CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

392 Effect of cART on Functional Connectivity in Cart-Naïve HIV-Infected Individuals Yuchuan Zhuang 1 ; MadalinaTivarus 2 ; Xing Qiu 2 ; LuWang 2 ; Emily Cosimano 2 ; AliciaTyrell 2 ; Amneris Luque 2 ; Mark E. Mapstone 2 ; Jianhui Zhong 2 ; Giovanni Schifitto 1 1 Univ of Rochester, Rochester, NY, USA; 2 Univ of Rochester Med Cntr, Rochester, NY, USA Background: Neuroimaging biomarkers provide an opportunity to investigate HIV-associated CNS injury at the time when clinical changes may be silent. In this study we assessed whether virologic response to combination antiretroviral therapy (cART) in naïve subjects was associated with changes in resting state fMRI. Methods: Subjects with confounding CNS disorders and history of CNS infection, other than HIV, were excluded. All subjects underwent a detailed neurocognitive and functional assessment. HIV infected subjects were scanned before starting cART and 12 weeks after initiation of treatment. Imaging protocol/analyses: Resting state fMRI was acquired on a 3T Siemens MAGNETOM Trio MRI scanner. Group ICA was used to identify the resting state networks and to determine differences in functional connectivity (FC) within the Default Mode Network (DMN) (Figure B). The anterior cingulate cortex (ACC) and posterior cingulate cortex (PCC) were used as ROIs to evaluate changes in FC between DMN and other brain areas. Statistical analyses were conducted using two-sample t test or paired t-test (HIV+ baseline vs. HIV+ after 12-week treatment). Multiple comparison correction was performed using values generated by AlphaSim (voxel p<0.01, cluster size > 64, which corresponds to a corrected p<0.05). Results: Cohort: 17 HIV+male, mean age 33 ± 3 years, 17 HIV- (10 female and 7 male) mean age 32 ± 3 years. At baseline, 10 of the HIV+ subjects had normal cognitive performance, 6 had asymptomatic impairment and 1 had mild neurocognitive disorder. The mean CD4 count and viral load at baseline were 479 ± 48 mm 3 and 121,804 ± 43586 copies/ml respectively. After 12 weeks, mean CD4 count and viral load were 636 ± 55/mm 3 and 899 ± 861 copies/ml. Imaging results: FC within the DMN network was significantly lower in HIV+ subjects at baseline as compared to HIV- subjects (Figure A). There were no significant differences in FC within DMN between HIV+ after 12 week treatment and HIV- subjects. There was a significant increase in FC between DMN and occipital lobe in the HIV+ after 12 week treatment as compared to HIV+ subjects at baseline (Figure C: seed in PCC, D: seed in ACC). Conclusions: Our results suggest that in relatively young cART naïve subjects with relatively preserved immune function, there are signs of decreased functional connectivity that tend to improve after 12 weeks of treatment.

393 HIV Infection in Astrocytes Via a CD4-Independent, CXCR4-Dependent Mechanism Guanhan Li 1 ; Caroline Anderson 1 ; Eugene O. Major 1 ; Avindra Nath 2 1 NINDS, NIH, Bethesda, MD, USA; 2 NIH, Bethesda, MD, USA

Poster Abstracts

Background: HIV reservoir in the brain represents a major barrier for curing HIV infection. As the most abundant, long-lived cell type, astrocytes play a critical role in maintaining the reservoir; however the mechanism of infection remains unknown. Here, we determine how viral transmission occurs from HIV-infected lymphocytes to astrocytes. Methods: Human astrocytes were exposed to HIV-infected lymphocytes and monitored by live-imaging, confocal microscopy, transmission and 3-demesional electron microscopies. A panel of receptor antagonists was used to determine mechanism of viral entry. A transwell culture systemwas used to test whether the infection of astrocytes could also be established by viral particles that were released from the infected lymphocytes and passed through the pores of transwell membranes. Results: We found that cell-to-cell contact resulted in efficient transmission of X4- or X4R5-using viruses from T lymphocytes to astrocytes. In co-cultures of astrocytes with HIV- infected lymphocytes, the interaction occurred through a dynamic process of attachment and detachment of the two cell types. Infected lymphocytes invaginated into astrocytes or the contacts occurred via filopodial extensions from either cell type, leading to formation of virological synapses. In the synapses, budding of immature or incomplete HIV particles from lymphocytes occurred directly onto the membranes of astrocytes. This cell-to-cell transmission could be almost completely blocked by anti-CXCR4 antibody and its antagonist. Furthermore, we also found that newly produced HIV particles from the infected lymphocytes on the top of transwells could go through the transwell membrane and infect astrocytes via the same mechanism. Conclusions: Cell-to-cell transmission is mediated by a unique mechanism by which immature viral particles initiate a fusion process in a CXCR4-dependent, CD4-independent manner; and newly produced, immature HIV can also infect astrocytes without cell-to-cell contact. These observations have important implications for developing approaches to prevent formation of HIV reservoirs in the brain. 394 A Unique In Vitro Neuropathic Phenotype of Clade D Transmitted/Founder Viruses Analise Gruenewald 1 ; Alexander J. Gill 1 ; George M. Shaw 2 ; Dennis L. Kolson 2 1 Univ of Pennsylvania, Philadelphia, PA, USA; 2 Perelman Sch of Med, Univ of Pennsylvania, Philadelphia, PA, USA Background: Expression of the anti-inflammatory, antioxidative phase II detoxifying enzyme heme oxygenase-1 (HO-1) is decreased in the brain prefrontal cortex of HIV+ individuals in association with increased neuroinflammation and neurocognitive impairment. In vitro HIV infection of monocyte-derived macrophages (MDM) induces HO-1 deficiency and increases release of neurotoxic levels of glutamate. We recently showed that HO-1 deficiency is consistently seen with HIV MDM infection by a broad panel of macrophage-tropic clade B HIV-1 strains (n=13). This suggests that induction of HO-1 deficiency is a highly conserved macrophage-tropic HIV phenotype that could be particularly relevant for neuropathogenesis driven by infection within the CNS myeloid reservoir. Methods: Human MDM (n=4 donors) were infected with viruses derived from infectious molecular clones of each of 5 clade D transmitted/founder (T/F) strains or the clade B 89.6 strain. Viral replication, glutamate concentration, and neurotoxicity of MDM culture supernatants were determined via reverse transcriptase (RT) activity, glutamate assay, and neuron-based MAP2 ELISA, respectively. Data were analyzed by ANOVA with Holm-Sidak post hoc test or Pearson’s correlation.

149

CROI 2016