Search

The role of the mammalian target of rapamycin (mTOR) was investigated in insulin responsive cell lines. mTOR was expressed at high levels in insulin responsive cell types and in 3T3-L1 cells mTOR expression levels increased dramatically as cells differentiated from fibroblasts into insulin responsive adipocytes. mTOR localized to membrane fractions in all cells tested and in 3T3-L1 adipocytes mTOR was specifically localized to microsomal membranes. Protein kinase activity directed towards mTOR was tightly associated with mTOR immunoprecipitates and this kinase activity was inhibited by FKBP12-rapamycin indicating it was due to an autokinase activity present in mTOR. The mTOR autokinase and the protein kinase activity of the p110 alpha isoform of PI 3-kinase shared several notable similarities; (a) both were maximally active in the presence of Mn2+ but also showed significant activity in the presence of Mg2+ (b) neither were inhibited by the presence of non-ionic detergent and (c) both were inhibited by wortmannin and LY294002, known inhibitors of the PI 3-kinase lipid kinase activity. These data taken together indicate the autokinase activity lay in the PI 3-kinase homology domain. In summary mTOR is a membrane anchored protein kinase that is active in conditions encountered in vivo and the fact it is highly expressed in insulin responsive cell types is consistent with a role in insulin signalling.

Isolated skeletal muscle from healthy individuals was used to evaluate the role of phosphoinositide 3-kinase (PI 3-kinase) in insulin signalling pathways regulating mitogen activated protein kinase (MAP-kinase) and protein kinase-B and to investigate whether MAP-kinase was involved in signalling pathways regulating glucose metabolism. Insulin stimulated glycogen synthase activity (approximately 1.7 fold), increased 3-o-methylglucose transport into human skeletal muscle strips (approximately 2 fold) and stimulated phosphorylation of the p42 ERK-2 isoform of MAP-kinase. This phosphorylation of p42 ERK2 was not blocked by the PI 3-kinase inhibitors LY294002 and wortmannin although it was blocked by the MAP-kinase kinase (MEK) inhibitor PD 98059. However, PD98059 (up to 20 micromol/l) did not block insulin activation of glycogen synthase or stimulation of 3-o-methylglucose transport. Wortmannin and LY294002 did block insulin stimulation of protein kinase-B (PKB) phosphorylation and stimulation of 3-o-methylglucose transport was inhibited by wortmannin (IC50 approximately 100 nmol/l). These results indicate that MAP-kinase is activated by insulin in human skeletal muscle by a PI 3-kinase independent pathway. Furthermore this activation is not necessary for insulin stimulation of glucose transport or activation of glycogen synthase in this tissue.

In summary, it is still unclear which type of PI 3-kinase is involved in the activation of glycogen synthase. However, there is a strong body of evidence to suggest class-1 PI 3-kinase activity is necessary for insulin stimulation of glucose transport and a growing body of evidence to suggest that insulin stimulation of this class of PI 3-kinase is sufficient to stimulate glucose transport. However, the mechanism by which PI 3-kinase lipid products stimulate glucose transport is still not clear.

