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.

HIV and the Politics of Molecular Epidemiology - Haiti Study Makes Waves

Today has seen a flood of press stories about a new study which strongly suggests that HIV-1 subtype B migrated into the US from Central Africa by way of Haiti. The stories were instigated by a press release from the University of Arizona, publicizing a paper by Michael Worobey and colleagues which is apparently due for online publication in PNAS later this week. Rather awkwardly, the paper does not seem to be online as yet (UPDATE 10/31: the PNAS article is now online, although to the journal's discredit they have not made it open access. UPDATE 11/1: I guess I wasn't the only person who thought the paper should be open access, to their credit PNAS has now made the PDF available free of charge). Anyone interested in finding out more details about the study can watch Michael Worobey's presentation at CROI earlier this year; just go to the webcast page and then scroll down to the session entitled "Oral Abstracts: Epidemiology: Transmission Dynamics and Risk." Click on "Play Video" and Worobey's presentation is the first in the session.

Sadly, the study is re-igniting controversies regarding Haitians somehow being blamed for HIV (as happened, with horrible consequences, in the early 1980s). You can hear an angry response to Worobey's presentation by a Haitian researcher at the end of the CROI webcast, and Jean Pape is quoted in an article in Science by Jon Cohen saying that the study authors "restate prejudices advanced 2 decades ago." However, it seems clear that the researcher's interest is in using molecular techniques to track the history of HIV's spread, not in judging any person, nation or ethnicity. Some press stories have highlighted the suggestion that one transmission event from Haiti around 1969 triggered the US epidemic, and made the leap that - if true - the individual in question must have been Haitian.  But the study makes no such implication; only that the genetics of early HIV sequences provides strong evidence that the virus circulated in Haiti prior to the US and that a single transfer event occurred; if a single individual did transfer the virus between the countries, they could have been of any ethnicity and the transfer was almost certainly an inevitability. The study authors do suggest that it might have been an infected Haitian worker returning from the Congo that transferred HIV from Africa to Haiti around 1966 but this is speculation; the molecular data supports the hypothesis that the virus moved between those two places around that time but says nothing about how the transfer occurred. It is dismal to realize that the stigma associated with HIV because of how it is transmitted remains strong enough to severely complicate the discussion of this type of study in 2007.

The Emergence of HIV/AIDS in the Americas and Beyond

Gilbert MTP, Rambaut A, Wlasiuk G, Spira TJ, Pitchenik AE & Worobey M

(2007) PNAS In press

HIV-1 group M subtype B was the first human immunodeficiency virus discovered and is the predominant variant of AIDS virus in most countries outside of sub-Saharan Africa. However, the circumstances of its origin and emergence remain unresolved. Here we propose a geographic sequence and timeline for the origin of subtype B and the emergence of pandemic HIV/AIDS out of Africa. Using HIV-1 gene sequences recovered from archival samples from some of the earliest known Haitian AIDS patients, we find that subtype B likely moved from Africa to Haiti in or around 1966 [1962-1970] then spread there for some years before successfully dispersing elsewhere. A ?pandemic? clade, encompassing the vast majority of non-Haitian subtype B infections in the US and elsewhere around the world, subsequently emerged after a single migration of the virus out of Haiti in or around 1969 [1966-1972]. Haiti appears to have the oldest HIV/AIDS epidemic outside sub-Saharan Africa and the most genetically diverse subtype B epidemic, which might present challenges for HIV-1 vaccine design and testing. The emergence of the pandemic variant of subtype B was an important turning point in the history of AIDS but its spread was likely driven by ecological rather than evolutionary factors. Our results suggest that HIV-1 circulated cryptically in the US for approximately twelve years before the recognition of AIDS in 1981.

