Azidophenantridinium compounds as photoaffinity labels of cholinergic proteins

The synthesis of diazidopropidium and diazidoethidium is described. The applicability of these compounds as photoaffinity labels for cholinergic proteins has been investigated: diazidopropidium inhibits neuromuscular transmission. This inhibition is reversible if the compound is applied in the dark but becomes irreversible after irradiation with white light. Inhibition is accompanied by a disappearance of miniature endplate potentials. Electrophysiological analysis of this effect indicates that diazidopropidium acts postsynaptically by blocking the acetylcholine receptors. At the molecular level the action of diazidopropidium and diazidoethidium on acetylcholinesterase has been investigated: both compounds appear to bind to a peripheral acetylcholine binding site of this enzyme. Binding of ¹²⁵I-labeled α-neurotoxin from Naja naja siamensis to purified membranes from Torpedo californica electric tissue rich in acetylcholine receptors is diminished after incubation and irradiation with diazidopropidium. About half of the toxin binding sites appear to be blocked by the photoaffinity label.     

Photoaffinity labeling of specific muscarinic antagonist binding sites of brain: I. Preliminary studies using two p-azidophenylacetate esters of tropine

Possible photoaffinity probes for muscarinic acetylcholine receptors have been explored for the first time: Specific [³H]-quinuclidinyl benzylate binding sites of several fractions from rat brain can be irreversibly inactivated by photoaffinity labeling with two p-azidophenylacetate esters of tropine. Inactivation of these sites depends on formation of a reversible complex with the azides prior to their photolytic conversion to the highly reactive nitrenes; it is dependent on ligand concentration and length of photolysis. Atropine and oxotremorine, but not d-tubocurarine, afford protection against photoinactivation.These findings suggest the utility of these and related azido derivatives as potent, selective photoaffinity ligands directed against binding sites for muscarinic antagonists.     

Photoaffinity labeling of A1-adenosine receptors

The ligand-binding subunit of the A1-adenosine receptor has been identified by photoaffinity labeling. A photolabile derivative of R-N6-phenylisopropyladenosine, R-2-azido-N6-p-hydroxyphenylisopropyladenosine (R-AHPIA), has been synthesized as a covalent specific ligand for A1-adenosine receptors. In adenylate cyclase studies with membranes of rat fat cells and human platelets, R-AHPIA has adenosine receptor agonist activity with a more than 60-fold selectivity for the A1-subtype. It competes for [3H]N6-phenylisopropyladenosine binding to A1-receptors of rat brain membranes with a Ki value of 1.6 nM. After UV irradiation, R-AHPIA binds irreversibly to the receptor, as indicated by a loss of [3H]N6-phenylisopropyladenosine binding after extensive washing; the Ki value for this photoinactivation is 1.3 nM. The p-hydroxyphenyl substituent of R-AHPIA can be directly radioiodinated to give a photoaffinity label of high specific radioactivity (125I-AHPIA). This compound has a KD value of about 1.5 nM as assessed from saturation and kinetic experiments. Adenosine analogues compete for 125I-AHPIA binding to rat brain membranes with an order of potency characteristic for A1-adenosine receptors. Dissociation curves following UV irradiation at equilibrium demonstrate 30-40% irreversible specific binding. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis indicates that the probe is photoincorporated into a single peptide of Mr = 35,000. Labeling of this peptide can be blocked specifically and stereoselectively by adenosine receptor agonists and antagonists in a manner which is typical for the A1-subtype. The results indicate that 125I-AHPIA identifies the ligand-binding subunit of the A1-adenosine receptor, which is a peptide with Mr = 35,000.     

Muscarinic Acetylcholine Receptors in Neuroblastoma Cells: Lack of Effect of Veratrum Alkaloids on Receptor Number

The effect of compounds that activate sodium channels on the number of muscarinic acetylcholine receptors in neuroblastoma NIE 115 cells has been investigated. The cells were used in electrically unexcitable ("control" cells) and excitable ("differentiated" cells) states. Although receptor assays using a single concentration of the radioligand [3H]scopolamine methyl chloride indicated a loss of receptors after a 6-h incubation of cells with veratrine, no true loss of receptors was seen with any of the compounds tested (veratridine, veratrine, aconitine) when full saturation analyses were performed in either control or differentiated cells. The apparent receptor loss seen with veratrine was due to a muscarinic receptor-active component of veratrine (not veratridine) occluded by the cells and released into the binding assays upon cell breakage. Veratridine and aconitine have a very low affinity for muscarinic acetylcholine receptors, and the binding of carbamoylcholine to the receptors is unaffected by tetrodotoxin, so that there is no evidence in this system for interaction between muscarinic receptors and sodium channels.     

