Increased translatable mRNA and decreased lipogenesis are responsible for the augmented secretion of lipid-deficient apolipoprotein E by hepatocytes from fasted rats

We examined the mechanism through which fasting selectively increases the secretion of apoE while it decreases the secretion of all lipoprotein lipids (Davis, R. A., Boogaerts, J. R., Borchardt, R. A., Malone-McNeal, M., and Archambault-Schexnayder, J. (1985) J. Biol. Chem. 260, 14137-14144). Livers were obtained from rats that were fed chow plus drinking water (control) and drinking water only (fasted) for three days. Livers were extracted for both total and poly(A) RNA. Using full length, nick-translated 32P-labeled cDNA probes for both apoE and beta-actin, the relative abundance was determined by slot blot hybridization assays. There was 2-fold more apoE mRNA in the livers of fasted rats. Furthermore, translation of poly(A) RNA using a reticulocyte lysate showed a similar 2.3-fold increase in the synthesis of immunoprecipitable [35S]methionine-labeled apoE. The 2-fold increase in translatable apoE mRNA correlates with a similar increase in apoE secretion. We also characterized the form of apoE secreted by hepatocytes from fasted cells. Cells were labeled with [35S]methionine, and the medium was separated by agarose 0.5m column chromatography. The majority of the apoE secreted cells from both control and fasted rats eluted in fractions that contained no detectable lipid. Furthermore, almost all of the increased apoE secreted by fasted cells was in these lipid-deficient fractions. The isoform distribution of apoE secreted by cells from both groups consisted of six major apoE isoforms. Consistent with previous results, treatment with neuraminidase transformed the acidic forms into the three most basic, suggesting that the three most acidic isoforms contain varying amounts of sialic acid. The isoform pattern of apoE secreted by cells from fasted rats was significantly enriched in two acidic isoforms, while it was significantly decreased in the major basic isoform. Moreover, when oleic acid (1 mM) was added to the culture medium to stimulate lipogenesis, the amount of apoE secreted with lipid increased as did the more basic isoforms. These data suggest that the secretion of lipid-deficient apoE by cells from fasted rats is the result of increased mRNA and a concomitant reduction in lipogenesis. Furthermore, the parallel shift of both the amount of apoE secreted associated with lipid as well as its isoform pattern to a more basic one by oleic acid suggests that the lipid availability plays a role in determining the lipid complement and sialic acid content of apoE secreted by the hepatocyte.     

Intrahepatic assembly of very low density lipoproteins. Varied synthetic response of individual apolipoproteins to fasting

Hepatocytes obtained from rats fed for 3 days chow (control) or drinking water only (fasted) were used to examine how metabolic state affects lipogenesis, apolipoprotein synthesis, and the capacity to secrete de novo synthesized triacylglycerol. The secretion of triacylglycerol (mass and 3H-labeled via 3H2O incorporation) by both groups of cells was constant for 30 h. Moreover, cells from fasted rats secreted triacylglycerol at rates which were markedly reduced (mass -84%; 3H-labeled -91%). To assess the relative capacities of the two groups of hepatocytes to augment triacylglycerol secretion in response to stimulated lipogenesis, cells were incubated with increasing concentrations of glucose. Control cells responded to glucose by increasing equally the synthesis and secretion of [3H] triacylglycerol. When cells from fasted rats were challenged with glucose, triacylglycerol secretion was not increased. Rather, it accumulated intracellularly. Double-reciprocal plot analysis of the capacity to augment triacylglycerol secretion in response to glucose showed that cells from fasted rats had a greater than 10-fold decrease in V'max. Moreover, fasting changed the synthesis and secretion of apolipoproteins selectively: secretion of low molecular weight apo-B was decreased 50%, large molecular weight apo-B was unchanged, and apo-E was increased 2-4-fold. Analysis of the lipoproteins from both groups of cells on Bio-Gel A-50m showed that the very low density lipoprotein secreted by cells from fasted rats was smaller. In addition, all of the increased de novo synthesized apo-E secreted by cells from fasted rats eluted after the triacylglycerol-rich lipoproteins. The combined data show that: 1) the synthesis of individual very low density lipoprotein apolipoproteins is independently regulated, and 2) the synthesis (availability) of apo-B determines the capacity of the hepatocyte to assemble/secrete triacylglycerol-rich very low density lipoprotein.     

