Very low (VLDL) and low density lipoproteins (LDL) were isolated from plasma of patients with the E3/3 phenotype which were divided into three groups based on their plasma triglyceride content: low (TG<200 mg/dl, TG(l)), intermediate (200<300 mg/dl, TG(i)300 mg/dl, TG(h)). The protein density (PD) on the VLDL and LDL surface was calculated from lipoprotein composition and protein location was studied by tryptophan fluorescence quenching by I(-) anions at 25 degrees C and 40 degrees C. A comparison of the TG(h) with the TG(l) group revealed a significant (<0.05) increase of the PD parameter as much as 21% for VLDL, but not for LDL where this parameter did not change for any group; generally, PD(LDL) values were 3.2-3.8-fold lower than PD(VLDL). In accordance with this difference, the tryptophan accessibility f in VLDL vs. LDL was lower at both temperatures. There were temperature-induced changes of the f parameter in opposite directions for these lipoproteins. The difference in f value gradually decreased for VLDL in the direction TG(l)TG(i)TG(h) while for LDL there was a U-shaped dependence for these groups. The Stern-Volmer quenching constant K(S-V) which is sensitive to both temperature and viscosity, did not change for VLDL, but K(S-V)(LDL) was 2-3-fold higher for the TG(i) group compared to the other two. The efficiencies of VLDL and LDL binding to the LDL receptor (LDLr) in vitro were compared by solid-phase assay free of steric hindrance observed in cell binding. The maximal number of binding sites did not change for either type of particles and between groups. The association constant K(a) and apolipoprotein (apo) E/apoB mole ratio values all increased significantly for VLDL, but not for LDL, in comparison of the TG(i+h) with the TG(l) group. Based on VLDL and LDL concentrations in serum and on the affinity constant values obtained in an in vitro assay, VLDL concentrations corresponding to 50% inhibition of LDL binding (IC(50)) were calculated in an assumption of the competition of both ligands for LDLr in vivo; the mean values of IC(50) decreased 2-fold when plasma TG exceeded 200 mg/dl. The functional dependences of K(a)(VLDL), IC(50) and apoE content in VLDL (both fractional and absolute) and in serum on TG content in the whole concentration range studied were fitted to a saturation model. For all five parameters, the mean half-maximum values TG(1/2) were in the range 52-103 mg/dl. The efficiency of protein-protein interactions is suggested to differ in normolipidemic vs. HTG-VLDL and apoE content and/or protein density on VLDL surface may be the primary determinant(s) of the increased binding of HTG-VLDL to the LDL receptor. ApoCs may compete with apoE for the binding to the VLDL lipid surface as plasma triglyceride content increases. The possible competition of VLDL with LDL for the catabolism site(s) in vivo, when plasma TG increases, could explain the atherogenic action of TG-rich lipoproteins. Moreover, the 'dual action' hypothesis on anti-atherogenic action of apoE-containing high density lipoproteins (HDL) in vivo is suggested: besides the well-known effect of HDL as cholesteryl ester catabolic outway, the formation of a transient complex of apoE-containing discs appearing at the site of VLDL TG hydrolysis by lipoprotein lipase with VLDL particles proposed in our preceding paper promotes the efficient uptake of TG-rich particles; in hypertriglyceridemia due to the diminished HDL content this uptake seems to be impaired which results in the increased accumulation of the remnants of TG-rich particles. This explains the observed increase in cholesterol and triglyceride content in VLDL and LDL, respectively, due to the CETP-mediated exchange of cholesteryl ester and triglyceride molecules between these particles.