1. n-Butyrate, a short chain fatty acid produced by colonic fermentation, induces differentiation in human neoplastic cell lines, and reduces expression in vitro of a sialyltransferase that glycosylates N-linked glycoproteins in hepatoblastoma cells. Gangliosides are amphipathic, sialylated glycosphingolipids that undergo profound changes in many transformed cells and may protect neoplastic cells from host immune surveillance. Colonic mucosal cells are exposed to luminal short-chain fatty acid concentrations of up to 80 mmol/l, and there is some evidence that short-chain fatty acids may alter ganglioside expression in colon cancer cells. 2. Because of the importance of gangliosides in cancer pathogenesis, we investigated the effects of n-butyrate on ganglioside expression of colonic (human and murine) and non-colonic cancer cells. 3. Three separate colon cancer cell lines (LS174T, T84 and MCA-38), when butyrate treated, demonstrated striking amplification of specific individual gangliosides. However, the total lipid-bound sialic acid content of gangliosides of butyrate-treated LS174T cells diminished. In contrast to earlier reports, n-butyrate did not mediate expression of all gangliosides and specifically did not mediate expression of G M3 . This effect persisted even after removal of butyrate. 4. In contrast, exposure of extracolonic cells to butyrate, including cervical cancer (HeLa) and laryngeal cancer (HEp-2) cell lines in this study and hepatoblastoma cells (Hep G2) in our previous work, caused no detectable changes in ganglioside expression. 5. In conclusion, our results indicate a relative tissue specificity of butyrate-mediated alterations in ganglioside expression that is not universal but is limited to specific gangliosides.

1. We have shown that serum total sialic acid is elevated in hypertriglyceridaemic patients showing the Frederickson's type IIB phenotype in comparison with normal subjects (2.30 ±0.34 versus 1.92 ± 0.32 mmol/l, P <0.02). Lipid-associated sialic acid was also elevated in the hypertriglyceridaemic group in comparison with the normal subjects (0.60 ±0.09 versus 0.35 ±0.04 mmol/l, P <0.001). 2. We measured five serum acute-phase proteins in the hypertriglyceridaemic patients and the normal subjects, namely α 1 -antichymotrypsin, α 1 -acid glycoprotein, α 2 -macroglobulin, C-reactive protein and haptoglobin. Serum α 2 -macroglobulin was significantly elevated in the hypertriglyceridaemic patients compared with the normal subjects (2.1 ±1.0 versus 1.5 ± 0.55 g/l, P <0.05) as was serum C-reactive protein (5.9 ±3.5 versus 3.5±1.9mg/l, P <0.05). There were, however, no significant differences in the serum concentrations of α 1 -antichymotrypsin, α 1 -acid glycoprotein or haptoglobin between the two groups. 3. There was a significant correlation between serum total sialic acid and serum α 1 -antichymotrypsin, α 1 -acid glycoprotein and haptoglobin (Spearman correlation coefficients 0.74, 0.50 and 0.76, respectively) in the normal subjects, and there was a significant correlation between serum total sialic acid and serum α 1 -antichymotrypsin and α 1 -acid glycoprotein (Spearman correlation coefficients 0.72 and 0.84, respectively) in the hypertriglyceridaemic patients.

1. Serum total sialic acid concentration, recently shown to be a cardiovascular risk factor, and also serum lipid-associated sialic acid concentration were measured in 15 patients with hypertriglyceridaemia (fasting serum triacylglycerol concentration > 2.3 mmol/l) showing a Frederickson's type IIB phenotype, 15 patients with hypercholesterolaemia showing a IIA phenotype and 15 age- and sex-matched normal control subjects. 2. Total serum sialic acid concentration was significantly raised in the hypertriglyceridaemic group (84.9 ± 21.5 versus 64.9 ± 20.8 mg/dl, P <0.03, Mann—Whitney U -test) compared with the normal control group, as was serum lipid-associated sialic acid concentration (23.0 ± 4.3 versus 12.0 ± 3.2 mg/dl, respectively, P< 0.001, Mann—Whitney U -test). 3. Serum total sialic acid concentration was also significantly elevated in the hypertriglyceridaemic group as compared with the IIA phenotype hypercholesterolaemic group (84.9 ± 21.5 versus 58.4 ± 11.7 mg/dl, P <0.03, Mann—Whitney U -test), as was serum lipid-associated sialic acid concentration (23.0 ± 4.3 versus 14.9 ± 4.7 mg/dl, P< 0.001, Mann—Whitney U -test). 4. We suggest that serum concentrations of both total sialic acid and lipid-associated sialic acid may be useful markers of cardiovascular risk which could, in part, be related to hypertriglyceridaemia.

