Gate-Tunable Atomically Thin Lateral MoS2 Schottky Junction Patterned by Electron Beam.

Among atomically thin two-dimensional (2D) materials, molybdenum disulfide (MoS2) is attracting considerable attention because of its direct bandgap in the 2H-semiconducting phase. On the other hand, a 1T-metallic phase has been revealed, bringing complementary application. Recently, thanks to top-down fabrication using electron beam (EB) irradiation techniques, in-plane 1T-metal/2H-semiconductor lateral (Schottky) MoS2 junctions were demonstrated, opening a path toward the co-integration of active and passive two-dimensional devices. Here, we report the first transport measurements evidencing the formation of a MoS2 Schottky barrier (SB) junction with barrier height of 0.13-0.18 eV created at the interface between EB-irradiated (1T)/nonirradiated (2H) regions. Our experimental findings, supported by state-of-the-art simulation, reveal unique device fingerprint of SB-based field-effect transistors made from atom-thin 1T layers.

Modulation of retinoid signalling through NGF-induced nuclear export of NGFI-B.

The retinoid-X receptor (RXR) regulates multiple hormonal pathways through heterodimerization with nuclear receptors such as the all-trans retinoic acid receptor (RAR). The orphan nuclear receptor NGFI-B (also called Nur77) can heterodimerize with RXR. Here we show that nerve growth factor (NGF) induces the phosphorylation of Ser 105 of NGFI-B in PC12 phaeochromocytoma cells, resulting in translocation of the NGFI-B-RXR heterodimer complex out of the nucleus using nuclear export signals within NGFI-B. As a consequence of the redistribution of RXR, the transcriptional activity of RXR-RAR is reduced. NGFI-B-mediated nuclear export of receptors may serve as a mechanism for crosstalk between NGF and retinoid pathways.

Integrin cytoplasmic domains mediate inside-out signal transduction.

We analyzed the binding of fibronectin to integrin alpha 5 beta 1 in various cells; in some cells fibronectin bound with low affinity (e.g., K562 cells) whereas in others (e.g., CHO), it bound with high affinity (Kd approximately 100 nM) in an energy-dependent manner. We constructed chimeras of the extracellular and transmembrane domains of alpha IIb beta 3 joined to the cytoplasmic domains of alpha 5 beta 1. The affinity state of these chimeras was assessed by binding of fibrinogen or the monoclonal antibody, PAC1. The cytoplasmic domains of alpha 5 beta 1 conferred an energy-dependent high affinity state on alpha IIb beta 3 in CHO but not K562 cells. Three additional alpha cytoplasmic domains (alpha 2, alpha 6A, alpha 6B) conferred PAC1 binding in CHO cells, while three others (alpha M, alpha L, alpha v) did not. In the high affinity alpha chimeras, cotransfection with a truncated (beta 3 delta 724) or mutated (beta 3(S752-->P)) beta 3 subunit abolished high affinity binding. Thus, both cytoplasmic domains are required for energy-dependent, cell type-specific affinity modulation. In addition, mutations that disrupted a highly conserved alpha subunit GFFKR motif, resulted in high affinity binding of ligands to alpha IIb beta 3. In contrast to the chimeras, the high affinity state of these mutants was independent of cellular metabolism, cell type, and the bulk of the beta subunit cytoplasmic domain. Thus, integrin cytoplasmic domains mediate inside-out signaling. Furthermore, the highly conserved GFFKR motif of the alpha subunit cytoplasmic domain maintains the default low affinity state.

Characterization of platelet aggregation induced by the human melanoma cell line HMV-I: roles of heparin, plasma adhesive proteins, and tumor cell membrane proteins.