Secretion of lipoprotein lipid and synthesis of fatty acids, cholesterol and triacylglycerol by hepatocytes of fasted-refed rats

Synthetic rates of fatty acid, cholesterol and triacylglycerols, and contents and secretion of lipoprotein lipids, were determined in hepatocytes of rats fed ad libitum a fat-containing stock diet or of rats fasted for 48 h and then refed for 24 or 48 h with stock diet or with a glucose-rich fat-free diet. When compared with the values for the ad libitum-fed rats, fatty acid synthesis was lower in fasted rats, slightly increased in rats refed with the stock diet, but several-fold elevated after refeeding the glucose-rich fat-free diet. Cholesterol synthesis was decreased in the fasted cells, and restored to the control level upon refeeding either diet. Triacylglycerol synthesis from exogenous oleate was greatly stimulated in the cells of fasted-refed rats above the rate in cells of the ad libitum-fed rats, the increase being considerably higher after refeeding the glucose-rich fat-free diet than the stock diet. The amount of triacylglycerol secreted by the cells was also elevated by the fasting-refeeding treatment, but the difference between the two diets was much less pronounced than seen for the lipids' synthetic rates. This imbalance may underlie the huge accumulation of this lipid observed in the heptatocytes after refeeding the rats for 48 h with the glucose-rich fat-free diet.     

Fatty acids alter time dependent loss of apolipoprotein E expression by primary cultures of rat hepatocytes

Although the phenomenon of intracellular apolipoprotein E (apoE) degradation has been reported in other cell types, the fate of newly synthesized apoE in the liver is not well understood. In the present study, we examined the expression (the balance of synthesis, secretion, and degradation) of apoE in primary cultures of rat hepatocytes and compared it with albumin, a typical secretory protein. Synthesis and secretion of [(35)S]apoE was diminished in primary hepatocytes cultured for more than 2 days, in agreement with an observed decrease in apoE mRNA. Cells cultured for 1 day and labeled for up to 4 hours secreted total protein, apoE, and albumin, linearly. The apparent rates of synthesis for apoE and albumin were similar (1,158 vs. 1,334 dpm/mg/min) but rates of their secretion differed significantly (225 vs. 1,159 dpm/mg/min). Pulse-chase experiments indicated that cell-associated [(35)S]albumin was secreted without degradation, whereas significant quantities of newly synthesized apoE were degraded. The overall synthesis and secretion of total proteins, including secretion of apoE, was enhanced by oleic acid (1 mmol/L). However, this effect may not be limited to oleic acid because other fatty acids showed a similar effect on apoE mRNA abundance. In control cells, apoE was found to associate with high density lipoproteins predominantly, although the fraction associated with very low density lipoprotein was increased in hepatocytes incubated with oleic acid. Overall, the findings from this study suggest that the level of apoE expression by primary hepatocytes is dependent on the age of the culture. The study also indicates that the phenomenon of apoE degradation occurs in primary hepatocytes.     

Expression of hepatic calcium-binding protein regucalcin mRNA is elevated by refeeding of fasted rats: Involvement of glucose,insulin and calcium as stimulating factors

The effect of refeeding on the expression of Ca²⁺-binding protein regucalcin mRNA in the liver of fasted rats was investigated. When rats were fasted overnight, the hepatic regucalcin mRNA level was reduced about 70% of that in feeding rats. Refeeding produced a remarkable elevation of hepatic regucalcin mRNA level (about 150–170% of fasted rats). Liver regucalcin concentration was appreciably increased by refeeding, although it was not altered by fasting. The oral administration of glucose (2 g/kg body weight) to fasted rats caused a significant increase in hepatic regucalcin mRNA level. Moreover, hepatic regucalcin mRNA level was clearly elevated by a single subcutaneous administration of insulin (10 and 100 U/kg) to fasted rats. The hormonal effect was not further enhanced by the simultaneous administration of calcium chloride (250 mg Ca/kg) to fasted rats, although calcium administration stimulated regucalcin mRNA expression in the liver. The present study suggests that the expression of hepatic regucalcin mRNA stimulated by refeeding is significantly involved in the action of insulin and/or calcium as stimulating factors.     

