Masao Iwamori

Glycolipids in the thymus of mice after administration of dexamethasone were compared with those in control mice. In parallel with a decrease in the tissue weight due to the disappearance of immature thymocytes in the cortex, the amounts of GlcCer, Gg₄Cer and GM1 decreased from 18 h after intraperitoneal administration of dexamethasone, but those of Gb₄Cer and Forssman glycolipid did not change, indicating the differential distribution of ganglio- and globo-series glycolipids in the thymus, GlcCer, Gg₄Cer and GM1 being on dexamethasone-sensitive cortical thymocytes, and Gb₄Cer and Forssman glycolipid on dexamethasone-resistant cells including thymic stromal cells, respectively. At the same time, a characteristic increase in GM3, whose amount per thymus and concentration per mg of thymus were increased 4-fold and 13-fold compared to those in the control mice, respectively, was observed at the onset of the decrease in tissue weight and was due to the increased activity of LacCer sialyltransferase with the enhanced expression of its gene and the concomitant decrease in cytosolic sialidase activity. One can suggest that endogenous accumulation of GM3 is involved in the dexamethasone-induced apoptosis of cortical thymocytes. On radiolabeling of the thymus with CMP-[¹⁴C]-NeuAc, the incorporation of radioactivity into GM3 was preferentially observed in the thymuses of dexamethasone-administered mice, but not in those of control mice, suggesting the possible involvement of plasma membrane-associated sialytransferase in GM3 synthesis in the thymuses of dexamethasone-administered mice. Published in 2005.

Glycolipids in epithelial tissues of the gastrointestinal tract act as receptors for enteric bacteria and are implicated in the activation of the intestinal immune system. To clarify the genes involved in the fucosylation of the major glycolipids, substrate glycolipids and fucosylated products were measured in tissues of wild-type and mutant mice lacking α(1,2)fucosyltransferase genes FUT1 or FUT2. Quantitative determination was performed by TLC-immunostaining for GA1 (Gg4Cer), FGA1 (fucosyl GA1), GM1 (II3NeuAc-Gg4Cer), FGM1 (fucosyl GM1), and Forssman glycolipids. Both FGM1 and FGA1 completely disappeared from the antrum, cecum, and colon of FUT2-null mice, but not those of FUT1-null and wild-type mice. Precursor glycolipids, GM1 and GA1, accumulated in tissues of FUT2-null mice, indicating that the FUT2-encoded enzyme preferentially participates in the fucosylation of GA1 and GM1 in these tissues. Female reproductive organs were similarly found to utilize FUT2 for the fucosylation of glycolipids FGA1 (uterus and cervix), and FGM1 (ovary), due to their absence in FUT2-null mice. In FUT1-null mice FGA1 was lost from the pancreas, but was present in wild-type and FUT2-null mice, indicating that FUT1 is essential for fucosylation of GA1 in the pancreas. Ulex europaeus agglutinin-I lectin histochemistry for α(1,2)fucose residues confirmed the absence of α(1,2)fucose residues from the apical surface of pancreatic acinar glands of FUT1-null mice. Ileum, epididymis, and testis retained specific fucosylated glycolipids, irrespective of targeted deletion of either gene, indicating either compensation for or redundancy of the α(1,2)fucosyltransferase genes in these tissues.

The regulation of estrogen activity through the formation and cleavage of sulfoconjugates of estrogens is known to be related to the progression and metastasis of estrogen‐dependent breast carcinomas, but the involvement of sulfoconjugates in the steroid stimulation of endometrial functions and the progression of endometrial adenocarcinomas is not clearly understood yet. Estrogen sulfotransferase (EST) in the uterine endometria during the follicular phase was more active than during the luteal phase, but estrogen sulfate (ES) sulfatase exhibited lower activity during the follicular phase than during the luteal phase. However, ES sulfatase activities in cancerous tissues were lower than those in normal endometria and endometrial adenocarcinoma‐derived cells, among which the activity was exceedingly high in Ishikawa cells, suggesting that ES sulfatase in Ishikawa cells contributes to the estrogen‐dependent growth of these cells. EST activities higher than that in Ishikawa cells were found in only 3 of 24 cancerous tissues. Reverse transcriptase‐polymerase chain reaction (RT‐PCR) analysis of the EST and ES sulfatase genes in carcinoma‐derived cells demonstrated the extensive expression of both genes in Ishikawa cells. The isolated EST gene was transfected into Ishikawa cells with a mammalian expression vector to establish cell clones with enhanced EST activity, and the estrogen‐dependent cell growth of the resultant cell clones was found to be abolished, due to the enhanced sulfoconjugation of estrogen. Since ES sulfatase activity in cancerous tissues was significantly lower than that in Ishikawa cells, it might be not involved in the enhancement of estrogen activity associated with the pathogenesis of endometrial adenocarcinoma tissues.