Photoaffinity probes for serotonin and histamine receptors. Synthesis and characterization of two azide analogues of ketanserin

Two azide analogues of ketanserin (6- and 7-azido-3-[2- [4-(4-fluorobenzoyl)-1-piperidinyl]ethyl]-2, 4(1H,3H)-quinazolinedione) were synthesized and tested as possible photoaffinity probes for serotonin-S2 and histamine-H1 receptors. In reversible binding experiments, the azides showed high affinity for both receptor types. When membrane preparations were incubated with nanomolar concentrations of 7-azidoketanserin and subsequently irradiated with UV light, both serotonin and histamine receptors became irreversibly blocked. This irreversible binding was dependent on azide concentrations and time of irradiation and did not change in the presence of the scavenger p-aminobenzoic acid. In contrast, irreversible blockade at low concentrations of 6-azidoketanserin was only obtained for histamine receptors. However, this blockade was abolished by addition of the scavenger p-aminobenzoic acid indicating that it was not due to a real photoaffinity mechanism. In the rat prefrontal cortex, irreversible blocking of serotonin receptors with 7-azidoketanserin could be inhibited by serotonin agonists or antagonists but not by histaminergic compounds. On the contrary, in the guinea pig cerebellum, inactivation of histamine receptors could be inhibited by histamine antagonists and histamine itself but not by serotonergic compounds. This provides a way for differential photolabeling of either of these receptors.     

Photoaffinity labeling of D-2 dopamine binding subunits from rat striatum, anterior pituitary and olfactory bulb with a new probe, [3H]azidosulpride

A novel photoaffinity probe [3H] azidosulpride has been developed for biochemical studies of D-2 dopamine receptors. This ligand binds to the receptors with high affinity (Kd = 3.1 +/- 0.2 nM) and, upon photoactivation, about 20% of the radioactivity bound to membranes becomes covalently incorporated. More than 90% of this irreversible binding is protectable by dopaminergic agents including D-2 selective compounds, whereas D-1 selective and non-dopaminergic compounds are ineffective. Analysis by sodium dodecylsulfate polyacrylamide gel electrophoresis reveals a single band at Mr = 85 kDa for labeled receptors in striatum, anterior pituitary or olfactory bulb, where pharmacologically distinct binding sites have been previously detected.     

Photoaffinity labelling of juvenile hormone binding proteins in the cockroach Leucophaeamaderae

The synthesis and testing of several diazocarbonyl JH analogs (diazo JHA) which act as photoaffinity labels for insect juvenile hormone binding proteins are described. The best competitor, 10,11-epoxyfarnesyl diazoacetate, has been shown to irreversibly reduce [³H]-JH III binding to both ovarian and hemolymph JHBP from Leucophaeamaderae after irradiation at 254 nm for 20 seconds. No loss of activity was observed after incubation of JHBP and diazo JHA without irradiation. Protection from photoinactivation by diazo JHA II was achieved by the presence of an equimolar amount of JH III during the photolysis. Photoaffinity labeled proteins show loss of binding capacity without alteration of the binding affinity. This is the first example of the use of a photoaffinity label in the study of JH action on a molecular level, and may become a valuable tool in the elucidation of JH-receptor-chromatin interactions.     

N-linked oligosaccharides are responsible for rat striatal dopamine D2 receptor heterogeneity

