Plasma Thrombospondin 1 Is Increased and Associated with Markers of Vasculopathy In Sickle Cell Disease

Abstract 946

Platelets are activated in patients with sickle cell disease (SCD), particularly in those with pulmonary hypertension (PH), high levels of intravascular hemolysis and during vaso-occlusive crisis [Villagra et al. Blood 2007 110:2166; Hu et al. Blood 2010 doi 10.1182/blood-2010-01-267112]. One of the major factors released by activated platelets is the platelet glycoprotein thrombospondin 1 (TSP1). Our group has discovered that TSP1 inhibits nitric oxide (NO) signaling through interaction with its ubiquitous receptor CD47 [Isenberg et al. J Biol Chem 2006; 281:26069] and has been implicated in atherogenesis, intimal hyperplasia, hypertension and PH in pre-clinical animal models [Moura R et al.; Circ Res 2008; 103:1181]. We therefore hypothesized that elevated TSP1 levels in SCD patients with activated platelets could inhibit NO-signaling via a novel TSP1-dependent mechanism and impair vascular function, predisposing patients to chronic vascular disease. We have tested the plasma collected from patients (n=96) and self-reported African American control subjects (n=19) enrolled in the NIH-pulmonary hypertension screening trial for levels of TSP1. This is one of the most comprehensively phenotyped cohorts of plasma samples from patients with SCD in steady state. Plasma from healthy controls and patients was assessed in parallel for levels of TSP1 by ELISA. We found a signficant difference in TSP1 levels between patients with SCD in steady state (median (IQ range) = 888.4 ng/ml (574.0-1161.0) and healthy controls (median (IQ range) = 515.5 ng/ml (398.6-701.9)) (Wilcoxon two-sample p=0.008). Unexpectedly, TSP1 levels positively correlated with LDL cholesterol (r=0.25, p=0.03), while the correlation with total cholesterol (r=0.20, p=0.07), and apolipoprotein B (r=0.20, p=0.07) did not reach statistical significance. Interestingly, we also observed a trend between systolic blood pressure and TSP values (p=0.07). TSP1 positively correlated with white blood cell count (r=0.28, p=0.01). As expected based on TSP1 localization in platelet alpha granules, TSP1 levels strongly correlated with platelet count (r=0.56, p=0.0001). LDL remained a significant predictor of elevated TSP levels in a logistic regression model (p=0.03). We next validated these results in prospectively collected plasma samples from patients with SCD at the University of Pittsburgh Medical Center presenting in steady state (n=13) and vaso-occlusive crisis (n=7), as well as healthy controls (n=8). To minimize the confounding effect of platelet activation and degranulation ex vivo, possibly leading to spurious elevation of TSP1 levels, we immediately processed these samples in laboratory facilities adjacent to the collection sites with a double centrifugation step to obtain platelet-poor plasma. Patients with SCD in crisis had significantly elevated TSP1 levels (median (IQ range) = 544.5 ng/ml (446.4-616.9) as compared to controls (median (IQ range) = 226.6 ng/ml (30.1-287.6). In this smaller cohort, patients in steady state SCD did not have significantly elevated TSP1 values as compared to healthy controls. The association of elevated TSP1 with dysregulated lipoprotein levels may indicate a role in SCD vascular pathobiology and is the subject of current studies. Interestingly, the trend between elevated blood pressure and TSP levels is consistent with our recent report that soluble TSP1 inhibits endothelial based NO formation and endothelial dependent vasodilation [Bauer et al, Cardiovasc Res 2010 doi10.1093/cvr/cvq218]. This association suggests a possible role for TSP1 in limiting vascular NO signaling in SCD. Ongoing prospective analysis will determine the strength of these preliminary associations. We are also conducting experiments to explore whether blockade of TSP1-CD47 in the BERK murine model of SCD restores endothelial function and modulates vasculopathy.