Search

We have used mRNA differential display PCR to search for genes induced in activated T cells and we identified a gene encoding an alpha2,6-sialyltransferase (ST6GalNAc IV) that is rapidly induced in lymphocytes after antigen or mitogen stimulation. The 3.6 kb full-length cDNA clone (MK45) obtained contained a single open reading frame encoding a 302 amino acid protein and a 2.5 kb 3' untranslated region. MK45 expression in in vivo-activated CD8 T cells reached the highest level 4 h after antigen triggering and then declined rapidly to nearly base levels within 45 h. Northern blot analysis further revealed that MK45 expression was also induced in LPS-activated B cells and antigen-triggered CD4 T cells in vitro. MK45 expression was low or undetectable in most other mouse tissues examined, when compared to activated lymphocytes. Importantly, the mRNA expression level of other sialyltransferases remained largely unchanged during the early stage of lymphocyte activation. Finally, increased ecto-sialyltransferase activity and an altered sialylation pattern were demonstrated on the cell surface of early activated CD8 T cells. Our report identifies a candidate sialyltransferase gene that is involved in the early alteration of the sialylation pattern of cell surface molecules in activated lymphocytes.

The CD22 antigen, although present in the cytoplasm of early and immature B cells, first appears on the cell surface of mature B lymphocytes. The phenotypic patterns and some functional properties of the surface-expressed CD22 antigen are well described, but little is known about its molecular structure. We have therefore investigated the relationship of the two CD22 glycoproteins (140/130 kd), and the influence of their complex N-glycosylation on surface expression and antibody recognition. Comparative peptide mapping of the 100 and 80 kd protein cores, obtained by endoglycosidase F treatment, revealed a common structure shared by both protein cores. In pulse-chase experiments the mature glycoproteins originated from two separate precursor molecules, indicating that the two proteins may be generated by different RNA processing. In cell lysates a CD22 specific polyclonal anti-serum recognized both molecules the size of the protein cores and glycosylated CD22 molecules, whereas in membrane preparations only the glycosylated forms were detected. The CD22 mAb HD39 reacted exclusively with the glycoprotein froms in either cellular preparation. The influence of glycosylation on surface expression and epitope recognition was investigated in more detail by applying various inhibitors of the glycosylation pathway. 1-Deoxymannojirimycin and Swainsonine, which block glycosylation at the high-mannose and hybrid-type stages respectively, modulated the CD22 antigen but did not alter its surface expression. Tunicamycin blocked de novo glycosylation and led to reduced surface recognition of the CD22 antigen. Together, these results suggested that comparatively simple oligosaccharide structures of high-mannose type are sufficient for surface expression of the CD22 antigen and for epitope recognition by mAb HD39. It is most likely that glycosylation is required to stabilize epitopes in the protein moiety recognized by CD22 mAb. Finally, we demonstrated the presence of glycosylated, cytoplasmic CD22 antigen in CD22 surface negative B-lineage ALL cells. This finding led us to conclude that complex glycosylation does not provide the determining signal for the switch from cytoplasmic to surface expression of the CD22 antigen.

The neutral glycosphingolipid (GSL) globotriaosylceramide (Gb3) of the globo-series was recently defined as the CD77 antigen. This B cell-associated antigen is characterized by its specific expression on germinal center B cells. In order to study the potential relation of the CD77 antigen and other GSLs to B cell activation we have performed a comprehensive analysis of the synthesis and expression of neutral GSL in tonsillar B lymphocytes. Monoglycosylceramide (GL1) and lactosylceramide (LacCer) comprised the largest portion of GSL in tonsillar B lymphocytes as detected by HPLC analysis. GSLs of the globo-series Gb3 and globotetraosylceramide (Gb4), were found in smaller amounts. Since other GSLs, like gangliotriaosylceramide (Gg3) and gangliotetraosylceramide (Gg4), could only be detected using highly sensitive antibody reactions, we assume that these GSLs occur in B cells only in minor amounts. When tonsillar B cells were density fractionated on Percoll, the light density cells, which correspond to activated cells, contained and expressed more of both globo-GSLs than cells in the higher density fraction. When the dense fraction of tonsillar B cells was activated in vitro by anti-mu/BCGF, synthesis of GL1, LacCer, Gb3, and Gb4 was biphasic, with maxima at 12 and 84 h. Surface expression of the CD77 antigen on the denser cells was strongly induced by anti-mu/BCGF during the first 24 h of cultivation followed by a rapid decline thereafter, mimicking synthesis. PMA treatment of this cell fraction caused an even stronger expression of the CD77 antigen, which lasted over 48 h of cultivation.(ABSTRACT TRUNCATED AT 250 WORDS)