Mechanism of action of ATP on intestinal epithelial cells: Cyclic AMP-mediated stimulation of active ion transport

ATP, ADP and AMP but not adenosine increased cyclic AMP in dispersed enterocytes prepared from guinea pig small intestine. This action of ATP was augmented by IBMX and was reproduced by App(NH)p or App(CH₂)p. ATP also increased the formation of cyclic [¹⁴C]AMP in enterocytes that had been preincubated with [¹⁴C]adenine. Gpp(NH)p and NaF each caused persistent activation of adenylate cyclase in plasma membranes from enterocytes and ATP caused significant augmentation of this persistent activation. In addition to increasing cellular cyclic AMP and agumenting Gpp(NH)p and NaF-stimulated persistent activation of adenylate cyclase, ATP increased the I_sc across mounted strips of small intestine and inhibited net absorption of fluid and electrolytes in segments of everted small intestine. These results indicate that intestinal epithelial cells possess a receptor that interacts with ATP and other adenine nucleotides and that receptor occupation by ATP causes activation of adenylate cyclase, increased cyclic AMP and changes in active ion transport across intestinal mucosa.     

Demonstration of adenylate cyclase activity in human red blood cell ghosts

The presence of adenylate cyclase (ATP pyrophosphate-lyase (cyclizing) EC 4.6.1.1) activity was demonstrated in human erythrocyte ghosts and was found to be around 3 pmol adenosine ′,5′-monophosphatase (cyclic AMP) · 2 h^?1 · mg^?1 protein. This enzymatic activity is strongly stimulated by NaF and 5′-guanylimidodiphosphate, is slightly stimulated by epinephrine, norephrine, soproterenol, and prostaglandin E, and is inhibited by calcium. The hormone stimulation is not potentiated by 5′-guanylylimidodiphosphate.     

Specific MG2+ and adenosine sites involved in a bireactant mechanism for adenylate cyclase inhibition and their probable localization on this enzyme's catalytic component.

The specificity of adenosine sites involved in adenylate cyclase inhibition (P sites) is identical on membrane-bound and on solubilized enzyme. Kinetic analysis indicates that in addition to a low affinity Mg²⁺ site involved in adenylate cyclase stimulation (Km = 10 mM), there is a high affinity Mg²⁺ site (Km = 2.10^?4M) involved together with P sites in a bireactant mechanism for triggering adenylate cyclase inhibition. Guanyl nucleotide-binding protein does not seem to be implicated in this inhibition. We were not able to separate the catalytic component of adenylate cyclase from P sites, either on a sucrose density gradient or in gel filtration experiments. It is suggested that P sites are located on the catalytic component of the enzyme.     

Guanosine 5', alpha-beta-methylene, triphosphate, a novel GTP analog, causes persistent activation of adenylate cyclase: evidence against pyrophosphorylation mechanism.

To test the hypothesis that guanine nucleotides activate adenylate cyclase by a covalent mechanism involving pyrophosphorylation of the enzyme, we studied the effect of a novel GTP analog, guanosine 5′, α-β-methylene triphosphate (Gp(CH₂)pp), with a methylene bond in the α-β-position that is stable to enzymatic hydrolysis. Gp(CH₂)pp was as effective as GTP in stimulating rat reticulocyte adenylate cyclase in the presence of isoproterenol. Previously only guanine nucleotides with modified terminal phosphates such as guanylyl 5′-imidodiphosphate (Gpp(NH)p) were thought capable of causing persistent activation of adenylate cyclase. Gp(CH₂)pp, however, caused persistent activation of rat reticulocyte and turkey erythrocyte adenylate cyclase. We conclude that guanine nucleotides do not activate adenylate cyclase by a pyrophosphorylation mechanism and that a modified γ-phosphate is not essential in guanine nucleotides for generation of the irreversibly-activated enzyme state.     

