Elevated glucose decreases the content of a basement membrane associated heparan sulphate proteoglycan in proliferating cultured porcine mesangial cells

Anionic biopolyelectrolytes of the syndecan/perlecan superfamily: physicochemical properties and medical significance

In the review article presented here, we demonstrate that the connective tissue is more than just a matrix for cells and a passive scaffold to provide physical support. The extracellular matrix can be subdivided into proteins (collagen, elastin), glycoconjugates (structural glycoproteins, proteoglycans) and glycosaminoglycans (hyaluronan). Our main focus rests on the anionic biopolyelectrolytes of the perlecan/syndecan superfamily which belongs to extracellular matrix and cell membrane integral proteoglycans. Though the extracellular domain of the syndecans may well be performing a structural role within the extracellular matrix, a key function of this class of membrane intercalated proteoglycans may be to act as signal transducers across the plasma membrane and thus be more appropriately included in the group of cell surface receptors. Nevertheless, there is a continuum in functions of syndecans and perlecans, especially with respect to their structural role and biomedical significance. HS/CS proteoglycans are receptor sites for lipoprotein binding thus intervening directly in lipid metabolism. We could show that among all lipoproteins, HDL has the highest affinity to these proteoglycans and thus instals a feedforward forechecking loop against atherogenic apoB100 lipoprotein deposition on surface membranes and in subendothelial spaces. Therefore, HDL is not only responsible for VLDL/IDL/LDL cholesterol exit but also controls thoroughly the entry. This way, it inhibits arteriosclerotic nanoplaque formation. The ternary complex 'lipoprotein receptor (HS/CS-PG) - lipoprotein (LDL, oxLDL, Lp(a)) - calcium' may be interpreted as arteriosclerotic nanoplaque build-up on the molecular level before any cellular reactivity, possibly representing the arteriosclerotic primary lesion combined with endothelial dysfunction. With laser-based ellipsometry we could demonstrate that nanoplaque formation is a Ca(2+)-driven process. In an in vitro biosensor application of HS-PG coated silica surfaces we tested nanoplaque formation and size in clinical trials with cardiovascular high-risk patients who underwent treatment with ginkgo or fluvastatin. While ginkgo reduced nanoplaque formation (size) by 14.3% (23.4%) in the isolated apoB100 lipid fraction at a normal blood Ca(2+) concentration, the effect of the statin with a reduction of 44.1% (25.4%) was more pronounced. In addition, ginkgo showed beneficial effects on several biomarkers of oxidative stress and inflammation. Besides acting as peripheral lipoprotein binding receptor, HS/CS-PG is crucially implicated in blood flow sensing. A sensor molecule has to fulfil certain mechanochemical and mechanoelectrical requirements. It should possess viscoelastic and cation binding properties capable of undergoing conformational changes caused both mechanically and electrostatically. Moreover, the latter should be ion-specific. Under no-flow conditions, the viscoelastic polyelectrolyte at the endothelium - blood interface assumes a random coil form. Blood flow causes a conformational change from the random coil state to the directed filament structure state. This conformational transition effects a protein unfurling and molecular elongation of the GAG side chains like in a 'stretched' spring. This configuration is therefore combined with an increase in binding sites for Na(+) ions. Counterion migration of Na(+) along the polysaccharide chain is followed by transmembrane Na(+) influx into the endothelial cell and by endothelial cell membrane depolarization. The simultaneous Ca(2+) influx releases NO and PGI2, vasodilatation is the consequence. Decrease in flow reverses the process. Binding of Ca(2+) and/or apoB100 lipoproteins (nanoplaque formation) impairs the flow sensor function. The physicochemical and functional properties of proteoglycans are due to their amphiphilicity and anionic polyelectrolyte character. Thus, they potently interact with cations, albeit in a rather complex manner. Utilizing (23)Na(+) and (39)K(+) NMR techniques, we could show that, both in HS-PG solutions and in native vascular connective tissue, the mode of interaction for monovalent cations is competition. Mg(2+) and Ca(2+) ions, however, induced a conformational change leading to an increased allosteric, cooperative K(+) and Na(+) binding, respectively. Since extracellular matrices and basement membranes form a tight-fitting sheath around the cell membrane of muscle and Schwann cells, in particular around sinus node cells of the heart, and underlie all epithelial and endothelial cell sheets and tubes, a release of cations from or an adsorption to these polyanionic macromolecules can transiently lead to fast and drastic activity changes in these tiny extracellular tissue compartments. The ionic currents underlying pacemaker and action potential of sinus node cells are fundamentally modulated. Therefore, these polyelectrolytic ion binding characteristics directly contribute to and intervene into heart rhythm.

