Liquid chromatography "on-flow" 1H nuclear magnetic resonance on native glycosphingolipid mixtures together with gas chromatography/mass spectrometry on the released oligosaccharides for screening and characterisation of carbohydrate-based antigens from pig lungs.

Glycosphingolipids were prepared from pig lung and pooled into two fractions with (i) < or = 3 sugar residues, and (ii) > or = 3 sugar residues. Oligosaccharides were prepared and used for gas chromatography, gas chromatography/mass spectrometry and matrix-assisted laser desorption/ionization mass spectrometry. The glycolipid fractions i and ii were further characterised and purified using a novel method based on high performance liquid chromatography "on-flow" proton nuclear magnetic resonance. The LC "on-flow" NMR technique showed good chromatographic separation and gave NMR spectral information which could be used as guidance for pooling of the separated mixture glycolipids. Conventional 1H NMR, thin layer immunostaining, gas chromatography, gas chromatography/mass spectrometry and matrix-assisted laser desorption/ionization mass spectrometry were used to characterise the glycolipids and to validate LC-NMR spectral data.

alpha-Gal epitopes in animal tissue glycoproteins and glycolipids.

alpha-Gal terminated saccharides are present on the cell surface both as glycolipids and glycoproteins in all mammals except Old World monkeys and humans. The structural diversity among identified saccharides terminated by this epitope in animal tissues is steadily increasing. The majority of these saccharides have the alpha-Gal linked to lactosamine but other core saccharides exist. The alpha-Gal terminated saccharides are recognized by the immune system as a specific antigen and antibodies directed to the alpha-Gal, which do not cross-react with the classic blood group B trisaccharide, are found in man and Old World monkeys. Similar to other complex carbohydrate cell surface antigens, the alpha-Gal epitope is heterogeneously distributed in different organs and in different cells within an organ. It is present on the vascular endothelium and it is the primary target for human naturally occurring antibodies following pig to primate/man xenotransplantation leading to hyperacute rejection of the graft. Important for the future will be to further structurally characterize this antigen system, its cellular/subcellular distribution, and to identify possible of additional glycosyltransferases, related to the already described alpha 1,3galactosyltransferase that may explain the structural diversity. Such information will be of importance in the studies of, for example, the pathogenesis of autoimmune diseases and for the production of genetically modified pigs to prevent xenograft rejection.

Leb glycolipids present in the lumen of the gastrointestinal tract of rats do not enter the plasma compartment.

We orally administered to rats several times more Leb glycolipids than is proportionally found in the gastrointestinal tract of humans. This was done in an effort to study two potential phenomena: the possibility that glycolipids in plasma may originate from glycolipids derived from the lumen of the gastrointestinal tract, and to investigate the potential to secondarily modify in vivo the glycolipid profile of gastrointestinal tract epithelial cells, a phenomenon clearly established for human erythrocytes, leukocytes, and platelets. We were able to establish that some of the orally administered glycolipids can be detected at the surface of the upper region mucosa of the gastrointestinal tract for more than 24 hours and are essentially excreted intact in stools in less than 72 hours. Some fecal degradation of the Leb glycolipids into Lea and H type 1 did occur. Although we clearly established that the glycolipids were present in the mucus layer adherent to the cell surface, we could not conclusively establish if the glycolipids had inserted into the epithelial cell membrane. This, however, could not be excluded. The fact that the fed glycolipids remained in the mucus layer of the upper region of the gastrointestinal tract for at least 24 hours may have some pharmacological value. Using sensitive techniques, including red cell serology, immunohistology, and immunochemistry of glycolipids isolated from plasma and red cells, there was no evidence that the fed Leb glycolipids reached the plasma compartment, thus suggesting that glycolipids present in the lumen of the gastrointestinal tract cannot reach the circulation.

An ELISA technique for quantitation of human xenoantibodies binding to pig cells: application in patients with pig kidneys extracorporeally connected to the circulation.