We investigated the in vitro mechanism of platelet aggregation induced by HMV-I human melanoma cells. HMV-I cells, in the absence of exogenous plasma proteins, induced platelet aggregation, followed by the release reaction. Heparin at an anticoagulant concentration had no effect on the aggregation. Calcium ion was essential for this tumor cell-platelet interaction and could not be replaced by magnesium. Among the adhesive proteins containing RGD sequences that have been reported to enhance experimental metastasis, fibrinogen and thrombospondin significantly enhanced the aggregation induced by HMV-I cells, fibronectin and von Willebrand factor inhibited it, and vitronectin had no effect. To identify the platelet-aggregating factor(s) of the tumor cells, we have developed a monoclonal antibody against HMV-I cells that can inhibit HMV-I cell-induced platelet aggregation. Immunoprecipitation analysis revealed that this antibody recognized an Mr 71,000 membrane protein. These results suggest that the association between the tumor cells and platelets is mediated by the Mr 71,000 membrane protein recognized by this monoclonal antibody.

CD44 variants but not CD44s cooperate with beta1-containing integrins to permit cells to bind to osteopontin independently of arginine-glycine-aspartic acid, thereby stimulating cell motility and chemotaxis.

The expression of osteopontin (OPN), CD44 variants, and integrins has been correlated with tumorigenesis and metastasis. Here we show that these proteins cooperate to enhance cell motility. First, we demonstrate that several different CD44 variants bind to OPN in an arginine-glycineaspartic acid-independent manner, but that the standard form of CD44 does not. These CD44 variants bind to both the amino- and COOH-terminal portions of OPN independently of the arginine-glycine-aspartic acid sequence, suggesting that multiple domains on OPN can be bound by the CD44 variants. Antibodies directed against the integrin beta1 subunit are able to inhibit this binding. The binding of CD44 variants to OPN is significantly augmented by both anti-CD44s and anti-CD44v antibodies. This augmentation by anti-CD44 antibodies is OPN specific and, again, can be blocked by anti-beta1 antibodies. Finally, we show that OPN binding by CD44 variants/beta1-containing integrins promotes cell spreading, motility, and chemotactic behavior.

Salutary effect of disopyramide on left ventricular diastolic function in hypertrophic obstructive cardiomyopathy.

OBJECTIVES: The purpose of this study was to estimate the effect of disopyramide on left ventricular diastolic function in patients with hypertrophic obstructive cardiomyopathy. BACKGROUND: Although disopyramide has been reported to lessen clinical symptoms in patients with hypertrophic obstructive cardiomyopathy, few data exist regarding its effect on diastolic function in these patients. METHODS: Thirteen patients with hypertrophic cardiomyopathy (six with and seven without left ventricular outflow obstruction) were examined. Before and after intravenous disopyramide, hemodynamic and angiographic studies were performed. RESULTS: In patients with outflow obstruction, pressure gradient at the outflow tract decreased from a mean +/- SD of 100 +/- 45 to 26 +/- 33 mm Hg (p < 0.01). Although systolic function was similarly impaired in both groups, the time constant of left ventricular pressure decay (tau) shortened from 56 +/- 10 to 44 +/- 8 ms (p < 0.01) and the constant of left ventricular chamber stiffness (kc) decreased from 0.049 +/- 0.017 to 0.038 +/- 0.014 m2/ml (p < 0.01) only in patients with outflow obstruction. Shortening in tau correlated best with decrease in left ventricular systolic pressure (r = 0.84, p < 0.01). In contrast, tau was prolonged from 52 +/- 10 to 64 +/- 11 ms (p < 0.01) and kc was unchanged in patients without outflow obstruction. CONCLUSIONS: The primary effects of disopyramide on the hypertrophied left ventricle were negative inotropic and negative lusitropic. However, left ventricular diastolic properties in patients with outflow obstruction were improved with a decrease in outflow pressure gradient. Relief of clinical symptoms in hypertrophic obstructive cardiomyopathy with disopyramide might be due in part to improvement of diastolic function, which appears secondary to the reduction in ventricular afterload.

Costimulatory signals distinctively affect CD20- and B-cell-antigen-receptor-mediated apoptosis in Burkitt's lymphoma/leukemia cells.