Changes in cultured endothelial cell glycosaminoglycans under hyperglycemic conditions and the effect of insulin and heparin

BACKGROUND

Heparan sulfate proteoglycans (HSPGs) contain glycosaminoglycan (GAG) chains made primarily of heparan sulfate (HS). Hyperglycemia in diabetes leads to endothelial injury and nephropathy, retinopathy and atherosclerosis. Decreased HSPG may contribute to diabetic endothelial injury. Decreased tissue HS in diabetes has been reported, however, endothelial HS changes are poorly studied.

OBJECTIVE

To determine total GAGs, including HS, in endothelium under hyperglycemic conditions and the protective effect of insulin and heparin.

METHODS

Confluent primary porcine aortic endothelial cells (PAECs) were divided into control, glucose (30 mM), insulin (0.01 unit/ml) and glucose plus insulin treatment groups for 24, 48 and 72 hours. Additionally, PAECs were treated with glucose, heparin (0.5 microg/ml) and glucose plus heparin for 72 hours. GAGs were isolated from cells and medium. GAG concentrations were determined by the carbazole assay and agarose gel electrophoresis.

RESULTS

GAGs were significantly increased only in control and glucose plus insulin groups at 72 versus 24 hours. Glucose decreased cell GAGs and increased medium GAGs, and insulin alone decreased cell GAGs at all times compared to control. In the glucose plus insulin group, cell GAGs were less than control at 24 hours, and greater than glucose or insulin alone at 48 and 72 hours while GAGs in medium were greater than control at all times and glucose at 72 hours. Heparin increased GAGs in glucose treated cells and medium.

CONCLUSION

High glucose and insulin alone reduces endothelial GAGs. In hyperglycemic conditions, heparin or insulin preserves GAGs which may protect cells from injury. Insulin is an effective diabetic therapy since it not only lowers blood glucose, but also protects endothelium.

Heparan sulfate proteoglycans in experimental models of diabetes: a role for perlecan in diabetes complications

Proteoglycans are ubiquitous extracellular proteins that serve a variety of functions throughout the organism. Unlike other glycoproteins, proteoglycans are classified based on the structure of the glycosaminoglycan carbohydrate chains, not the core proteins. Perlecan, a member of the heparan sulfate proteoglycan (HSPG) family, has been implicated in many complications of diabetes. Decreased levels of perlecan have been observed in the kidney and in other organs, both in patients with diabetes and in animal models. Perlecan has an important role in the maintenance of the glomerular filtration barrier. Decreased perlecan in the glomerular basement membrane has a central role in the development of diabetic albuminuria. The involvement of this proteoglycan in diabetic complications and the possible mechanisms underlying such a role have been addressed using a variety of models. Due to the importance of nephropathy among diabetic patients most of the studies conducted so far relate to diabetes effects on perlecan in different types of kidney cells. The various diabetic models used have provided information on some of the mechanisms underlying perlecan's role in diabetes as well as on possible factors affecting its regulation. However, many other aspects of perlecan metabolism still await full elucidation. The present review provides a description of the models that have been used to study HSPG and in particular perlecan metabolism in diabetes and some of the factors that have been found to be important in the regulation of perlecan.

Protective effect of aminoguanidine upon capillary and submesothelial anionic sites

This study evaluates albuminuria and peritoneal permeability to albumin in control and diabetic rats, as well as in diabetic animals treated with subcutaneously injected aminoguanidine hydrochloride (Ag) (5 mg/100 g/day), during a follow-up period of 6 months. Aminoguanidine effectively prevented albuminuria and albumin extravasation in the mesenteric interstitial tissue (control, 0.43 +/- 0.11 microg EB/100 g of dry tissue, Ag, 0.60 +/- 0.44; untreated diabetic animals, 1.22 +/- 0.73; control and Ag group vs untreated diabetic rats, P < 0.001). Albumin D/P ratio of the aminoguanidine-exposed rats (0.017 +/- 0.011) was higher than that of controls (0.008 +/- 0.002), but significantly lower (P < 0.001) than values observed in the untreated group of animals (0.046 +/- 0.003). Administration of aminoguanidine preserved both submesothelial and subendothelial electronegative charges. For capillary basement membrane (BM), control at zero time, 32 +/- 4 AS/microm BM; control at 6 months, 33.4; aminoguanidine-treated rats, 35 +/- 2. For submesothelial BM, control at zero time, 33 +/- 3; control at 6 months, 32 +/- 3; aminoguanidine-treated rats, 35 +/- 3. Splitting and thickening of both basement membranes were not prevented by the therapeutic intervention. We conclude that the shielding effect of aminoguanidine upon the permselectivity capabilities of the endothelial and mesothelial monolayers appears to be connected, basically to the preservation of anionic fixed charges.

