In vitro studies of the testis: the effect of LH on cyclic nucleotides.

This study evaluated the relationship between LH, cyclic AMP, cyclic GMP, and testosterone using in vitro incubation of decapsulated rat testes and sampling incubation medium. With added LH (1.0, 5.0, 100, and 500 mIU/ml) there were statistically significant increases in cyclic AMP at 5 mIU/ml or more LH, and progressively greater titers of this nucleotide were produced as LH was increased. For cyclic GMP all levels of added LH caused significant increments in titers of nucleotide; however, peak cyclic GMP concentrations occurred with 5 mIU/ml of LH. The addition of 10(-3) and 10-(4)M 8-bromo-cyclic AMP caused significant increases in testosterone production, while no changes in production of this androgen were found with 10(-3), 10(-4), or 10(-5)M 8-bromo-cyclic GMP. Neither cyclic AMP nor cyclic GMP titers were altered by the addition of 1 to 50 micrograms/ml of testosterone to medium bathing the rat testes. The dose response curves of cyclic AMP and cyclic GMP to LH are different. Progressive increments in added LH cause parallel increases of cyclic AMP and a biphasic change of cyclic GMP, 8-bromo-cyclic GMP does not cause testosterone generation, suggesting that cyclic GMP does not result in androgen synthesis. However, cyclic GMP may be involved in other Leydig cell functions.     

On the role of cyclic AMP and cyclic GMP in steroid production by bovine cortical cells

The time course of corticotropin-induced steroidogenesis and changes in intracellular cyclic AMP and cyclic GMP levels were investigated in isolated bovine adrenocortical cells prepared by trypsin digestion. Corticotropin produced a pea a peak rise in cyclic AMP during the first 5 min of stimulation and enhanced steroid production after 15 min. Corticotropin also caused a decrease in cortical cyclic GMP at 5 min; this decrease in cyclic GMP reverted to a 2–3 fold increase at 15–30 min which gradually subsided by 60 min. A steroidogenic concentration of prostaglandin E₂ also produced an elevation in the levels of both nucleotides, but the rise in cyclic GMP preceded the rise incyclic AMP. These results suggest that the relative amount of cyclic AMP and cyclic GMP, rather than the absolute levels of cyclic AMP, may be a key factor in the regulation of steroidogenesis.     

On the role of cyclic AMP and cyclic GMP in steroid production by bovine cortical cells.

The time course of corticotropin-induced steroidogenesis and changes in intracellular cyclic AMP and cyclic GMP levels were investigated in isolated bovine adrenocortical cells prepared by trypsin digestion. Corticotropin produced a peak rise in cyclic AMP during the first 5 min of stimulation and enhanced steroid production after 15 min. Corticotropin also caused a decrease in cortical cyclic GMP at 5 min; this decrease in cyclic GMP reverted to a 2-3 fold increase at 15-30 min which gradually subsided by 60 min. A steroidogenic concentration of prostaglandin E2 also produced an elevation in the levels of both nucleotides, but the rise in cyclic GMP preceded the rise in cyclic AMP. These results suggest that the relative amounts of cyclic AMP and cyclic GMP, rather than the absolute levels of cyclic AMP, may be a key factor in the regulation of steroidogenesis.     

Release of arachidonic acid and the effects of corticosteroids on steroidogenesis in rat testis Leydig cells.

The release of arachidonic acid by luteinizing hormone (LH) and the effects of inhibiting phospholipase A2 (PLA2) in vivo and in vitro on LH stimulated steroidogenesis in rat testis Leydig cells has been investigated. It was found that arachidonic acid is rapidly incorporated into phospholipids and is released within 1 min after addition of LH. The effects of treating adult rats with dexamethasone and human chorionic gonadotropin (hCG) in vivo on steroidogenesis and prostaglandin synthesis in Leydig cells isolated 6 h later were determined. It was found that hCG caused a marked increase in prostaglandin F2 alpha formation which was inhibited by treatment with dexamethasone. LH-stimulated testosterone production was inhibited in the hCG treated rats and dexamethasone caused a further decrease. Treatment with dexamethasone alone also caused a decrease in the response to LH. HCG, but not dexamethasone, had similar inhibitory effects on LH-stimulated cyclic AMP production. Similarly, the PLA2 inhibitors quinacrine, dexamethasone and corticosterone, added to the Leydig cells in vitro, inhibited LH-stimulated testosterone production but not cyclic AMP production. 11-Dehydrocorticosterone also inhibited LH-stimulated testosterone production, but higher concentrations were required to give 50% inhibition compared to corticosterone (50 and 25 microM, respectively). Ring A-reduced metabolites of corticosterone and progesterone were also found to inhibit LH-stimulated steroidogenesis. The results obtained in this and previous studies are consistent with the activation of PLA2, (either directly by LH and/or via cyclic AMP), which results in the release of arachidonic acid and the formation of leukotrienes, which stimulate steroidogenesis in the Leydig cell. This study also indicates that corticosteroids and their metabolites may exert inhibitory effects at other sites in the steroidogenic pathways, in addition to PLA2.     