CD20 is a B-cell differentiation antigen and known to induce apoptosis in Burkitt's lymphoma/leukemia (BL) cells upon antibody-mediated crosslinking. We examined the biological effect of CD20 crosslinking on BL cell lines and observed that apoptosis induction is accompanied by activation of multiple caspases, including caspase-8, -9, -3, -2, and -7. Further investigation revealed a clear synergism between apoptosis mediated by CD20 and by B-cell antigen receptor (BCR). Examination of the effect of simultaneous crosslinking of other cell surface molecules with crosslinking of CD20 or BCR on apoptosis induction showed that these molecules had either a synergistic or inhibitory effect on induction of apoptosis. It is worth noting that some molecules had a different effect on CD20- and BCR-mediated apoptosis. Simultaneous crosslinking of the molecules CD10, CD22, CD72, and CD80 inhibited BCR-mediated apoptosis, but enhanced CD20-mediated apoptosis. Further studies revealed that regulation of CD20-induced apoptosis by other costimulatory molecules is achieved by modification of caspase activation. CD20-mediated apoptosis in BL cells may provide not only a model for understanding the mechanism regulating clonal selection of B cells but a new therapeutic strategy for BL patients.

Involvement of tyrosine phosphorylation in inhibition of fMLP-induced PLD activation by N-acetyl-L-cysteine in differentiated HL60 cells.

N-acetyl-L-cysteine (NAC), which is known as a multipotential agent; an antioxidant, a thiol reagent, or a tyrosine kinase inhibitor, inhibited N-formyl-methionyl-leucyl-phenylalanine (fMLP)-induced phospholipase D (PLD) activation in HL60 cells in a concentration-dependent manner (IC50 = 2 mM). Its inhibitory mechanism was examined in this study to gain insight into the regulation of PLD activity. NAC had no direct effect on membrane PLD activity in an in vitro assay system. fMLP-induced formation of inositol phosphates via phospholipase C (PLC) was not affected by the drug, suggesting that the receptor-G protein coupling was not inhibited. H2O2, which is known to induce PLD activation in several types of cells, failed to activate PLD in HL60 cells. Pretreatment of 3-amino-1,2,4-triazole (ATZ), a catalase inhibitor, did not enhance fMLP-induced PLD activation. NAC inhibited fMLP-induced tyrosine phosphorylation of several protein bands (42, 44, 64, and 138 kDa) in a concentration-dependent manner. The temporal and concentration-dependent inhibitory profiles for tyrosine phosphorylation of 64- and 138-kDa proteins were well correlated with PLD activation. However, thiol reagents, 1 mM 2,3-dimercapto-l-propanol (2,3-DMP), 1 mM dithiothreitol (DTT), and 2 mM cysteine also did not suppress protein tyrosine phosphorylation or PBut formation by fMLP. Wortmannin, a selective phosphatidylinositol 3-kinase (PI 3-kinase) inhibitor, inhibited these two tyrosine phosphorylation bands. These results suggest that NAC inhibits fMLP-induced PLD activation through blockage of protein tyrosine phosphorylation, which is located at the downstream of PI-3 kinase.

Globotriaosyl ceramide (CD77/Gb3) in the glycolipid-enriched membrane domain participates in B-cell receptor-mediated apoptosis by regulating lyn kinase activity in human B cells.

The role of CD77 expressed on a fraction of germinal center B cells, also known as glycosphyngolipid Gb3, and as a functional receptor for Shiga toxins (Stx) in B-cell receptor (BCR)-mediated apoptosis was investigated. Using Stx1-sensitive Burkitt's lymphoma Ramos cells as an in vitro model of CD77(+) germinal center B cells, intracellular signaling events mediated by either Stx1 or anti-CD77 antibody were examined immunobiochemically and immunocytologically. We observed prompt activation of Lyn and Syk kinases leading to increased binding of these proteins to surface IgM (sIgM) in Ramos cells after Stx1 treatment. We also observed microscopic colocalization of CD77 and sIgM after stimulation with Stx1. Along with the synergism between the cross-linking of CD77 and that of sIgM in their effect on apoptosis induction, it was highly probable that CD77 cross-linking induces activation of the BCR signaling cascade. Analysis using sucrose density gradient centrifugation suggested that Stx1 binding to CD77 induced recruitment and activation of Lyn in the glycolipid-enriched membrane (GEM) fractions. Once activated, however, Lyn seemed to acquire an increased detergent solubility and moved outside of the GEM fractions. This study describes the participation of the GEM domain in BCR-signaling cascade and suggests a possible role of CD77 as a regulator of BCR-induced apoptosis in human B cells.