The role of phosphoinositide 3-kinase (PI 3-kinase) in insulin signaling was evaluated in human skeletal muscle. Insulin stimulated both antiphosphotyrosine-precipitable PI 3-kinase activity and 3-O-methylglucose transport in isolated skeletal muscle (both approximately 2-3-fold). Insulin stimulation of 3-O-methylglucose transport was inhibited by the PI 3-kinase inhibitor LY294002 (IC50 = 2.5 microM). The PI 3-kinase adapter subunits were purified from muscle lysates using phosphopeptide beads based on the Tyr-751 region of the platelet-derived growth factor receptor. Immunoblotting of the material adsorbed onto the phosphopeptide beads revealed the presence of p85alpha, p85beta, p55(PIK)/p55gamma, and p50 adapter subunit isoforms. In addition, p85alpha-NSH2 antibodies recognized four adapter subunit variants of 54, 53, 48, and 46 kDa, the latter corresponding to the p50 splice variant. Serial immunoprecipitations demonstrated that these four proteins were associated with a large proportion of the total PI 3-kinase activity immunoprecipitated by p85alpha-NSH2 domain antibodies. Antibodies to p85beta, p55(PIK)/p55gamma, and the p50 adapter subunit also immunoprecipitated PI 3-kinase activity from human muscle lysates. A large proportion of the total cellular pool of the 53-kDa variant, p50, and p55(PIK) was present in antiphosphotyrosine immunoprecipitates from unstimulated muscle, whereas these immunoprecipitates contained only a very small proportion of the cellular pool of p85alpha, p85beta, and the 48-kDa variant. Insulin greatly increased the levels of the 48-kDa variant in antiphosphotyrosine immunoprecipitates and caused smaller -fold increases in the levels of p85alpha, p85beta, and the 53-kDa variant. The levels of p50 and p55(PIK) were not significantly changed. These properties indicate mechanisms by which specificity is achieved in the PI 3-kinase signaling system.

Insulin action on adipocytes was studied in the offspring of mothers who had been fed either a control (20% protein) or a low (8%)-protein diet during pregnancy and lactation. Adipocytes isolated from low-protein offspring had significantly higher basal and insulin-stimulated glucose uptakes than controls. This may be related to a threefold increase in insulin receptors in low-protein adipocytes. Consistent with these observed changes in glucose transport, adipocytes from low-protein animals had significantly higher basal and insulin-stimulated insulin receptor substrate (IRS)-1-associated phosphatidylinositol 3-kinase (PI 3-kinase) activities. There was also more p85-associated PI 3-kinase activity in these adipocytes. There was no difference in expression in the p85 regulatory subunit or the p110-alpha catalytic subunit of PI 3-kinase. In contrast, there was a sixfold reduction in the p110-beta catalytic subunit of PI 3-kinase in adipocytes from low-protein animals. These results suggest that poor fetal nutrition during pregnancy and lactation can have long-term effects on glucose transport and on the expression of key components of the insulin signaling pathway in adipocytes.

It is becoming well established that poor fetal and early postnatal growth can have long-term effects on adult health, including susceptibility to non-insulin-dependent diabetes mellitus, cardiovascular disease and hypertension. It is suggested that this results from poor nutrition during early life having permanent effects on the structure and metabolism of certain organs and tissues. In the present study we investigated the effect of a low-protein diet during pregnancy and lactation on adipocyte properties and glucose tolerance. Rat dams were fed on a diet containing either 200 (control) or 80 (low protein) g protein/kg during pregnancy and lactation. In addition cross-fostering techniques were employed to enable a separate evaluation of the prenatal and postnatal periods. All offspring were weaned onto a 200 g protein/kg diet at 21 d of age and then studied at 6 weeks of age. The mothers' protein supply during lactation appeared to be the most critical time window for long-term growth. In contrast, the offspring of mothers fed on a low-protein diet during pregnancy or lactation were significantly more glucose tolerant than controls, suggesting that both time windows can have long-term effects on glucose tolerance. In addition offspring of mothers fed on a low-protein diet during pregnancy or lactation had significantly smaller adipocytes than controls. However the largest reduction in adipocyte size was observed when there was a low-protein diet during both pregnancy and lactation. The amount of insulin receptor present in adipocyte membranes was increased in the three animal groups that had been exposed to the low-protein diets while levels of the insulin responsive glucose transporter (GLUT 4) were similar in adipocyte membranes from all groups.