1. Sialic acid moieties of erythrocyte membrane glycoproteins are the principal determinants of the negative charge on the cell surface. The resultant electrostatic repulsion between the cells reduces erythrocyte aggregation and hence the low shear rate viscosity and yield stress of blood. 2. Using g.c.–m.s., a decrease in sialic acid content has been observed in the major erythrocyte membrane glycoprotein, glycophorin A, obtained from nine diabetic patients compared with that from seven normal control subjects [median (range): 3.30 (0.01–11.90) versus 18.60 (3.20–32.60) μg/100 μg of protein, P <0.02]. 3. Erythrocyte aggregation, measured by viscometry as the ratio of suspension viscosity to supernatant viscosity ( L s / S ) in fibrinogen solution, was increased in ten diabetic patients compared with ten normal control subjects (mean ± sem , 37.6 ± 1.3 versus 33.8 ± 0.6, P <0.02). 4. In the patients in whom both viscometry and carbohydrate analysis were performed, the decrease in erythrocyte glycophorin sialylation and the increase in erythrocyte aggregation in fibrinogen solution were related statistically( L S / S correlated negatively with glycophorin sialic acid content, r = 0.73, P <0.05). 5. Decreased glycophorin sialylation provides an explanation at the molecular level for increased erythrocyte aggregation and it may be important in the pathogenesis of vascular disease in diabetes.

1. The activities of enzymes degrading human colonic mucin were examined in faecal specimens from healthy subjects and patients with inflammatory bowel disease. 2. The activity of sialidase was measured using a new physiological substrate related to mucus glycoproteins. In addition, acylneuraminate pyruvate-lyase ( N -acetylneuraminate lyase; EC 4.1.3.3.) and a novel O -acetylsialic acid esterase (sialate O -acetylesterase; EC 3.1.1.35) were detected. 3. The O -acetylsialic acid esterase activity was readily detectable in partially purified fractions after Sephadex G-100 chromatography. 4. Patients with inflammatory bowel disease showed significant increases in acylneuraminate pyruvate-lyase and proteinase activity but sialidase activity did not differ from normal. The activity of these enzymes in neutrophils could not account for the differences observed.

1. Transient mild dermal inflammation was produced in rats by the subcutaneous injection of either carageenan or monosodium urate crystals. The activities of enzymes of the sialic acid metabolic pathway were measured in liver and colon at 8 h, 3 days and 7 days. 2. In both liver and colon the UDP- N -acetyl- d -glucosamine 2-epimerase and asialo-α 1 -acid glycoprotein sialyltransferase activities increased relative to controls while that of CMP-sialic acid hydrolase decreased. Similar changes occurred for both carageenan and monosodium urate crystals. 3. These results indicate that the acute phase response to a relatively minor inflammatory lesion causes significant changes in a distant and apparently unrelated tissue.

1. Glomerular sialic acid was chemically measured in rats with experimental proteinuria induced by N,N '-diacetylbenzidine (DAB) or with autoimmune Heymann nephritis. 2. In DAB nephrosis and in Heymann nephritis the relative amount of glomerular protein was increased. In DAB nephrosis the quantity of sialic acid expressed per amount of protein was decreased, but expressed per amount of DNA, which reflects the number of cells, there was no significant change. In Heymann nephritis the amount of sialic acid was not significantly altered when expressed per amount of protein or per amount of DNA. 3. In individual animals, the amount of glomerular sialic acid expressed per amount of protein or per amount of DNA did not correlate with the severity of proteinuria. An increase in the total number of glomerular cells was found only in Heymann nephritis. 4. The results suggest that proteinuria in the reported models is not the result of changes in glomerular sialic acid.

1. The organic matrices of 12 kidney stones containing calcium and two composed of uric acid were solubilized, with ethylenediaminetetra-acetate for the former and sodium hydroxide for the latter. 2. The solubilized matrices and residual fine particulate material were examined for sialic acid by the thiobarbituric acid method. 3. Sialic acid was found in every stone in either the soluble and/or insoluble material. 4. The identity of sialic acid was confirmed by the absorption spectrum of the colour produced and by its release from the protein by neuraminidase. 5. The presence of sialic acid in all stones despite widely varying composition suggests that it may be passively deposited.

1. Ferritin has been partially purified from the serum of patients with idiopathic haemochromatosis. 2. Incubation with neuraminidase of this partially purified serum ferritin eliminated much of the microheterogeneity of the protein so that only ferritin of isoelectric point approximately 5·8 was present. 3. There was no change in the total amount of ferritin present (measured immunologically) or in the percentage of ferritin binding to concanavalin A. 4. Incubation of liver, spleen or heart ferritin with neuraminidase did not change the isoelectric focusing patterns.

1. The glomerular sialic acid content of kidneys from diabetic subjects, aged subjects and healthy young individuals was determined. 2. Similar mean sialic acid contents were found in the glomeruli of elderly diabetic subjects, and in age- and sex-matched non-diabetic subjects. 3. Higher mean sialic acid content was found in glomeruli from healthy, young non-diabetic subjects. 4. These results, in conjunction with other reported work, indicate that low sialic acid content may result from both diabetic and age changes in the glomeruli.