A quantitative ELISA technique for determination of human anti-pig xenoantibody number in serum samples has been established using pig lymphocytes and pig/rabbit erythrocytes as target cells and a pool of serum from human blood group AB donors. The number of low affinity antibodies binding to the cells was determined by quantitation following the use of aqueous washing of the cells and separation of bound and unbound antibodies with the phthalate oil method. The efficiency of different soluble Gal(alpha)1-3Gal-terminating di- and tri-saccharides to inhibit antibody binding was tested and found to vary between 70-90% at a saccharide concentration of 10 mg/ml. The assay was used to evaluate the antibody changes in two patients who, after plasmapheresis treatments, had pig kidneys extracorporeally connected to their blood circulation. The number of anti-pig IgM/IgG antibodies bound to each pig lymphocyte were reduced from 5,600/13,200 to 1,300/3,100 in patient 1 and from 1,200/6,500 to 500/2,100 in patient 2 by three consecutive daily plasmapheresis treatments. Although the lymphocytotoxic titers were reduced to very low levels, the antibody numbers still present in the blood of patient 1 caused a hyperacute rejection of the pig kidney. However, the antibody levels in patient 2 did not cause rejection of this kidney during 15 min perfusion time. A strong anti-pig antibody response 3 weeks after the perfusion experiment was found in patient 1 as shown by 27,600/245,300 IgM/IgG molecules bound to pig lymphocytes corresponding to an increase of lymphocytotoxic titer from 8 to 512. The second patient showed a much weaker immune response with 1,400/19,800 IgM/IgG antibodies corresponding to a lymphocytotoxic titer increase from 8 to 32. The use of this quantitation technique enables more accurate investigation of antibody binding to xenogenic target cells than conventional titration techniques.

Point mutations and deletion responsible for the Bombay H null and the Reunion H weak blood groups.

OBJECTIVE: Definition of the molecular basis of the Reunion and the Bombay red cell and salivary H-deficient phenotypes. METHODS: Sequence and expression of FUT1 and FUT2 genes from H-deficient individuals. Family segregation analysis of the mutations responsible for the fucosyltransferase defects of H, secretor and Lewis systems. RESULTS: The Indian red cell H null Bombay phenotype depends on a new mutation of the FUT1 gene. T725-->G changing Leu242-->Arg. Their salivary nonsecretor phenotype is secondary to a complete deletion of the FUT2 gene. The red cell H weak Reunion phenotype depends on another new mutation of FUT1, C349-->T which induces a change of His117-->Tyr. Their salivary nonsecretor phenotype is due to the known Caucasian inactivating mutation G428-->A. CONCLUSION: Single prevalent FUT1 and FUT2 point mutations and a deletion are responsible for the Indian Bombay H null and the Reunion H weak phenotypes found on Reunion island. This is in contrast with other H-deficient phenotypes where sporadic nonprevalent inactivating mutations are the rule.

Glycosphingolipid expression in pig aorta: identification of possible target antigens for human natural antibodies.

Total non-acid glycosphingolipids were isolated from the aortas of more than 80 pigs. The glycolipids were separated by HPLC, analysed by thin-layer chromatography, and tested for reactivity with monoclonal anti-blood group antibodies. The fractions were structurally characterized by NMR spectroscopy and mass spectrometry. Reactivity with both anti-blood group A and H antibodies was seen. The major glycosphingolipid constituents were globotri- and globotetraosylceramides and blood group H pentaglycosylceramides based on type 1 and type 2 core saccharide chains. Globopentaosylceramides, blood group H hexaglycosylceramides based on type 4 chain, and blood group A hexaglycosylceramides based on type 1 core chain were also present. Two structures, that may be important targets for human antibodies initiating hyperacute rejection following pig to human xenotransplantation, were present as minor constituents compared to the blood group components. These were Galalpha1,3neolactotetraosylceramide and a Galalpha1, 3Lexstructure. A Leb/Y hexaglycosylceramide was also present.

