Our research program aims to identify modifiable factors that promote atherogenesis by increasing levels of atherogenic lipoproteins and/or by impairing HDL?s ability to remove cholesterol from artery wall macrophages. In contrast to most investigators, we first identify proteins and lipoproteins that predict cardiovascular disease risk (CVD) risk in humans and then perform mechanistic studies in mice based on those results. A major component of this approach has been to devise high-throughput state-of-the-art analytical methods for use in clinical studies. We have a particular interest in diabetic atherosclerotic disease because there are no well- established lipoprotein risk factors in type 1 diabetic patients (T1DM) and because type 2 diabetic patients (T2DM) treated with statins still have a substantial risk for CVD. Our compelling preliminary data suggest that small HDL particles altered by the diabetic milieu strongly predict future CVD events in healthy T1DM patients (n=181, P=0.0008). In parallel, we isolated HDL from 19 T2DM subjects and 20 control subjects and showed that small HDL?s cholesterol efflux capacity?a proposed metric of HDL?s cardioprotective effects?was markedly impaired in the T2DM subjects. These observations point to two specific defects in HDL of T1DM and T2DM patients?size and inability to remove cholesterol from the artery wall?that may increase CVD risk. The demonstration that small HDL particles are markedly elevated in healthy T1DM patients who subsequently experience CVD events contradicts the dogma that high levels of HDL are always cardioprotective. A major goal of this proposal is to identify the mechanisms underlying the association of small HDL with CVD. First, we propose to confirm and extend our findings in two large clinical studies of T1DM and T2DM subjects. We also plan to explore the hypothesis that lipolysis of triglyceride-rich lipoproteins couples the generation of highly atherogenic remnant lipoproteins (RLPs) with remodeling of HDL into small, dysfunctional particles. Second, we will use a mouse expressing human APOA1, HDL?s major protein, to test the role of phospholipid transfer protein (PLTP) in the formation of RLPs and small HDL particles in diabetic mice. We base this approach on the demonstration that PLTP drives the generation of both small and large HDL particles in mice, that PLTP increases hepatic production of VLDL (the precursor of RLPs), and that PLTP associates in Mendelian randomization studies with increased CVD risk in concert with increased HDL-cholesterol levels and a larger percentage of small HDL. Importantly, PLTP also predicted incident CVD in our study of T1DM patients. Collectively, our proposed experiments will provide a powerful test of the hypothesis that a high level of small, dysfunctional HDL is a marker, and perhaps a mediator, of increased CVD risk in patients with diabetes.