The glycoprotein nature of the binding subunit of the dopamine D2 receptor in rat striatum has been examined by photoaffinity labeling receptor preparations with N-(p-azido-m-[125I]iodophenethyl)spiperone followed by treatment of crude membrane receptor or receptor fractions isolated from sodium dodecyl sulfate (SDS) polyacrylamide gels with endo- and exoglycosidases. The major photoaffinity labeled protein migrates as a heterogeneous species on 10% SDS polyacrylamide gels and ranges from 130,000 to 75,000 relative molecular mass (Mr). This heterogeneity can be explained by glycosylation of the receptor by complex-type N-linked oligosaccharides. Three fractions of labeled receptor were isolated from SDS polyacrylamide gels over a range of 130,000 to 75,000 Mr; after digestion with peptide-N4-[N-acetyl-beta-glucosaminyl] asparagine amidase, all fractions yielded a single peptide approximately 40,000 Mr. Treatment of photoaffinity labeled membranes with alpha-mannosidase was without effect. The dopamine D2 receptor appears to contain substantial amounts of sialic acid as treatment of photoaffinity labeled membranes with neuraminidase increased the receptor mobility on SDS polyacrylamide gels to a species of 50,000-54,000 Mr. Treatment of the receptor with neuraminidase followed by endo-alpha-N-acetylgalactosaminidase did not change the electrophoretic migration pattern from that seen after neuraminidase treatment alone, suggesting that the binding peptide contains no serine- or threonine-linked oligosaccharides. A smaller binding peptide of approximately 31,000 Mr is also apparent in crude photoaffinity labeled membranes. This material also contains N-linked oligosaccharide. Complete removal of N-linked oligosaccharide from the dopamine D2 receptor did not change the rank order potency of agonist and antagonist compounds to compete for [3H]spiperone binding to crude membrane fractions. The dopamine D2 receptor represents a highly glycosylated neural receptor.     

The apparent target size of rat brain benzodiazepine receptor, acetylcholinesterase, and pyruvate kinase is highly influenced by experimental conditions

Radiation inactivation is a method to determine the apparent target size of molecules. In this report we examined whether radiation inactivation of various enzymes and brain receptors is influenced by the preparation of samples preceding irradiation. The apparent target sizes of endogenous acetylcholinesterase and pyruvate kinase from rat brain and from rabbit muscle and benzodiazepine receptor from rat brain were investigated in some detail. In addition the target sizes of alcohol dehydrogenase (from yeast and horse liver), beta-galactosidase (from Escherichia coli), lactate dehydrogenase (endogenous from rat brain), and 5-HT2 receptors, acetylcholine muscarine receptors, and [35S] butyl bicyclophosphorothionate tertiary binding sites from rat brain were determined. The results show that apparent target sizes are highly influenced by the procedure applied for sample preparation before irradiation. The data indicate that irradiation of frozen whole tissue as opposed to lyophilized tissue or frozen tissue homogenates will estimate the smallest and most relevant functional target size of a receptor or an enzyme.     

Interaction of myasthenic serum globulin with the acetylcholine receptor.

A serum factor from patients with myasthenia gravis which inhibited the binding of 125I-labeled alpha-bungarotoxin to acetylcholine receptor extracted with Triton X-100 from rat muscle has been studied in detail. The inhibitory activity was localized to the IgG fraction based upon the fractionations by sodium sulfate precipitation and DEAE chromatography as well as reaction with anti-IgG globulin. The myasthenic globulin inhibited toxin binding to receptors extracted from degenerated muscle but did not inhibit toxin binding to normal junctional receptors. At saturation levels of myasthenic globulin, the number of denervated acetylcholine receptors available for toxin binding was reduced approx. 50 percent. The myastehnic globulin was found to bind to denervated acetylcholine receptors but not to normal acetylcholine receptors by a radioimmunoassay technique in which myasthenic globulin incubated with 125I-labeled alpha bungarotoxin-receptor complexes was precipitated by anti-IgG serum. The globulin binding was saturable over the same range as inhibition of toxin binding. The data suggest that the myasthenic IgC binds to a site on the receptor complex juxtaposed to the acetylcholine receptor site. The myasthenic globulin appears to be a useful probe for investigation differences between acetylcholine receptors extracted from normal and denervated muscle and for investigating the pathogenesis of myasthenia gravis.     