Cyclic AMP metabolism in mouse parotid glands properties of adenylate cyclase,protein kinase and phosphodiesterase

Catecholamines induce unique growth and secretory responses in salivary glands. An analysis of three enzyme activities involved in cyclic AMP metabolism was carried out to identify the specificity of these responses for salivary glands.Although parotid adenylate cyclase has an unusually high specific activity, its kinetic properties and responses to NaF, guanine nucleotides, and isoproterenol are similar to other tissues not stimulated to grow after isoproterenol stimulation. Solubilized adenylate cyclase was separated from other membrane proteins by isoelectric focusing on polyacrylamide gels. There was a single broad peak of activity with a pI of 5.9. Parotid protein kinase has a subcellular distribution and substrate preference similar to hepatic protein kinase. Activation by cyclic AMP is also similar to that reported for other tissues, with a K_a of 1.2·10^?7 M. Parotid cyclic AMP and cyclic GMP phosphoriesterases are a heterogeneous group of enzymes with relatively low specific activity as compared with mouse pancreas, liver and brain. Isoelectric focusing of supernatant phosphodiesterases revealed at least six peaks of enzyme activity in the pI range of 4–6.Previous reports of a large increase in parotid cyclic AMP levels after in vivo administration of catecholamines and specific growth and secretion could be the result of a relatively high specific activity adenylate cyclase associated with low specific activity cyclic AMP phosphodiesterases.     

Properties of β-adrenergic receptors in untreated and butyrate-threated HeLa cells

HeLa cells contain receptors on their surface which are β-adrenergic in nature. The binding of (?)-[³H]dihydroalprenolol is rapid, reversible, stereo-specific and of relatively high affinity. The HeLa cells also contain an adenylate cyclase which is activated by (?)-isoproterenol > (?)-epinephrine > (?)-norepinephrine. The adenylate cyclase of HeLa is also activated by guanyl-5′-yl-imidodophosphate (Gpp(NH)p), a nonhydrolyzable analogue of GTP. Inclusion of both (?)-isoproterenol and Gpp(NH)p leads to approximately additive rathen than synergistic activation of adenylate cyclase. After treatment of HeLa cells with 5 mM sodium butyrate there is an increase in the number of β-adrenergic receptors, but not in their affinity, which is reflected in an increased ability of (?)-isoproterenol to activate adenylate cyclase. Other properties of the β-adrenergic receptor including association and dissociation rates, temperature optimum of adenylate cyclase and response to Gpp(NH)p are relatively unaffected by butyrate pretreatment of the cells.     

Roles of GTP and GDP in the regulation of the thyroid adenylate cyclase system

Effects of guanine nucleotides on the adenylate cyclase activity of thyroid plasma membranes were investigated by monitoring metabolism of the radiolabeled nucleotides by thin-layer chromatography (TLC). When ATP was used as substrate with a nucleotide-regeneratign system, TSH stimulated the adenylate cyclase activity in the absence of exogenous guanine nucleotide. Addition of GTP and GDP equally enhanced the TSH stimulation. Effects of GTP and GDP were indistinguishable in regard to their inhibitory effects on NaF-stimulated activities. The results from TLC suggested that GDP could be converted to GTP by a nucleotide-regenerating system. Even in the absence of nucleotide-regenerating system, addition of GDP to the adenylate cyclase assay mixture int he parallel decrease in ATP levels and formation of GTP indicating that thyroid plasma membrane preparatiosn possessed a transphosphorylating activity. When an ATP analog, App[NH]p, was used as substrate without a nucleotide-regenerating system, no conversion of GDP to GTP was observed. Under such conditions, TSH did not stimulate the adenylate cyclase activity unless exogenous GTP or Gpp[NH]p was added. GDP no longer supported TSH stimulation and caused a slight decrease in the activity. GDP was less inhibitory than Gpp(NH)p to the NaF-stimulated adenylate cyclase activity. These results suggest: (1) TSH stimulation of thyroid adenylate cyclase is absolutely dependent on the regulatory nucleotides. (2) In contrst to GTP, GDP cannot support the coupling of the receptor-TSH complex to the catalytic componenet of adenylate cyclase. (3) The nucleotide regulatory site is more inhibitory to the stimulation of the enzyme by NaF when occupied by Gpp[NH]p than GDP.     