High glucose modifies heparansulphate synthesis by mouse glomerular epithelial cells

BACKGROUND

Alterations in proteoglycan metabolism are involved in the pathogenesis of diabetic nephropathy. The aim of this study is to evaluate the effects of high glucose on proteoglycan production and to find a reliable in vitro model for the study of diabetic nephropathy.

METHODS

A clone of mouse glomerular epithelial cells was cultured in media containing elevated (30 mmol) and physiological (5 mmol) glucose, or iso-osmolar (30 mmol) mannitol concentrations. We evaluated the synthesis of 35SO4-labeled molecules and the amount of proteoglycans by Sepharose CL6B and DEAE-Sephacel chromatographies.

RESULTS

A clear decrease (56%) in total cell-layer proteoglycan synthesis was induced by 30 mmol glucose, in comparison with normal glucose. A reduction of 25% in medium associated proteoglycan synthesis was observed in high glucose cultured cells. After Sepharose CL6B, in cells cultured in high glucose, cell layer heparansulphate proteoglycan-I (Kav 6B 0. 04) synthesis was reduced by about 81%, heparansulphate proteoglycan-II (Kav 6B 0.21) by about 87% and heparansulphate glycosaminoglycan (Kav 0.4-0.8) by about 91%, respectively. In mannitol-incubated cells the reductions observed were less evident and not significantly different from those in normal glucose.

CONCLUSIONS

These results indicate that (1) glomerular epithelial cells play a central role in proteoglycan synthesis, (2) high glucose modifies the amount and influences the different species production of these macromolecules, while osmotic forces seem to be only partially involved in these effects, and (3) this cellular clone of glomerular epithelial cells can represent a reliable in vitro model for the study of the mechanisms involved in diabetic nephropathy.

Growth factors and the kidney in diabetes mellitus

Nephromegaly and mesangial matrix expansion observed in the diabetic kidney are all clues of a role of growth factors in the pathogenesis of these lesions. A growing body of evidence shows that changes in (1) insulin-like growth factor I regulation, and (2) the transforming growth factor beta loop exist in the kidney in the diabetic hypertrophic kidney and in diabetic glomerulosclerosis. However, other growth factors may be involved in some diabetic renal changes. The abnormalities in growth factor content and regulation, the role of growth factors in the diabetic kidney, and the effect of hyperglycemia and advanced glycosylation end products on growth factors in the kidney are reviewed.

Markers of diabetic nephropathy

It is well established that the detection of microalbuminuria in a patient with diabetes mellitus indicates the presence of glomerular involvement in early renal damage. Recent studies have demonstrated that there is also a tubular component to renal complications of diabetes, as shown by the detection of renal tubular proteins and enzymes in the urine. In fact, tubular involvement may precede glomerular involvement, as several of these tubular proteins and enzymes are detectable even before the appearance of microalbuminuria. This review looks at the studies reported so far on serum and urinary markers of diabetic nephropathy, both glomerular and tubular, and their roles in the early detection of renal damage. The advantages and disadvantages of some of these markers are also discussed. The markers reviewed include (1) glomerular--transferrin, fibronectin, and other components of glomerular extracellular matrix, and (2) tubular--low molecular weight proteins (beta 2 microglobulin, retinol binding protein, alpha 1 microglobulin, urine protein 1), other proteins such as Tamm-Horsfall protein, beta 2 glycoprotein-1, urinary enzymes (N-acetyl-beta-D-glucosaminidase, cholinesterase, gamma glutamyltranspeptidase, alanine aminopeptidase), and tubular brush-border antigen.