Electrospray ionization and collision-induced dissociation time-of-flight mass spectrometry of neutral glycosphingolipids.

A series of native naturally occurring neutral glycosphingolipids has been analysed by electrospray ionization tandem mass spectrometry using a hybrid magnetic sector-TOF instrument. The collision-induced dissociation products of precursor ions were detected by an orthogonal acceleration time-of-flight mass spectrometer as the second analyser. Glycosphingolipids, with mono- to hexa-saccharide chain lengths with different ceramide constituents, were studied. The result of electrospray ionization in the positive ion mode generally showed singly charged molecular ions with Na+ as adduct, [M + Na]+. The sensitivity of the electrospray ionization was greatly enhanced by addition of NaCl, LiCl (forming [M + Li]+) or KCl (yielding [M + K]+) to the sample. A comparison between the collision-induced dissociation of precursor molecular ions of monoglycosylceramides, using Na+, Li+ and K+ as adducting species, showed that the intensity of the fragment ions and the extent of the daughter ion fragmentation of the molecular ions, are dependent on the type of adduct used. The daughter ion spectra of Li+ adduct ions showed intense sequence fragment ions, both of the saccharide chain and the ceramide moiety, and were superior to those obtained using Na+ or K+. The collision-induced dissociation spectra of the [M + Li]+ ions, of glycosphingolipids containing di- to hexasaccharides, are also presented. Proposed possible fragments, resulting from the CID of the molecular ions [M + Li]+ of monoglycosylceramides, are shown.

Rapid and sensitive GC/MS characterization of glycolipid released Galalpha1,3Gal-terminated oligosaccharides from small organ specimens of a single pig.

Pig to human xenotransplantation is considered a possible solution to the prevailing chronic lack of human donor organs for allotransplantation. The Galalpha1,3Gal determinant is the major porcine xenogeneic epitope causing hyperacute rejection following human antibody binding and complement activation. In order to characterize the tissue distribution of Galalpha1,3Gal-containing and blood group-type glycosphingolipids in pig, acid and nonacid glycosphingolipids were isolated from the kidney, small intestine, spleen, salivary gland, liver, and heart of a single pig obtained from a semi-inbred strain homozygous at the SLA locus. Glycolipids were analyzed by thin-layer immunostaining using monoclonal antibodies, and following ceramide glycanase cleavage as permethylated oligosaccharides by gas chromatography, gas chromatography-mass spectrometry, and matrix-assisted laser desorption/ionization mass spectrometry. The kidney contained large amounts of Galalpha1,3Gal-containing penta- and hexasaccharides having carbohydrate sequences consistent with the Galalpha1,3nLc4and Galalpha1,3Lexstructures, respectively. The former structure was tentatively identified in all organs by GC/MS. The presence of extended Galalpha1,3Gal-terminated structures in the kidney and heart was suggested by antibody binding, and GC/MS indicated the presence of a Galalpha1,3nLc6structure in the heart. The kidney, spleen, and heart contained blood group H pentaglycosylceramides based on type 1 (H-5-1) and type 2 (H-5-2) chains, and H hexaglycosylceramides based on the type 4 chain (H-6-4). In the intestine H-5-1 and H-6-4 were expressed, in the salivary gland H-5-1 and H-5-2, whereas only the H-5-1 structure was identified in the liver. Blood group A structures were identified in the salivary gland and the heart by antibody binding and GC/MS, indicating an organ-specific expression of blood group AH antigens in the pig.

Chemical and lectin-gold electron microscopical studies of the expression of the Galalpha1-determinant in the pig aorta.

