[Genetic susceptibility to norovirus infection]. / Sensibilité génétique aux infections à norovirus.

Norovirus (NoV) are the most common cause of acute gastroenteritidis in humans worldwide. They are transmitted through consumption of contaminated food, or mostly by direct person-to-person contact. However, susceptibility to NoV infection is variable. NoVs recognize carbohydrate ligand, including A, B, H and Lewis histoblood group antigen (HBGAs) for attachment to human epithelial cells. Synthesis of these HBGAs requires various glycosyltransferase encoded by the ABO, FUT2, FUT3 genes. The presence of distinct carbohydrates structures dependent upon the combined polymorphism at the FUT2, FUT3 and ABO loci influences susceptibility to NoV infection. NoV-glycan interactions studies show that different strains recognize specific HBGAs. Together with herd immunity, HBGAs play a major role in the epidemiology and evolution of NoVs.

Qualitative and quantitative analysis of the binding of GII.4 norovirus variants onto human blood group antigens.

Noroviruses (NoVs) are one of the leading causes of gastroenteritis in children and adults. For the last 2 decades, genogroup II genotype 4 (GII.4) NoVs have been circulating worldwide. GII.4 NoVs can be divided into variants, and since 2002 they have circulated in the population before being replaced every 2 or 3 years, which raises questions about the role of their histo-blood group antigen (HBGA) ligands in their evolution. To shed light on these questions, we performed an analysis of the interaction between representative GII.4 variants and HBGAs, and we determined the role of selected amino acids in the binding profiles. By mutagenesis, we showed that there was a strict structural requirement for the amino acids, directly implicated in interactions with HBGAs. However, the ablation of the threonine residue at position 395 (ΔT395), an epidemiological feature of the post-2002 variants, was not deleterious to the binding of the virus-like particle (VLP) to the H antigen, while binding to A and B antigens was severely hampered. Nevertheless, the ΔT395 VLPs gained the capacity to bind to the Lewis x and sialyl-Lewis x antigens in comparison with the wild-type VLP, demonstrating that amino acid residues outside the HBGA binding site can modify the binding properties of NoVs. We also analyzed the attachment of baculovirus-expressed VLPs from six variants (Bristol, US95/96, Hunter, Yerseke, Den Haag, and Osaka) that were isolated from 1987 to 2007 to phenotyped saliva samples and synthetic HBGAs. We showed that the six variants could all attach to saliva of secretors irrespective of the ABO phenotype and to oligosaccharides characteristic of the secretor phenotype. Interestingly, Den Haag and Osaka variants additionally bound to carbohydrates present in the saliva of Lewis-positive nonsecretors. The carbohydrate binding profile and the genetic and mutagenesis analysis suggested that GII.4 binding to Lewis x and sialyl-Lewis x antigens might be a by-product of the genetic variation of the amino acids located in the vicinity of the binding site. Analysis of the binding properties for the six variants by surface plasmon resonance showed that only post-2002 variants (i.e., Hunter, Yerseke, Den Haag, and Osaka) presented strong binding to A and B antigens, suggesting that the GII.4 evolution could be related to an increased affinity for HBGAs for the post-2002 variants. The combination of increased affinity for ABH antigens and of a newly acquired ability to recognize glycans from Lewis-positive nonsecretors could have contributed to the epidemiological importance of strains such as the Den Haag GII.4 subtype.

Infection by Helicobacter pylori expressing the BabA adhesin is influenced by the secretor phenotype.

Helicobacter pylori (Hp) infects half the world's population and causes diverse gastric lesions, from gastritis to gastric cancer. Our aim was to evaluate the significance of secretor and Lewis status in infection and in vitro adherence by Hp expressing BabA adhesin. We enrolled 304 Hp-infected individuals from Northern Portugal. Gastric biopsies, blood and saliva were collected. Polymerase chain reaction (PCR) and immunofluorescence were used to detect BabA+ Hp in gastric biopsies. In vitro adherence by a BabA expressing Hp strain to gastric biopsies was performed. Secretor status was identified by Ulex, a lectin that recognizes secretor-dependent glycan structures in saliva and in gastric mucosa, and by Lewis(a/b) antibodies, and indirectly by identification of an inactivating mutation in the FUT2 gene (G428A). BabA status of infecting Hp was associated with CagA and VacAs1 (p < 0.05), intercellular localization of Hp (p < 0.01) and the presence of intestinal metaplasia (p < 0.05) and degenerative alterations (p < 0.005) in the biopsies. BabA was associated (p < 0.05) with Ulex staining of gastric biopsies and, although not significantly, to absence of homozygosity for FUT2 G428A inactivating polymorphism. In vitro Hp adherence was higher in cases wild-type or heterozygous for FUT2 G428A mutation (p < 0.0001), cases staining for Ulex (p < 0.0001) and a(-)b+ and a(-)b(-) secretor phenotypes (p < 0.001). In conclusion, BabA+ Hp infection/adhesion is secretor-dependent and associated with the severity of gastric lesions.