Posttranscriptional effects of glucose on proteoglycan expression in mesangial cells

Hyperglycemic conditions are known to increase mRNA and protein levels of several extracellular matrix molecules in cultured mesangial cells, but accompanying increases in proteoglycan mRNA have not been found, and there are discrepant reports of normal or decreased proteoglycan synthesis with or without undersulfation in diabetic kidneys and hyperglycemic cultures. We examined the effects in proliferating cells of glucose on [35S]Sulfate incorporation into heparan and dermatan sulfates and on mRNA levels of decorin, biglycan, and basement membrane perlecan. In both mesangial cells and vascular smooth muscle cells, 30 mmol/L glucose caused a decrease of 15% to 25% in the amount of sulfate incorporated into each proteoglycan in cultures confluent for 1 to 4 days, compared with 10 mmol/L glucose. The effect showed no specificity for the class of proteoglycan and was not a consequence of changes in total protein synthesis, which increased, or cell proliferation, which was unaffected. No decrease in charge density of any of the proteoglycan fractions was observed by ion-exchange chromatography. Therefore, the decrease in labeling was due to a decrease in synthesis and not undersulfation. mRNA levels for biglycan and perlecan increased slightly and transiently, and these changes cannot account for the decreased synthesis. Decorin mRNA was detected only in smooth muscle cells, where it and biglycan were differentially affected by glucose, apparently at the transcriptional level; stabilities of the two messages were unaffected by glucose. Although transforming growth factor-beta 1 (TGF-beta 1) mRNA levels increased in response to glucose, the cytokine did not appear to regulate proteoglycan synthesis, because structural changes in proteoglycans elicited by addition of TGF-beta 1 to the culture medium did not occur in the hyperglycemic cultures. On the other hand, inhibition and downregulation of protein kinase C (PKC), while decreasing net sulfate incorporation into mesangial cell proteoglycans, prevented the effect of high glucose. We conclude that a high glucose concentration causes a general decrease in the synthesis of all classes of proteoglycans at a posttranscriptional level, and can do so without affecting the charge density of individual proteoglycan molecules.

Increased peritoneal permeability to albumin in streptozotocin diabetic rats

The mechanism behind the increased peritoneal permeability to albumin in diabetics is still unclear. In this study, streptozotocin diabetic rats developed albuminuria and significantly increased D/P of albumin after the fourth week of disease, reaching peak levels at the end of the 24 week period of follow-up. Coincidentally, extravasation of albumin to the interstitial tissue was evaluated with the Evans-blue method. Age-matched control rats showed Evans-blue concentrations of 0.023 +/- 0.013 micrograms/100 mg of dry tissue, whereas in diabetics the numbers were 1.22 +/- 0.719 micrograms (P < 0.001). Perfusion with Ruthenium-Red (RR) done in control at zero time, and in age-matched intact as well as in diabetic rats after 24 weeks of disease showed that the density distribution of capillary subendothelial anionic sites was significantly lower for diabetics (13 +/- 3/microns basement membrane vs. 31 +/- 3 and 34 +/- 4 in control groups; P < 0.001). Similar findings were made on the mesenteric submesothelial basement membrane. Mean density of RR decorated anionic sites was 12 +/- 2/microns basement membrane in diabetics, whereas those observed in both control groups were 31 +/- 2 and 31 +/- 3/microns (P < 0.001). Therefore, this reduced density of microvascular and submesothelial negative charges, equivalent to that induced by diabetes in other capillary beds, appears to be at the origin of the decreased permselectivity of the diabetic peritoneum for anionic serum albumin.

Increased hyaluronan production in the glomeruli from diabetic rats: a link between glucose-induced prostaglandin production and reduced sulphated proteoglycan

Exposure in vivo or in vitro to elevated glucose increases production of vasoactive prostaglandins by glomeruli and mesangial cells. This study aimed to determine whether this increased prostaglandin production could provide a link with later structural changes in diabetic nephropathy. Glomerular cores were prepared from control rats and streptozotocin-diabetic rats (3 weeks' duration). Over 24 h in culture hyaluronan production from diabetic glomerular cores was higher than production from control glomerular cores whether maintained in 5.6 mmol/l glucose (105.6 +/- 15.5 vs 53.6 +/- 8.5 ng hyaluronan per 250 glomerular cores, p < 0.001); in 25 mmol/l glucose (149.3 +/- 34.8 vs 62.7 +/- 7.8 ng hyaluronan per 250 glomerular cores, p < 0.01); or in 45 mmol/l glucose (176.8 +/- 23.3 vs 102.0 +/- 17.9 ng hyaluronan per 250 glomerular cores, p < 0.01). At 5.6 mmol/l glucose, exposure in vitro to prostaglandin E2 caused an increase in hyaluronan production [maximal at 10(-9) mol/l prostaglandin E2, 237 +/- 19 vs 42 +/- 4, ng hyaluronan per 250 glomerular cores, p < 0.001 (control) and 195 +/- 7 vs 103 +/- 5, ng hyaluronan per 250 glomerular cores, p < 0.001 (diabetic)]. In both control and diabetic glomerular cores hyaluronan production was reduced significantly by the cyclooxygenase inhibitor indomethacin (10(-5) mol/l) [24.7 +/- 3.33 vs. 40.25 +/- 4.11 ng hyaluronan per 250 glomerular cores, p < 0.05 (control) and 36.5 +/- 6.25 vs 118.0 +/- 22.6, p < 0.01 (diabetic)]. A direct spectrophotometric microassay was used to determine the concentration of sulphated glycosaminoglycans derived from papain-digested glomerular core proteoglycans.(ABSTRACT TRUNCATED AT 250 WORDS)

