Hepatic endocytosis of various types of mannose-terminated albumins. What is important, sugar recognition, net charge, or the combination of these features.

We synthesized several para-aminophenyl (pap-) mannose-terminated albumins with varying sugar density (Man7-HSA, Man22-HSA, and Man40-HSA) and compared hepatic uptake with (thio-)mannose-terminated bovine serum albumin (Man-43-AI-BSA) The rate of uptake in isolated perfused rat livers was found to be positively correlated with the sugar density (Man40-HSA = Man22-HSA greater than Man7-HSA greater than HSA). Immunohistochemical staining of liver sections showed for both types of neoglycoproteins that uptake occurred in nonparenchymal cells only. Competition experiments with a 500-fold excess of mannan, a known ligand for the mannose/N-acetylglucosamine receptor, that is predominantly localized in endothelial cells, showed complete inhibition of the (thio-)Man43-AI-BSA uptake. In the case of (pap-)mannose-terminated albumins, however, the extent of inhibition by mannan was moderate and decreased markedly with increasing sugar density, being only 20% for (pap-)Man40-HSA. Therefore, we hypothesized that one or more additional removal systems contributed to the clearance of these (pap-)mannose glycoproteins. We found that net negative charge of the (pap-)mannose albumins clearly increased with increasing sugar density, as shown on fast protein liquid chromatography anion-exchange chromatograms. To determine whether the scavenger receptor system that is also mainly present on endothelial cells is involved, we performed competition studies with strongly negatively charged substrates, such as dextran sulfate and formaldehyde-treated human serum albumin (fHSA). An excess of dextran sulfate (500 kDa), indeed blocked the (pap-)mannose-albumin uptake for more than 95%. Dextran sulfate completely inhibited the hepatic uptake of mannan as well, indicating that the polyanion does not discriminate between the scavenger system and the mannose receptor system and should be regarded as an aspecific inhibitor of receptor-mediated endocytotic pathways. Surprisingly, a 500-fold excess of fHSA only moderately (20%) inhibited the clearance of (pap-)Man40-HSA in spite of its high affinity for the scavenger receptor. However, a combination of mannan and fHSA strongly inhibited the uptake of (pap-)Man22-HSA (90%) and to a lesser extent (pap-)Man40-HSA (80%), indicating that a third uptake mechanism may exist that recognizes both mannose groups (or other sugars) and net negative charge. This so far unnoticed receptor system apparently is strongly affected by dextran sulfate and, as shown by immunohistochemistry, is mainly localized on Kupffer cells rather than on the endothelial cells of the liver.