CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

355 Clinical Administration of Vorinostat Increases NK Cell Capacity to Produce IFN-g Carolina Garrido ; Natalia Soriano-Sarabia; Brigitte Allard; Katherine Sholtis; Nancie M. Archin; David M. Margolis Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA

Background: Vorinostat (VOR) can upregulate HIV RNA expression in latently infected resting CD4 + T cells from treated aviremic HIV-infected patients. However, to achieve viral eradication, latency reactivation must be followed by immune clearance, and thus it is essential to investigate whether reactivation therapies have an impact on immune function. While VOR effect on T cell function has been studied, here we evaluate the impact of in vivo administration of VOR on Natural Killer (NK) cell function and phenotype. Methods: Five participants received 22 doses of VOR over 8 weeks (Archin JID, 2014). Cryopreserved PBMC from these patients were used to assess the impact of VOR on CD4, CD8,

NK and NKT cell frequency, proliferation and activation. In addition, NK cells were isolated to analyze NK phenotype, including the expression of activating receptors CD16, NKG2D, NKp30, NKp46 and DNAM-1. NK degranulation (CD107a expression) and IFN-g production after culture with the target cell line K562 was measured. Cells were studied before VOR administration, after 11 doses and after 22 doses, and differences assessed using the Mann Whitney test. Results: The frequency of CD3 + CD4 + , CD3 + CD8 + or CD3 + CD56 + cells were unaltered by VOR. However, a modest increase in NK cell (CD3 - CD56 + ) frequency was observed after VOR treatment (p=0.059 and p<0.05 after 11 and 22 doses). Interestingly, we observed a significant decrease of Ki67 expression in all cell subpopulations after 11 VOR doses, which persisted after 22 doses (p<0.01). Expression of the activation marker CD69 remained stable during VOR treatment. Phenotypic analysis on isolated NK cells showed that CD16, DNAM-1 and NKp30 expression did not change upon VOR exposure, while NKG2D and NKp46 expression increased (p<0.05). Regarding NK cell function, we observed a trend towards an increase in degranulation after VOR exposure (p=NS), and remarkably, a significant increase in IFN-g production after VOR administration (p<0.01). Conclusions: NK cell capacity to produce IFN-g was improved after in vivo treatment with VOR. In addition, activating receptors NKG2D and NKp46 were upregulated. Interestingly, Ki67 expression was significantly reduced after 11 and 22 doses of in vivo VOR in all cell subpopulations. These results suggest in vivo VOR dosing is not reducing NK cell function, and could even provide some enhancement. We are currently testing this hypothesis by evaluating the potential of NK cells exposed to VOR to reduce viral production in latency clearance assays.

356 CCR5 Gene-Edited Cells Undergo Positive Selection in SHIV-Infected Nonhuman Primates ChristopherW. Peterson 1 ; JianbinWang 2 ; Michael Holmes 2 ; Patricia Polacino 3 ; Shiu-Lok Hu 3 ; Robert Murnane 4 ; Audrey Baldessari 4 ; Hans-Peter Kiem 1 1 Fred Hutchinson Cancer Rsr Cntr, Seattle, WA, USA; 2 Sangamo BioScis, Richmond, CA, USA; 3 Univ of Washington, Seattle, WA, USA; 4 Washington Natl Primate Rsr Cntr, Seattle, WA, USA Background: Nuclease-mediated gene editing holds great promise in the cure of HIV infection, but the efficacy of this approach in patients is unclear. We have targeted the CCR5 locus in hematopoietic stem cells (HSCs) and assessed engraftment following autologous transplant in the pigtailed macaque. CCR5 disruption in this model should directly protect against infection with simian/human immunodeficiency virus (SHIV). We are evaluating the extent to which CCR5-disrupted cell progeny impact SHIV infection in naïve animals, and viral reservoirs in previously infected, combination antiretroviral therapy (cART)-suppressed animals. Methods: Animals are challenged with CCR5-tropic, HIV enveloped SHIV, and suppressed by three-drug cART following viral set point. Zinc Finger Nucleases (ZFNs) are used to disrupt CCR5 in autologous HSCs; these stem cells and their progeny are subsequently measured ex vivo and in vivo . In previously infected and suppressed animals, assays are conducted to measure the size of the latent SHIV reservoir before and after transplant. Results: In SHIV-naïve animals, we observe approximately 5% steady state bulk disruption of CCR5 in vivo , which persists following SHIV challenge. This approach is equally feasible in SHIV-infected, cART-suppressed animals. Virus-dependent selection for CCR5-disrupted HSC progeny is observed, namely in memory T-cell subsets. Viral reservoir assays demonstrate that sites of viral persistence are present in cART-suppressed animals, and are impacted by the transplant procedure. Conclusions: We have built on our unprecedented demonstration of successful long-term engraftment of CCR5 gene-edited HSCs in vivo by showing that these cells can undergo virus-dependent positive selection. We are currently developing gene-editing approaches to enrich for CCR5-edited cells without the need for ongoing viral replication, using adeno-associated virus (AAV) to knock in a chemoselection marker at the disrupted CCR5 locus. This should allow for significant enrichment of infection- resistant cells in vivo under the cover of ongoing cART. These results have important implications for cART-independent control of viremia in HIV + in patients, and should be a promising component of combinatorial HIV cure strategies. 357 Predicting Determinants of Long-TermHIV Control With Gene Therapy Strategies Alison L. Hill 1 ; Lily Hu 1 ; Louise Hogan 2 ; Martin A. Nowak 1 ;Timothy J. Henrich 2 1 Harvard Univ, Boston, MA, USA; 2 Univ of California San Francisco, San Francisco, CA, USA Background: Gene therapy to render lymphocytes resistant to HIV infection is a proposed strategy to achieve long-term antiretroviral-free remission (e.g. CCR5 modification or delivery of antiviral gene products). Preliminary in vivo studies of gene therapy, however, have had limited success, and there is a need to elucidate the conditions under which modified cells have sufficient selective advantages to reduce target cell density below a critical level required to maintain infection. Methods: We designed a mathematical model of the competition between wildtype and genetically-modified CD4 T cells in vivo and the accompanying dynamics of HIV infection. The model was parameterized using data on lymphocyte kinetics, HIV viral dynamics, and the effects of modification. The dynamics were analyzed to determine the conditions under which gene therapy strategies will be successful, and to better understand factors that could be manipulated to improve outcomes. Results: In summary, complete control of HIV with these strategies is difficult. Under a range of model structures and parameters, the most likely outcome is that modified cells co-exist with wild-type cells below the level required to prevent HIV persistence. To obtain viral control off ART, edited CD4+ T cells must have a higher proliferation rate or a longer lifespan even in the absence of virus, or, edited hematopoietic stem cells must be included. The enrichment level of edited cells is highly dependent on the strength of competition between cells for homeostatic proliferation signals. Interestingly, lower thymic contribution to wild-type T cell levels makes invasion easier and engraftment higher, promoting viral control. Higher viral fitness in wildtype cells can make it easier for edited cells to expand initially but harder to control infection. The potential benefit can be boosted if edited cells are also resistant to causes of bystander cell death that are not viral-load-dependent, or if they provide enhanced immune effector function. Using this model to interpret recent clinical data, observed outcomes may be more consistent with transfusion-mediated changes in cell dynamics than due to resistance of edited cells. Conclusions: A mathematical model of gene therapy strategies for HIV demonstrates that viral control is possible only under a narrow range of conditions, and that further measurement and manipulation of immunological dynamics during engraftment may be necessary to improve outcomes.

Poster Abstracts

136

CROI 2016