Structural organization of nicotinic acetylcholine receptors

Nicotinic acetylcholine receptor of the electric ray Torpedo is the most comprehensively characterized neurotransmitter receptor. It consists of five subunits (alpha2beta gammadelta) amino acid sequences of which were determined by cDNA cloning and sequencing. The shape and size of the receptor were determined by electron cryomicroscopy. It has two agonist/competitive antagonist binding sites which are located between subunits near the membrane surface. The receptor ion channel is formed by five transmembrane helices (M2) of all five subunits. The position of the binding site for noncompetitive ion channel blockers was found by photoaffinity labelling and site-directed mutagenesis. The intrinsic feature of the receptor structure is the position of the agonist/competitive antagonist binding sites in close vicinity to the ion channel spanning the bilayer membrane. This peculiarity may substantially enhance allosteric transitions transforming the ligand binding into the channel opening and physiological response. Muscle nicotinic acetylcholine receptors from birds and mammals are also pentaoligomers consisting of four different subunits (alpha2beta gammadelta or alpha2beta epsilondelta) with high homology to the Torpedo receptor. Apparently, the pentaoligomeric structure is the main feature of all nicotinic, both muscle and neuronal, receptors. However, the neuronal receptors are formed only by two subunit types (alpha and beta) or are even pentahomomers (alpha7 neuronal receptors). All nicotinic receptors are ligand-gated ion channel, the properties of the channels being essentially determined by amino acid residues forming M2 transmembrane fragments.     

Direct photoaffinity labeling of leukotriene binding sites

Due to their conjugated double bonds the leukotrienes themselves are photolabile compounds and may therefore be used directly for photoaffinity labeling of leukotriene binding sites. Cryofixation eliminates unspecific labeling taking place in solution by photoisomers and photodegradation products of leukotrienes. After fixation of receptor ligand interactions by shock-freezing of the samples, irradiation-induced highly reactive excited states and/or intermediates can form covalent bonds with the respective binding site in the frozen state. After cryofixation of a solution of albumin incubated with [3H8]leukotriene E4, irradiation at 300 nm resulted in time-dependent incorporation of radioactivity into the protein. Photoaffinity labeling of rat as well as of human blood serum with [3H8]leukotriene E4 after cryofixation revealed that only one polypeptide with an Mr of 67,000 was labeled. This polypeptide was identified as albumin. Photoaffinity labeling of rat liver membrane subfractions enriched with sinusoidal membranes resulted in the labeling of a polypeptide with an apparent Mr of 48,000, whereas no polypeptide was predominantly labeled in the subfraction enriched with canalicular membranes. Photoaffinity labeling of isolated hepatocytes disclosed different leukotriene E4 binding polypeptides. In the particulate fraction of hepatocytes a polypeptide with an apparent Mr of 48,000 was labeled predominantly, whereas in the soluble fraction several polypeptides were labeled to a similar extent. One of these, with an apparent Mr of 25,000, was identified as subunit 1 of glutathione transferases by immunoprecipitation. The method of direct photoaffinity labeling in the frozen state after cryofixation using leukotrienes as photoactivatable compounds, as exemplified by leukotriene E4, may be most useful for the identification and characterization of various leukotriene binding sites, including receptors, leukotriene-metabolizing enzymes, and transport systems.     

Endogenous low molecular weight inhibitors of cholinergic binding sites

Endogenous low molecular weight compounds which inhibit ligand binding to nicotinic acetylcholine receptors of neuronal membranes have been isolated from insect nervous tissue. Two distinct heat-stable, cationic inhibitory compounds with molecular weights of about 700-500 Da and below 500 Da have been identified. The active material was found to competitively inhibit [¹²⁵I]α-bungarotoxin and [³H]acetylcholine binding in a reversible, dose dependent manner. Comparative binding studies revealed that the active material also inhibits [¹²⁵I]α-bungarotoxin and [³H]acetylcholine binding in vertebrate brain, but not in the electric tissue of Torpedo. These results suggest that the endogenous inhibitors may function as modulators specific for neuronal acetylcholine receptors.     