In the xenotransplantation research field, pig aortic endothelial cells are frequently used in different model systems, e.g., for the study of the xenoantibody-antigen reaction. The Gal(alpha1),3Gal determinant is the major target for human xenoreactive antibodies in pig tissue. Characterisation of the Gal(alpha1)- distribution in pig aortic endothelial cells is thus important for understanding the reaction occurring at the endothelial cells during the xenorejection. We have determined the complete structure of the major Gal(alpha1),3Gal terminated glycolipid, Gal(alpha1),3nLc4Cer. Structural studies were performed on isolated glycosphingolipids by mass spectrometry and NMR spectroscopy. The results show a predominance of the pentasaccharide among the Gal(alpha1)-terminated glycolipids but also the presence of several Gal(alpha1)-terminated glycolipids with extended carbohydrate core chains. Ultrastructural localisation of the Galalpha1-antigen in pig aorta was done by lectin-gold electron microscopic studies of aortic wall sections using the Griffonia simplicifolia isolectin B4. Gal(alpha1)-determinants are predominantly localised on the luminal surface of pig aortic endothelial cells and endothelial cells of vasa vasorum and, to a lesser extent, vascular subendothelium.

Biochemical and enzymatic characterization of blood group ABH and related histo-blood group glycosphingolipids in the epithelial cells of porcine small intestine.

Non-acid glycosphingolipids were isolated from small intestinal epithelial cells of a single blood group A pig. One very predominant blood group compound was obtained chemically pure upon HPLC fractionation. It was characterized by mass spectrometry and 1H NMR spectroscopy to be the type 1 chain blood group A hexaglycosylceramide. Support for the presence of minute amounts of additional A glycolipids was obtained by mass spectrometry and immunostaining of TLC plates with anti-A antibodies specific for A type 2 chain, A type 3 and 4 chain, and the ALe(b) determinant. Among precursor chains, globoside (type 4) and lactotetraosylceramide (type 1) were immunologically identified, whereas no neolactotetraosylceramide (type 2) and gangliotetraosylceramide reactivities were detected. We addressed the question whether the predominant expression of type 1 chain based A glycolipids reflects a restricted glycolipid precursor chain specificity of the alpha 1-2 fucosyl- and/or the alpha 1-3 N-acetylgalactosaminyltransferases, or if the biosynthesis of the precursor chains themselves is regulated. All precursor core saccharides, lacto- (type 1), neolacto-(type 2), and gangliotetraosylceramide as well as globopentaosylceramide (type 4), could serve as acceptors for fucose in vitro when a crude microsomal fraction obtained from mechanically released, porcine intestinal epithelial cells was used as an enzyme source. Under the same conditions an N-acetylgalactosamine residue could be transferred to the blood group H structures based on these core saccharide chains. Lactotriaosylceramide, but not gangliotriaosylceramide, could serve as an acceptor for UDP-galactose. When the product was digested with beta-galactosidase (EC 3.2.1.23) from S.pneumoniae, under conditions where it specifically cleaves Gal beta 1-4 residues, approximately 40% of the radioactivity was cleaved off, indicating that a substantial amount of neolactotetraosylceramide was made in vitro, as opposed to the predominance of lactotetraosylceramide-based structures found in vivo.

Blood group glycosphingolipid expression in kidney of an individual with the rare blood group A1 Le(a-b+) p phenotype: absence of blood group structures based on the globoseries.

Total neutral glycolipid fractions were isolated from kidney and ureter tissue obtained at autopsy of an individual of the rare blood group A1 Le(a-b+) p. The amount of glycolipids isolated were 3.7 and 2.5 mg g-1 dry tissue weight for the kidney and ureter tissue, which is in the range of reference blood group P kidneys. Part of the kidney glycolipid fraction was subfractionated by HPLC. Glycolipid compounds were structurally characterized by thin-layer chromatography (chemical detection and immunostaining with monoclonal antibodies), proton NMR spectroscopy and mass spectrometry. Globotriaosyl- and globotetraosyl-ceramides, which are the major compounds in kidneys of P individuals, were absent in the p kidney, and a comparatively increased amount of monoglycosyl- and lactosylceramides was found. A shift to longer fatty acyl chains in the ceramide part of lactosylceramides was noted. Elongated globoseries compounds with five to seven sugar residues, including the blood group A type 4 chain structure, were lacking. A slight increase in neolactotetraosyl- and blood group X pentaglycosyl-ceramides was noticed. The study confirms an enzymatic block in the conversion of lactosylceramide to elongated globoseries compounds in the kidney tissue similar to that of erythrocytes of p individuals.