Protective effect of heparin on renal glomerular anionic sites of streptozotocin-injected rats

Albuminuria at the rate of 500 micrograms/24 h was observed in rats treated with 50 mg/kg body wt. i.v. streptozotocin (STZ). When STZ was combined with heparin, administered twice daily (250 IU/kg subcutaneous injection) after 24 h following STZ injection, the daily urinary albumin excretion (U-AE) was less than half the amount found in non-heparinized rats (280.3 micrograms/24 h). The observation period in both instances was 8 weeks. When animals were permitted to develop albuminuria over the first 4 weeks, subsequent heparin administration for another 4 weeks lowered U-AE from 500 micrograms/24 h to less than half the amount (227.8 micrograms/24 h). A significantly negative correlation (P < 0.001) existed between U-AE and the number of anionic sites (AS) in the lamina rara externa of glomerular basement membranes, as visualized by electron microscopy. The number of AS per 1000 nm in STZ-injected rats (15.5 +/- 0.2) was lower than that in control rats (23.1 +/- 0.6); however, in heparinized animals, regardless of STZ, the values were not significantly different from normal. Heparin by itself had no effect on the number of AS in normal rats. All STZ-injected animals became hyperglycemic (400-550 mg/dl), but received no insulin. Heparin had no effect on plasma glucose levels. From these results, it is concluded that heparin suppressed both of an increase in U-AE and a decrease in AS count of glomerular basement membranes in STZ-injected rats.

Proteoglycans of basement membranes

Proteoglycans carrying either heparan sulfate and/or chondroitin sulfate side chains are typical constituents of basement membranes. The most prominent proteoglycan (perlecan) consists of a 400-500 kDa core protein and three heparan sulfate chains. Electron microscopy and cDNA sequencing show a complex and elongated domain structure for the core protein which in part is homologous to that of the laminin A chain. This structure may be varied by alternative splicing and proteolysis. Integration into basement membranes probably occurs by heparan sulfate binding to laminin and collagen IV, core protein binding to nidogen and by limited self assembly. The proteoglycan is in addition a cell-adhesive protein which is recognized by beta 1 integrins. Several more proteoglycans with smaller core proteins (10-160 kDa) apparently exist in basement membranes but are less well characterized. Biological functions include control of filtration through basement membranes and binding of growth factors and protease inhibitors.

Proteoglycans of basement membranes

Proteoglycans carrying either heparan sulfate and/or chondroitin sulfate side chains are typical constituents of basement membranes. The most prominent proteoglycan (perlecan) consists of a 400-500 kDa core protein and three heparan sulfate chains. Electron microscopy and cDNA sequencing show a complex and elongated domain structure for the core protein which in part is homologous to that of the laminin A chain. This structure may be varied by alternative splicing and proteolysis. Integration into basement membranes probably occurs by heparan sulfate binding to laminin and collagen IV, core protein binding to nidogen and by limited self assembly. The proteoglycan is in addition a cell-adhesive protein which is recognized by beta 1 integrins. Several more proteoglycans with smaller core proteins (10-160 kDa) apparently exist in basement membranes but are less well characterized. Biological functions include control of filtration through basement membranes and binding of growth factors and protease inhibitors.