Multistep regulation of Leydig cell function

LH controls Leydig cell steroidogenesis by interaction with specific membrane receptors initiating membrane coupling events. Stimulation of the androgen pathways occurs mainly through cAMP mediated mechanism including LH induced guanyl nucleotide binding, membrane phosphorylation and adenylate cyclase activation. cAMP dependent kinase activation presumably causes phosphorylation of key proteins of the steroidogenic pathway and consequent increase in testosterone production. The hormone also appears to facilitate the androgen stimulus by a cyclic AMP independent mechanism located at the plasma membrane or intracellular sites. The stimulatory event can be negatively influenced by the action of certain peptide hormones (i.e. angiotensin II) through the guanyl nucleotide inhibitory subunit of adenylate cyclase (Gi). In recent studies we have presented evidence for a Ca2+ sensitive kinase system present in purified cell membranes. Gpp(NH)p, GTP, and phospholipid in presence of nanomolar Ca2+ induce phosphate incorporation into Mr 44,500 substrate with marked inhibition at microM Ca2+. Similarly a biphasic pattern of activation was observed with adenylate cyclase activity. Membrane phosphorylation may be a modifier of LH-stimulated adenylate cyclase activity and possibly other LH induced actions in the activated Leydig cell membrane. Furthermore we have defined the stimulatory effects of forskolin on all Leydig cell cyclic AMP pools and have provided additional evidence of functional compartmentalization and/or cAMP independent facilitory stimulus of steroidogenesis by the trophic hormone. The demonstration of a novel high affinity inhibitory action of forskolin upon adenylate cyclase activity and cyclic AMP generation mediated by the Gi subunit of adenylate cyclase has provided a new approach for direct evaluation of functional inhibitory influence of Gi subunit in the Leydig cell. The cultured fetal Leydig cell system has provided a useful model to elucidate mechanisms involved in the development of gonadotropin induced estradiol mediated desensitization of steroidogenesis. We have isolated from the fetal testis a small population (2-5% of total) of transitional cells with morphological characteristics of cells found in 15 day postnatal testis but functional capabilities of the adult cell. We have also demonstrated after appropriate treatment (i.e. estrogen, and frequent or a high gonadotropin dose) the emergence of a functional adult-like cell type from the fetal Leydig cell population.     

The binding of cyclic AMP to renal brush border membranes.

The binding of cyclic AMP to the proximal tubule luminal (brush border) membrane isolated from the rabbit renal cortex was studied. The rate of binding was dependent on temperature; at 37 degrees equilibrium was attained in 45 min, whereas at 0 degrees 120 min was required. The final levels of binding were identical. The binding of 3H-cyclic AMP was reversed by dilution or addition of unlabeled cyclic nucleotide. Debinding was markedly temperature sensitive. Binding was only partially saturable with respect to cyclic AMP concentration, apparently with more than one binding site. The cyclic AMP bound to the membrane was recovered unchanged. When bound to the membrane cyclic AMP was resistant to hydrolysis by endogenous membrane or exogenously added phosphodiesterase. The binding to the membranes was relatively specific for cyclic AMP, although other cyclic purine nucleotides inhibited, cyclic IMP greater than dibutyryl cyclic AMP greater than cyclic GMP. The renal membranes did bind cyclic GMP, but this binding was relatively non-specific. Hormones and drugs, that mediate cyclic AMP generation or renal function, as well as other compounds common to the proximal tubule were without significant effect on cyclic AMP binding. Binding was inhibited by sulfhydryl reacting agents and this inhibition could be blocked and partially reversed by mercaptoethanol.     