Physiological regulation of acetyl-CoA carboxylase gene expression: effects of diet, diabetes, and lactation on acetyl-CoA carboxylase mRNA

We measured acetyl-CoA carboxylase mRNA levels in various tissues of the rat under different nutritional and hormonal states using a cDNA probe. We surveyed physiological conditions which are known to alter carboxylase activity, and thus fatty acid synthesis, to determine whether changes in the levels of carboxylase mRNA are involved. The present studies include the effects of fasting and refeeding, diabetes and insulin, and lactation on carboxylase mRNA levels. Northern blot analysis of liver RNA revealed that fasting followed by refeeding animals a fat-free (high carbohydrate) diet dramatically increased the amount of carboxylase mRNA compared to the fasted condition. These changes in the level of mRNA correspond to changes in the activity and amount of acetyl-CoA carboxylase. Acetyl-CoA carboxylase mRNA levels in epididymal fat tissue decreased upon fasting and increased to virtually normal levels after 72 h of refeeding, closely resembling the liver response. The amount of acetyl-CoA carboxylase mRNA decreased markedly in epididymal fat tissue of diabetic rats as compared to nondiabetic animals. However, 6 h after injection of insulin the mRNA level returned to that of the nondiabetic animals. Gestation and lactation also affected the levels of carboxylase mRNA in both liver and mammary gland. Maximum induction in both tissues occurred 5 days postpartum. These studies suggest that these diverse physiological conditions affect fatty acid synthesis in part by altering acetyl-CoA carboxylase gene expression.     

Secretion and uptake of nascent hepatic very low density lipoprotein by perfused livers from fed and fasted rats

Livers from fed or 24-hr fasted male rats were perfused in a recycling system. VLDL labeled with [1-14C]oleate (95% in triglyceride), produced in separate perfusions of livers from fed rats, was added to the medium as a pulse. Uptake of VLDL 14C-labeled triglyceride by livers from fasted rats was less than that from fed rats regardless of addition of oleate. During the interval in which radioactive triglyceride was taken up, the mass of triglyceride in the medium increased, indicative of the synthesis and net secretion of triglycerides. The rates of secretion of VLDL and uptake of VLDL were both more rapid in livers from fed rats in comparison to those from fasted animals. It was calculated that about 50% of the triglyceride synthesized and secreted by the liver was taken back by livers from fed rats. The VLDL from livers of fasted rats did not contain any apoE detectable by SDS gel electrophoresis or by radioimmunoassay when no fatty acid or 166 mumol of oleic acid was infused. In contrast, apoE comprised 6% of the VLDL apoprotein derived from perfusion of livers from fed animals in the absence of added fatty acid, and 20% when the fed livers were infused with 166 mumol of oleic acid. However, the net output (accumulation) of apoE by fasted liver was only two-thirds that from fed livers. When lipoprotein-free rat plasma containing apoE (4 mg/dl) was used in place of bovine serum albumin, the VLDL secreted by livers from either fed or fasted rats contained apoE and was taken up to a similar extent by such livers. These data suggested that the apoE of the d greater than 1.21 g/ml fraction was transferred to newly secreted VLDL which then stimulated uptake of the VLDL by livers from fasted rats. With further stimulation of secretion of VLDL triglyceride by infusion of 332 mumol of oleic acid/hr, the percent of apoE in the VLDL secreted by livers from fasted rats increased to 20%, which was similar to that of the VLDL produced by livers from fed rats when either 166 or 332 mumol/hr was infused. These data suggest a relationship between rates of hepatic secretion of VLDL (TG) and apoE, and the association of apoE with the secreted VLDL. During fasting, reduced secretion of both VLDL and apoE resulted in a VLDL particle that was considerably diminished in content of apoE and, therefore, that would be taken up by the liver at a reduced rate, in comparison to that observed in the fed animal.     

