Molecular weight determination of lipoprotein(a) [Lp(a)] in solutions containing either NaBr or D2O: relevance to the number of apolipoprotein(a) subunits in Lp(a).

Molecular weight determination of low-density lipoprotein (LDL) is usually performed in solutions containing high concentrations of salt (up to 13.4 M NaBr) by sedimentation velocity and diffusion experiments, because it does not preferentially bind salt or water. Considering that lipoprotein(a) [Lp(a)] is structurally similar to LDL, differing only by the presence of Apo(a), the molecular weight, M, of Lp(a) has also been measured in solutions containing high concentrations of NaBr. We questioned the suitability of this practice by comparing the apparent molecular weight, Mapp, and partial volume, phi', of Lp(a) determined by sedimentation and flotation equilibrium in a three-component system containing NaBr with the analogous parameters, M and partial specific volume, v, determined in a two-component system containing D2O. LDL served as a control. In agreement with previous findings obtained with different methods, our results indicate no significant differences in M and v of four different LDL samples and apparently no significant preferential binding of solvent components. In contrast, values of Mapp and phi' of Lp(a) evaluated in NaBr are significantly greater than M and v. Preferential binding of solvent components appeared to be a function of Apo(a) mass or the number of kringle IV domains, as expressed by increasing percentage differences between the two sets of parameters, ranging from 4 to 13% in M and 0.2 to 0.5% in v of Lp(a) species having Apo(a) with 15-27 kringle IV domains. Furthermore, our results indicate that the variable Apo(a) kringle IV domains are more involved in this process than the constant domain of Apo(a). These findings indicate that the Lp(a) molecular weight should be determined in D2O and that high concentrations of NaBr should be avoided as their use would lead to overestimated molecular weights and partial specific volumes. Application of this method to the question of how much Apo(a) is released upon the reduction of Lp(a) led to the conclusion that Lp(a) contains only one Apo(a) molecule.