Fluorescent carbohydrate probes for cell lectins.

Fluorescein labeled carbohydrate (Glyc) probes were synthesized as analytical tools for the study of cellular lectins, i.e. SiaLe(x)-PAA-flu, Sia2-PAA-flu, GlcNAc2-PAA-flu, LacNAc-PAA-flu and a number of similar ones, with PAA a soluble polyacrylamide carrier. The binding of SiaLe(x)-PAA-flu was assessed using CHO cells transfected with E-selectin, and the binding of Sia2-PAA-flu was assessed by COS cells transfected with siglec-9. In flow cytometry assays, the fluorescein probes demonstrated a specific binding to the lectin-transfected cells that was inhibited by unlabeled carbohydrate ligands. The intense binding of SiaLe(x)-PAA-3H to the E-selectin transfected cells and the lack of binding to both native and permeabilized control cells lead to the conclusion that the polyacrylamide carrier itself and the spacer arm connecting the carbohydrate moiety with PAA did not contribute anymore to the binding. Tumors were obtained from nude mice by injection of CHO E-selectin or mock transfected cells. The fluorescent SiaLe(x)-PAA-flu probe could bind to the tumor sections from E-selectin positive CHO cells, but not from the control ones. Thus, these probes can be used to reveal specifically the carbohydrate binding sites on cells in culture as well as cells in tissue sections. The use of the confocal spectral imaging technique with Glyc-PAA-flu probes offered the unique possibility to detect lectins in different cells, even when the level of lectin expression was rather low. The confocal mode of spectrum recording provided an analysis of the probe localization with 3D submicron resolution. The spectral analysis (as a constituent part of the confocal spectral imaging technique) enabled interfering signals of the probe and intrinsic cellular fluorescence to be accurately separated, the distribution of the probe to be revealed and its local concentration to be measured.

ABH and Lewis histo-blood group antigens, a model for the meaning of oligosaccharide diversity in the face of a changing world.

Antigens of the ABH and Lewis histo-blood group family have been known for a long time. Yet their biological meaning is still largely obscure. Based on the available knowledge about the genes involved in their biosynthesis and about their tissue distribution in humans and other mammals, we discuss here the selective forces that may maintain or propagate these oligosaccharide antigens. The ABO, alpha 1,2fucosyltransferase and alpha 1,3fucosyltransferase enzyme families have been generated by gene duplications. Members of these families contribute to biosynthesis of the antigens through epistatic interactions. We suggest that the highly polymorphic genes of each family provide intraspecies diversity that allows coping with diverse and rapidly evolving pathogens. In contrast, the genes of low frequency polymorphism are expected to play roles at the cellular level, although they may be dispensable at the individual level. In addition, some members of these three gene families are expected to be functionally redundant and may either provide a reservoir for additional diversity in the future or become inactivated. We also discuss the role of the ABH and Lewis histo-blood group antigens in pathologies such as cancer and cardiovascular diseases, but argue that it is merely incidental and devoid of evolutionary impact.

ABH and Lewis histo-blood group antigens in cancer.

Antigens of the ABH and Lewis histo-blood group family can be found on many normal cells, mainly of epithelial type. In carcinomas, altered expression of the various carbohydrate epitopes of this family occur, and are often strongly associated with either a good or bad prognosis. A review of the available data on these tumor-associated markers, their biosynthesis and their prognostic value is proposed here. For a long time it has been unclear whether their presence could affect the behavior of carcinoma cells. Recent data, however, indicate that they play biological roles in the course of tumor progression. The presence of sialyl-Le(a) or sialyl-Le(x), which are ligands for selectins, promotes the metastatic process by facilitating interaction with the endothelium of distant organs. The loss of A and B antigens increases cellular motility, while the presence of H epitopes increases resistance to apoptosis by mechanisms that remain to be defined. The Le(y) antigen has procoagulant and angiogenic activities. All these observations are used to present a model that may account for the described associations between the presence or loss of these markers and the outcome of disease. Finally, their potential clinical applications as tumor-associated markers or as targets of immunotherapy are reviewed.