Vitamin A status and its metabolism contribute to the regulation of hepatic genes during the cycle of fasting and refeeding in rats

Vitamin A (VA) status and its metabolism affect hepatic metabolic homeostasis. We investigated if VA status and metabolism contribute to energy metabolism and expression of hepatic genes in the cycle of fasting and refeeding. Zucker lean rats with VA sufficient (VAS) or VA deficient (VAD) status were respectively grouped as: ad libitum (VAS-AD or VAD-AD), 48-h fasted (VAS-Fasted or VAD-Fasted), 48-h fasted and refed a VAS diet (VAS-Refed-VAS or VAD-Refed-VAS), or refed a VAD diet (VAS-Refed-VAD or VAD-Refed-VAD) for 6 h. Respiratory exchange ratio (RER) of rats fed the VAS or VAD diet was monitored for 6 weeks. From week four, rats fed the VAS diet had higher RER than those fed the VAD diet. VAS-Refed rats had higher plasma levels of glucose, triglyceride, insulin and leptin than VAD-Refed rats. The mRNA and protein levels of hepatic genes for fuel metabolism in the fasting and refeeding cycle were determined using real-time polymerase chain reaction and immunoblot, respectively. The mRNA levels of glucokinase (Gck), sterol regulatory element-binding protein 1c (Srebp-1c), and fatty acid synthase (Fas) were lowered in VAS-Fasted and VAD-Fasted rats, and increased in VAS-Refed-VAS, VAS-Refed-VAD and VAD-Refed-VAS, but not VAD-Refed-VAD, rats. The ACL and FAS protein levels only dropped in VAS-Fasted rats and increased in VAS-Refed-VAS rats. The GK protein level decreased only in VAS-Fasted rats, and increased in VAS-Refed-VAS, VAS-Refed-VAD and VAD-Refed-VAS (but not VAD-Refed-VAD) rats. We conclude that VA status and its metabolism in the fasting and refeeding cycle contribute to the regulation of hepatic gene expression in rats.     

Influence of fasting on glycogen depletion in rats during exercise

We recently observed that a 24-h fasted group of rats could run longer than an ad libitum fed control group before becoming exhausted. Because of the demonstrated importance of glycogen levels and free fatty acid availability during endurance exercise, we have investigated several parameters of carbohydrate and lipid metabolism in exercised and nonexercised rats that were either fed ad libitum or fasted for 24 h. A 24-h fast depleted liver glycogen, lowered plasma glucose concentration, decreased muscle glycogen levels, and increased free fatty acid and beta-hydroxybutyrate concentrations in plasma. During exercise the fasted group had lower plasma glucose concentration, higher plasma concentration of free fatty acids and beta-hydroxybutyrate, and a lower muscle glycogen depletion rate than did the ad libitum fed group. Since fasted rats were able to continue running even when plasma glucose had dropped to levels lower than those of fed-exhausted rats, it seems unlikely that blood glucose level, per se, is a factor in causing exhaustion. These results suggest that fasting increases fatty acid utilization during exercise and the resulting "glycogen sparing" effect may result in increased endurance.     

Effect of ethanol on hepatic phosphatidate phosphohydrolase: dose-dependent enzyme induction and its abolition by adrenalectomy and pyrazole treatment

Protein synthesis, measured as the incorporation of [¹⁴C]valine into cell proteins and into proteins secreted into the medium, and albumin production were studied in isolated rat liver hepatocytes. Protein synthesis was substantially higher in cells from fed rats than in cells from fasted rats. Addition of carbohydrates or amino acids increased protein synthesis in cells from fasted rats, whereas no effect was seen in cells from fed rats. Addition of oleate had no effect on protein synthesis. Ethanol inhibited protein synthesis in cells from fasted rats, whereas no or only small effect was seen in cells from fed rats. Simultaneous addition of carbohydrates diminished the inhibitory effect of ethanol, whereas addition of oleate increased the inhibitory effect of ethanol. It is suggested that the rate of protein synthesis in cells from fasted rats could be restricted by lack of precursors for synthesis of nonessential amino acids. The effect of ethanol is explained by an inhibition of gluconeogenesis.     

Dietary protein and the control of fatty acid synthesis in rat adipose tissue

Fatty acid synthesis in adipose tissue normally proceeds at a high rate when fasted animals are refed a diet containing carbohydrate, protein, and low levels of fat. This study investigated the effect of omitting protein from the refeeding diet. Rats were fasted for 48 hr and refed either a protein-free diet or a balanced diet, and the rate of fatty acid synthesis from glucose, pyruvate, lactate, and aspartate was measured. Refeeding the animals a diet devoid of protein resulted in a low rate of fatty acid synthesis from each of these substrates as well as a reduction in carbon flow over the citrate cleavage pathway. The activities of glucose-6-phosphate dehydrogenase, 6-phosphogluconate dehydrogenase, NADP-malate dehydrogenase, and ATP-citrate lyase were also reduced in epididymal fat pads from these rats. On the other hand, adipose tissue phosphoenolpyruvate carboxykinase activity was five times as great as that in tissue from animals refed a balanced diet. This difference could be eliminated if actinomycin D was injected coincident with refeeding. Refeeding rats diets high in carbohydrate is not, therefore, capable of inducing high rates of fatty acid synthesis in adipose tissue in the absence of dietary proteins. Thus, liver and adipose tissue respond differently to dietary protein.     

