Plasma acylation stimulating protein concentration and subcutaneous adipose tissue C3 mRNA expression in nondiabetic and type 2 diabetic men.

We studied the effect of an oral fat load on plasma acylation stimulating protein (ASP) concentrations in 9 lean healthy (age 59+/-2 years, body mass index [BMI] 23.2+/-0.4 kg/m(2); both mean+/-SEM), 9 obese nondiabetic (58+/-2 years, BMI 29.4+/-0.5 kg/m(2)), and 12 type 2 diabetic (60+/-2 years, BMI 29.6+/-1.0 kg/m(2)) men. Because ASP is a cleavage product of complement protein C3 (C3adesArg) and its secretion is regulated by insulin, we also examined the subcutaneous adipose tissue expression of C3 mRNA before and after a 240-minute euglycemic hyperinsulinemic clamp in a subgroup of these men. Plasma ASP concentration and adipose tissue C3 mRNA expression were higher in the obese groups than in the lean men. Plasma ASP concentration did not change significantly after the fat load. Fasting plasma ASP concentration and C3 mRNA expression were correlated negatively with insulin sensitivity and positively with the magnitude of postprandial lipemia in nondiabetic but not in type 2 diabetic men. The expression of C3 mRNA was not regulated by insulin. These data suggest that ASP is associated with whole-body glucose and lipid metabolism in nondiabetic individuals, whereas metabolic disturbances in diabetes may overcome the regulatory role of ASP in lipid and glucose metabolism.

Subcutaneous adipose tissue expression of plasminogen activator inhibitor-1 (PAI-1) in nondiabetic and Type 2 diabetic subjects.

OBJECTIVE: Increased plasma levels of plasminogen activator inhibitor-1 (PAI-1) have been suggested to be a part of the insulin resistance syndrome, and recent data suggest that adipose tissue participates in the production of PAI-1. We examined the expression and insulin regulation of subcutaneous adipose tissue PAI-1 mRNA and its relationship to insulin sensitivity. DESIGN: A cross-sectional study involving five lean (60.0+/-3.1 years, BMI 23.5+/-0.5 kg/m(2)) and six obese nondiabetic men (56.0+/-3.1 years, BMI 30.4+/-0.7 kg/m(2)), and six obese Type 2 diabetic men (61.4+/-3.2 years, BMI 31.8+/-1.0 kg/m(2)). MEASUREMENTS: Subcutaneous adipose tissue PAI-1 mRNA and insulin sensitivity were quantified using RT-competitive PCR and euglycemic hyperinsulinemic clamp technique, respectively. RESULTS: Subcutaneous adipose tissue PAI-1 mRNA levels were higher in obese nondiabetic and Type 2 diabetic men than in lean nondiabetic men. PAI-1 mRNA levels decreased in the three groups during a 240-min euglycemic hyperinsulinemic clamp (P<0.05 for all groups), and a similar reduction was observed during a 240-min saline control study indicating that adipose tissue PAI-1 gene expression has diurnal variation and is not acutely controlled by hyperinsulinemia. The basal PAI-1 mRNA levels correlated positively with BMI, and waist-to-hip ratio; and negatively with whole-body glucose disposal rate in nondiabetic men. CONCLUSIONS: Subcutaneous adipose tissue PAI-1 mRNA expression is increased in obese nondiabetic or in Type 2 diabetic men. Subcutaneous adipose tissue PAI-1 mRNA expression is increased in proportion to visceral obesity and to the level of whole-body insulin resistance. Subcutaneous adipose tissue PAI-1 mRNA expression is not acutely regulated by insulin, and it is subject to a diurnal variation.

Subcutaneous adipose tissue expression of plasminogen activator inhibitor-1 gene during very low calorie diet in obese subjects.