The effects of insulin and phorbol 12-myristate 13-acetate (PMA) on the levels of cellular phosphoinositides were investigated in 3T3-L1 adipocytes. Stimulation for 4 min with PMA (1 microM) or insulin (10 nM) increased levels of PtdIns(3,4,5)P3 approx. 2-fold and 6-fold respectively. PMA also had a small effect on the cellular levels of PtdIns4P, whereas insulin had no effect on PtdIns4P levels; levels of PtdIns(4,5)P2 and PtdIns3P were not significantly affected by either agent. Insulin increased the levels of the p85 alpha subunit of phosphoinositide (PI) 3-kinase associated with membranes, whereas PMA decreased levels of membrane-associated p85 alpha. PMA did not increase PI 3-kinase activity in anti-phosphotyrosine or anti-p85 immunoprecipitates. The stimulation of glucose transport by insulin or PMA was blocked by 100 nM wortmannin or 10 ng/ml LY294002, indicating that PI 3-kinase is essential for stimulation by both agents. In summary, these results demonstrate: (1) that PMA and insulin stimulate PtdIns(3,4,5)P3 production by distinct mechanisms in 3T3-L1 adipocytes, and (2) that stimulation of PtdIns(3,4,5)P3 production by PMA is likely to be important in signalling pathways leading from PMA stimulation to end-point responses such as glucose transport.

To understand how the stimulation of phosphoinositide 3-kinase (PI 3-kinase) by different growth factors can activate different subsets of downstream responses, growth-factor regulation of PI 3-kinase activity at different intracellular locations was investigated in 3T3-L1 adipocytes. Insulin caused a large stimulation of glucose transport and stimulated recruitment of transferrin receptors to the plasma membrane (PM) in these cells, whereas platelet-derived growth factor (PDGF)-bb was virtually without effect on these responses. Subcellular fractionation studies after stimulation with PDGF-bb or insulin revealed a differential effect of these growth factors on subcellular localization of PI 3-kinase activity. PDGF was more effective than insulin in stimulating PI 3-kinase activity and recruiting the p85 alpha PI 3-kinase adaptor subunit in the fraction containing the PM. However, in the microsomal fraction insulin significantly increased PI 3-kinase activity and p85 alpha levels, whereas PDGF was almost without effect. In the microsomal membrane fraction the insulin-stimulated recruitment of p85 alpha closely matched the increase PI 3-kinase activity, indicating that insulin stimulation of PI 3-kinase in this fraction is largely due to recruitment of PI 3-kinase enzyme rather than alterations in specific activity. Insulin-stimulated recruitment of p85 alpha to the microsomal membranes was not inhibited by wortmannin, indicating that PI 3-kinase activity was not required for this process. A further level of compartment-specific regulation of PI 3-kinase in response to PDGF was revealed by the finding that tyrosine phosphorylation of the p85 alpha adaptor was restricted to the PM-containing fraction. Insulin had no effect on p85 tyrosine phosphorylation in either fraction. In summary, these results suggest a basis by which insulin and PDGF could both use PI 3-kinase signalling cascades but achieve different signalling outcomes.

Phosphoinositide 3-kinases (PI 3-kinases) and their 3-phosphoinositide products were identified initially as components of intracellular signalling pathways emanating from cell surface receptors. A new role for 3-phosphoinositides in the constitutive movement o f proteins from one intracellular compartment to another was proposed with the discovery of homology between the product of a yeast gene important for vacuolar sorting, Vps34p, and a mammalian PI 3-kinase. Recent studies have implicated PI 3-kinase as an essential component in membrane traffic at specific steps o f the trans-Golgi-network-endosomal pre-lysosomal system. Evidence largely emerging from the insulin-stimulated glucose transport system suggests that PI 3-kinase may also mediate the effects o f growth factors on membrane traffic events. These studies suggest a possible link between growth-factor-stimulated and constitutive membrane traffic in the endosomal system.