Spermine is not essential for growth of Saccharomyces cerevisiae: identification of the SPE4 gene (spermine synthase) and characterization of a spe4 deletion mutant.

Spermine, ubiquitously present in most organisms, is the final product of the biosynthetic pathway for polyamines and is synthesized from spermidine. In order to investigate the physiological roles of spermine, we identified the SPE4 gene, which codes for spermine synthase, on the right arm of chromosome XII of Saccharomyces cerevisiae and prepared a deletion mutant in this gene. This mutant has neither spermine nor spermine synthase activity. Using the spe4 deletion mutant, we show that S. cerevisiae does not require spermine for growth, even though spermine is normally present in the wild-type organism. This is in striking contrast to the absolute requirement of S. cerevisiae for spermidine for growth, which we had previously reported using a mutant lacking the SPE3 gene (spermidine synthase) [Hamasaki-Katagiri, N., Tabor, C. W., Tabor, H., 1997. Spermidine biosynthesis in Saccharomyces cerevisiae: Polyamine requirement of a null mutant of the SPE3 gene (spermidine synthase). Gene 187, 35-43].

Demonstration of a rhodopsin-retinochrome system in the stalk eye of a marine gastropod, Onchidium, by immunohistochemistry.

The stalk eye of Onchidium sp. (Gastropoda, Mollusca) is the principal photoreceptor in a multiple photoreceptive system that consists of the stalk and dorsal eyes, dermal photoreceptor cells, and photosensitive neurons. To examine the localization of photopigments, the stalk eyes were immunostained with specific antibodies to rhodopsin, retinochrome, and retinal-binding protein (RALBP), which had been generated against squid retinal proteins. The retina of the stalk eye was divided into villous, pigmented, somatic, and neural layers. It was comprised mainly of two types of visual and pigmented supportive cells. The type 1 visual (VC1) cell was characterized by well-developed microvilli on its apical protrusion and photic vesicles in the cytoplasm. The photic vesicles were specifically blackened by prolonged osmification. The type 2 visual (VC2) cell had less numerous, shorter microvilli on its concave apical surface and lacked photic vesicles. The anti-squid rhodopsin antiserum was localized specifically to the villous layer that corresponded to the VC1 microvilli. With the anti-retinochrome peptide antibody, the somatic layer showed specific but patchy, positive staining that corresponded to the cytoplasm of the VC1 cells. Because the photic vesicles are known to contain retinochrome, these results indicate that this retinochrome is localized in the VC1 cytoplasm. Anti-RALBP antibody stained the supranuclear cytoplasm to the distal cytoplasm of VC1 cells. This is the first demonstration of the localization of RALBP in the Gastropoda Onchidium stalk eye. In squid retina that were immunostained as positive controls, the anti-rhodopsin antibody stained rhabdomeric microvilli, the anti-retinochrome antibody stained the inner segment and the basal region of the outer segment, and the anti-RALBP antibody stained the outer and inner segments, respectively. These results suggest that the rhodopsin-retinochrome system that has been established in cephalopod eyes is present in the Onchidium stalk eye.

Heparin-stimulated expression of extracellular-superoxide dismutase in human fibroblasts.

