The surface of alpha-subunit loop 1 distant from the subunit interface is exposed in the hCG lutropin receptor complex

Interactions of the placental glycoprotein hormone human choriogonadotropin (hCG) with lutropin receptors (LHR) are required for maintenance of early pregnancy. Knowledge of how hCG interacts with LHR is useful for understanding the mechanism of receptor function, an issue of considerable debate. A large surface of hCG remains exposed after the hormone binds the LHR and can be readily detected with monoclonal antibodies. Here we show that the surface of hCG alpha-subunit loop 1 furthest from the beta-subunit interface can also be recognized by a monoclonal antibody when hCG is bound to the LHR. This extends the area of hCG known to be exposed in the hormone receptor complex, an observation that further restricts models of hCG-LHR interaction.     

Identification of the retinal cyclic GMP phosphodiesterase inhibitory gamma-subunit interaction sites on the catalytic alpha-subunit.

Retinal rod outer segment phosphodiesterase (PDE) consists of two similar catalytic subunits (alpha and beta) and two identical inhibitory subunits (gamma 2). A trypsin-activated soluble PDE exhibiting the ability to be reinhibited by PDE gamma was shown by peptide antisera to retain both N and C termini. Synthetic peptides corresponding to residues 16-30, 78-90, 389-403, and 535-563 of PDE alpha used in a PDE activity assay with trypsin-activated PDE partially prevented inhibition by exogenous PDE gamma; however, only competitions by peptides 16-30 and 78-90 (corresponding to PDE alpha 16-30 and 78-90) were concentration-dependent below 100 nmol of peptide. Binding studies using radio-immunoassays and PDE alpha peptides confirmed that peptides 16-30 and 78-90 (corresponding to PDE alpha 16-30 and 78-90, respectively) were able to bind PDE gamma. Additionally, peptides corresponding to the PDE alpha region 453-534 bound PDE gamma in the binding assay. This suggests that several regions on PDE alpha interact with the PDE gamma inhibitor. While some regions may be involved in binding to PDE gamma, other sites may be involved in PDE gamma inhibition of catalytic activity. Our results suggest that the major regions of PDE alpha that interact with PDE gamma reside within the N terminus (16-30 and 78-90), with weaker interaction regions within or near the hypothesized catalytic domain (453-563). Sequence analysis of three retinal phosphodiesterases (rod outer segment alpha, beta, and cone outer segment alpha') revealed the highest region of dissimilarity in the N and C termini.     

Differential sorting of lutropin and the free alpha-subunit in cultured bovine pituitary cells.

The glycoprotein hormones lutropin (LH) and follitropin (FSH) are both synthesized by gonadotrophs in the anterior pituitary but are stored in separate secretory granules prior to secretion. Despite having highly homologous beta-subunits and alpha-subunits with the identical amino acid sequence, the Asn-linked oligosaccharides on LH terminate with SO4-GalNAc while those on FSH terminate with sialic acid-Gal. In addition to LH and FSH, gonadotrophs secrete uncombined (free) alpha-subunit which bears the same sulfated oligosaccharides as LH. We have examined the synthesis and secretion of LH and free alpha-subunit in primary cultures of bovine pituitary cells in order to determine if the sulfated oligosaccharides have any impact on sorting. Our results show that newly synthesized free alpha-subunit is secreted exclusively via the constitutive pathway with a t1/2 of 1.8 h and is never found in dense-core secretory granules. In contrast, LH dimer is secreted by both the constitutive and the regulated pathways. Constitutive secretion and arrival in a dense secretory granule both occur with t1/2 values of 1-1.5 h for newly synthesized LH. Sulfation occurs immediately prior to arrival of LH in the secretory granule and is followed by a period of 1-1.5 h before the LH-containing granules become sensitive to release by gonadotropin releasing hormone. As a result the t1/2 for LH secretion in the presence of gonadotropin releasing hormone is 3.5 h. Sulfation of the free alpha-subunit oligosaccharides is not, therefore, sufficient to direct the alpha-subunit to secretory granules, and the information required for directing LH to granules must reside either in the beta-subunit or the alpha beta-complex.     

Effects of beta subunit acylation on lutropin receptor site binding.

The effects of various modifications on the beta subunit of lutropin have been studied using the binding characteristics of the reconstituted hormone in the rat testicular radioligand assay. Conditions for iodinating lutropin and lutropin derivatives were determined which resulted in 15 per cent specific binding when tested immediately and retention of 6 to 7 per cent specific binding even after storage for 6 months. Acetimidinyl, acetyl, and carbamyl derivatives of the beta subunit were prepared and combined with unmodified alpha subunit to form reconstituted lutropin. Modification of the beta subunit was shown to have no effect on the time course of binding to testicular receptors or, with one exception, on the extent of receptor saturation. Very high concentrations of lutropin reconstituted with acetylated beta subunit showed an anomalous binding behavior. Scatchard plots of the binding data support the view that the native hormone has a unique receptor affinity which is irreversibly disrupted by separation of subunits and that derivatization of the beta subunit does not alter this parameter further. These data also suggest that there are no significant differences in the amino groups modified on the beta subunit. Competition and preincubation tests for receptor sites that reacted only with modified lutropin and not with the native hormone were negative.     

