Effects of pancreas transplantation on distribution and composition of plasma lipoproteins

Abstract

In type I (insulin-dependent) diabetic patients, peripheral hyperinsulinemia due to subcutaneous insulin treatment is associated with increased high-density lipoprotein (HDL) cholesterol, and also with an altered surface composition of HDL. Pancreas grafts also release insulin into the systemic rather than into the portal venous system, giving rise to pronounced peripheral hyperinsulinemia. We hypothesized that if peripheral hyperinsulinemia is responsible for high HDL cholesterol and/or altered surface composition of HDL in diabetic subjects, similar changes in the lipid profile should be present in pancreas-kidney transplant recipients (PKT-R). Using zonal ultracentrifugation, we isolated HDL2, HDL3, very-low-density lipoprotein (VLDL), intermediate-density lipoprotein (IDL), and low-density lipoprotein (LDL) from fasting plasma of 14 type I diabetic PKT-R, eight nondiabetic kidney transplant recipients (KT-R), and 14 healthy control subjects and determined the level and composition of the above lipoproteins. HDL2 cholesterol was increased in PKT-R as compared with KT-R and healthy controls (both P < .05), whereas HDL3 cholesterol was unchanged. However, an altered lipoprotein surface composition was evident in PKT-R: HDL2, HDL3, and LDL were enriched in unesterified cholesterol ([UC] PKT-R v KT-R, P=.13, P < .005, and P < .05, respectively; PKT-R v controls, all P < .005); HDL2 was enriched in phospholipids; and LDL was depleted of phospholipid. KT-R, in contrast, showed no changes in lipoprotein surface composition but a substantial triglyceride enrichment of HDL2 as compared with PKT-R and healthy controls (both P < .05). LDL size as determined by gradient gel electrophoresis was increased in PKT-R compared with controls (P < .005). The plasma concentration of cholesteryl ester (CE) transfer protein (CETP), involved also in phospholipid transfer, was increased in both transplant groups compared with healthy controls (both P < .05). Insulin concentrations in fasting plasma were directly related to CETP levels and to the weight-percentage of UC in HDL3, and inversely to the weight-percentage of phospholipids in LDL (all P < .05). We explain the increase in HDL2 cholesterol and LDL size in PKT-R by their high lipoprotein lipase (LPL) activity conferring an excellent capacity to clear chylomicron triglycerides. Effective handling of postprandial triglycerides, high HDL2 cholesterol, and predominance of LDL pattern A, respectively, are established indicators of a low risk of atherosclerosis. However, it is presently unclear what effects the compositional changes on the surface of HDL and LDL may have on cardiovascular risk in clinically stable PKT-R.