OBJECTIVE: To determine whether changes in subcutaneous adipose tissue plasminogen activator inhibitor-1 (PAI-1) expression influence plasma PAI-1 level during weight loss in obese humans. DESIGN: Study of the variations of PAI-1 levels both in plasma and in subcutaneous abdominal adipose tissue in 15 volunteer non-diabetic obese subjects, body mass index (BMI) 40.4.+/-1.9 kg/m2, aged 48+/-3 y, before and after a 3 week very low calorie diet (VLCD) programme (3.9+/-0.1 MJ/day). MEASUREMENTS: Plasma and adipose tissue PAI-1 protein levels were measured by enzyme-linked immunosorbent assay and PAI-1 mRNA levels were quantified by quantitative RT-competitive PCR. RESULTS: VLCD induced weight loss (5.8+/-0.8 kg) and decreased plasma PAI-1 concentration (-26% (P<0. 01)). Surprisingly, PAI-1 mRNA and protein abundance in subcutaneous adipose tissue increased by 87% (P<0.05) and by 44% (P<0.01), respectively. CONCLUSION: These data indicate thus that changes in subcutaneous adipose tissue PAI-1 expression are not involved in the decrease of plasma PAI-1 levels during VLCD in obese subjects. International Journal of Obesity (2000)24, 70-74

Peroxisome proliferator activated receptor-gamma, leptin and tumor necrosis factor-alpha mRNA expression during very low calorie diet in subcutaneous adipose tissue in obese women.

BACKGROUND: PPAR gamma, leptin and TNF alpha are three major factors that play a key role in influencing adipocyte differentiation and both adipose tissue function and metabolism. However, the regulation of these three genes during a dynamic period of weight loss is unknown. We therefore investigated the concomitant regulation of the mRNA expression of PPAR gamma, leptin and TNF alpha in adipose tissue during a 21-day very low calorie diet (VLCD) in 12 non-diabetic obese women. METHODS: The mRNA levels of PPAR gamma, leptin and TNF alpha were quantified by quantitative RT-competitive PCR in abdominal subcutaneous adipose tissue before and during VLCD (940 kcal/day). RESULTS: VLCD induced weight loss (approximately 6 kg) and improved insulin sensitivity. Simultaneously, VLCD induced the reduction in the adipose tissue mRNA abundances of PPAR gamma (-13%, p < 0.05) and of leptin (-58%, p < 0.005), whereas TNF alpha mRNA levels increased (+78%, p < 0.005). PPAR gamma and leptin mRNA levels were correlated before (r = 0.778, p < 0.01) and after VLCD (r = 0.797, p < 0.01). Serum HDL-cholesterol concentrations were positively associated with PPAR gamma (r = 0.696, p < 0.03) and leptin (r = 0.806, p < 0.01) mRNA levels. CONCLUSIONS: The increase in TNF alpha mRNA levels suggested that a local increased expression of this cytokine in adipose tissue might play a role in the control of the fat mass during weight loss. PPAR gamma and leptin mRNA levels were positively associated both before and after VLCD, suggesting that common regulatory mechanism(s) might control their expression. More strikingly, we found strong positive correlations between circulating HDL-cholesterol and both PPAR gamma and leptin mRNA levels, suggesting the existence of physiological links between circulating lipoprotein metabolism and adipose tissue function.

Insulin acutely regulates the expression of the peroxisome proliferator-activated receptor-gamma in human adipocytes.

Peroxisome proliferator-activated receptor (PPAR)-gamma is one of the key actors of adipocyte differentiation. This study demonstrates 1) that PPAR-gamma mRNA expression is not altered in subcutaneous adipose tissue (n = 44) or in skeletal muscle (n = 19) of subjects spanning a wide range of BMIs (20-53 kg/m2) and 2) that insulin acutely increases PPAR-gamma mRNA expression in human adipocytes both in vivo and in vitro. The effect of insulin was investigated in abdominal subcutaneous biopsies obtained before and at the end of a 3-h euglycemic-hyperinsulinemic clamp. Insulin significantly increased PPAR-gamma mRNA levels in lean subjects (88 +/- 17%, n = 6), in type 2 diabetic patients (100 +/- 19%, n = 6), and in nondiabetic obese patients (91 +/- 20%, n = 6). Both PPAR-gamma1 and PPAR-gamma2 mRNA variants were increased (P < 0.05) after insulin infusion. In isolated human adipocytes, insulin induced the two PPAR-gamma mRNAs in a dose-dependent manner, with half-maximal stimulation at a concentration of approximately 1-5 nmol/l. However, PPAR-gamma2 mRNA was rapidly (2 h) and transiently increased, whereas a slow and more progressive induction of PPAR-gamma1 was observed during the 6 h of incubation. In explants of human adipose tissue, PPAR-gamma protein levels were significantly increased (42 +/- 3%, P < 0.05) after 12 h of incubation with insulin. These data demonstrate that PPAR-gamma belongs to the list of the insulin-regulated genes and that obesity and type 2 diabetes are not associated with alteration in the expression of this nuclear receptor in adipose tissue.