The role of Ca2+ in steroidogenesis in Leydig cells. Stimulation of intracellular free Ca2+ by lutropin (LH), luliberin (LHRH) agonist and cyclic AMP.

The requirements of purified rat Leydig cells for intra- and extra-cellular Ca2+ during steroidogenesis stimulated by LH (lutropin), cyclic AMP analogues and LHRH (luliberin) agonist were investigated. The intracellular Ca2+ concentrations ([Ca2+]i) were measured by using the fluorescent Ca2+ chelator quin-2. The basal [Ca2+]i was found to be 89.4 +/- 16.6 nM (mean +/- S.D., n = 25). LH, 8-bromo cyclic AMP and dibutyryl cyclic AMP increased [Ca2+]i, by 300-500 nM at the highest concentrations of each stimulator, whereas LHRH agonist only increased [Ca2+]i by a maximum of approx. 60 nM. Low concentrations of LH (less than 1 pg/ml) and all concentrations of LHRH agonist increased testosterone without detectable changes in cyclic AMP. With amounts of LH greater than 1 pg/ml, parallel increases in cyclic AMP and [Ca2+]i occurred. The steroidogenic effect of the LHRH agonist was highly dependent on extracellular Ca2+ concentration ([Ca2+]e), whereas LH effects were only decreased by 35% when [Ca2+]e was lowered from 2.5 nM to 1.1 microM. No increase in [Ca2+]i occurred with the LHRH agonist in the low-[Ca2+]e medium, whereas LH (100 ng/ml) gave an increase of 52 nM. It is concluded that [Ca2+]i can be modulated in rat Leydig cells by LH via mechanisms that are both independent of and dependent on cyclic AMP, whereas LHRH-agonist action on [Ca2+]i is independent of cyclic AMP. The evidence obtained suggests that, at sub-maximal rates of testosterone production, Ca2+, rather than cyclic AMP, is the second messenger, whereas for maximum steroidogenesis both Ca2+- and cyclic-AMP-dependent pathways may be involved.     

Gonadotropin binding and stimulation of cyclic adenosine 3':5'-monophosphate and testosterone production in isolated Leydig cells.

Gonadotropin binding and stimulation of cyclic adenosine 3':5'-monophosphate (cyclic AMP) formation and testosterone synthesis were studied in collagenase-dispersed interstitial cells from the adult rat testis. Binding of 125I-human chorionic gonadotropin (hCG) by isolated Leydig cells was of high affinity (Ka = 10(10) M-1) and low capacity, equivalent to approximately 6000 sites/cell. The binding data were consistent with the presence of a single order of receptors, with no interaction between binding sites. Stimulation of testosterone synthesis by increasing concentrations of hCG was completely dissociated from changes in cyclic AMP formation, and maximum activation of steroidogenesis was induced by hCG concentrations which had no effect upon cyclic AMP production. Kinetic analysis of gonadotropin-induced responses in dispersed Leydig cells also showed a marked dissociation between steroidogenesis and cyclic nucleotide formation. Low concentrations of hCG caused maximum stimulation of testosterone production which was not accompanied by a rise in cyclic AMP formation at any time after addition of gonadotropin. Higher concentrations of hCG caused marked elevations of cyclic AMP at progressively earlier time intervals, but did not alter the 20 to 30 min lag period required for induction of testosterone synthesis. These observations indicated that occupancy of gonadotropin receptors occurs over a much wider range of hCG concentration than that required for maximum steroidogenesis.     

Elevated inhibin concentration in the follicular fluid of dairy cows with chronic cystic ovarian disease

This study compared serum and follicular fluid inhibin and gonadotropin profiles between chronic cystic ovarian diseased (CCOD) and normal cyclic dairy cows. Blood samples and follicular fluid were collected from CCOD cows (n=15) and cyclic cows in the follicular phase of the estrous cycle (control, n=6) and analyzed for inhibin, follicle stimulating hormone (FSH) and luteinizing hormone (LH) concentrations. There was a significant increase in inhibin and a decrease in FSH and LH concentrations in the follicular fluid of CCOD cows compared with those of cyclic cows (P < 0.05). Mean serum inhibin, FSH and LH concentrations between CCOD and cyclic cows were not differnt (P > 0.05), however, there was a tendency for serum inhibin to be higher and FSH to be lower in CCOD cows compared to cyclic animals (P < 0.1). The FSH pulse frequency also was lower in CCOD cows than in cyclic cows (P < 0.05). These data suggest that increased production of inhibin from cystic follicles of CCOD cows alters pituitary FSH secretion and subsequently reduces the concentration of FSH in follicular fluid. As a result, decreased FSH stimulation at the ovarian level could ultimately lead to the reduction in follicular LH and FSH receptor concentrations, resulting in abnormal follicular steroidogenesis in CCOD dairy cows.     