Anti-carbohydrate antibodies associated with hyperacute rejection in a vascularized mouse heart-to-rat xenotransplantation model.

Specificity of immune reactions has always been sought, because it facilitates intervention with unwanted mechanisms. Specific carbohydrate antigens have been proposed to be targets of antibodies in early immune responses in pig-to-man xenografts. This work was undertaken to determine carbohydrate structure for antibody response in the experimental xenograft model mouse-to-rat. Glycolipids were prepared from nine different mouse organs and separated for carbohydrate size on thin layer plates. Sera taken from normal untreated rats showed only weak or absent IgM antibody-binding to the separated mouse glycolipids. This is in accordance with the observation that mouse heart grafts are not hyperacutely rejected by the rat. However, sera taken from mouse heart xenografted rats show clear IgG and IgM antibody binding to neutral glycolipids migrating in the five-sugar region of the thin-layer plate. These rats have previously been reported to hyperacutely reject a second xenograft. Glycolipids with this particular mobility and immunostaining properties are the dominant ones in the mouse caval vein preparation, which probably represents a rather pure vascular structure. The target antigen structure was identified, by mass spectrometry and proton nuclear magnetic resonance spectroscopy, to be the Forssman pentaglycosylceramide. A commercial monoclonal antibody directed toward the Forssman antigen bound the same biochemical structure as the antibodies derived from the mouse heart-xenografted rats. Most of the IgM activity, but very little of the IgG activity was adsorbed using the Forssman terminal disaccharide solid phase.

Extracorporeal ("ex vivo") connection of pig kidneys to humans. I. Clinical data and studies of platelet destruction.

The pioneering experiment by Welsh et al. (Immunological Lett 1991:29:167-170) connecting a pig kidney to the human circulation has been repeated in a modified manner. Two volunteer dialysis patients were pretreated by daily plasmapheresis on days -2,-1, and 0 to remove the naturally occurring anti-pig xenoantibodies. The anti-pig lymphocytotoxic liters were reduced from 1:8 to 1:2 in patient 1 and from 1:8 to 1:1 in patient 2. No steroids or immunosuppressive drugs were administrated before or during the experiments. A sterile pig kidney was extracorporeally ("ex vivo") connected to the patients a/v fistula using an arterial and a venous pump similar to a dialysis. The two experiments gave different results. In the first experiment the perfusion pressure was kept at 100 mmHg for the initial 25 min by reducing the pump speed until the minimum blood flow of 30 ml/min was reached. Thereafter, the pressure rose continuously and the experiment was terminated at 65 min at a perfusion pressure of 200 mmHg. The patient did not feel any discomfort during the perfusion. In the second experiment, a stable blood flow of 200 ml/min was reached at a pressure of 100 mmHg after a few minutes. The perfusion was terminated at 15 min when the patient developed chest and abdominal pain, hypotension, and electrocardiographic signs of myocardial ischemia. The patient recovered quickly. In the first experiment, small volumes of clear urine was produced until the pressure rose above 100 mmHg, which resulted in hematuria. In the second experiment clear urine (4 ml/min) was produced. (51)Chromium clearance values were after 15 min <1 ml/min for kidney 1 and 12 ml/min (8 ml/min/100 g) for kidney 2. A drastic reduction in platelet count (128 to 48 and 64 to 8 × 10(9)/1, respectively) during the passage through the kidney was found in blood samples collected simultaneously before and after the organ. No change in hemoglobin values and leucocyte counts were found. Light- and electron-microscopical analysis of the kidney tissues revealed for kidney 1 focal areas with obliteration of the glomerular and peritubular capillaries by platelets and PMN cells and severe damage of the endothelial cells comparable to a picture of a hyperacute rejection. In kidney 2, all vessels were patent but in the capillaries large amount of membrane fragments were detected by electron microscopy and a discrete damage of the endothelial cells were seen in some segments. No intact platelets were present in the vascular tree. These human experiments support the hypothesis that hyperacute rejection of pig to human xenografts is delayed in time by removal of the preformed anti-pig xenoantibodies. A new finding was a very rapid destruction of platelets occurring in the kidney of patient 2 who had very low liters of xenoantibodies. The humoral immune response is described in detail in an accompanying paper (Rydberg et al., this issue).