Transformation-associated expression of Le(b) (Lewis antigen-b) or Le(Y) in human colorectal carcinomas has been well described. To examine the expression of glycosphingolipids (GSLs) bearing Lewis-phenotypes in human gynecological carcinoma-derived cells, we determined the concentrations of all GSLs. Although neither Le(b) nor Le(Y) was present in HEC-108 cells established from the poorly differentiated type of endometrial adenocarcinoma, other cell lines from moderately or well-differentiated types expressed either Le(b) or Le(Y), or both, at concentrations of 0.01 to 0.03 microg per mg of dry cells, which comprised 0.3 to 1.3% of the total GSLs. In the cervical and ovarian carcinoma-derived cell lines, Lewis phenotypes tended to be carried by nLc(4)Cer, which was accumulated in the cells without sialylation or fucosylation. These results indicated that expression of Le(b)- or Le(Y)-phenotypes was strongly dependent on the metabolic ability to supply the precursor GSLs. Both Le(b) and Le(Y) were successfully detected by monoclonal antibody MSN-1, which was a useful probe for the simultaneous detection of Le(b) and Le(Y). On application of MSN-1, either Le(b) or Le(Y) was detected in tissues from patients with well- and moderately differentiated types of endometrial adenocarcinoma at concentrations of 0.01 to 0.04 microg per mg of dry tissues, but not in the tissues of poorly differentiated type. Normal endometria at the follicular and luteal phases also contained the antigens, but the concentrations and the frequency of antigen expression were lower than those in the well- and moderately differentiated types of endometrial adenocarcinoma.

Thrombin and plasmin, both of which are serine proteases in the plasma of vertebrates, play essential roles in blood clotting and fibrinolysis, respectively, and regulation of their activities is important to suppress the excessive reactions within the vascular network and to prevent tissue injury. Along with the peptidic inhibitors belonging to the serpin family, we found that cholesterol sulfate (CS), which is present at the concentration of 2.0±1.2 nmol/ml in human plasma, was a potent inhibitor of both plasma thrombin and plasmin. Thrombin, as determined both using a chromogenic substrate and the natural substrate, fibrinogen, was inactivated upon reaction with CS in a dose-dependent manner, but not in the presence of the structually related steroid sulfates, I³SO₃-GalCer and II³NAα-LacCer, suggesting that both the sulfate group and the hydrophobic side chain of CS are necessary for the inhibitory activity of CS. Preincubation of thrombin with CS at 37°C for 10 min was required to achieve maximum inhibition, and virtually complete inhibition was achieved at a molar ratio of CS to thrombin of 18:1. CS-treated thrombin had the same K_m and a lower V_max than the original enzyme, and a higher molecular weight. The molecular weight and activity of the original enzyme were not observed on the attempted separation of the CS-treated enzyme by gel permeation chromatography and native PAGE, indicating that the inactivation of thrombin by CS is irreversible. In contrast, CS was readily liberated from the enzyme by SDS-PAGE, suggesting that hydrophobic interactions are involved in the CS-mediated inactivation of thrombin. When acidic lipids were reacted with thrombin after dissolving them in DMSO, I³SO₃-GalCer, steroid sulfates and II³NAα-LacCer, as well as CS, but not SDS and sodium taurocholate, exhibited inhibitory activity, probably due to micellar formation facilitating interaction between thrombin and negatively charged lipids. On the other hand, plasmin, as determined using a chromogenic substrate, was more susceptible to acidic lipids than thrombin. CS, I³SO₃-GalCer and II³NAα-LacCer, all of which are present in serum, inhibited the activity of plasmin in aqueous media, as well as in DMSO-mediated lipid solutions. Thus, acidic lipids in plasma were demonstrated to possess regulatory activity as endogenous detergents toward both enzymes for blood clotting and fibrinolysis.