Stimulation of cyclic AMP production in the rat ovary by luteinizing hormone: Independence of prostaglandin mediation

In cubation of intact juvenile rat ovaries with luteinizing hormone (LH; 10 μ g/ml) for 20 min caused a ten-fold rise in cyclic AMP concentration, without increasing the activity of “prostaglandin synthetase” in the tissue. Flufenamic acid [N-(α,α,α- trifluoro-m-tolyl) anthranilic acid], aspirin or indomethacin (100 μ g/ml of each added to the incubation medium) inhibited “prostaglandin synthetase” activity by 90%, 97% and 70%, respectively, but did not prevent the stimulatory effect of LH on cyclic AMP formation. Prostaglandin E₂ (10 μ g/ml) also stimulated cyclic AMP formation in vitro, but this action was abolished by flufenamic acid. These findings argue against the hypothesis proposed in the literature that prostaglandins of the E-type are essential for the LH effect on ovarian adenylate cyclase and thus serve as obligatory mediators of cyclic AMP-dependent actions of LH on the ovary.     

Interrelationships between prostaglandins, cyclic AMP and steroids in ovulation.

Ovulation is a complex phenomenon, involving a series of biochemical events within the ovary, leading to the rupture of the follicle. This paper summarizes recent studies in our laboratory of some of these biochemical changes using the rabbit as an experimental model. It has been shown in our laboratory that isolated Graafian follicles obtained from oestrous rabbits synthesize steroids and cyclic AMP when incubated in vitro. Luteinizing hormone added to the incubation medium increased steroidogenesis and cyclic AMP synthesis many fold. When follicles were isolated from rabbits at different times following the ovulatory stimulus (mating or HCG injection) it was found that the in vitro response to LH in terms of steroidogenesis and cylcic AMP synthesis was lost as ovulation approached. In contrast, when prostaglandins (PGF and PGE) were measured in rabbit Graafian follicles it was found that the PGF and PGE levels increased as ovulation approached. From these data and from reports in the literature, we have developed a hypothetical model for ovulation in the rabbit which may help in a better understanding of the ovulatory process.     

Modulation and role of Ca2+ in LH and LHRH agonist action in rat Leydig cells

The results of our recent studies on purified rat Leydig cells indicate that there are no major qualitative differences in the stimulating effects of LH and LHRH agonists on steroidogenesis via mechanisms that are dependent on calcium. This was demonstrated by using inhibitors of calmodulin and the lipoxygenase pathways of arachidonic acid metabolism. Using the fluorescent indicator quin-2, it was shown that LH and LHRH agonist increase intracellular calcium levels; LH was more potent than LHRH agonist (max increase in concentrations obtained were 500 nM and 60 nM respectively). This difference was probably the result of a direct effect of cyclic AMP (whose production is stimulated by LH but not by LHRH) because cyclic AMP analogues were as potent as LH in increasing calcium levels. These studies indicate a major role for calcium in the control of steroidogenesis in testis Leydig cells.     

The role of calcium in luteinizing hormone-releasing hormone agonist (ICI 118630)-stimulated steroidogenesis in rat Leydig cells.

The luteinizing hormone-releasing hormone (LHRH) agonist ICI 118630 was found to increase testosterone production in purified rat testis Leydig cells in a concentration- and time-dependent manner, but no consistent changes in cyclic AMP levels were detectable. The stimulation of steroidogenesis by LHRH agonist was found to be dependent on the concentration of Ca2+ in the incubation medium; at least 1 mM was required. The calcium ionophore A23187 mimicked the effects of the LHRH agonist on steroidogenesis, and addition of both compounds together did not further increase testosterone production. The calcium ionophore caused a small increase in cyclic AMP which was independent of the concentration of the ionophore and of the calcium concentrations. The evidence obtained in this study indicates that LHRH agonist-stimulated steroidogenesis in rat testis Leydig cells is primarily mediated by calcium and not cyclic AMP.     

