Studies were carried out to determine whether apolipoprotein (apo) A-II, like apoA-I, can recruit phospholipid and cholesterol from cell membranes, thereby forming nascent apoA-II-specific HDL. ApoA-II and apoA-I were purified from plasma and each was incubated with CHO cells at a concentration of 10 micrograms/ml. Lipid-containing complexes were isolated from the medium in both cases; the composition of the apoA-II- and apoA-I-specific complexes were similar where percent protein, phospholipid, and cholesterol were 35 +/- 3, 38 +/- 2, and 25 +/- 1 for apoA-II, respectively, and 40 +/- 2, 35 +/- 1, and 24 +/- 2 for apoA-I, respectively. On a per mole of apolipoprotein basis, apoA-I recruited significantly more phospholipid and cholesterol than dimeric apoA-II suggesting that apoA-I with its greater number of alpha helices binds more lipid. By electron microscopy, nascent apoA-II- and apoA-I-specific particles were predominantly discoidal in morphology. ApoA-II complexes were unique in their nondenaturing polyacrylamide gradient gel size distribution as six distinct populations of particles with diameters of 8.1, 9.3, 10.4, 11.8, 13.1, and 14.6 nm were routinely noted, compared with apoA-I which formed only three major populations with diameters of 7.3, 9.2, and 11.0 nm. Nascent apoA-I complexes incubated with purified lecithin:cholesterol acyltransferase (LCAT) were transformed into predominantly 8.4 nm particles. The latter is similar in size to plasma HDL3a, LpA-I particles, suggesting that extracellularly assembled apoA-I-lipid complexes can directly give rise to a major plasma LpA-I subpopulation upon interaction with LCAT. Unlike apoA-I, apoA-II-lipid complexes could not serve as substrates for LCAT and did not undergo transformation. This study also demonstrates, for the first time, that apoA-II and apoA-I show a preference in phospholipid recruitment from membranes. Although phosphatidylcholine is the major phospholipid removed by both apolipoproteins, apoA-II preferentially recruits phosphatidylethanolamine (PE) as its second most abundant phospholipid while apoA-I recruits sphingomyelin. As PE is usually associated with the inner leaflet of the membrane, it is likely that dimeric apoA-II, compared with apoA-I, can penetrate farther into the membrane and extract PE. This ability of apoA-II to insert more deeply into the lipid milieu may explain the known ability of apoA-II to resist dissociation from the mature HDL particle.