S-15261, a new anti-hyperglycemic agent, reduces hepatic glucose production through direct and insulin-sensitizing effects.

S-15261 is a new oral anti-hyperglycemic agent that increases insulin sensitivity in various insulin-resistant animal models. The aim of this study was to determine the short- and long-term effects of S-15261 and its metabolites (S-15511 and Y-415) on fatty acid and glucose metabolism in hepatocytes isolated from 24-h starved rats. During short-term exposure (1h) neither S-15261 nor its metabolites affected fatty acid oxidation whatever the concentration used. By contrast, S-15261 and its two metabolites reduced the rates of glucose production from lactate/pyruvate and dihydroxyacetone. Using crossover plot analysis, it was shown that Y-415 reduced hepatic gluconeogenesis upstream the formation of dihydroxyacetone phosphate. After 48 h in culture, S-15261 and its two metabolites reduced the rates of glucose production from lactate/pyruvate secondarily to a decrease in PEPCK and Glc-6-Pase mRNA levels. A part of these effects on gene expression could be due to a drug-induced reduction in PGC-1 gene expression. When hepatocytes were cultured in the presence of a submaximal concentration of insulin (10(-9)M), S-15261, through its metabolite S-15511, enhanced insulin sensitivity both on gene expression (PEPCK, Glc-6-Pase, PGC-1) and on gluconeogenesis. Furthermore, S-15261 and S-15511 induced the expression of GK and FAS genes as the result of an increased in SREBP-1c mRNA levels. Finally, S-15511 enhanced the stimulatory effect of insulin on GK mRNA level through an additional increase in SREBP-1c gene expression. In conclusion, this work reveals that S-15261 via its metabolites reduces hepatic glucose production through direct and insulin-sensitizing effects on genes encoding regulatory proteins of hepatic glucose metabolism.

Transduction of leptin growth signals in placental cells is independent of JAK-STAT activation.

We have characterized the transduction pathways induced by leptin in the placenta, using human BeWo cells that express endogenous leptin receptors and synthesize leptin in a regulated manner. We first examined if the JAK-STAT phosphorylation cascade was functional in these cells. Phosphorylated JAK2 was primarily bound to a short 106kDa leptin receptor isoform and to a lesser extent to a 210kDa molecule. Leptin neither enhanced JAK2 phosphorylation nor activated STAT3 and STAT1 proteins indicating that JAK2 is constitutively activated and that the JAK-STAT transduction pathway is not recruited by leptin in BeWo cells. By contrast, leptin stimulated the transcription of the c-fos gene (3-fold) and cell proliferation (2-fold) as measured by DNA synthesis. Both effects were dependent on the rapid phosphorylation of p42-44 MAPK but not p38 MAPK. We conclude that a functional JAK-STAT pathway is not required for leptin to transduce proliferative signals in human placental cells. These findings extend the physiological action of leptin beyond its central effects, to the control of placental gene transcription and cell proliferation.

Transcriptional effect of hypoxia on placental leptin.

We observed recentlyl that placental leptin is markedly increased in preeclampsia. Since this disorder is associated with vascular disorders, we have tested the hypothesis that hypoxia regulates leptin expression. We show that hypoxia increased leptin mRNA and secretion in trophoblast-derived BeWo cells. This effect was mediated through leptin promoter activation. 5' deletion analysis allowed us to delineate two regions containing 1.87 kb and 1.20 kb of the promoter which conferred respectively high and low responsiveness to hypoxia. These data indicate that leptin is up-regulated by hypoxia through a transcriptional mechanism likely to involve distinct hypoxia-responsive cis-acting sequences on the promoter.

Dissociation between insulin-mediated signaling pathways and biological effects in placental cells: role of protein kinase B and MAPK phosphorylation.

Beyond the presence of insulin receptors, little is known of the mechanisms underlying the biological effects of insulin in the placenta. We show that phosphorylation of MAPK and protein kinase B were enhanced 286 +/- 23% and 393 +/- 17% upon insulin stimulation of JAr placental cells. MAPK activation was prevented by pretreatment with PD98059 but was unaffected by wortmannin. Insulin stimulation of protein kinase B phosphorylation was abolished by pretreatment with wortmannin, suggesting that it is dependent on phosphatidylinositol 3- kinase activation. Despite protein kinase B phosphorylation, GLUT4 translocation, glucose uptake, and glycogen synthesis were not stimulated by insulin. By contrast, glycogen synthesis was stimulated 20-fold in cells incubated with 11 mM glucose. Mitogenesis assessed by incorporation of [(3)H]thymidine into DNA was enhanced 1.9-fold in response to insulin. Stimulation of DNA synthesis was inhibited by pretreatment with PD98059 but was insensitive to wortmannin. These results indicate that stimulation of mitogenesis is one major biological effect of insulin in placenta cells that implicates the MAPK signaling pathway. Phosphatidylinositol 3-kinase- dependent protein kinase B activation is not sufficient to stimulate glucose transport and glycogen synthesis, highlighting the placenta as a nonclassic target of insulin for the regulation of glucose metabolism.