Presence of “Ra” and “P”-site receptors for adenosine coupled to adenylate cyclase in cultured vascular smooth muscle cells

The existence of adenosine receptors coupled to adenylate cyclase in cultured vascular smooth muscle cells from rat aorta is demonstrated in these studies. Adenosine, N⁶-phenylisopropyladenosine, adenosine N′-oxide and 2-chloroadenosine stimulated adenylate cyclase in a concentration dependent manner. The stimulation was dependent on the presence of guanine nucleotides and was blocked by 3-isobutyl-1-methylxanthine. In contrast, 2′ deoxyadenosine inhibited adenylate cyclase activity. Adenosine and 2-chloroadenosine showed a biphasic effect on adenylate cyclase, stimulation occurred at low concentrations. The activation of adenylate cyclase by N⁶-phenylisopropyladenosine was also dependent on the Mg²⁺ concentration. The data suggest that vascular smooth muscle cells have both “Ra” and “P” receptors for adenosine, and it can be postulated that the relaxant effect of adenosine on vascular smooth muscle may be mediated by its interaction with “Ra” receptors associated with adenylate cyclase.     

Specific binding of the calcium-dependent regulator protein to brain membranes from the guinea pig

The interaction of a calcium-dependent regulator protein (CDR) of brain adenylate cyclase (EC 4.6.1.1) with synaptic membranes from guinea pig brain was examined using ¹²⁵I-CDR as a tracer molecule. ¹²⁵I-CDR binding was reversible, saturable, and temperature sensitive. The same Ca²⁺ and Mg²⁺ dependence was observed for ¹²⁵I-CDR binding and for brain adenylate cyclase activation by CDR.     

Activation by a calcium-binding protein of guanylate cyclase in Tetrahymena pyriformis.

A Ca²⁺-binding protein (TCBP), which was isolated from $\text{Tetrahymena pyriformis}$, enhanced about 20-fold particulate-bound guanylate cyclase activity in $\text{Tetrahymena}$ cells in the presence of a low concentration of Ca²⁺, while the adenylate cyclase activity was not increased. The enhancement was eliminated by ethylene glycol-bis (β-aminoethyl ether)-N,N′-tetraacetic acid. The enzyme activity was not stimulated by rabbit skeletal muscle troponin-C, the Ca²⁺-binding component of troponin, or other some proteins. In the presence of TCBP, stimulating effect of calcium ion on the enzyme activity was observed within the range of pCa 6.0 to 4.6, and was immediate and reversible.     

Evidence for two types of β-adrenergic-sensitive adenylate cyclase activities in bovine cerebellum

A latent, as well as an expressed form of adenylate cyclase coupled to β-adrenergic receptors is present in intact crude synaptosomal preparations from bovine cerebellum. The latent adenylate cyclase activity was assayed in Krebs-Ringer buffer by [³H]adenine labeling and was found to be coupled to a β₁-like adrenergic receptor. The externally accessible adenylate cyclase assayed in the same with [³H]ATP was stimulated via β₂-adrenergic receptors.     

Multiple effects of guanine nucleotides on human platelet adenylate cyclase

We report that the adenylate cyclase system in human platelets is subject to multiple regulation by guanine nucleotides. Previously it has been reported that GTP is either required for or has little effect on the response of the enzyme to prostaglandin E₁. We have found that when platelet lysates were prepared in the presence of 5 mM EDTA, GTP lowered the basal and prostaglandin E₁-stimulated adenylate cyclase activity when the enzyme was assayed in the presence of Mg²⁺. The basal and prostaglandin E₁-stimulated adenylate cyclase activities were also increased by washing, which presumably removes endogenous GTP. The analog, guanyl-5′-yl-imidodiphosphate mimics the inhibitory effect of GTP on prostaglandin E₁-stimulated adenylate cyclase activity but it stimulates basal enzyme activity. The onset of the inhibitory effect of GTP on the adenylate cyclase system is rapid (1 min) and is maintained at a constant rate during incubation for 10 min. GTP and guanyl-5′-yl-imidodiphosphate were noncompetitive inhibitors of prostaglandin E₁. An increase in the concentration of Mg²⁺ gradually reduces the effect of GTP while having little influence on the effect of guanyl-5′-yl-imidodiphosphate. Neither the substrate concentration nor the pH (7.2–8.5) is related to the inhibitory effect of guanine nucleotides. The inhibition by nucleotides was found to show a specificity for purine nucleotides with the order of potency being guanyl-5′-yl-imidodiphosphate > dGTP > GTP > ITP > XTP > CTP > TTP. The inhibitory effect of GTP is reversible while the effect of guanyl-5′-yl-imidodiphosphate is irreversible. The GTP inhibitory effect was abolished by preparing the lysates in the presence of Ca²⁺. However, the inhibitory effect of guanyl-5′-yl-imidodiphosphate persisted. Substitution of Mn²⁺ for Mg²⁺ in the assay medium resulted in a diminution of the inhibitory effect of GTP on basal activity and converted the inhibitory effect of GTP on prostaglandin E₁-stimulated activity to a stimulatory effect. At a lower concentration of Mn²⁺ (less than 2 mM) guanyl-5′-yl-imidodiphosphate inhibited prostaglandin E₁-stimulated adenylate cyclase activity, but at a higher concentration of Mn²⁺, it caused an increase in enzyme activity exceeding that occuring in the presence of prostaglandin E₁. In the presence of Mn²⁺, dGTP mimics the effect of GTP and is 50% as effective as GTP. Our data suggest that the inhibitory effect of GTP on prostaglandin E₁-stimulated adenylate cyclase is mainly due to its direct effect on the enzyme itself, whereas the stimulatory effect of GTP on prostaglandin E₁-stimulated adenylate cyclase is due to enhancement of the coupling between the prostaglandin E₁ receptor and adenylate cyclase. These studies also indicate that the method of preparation of platelet lysates can profoundly alter the nature of guanine nucleotide regulation of adenylate cyclase.     