Acute Infection is a Major Contributor to HIV Transmission

A new study and associated commentary are available free online from the Journal of Infectious Diseases:

The Journal of Infectious Diseases 2007;195:000

Primary HIV Infection, Phylogenetics, and Antiretroviral Prevention

Deenan Pillay and Martin Fisher

The Journal of Infectious Diseases 2007;195:000

High Rates of Forward Transmission Events after Acute/Early HIV-1 Infection

Bluma G. Brenner,1 Michel Roger,2 Jean-Pierre Routy,3 Daniela Moisi,1 Michel Ntemgwa,1 Claudine Matte,2 Jean-Guy Baril,4 Réjean Thomas,5 Danielle Rouleau,2 Julie Bruneau,6 Roger Leblanc,7 Mario Legault,8 Cecile Tremblay,9 Hugues Charest,10 Mark A. Wainberg,1 and the Quebec Primary HIV Infection Study Groupa

1McGill AIDS Centre–Jewish General Hospital, 2Centre Hospitalier de Université de Montréal (CHUM)–Hôpital Notre-Dame, 3McGill University Health Centre, 4Clinique Médicale du Quartier Latin, 5Clinique Médicale l'Actuel, 6CHUM–Hôpital St. Luc, 7Clinique Médicale Goldberg, LeBlanc, & Rosengren, 8Fonds de la Recherche en Santé du Québec–SIDA Network, 9CHUM–Hôtel Dieu, and 10Institut National de Santé Publique du Québec, Montreal, Canada

Background: A population-based phylogenetic approach was used to characterize human immunodeficiency virus (HIV)–transmission dynamics in Quebec.

Methods: HIV-1 pol sequences included primary HIV infections (PHIs; <6 months after seroconversion) from the Quebec PHI cohort (1998–2005; n = 215) and the provincial genotyping program (2001–2005; n = 481). Phylogenetic analysis determined sequence interrelationships among unique PHIs (n = 593) and infections from untreated (n = 135) and treated (n = 660) chronically infected (CI) potential transmitter populations (2001–2005). Clinical features, risk factors, and drug resistance for clustered and nonclustered transmission events were ascertained.

Results: Viruses from 49.4% (293/593) of PHIs cosegregated into 75 transmission chains with 2–17 transmissions/cluster. Half of the clusters included 2.7 ± 0.8 (mean ± SD) transmissions, whereas the remainder had 8.8 ± 3.5 transmissions. Maximum periods for onward transmission in clusters were 15.2 ± 9.5 months. Coclustering of untreated and treated CIs with PHIs were infrequent (6.2% and 4.8%, respectively). The ages, viremia, and risk factors were similar for clustered and nonclustered transmission events. Low prevalence of drug resistance in PHI supported amplified transmissions at early stages.

Conclusions: Early infection accounts for approximately half of onward transmissions in this urban North American study. Therapy at early stages of disease may prevent onward HIV transmission.

Initial Events in Vaginal HIV Transmission

An open access article just published in the journal Immunity offers new insights into the vaginal transmisison of HIV. A group of researchers from the University of Washington report the development of a human organ culture system that allows the behavior of HIV to be tracked in intact, viable vaginal outer epithelium. The researchers, led by Florian Hladik, use the system to show that HIV initially targets two different cell types: CD4 T cells, which it enters by direct binding to CD4 and the CCR5 coreceptor leading to productive infection, and Langerhans cells, which it enters via a process called endocytosis (in which the cell draws in material from outside). After HIV is endocytosed by the Langerhans cells it remains viable for several days in an internal compartment of the cell called the cytoplasm, but productive infection does not ensue. Previous research had suggested that HIV infection of CD4 T cells occurs only after transfer of virus from Langerhans cells but this study demonstrates clearly that the virus can enter both cell types simultaneously. About half the CD4 T cells from five tissue donors showed direct binding of virus and this proportion was reduced to ~6% in the presence of antibodies to CCR5 and CD4.

Langerhans cells are antigen-presenting cells (APCs) and their job is to transport foreign material to the lymph nodes for presentation to T cells, so these data also support previous evidence suggesting that APCs can act as Trojan horses that bring HIV into contact with T cells during the process of antigen presentation. Hladik and colleagues found that both infected CD4 T cells and Langerhans cells carrying HIV migrated out of the vaginal epithelium. This finding strongly suggests that, in vivo, these cells are capable of migrating to lymphoid tissue where APC/CD4 T cell interactions and CD4 T cell activation amplify and spread HIV infection systemically.

The researchers conclude by noting that the results have implications for strategies aiming to block vaginal transmission of HIV. Such approaches will likely need to try and block uptake of virus by Langerhans cells in addition to preventing entry into CD4 T cells.