The effects of fasting and refeeding on serum parathormone and calcitonin concentrations in young and old male rats.

Although fasting and refeeding reveal the existence of age-related changes in carbohydrate and lipid metabolism, the effects of aging on mineral metabolism in refed animals are unknown. We therefore investigated hormonal regulation of calcium metabolism in young (4 months) and old (26 months) male rats fasted for 48 hours and then refed for 4 or 24 hours. Serum concentrations of total and ionized calcium and parathormone were similar in control young and old rats. Serum calcitonin level was higher, and the concentrations of albumin and inorganic phosphate and alkaline phosphatase activity were lower in fed old rats. In young fasted rats, the serum ionized and total calcium was decreased, and phosphate concentration was increased. In old rats, fasting resulted in the increase of serum parathormone level. Fasting reduced serum alkaline phosphatase activity to a similar extent in both age groups. In young rats, refeeding for 24h normalized serum calcium and phosphate levels and alkaline phosphatase activity, and decreased serum concentrations of PTH and calcitonin. In old refed rats, serum calcitonin concentration was raised by 77% compared to fed or fasted animals, whereas parathormone levels were normalized. Our results indicate that old fasted or refed rats maintain normal serum calcium concentration in a different way than young animals, possibly through the increase in serum levels of parathormone and/or calcitonin. Thus, dietary manipulations such as fasting and refeeding constitute an interesting model for the investigation of the effects of aging on the hormonal regulation of serum calcium level.     

Use of a monoclonal antibody to distinguish between precursor and mature forms of human lysosomal alpha-glucosidase

The levels of functional mRNA encoding glucose-6-phosphate dehydrogenase (G6PDH; EC 1.1.1.49) were examined in hepatocytes from fasted and fasted/carbohydrate-refed rats and in hepatocytes inoculated into primary culture. Functional G6PDH mRNA was assessed in a cell-free protein synthesis system in vitro. We observed that hepatocytes from fasted/carbohydrate-refed rats had a 12-fold higher level of mRNA than did hepatocytes from fasted rats. The possibility that the adrenal glucocorticoids and insulin were responsible for the increase in G6PDH mRNA in refed rats was examined by studying the effect of insulin and the synthetic glucocorticoid, dexamethasone, on the level of functional G6PDH mRNA in primary cultures of rat hepatocytes maintained in a chemically defined medium. Hepatocytes from fasted rats were inoculated into primary culture and maintained for 48 h either in the absence of hormones or in the presence of insulin alone, dexamethasone alone or both hormones together. We observed that dexamethasone alone caused a fourfold increase in G6PDH mRNA while insulin caused about a twofold increase. Both hormones together elicited an increase that was additive. A comparison of functional G6PDH mRNA levels with the effect of the hormones on G6PDH activity and relative rate of enzyme synthesis suggests that the glucocorticoid elevates the level of G6PDH mRNA within the cell without causing a concommitant increase in the rate of synthesis of the enzyme or the level of G6PDH activity. The results obtained with the primary cultures of hepatocytes indicate that insulin and the glucocorticoids are probably involved with the regulation of hepatic G6PDH mRNA. However, involvement of other hormones, such as thyroid hormone, seems likely since the induced levels of G6PDH mRNA in hepatocytes in culture was one-third of that observed in refed rats.     