Effects of forskolin on adenylate cyclase,cyclic AMP,protein kinase and intermediary metabolism of the thyroid gland

Forskolin (40 μM) stimulated adenylate cyclase activities of bovine thyroid plasma membranes without pthe addition of guanine nucleotides. GDP had little effect on the forskolin-stimulated adenylate cyclase activity while Gpp[NH]p (0.1–1.0 μM) decreased it. In the presence of TSH (10 mU/0.11), Gpp[NH]p no longer caused inhibition. Forskolin did not affect phosphodiesterase activities of thyroid homogenates. Forskolin (10 μM) rapidly increased cAMP levels in bovine thyroid slices both in the absence and presence of a phosphodiesterase inhibitor. The effect of TSH (50 mU/ml) on cAMP levels was additive or greater than additive to that of forskolin. An initial 2-h incubation of slices with forskolin did not decrease their subsequent cAMP responses to either forskolin and/or TSH while similar treatment of slices with TSH induced desensitization of the cAMP response to TSH, but not to forskolin. Forskolin (10 μM) as well as TSH (50 mU/ml) activated cAMP-dependent protein kinase of slices in the absence of a phosphodiesterase inhibitor. Although forskolin activated the adenylate cyclase cAMP system, it did not stimulate iodide organification or glucose oxidation, effects which have been attributed to cAMP. In fact, forskolin inhibited these parameters and ³²P incorporation into phospholipids as well as their stimulation by TSH. These results indicate that an increase in cAMP levels and cAMP-dependent protein kinase activity in thyroid slices may not necessarily reproduce the effects of TSH on the thyroid.     

Effects of calcium on a parathyroid hormone-sensitive adenylate cyclase inhibitor

Inhibition of parathyroid hormone (PTH)-sensitive adenylate cyclase by {Nle-8, Nle-18, Tyr-34} bPTH-(3–34) amide was studied in thyroparathyroid-ectomized dogs. The inhibitory effect was shown to be markedly enhanced by the addition of calcium ions into the $\text{in}\text{,}\text{vitro}$ assay system. At 0.1 mM Ca²⁺, complete inhibition by the antagonist was obtained. Chelation of exogenous Ca²⁺ by EGTA eliminated the Ca²⁺-induced inhibition. Both the basal and hormone-stimulated activities were decreased in the presence of 0.1 mM Ca²⁺, whereas the addition of EGTA increased both activities. Our results suggest that Ca²⁺ modulates canine renal PTH-sensitive adenylate cyclase and its inhibition by substituted bPTH-(3–34).     