Extracellular-superoxide dismutase (EC-SOD) is the major SOD isozyme in the arterial wall and may be important for antioxidation capability of the vascular wall and normal vascular function. EC-SOD is expressed in various cell types in the vascular wall such as fibroblasts, smooth muscle cells and macrophages, and the synthesis of EC-SOD by human fibroblasts is known to be highly responsive to various inflammatory cytokines, although there is no response to oxidative stress. Heparin is a highly sulfated glycosaminoglycan with many functions such as antithrombotic, antilipemic and antiatherosclerotic effects. Another less well-known function of heparin is regulation of protein synthesis. In this study, we measured the induction of EC-SOD after treatment with heparin to understand the role of heparin in the antiatherosclerotic response of fibroblasts. Heparin induced EC-SOD expression at both the mRNA and protein levels. Heparin showed the greatest stimulatory effect and heparan sulfate showed moderate effects. The effect of chondroitin sulfate A was not clear. In contrast, desulfated heparin and chondroitin sulfate C did not increase EC-SOD expression. The stimulatory effect seemed to increase roughly with the degree of glycosaminoglycan sulfation. The enhanced expression of EC-SOD by heparin must contribute to the antiatherosclerotic effect of heparin.

Specific cross-reaction of IgG anti-phospholipid antibody with platelet glycoprotein IIIa.

To study the pathological functions of anti-phospholipid (anti-PL) antibodies, we have analyzed their effect on platelet function. We identified an IgG anti-PL mAb, designated PSG3, which cross-reacted specifically with glycoprotein (GP) IIIa in human platelets and inhibited platelet aggregation. PSG3 bound also to certain polyanionic substances, such as double-stranded DNA, heparan sulfate, dextran sulfate and acetylated-LDL, but not to other polyanionic substances. The binding of PSG3 to GPIIIa was completely inhibited by heparan sulfate and dextran sulfate, indicating that PSG3 recognizes a particular array of negative charges expressed on both GPIIIa and the specified polyanionic substances. Since neither neuraminidase- nor endoglycopeptidase F-treatment of GPIIIa had any significant effect on the binding of PSG3, this array must be located within the amino acid sequence of GPIIIa but not in the carbohydrate moiety. Reduction of the disulfide bonds in GPIIIa greatly reduced its reactivity, suggesting that the negative charges in the epitope are arranged in a particular conformation. PSG3 inhibited platelet aggregation induced by either ADP or collagen, it also inhibited fibrinogen binding to activated platelets in a dose-dependent fashion. PSG3, however, did not inhibit the binding of GRGDSP peptide to activated platelets. These results suggest that the PSG3 epitope on GPIIIa contains a particular array of negative charges, and possibly affects the fibrinogen binding to GPIIb/IIIa complex necessary for platelet aggregation.

Localization of von Willebrand factor and thrombin-interactive domains on human platelet glycoprotein Ib.

Platelet membrane glycoprotein Ib (GPIb) functions as receptors for thrombin and von Willebrand factor (vWF) in the presence of ristocetin. To precisely locate the domains on GPIb interacting with vWF and thrombin, we prepared several peptides that have amino acid sequences analogous to that of the GPIb alpha-chain and examined their effects on ristocetin-induced (vWF-dependent) and thrombin-induced platelet aggregations. A peptide extending from residues Asp235 to Lys262 showed the strongest inhibitory effect on ristocetin-induced platelet agglutination, and a group of overlapping peptides composed of 24-28 amino acid residues representing sequences extending from Phe216 to Asp274 was found to inhibit platelet aggregation induced by thrombin. Other peptides did not inhibit platelet aggregations. Moreover, the binding to platelets of the monoclonal anti-GPIb antibody (TM60) which had been shown to inhibit both ristocetin- and thrombin-induced platelet aggregations was strongly inhibited by a peptide extending from Asp249 to Asp274. These data demonstrate that the vWF-binding domain exists in a small region between residues Asp235 and Lys262; the thrombin-interacting domain, in contrast, is located between residues Phe216 and Ala274, with a possible center of interaction in the sequence from Phe216 to Thr240 on the GPIb alpha-chain, and thrombin binding requires a relatively strict conformation in this domain.

Effect of wortmannin and 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002) on N-formyl-methionyl-leucyl-phenylalanine-induced phospholipase D activation in differentiated HL60 cells: possible involvement of phosphatidylinositol 3-kinase in phospholipase D activation.

