CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

binds 3DL1 with a high and 3DS1 with a low intensity. Z27 staining cannot distinguish whether Z27 high NK cells express 3DL1 only or include NK cells that also express 3DS1. We developed an approach to address this question. Methods: Peripheral blood mononuclear cells from 3DL1/S1 htz and, as controls, 3DL1 and 3DS1 homozygotes were stained with anti-CD3, anti-CD56 and Z27. CD3 - CD56 + NK cells were sorted into Z27 high , Z27 dim and Z27 neg sub-populations. RNA was extracted from each subset and cDNA was prepared from RNA. Real-time (RT)-PCR was used to determine the presence or absence of 3DS1 and 3DL1 transcripts in sorted cells. Results: First, we verified the specificity of 3DS1 , 3DL1 primer sets. We verified that the sorting gates distinguishing Z27 dim from Z27 neg populations were correctly set by showing that no 3DS1 transcripts were present in the Z27 neg subset from either 3DS1 hmz or 3DL1/S1 htz. Using RT- PCR and the 3DS1 specific primers we found that the Z27 dim subset of 3DS1 hmz and both the Z27 dim and Z27 high subsets of 3DL1/S1 htz amplified 3DS1 transcripts. We found 3DL1 primers amplified transcripts from Z27 high , but not from the Z27 dim subset of 3DL1/S1 htz. Conclusions: 3DL1 and 3DS1 can be co-dominantly expressed on NK cells. The accepted theoretical paradigm posits that engagement of the inhibitory 3DL1 NK receptor by its ligand, HLA-Bw4, is important in licensing NK cells for function. Engagement of the activating 3DS1 NK receptor should tune down the potency of NK cell licensing and the functional potential of 3DS1+ NK cells. This raises questions regarding the mechanism underlying the association between 3DS1+*80I carriage and HIV control. Our results provide a framework for improving our understanding of the role played by the 3DL1 and 3DS1 receptors in determining the function of NK cell subsets in general and the anti-HIV function of these cells in particular. 278 Disruption of Innate-Like Unconventional T-Cell Subsets in HIV-Infected Children Background: A subset of T cells comprised of mucosal associated invariant T (MAIT), natural killer T (NKT), and γδ T cells exhibit features of innate cells such as invariant TCR and recognition of non-peptide antigens. These unconventional T cell subsets can secrete pro-inflammatory cytokines or display cytotoxic activity in response to bacterial and fungal derived glycolipids or metabolites. HIV+ adults have lower MAIT and NKT and higher γδ T cells in the peripheral blood. Less is known about their changes in HIV-infected children. We sought to determine whether innate-like T cells are disrupted in treated and untreated HIV+ children. Methods: We evaluated peripheral blood samples of 76 perinatally-infected HIV+ and 40 HIV- children between 3-12 years old fromMombasa, Kenya. The HIV+ cohort included 39 ART naïve (ART-) and 37 children on ART (ART+). Cryopreserved PBMCs were thawed and stained with surface antibodies CD3, CD4 and CD8 with V α 7.2 and CD161 (MAIT), γδ TCR, and V α 24J α 18 TCR (NKT) then analyzed by flow cytometry. Plasma sCD14 was measured by ELISA. Statistical analysis was performed on GraphPad Prismwith Mann-Whitney or Spearman’s correlation tests. Results: HIV+ children had decreased levels of MAIT in CD8+ T cells (ART- p=0.0156; ART+ p=0.0137), which correlated positively with CD4:CD8 ratios (r=0.37, p=0.0018). ART+ had lower MAIT levels in CD4+ T cells (p=0.0090); both HIV+ groups maintained CD4-CD8- MAIT cell frequencies and MAIT cell numbers in total T cells. Plasma sCD14 levels were higher in HIV+ (p=0.0068). In HIV+ children CD8+MAIT cells inversely correlated with sCD14 (r= -0.33, p=0.0049). NKT cells were also lower in ART- (p=0.0001) and ART+ (p=0.0140) and correlated with CD4:CD8 (r=0.25, p=0.0417). Conversely, γδ T cells were persistently elevated in HIV+ (ART- p=0.0064; ART+ p<0.0001) and correlated positively with sCD14 (r=0.36, p=0.0025). Interestingly, in HIV+ children γδ T cells negatively correlated with CD8+MAIT (r= -0.36, p=0.0024) and NKT (r= -0.27, p=0.0063) cells and the ratio of γδ T/MAIT cells positively correlated with sCD14 (r=.0.37, p=0.0019). Conclusions: CD8+MAIT and NKT cells are relatively lower in HIV+ children despite ART compared to HIV- children. This decline in innate-like T cells worsens with diminished immune status. Loss of MAIT cells from the periphery may reflect migration to mucosal tissues in response to microbial translocation. Increases in γδ T cells may be a compensatory response to decreased MAIT and NKT cells in HIV infected children. 