Catecholamine and guanine nucleotide activation of skeletal muscle adenylate cyclase

Activation of adenylate cyclase by guanine nucleotide and catecholamines was examined in plasma membranes prepared from rabbit skeletal muscle. The GTP analog, 5′-guanylyl imidodiphosphate caused a time and temperature-dependent activation of the enzyme which was persistent, the K_a was 0.05 μM. 5′-Guanylyl imidodiphosphate binding to the membranes was time and temperature dependent, K_D 0.07 μM. Beta adrenergic amines accelerated the rate of 5′-guanylyl imidodiphosphate activation of the enzyme with an order of potency isoproterenol ≈ soterenol ≈ salbutamol > epinephrine ? norepinephrine. Catecholamine activation was antagonized by propranolol and the β₂ antagonist butoxamine; the β₁ antagonist practolol was inactive. [³H]Dihydroalprenolol bound to the membranes and binding was antagonized by β adrenergic agonists with an order of potency similar to the activation of adenylate cyclase and was antagonized by butoxamine but not by practolol. The data are consistent with the idea that adenylate cyclase in skeletal muscle plasma membranes is coupled to adrenergic receptors of the β₂ type.     

Adenylate cyclase system of human skeletal muscle: Characteristics of catecholamine stimulation and nucleotide regulation

The subcellular localization of adenylate cyclase was examined in human skeletal muscle. Three major subcellular membrane fractions, plasmalemma, sarcoplasmic reticulum and mitochondria, were characterized by membrane-marker biochemical studies, by dodecyl sulfate polycrylamide gel electrophoresis and by electron microscopy. About 60% of the adenylate cyclase of the homogenate was found in the plasmalemmal fraction and 10–14% in the sarcoplasmic reticulum and mitochondria. When the plasmalemmal preparation was subjected to discontinuous sucrose gradients, the distribution of adenylate cyclase in different subfractions closely paralleled that of (Na⁺ + K⁺)-ATPase. The highest specific activity was found in a fraction which setteled at the 0.6–0.8 M sucrose interface. The electron microscopic study of this fraction revealed the presence of flattened sacs of variable sizes and was devoid of mitochondrial and myofibrillar material. The electron microscopy of each fraction supported the biochemical studies with enzyme markers. The three major membrane fractions also contained a low K_m phosphodiesterase activity, the highest specific activity being associated with sarcoplasmic reticulum.The plasmalemmal adenylate cyclase was more sensitive to catecholamine stimulation than that associated with sarcoplasmic reticulum or mitochondria. The catecholamine-sensitive, but not the basal, enzyme was further stimulated by GTP. The plasmalemmal adenylate cyclase had typical Michaelis-Menten kinetics with respect to ATP and the apparent K_m for ATP was approx. 0.3. mM. The pH optimum for that enzyme was 7.5. The enzyme required Mg²⁺, and the concentration to achieve half-maximal stimulation was approx. 3 mM. Higher concentrations of Mg²⁺ (about 10 mM) were inhibitory. Solubilization of the plasmalemmal membrane fraction with Lubrol-PX resulted in preferential extraction of 106 000- and 40 000-dalton protein components. The solubilized adenylate cyclase lost its sensitivity for catecholamine stimulation, and the extent of fluoride stimulation was reduced to one-sixth of that of the intact membranes. It is concluded that the catalytically active and hormone-sensitive adenylate cyclase is predominantly localized in the surface membranes of the cells within skeletal muscle. (That “plasmalemmal” fraction is considered likely to contain, in addition to plasmalemma of muscle cells, plasmalemma of bloodvessel cells (endothelium, and perhaps smooth muscle) which may be responsible for a certain amount of the adenylate cyclase activity and other propertiesobserved in that fraction.)The method of preparation used in this study provides a convenient material for evaluating the catecholamine-adenylate cyclase interactions in human skeletal muscle.     

Stimulation of brain adenylate cyclase activity by the undecapeptide substance P and its modulation by the calcium ion

Synthetic substance P stimulated adenylate cyclase activity in particulate preparations from rat and human brain.The concentration of substance P for half maximal stimulation in rat brain was 1.8 · 10⁻⁷ M.The stimulatory effect of substance P on the rat brain adenylate cyclase activity was 88% compared with 48% by noradrenalin, 163% by prostaglandin E₁ and 184% by prostaglandin E₂.Both the basal and substance P-stimulated adenylate cyclase activity in rat brain were inhibited by concentration of Ca²⁺ above 10⁻⁶ M.The chelating agent ethyleneglycol-bis-(β-aminoethylether)-N,N′-tetraacetic acid at a concentration of 0.1 mM reduced the basal adenylate cyclase activity by 64% and eliminated the substance P-stimulated activity.The inhibition by ethyleneglycol-bis-(β-aminoethylether)-N,N′-tetraacetic acid was completely reversed by increasing concentrations of Ca²⁺.     