Dietary and hormonal regulation of L-type pyruvate kinase gene expression in rat small intestine

L-type pyruvate kinase is an enzyme of the glycolytic pathway whose activity and mRNA levels fluctuate in the small intestine according to dietary status. Both the enzyme activity and mRNA concentration decline during fasting and increase upon refeeding either a glucose-rich or a fructose-rich diet. Using a single-strand M 13 phage complementary to L-type pyruvate kinase mRNA as probe, we determined the level of the mRNA in the small intestine of normal, adrenalectomized, thyroidectomized, diabetic and glucagon-treated or cAMP-treated animals refed either a glucose-rich or a fructose-rich diet. The specific mRNA is present in the small intestine of normal fasted rats and increases twofold and threefold on refeeding glucose and fructose respectively. However, the hormonal control of the gene expression differs according to the dietary carbohydrate. The L-type pyruvate kinase mRNA increase, induced by glucose feeding, is hormone-dependent and requires the presence of thyroid hormones and insulin. In fructose-fed rats a certain level of mRNA increase occurs regardless of the hormonal status of the animals, but the full induction of the mRNA by fructose requires the presence of glucocorticoids, thyroid hormones and insulin. Thus, the hormonal regulation of L-type pyruvate kinase gene expression in the small intestine is largely similar to that described in normal rat liver but the basal mRNA level and the stimulation of the mRNA increase by fructose are higher in the small intestine.     

Hepatic triglyceride secretion in relation to lipogenesis and free fatty acid mobilization in fasted and glucose-refed rats

Plasma triglyceride concentrations were significantly lowered by a single feeding of glucose to rats that had been fasted for 22 hr. Three feedings of glucose produced a similar effect. In the glucose-refed animals mobilization of free fatty acids from adipose tissue was impaired more rapidly than hepatic lipogenesis was restored from its low fasting level. These effects of glucose were shown by both a 50% fall in plasma free fatty acid concentration and an 84% decrease in free fatty acid release by isolated epididymal fat pads within 30 min after a single refeeding of glucose. Hepatic lipogenesis from either acetate-1-(14)C or glucose-U-(14)C was not restored even after glucose had been fed three times at hourly intervals. Triton-induced hypertriglyceridemia was used to measure the hepatic triglyceride secretory rate; it was found that glucose refeeding decreased this rate in all but one of several experiments. This decreased secretion rate was sufficient to account for the nearly complete disappearance of triglyceride in very low density lipoproteins (d < 1.019) that occurred within 1 hr after a single glucose intubation.     

Regulation of phosphoenolpyruvate carboxykinase mRNA in mouse liver, kidney, and fat tissues by fasting, diabetes, and insulin.

Previous investigations of the phosphoenolpyruvate carboxykinase (PEPCK) gene have been conducted using rats. In a recent comparative study, we investigated, for the first time, the effects of fasting, refeeding, alloxan-induced diabetes, and insulin treatment on the levels of PEPCK mRNA in mouse liver, kidney, and adipose tissues. As in rats, fasting and diabetes induced, while insulin repressed, hepatic PEPCK mRNA. In contrast, the response of renal PEPCK mRNA to fasting, refeeding, and diabetes in mice differed quantitatively with that in rats: fasting caused a twofold increase in mice and a fourfold increase in rats. Moreover, diabetes, which induces renal PEPCK mRNA indirectly by causing acidosis in rats, was without effect in mice. In adipose tissue, the results of previous studies in both rats and mice have shown that the amount of PEPCK protein and its rate of synthesis are increased by fasting and diabetes and decreased by refeeding and insulin treatment. Thus, it was surprising to find that fasting, refeeding, alloxan-induced diabetes, and insulin treatment had no effect on adipose tissue PEPCK mRNA in either rats or mice.     

Recombinant leptin in the hypothalamic response to late fasting

The aim of the current study was to gain further insight into the implication of leptin in the regulation of hypothalamic gene expression during long-term food deprivation with emphasis on phase 3 of fasting (P3, late protein breakdown). Among plasma parameters, glucose, non-esterified fatty acids, and insulin levels tended to be decreased by leptin infusion, whilst corticosterone levels remained unchanged. From Northern blot analysis, NPY, AGRP, and MCH mRNA gene expressions were differentially regulated during prolonged fasting in leptin-perfused rats. In comparison with fed animals, NPY, AGRP, and MCH mRNA levels in P3 rats treated with leptin either remained stable or increased slightly. Regarding anorexigenic peptides (CART and POMC) and prepro-OX, fasting with leptin induced only slight changes in gene expression. Similar data have been obtained in leptin-treated fasted rats at various doses within the physiological range. We conclude that leptin and particularly low levels of plasma leptin can reasonably be considered as a constituent of a signal triggering the fasting-induced enhanced drive for refeeding in P3.