In summary, over-expression of GLUT4 selectively in fat causes increased flux of glucose into adipocytes and leads to increases in either the replication of immature pre-adipocytes or their differentiation into mature adipocytes resulting in an increase in fat cell number. This is the first model in which obesity is accounted for entirely by adipocyte hyperplasia and, therefore, is useful for studying the mechanisms involved in controlling fat cell number in vivo. GLUT4 over-expression in adipocytes of transgenic animals also increased whole- body insulin sensitivity. However, GLUT4 over-expression exclusively in adipocytes did not protect them from insulin resistance in vivo induced by high-fat feeding, in spite of the fact that insulin resistance was prevented at the level of the adipocyte. Interestingly, GLUT4 over-expression in fat protected the animals from developing further obesity when fed on a high-fat diet. It is possible that this failure to increase adiposity further is due to enhanced partitioning of glucose into fat, which may result in decreased glucose supply to muscle. This in turn may cause diversion of lipid to muscle to be oxidized as fatty acid. This diversion of lipid could result in protection against increased fat deposition in adipocytes. Further studies will be required in order to understand the molecular mechanisms by which GLUT4 over-expression in adipose tissues affects nutrient partitioning between muscle and adipose tissue and what the consequences of this are for whole-body fuel metabolism.

To determine whether fructose can be utilized as a metabolic substrate for skeletal muscle in man, we investigated its incorporation into glycogen, its oxidation and lactate production in isolated human skeletal muscle. Rates of fructose oxidation and incorporation into glycogen increased in the presence of increasing fructose concentrations (0.1-1.0 mM). Lactate production increased 3-fold when extracellular fructose was increased from 0.1 to 0.5 mM. Cytochalasin B, a competitive inhibitor of hexose transport mediated by the GLUT1 and GLUT4 facilitative glucose transporters, completely inhibited insulin-stimulated glucose incorporation into glycogen and glucose oxidation (P < 0.01), but did not alter fructose incorporation into glycogen or fructose oxidation. Insulin (1000 mu-units/ml) increased glucose incorporation into glycogen 2.7-fold and glucose oxidation 2.3-fold, whereas no effect on fructose incorporation into glycogen or fructose oxidation was noted. A physiological concentration of glucose (5 mM) decreased the rate of 0.5 mM fructose incorporation into glycogen by 60% (P < 0.001), whereas fructose oxidation was not altered in the presence of 5 mM glucose. Irrespective of fructose concentration, the majority of fructose taken up underwent non-oxidative metabolism. Lactate production accounted for approx. 80% of the fructose metabolism in the basal state and approx. 70% in the insulin (1000 mu-units/ml)-stimulated state. In the presence of 5 mM glucose, physiological concentrations of fructose could account for approximately 10-30% of hexose (glucose + fructose) incorporation into glycogen under non-insulin-stimulated conditions. In conclusion, fructose appears to be transported into human skeletal muscle via a carrier-mediated system that does not involve GLUT4 or GLUT1. Furthermore, under physiological conditions, fructose can significantly contribute to carbohydrate metabolism in human skeletal muscle.

The role of PI 3-kinase in transferrin receptor recycling was investigated in 3T3-L1 adipocytes. Wortmannin, a specific PI 3-kinase inhibitor, blocked insulin stimulated recruitment of transferrin receptors to the plasma membrane (IC50 approximately 5-15 nM) with a similar potency to its inhibitory effects on PI 3-kinase and glucose transport. Wortmannin also blocked insulin stimulated recruitment of IGF-2 receptors to the cell surface. However, wortmannin (up to 200 nM) and another PI 3-kinase inhibitor, LY294002 (3 micrograms/ml), did not block transferrin receptor endocytosis. This demonstrates that while PI 3-kinase may be necessary for insulin stimulated exocytosis, it is not necessary for endocytosis of transferrin receptors.