Defective regulation of phosphatidylinositol-3-kinase gene expression in skeletal muscle and adipose tissue of non-insulin-dependent diabetes mellitus patients.

We investigated the regulation of the mRNA expression of the insulin receptor, insulin receptor substrate-1 (IRS-1) and p85alpha-phosphatidylinositol-3-kinase (PI-3K), three major actors of insulin action, in skeletal muscle from 10 healthy lean volunteers, 13 obese patients with Type II (non-insulin-dependent) diabetes mellitus and 7 non-diabetic obese subjects. The in vivo regulation by insulin was studied using a 3-h euglycaemic, hyperinsulinaemic clamp. There were no differences in the basal concentrations of the three mRNAs in skeletal muscle between groups. Insulin infusion produced a twofold reduction in insulin receptor substrate-1 mRNA expression in the three groups (p<0.02). In contrast, insulin increased p85alpha-phosphatidylinositol-3-kinase mRNA expression in muscle from non-diabetic subjects (+98+/-22% in lean and +127+/-16% in obese, p<0.02) but this effect was totally impaired in Type II diabetic patients (+5+/-12%, NS). A similar defect in insulin action on p85alpha-phosphatidylinositol-3-kinase mRNA expression was observed in abdominal subcutaneous adipose tissue (+138+/-25%, p<0.01 in lean and +46+/-14%, p<0.02 in obese and +29+/-11%, NS in Type II diabetic patients). The lack of action of insulin on p85alpha-phosphatidylinositol-3-kinase mRNA in diabetic subjects was probably not due to a deleterious effect of hyperglycaemia since improvement of the glycaemic control for 10 days did not restore the response in muscle or in adipose tissue. This study provides evidence for a defect in the regulation by insulin of PI-3K gene expression in Type II diabetic patients, thus reinforcing the concept that alterations at the gene expression might be involved in the pathogeny of Type II diabetes.

Tissue distribution and quantification of the expression of mRNAs of peroxisome proliferator-activated receptors and liver X receptor-alpha in humans: no alteration in adipose tissue of obese and NIDDM patients.

Members of the peroxisome proliferator-activated receptor (PPAR) family might be involved in pathologies with altered lipid metabolism. They participate in the control of the expression of genes involved in lipid metabolism and adipocyte differentiation. In addition, thiazolidinediones improve insulin resistance in vivo by activating PPAR gamma. However, little is known regarding their tissue distribution and relative expression in humans. Using a quantitative and sensitive reverse transcription (RT)-competitive polymerase chain reaction (PCR) assay, we determined the distribution and relative mRNA expression of the four PPARs (alpha,beta, gamma1, and gamma2) and liver X receptor-alpha (LXR alpha) in the main tissues implicated in lipid metabolism. PPAR alpha and LXR alpha were mainly expressed in liver, while PPAR gamma1 predominated in adipose tissue and large intestine. We found that PPAR gamma2 mRNA was a minor isoform, even in adipose tissue, thus causing question of its role in humans. PPAR beta mRNA was present in all the tissues tested at low levels. In addition, PPAR gamma mRNA was barely detectable in skeletal muscle, suggesting that improvement of insulin resistance with thiazolidinediones may not result from a direct effect of these agents on PPAR gamma in muscle. Obesity and NIDDM were not associated with change in PPARs and LXR alpha expression in adipose tissue. The mRNA levels of PPAR gamma1, the predominant form in adipocytes, did not correlate with BMI, leptin mRNA levels, or fasting insulinemia in 29 subjects with various degrees of obesity. These results indicated that obesity is not associated with alteration in PPAR gene expression in abdominal subcutaneous adipose tissue in humans.

Identification of profilin as an IgE-binding component in latex from Hevea brasiliensis: clinical implications.