Comparison of the three rat GDP-L-fucose:beta-D-galactoside 2-alpha-L-fucosyltransferases FTA, FTB and FTC.

The complete coding sequences of three rat alpha1,2fucosyltransferase genes were obtained. Sequence analysis revealed that these genes, called FTA, FTB and FTC, were homologous to human FUT1, FUT2 and Sec1, respectively. A distance analysis between all alpha1,2fucosyltransferase sequences available showed that the two domains of the catalytic region evolved differently with little divergence between the FUT2 and Sec1 N-terminal domains, quite distant from that of FUT1. At variance, FUT1 and FUT2 C-terminal domains were less distant while a high evolutionary rate was noted for Sec1 C-terminal domain. Whereas FTA and FTB encode typical glycosyltransferases, FTC lacks the homologous start codon and encodes a protein devoid of intracellular and transmembrane domains. It is located on rat chromosome 1q34. Transfection experiments revealed that unlike FTA and FTB, FTC does not generate enzyme activity. Analysis by flow cytometry showed that H type 2 epitopes were synthesized in Chinese hamster ovary cells transfected by both FTA and FTB cDNA, but only FTB transfectants possessed H type 3 determinants. In REG rat carcinoma cells, both FTA and FTB allowed synthesis of H type 2 and H type 3 at the cell surface. Western blots showed that, in both cell types, FTA was able to synthesize H type 2 epitopes on a larger set of glycoproteins than FTB. Analysis of the kinetic parameters obtained using small oligosaccharides revealed only a slight preference of FTA for type 2 over other types of acceptor substrates, whereas FTB was barely able to fucosylate this substrate.

Binding of rabbit hemorrhagic disease virus to antigens of the ABH histo-blood group family.

The ability of rabbit hemorrhagic disease virus to agglutinate human erythrocytes and to attach to rabbit epithelial cells of the upper respiratory and digestive tracts was shown to depend on the presence of ABH blood group antigens. Indeed, agglutination was inhibited by saliva from secretor individuals but not from nonsecretors, the latter being devoid of H antigen. In addition, erythrocytes of the rare Bombay phenotype, which completely lack ABH antigens, were not agglutinated. Native viral particles from extracts of infected rabbit liver as well as virus-like particles from the recombinant virus capsid protein specifically bound to synthetic A and H type 2 blood group oligosaccharides. Both types of particles could attach to adult rabbit epithelial cells of the upper respiratory and digestive tracts. This binding paralleled that of anti-H type 2 blood group reagents and was inhibited by the H type 2-specific lectin UEA-I and polyacrylamide-conjugated H type 2 trisaccharide. Young rabbit tissues were almost devoid of A and H type 2 antigens, and only very weak binding of virus particles could be obtained on these tissues.

Tk, a new colon tumor-associated antigen resulting from altered O-glycosylation.

Erythrocyte polyagglutination antigens T and Tn are truncated O-glycan chains that are also carcinoma-associated antigens. We investigated whether Tk polyagglutination antigen could similarly be a carcinoma-associated marker and a target of immunotherapy. Monoclonal antibody LM389 was raised against Tk erythrocytes and tested by immunohistochemistry. LM389 strongly reacted with 48% human colorectal carcinomas. Labeling of normal tissues was visible on epithelial cells, mainly digestive, but was confined at a supranuclear level. Expression of the antigen on cloned human carcinoma cells correlated with sialosyl-Tn expression. O-Sialoglycoprotein endopeptidase treatment revealed that on carcinomas and cell lines, the epitope was present on O-glycans. Antibody specificity was determined using synthetic carbohydrates. Direct binding and inhibition studies indicated that LM389 best ligands were terminated by two branched N-acetylglucosamine units. Screening of murine cellular cell lines with LM389 allowed development of an experimental model with Tk-positive and -negative cells in syngeneic BDIX rats. Vaccination of rats with Tk erythrocytes provided a protection against growth of rat Tk-positive, but not of Tk-negative, tumor cells in association with the development of antibodies. Taken together, the results indicate that Tk polyagglutination antigen is a new colorectal carcinoma-associated antigen, absent from the normal cell surface, resulting from alteration of O-glycans biosynthesis and with potential as a target of immunotherapy.

Susceptibility of rat colon carcinoma cells to lymphokine activated killer-mediated cytotoxicity is decreased by alpha1,2-fucosylation.