279 HIV-1 Alters Innate Immune Response to BCG , Which Differs From SIV Infected Mangabeys Melanie Gasper; David Sherman; Donald Sodora Seattle Biomed, Seattle, WA, US Background: Coinfection with M. tuberculosis, as well as other Mycobacteria, is a leading cause of morbidity and mortality for HIV-infected patients. Here we set out to understand functional immune responses to an opportunistic pathogen, M.bovis BCG , during both pathogenic HIV and non-pathogenic SIV infections (natural SIV host sooty mangabeys). M. Bovis BCG is a live-attenuated vaccine strain of Mycobacteria, which has demonstrated opportunistic potential. Methods: Multiparameter flow cytometry and gene expression analysis, utilizing a Nanostring platform (248 genes), was used to evaluate the ability of monocytes and peripheral blood mononuclear cells (PBMC) to respond to 4 hour ex vivo BCG stimulation. KEGG analysis was used to identify pathways altered by changes in gene expression. Uninfected and HIV/SIV infected humans and sooty mangabeys were analyzed. Results: HIV infection was associated with a reduced percentage of whole blood monocytes producing IL-12 in response to BCG (1.8% compared to 4% in uninfected donors; p=0.02). In contrast, SIV+mangabeys had a higher percentage of monocytes producing IL-12 and TNF- α in response to BCG compared to uninfected mangabeys, suggesting a maintained immune response. In addition, principal component analysis showed that although both species mounted a robust response to BCG, HIV-infected donors had a unique BCG-induced gene expression profile that distinguished them from uninfected age-matched donors. In contrast, SIV-infection status did not alter the gene expression clustering pattern between SIV+ and SIV- mangabeys. KEGG analysis demonstrated that genes with increased expression at baseline in HIV+ donors were significantly enriched within the NK cell cytotoxicity pathway (hypergeometric test; p=0.007), this difference was not seen between SIVneg and SIV+mangabeys. Additionally, a statistical interaction between HIV disease status and the response to BCG was found in a number of genes related to NK cell function, including NKp46, IFNg and perforin, suggesting that HIV-infection altered the NK cell response to bacterial stimulation. Conclusions: Overall, these data provide evidence that innate immune responses (including monocyte IL-12 production and NK cell activity) to the BCG opportunistic pathogen are Alka Khaitan 1 ; Adam Kravietz 1 ; IlmetTiina 1 ; Mussa Mwamzuka 2 ; Fatma Marshed 2 ; CihanTastan 1 ; Aabid Ahmed 2 ; Bill Borkowsky 1 ; Derya Unutmaz 1 1 New York University School of Medicine, New York, NY, US; 2 Bomu Hospital, Mombasa, Kenya; 3 New York University School of Medicine, New York, NY, US

Poster Abstracts

likely to be critical for resistance to opportunistic infections following lentiviral (HIV/SIV) infection. 280 High-Level Replication Allows SHIVs to Overcome the Macaque Interferon Response David F. Boyd ; Amit Sharma; Julie M. Overbaugh Fred Hutchinson Cancer Research Center, University of Washington, Seattle, WA, US

Background: SIV/HIV chimeric viruses (SHIVs) encoding HIV-1 envelope (Env) have been studied in macaques as models of HIV-1 infection in humans. Typically, the initial chimeric SHIVs generated in vitro replicate poorly in macaques, but a subset has been successfully adapted to replicate in macaques through animal passage. The resulting SHIVs cause persistent infection in macaques that mimics many aspects of pathogenic HIV-1 infection. The SHIV/macaque model is limited because the few pathogenic SHIVs currently in use

237

CROI 2015