Phospholipase D (PLD) plays an important role in neutrophil activation. However, despite various proposed mechanisms, its detailed regulatory mechanism is not fully understood. The functional coupling between phosphatidylinositol 3-kinase (PI 3-kinase) and PLD was investigated in N-formyl-methionyl-leucyl-phenylalanine (fMLP)-stimulated human promyelocytic leukemia HL60 cells, using wortmannin, a fungal metabolite that is known as a selective inhibitor for phosphatidylinositol 3-kinase. Treatment of cells with this drug inhibited the formation of both phosphatidylinositol 3,4,5-trisphosphate (PIP3), a product of PI 3-kinase, and phosphatidylbutanol (PBut), the specific product of transphosphatidylation due to PLD in the presence of butanol, with similar concentration dependence (IC50 = 30-70 nM). Another PI 3-kinase inhibitor, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002) also inhibited PBut formation in a concentration-dependent manner. However, wortmannin failed to inhibit phorbol 12-myristate 13-acetate-induced PLD activation in whole cells and membrane PLD activity in an in vitro assay system, indicating that inhibition of fMLP-induced PLD activation by wortmannin was not due to its direct effect on PLD activity. These results suggest that a major part of inhibition of PLD activation by wortmannin might be mediated through its effect on PI 3-kinase.

A double cysteine trkA mutant exhibiting reduced NGF binding and delayed Erk signaling.

The NGF receptor trkA is a tyrosine kinase receptor comprising an extracellular domain with a ligand-binding site, a transmembrane-spanning domain (TMD), and an intracellular domain composed of a juxtamembrane region (JMR), a tyrosine kinase domain, and a short carboxy-terminal tail. Nerve growth factor (NGF) binds and activates this receptor, leading to phosphorylation of signaling substrates involved in neuronal proliferation, differentiation, and survival. Human trkA contains one cysteine residue in the TMD (C423) and another, separated by 12 residues, in the JMR (C436). We hypothesized that the removal of one or both of the cysteines would affect NGF-induced signaling of the trkA receptor. Here we show that NGF induces rapid receptor autophosphorylation in a wild-type, trkA-expressing clone (WT11), in a single cysteine trkA mutants (C423T or C436A), but lower autophosphorylation activity in a double-cysteine trkA mutant (C423T/C436A). WT11 and SM cells had similar binding affinity, but that of DM cells was lower, according to the NGF radioreceptor assay. NGF-induced Erk phosphorylation was rapid in WT11 and C423T cells, but delayed in C436A and C423T/C436A cells. NGF induced [3H]thymidine incorporation into WT11 and SM cells, but had no effect on DM cells. However, basic fibroblast growth factor (bFGF) induced rapid phosphorylation of Erk1/2, and [3H]thymidine incorporation in NIH3T3, WT11, single mutant (SM), and double mutant (DM) cells, suggesting that the impaired NGF-induced Erk phosphorylation and thymidine incorporation observed in DM cells are due to the double-cysteine mutations in the trkA receptor. Cumulatively, our findings support a model in which Cys436 of the trkA is responsible for the rapid transfer of the transmembrane occupancy signal to the SHC adaptor protein for activation of the Ras-Erk pathway and DNA synthesis.

The dysfunction of calcium-ATPase pump in double negative T cells of autoimmune-prone mice.

Double negative (DN) T cells expanding in peripheral lymphoid tissues in mice bearing lymphoproliferation (lpr) gene are generally unresponsive to mitogens, antigens, and anti-T cell receptor (TCR) or anti-CD3 monoclonal antibodies (mAb). In response to the stimulation with 0.125-5.0 microM ionomycin, control T cells sustained an increase in intracellular free calcium ([Ca2+]i), while DN lpr T cells showed a gradual fall following initial rapid increase in [Ca2+]i. Such gradual fall in [Ca2+]i was overcome by the addition of endoplasmic and sarcoplasmic reticulum Ca(2+)-ATPase inhibitor or high dose (10 microM) of ionomycin. The requirement of high concentration of calcium ionophore for the sustained increase of [Ca2+]i in lpr DN T cells is due to dysfunction of Ca(2+)-ATPase pump.

