiently expressed during normal T cell activation and therefore researchers need to carefully distinguish between PD-1 expression as a marker of T cell activation and PD-1 expression as a marker of T cell exhaustion. The study authors note that the correlations between PD-1 expression and viral load that have been reported previously may relate to the close correlation they observed between CD38 (a T cell activation marker) and PD-1 expression.

Nat Immunol. 2007 Sep 30; [Epub ahead of print]

Upregulation of CTLA-4 by HIV-specific CD4(+) T cells correlates with disease progression and defines a reversible immune dysfunction.

Kaufmann DE, Kavanagh DG, Pereyra F, Zaunders JJ, Mackey EW, Miura T, Palmer S, Brockman M, Rathod A, Piechocka-Trocha A, Baker B, Zhu B, Le Gall S, Waring MT, Ahern R, Moss K, Kelleher AD, Coffin JM, Freeman GJ, Rosenberg ES, Walker BD.

In progressive viral infection, antiviral T cell function is impaired by poorly understood mechanisms. Here we report that the inhibitory immunoregulatory receptor CTLA-4 was selectively upregulated in human immunodeficiency virus (HIV)-specific CD4(+) T cells but not CD8(+) T cells in all categories of HIV-infected subjects evaluated, with the exception of rare people able to control viremia in the absence of antiretroviral therapy. CTLA-4 expression correlated positively with disease progression and negatively with the capacity of CD4(+) T cells to produce interleukin 2 in response to viral antigen. Most HIV-specific CD4(+) T cells coexpressed CTLA-4 and another inhibitory immunoregulatory receptor, PD-1. In vitro blockade of CTLA-4 augmented HIV-specific CD4(+) T cell function. These data, indicating a reversible immunoregulatory pathway selectively associated with CD4(+) T cell dysfunction, provide a potential target for immunotherapy in HIV-infected patients.

AIDS. 2007 Oct 1;21(15):2005-2013.

PD-1 expression on human CD8 T cells depends on both state of differentiation and activation status.

Sauce D, Almeida JR, Larsen M, Haro L, Autran B, Freeman GJ, Appay V.

OBJECTIVE AND DESIGN:: PD-1 expression on HIV-specific CD8 T cells was recently reported to reflect functional exhaustion, resulting in uncontrolled HIV-1 replication. Assessing PD-1 expression on T cells may be highly relevant in T-cell immunology and vaccine monitoring. However, this requires us to gain further insights into the significance of PD-1 expression on CD8 T cells in humans. METHODS:: We performed a detailed analysis of PD-1 expression pattern on various CD8 T cell subsets from healthy or HIV infected donors. RESULTS:: PD-1 expression has two facets in vivo. On the one hand, it is linked to T-cell differentiation: PD-1 is up-regulated on early/intermediate differentiated subsets, which include HIV and Epstein-Barr virus-specific CD8 T-cell populations, but is down-regulated during late stages of differentiation. On the other hand, it is linked to T-cell activation: on PD-1 positive cells, PD-1 over-expression occurs along with the up-regulation of activation markers such as CD38 or HLA-DR. CONCLUSIONS:: PD-1 expression on CD8 T cells, including those specific for HIV, can be related both to their differentiation stage and their activation status. It is important to consider these findings when assessing the expression of PD-1 on T cells.