Azido auxins : photolysis in solution and covalent binding to soybean

The potential of three auxin analogs, 4-, 5-, and 6-azidoindole-3-acetic (4-N₃IAA, 5-N₃IAA, and 6-N₃IAA), as photoaffinity labeling agents for the detection and isolation of auxin receptors was assessed by irradiating these compounds at 365 nm on TLC plates, in solution, and in contact with soybean (Glycine max L. Merr. var. Wayne) hypocotyl. Photolysis on TLC plates produces immobile spots, indicating extremely polar or covalent binding of the photoproducts to the plates. On irradiation in buffer or in buffer containing sucrose, all three compounds decompose at rates that are first order in N₃IAA to give fluorescent solutions. Photolysis through a Pyrex filter is slower than that through quartz, but the filter prevents tissue damage and allows a given dose of irradiation to photolyze all three N₃IAAs to the same extent. The effects of photolysis of these compounds in vivo were evaluated with a straight growth assay using etiolated soybean hypocotyl segments. According to this assay, the photoproducts of the N₃IAAs possess little auxin activity. Irradiation of soybean hypocotyl tissue after 1-hour exposure to 4-N₃IAA in the dark causes the tissue to grow during 12 hours as much as tissue that is continuously exposed to 4-N₃IAA in the dark for this period, suggesting that, on photolysis, this auxin analog binds irreversibly to an auxinsensitive site. Although the fluorescence intensity of the photolyzed N₃IAAs is weak enough to require another method of detecting the bound analog under physiological conditions, the evidence for covalent binding of the N₃IAAs on photolysis implies that these compounds will be satisfactory photoaffinity labeling agents.     

Actions of the insecticide 2(nitromethylene)tetrahydro-1,3-thiazine on insect and vertebrate nicotinic acetylcholine receptors

The nitromethylene heterocyclic compound 2(nitromethylene)tetrahydro)1,3-thiazine (NMTHT) inhibits the binding of [125I]alpha-bungarotoxin to membranes prepared from cockroach (Periplaneta americana) nerve cord and fish (Torpedo californica) electric organ. Electrophysiological studies on the cockroach fast coxal depressor motorneuron (Df) reveal a dose-dependent depolarization in response to bath-applied NMTHT. Responses to ionophoretic application of NMTHT onto the cell-body membrane of motorneuron Df are suppressed by bath-applied mecamylamine (1.0 x 10(-4) M) and alpha-bungarotoxin (1.0 x 10(-7) M). These findings, together with the detection of a reversal potential close to that estimated for acetylcholine, provide evidence for an agonist action of this nitromethylene on an insect neuronal nicotinic acetylcholine receptor. The binding of [3H]H12-histrionicotoxin to Torpedo membranes was enhanced in the presence of NMTHT indicating an agonist action at this vertebrate peripheral nicotinic acetylcholine receptor. NMTHT is ineffective in radioligand binding assays for rat brain GABAA receptors, rat brain L-glutamate receptors and insect (Musca domestica) L-glutamate receptors. Partial block of rat brain muscarinic acetylcholine receptors is detected at millimolar concentrations of NMTHT. Thus nitromethylenes appear to exhibit selectivity for acetylcholine receptors and exhibit an agonist action at nicotinic acetylcholine receptors.     

Mutational analysis of ligand-induced activation of the Torpedo acetylcholine receptor.

A number of studies have demonstrated that a major portion of the ligand binding site of the Torpedo nicotinic acetylcholine receptor is near cysteines 192 and 193 of the alpha subunit. The role of conserved tyrosine and aspartate residues within this region in ligand binding and receptor activation was investigated using a combination of site-directed mutagenesis and expression in Xenopus oocytes. Wild-type receptors are half-maximally activated (K1/2) by 20 microM acetylcholine with a Hill coefficient, n, of 1.9. Substitution of alpha Y190 and alpha Y198 with phenylalanines (alpha Y190F, alpha Y198F) or alpha D200 with asparagine (alpha D200N) shifts the K1/2 to 408, 117, and 75 microM, respectively, with no effect on the Hill coefficient. To further study the effects of these mutations on activation, the responses of the receptors to the partial agonists phenyltrimethylammonium (PTMA) and tetramethylammonium (TMA) were examined. Wild-type receptors are half-maximally activated by 73 microM PTMA and 2 mM TMA. In contrast, alpha Y190F, alpha Y198F, and alpha D200N receptors are not activated by PTMA and TMA by concentrations of up to 500 microM or 5 mM, respectively. However, PTMA and TMA do act as competitive antagonists of the mutant receptors, an indication that the binding of these compounds is not abolished by these mutations. Comparison of the the Ki values for TMA and PTMA inhibition with the K 1/2 values for TMA and PTMA activation of wild-type receptors indicates that the affinities of these compounds are similar in wild-type and mutant receptors. Therefore, alpha Y190F, alpha Y198F, and alpha D200N mutations do not significantly alter the affinity of the ligand binding site; rather, these mutations appear to interfere with the coupling of ligand binding to channel opening.     