The presence of alpha1,2-fucosylated glycans at the surface of rat colon carcinoma cells has been associated with an increased tumorigenicity and resistance to natural killer/lymphokine activated killer (NK/LAK) cytotoxicity. We now report that transfection of rat alpha1,2-fucosyltransferases cDNA (FTA and FTB) into REG cells, which are spontaneously devoid of this enzymatic activity, allows expression of histo-blood group H antigen and increases their resistance to LAK, but not NK cell lysis. Conversely, transfection of PRO cells, which spontaneously express alpha1, 2-fucosyltransferase activity, with the FTA cDNA in the antisense orientation decreases expression of the H antigen together with their resistance to LAK cell lysis, but again, not to NK cell lysis. Furthermore, REG cells that are rejected by immunocompetent syngeneic rats are similarly rejected by rats depleted of NK cells by antibody 3.2.3, directed against the NKR-P1 molecule. Thus, the rejection of REG cells by immunocompetent rats and their earlier reported increased tumorigenicity after transfection with an alpha1, 2-fucosyltransferase cDNA cannot be ascribed to NK cell sensitivity or resistance, respectively. The increased resistance to LAK cell lysis, however, may be relevant to tumor progression.

alpha1,2Fucosyltransferase increases resistance to apoptosis of rat colon carcinoma cells.

Accumulation of histo-blood group antigens such as Lewis b, Lewis Y and H in colon cancer is indicative of poor prognosis. It is accompanied by increase in alpha1,2fucosyl-transferase activity, a key enzyme for synthesis of these antigens. Using a model of colon carcinoma, we previously showed that alpha1,2fucosylation increases tumorigenicity. We now show that tumorigenicity inversely correlates with the cells' sensitivity to apoptosis. In addition, poorly tumorigenic REG cells independently transfected with three different alpha1,2fucosyltransferase cDNAs, the human FUT1, the rat FTA and FTB were more resistant than control cells to apoptosis induced in vitro by serum deprivation. Inversely, PRO cells, spontaneously tumorigenic in immunocompetent syngeneic animals and able to synthesize alpha1,2fucosylated glycans, became more sensitive to apoptosis after transfection with a fragment of the FTA cDNA in the antisense orientation. Expression of alpha1,2fucosyl-transferase in poorly tumorigenic REG cells dramatically enhanced their tumorigenicity in syngeneic rats. However, in immunodeficient animals, both control and alpha1,2fuco-syltransferase transfected REG cells were fully tumorigenic and metastatic, indicating that the presence of alpha1,2fucosylated antigens allowed REG tumor cells to escape immune control. Taken together, the results show that increased tumorigenicity mediated by alpha1,2fucosyl-ation is associated to increased resistance to apoptosis and to escape from immune control.

Role for alpha1,2-fucosyltransferase and histo-blood group antigen H type 2 in resistance of rat colon carcinoma cells to 5-fluorouracil.

5-Fluorouracil (5-FU) is a drug of standard use in chemotherapy of colon carcinoma. However, its efficacy is limited by inherent and acquired cell resistance. Major changes in histo-blood group antigenic expression, at times associated with poor prognosis, occur on colon cancer cells. To assess whether these antigens might play a role in the resistance to 5-FU, a rat model of colon carcinoma was used. We observed that in vivo treatment of tumors with the drug increased expression of antigen H type 2. The increase was also observed after in vitro short-term exposure to 5-FU, as well as on a cell-resistant variant selected by continuous exposure to the drug, and was accompanied by an increase in alpha1,2-fucosyltransferase activity, the key enzyme involved in synthesis of H antigens. Transfection of cells devoid of this enzymatic activity by an alpha1, 2-fucosyltransferase cDNA allowed expression of H type 2 antigen and increased resistance to 5-FU. Inversely, transfection of cells which possess enzymatic activity by a cDNA in anti-sense orientation reduced both H type 2 cell-surface antigen and resistance to the drug. These results demonstrate that, in this experimental model, alpha1,2-fucosyltransferase and H type 2 antigen are involved in cellular resistance to 5-FU.

Increased tumorigenicity of rat colon carcinoma cells after alpha1,2-fucosyltransferase FTA anti-sense cDNA transfection.