GLUT-4 expression varies widely among normal humans and those with obesity and diabetes. Using the alpha P2 promoter/enhancer ligated to the human GLUT-4 gene, we created transgenic mice to study the impact of alterations in GLUT-4 expression selectively in adipocytes on glucose homeostasis and body composition. Here we investigated molecular mechanisms for enhanced glucose tolerance and obesity in these mice. [U-14C]glucose incorporation into triglycerides, glyceride-glycerol, glyceride-fatty acids, CO2, and lactate was measured in adipocytes incubated at 3, 0.5, and 3 microM glucose with or without maximally stimulating insulin. In nontransgenic and transgenic mice, the major pathway for glucose metabolism shifts from lipogenesis at tracer glucose concentration to glycolysis at physiological glucose concentration. In transgenic adipocytes incubated at 3 microM glucose, metabolism via all major pathways is enhanced by 8.6- to 38-fold in the absence of insulin and 3- to 13-fold in the presence of insulin. At physiological glucose concentration, constitutive metabolism to triglycerides, CO2, and lactate is two- to threefold greater in transgenic than in nontransgenic adipocytes. De novo fatty acid synthesis is preferentially increased: 31-fold for basal and 21-fold for insulin-stimulated compared with nontransgenic adipocytes. Thus overexpression of GLUT-4 in adipocytes of transgenic mice results in increased glucose metabolism in all major pathways, with differential regulation of the pathways involved in lipogenesis.

High fat feeding is associated with impaired insulin action, an obese body composition, and down-regulation of glucose transporter-4 (GLUT4) expression in adipocytes. We recently showed that overexpression of GLUT4 selectively in adipocytes of transgenic mice using the aP2 (fatty acid-binding protein) promoter/enhancer results in enhanced glucose tolerance and adipocyte hyperplasia. Here, we fed these GLUT4-overexpressing transgenic mice a high fat (55%) or a low fat (10%) diet for 13-15 weeks to determine the role of alterations in GLUT4 expression in adipocytes in the development of insulin resistance and obesity, which are characteristic of high fat consumption. In nontransgenic mice, high fat feeding results in 45-50% reduction of GLUT4 levels in white and brown adipose tissue, with a parallel decrease in insulin-stimulated glucose transport. In transgenic mice receiving the low fat diet, GLUT4 is overexpressed 20-fold in white and 4-fold in brown adipose tissue. Glucose transport in epididymal adipocytes is increased 20-fold in the basal state and 6-fold in the insulin-stimulated state. Even after transgenic mice are fed a high fat diet, GLUT4 expression and glucose transport in their adipocytes remains 14- to 30-fold greater than that in nontransgenic mice receiving the same diet. Despite these marked effects at the adipose cell level, glucose tolerance is not improved, probably due to insulin resistance in skeletal muscle and liver, where the transgene is not expressed. During the low fat diet, transgenic mice have 80% more body lipid than nontransgenics. High fat feeding increases body lipid 76% and adipocyte size 65% in nontransgenic mice, but has no effect in transgenic mice. Thus, overexpression of GLUT4 selectively in adipocytes protects against a further increase in adiposity. Furthermore, by using a heterologous promoter, high level overexpression of GLUT4 can be maintained even under metabolic conditions where it is normally down-regulated in adipocytes. This overexpression results in markedly increased glucose transport at the cellular level, but adipose-specific GLUT4 overexpression does not prevent the decrease in glucose tolerance associated with high fat feeding.

We have investigated the involvement of phosphoinositide (PI) 3-kinase and p70 ribosomal protein-S6 kinase (p70s6k) in mediating insulin stimulation of glycogen synthesis in 3T3-L1 adipocytes using specific inhibitors. Wortmannin inhibited PI 3-kinase activity (IC50 approximately 10 nM), inhibition being complete at 100 nm. Wortmannin (100 nM) completely blocked the ability of insulin to activate glycogen synthase in 3T3-L1 adipocytes and the ability of insulin to stimulate glucose incorporation into glycogen in 3T3-L1 fibroblasts. Rapamycin, which blocks insulin-stimulated activation of p70s6k, decreased insulin activation of glycogen synthase in a dose-dependent manner (IC50 approximately 0.8 ng/ml), with a maximum approx. 75% inhibition of insulin's stimulatory effect. Rapamycin inhibited insulin-stimulated glucose incorporation into glycogen to a similar extent and with similar dose-dependency, while having no effect on insulin-stimulated glucose transport. We conclude that PI 3-kinase and p70s6k are involved in the signalling pathways by which insulin stimulates glycogen synthase in 3T3-L1 adipocytes.