Prenatal leptin production: evidence that fetal adipose tissue produces leptin.

In the adult, circulating leptin is highly correlated to adipose tissue mass. Whether such a relationship exists prenatally is unknown, because the actual source of fetal leptin has not been determined. In the present study, we have assessed the placental contribution to fetal and maternal circulating leptin concentrations and determined whether fetal adipose tissue produces leptin. The rate of leptin production in dually perfused human placenta was 0.036 ng/min.g. Ninety-five percent of the leptin released was delivered into the maternal circulation, vs. only 5% on the fetal side. Leptin messenger RNA and protein were detected in adipose tissue biopsies of 20-38 week human fetuses. However, leptin concentration was twice lower in fetal (0.22 +/- 0.11 ng/mg protein, n = 6) than in adult (0.49 +/- 0.12 ng/mg protein, n = 8) adipose tissue. Umbilical leptin levels closely reflected ponderal index at birth over a wide range of birth weights (1.6--4.1 kg). In sharp contrast, maternal and placental leptin concentrations were increased in pregnancies associated with fetal growth retardation. We conclude that umbilical leptin levels are independent of placental leptin production and can be taken as a marker of fat mass in human fetuses. By contrast, placental leptin production makes a substantial contribution to maternal circulating leptin levels during pregnancy.

Comparative in vivo approaches for selective adenovirus-mediated gene delivery to the placenta.

Gene delivery to the placenta is one potential way of specifically modifying placental biological processes and fetal development. The aim of this study was to determine the most efficient and least invasive route of placental adenovirus delivery. The feasibility of adenovirus-mediated gene transfer to the rat placenta was addressed by maternal intravenous or direct intraplacental injection of adenoviral vectors expressing the glucose transporter GLUT3, a noncirculating integral membrane protein. Both routes led to transgene expression in the placenta. However, direct intraplacental delivery on day 14 of gestation yielded a higher transduction efficiency than maternal intravenous administration, and markedly reduced transgene expression in maternal liver. Most importantly, the amount of the GLUT3 transgene and the adenovirus itself in fetal tissues was only 1 to 3% of that found in the placenta. These results indicate that the nature of the transgene and the route of adenovirus administration are key parameters in selective placental somatic gene transfer. This novel strategy may prove useful for modifying a placental function without altering the fetal genome.

Unexpected expression of glucose transporter 4 in villous stromal cells of human placenta.

Glucose transporter 4 (GLUT4) protein expression was characterized in human and rodent term placentas. A 50-kDa protein was detected, by immunoblotting, in term human placenta at levels averaging 25% of those found in white adipose tissue. It was also present, albeit at lower levels, in mouse and rat placentas. The specificity of the 50-kDa signal was established by using skeletal muscle and placental tissues obtained from GLUT4-null mice as controls. Indirect immunohistochemistry, performed in human placentas, showed that intravillous stromal cells were conspicuously labeled by GLUT4 and revealed colocalization of GLUT4 transporters with insulin receptors. This study provides the first evidence that the insulin-responsive GLUT4 glucose transporter is present in human and rodent hemochorial placentas. Placental GLUT4 gene and protein levels were not modified in human pregnancy complicated by insulin-dependent diabetes mellitus. The significance of the high level of GLUT4 protein in human placenta remains to be elucidated, because, so far, this organ was not considered to be insulin-sensitive, with regard to glucose transport.

Hexokinase isoenzymes in the rat placenta.

The phosphorylation of glucose to glucose-6-phosphate, the first enzymatic step for glucose utilization is catalysed by a family of four hexokinase isoenzymes (HKI-IV) which display a tissue-specific distribution. The expression of HK isoenzymes was investigated in the rat placenta. High levels of HKI and HKII mRNA were found in the junctional and the labyrinthine zones. HKIII mRNA was present at low levels in the junctional zone and glucokinase (HKIV) mRNA was not detected, indicating that HKI and HKII are the two major placental HK isoenzymes. HKII activity was increased in placenta of insulinopenic diabetic rats. This regulation is likely to support the increase in glucose utilization and storage characteristics of the enlarged placentae of diabetic rats.

