Abstract HIV infection is associated with increased risk for atherosclerosis and cardio-vascular disease (CVD). This risk does not subside even when HIV load is suppressed to undetectable levels by combined anti-retroviral therapy (cART). Reasons for persistent risk of CVD in cART-treated subjects are not fully understood. Although cART metabolic effects may drive some of this excess risk, this cannot be the only or the main reason, since new generation of anti-retroviral drugs have reduced metabolic side-effects, and cART-naïve HIV-infected subjects also have increased atherosclerotic CVD risk. We, and others, have shown that some of the virus-mediated CVD risk may involve dysregulation of high density lipoprotein structure and reverse cholesterol transport function. Another contributing factor is persistent activation of the innate immune system, presumably due to disruption of the gut barrier and bacterial leakage. However, it remains unclear why these factors do not subside after HIV replication had been brought down to undetectable levels by cART. We hypothesize that HIV-associated atherosclerosis is caused by a two-hit mechanism: HIV replication during the early, untreated phase of infection induces innate memory in myeloid cells increasing their responsiveness to TLR ligands, which persists after cART initiation, so that even low levels of bacterial translocation through the incompletely recovered gut mucosa lead to persistent inflammation and CVD. This hypothesis is based on published literature showing trained innate immunity after exposure of monocytes to fungal cell wall β-glucans, and on our preliminary evidence that HIV Nef drives trained immunity of human monocytes. Here, we propose to test this hypothesis using blood and endoscopic samples from an interventional trial conducted by one of the PD/PIs of this proposal (aim 1). In this trial, an HIV-positive cohort on stable cART is randomized to placebo or teduglutide, a glucagon-like peptide 2 (GLP-2) analog that increases the tightness of gut epithelial barrier and reduces intestinal leakage. We will use advanced epigenomic, transcriptomic and cell biology methods applied to monocytes from these subjects to test if activation of the innate immune memory program is associated with arterial inflammation and coronary atherosclerotic disease, and whether GLP-2 treatment reverses trained memory. In aim 2 we will complement the human cohort studies with experiments in mouse models to determine specific viral and host factors driving the long lasting innate immune memory, and identify their mechanisms of action. Proposed studies will provide mechanistic insight into causes of CVD risk in HIV- infected subjects and will likely inform future therapeutic and preventative strategies to reduce CVD in this population.