Considering the high occurrence of profilin as an allergen in many plant species, the assumption was made that profilin might be an allergen in Hevea brasiliensis, a member of the latex producing Euphorbiaceae family. Using IgE-binding inhibition by purified profilins we demonstrated that profilin is an IgE-binding component in the cytosolic fraction of natural latex and, to a lower extent, in the rubber fraction. Thirty-five out of 36 sera containing IgE to ragweed-profilin reacted with profilin from latex, indicating structural homologies between profilins from latex and ragweed. A large percentage (59%) of these sera were found to be positive in CAP latex assay. The preincubation of these sera with purified ragweed profilin greatly inhibited the CAP latex. Because profilin is also present in banana extract, it is likely to be involved in cross-sensitivity to banana and latex. In a group of 19 individuals allergic to latex only two had anti-profilin IgE antibodies. Profilin was barely detectable on glove extract immunoblots, whereas some sera from patients allergic to latex reacted with a 15 kDa allergen which was not profilin. Consequently, IgE antibodies to latex-profilin is a questionable factor for sensitization of occupationally-exposed patients; however, sensitization to profilin should be taken into account when interpreting the results of latex IgE antibody assays.

Purification and characterization of an allergen from celery immunochemically related to an allergen present in several other plant species. Identification as a profilin.

The purification of a 15 kD allergen from celery was obtained by a four step procedure. Evidence for at least two isoallergenic forms was obtained after analysis by two-dimensional-electrophoresis. A rabbit polyclonal antibody raised against this purified allergen allowed us to confirm our precedent results on the occurrence of allergenically and molecular mass-related components in celery and birch and mugwort pollens. In addition such components were also present in numerous other species like Cynodon dactylon, Sorghum halopense, Poa pratensis, Ambrosia elatior and in apple and carrot. The 15 kD allergen was identified as profilin by use of a specific rabbit polyclonal antibody that recognized a recombinant birch profilin. In addition, the purified celery allergen binds IgE from sera of patients allergic to birch profilin. These results reinforce the concept of profilin as a panallergen responsible in some patients for cross-allergic manifestations to various and unrelated species of grasses, weeds, trees, vegetables and fruits.

A study of allergens in celery with cross-sensitivity to mugwort and birch pollens.

Sixty-one sera with positive RAST to mugwort pollen (Artemisiae vulgaris) were submitted to RASTs for birch pollen (Betula verrucosa) and celery (Apium graveolens). In 36 cases RAST results were positive for celery. In addition, 23 sera presented specific IgE to birch pollen. The binding of specific IgE to individual allergens in celery, mugwort pollen and birch pollen was studied by the immunoblotting technique. This involved electrophoretic separation of allergenic extracts, electrotransfer of proteins onto nitrocellulose sheets and sensitive immunoenzymatic detection. Eighteen sera had specific IgE binding to two celery components of molecular weight around 15 kD. All these sera also detected a 15 kD allergen in mugwort and two allergens in birch of 14 kD and 16 kD molecular weight. The sera that did not detect the 15 kD bands in celery failed to react with both the 15 kD mugwort component and the 14 and 16 kD birch components. Specific cross-inhibitions of the detection of these allergens on immunoblots were obtained by pre-incubation of the sera with crude extract of the three species. These results strongly suggest that such allergens display some structural identity and that they could be at the origin of some cases of crossed hypersensitivity to celery, mugwort pollen and birch pollen.

Growth and differentiation of human keratinocytes on extracellular matrix.

Extracellular matrices (ECM) have been reported to enhance epithelial cell attachment and proliferation as well as to induce differentiation in vitro. Since ECM components are physiological constituents of the dermoepidermal basement membrane, we studied the growth and differentiation of human keratinocytes on ECM in order to determine the benefits of culturing epidermal epithelial cells (keratinocytes) on reconstituted basement membranes. Disaggregated epidermal cells were grown in primary and subcultures in liquid medium; the attachment of the cells was greatly enhanced by ECM and noted within the first few hours after seeding; cells formed small islets that reached confluence within 2-12 days depending upon the plating density and the type of culture (primary or passages). Histological and ultrastructural cross-sections of the cultures clearly indicated that a multilayered epithelium can be obtained including a basal cell layer, several intermediate cell layers with cytoplasmic organelles, intermediate size filaments, desmosomes, and keratohyaline granules, and an upper layer of anucleated cells. Using immunofluorescence, both pemphigus and pemphigoid (basal membrane zone) antigens were expressed. The keratin pattern noted indicated that these epithelia differentiate and keratinize but do not express a complete program of keratinization, a finding usually noted when cells are grown submersed. These data show that ECM favor epidermal cell proliferation and differentiation and suggest that they may be used to obtain large amounts of epidermal equivalent suitable for grafting and/or in vitro studies.