The induction of J11d antigen on double negative T cells of MRL/Mp-Lpr/Lpr mice by high dose calcium ionophore.

Mice homozygous for the lymphoproliferation (lpr) gene spontaneously develop autoimmune syndrome. These mice were characterized by the massive accumulation of double negative (DN) T cells. Although peripheral T cells in normal mice do not express J11d antigen, those abnormal DN T cells in autoimmune-prone mice express J11d antigen. In this study, the mechanisms that control the expression of J11d antigen are analyzed. High concentration of calcium ionophore alone induces the expression of J11d antigen, but not of CD4, CD8, and activation antigens such as interleukin 2 receptor as well as transferrin receptor by J11d- DN T cells from lpr mice. The expression of J11d antigen is primarily regulated at the transcription level rather than the post transcription level. Experiments using metabolic inhibitors reveal that the induction of J11d antigen requires the activation of not only a Ca2+/calmodulin- but also protein kinase C-dependent signaling pathway. Furthermore, J11d- DN thymocytes from control mice share the similar functional property with DN lpr T cells in J11d antigen inducibility.

Nitrobenzylthioinosine (NBT), a nucleoside transport inhibitor, protects against Shiga toxin cytotoxicity in human microvascular endothelial cells.

Infections with Shiga toxin (Stx)-producing Escherichia coli (STEC) cause microvascular endothelial cell damage, resulting in hemorrhagic colitis and hemolytic uremic syndrome. The prevention of endothelial cell damage is therefore a crucial step in overcoming this disorder. Here, we report that nitrobenzylthioinosine (NBT), a nucleoside transport inhibitor, has a protective effect against the cytotoxicity of Stxs in human microvascular endothelial cells (HMVECs). The relative viability of cells treated with 1.5-15 pM of Stx1 was reduced to 10-20% of that without Stx1. However, the viability of cells treated with NBT (10-100 microM) remained higher than 80%, even in the presence of Stx1. NBT also protected against Stx1 cytotoxicity in sodium butyrate-treated hypersensitive HMVECs. The protective effect of NBT against Stx cytotoxicity may be due to the depletion of ATP in the cells, thereby inhibiting the entry of Stx1.

Microheterogeneity and oligosaccharide chains on the beta chains of HLA-DR, human major histocompatibility complex class II antigen, analyzed by the lectin-nitrocellulose sheet method.

The beta chain of human histocompatibility complex class II antigen, HLA-DR, showed 4 to 5 microheterogeneous spots on a gel obtained by two-dimensional polyacrylamide gel electrophoresis. The types of oligosaccharide chains on the beta chains were analyzed by the lectin-nitrocellulose sheet method for each microheterogeneous spot with 3 cell lines of two haplotypes (HLA-DR 4,4, and 3,3). Two kinds of oligosaccharide chains were observed and were essentially the same in the microheterogeneous spots from all three cell lines. One, the oligosaccharide chain on the most basic spot (beta 1), was stained with peroxidase-coupled concanavalin A (Con A-P.O.) but not with peroxidase-coupled wheat germ agglutinin and was sensitive to endo-beta-N-acetylglucosaminidase H (endo H), indicating that it was a high-mannose type. The oligosaccharide chains on other spots that were not stained with Con A-P.O. but were stained with peroxidase-coupled Ricinus communis agglutinin were resistant to endo H. beta 2 and beta 3 were stained with E-PHA. Thus, they probably had bisected biantennary and others probably had multiantennary complex-type oligosaccharides. Sialidase experiments showed that the charge heterogeneity was due to post-translational sialylation of the oligosaccharide chains. In pulse-chase experiments, the most basic spot of beta chain (beta 1) was labeled first, beta 2 and beta 3 were labeled next, and beta 4 was labeled last. These labeling characters accorded well with the results on the oligosaccharide types mentioned above.