Accumulation of histo-blood group antigens such as Lewis b, Lewis Y and H increases tumor cell motility and tumorigenesis. Alpha1,2-fucosylation is a key step in the synthesis of these antigens. Two alpha1,2-fucosyltransferases, expressed in colorectal carcinomas, have been characterized (FUT1 and FUT2 in humans, FTA and FTB in rats). To define the relative contribution of each of these enzymes in tumor cell behavior, we have used an anti-sense transfection approach in rat colon carcinoma PROb cells, which synthesize mRNA encoding for both enzymes. We have previously reported that anti-sense transfection of a cDNA fragment of the FTB enzyme decreased H antigenic cell-surface levels and concomitantly decreased tumorigenicity. H antigens, detected by antibodies specific for H type 1, 3 or 4, were detected only on a splice variant of CD44 containing the product of exon v6. We now report the anti-sense transfection of an FTA cDNA fragment into PROb cells, which resulted in decreased enzymatic activity on a type 2 precursor and decreased cell-surface H type 2 antigen exclusively. Compared to controls, FTA anti-sense-transfected cells were significantly more tumorigenic in syngeneic animals but not in immunodeficient SCID mice. The UEA-I lectin, specific for H type 2, revealed that these structures were present on the CD44v6 variant and on an uncharacterized 80-kDa glycoprotein. Our results indicate that FTA and FTB fucosylate distinct glycan chains in the same cell, leading to opposite effects, under control of the immune system.

A rat experimental model for the design of vaccines against tumor associated antigens Tn and Sialyl-Tn.

Clones either strongly or barely expressing the Tn and Sialyl-Tn antigens were isolated from a rat colon carcinoma cell line. Expression of the antigens in normal rat tissues was very restricted and vaccination using Ovine Submaxillary Mucin as the immunogen could delay growth of the Sialyl-Tn positive cells, but not of the Sialyl-Tn negative cells in syngeneic rats. The model should be useful for testing new anti-Tn or Sialyl-Tn vaccination protocols.

Detection of a potential receptor for the H-blood-group antigen on rat colon-carcinoma cells and normal tissues.

Up-regulation of the synthesis of carbohydrate tumor-associated antigens terminated by the disaccharide Fucalpha1-2Gal is frequent in colon carcinoma and associated with poor prognosis. There is evidence that Fucalpha1-2Gal (H-disaccharide) structures increase cancer-cell motility and tumorigenicity by as-yet unknown mechanisms. Using polyacrylamide-based neoglycoconjugates, we looked for a potential receptor for this disaccharide, and observed that a neoglycoconjugate probe containing the H-disaccharide could bind rat colon-carcinoma cells in a dose-dependent manner, whereas very little binding was evidenced when a probe containing glucose was used. Binding of the H-disaccharide probe could be inhibited by the free H-disaccharide as well as by unlabeled neoglycoconjugates containing a terminal H-disaccharide. The best inhibitor was the H-type-1 trisaccharide neoglycoconjugate. Histochemical detection of the potential H-receptor was performed on rat normal tissues and in situ 1,2-dimethylhydrazine-induced colon carcinomas. A strong binding of the H-disaccharide probe was evidenced on most tumors that could be partly inhibited by the trisaccharide Fucalpha1-2Galbeta1-4Glc and by the unlabeled H-disaccharide neoglycoconjugate, indicating carbohydrate specificity of the binding. Staining of normal colonic mucosa was much weaker. Strong staining was also observed on some normal tissues, such as the spleen or lymph nodes, while others, such as lungs or liver, were negative. Probes containing glucose or the Lewis-a trisaccharide did not stain tumors or normal tissues. These results provide preliminary evidence for the existence of H-specific binding sites, the number of which increases in colon carcinoma.

Carbohydrate-based probes for detection of cellular lectins.

Carbohydrate (spacered saccharide residue, Glyc) probes with various tags were synthesized as analytical tools for study of cellular lectins, i.e., Glyc-polyacrylamide-3H, Glyc-PAA-biotin, Glyc-PAA-fluorescein (flu), and Glyc-PAA-digoxigenin, where PAA is a soluble polyacrylamide carrier of approximately 30 kDa. Binding of all types of probes, where Glyc is the sialyl Lewis X (SiaLeX) tetrasaccharide or a blank saccharide, was assessed using Chinese hamster ovary (CHO) cells either transfected with the E-selectin cDNA or mock-transfected. High binding of SiaLeX-PAA-3H to E-selectin-transfected cells and absence of binding to control cells (both native and permeabilized) allowed the conclusion that the polyacrylamide carrier and the spacer arm do not contribute significantly to the binding. The biotinylated probe showed a high level of nonspecific binding in cell enzyme-linked assays. A similarly built digoxigenin-labeled probe was significantly better. In flow cytometry assays, the fluorescein probe demonstrated a specific binding to E-selectin-transfected cells of a similar level to that given by an anti-E-selectin antibody. In addition, it could be inhibited by the anti-E-selectin antibody, further demonstrating specificity. Tumors were obtained from nude mice by injection of CHO E-selectin or mock-transfected cells. The fluorescent SiaLeX-PAA-flu probe could bind to tumor sections from E-selectin-positive CHO cells, but not from control CHO cells. These probes can thus be used to reveal specifically complex carbohydrate-binding sites on cells either in culture or on tissue sections.

