In vitro evidence that carbohydrate moieties derived from uromodulin, an 85,000 dalton immunosuppressive glycoprotein isolated from human pregnancy urine, are immunosuppressive in the absence of intact protein.

Our laboratory recently reported the purification of a unique immunosuppressive glycoprotein isolated from human pregnancy urine (7). This glycoprotein, which we term uromodulin, has a m.w. of 85,000 as assessed on SDS-PAGE and is 30% carbohydrate. Uromodulin blocks in vitro antigen-specific T cell proliferation to recall antigens such as tetanus toxoid at concentrations as low as 100 pM. This glycoprotein also blocks the in vitro generation of spontaneous monocyte-mediated cytotoxicity (7, 36). Recent evidence strongly suggests that the primary action of uromodulin is to act as a specific ligand and modulator of IL 1 (10, 33). We now report additional biochemical characterization of uromodulin, and based on three independent lines of evidence, find that its immunologic activity appears to result from its glycosylation. First, measures to alter the tertiary folding of the protein backbone of uromodulin, including succinylation or reduction and carboxymethylation, fail to significantly affect its in vitro bioactivity. Second, after extensive digestion of intact uromodulin with pronase, the majority of the in vitro bioactivity can be recovered in a single carbohydrate-rich fraction. Finally, digestion with N-glycanase (N-glycosidase F-, an enzyme specific for N-asparagine-linked oligosaccharides) and subsequent purification on thin layer chromatography yields a single complex oligosaccharide that appears to be responsible for the majority of the in vitro immunosuppression mediated by uromodulin. These data suggest that uromodulin displays N-linked carbohydrate sequences capable of down-regulating antigen-specific T cell responses in vitro. It has been suggested that endogenous lectins may play an important role as recognition molecules in mammalian, as well as more primitive immune systems (23, 24). Our in vitro biologic data strongly suggest that the carbohydrate portion of uromodulin is an excellent candidate to function as a potential lectin receptor.