Corticotropin and angiotensin induce dephosphorylation of a Mr-20,000 protein in bovine adrenal cells.

ACTH (1-24) and Angiotensin II, both able to activate steroidogenesis in bovine fasciculata-reticularis cells, each reduced the [32P] incorporation in a cytosolic Mr-20,000 pI 6.8 protein in this cell. Cells preincubated with Sar1-Angiotensin prevented the effect of Angiotensin. Angiotensin 10(-8)M and ACTH 10(-10)M led to an almost complete disappearance of the corresponding radioactive spot on the autoradiograph. The effect was observed as soon as 2 minutes after addition of hormones to the cells. Other activators of steroidogenesis such as 8-bromocyclicAMP (8-BrcAMP), 4 beta-Phorbol-12 beta-Myristate-13 alpha-acetate (PMA) and [9-tryptophan (o-nitrophenylsulfenyl)] substituted ACTH (NPS-ACTH), also reduced the labeling of the Mr-20,000 polypeptide. On the other hand, this effect was not reproduced by insulin or human growth hormone (hGH). On 2-D gels from control, the coincidence of this polypeptide with phosphorylated myosin light chain was not observed. We suggest that the apparent dephosphorylation of this polypeptide may represent a common effect of all steroidogenic agents regardless of their seemingly distinct early actions.

In vitro interactions between Sertoli cells and steroidogenic cells.

Both the cell and the species specificities of the steroidogenic potentiating activity (SPA) of Sertoli cells on Leydig cells were studied using a coculture system. Coculture of purified pig Leydig cells with rat or pig Sertoli cells in the presence of FSH led in both cases, to a significant increase in hCG receptor number and in hCG-stimulated testosterone production. Similarly, coculture of bovine adrenal cells with rat or pig Sertoli cells enhanced the steroidogenic response of adrenal cells to ACTH and angiotensin II. Such effects were not observed when pig Leydig cells or bovine adrenal cells were cocultured with bovine aortic endothelial cells. Coculture of Sertoli and Leydig cells in the presence of hCG, resulted in a significant increase in FSH receptor number and in FSH-induced plasminogen activator activity. Such effects did not occur when Sertoli cells were cocultured with either adrenal or aortic endothelial cells.

Activation of microsomal cAMP-dependent protein kinase isoenzyme I by ACTH1-24 in bovine adrenal cells.

In microsomes of bovine fasciculata reticularis cells incubated with or without 10(-8) M ACTH during 20 min, we measured covalent and non covalent cAMP binding under exchange or non-exchange conditions and cAMP-kinase activity. ACTH induced a decrease in cAMP-kinase activity and in the number of free cAMP binding sites. These results indicate an activation by ACTH of a part of microsomal cAMP-dependent protein kinase. Photoaffinity labeling of microsomal protein with 8-azido-cAMP revealed the presence of both cAMP-kinase isoenzyme I and II in this cellular fraction. Using this method, it was demonstrated that ACTH1-24 caused a preferential and nearly complete activation of microsomal protein kinase I.

Distribution and characterization of cAMP-dependent protein kinase isoenzymes in bovine adrenal cells.

Subcellular localization and characterization of cAMP-kinase isoenzymes in fasciculata reticularis bovine adrenal cells has been investigated. Different subcellular fractions were purified on a Percoll gradient and characterized by marker enzymes. cAMP-kinase was located principally in cytosol and microsomes. In the low-speed particulate fractions cAMP-kinase was found associated mainly with plasma membrane but not with mitochondria. Characterization of isoenzyme patterns in subcellular fractions by conventional DEAE-cellulose chromatography and by anion-exchange HPLC gives essentially the same results. Isoenzyme I appears to be the main enzyme in cytosol whereas isoenzyme II predominates in solubilized microsome and plasma membrane enriched fraction. Photoaffinity labelling of chromatographic fractions demonstrated that HPLC separates both cAMP binding subunits. Photoaffinity labelling of the different subcellular fraction by 8-azido-[32P]cAMP confirmed the data obtained by anion-exchange chromatography. However, in microsomes this method revealed the presence of both isoenzymes and the preferential solubilization of isoenzyme II by Triton X-100. In summary, our results indicate a subcellular compartmentalization of cAMP-kinase in bovine adrenal cells with a preferential localization of isoenzyme I in cytosol and of isoenzyme II in membrane. However, the relation between the distribution and the role of each isoenzyme has so far not been documented.