5-Fluorouracil-resistant colonic tumors are highly responsive to sodium butyrate/interleukin-2 bitherapy in rats.

Advanced colorectal cancer is generally refractory to 5-fluorouracil (5-FU) chemotherapy. This is linked to the emergence of resistant cell populations, probably due to a selection process. The identification of molecular markers and the improvement of alternative therapies thus remain important. We have used as an experimental model a rat colon cancer cell line (PROb), which exhibits features similar to those of the human situation. 5-FU treatment of rats bearing PROb tumors enhanced their survival but did not lead to cure. A PROb 5-FU-resistant subline (PRObR1) was obtained by continuous in vitro exposure to 5-FU. Resistance to 5-FU was accompanied by a 2-fold increase in thymidylate synthase activity and a substantially higher incorporation of thymidine in the presence of 5-FU, compared with parental PROb cells. Unexpectedly, in syngeneic rats, PRObR1 tumors exhibited delayed growth when compared with parental PROb tumors. This was ascribed to an increased sensitivity of PRObR1 cells to host immune response since no growth delay was observed in immunocompromised nude mice and since there was no detectable difference in proliferation rates between PROb and PRObR1 cells. 5-FU treatment was inefficient in prolonging the survival of rats bearing PRObR1 tumors. In contrast, an immunotherapeutic protocol combining sodium butyrate and recombinant interleukin-2 (NaBut/rIL-2) cured 80% of the rats bearing established PRObR1 tumors. Our results suggest that NaBut/rIL-2 treatment is efficient against 5-FU-chemoresistant rat colon cancer.

Increase of rat colon carcinoma cells tumorigenicity by alpha(1-2) fucosyltransferase gene transfection.

Fucosylated histo-blood group antigens such as Lewis b, Lewis Y, and H accumulate in colon carcinoma and this is accompanied by a clear increase in alpha(1-2)fucosyltransferase activity, a key enzyme for the biosynthesis of these antigens. Yet the biological significance of alpha(1-2) fucosylated structures is not well defined. We have transfected a poorly tumorigenic rat colon carcinoma cell line with the human H blood group alpha(1-2)fucosyltransferase cDNA. This resulted in cell surface expression of H antigens with a concomitant decrease of sialic acid substituted and free beta-galactosides. Immunoprecipitation experiments showed that H antigens were essentially borne by variants of CD44 carrying amino acid sequences encoded by exon v6. The transfected cells showed increased motility in a wound healing assay, without changing their proliferation rates. Parental and control cells transfected with an empty vector formed small tumors that always regressed after 30 days when injected subcutaneously to syngeneic rats. In contrast, alpha(1-2)fucosyltransferase transfectants were able to form progressive tumors. Increased tumorigenicity was also visible in nude mice. These results demonstrate that alpha(1-2)fucosylated antigens contribute directly to aggressiveness of colon carcinoma cells. This could occur by altering a function of CD44 variants.

Immunohistological analysis of antibodies against ABH and other glycoconjugates in normal human pyloric and duodenal mucosae.

Eighty-eight monoclonal antibodies were tested by immunohistochemistry on human gastro-duodenal mucosae of known ABO, Lewis and Secretor phenotypes. Antibodies were classified among anti-A, anti-B, anti-AB, anti-H and other anti-glycoconjugates (I, i, T, Tn, Lewis, P1, Tk). Anti-A, B and AB antibodies were subdivided into subgroups with "broad" or "restricted" reactivity according to the extent of epithelial cell labeling. Anti-H antibodies were classified in accordance to their degree of sensitivity to the secretor phenotype. Among anti-T and anti-I antibodies, only one of each showed positive staining of epithelial cells. All anti-Lewis antibodies had distinct reactivities, although, they were clearly anti-Lewis reagents. Some anti-P1 antibodies labeled epithelial cells, irrespective of the ABO, Lewis and secretor phenotypes. One anti-Tk stained the Golgi apparatus of most epithelial cells, irrespective of the individual's phenotype. In conclusion, some of the antibodies tested were defined as very useful reagents for immunohistochemistry showing both specificity and sensitivity.