Extracorporeal ("ex vivo") connection of pig kidneys to humans. II. The anti-pig antibody response.

Pig kidneys were extracorporeally "ex vivo" connected to the circulation of two volunteer male dialysis patients (Breimer et al., this issue). The patients were pretreated by daily plasmapheresis for 3 consecutive days, which reduced the anti-pig lymphocytotoxic titer from 8 to 2 in the first patient and from 8 to 1 in the second patient. The anti-pig hemagglutinating titers were reduced from 32 to 4 in the first patient and from 2 to 1 in the second patient. No drugs, except heparin, were given. The perfusion lasted for 65 min in patient 1 and the experiment was terminated due to increased vascular resistance in the pig kidney. Ultrastructural investigation showed a picture similar to a hyperacute vascular rejection. Immunohistochemical studies showed a weak staining of IgM antibodies, but no IgG in the small arteries and glomeruli. The pig kidney of patient 2 was perfused for 15 min and the experiment terminated due to serious side effects of the patient. Light and electron microscopical investigation showed virtually no structural changes of the kidney tissue and immunostaining for human antibodies was negative. In both patients, serum samples collected 2-5 weeks postperfusion showed a strong anti-pig antibody titer rise (up to 512) which thereafter declined but stabilized on a higher level than before the experiment. The antibody response in the two patients was different. In patient 1, the major anti-pig antibodies directed to carbohydrate antigens were of IgG (IgG1 and IgG2 subclasses) type, while the IgM response was less prominent and virtually no IgA antibodies were produced. Despite the short duration of the perfusion in patient 2, a humoral immune response was seen that was mainly confined to the IgA immunoglobulin class (IgA1 subclass). Blood group glycospingolipid fractions, prepared from the contralateral kidney of the donor pigs, were used for immunostaining with patient serum samples. In both patients, the antibodies produced after the perfusion, mainly recognized the Galα1-3Gal epitope both as part of the "linear B" pentasaccharide but also on more complex carbohydrate structures. Patient 1 was HLA-immunized before the experiment due to a kidney allograft and had a panel reactivity of 85% before the perfusion. No change in the panel reactivity of HLA-antibodies was found after the perfusion experiments. Patient 2 had no HLA antibodies before and remained negative after the perfusion. Patient serum samples collected before and after the perfusion were tested for reactivity against human endothelial cell lines. No antibodies were generated.

Structural and immunochemical identification of Leb glycolipids in the plasma of a group O Le(a-b-) secretor.

Total non-acid glycosphingolipids were isolated from the plasma of a healthy red blood cell group O Le(a-b-) salivary ABH secretor individual. Glycolipids were fractionated by HPLC and combined into eight fractions based on chromatographic and immunoreactive properties. These glycolipid fractions were analysed by thin-layer chromatography and tested for Lewis activity with antibodies reactive to the type 1 precursor (Le(c)), H type 1 (Le(d)), Le(a) and Le(b) epitopes. Fractions were structurally characterized by mass spectrometry (EI-MS and LSIMS) and proton NMR spectroscopy. Expected blood group glycolipids, such as H type 1, (Fuc alpha 1-2Gal beta 1-3GlcNac beta 1-3Gal beta 1-4Glc beta 1-1Cer) were immunochemically and structurally identified. Inconsistent with the red cell phenotype and for the first time, small quantities of Le(b) blood group glycolipids (Fuc alpha 1-2Gal beta 1-3(Fuc alpha 1-4)GlcNAc beta 1-4Glc beta 1-1Cer) were immunochemically and structurally identified in the plasma of a Lewis-negative individual. These findings confirm recent immunological evidence suggesting the production of small amounts of Lewis antigens by Lewis negative individuals.