ACTH-induced change of cAMP-dependent protein-kinase repartition in bovine adrenal cells.

Acute ACTH stimulation of isolated adrenal cells produced a modification of the subcellular distribution of cAMP-dependent protein kinase. Within 20 min, the protein-kinase ratio in all subcellular fractions, particularly in the 175 000 x g supernatant, was increased. Total protein-kinase activity, as well as the specific activity of both the homogenate and particulate enzymatic activities, was decreased while that of the 175 000 x g supernatant was increased. However, the increase of the soluble kinase activity represented only 29-46% of the lost particulate activity. On the other hand, under exchange conditions, the cAMP-binding capacity of all subcellular fractions was similar in control and ACTH-treated cells, except in the 175 000 x g pellet in which it was slightly decreased in ACTH-treated cells. These modifications were observed for supraphysiological (10(-8) M), as well as for physiological (10 (-11) M), concentrations of ACTH. These observations suggest that, after ACTH stimulation, a liberation of free catalytic sub-unit occurs from particulate into the soluble cell compartment, which shows an activation of particulate cAMP-dependent protein-kinase activity.

Modification of lymphocyte DNA synthesis by alpha 1-antitrypsin.

Purified human alpha 1-antitrypsin (alpha 1-AT) was shown to inhibit 3H-thymidine incorporation into mouse or human lymphocytes stimulated by various mitogens or by allogeneic cells. In the mouse, both B- and T-cell responses were affected. In the human, proliferative responses of peripheral blood lymphocytes, thymocytes and T-enriched tonsillar lymphocytes to phytohaemagglutinin were inhibited as well as that of tonsillar lymphocytes to Salmonella typhi-murium lipopolysaccharide. Spontaneous 3H-thymidine incorporation was moderately and inconstantly decreased, without evidence of altered cell viability. The inhibitory effect of alpha 1-AT appears to be related to its protease inhibitory capacity. These data bring further evidence for the role of proteolytic enzymes in the early events of lymphocyte activation, and support the hypothesis that serum inhibitors of proteases may contribute to the modulation of the immune response.

[Intracellular cyclic AMP in vesicular stomatitis or Sendaï virus infected cells (author's transl)]. / Variations des concentrations de cAMP intracellulaire de différents types de cellules après infection par le virus Sendaï ou le virus de la stomatite vésiculeuse.

In KB cells, MRC5 and adult skin fibroblasts infected by low doses of Sendaï virus, intracellular cyclic AMP levels rose and fell in the first hours following infection, then remained lower than basal level during at least 2 days in KB cells and adult skin fibroblasts. When compared to other viruses or cAMP inducers previously described, this effect appeared specific of Sendaï virus. Mechanisms and roles of cAMP variations are discussed. VSV-infected KB cells showed slightly decreased cAMP levels during the first hours following infection.

[Behaviour of the prolyl hydroxylase to collagen synthesis in KB cells]. / Place de la prolyl hydroxylase dans la synthèse du collagène de cellules KB.

KB cells, which synthetized collagen at a low rate, shown a prolyl hydroxylase activity at the same rate that fibroblast. The relationship between collagen synthesis and prolyl hydroxylase activity in these cells was discussed.

[Adenylate cyclase activation and its effects on intracellular cAMP in infected KB cells during trypsin-induced infectious Sendaï virus production (author's transl)]. / Activation de l'adénylate cyclase et ses répercussions sur le taux intracellulaire d'AMP cyclique dans les cellules KB infectées produisant du virus Sendaï infectieux sous l'influence de la trypsine.

KB cells infected by Sendaï virus can produce infectious virus if they are trypsinated twice over 24 h. Adenylate cyclase activity in infected KB cells is higher and more strongly activated by trypsin than that of control cells, but intracellular concentration of cAMP is the same, except during a short time after trypsinations, especially after the second trypsination which causes infectious virus production. During this short time, intracellular cAMP is slightly higher in infected cells. This miseffect of adenylate cyclase activation on intracellular cAMP concentrations might be related to an increased cell permeability caused by trypsin.