Overexpression of placental leptin in diabetic pregnancy: a critical role for insulin.

Leptin, a small peptide produced by adipocytes, is implicated in an increasing number of endocrine regulations, including adiposity, satiety, puberty, and fertility. Although the factors involved in controlling maternal and fetal weight gain during pregnancy have not been fully elucidated, leptin has recently emerged as such a potential factor. In our study, we report the presence of high amounts of leptin mRNA and immunoreactive protein in the human placenta, establishing the placental synthesis of this hormone. A large (three- to fivefold) augmentation in leptin mRNA and protein was found in placentas from insulin-treated diabetic women. This finding was associated with increased concentrations of leptin and insulin in venous cord blood without modification of maternal circulating leptin levels. These data provide evidence that the placenta is a site for regulated leptin production in utero. Insulin is likely to play a critical role in this regulation, thus emphasizing the importance of placental leptin signaling in diabetic pregnancy.

The GLUT3 glucose transporter isoform is differentially expressed within human placental cell types.

The cellular localization of GLUT3 messenger ribonucleic acid (mRNA) and protein was examined in human term placenta using a combination of methodologies. In situ hybridization indicated that GLUT3 mRNA was present in the trophoblast cell layer and in vascular endothelium with a heterogeneous distribution pattern. GLUT3 protein migrating at an apparent molecular mass of 49 kDa was detected by immunoblotting in membranes from whole placenta and endothelial cells derived from intraplacental microvessels, but not in isolated trophoblast cells. This cell-specific pattern of expression was confirmed by immunocytochemical studies showing a prominent localization of GLUT3 protein in vascular endothelium. These findings indicate a differential distribution of GLUT3 mRNA and protein in the human placenta. Based on its cell-specific distribution at the fetal interface, GLUT3 protein could be of cardinal importance in the transport of glucose from the placenta to the fetal circulation.

Rheogenic Cl- conductance and Cl(-)-Cl(-)-exchange activities in guinea pig jejunal basolateral membrane vesicles.

We describe a method for the simultaneous purification of apical (brush-border membrane) and basolateral membrane (BLM) vesicles from the same sample of guinea pig jejunum. We applied functional tests to demonstrate the absence of reciprocal cross contamination between the two vesicle preparations. By using the BLM vesicles and a rapid filtration technique, we quantified 36Cl uptake under conditions of equilibrated pH (pHout = pHin = 7.5). The presence of 200 mM cis of either Na+ or K+, or an equimolar mixture of both, significantly increased the initial Cl- entry rate. In the presence of K+, valinomycin further increased Cl- uptake, but no Cl- uphill transport was ever observed under any of the conditions. All the increases were abolished by voltage clamping, indicating that the alkali-metal ions act by creating an inside-positive membrane potential capable of stimulating a Cl(-)-conductance pathway. In the absence of K+, BLM vesicle preloading to obtain a [Cl-]out/[Cl-]in = 16/200 mM gradient (delta Cl-) resulted in a 500% increase in the initial 36Cl- entry rate, accompanied by a transient Cl- accumulation, with an overshoot at approximately 5 min. In the presence of both a positive-inside electrical gradient (delta psi) and a delta Cl-, the initial Cl- uptake rate was increased by 800%, indicating that the effects of delta psi and of delta Cl- are additive. The delta Cl- effect was blocked, but only partially, by short-circuiting the membrane potential with equilibrated K+ and valinomycin, thus indicating that it has both rheogenic and electroneutral components. We conclude that Cl- influx across the guinea pig intestinal BLM involves a Cl(-)-conductance pathway plus a distinct Cl(-)-Cl(-)-exchange system, exhibiting both electroneutral and rheogenic components. Alternatively, the possibility can also be entertained that the conductance and the exchange pathways share a common molecular basis, e.g., a nonobligatory Cl(-)-Cl- exchanger or rheogenic uniport.

Trans-potassium effects on the chloride/proton symporter activity of guinea-pig ileal brush-border membrane vesicles.