To gain insight into the molecular pathogenesis of obesity and specifically the role of nutrient partitioning in the development of obesity, we overexpressed the insulin-responsive glucose transporter (GLUT4) in transgenic mice under the control of the fat-specific aP2 fatty acid-binding protein promoter/enhancer. Two lines of transgenic mice were generated, which overexpressed GLUT4 6-9-fold in white fat and 3-5-fold in brown fat with no overexpression in other tissues. In vivo glucose tolerance was enhanced in transgenic mice. In isolated epididymal, parametrial, and subcutaneous adipose cells from transgenic mice, basal glucose transport was 20-34-fold greater than in nontransgenic littermates. Insulin-stimulated glucose transport was 2-4-fold greater in cells from transgenic mice. Total body lipid was increased 2-3-fold in transgenic mice overexpressing GLUT4 in fat. Surprisingly, fat cell size was unaltered and fat cell number was increased > 2-fold. This is the first animal model in which increased fat mass results solely from adipocyte hyperplasia and it will be a valuable model for understanding the mechanisms responsible for fat cell replication and/or differentiation in vivo.

To investigate the tissue distribution of the GLUT3 glucose transporter isoform in human tissue we produced affinity purified antibodies to the COOH terminus of the human GLUT3. Both antibodies recognize a specific GLUT3 band in oocytes injected with GLUT3 mRNA but not in those injected with H2O or GLUT1, 2, 4, 5 mRNA. This immunoreactive band in GLUT3 injected oocytes is photolabelled by cytochalasin-B in the presence of L- but not D-glucose indicating that it is a glucose transporter. A high cross reactivity between the human GLUT3 antibodies and a 43 kDa cytoskeletal actin band was identified in all oocyte lysates and many human tissues. However, the specific GLUT3 band could be distinguished from the actin band by carbonate treatment which preferentially solubilized the actin band. Using these antibodies we show that GLUT3 is present as a 45-48 kDa protein in human brain with lower levels detectable in heart, placenta, liver and a barely detectable level in kidney. No GLUT3 was detected in membranes from any of 3 skeletal muscle groups investigated. We conclude that a major role of GLUT3 in humans is as the brain neuronal glucose transporter.

A recent study by C.F. Burant et al. (13) demonstrates that GLUT5 is a high-affinity fructose transporter with a much lower capacity to transport glucose. To characterize the potential role of GLUT5 in fructose and glucose transport in insulin-sensitive tissues, we investigated the distribution and insulin-stimulated translocation of the GLUT5 protein in human tissues by immunoblotting with an antibody to the COOH-terminus of the human GLUT5 sequence. GLUT5 was detected in postnuclear membranes from the small intestine, kidney, heart, four different skeletal muscle groups, and the brain, and in plasma membranes from adipocytes. Cytochalasin-B photolabeled a 53,000-M(r) protein in small intestine membranes that was immunoprecipitated by the GLUT5 antibody; labeling was inhibited by D- but not L-glucose. N-glycanase treatment resulted in a band of 45,000 M(r) in all tissues. Plasma membranes were prepared from isolated adipocytes from 5 nonobese and 4 obese subjects. Incubation of adipocytes from either group with 7 nM insulin did not recruit GLUT5 to the plasma membrane, in spite of a 54% insulin-stimulated increase in GLUT4 in nonobese subjects. Thus, GLUT5 appears to be a constitutive sugar transporter that is expressed in many tissues. Further studies are needed to define its overall contribution to fructose and glucose transport in insulin-responsive tissues and brain.

