Mice are frequently used in models for the study of immunological processes related to inflammation. Since it is known that the degree of fucosylation of human acute phase proteins (APPs) is altered as a consequence of an inflammatory response, we have undertaken this study to gain more insight into the fucosylation of acute phase proteins as it occurs in mouse liver. Mice carrying the cluster of the three genes encoding human α1-acid glycoprotein (AGP), one of the well known APPs, were used and the fucosylation of AGP was assessed. A complete absence of fucosylation on the transgenic human AGP was found, which is in sharp contrast to AGP in human serum, of which a major proportion is normally α3-fucosylated. Remarkably, a large proportion of mouse AGP did contain fucose residues. Fucosylation was also detected on another APP, mouse protease inhibitor (PI).
α3-Fucosylation of the transgenic human AGP can be achieved in vitro, using an α3/4-fucosyltransferase (α3/4-FucT) isolated from human milk, showing that the glycoprotein is not intrinsically resistant to fucosylation. Upon subsequent measurement of the activities of the possible fucosyltransferases present in liver membranes of parent and transgenic mice, only an N-linked-core α6-FucT and no α2-, α3- or α4-FucT activity was detected. This indicates that fucose residues found on the mouse serum proteins AGP and PI, which are synthesized in the liver, are most probably in α6-linkage to the core chitobiosyl unit. Interestingly, both α6- and α3-FucT activity was detectable in human liver membranes. None of the above mentioned findings were influenced by the induction of an acute phase response by administration of bacterial lipopolysaccharide. This study shows that: (a) α6-FucT is probably a protein specific-glycosyltransferase, since mouse AGP, but not human AGP, may be used as an acceptor; (b) in contrast to human liver, mouse liver does not express any α3-FucT-activity, thereby making the mouse incapable of producing the Sialyl Lewisx epitope on APPs, which is an important part of the inflammatory reaction in humans. This last finding indicates that the mouse is not suitable as a model for the study of those phenomena related to inflammation in humans, in which glycosylation of acute phase proteins could play a significant role. © 1998 Rapid Science Ltd