To investigate the inhibitory effect of trans potassium on the Cl-/H+ symporter activity of brush-border membrane vesicles from guinea pig ileum, we measured both 36Cl uptake and, by the pyranine fluorescence method, proton fluxes, in the presence of appropriate H+ and K+ gradients. In the absence of valinomycin, a time-dependent inhibitory effect of chloride uptake by trans K+ was demonstrated. This inhibition was independent of the presence or absence of any K+ gradient. Electrical effects cannot be involved to explain these inhibitions because the intrinsic permeability of these vesicles to Cl- and K+ is negligibly small. Rather, our results show that, in the absence of valinomycin, the inhibitory effect of intravesicular K+ involves an acceleration of the rate of dissipation of the proton gradient through an electroneutral exchange of trans K+ for cis H+, catalyzed by the K+/H+ antiporter also present in these membranes. Valinomycin can further accelerate the rate of pH gradient dissipation by facilitating an electrically-coupled exchange between K+ and H+. To evaluate the apparent rate of pH-dissipating, downhill proton influx, we measured chloride uptake by vesicles preincubated in the presence of alkaline-inside pH gradients (pHout/pHin = 5.0/7.5), charged or not with K+. In the absence of intravesicular K+, proton influx exhibited monoexponential kinetics with a time constant k = 11 s-1. Presence of 100 mM K+ within the vesicles significantly increased the rate of pH gradient dissipation which, furthermore, became bi-exponential and revealed the appearance of an additional, faster proton influx component with k = 71 s-1. This new component we interpret as representing the sum of the electroneutral and the electrically-coupled exchange of trans K+ for cis H+, mentioned above. Finally, by using the pH-sensitive fluorophore, pyranine, we demonstrate that, independent of the absence or presence of a pH gradient, either vesicle acidification or alkalinisation can be generated by adding, respectively, Cl- or K+ to the extravesicular medium. Such results confirm the independent existence of both Cl-/H+ symporter and K+/H+ antiporter activities in our vesicle preparations, the relative activity of the former being larger under the conditions of the present experiments. The possible interplay of these two proton-transfer mechanisms in the regulation of the intracellular pH is discussed.

Chloride transport in control and cystic fibrosis human skin fibroblast membrane vesicles.

Plasma membrane vesicles were isolated from either cystic fibrosis (CF) or non-CF cultured fibroblasts derived from skin biopsies of either foetus, child or adolescent human donors. The total membrane yield was essentially identical for either CF or control membranes. By using a rapid filtration technique, 36Cl uptake by these vesicles was quantitated in the absence and presence of alkali-metal ion-, electrical- and/or pH gradients. In the absence of a pH gradient (pHout = pHin = 7.5), Cl uptake took place downhill in both cases. Either cis K+, cis Na+ or an equimolar mixture of cis Na+ plus K+ caused Cl uptake activation. In the presence of an alkaline-inside pH gradient (pHout/pHin = 5.5/7.5), Cl uptake exhibited an apparent overshoot independently of the presence or absence of any metal-ion gradient. The observed potassium-, sodium- and proton-dependent Cl influx rates were all unaffected by voltage clamping, indicating the existence in these vesicles of electroneutral symport systems of the type Cl-/H+, Cl-/K+ and/or Cl-/Na+; but not 2 Cl-/Na+/K+. In the presence of an inward-directed K+ gradient, valinomycin further increased Cl uptake, both in the presence and in the absence of a pH gradient, indicating the presence of a rheogenic Cl uniport. In absolute quantitative terms, the two different modes (rheogenic and electroneutral) of Cl transport evinced in these vesicles were about 45% lower in CF than in control skin fibroblasts. However, qualitatively, there was no difference between normal and CF cells. The evidence obtained indicates that the CF defect, which is expressed in fibroblast plasma membranes, does not affect specifically either the rheogenic or the electroneutral Cl transport systems. Rather, the CF cells appear to give a smaller yield of closed, functional vesicles, reflected by a significantly smaller apparent intravesicular volume. Because it also affects the transport of D-glucose and L-alanine, this anomaly could be the consequence of a generalized membrane defect characterizing CF fibroblasts.

pH-dependent inhibitory effects of tris and lithium ion on intestinal brush-border sucrase.