Differential modulation of mitogenic and metabolic actions of insulin-like growth factor I in rat glomerular mesangial cells in high glucose culture

In order to explore the possible contribution of insulin-like growth factor I to the development of diabetic nephropathy, the effect of glucose on the mitogenic and metabolic actions of insulin-like growth factor I in cultured rat glomerular mesangial cells was examined. The stimulation of [3H]-thymidine incorporation by insulin-like growth factor I in the cells exposed to high concentrations (55 mmol/l) of glucose (4.6 +/- 1.3 fold stimulation) was significantly suppressed as compared with that in the cells cultured in 11 mmol/l glucose (17.5 +/- 0.8 fold). In contrast, [3H]-amino-isobutylic acid uptake into the mesangial cells was significantly enhanced by glucose (2.03 +/- 0.03 nmol.mg protein-1. 15 min-1 at 55 mmol/l glucose vs 0.59 +/- 0.01 at 11 mmol/l glucose), while 2-deoxyglucose uptake remained unchanged. [125I]-insulin-like growth factor I binding was slightly but significantly increased in the cells exposed to high concentrations of glucose. Thus, glucose may modulate the mitogenic and metabolic actions of insulin-like growth factor I differently in cultured mesangial cells probably at the post-insulin-like growth factor I receptor level. These results may indicate that the differential modulation of the actions of insulin-like growth factor I by glucose could result in the increase in amino acid uptake and decrease in the cell proliferation in the mesangial cells, possibly leading to enhanced mesangial matrix synthesis with a relatively small increase in mesangial cell volume as seen in diabetic nephropathy.

Alterations of glomerular matrix proteins in the pathogenesis of diabetic nephropathy

Diabetic late complications are characterized by morphological and biochemical alterations of the extracellular matrix. In particular, longstanding diabetes causes quantitative and qualitative changes in basement membrane structure of retinal and renal capillaries. Immunohistochemical investigations of diabetic kidneys with diffuse glomerulosclerosis show increased collagen type IV deposition in the mesangial matrix and decreased heparan sulfate proteoglycan content in the mesangial matrix and glomerular basement membrane as well. In nodular glomerulosclerosis normal basement membrane components are decreased or absent while the occurrence of collagen type III in this stage has been interpreted as an irreversible alteration of the glomerular structure. These changes seem to be the underlying cause for the alterations in renal functions like persistent albuminuria and proteinuria. Increased intra- and extracellular levels of glucose and its derivatives are thought to be responsible for diabetic tissue dysfunction although there are reports on possible genetic defects causing increased susceptibility to develop diabetic nephropathy. Recent results, however, focus on the role of glucose-induced cytokine secretion as mediator for altered metabolism of glomerular matrix proteins. In vitro studies with cultured kidney cells have shown that the glucose-induced dysregulation of the basement membrane synthesis may be mediated by a glucose dependent activation of protein kinase C. Alternatively or synergistically, the formation of AGE products formed after prolonged exposure of matrix proteins to elevated glucose may also lead to cytokine secretion subsequently inducing synthesis of extracellular matrix proteins. Studies in experimental animals confirm the diabetes induced dysregulation of the synthesis of extracellular matrix components on the molecular level.

Increases in plasma lysosomal enzymes in type 1 (insulin-dependent) diabetes mellitus: relationship to diabetic complications and glycaemic control

Lysosomal enzymes degrade membrane glycoconjugates, and increased circulating enzyme activity may be an important mechanism in the pathogenesis of diabetic microangiopathy. We have assayed a profile of seven lysosomal enzyme activities (nmol.h-1.ml-1) in platelet-free plasma from 54 Type 1 (insulin-dependent) diabetic subjects (median age 31 years) and 42 matched normal control subjects. A significant increase in median (interquartile range) enzyme activity was measured in diabetic compared to control subjects for beta-D-glucuronidase, 121 (97.7-171) vs 88.8 (62.8-113), p less than 0.001; beta-D-Nacetylglucosaminidase, 693 (568-799) vs 568 (462-686), p less than 0.001; alpha-D-mannosidase, 23.8 (16.7-28.9) vs 14.5 (10.1-20.0), p less than 0.001; and beta-D-galactosidase, 6.94 (6.11-9.99) vs 6.66 (4.78-8.33), p less than 0.04. In contrast, alpha-L-fucosidase, alpha-D-galactosidase and beta-D-mannosidase activities were similar in diabetic and control subjects. None of the enzyme activities differed significantly (p less than 0.05) between 24 diabetic patients with clinical complications and 30 complication-free diabetic patients with similar glycaemic control which does not support the hypothesis that enzyme increases in diabetes arise simply by leakage from damaged tissues. In the diabetic subjects HbA1, median (interquartile range) 9.10 (7.40-10.60), was significantly related to beta-D-glucuronidase (rs = 0.56, p less than 0.001) and beta-D-Nacetylglucosaminidase (rs = 0.55, p less than 0.001). We have therefore demonstrated in diabetic subjects an increase in certain lysosomal glycosidases, that correlates with glycaemic control.(ABSTRACT TRUNCATED AT 250 WORDS)