Expression of Lewis histo-blood group glycolipids in the plasma of individuals of Le(a+b+) and partial secretor phenotypes.

Red cell Lewis antigens are carried by glycosphingolipids passively absorbed from plasma. Plasma was collected from a spectrum of individuals with normal and unusual Lewis/secretor phenotypes in order to investigate the glycolipid basis for the unusual phenotypes. Samples were obtained from: a Le(a+b-) ABH nonsecretor who secreted Lewis substances; a Le(a+b-) partial secretor; Le(a+b+) partial secretors; Le(a+b+) secretors; and a full range of normal Lewis/secretor phenotypes as controls. The Le(a+b+) samples represented Polynesian, Asian and Réunion Island ethnic backgrounds. Nonacid glycolipids were prepared, separated by thin-layer chromatography, and then immunostained with potent monoclonal antibodies of known specificity. Despite different serological profiles of the Le(a+b-) and Le(a+b+) Polynesian samples, their plasma glycolipid expressions were very similar, with both Le(a) and Le(b) co-expressed. The copresence of Le(a) and Le(b) in Le(a+b+) samples is in marked contrast to Caucasians with normal Lewis phenotypes, who have predominantly either Le(a) or Le(b). These results suggest that there is a range of the secretor transferases in different individuals, possibly due to different penetrance or to several weak variants. We also show that Lewis epitopes on longer and/or more complex core chains appear to be predominant in the Polynesian Le(a+b+) samples. The formation of these extended glycolipids is compatible with the concept that in the presence of reduced secretor fucosyltransferase activity, increased elongation of the precursor chain occurs, which supports the postulate that fucosylation of the precursor prevents or at least markedly reduces chain elongation.

Immunochemical and immunohistological expression of Lewis histo-blood group antigens in small intestine including individuals of the Le(a+b+) and Le(a-b-) nonsecretor phenotypes.

Histological samples and total non-acid glycosphingolipids were prepared from small intestine of human cadavers with the Le(a+b+) and Le(a-b-) nonsecretor phenotypes and contrasted with the more common Lewis phenotypes. Glycolipid fractions were analysed by thin-layer chromatography and tested for Lewis activity with monoclonal antibodies reactive to Lewis epitopes. Paraffin-embedded small intestine sections were also fluorescently immunostained with anti-Lewis antibodies. Unlike the common Lewis positive phenotypes, we were immunochemically able to demonstrate the copresence of large amounts of Lea and Leb glycolipids in the Le(a+b+) sample. In addition we demonstrated increased formation of extended Lewis structures in this phenotype. By immunohistochemistry Lea, Leb and type 1 precursor chain epitopes could be demonstrated in the brush border. These results show that the expression of the Le(a+b+) phenotype at the erythrocyte phenotyping level parallels the small intestinal expression of this phenotype, and the patterns of Lewis antigen expressions are unique to this phenotype. By immunohistochemistry and immunochemistry we also demonstrated the presence of trace amounts of Lewis active glycoconjugates in the small intestine of the Le(a-b-) nonsecretor and Le(a+b-) samples. In the Le(a-b-) nonsecretor Lea and Leb activity was absent and type 1 precursor was present in brush border, while Leb activity was immunohistologically demonstrated in the Golgi apparatus of the deep glands. Trace amounts of both Lea and Leb glycolipids were identified in this sample. In parallel trace Leb activity could also be detected in the glycolipids of the Le(a+b-) sample and could be immunohistologically demonstrated to be fully expressed in occasional cells in the deep glands of the small intestine, a pattern quite dissimilar to that of the Le(a-b-) nonsecretor.(ABSTRACT TRUNCATED AT 250 WORDS)