Direct and energy-transfer photolabelling of brain muscarinic acetylcholine receptors

Efficient photolabelling of muscarinic acetylcholine receptors was obtained using either two aryldiazonium salts or an azido derivative. These probes did not discriminate between muscarinic binding subtypes or affinity states and became irreversibly bound to the receptor sites, in an entirely atropine-protectable manner, upon ultraviolet irradiation. The extent of labelling was dependent both on probe concentration and on time of irradiation and reached up to 80% of the receptor population, under optimal alkylating conditions. In contrast to the azido derivative, both diazonium salts behave as potent irreversible labels of muscarinic receptors, provided energy-transfer photolabelling conditions were followed. Such an indirect activation of diazonium ligands, through an energy transfer from photoexcited tryptophan residues, has been previously found to increase the site-specificity and the rate of labelling of other acetylcholine binding proteins. Analogies in the photolabelling process of acetylcholinesterase or of nicotinic and muscarinic receptors by the two diazonium salts are discussed. Altogether, these findings suggest that these new probes may be promising tools to investigate the location and the topography of the agonist-antagonist binding domain on purified muscarinic receptors, through amino acid and/or sequence analyses of radioactive, photolabelled residues.     

Nicotinic acetylcholine receptor probed with a photoactivatable agonist: improved labeling specificity by addition of CeIV/glutathione. Extension to laser flash photolabeling.

The molecular structure of Torpedo marmorata acetylcholine binding sites has been investigated previously by photoaffinity labeling. However, besides the nicotine molecule [Middleton et al. (1991) Biochemistry 30, 6987-6997], all other photosensitive probes used for this purpose interacted only with closed receptor states. In the perspective of mapping the functional activated state, we synthesized and developed a new photoactivatable agonist of nAChR capable of alkylation of the acetylcholine (ACh) binding sites, as reported previously [Kotzyba-Hibert et al. (1997) Bioconjugate Chem. 8, 472-480]. Here, we describe the setup of experimental conditions that were made in order to optimize the photolabeling reaction and in particular its specificity. We found that subsequent addition of the oxidant ceric ion (CeIV) and reduced glutathione before the photolabeling step lowered considerably nonspecific labeling (over 90% protection with d-tubocurarine) without affecting the binding properties of the ACh binding sites. As a consequence, irradiation at 360 nm for 20 min in these new conditions gave satisfactory coupling yields (7.5%). A general mechanism was proposed to explain the successive reactions occurring and their drastic effect on the specificity of the labeling reaction. Last, these incubation conditions can be extended to nanosecond pulsed laser photolysis leading to the same specific photoincorporation as for usual irradiations (8.5% coupling yield of ACh binding sites, 77% protection with carbamylcholine). Laser flash photocoupling of a diazocyclohexadienoyl probe on nAChR was achieved for the first time. Taken together, these data indicate that future investigation of the molecular dynamics of allosteric transitions occurring at the activated ACh binding sites should be possible.     

Synthesis and binding properties of specific photoaffinity ligands for mu and delta opioid receptor subtypes

We report the synthesis and binding properties of specific photoaffinity ligands for mu and delta opioid receptor subtypes. These ligands are derived from DAGO: Tyr-D-Ala-Gly-NMePhe-Gly-ol, a mu selective probe and DTLET: Tyr-D-Thr-Gly-Phe-Leu-Thr, a delta selective probe by modifying the Phe 4 residue. These modifications are: i) a nitro group on the para position of Phe ring as Phe(4 NO2) or Nip, ii) an azido group as Phe(4 N3) or AZ. Pharmacological responses on mouse vas deferens (delta sites) and guinea pig ileum (mu sites), as well as competition experiments with [3H] DAGO and [3H] DTLET on crude rat brain membranes have been performed. The nitro group on the phenyl ring of the Phe residue preserves the affinity and selectivity of each probe: NipDAGO for the mu sites, NipDTLET for the delta ones. However the nitro probes do not appear to be photoactivable by u.v. irradiation. Likewise, azidation of the phenyl ring of the Phe residue does not change the receptor selectivity of each probe, but AZDAGO has less affinity than its parent molecule DAGO, while AZDTLET has more affinity than DTLET. These compounds are photoactivable and provide an efficient tool to characterize and isolate the different receptor subtypes, especially the delta site.