Preparation of renal cortex basal-lateral and brush border membranes. Localization of adenylate cyclase and guanylate cyclase activities

Luminal brush border and contraluminal basal-lateral segments of the plasma membrane from the same kidney cortex were prepared. The brush border membrane preparation was enriched in trehalase and γ-glutamyltranspeptidase, whereas the basal-lateral membrane preparation was enriched in (Na⁺ + K⁺)-ATPase. However, the specific activity of (Na⁺ + K⁺)-ATPase in brush border membranes also increased relative to that in the crude plasma membrane fraction, suggesting that (Na⁺ + K⁺)-ATPase may be an intrinsic constituent of the renal brush border membrane in addition to being prevalent in the basal-lateral membrane. Adenylate cyclase had the same distribution pattern as (Na⁺ + K⁺)-ATPase, i.e. higher specific activity in basal-lateral membranes and present in brush border membranes. Adenylate cyclase in both membrane preparations was stimulated by parathyroid hormone, calcitonin, epinephrine, prostaglandins and 5′-guanylylimidodiphosphate. When the agonists were used in combination enhancements were additive. In contrast to the distribution of adenylate cyclase, guanylate cyclase was found in the cytosol and in basal-lateral membranes with a maximal specific activity (NaN₃ plus Triton X-100) 10-fold that in brush border membranes. ATP enhanced guanylate cyclase activity only in basal-lateral membranes. It is proposed that guanylate cyclase, in addition to (Na⁺ + K⁺)-ATPase, be used as an enzyme “marker” for the renal basal-lateral membrane.     

Solubilization of adenylate cyclase of brain membranes by lipid peroxidation

Adenylate cyclase in the membrane fractions of bovine and rat brains, but not in rat liver plasma membranes, was solubilized by treatment with Fe²⁺ (10 μM) plus dithiothreitol (5 mM). Solubilization of the enzyme by these agents was completely prevented by simultaneous addition of N,N′-diphenyl-p-phenylenediamine (DPPD), an inhibitor of lipid peroxidation. Ascorbic acid also solubilized the enzyme from the brain membranes. Lipid peroxidation of the brain membranes was characterized by a selective loss of phosphatidylethanolamine. Solubilization of membrane-bound enzymes by Fe²⁺ plus dithiothreitol was not specific for adenylate cyclase, because phosphodiesterase, thiaminediphosphatase and many other proteins were also solubilized. Solubilized adenylate cyclase had a high specific activity and was not activated by either NaF, 5′-guanylyl imidodiphosphate (Gpp[NH]p) or calmodulin. These results suggested that lipid peroxidation of the brain membranes significantly solubilized adenylate cyclase of high specific activity.     

Lack of adenylate and guanylate cyclases responsiveness to hormones in a spontaneous murine thyroid tumor

Animals with tumors were obtained from Dr. ZAJDELA and belong to sublines (XVIInc/Z/E) in which some individuals (TT) developed after 15 months thyroid tumors weighing between 150 and 1200 mg. Hyperplasia affects thyrocytes which do not present a follicular structure. The purpose of our work was to assay the action of various effectors on the adenylate and guanylate cyclase system in vitro. The following results have been obtained: the cyclic-AMP content of tumor tissue is not raised either by TSH or PGE₂. Nevertheless, TSH enhances the phosphatidylinositol phosphate turnover (phospholipid effect) as in normal tissue. This latter observation points at the existence of functional TSH receptors in tumor cells. The study of adenylate cyclase activity of the tumor homogenate shows the presence of this enzyme and its responsiveness to NaF and GppNHp. Unexpectedly, the cyclase is also sensitive to the stimulation by TSH.A tentative interpretation of these facts is that no component of the cyclase is missing, but that they are physically separated. The homogeneization allows the various components to interact productively.A parallel study was devoted to cyclic-GMP. Carbamylcholine fails to increase the cyclic-GMP content of the tumor tissue, whereas it has the described phospholipid effect on phosphatidylinositol. Nevertheless, there is no deficiency in the guanylate cyclase activity, since nitroprusside enhances strongly the cyclic-GMP content of the tumor.To conclude, the murine thyroid tumor presents a genetic alteration that results in the uncoupling of effector binding and catalytic stimulation of adenylate and guanylate cyclase.