Northern blot analysis of human tissues has demonstrated the expression of the brain-type glucose transporter isoform (GLUT 3) in liver, muscle and fat, raising the possibility that this transporter isoform may play a role in the regulation of glucose disposal in these tissues in response to insulin. We have raised an anti-peptide antibody against the C-terminal 13 amino acids of the murine homologue of this transporter isoform, and determined its tissue distribution in mouse tissues and murine-derived cell lines. The antibodies recognise a glycoprotein of about 50 kilodaltons, expressed at high levels in murine brain. In contrast to human tissues, the expression of GLUT 3 in mice is restricted to the brain, and no immunoreactivity was observed in either liver, fat or muscle membranes, or in murine 3T3-L1 fibroblasts or adipocytes. In contrast, high levels of expression of this isoform were observed in the NG 108 neuroblastoma x glioma cell line, a hybrid cell derived from rat glioma and mouse neuroblastoma cells. Taken together, these data suggest that the expression of GLUT 3 in rodents is restricted to non-insulin responsive neuronal cells and hence it is likely that the factors regulating the expression of this transporter in rodents differ to those in humans.

The nature of the dietary component responsible for adipocytes having the ability to respond to Glucose Tolerance Factor (GTF) was investigated. Rats were raised on either a control diet or one of three diets differing only in the protein source (torula yeast, brewer's yeast, or casein). Only in adipocytes from rats fed the torula yeast diet did a GTF fraction prepared from brewer's yeast potentiate the action of suboptimal concentrations of insulin in the incorporation of label from D-[1-14C]-glucose and D-[U-14C]-glucose into CO2 and fatty acids. It was concluded that this potentiation was not the result of a deficiency of GTF activity in torula yeast, because a GTF fraction prepared from torula yeast had similar insulin potentiating activity. Differences in response among diets were not owing to differences in levels of amino acids or owing to concentrations of 22 (Al, As, B, Ca, Cd, Co, Cr, Cu, Fe, K, Mg, Mo, Na, Ni, P, Pb S, Se, Si, Sn, Sr, Zn) of the 23 trace elements investigated. The level of Mn was low in all diets, but particularly low in the torula yeast diet. Mn deficiencies have previously been implicated in perturbations of glucose metabolism, so that it is possible that this deficiency may be responsible for the effects attributed to the torula yeast diet.

A double-blind, crossover study was undertaken in general practice to compare the efficacy and tolerability of a new controlled-release indomethacin with sustained-release diclofenac sodium in patients with osteoarthritis. Eighty-four patients were randomly allocated to receive controlled-release indomethacin tablets (75 mg) or sustained-release diclofenac sodium tablets (100 mg) at night for a period of 4 weeks before being crossed-over to receive the alternative treatment for a further 4 weeks. Pain scores for day and night, duration of morning stiffness, requirement for escape analgesia, and treatment preference were similar for both treatments. There was no significant difference between treatments for incidence and severity of side-effects. It was concluded that controlled-release indomethacin tablets (75 mg) given as a single night-time dose were as efficacious and well tolerated as sustained-release diclofenac sodium (100 mg).

1. Cationic fractions were isolated from a low chromium (less than 0.2 ppm) commercial yeast extract in an attempt to purify the material responsible for glucose tolerance factor (GTF) activity observed in a standard yeast assay system. 2. Following previously described procedures a fraction with GTF activity but containing negligible chromium was isolated, which on further purification was found to be composed of many separate small basic peptides. 3. Much of the activity of the yeast GTF material in the yeast assay could be attributed to the presence of basic peptides and free amino acids acting as nitrogen sources for the yeast. 4. Additional activity was present in the yeast GTF sample, which was not due to a synergistic effect of the mixed amino acids and peptides although the component of the yeast extract responsible for this activity was not identified. 5. The results show that the GTF fractions isolated according to most previously published procedures are highly impure, and conclusions drawn about the nature of GTF based on these isolates must remain open to question. 6. The activity due to the presence of peptides and amino acids is a major cause of lack of specificity of the yeast systems as an assay for GTF.