Cyclosporine A transfer between high- and low-density lipoproteins: independent from lipid transfer protein I-facilitated transfer of lipoprotein-coated phospholipids because of high affinity of cyclosporine a for the protein component of lipoproteins

Abstract

The objectives of this study were to determine if lipid transfer protein I (LTP I)-facilitated phospholipid (PC) transfer activity regulates the plasma lipoprotein distribution of cyclosporine (CSA) and if the association of CSA with high-density lipoproteins (HDL) is due to the high protein and/or alterations in coat lipid content of HDL. To assess if LTP I-facilitated PC transfer activity regulates the plasma lipoprotein distribution of CSA, (14)C-PC- or (3)H-CSA-enriched HDL or low-density lipoproteins (LDL) were incubated in T150 buffer [pH 7.4, containing a (14)C-PC- or (3)H-CSA-free lipoprotein counterpart +/- exogenous LTP I (1.0 microg protein/mL)] or in delipidated human plasma that contained 1.0 microg protein/mL of endogenous LTP I in the presence or absence of a monoclonal antibody TP1 (30 microg protein/mL) directed against LTP I for 90 min at 37 degrees C. To assess the influence of HDL subfraction lipid composition and structure on the plasma distribution of CSA, CSA at 1000 ng of drug/mL of plasma was incubated in human plasma pretreated for 24 h with a lecithin:cholesterol acyltransferase (LCAT) inhibitor, dithionitrobenzoate (DTNB; 3 mM). To assess the binding of CSA to apolipoproteins AI, AII, and B, increasing concentrations of CSA were added to a constant concentration of either apolipoprotein AI, AII, or B. Equilibrium dialysis was used to determine free and bound fractions and Scatchard plot analysis was used to determine binding coefficients. To assess the influence of hydrophobic core lipid volume on the plasma distribution of CSA, CSA was incubated in plasma from patients with well-characterized dyslipidemias. The hydrophobic core lipid volume (CE + TG) within each lipoprotein subfraction was correlated to the amount of CSA recovered in each plasma sample from the different human subjects. The percent transfer of PC from LDL to HDL was different than the percent transfer of CSA in T150 buffer or human plasma source. In the presence of TP1, only PC transfer from LDL to HDL decreased. For plasma incubated with CSA and separated into HDL(2) and HDL(3), 35-50% of drug originally incubated was recovered in the HDL(3) fraction, with the remaining drug being found within the other fractions. When CSA was incubated in plasma pretreated with DTNB, the percentage of CSA recovered in the HDL(3) and HDL(2) fractions was not significantly different compared with that in the HDL(3) and HDL(2) fractions from untreated control plasma. CSA distribution into HDL inversely correlated with the hydrophobic core lipid volume of HDL, whereas distribution into LDL and triglyceride-rich lipoproteins directly correlated with their respective hydrophobic core lipid volumes. We further observed that CSA has high binding affinity and multiple binding sites with apolipoproteins AI (k(d) = 188.9 nM; n = 2), AII (k(d) = 184.7 nM; n = 2), and B (k(d) = 191 nM; n = 3). These findings suggest that the transfer of CSA between different lipoprotein particles is not influenced by LTP I-facilitated PC transfer activity probably because of the high affinity of CSA for the protein components of HDL and LDL.