Tris and two of its hydroxylated amine analogs were examined in a metal-free, universal n-butylamine buffer, for their interaction with intestinal brush border sucrase. Our recent three-proton-families model (Vasseur, van Melle, Frangne and Alvarado (1988) Biochem. J., 251, 667-675) has provided the sucrase pK values necessary to interpret the present work. At pH 5.2, 2-amino-2-methyl-l-propanol (PM) causes activation whereas Tris has a concentration-dependent biphasic effect, first causing activation, then fully competitive inhibition. The amine species causing activation is the protonated, cationic form. The difference between the two amines is related to the fact that Tris has a much lower pKa value than PM (respectively, 8.2 and 9.8). Even at pH 5.2, Tris (but not PM) exists as a significant proportion of the free base which, by inhibiting the enzyme fully competitively, overshadows the activating effect of the cationic, protonated amine. Above pH 6.8, both Tris and PM act as fully competitive inhibitors. These inhibitions increase monotonically between pH 6.5 and 8.0 but, above pH 8, inhibition by 2.5 mM Tris tends to diminish whereas inhibition by 40 mM PM increases abruptly to be essentially complete at pH 9.3 and above. As pH increases from 7.6 to 9.0, the apparent affinity of the free amine bases decreases whereas that of the cationic, protonated amines, increases. In this way, the protonated amines replace their corresponding free bases as the most potent inhibitors at high pH. The pH-dependent inhibition by 300 mM Li+ is essentially complete at pH 8, independent of the presence or absence of either 2.5 mM Tris or 40 mM PM. Even at pH 7.6, an excess (300 mM) of Li+ causes significant increases in the apparent Ki value of each Tris, PD (2-amino-2-methyl-1-3-propanediol) and PM, suggesting the possibility of a relation between the effects of Li+ and those of the hydroxylated amines which in fact are mutually exclusive inhibitors. The inhibitory results are interpreted in terms of a mechanistic model in which the free bases bind at two distinct sites in the enzyme's active center. Binding at the glucosyl sub-site occurs through the amine's free hydroxyl groups. This positioning facilitates the interaction between the lone electron pair of the deprotonated amino group with a proton donor in the enzyme's active center, characterized by a pK0 around 8.1. When this same group deprotonates, then the protonated amines acting as proton donors replace the free bases as the species giving fully competitive inhibition of sucrase.

Electroneutral, HCO3(-)-independent, pH gradient-dependent uphill transport of Cl- by ileal brush-border membrane vesicles. Possible role in the pathogenesis of chloridorrhea.

By applying a rapid filtration technique to isolated brush border membrane vesicles from guinea pig ileum, 36Cl uptake was quantified in the presence and absence of electrical, pH and alkali-metal ion gradients. A mixture of 20 mM-Hepes and 40 mM-citric acid, adjusted to the desired pH with Tris base, was found to be the most suitable buffer. Malate and Mes could be used to replace the citrate, but succinate, acetate and maleate proved to be unsuitable. In the absence of a pH gradient (pHout:pHin = 7.5:7.5), Cl- uptake increased slightly when an inside-positive membrane potential was applied, but uphill transport was never observed. A pH gradient (pHout:pHin = 5.0:7.5) induced both a 400% increase in the initial Cl- influx rate and a long-lasting (20 to 300 s) overshoot, indicating that a proton gradient can furnish the driving force for uphill Cl- transport. Under pH gradient conditions, initial Cl- entry rates had the following characteristics. (1) They were unaffected by cis-Na+ and/or -K+, indicating the absence of Cl-/K+, Cl-/Na+ or Cl-/K+/Na+ symport activity. (2) Inhibition by 20-100 mM-trans-Na+ and/or -K+ occurred, independent of the existence of an ion gradient. (3) Cl- entry was practically unaffected by short-circuiting the membrane potential with equilibrated potassium and valinomycin. (4) Carbonyl cyanide m-chlorophenylhydrazone was strongly inhibitory and so, to a lesser extent, was 4-acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid [(SITS)], independent of the sign and size of the membrane potential. (5) Cl- entry was negligibly increased (less than 30%) by either trans-Cl- or -HCO3-, indicating the absence of an obligatory Cl-/anion antiport activity. In contrast, the height of the overshoot at 60 s was increased by trans-Cl-, indicating time-dependent inhibition of 36Cl efflux. That competitive inhibition of 36Cl fluxes by anions is involved here is supported by initial influx rate experiments demonstrating: (1) the saturability of Cl- influx, which was found to exhibit Michaelis-Menten kinetics; and (2) competitive inhibition of influx by cis-Cl- and -Br-. Quantitatively, the conclusion is warranted that over 85% of the total initial Cl- uptake energized by a pH gradient involves an electroneutral Cl-/H+ symporter or its physicochemical equivalent, a Cl-/OH- antiporter, exhibiting little Cl- uniport and either Cl-/Cl- or Cl-/HCO3- antiport activities.(ABSTRACT TRUNCATED AT 400 WORDS)