Receptor-mediated gonadotropin action in the ovary. Regulatory role of cyclic nucleotide phosphodiesterase(s) in intracellular adenosine 3′:5′-cyclic monophosphate turnover and gonadotropin-stimulated progesterone production by rat ovarian cells

The regulatory role of cyclic nucleotide phosphodiesterase(s) and cyclic AMP metabolism in relation to progesterone production by gonadotropins has been studied in isolated rat ovarian cells. Low concentrations of choriogonadotropin (0.4-5ng/ml) increased steroid production without any detectable increase in cyclic AMP, when experiments were carried out in the absence of phosphodiesterase inhibitors. The concentration of choriogonadotropin (10ng/ml) that stimulated progesterone synthesis maximally resulted in a minimal increase in cyclic AMP accumulation and choriogonadotropin binding. Choriogonadotropin at a concentration of 10ng/ml and higher, however, significantly stimulated protein kinase activity and reached a maximum between 250 and 1000ng of hormone/ml. Higher concentrations (50-2500ng/ml) of choriogonadotropin caused an increase in endogenous cyclic AMP, and this increase preceded the increase in steroid synthesis. Analysis of dose-response relationships of gonadotropin-stimulated cyclic AMP accumulation, progesterone production and protein kinase activity revealed a correlation between these responses over a wide concentration range when experiments were performed in the presence of 3-isobutyl-1-methylxanthine. The phosphodiesterase inhibitors papaverine, theophylline and 3-isobutyl-1-methylxanthine each stimulated steroid production in a dose-dependent manner. Incubation of ovarian cells with dibutyryl cyclic AMP or 8-bromo cyclic AMP mimicked the steroidogenic action of gonadotropins and this effect was dependent on both incubation time and nucleotide concentration. Maximum stimulation was obtained with 2mm-dibutyryl cyclic AMP and 8-bromo cyclic AMP, and this increase was close to that produced by a maximally stimulating dose of choriogonadotropin. Other 8-substituted derivatives such as 8-hydroxy cyclic AMP and 8-isopropylthio cyclic AMP, which were less susceptible to phosphodiesterase action, also effectively stimulated steroidogenesis. The uptake and metabolism of cyclic [(3)H]AMP in ovarian cells was also studied in relation to steroidogenesis. When ovarian cells were incubated for 2h in the presence of increasing concentrations of cyclic [(3)H]AMP, the radioactivity associated with the cells increased almost linearly up to 250mum-cyclic [(3)H]AMP concentration in the incubation medium. The (3)H label in the cellular extract was recovered mainly in the forms ATP, ADP, AMP, adenosine and inosine, with cyclic AMP accounting for less than 1% of the total tissue radioactivity. Incubation of cyclic AMP in vitro with ovarian cells resulted in a rapid breakdown of the nucleotide in the medium. The degradation products in the medium have been identified as AMP, adenosine and inosine. The rapid degradation of cyclic AMP by phosphodiesterase(s) makes it difficult to correlate changes in cyclic AMP concentrations with steroidogenesis. These observations thus provide an explanation for the previously observed lack of cyclic AMP accumulation under conditions in which low doses of choriogonadotropin stimulated steroidogenesis without any detectable changes in cyclic AMP accumulation.     

Differential actions of gangliosides on gonadotropin and cholera enterotoxin stimulated adenosine3':5' cyclic monophosphate dependent protein kinase in isolated rat ovarian cells.

Rat ovarian cells were exposed to cholera enterotoxin, and the effect on progesterone synthesis as well as on protein kinase stimulation was examined. Cholera enterotoxin stimulated ovarian steroidogenesis in a dose dependent manner similar to that of hCG. The stimulation of protein kinase by cholera toxin was followed by a lag period, whereas hCG effect was immediate. Mixed gangliosides, when added to the incubation medium, blocked the cholera enterotoxin-stimulated protein kinase activity and abolished the decrease in exogenous [³H] cyclic AMP receptor activity brought about by the toxin. In contrast, under similar experimental conditions ganglioside addition elicited no effect on protein kinase activation produced by hCG or LH. The data suggest that gangliosides do not appear to be directly involved in gonadotropin binding to ovarian cell membrane and subsequent mediation of physiological response.     

Control of steroidogenesis in Leydig cells.

Luteinizing hormone (LH) interacts with its plasma membrane receptor to stimulate steroidogenesis not only via cyclic AMP but also other pathways which include arachidonic acid and leukotrienes and regulation of chloride and calcium channels. The same stimulatory pathways may lead to desensitization and down-regulation of the LH receptor and steroidogenesis. The LH receptor exists in a dynamic state, being truncated, or internalized, degraded or recycled. Desensitization is controlled by protein kinase C (PKC) in the rat and by cyclic AMP dependent protein kinase and PKC in the mouse Leydig cells. Using an adapted anti-sense oligonucleotide strategy we have shown that the cytoplasmic C-terminal sequence of the LH receptor is essential for desensitization to occur. In contrast, these sequences of the LH receptor are not required for the stimulation of cyclic AMP and steroid production. We have also shown that the extracellular domain of the LH receptor is secreted from the Leydig cell and may act as a LH-binding protein.     

Evidence for two independent pathways in the stimulation of steroidogenesis by luteinizing hormone involving chloride channels and cyclic AMP

The possible role of chloride channels in luteinizing hormone (LH) action on steroidogenesis in rat Leydig cells had been investigated. A chloride channel blocker, SITS (4-acetamido-4'-isothiocyanatostilbene-2,2'-disulphonic acid), inhibited LH-stimulated steroidogenesis at low (less than or equal to 1 ng/ml), but not at high (100 ng/ml) LH concentrations. In addition, dibutyryl cyclic AMP- and forskolin-stimulated steroidogenesis was unaffected by SITS. The removal of extracellular chloride potentiated steroidogenesis stimulated by submaximal but not maximal doses of LH. These results suggest that at low levels of LH, steroidogenesis depends on chloride channels whereas with high levels, cyclic AMP is the mediator of LH action.     

Evidence for the requirement of proteolysis in LH stimulated cyclic AMP production and steroidogenesis in Leydig cells.

We have investigated the effect of protease activity on cyclic AMP production and steroidogenesis in rat testis, mouse testis and mouse tumour Leydig (MA10) cells. LH-, dibutyryl cyclic AMP-, and forskolin-stimulated steroidogenesis, but not 22R(OH) cholesterol conversion to pregnenolone, was inhibited by protease inhibitors. In mouse Leydig cells, LH but not forskolin or cholera toxin stimulated cyclic AMP production was inhibited by protease inhibitors. These results suggest that steroidogenesis in Leydig cells requires proteolysis before the conversion of cholesterol to pregnenolone. In the mouse but not rat Leydig cells, LH-stimulated cyclic AMP production is also dependent on proteolysis.     

The effects of calcium ionophore A23187 on interstitial cell steroidogenesis

The present study was performed to evaluate the effects of calcium ionophore A23187 on adenosine 3',5'-monophosphate (cyclic AMP) and testosterone production in rat interstitial cells. Interstitial cells were incubated in Krebs-Ringer solution with varying amounts of luteinizing hormone, pregnenolone, or A23187. Cyclic AMP and testosterone were measured in the incubation medium after 4 h incubation. A23187 (0.01--10 microgram/ml) caused progressive increases of cyclic AMP formation (from 0.18 +/- 0.02 (S.E.) pmol/10(6) cells for the control of 0.42 +/- 0.02 pmol/10(6) cells, P less than 0.025), while testosterone production remained unaltered. When varying amounts of A23187 were added concomitantly with luteinizing hormone (5 IU/l), A23187 inhibited luteinizing hormone-induced steroidogenesis in a dose-dependent manner, but it had no effect on luteinizing hormone-induced cyclic AMP formation. When pregnenolone (10(-6) M) was added to the cells, testosterone formation increased from 1.50 +/- 0.22 to 8.46 +/- 1.65 ng/10(6) cells. A23187 (1 microgram/ml) had no discernable effect on the conversion of pregnenolone to testosterone. The main effect of increased cytosol calcium on steroidogenesis seems to be at the steps beyond adenylate cyclase-cyclic AMP. These results suggest that calcium is important for the conversion of cholesterol to pregnenolone, while the steps beyond pregnenolone are relatively independent of Ca2+.