Characterization of cleavage products in selected human lutropin preparations. A protease-sensitive site in human lutropin beta subunit

Low molecular weight fragments derived from the beta subunit of human lutropin have been frequently observed. These fragments are detected by polyacrylamide gel electrophoresis in sodium dodecyl sulfate following reduction of the disulfide bonds. A sample of human lutropin was identified that had a major portion of its beta subunit showing this proteolytic nick. Over 83% of the subunit was nicked based on reduction, carboxymethylation, and isolation of the low molecular weight fragments. This preparation had 53% of the activity of an intact human lutropin (radioligand assay). The proteolytic nick in the subunit was shown by N-terminal sequencing of the C-terminal fragments to be derived from three clips in a hexapeptide region (residues 44-49) characterized by hydrophobic alkyl side chains. Specific clips were on the amino side of Leu-45 (8%), Val-48 (45%) and Leu-49 (47%). Thus the proteolytic activity, presumably derived from the pituitary during processing, has a substrate specificity reminiscent of the bacterial protease, thermolysin.     

The amino acid sequence of the rabbit lutropin beta subunit

The amino acid sequence of the beta subunit of rabbit lutropin (lLH) has been determined. The amino terminus of about 97% of the beta subunit has a two amino acid extension (pyro-Glu-Pro) compared to other lutropin beta sequences. Overlapping peptides from trypsin and chymotrypsin digestions of the performic acid-oxidized beta subunit and trypsin digestion of the S-aminoethylated cysteine beta subunit were isolated by chromatography on TSK Fractogel 40F and high-pressure liquid chromatography (HPLC). Sequencing was by a combination of the dansyl-Edman method and the direct Edman method. Amide placements were established by HPLC analysis of the PTH amino acid derivatives. The proposed sequence of lLH subunit is: This sequence is highly homologous to the other known lutropin beta subunits, especially rat and pig lutropin beta (91%). Partial cleavage of the peptide bond between Asp-79 and Pro-80 was observed during cyanogen bromide treatment. Rabbit thyrotropin and thyrotropin beta subunit copurified with lLH and lLH except at a final chromatography on Sephadex G-100.     

A carboxyl-terminal sequence in the lutropin beta subunit contributes to the sorting of lutropin to the regulated pathway

Although synthesized in the same pituitary gonadotropes, the secretion profiles of lutropin (LH) and follitropin (FSH) differ. LH is secreted through a regulated pathway and associated with a bolus release at mid-estrous cycle. In contrast, the majority of FSH is secreted constitutively with an incremental increase until ovulation. Both share an identicalalpha subunit, and thus thebeta subunit contains determinants for sorting into the regulated pathway. Previously, we demonstrated that a hydrophobic carboxyl-terminal heptapeptide of the LHbeta subunit (Leu-Ser-Gly-Leu-Leu-Phe-Leu), not found in the FSHbeta subunit, influences the intracellular behavior of the LH dimer. To test the hypothesis that the peptide contributes to differential sorting, we monitored the fates of LH and LHDeltaT (LHbeta subunit lacking the carboxyl-terminal seven amino acids) dimers in the rat somatotrope-derived GH(3) cell line in which both the regulated and constitutive secretory pathways operate. Pulse-chase labeling demonstrated that the LHDeltaT dimer was diverted to the constitutive pathway, resulting in a significant decrease in the corresponding intracellular pool. Forskolin stimulated LH dimer release 3-fold, which was accompanied by a parallel decrease of intracellular LH; only marginal forskolin stimulation of LHDeltaT was seen. Immunofluorescence after cycloheximide treatment demonstrated decreased retention of LHDeltaT compared with LH, consistent with increased constitutive secretion of LHDeltaT. We also demonstrated that fusing the heptapeptide to the carboxyl terminus of the FSHbeta subunit resulted in an increased regulated secretion of this FSH analog compared with wild-type FSH. These data are the first to identify a novel structural determinant responsible for the sorting of a member of the glycoprotein hormone family into the regulated secretory pathway.     

Location of ligand-binding sites on the nicotinic acetylcholine receptor alpha-subunit

The portions of the Torpedo californica nicotinic acetylcholine receptor (AChR) alpha-subunit that contribute to the allosteric antagonist-binding site and to the agonist-binding site have been localized by affinity labeling and proteolytic mapping. [3H]Meproadifen mustard was employed as an affinity label for the allosteric antagonist-binding site and [3H]tubocurare as a photoaffinity label for the agonist-binding site. Both labels were found in a 20-kDa proteolytic fragment generated from the AChR alpha-subunit by Staphylococcus aureus V8 protease. This 20-kDa peptide also contains the 3H-labeled 4-(N-maleimido)-alpha-benzyltrimethylammonium iodide-reactive site and binds 125I-alpha-bungarotoxin. N-terminal sequencing established that the 20-kDa fragment began at Ser-173 of the alpha-subunit. Fluorescein isothiocyanate-conjugated concanavalin A could be bound to the second of the two major V8 cleavage products, an 18-kDa peptide. This peptide was also sensitive to treatment with endo-beta-N-acetyl-glucosaminidase H, consistent with the presence of N-linked carbohydrate on this fragment. The N terminus of this peptide was found to be Val-46 of the alpha-subunit sequence. Experiments designed to map disulfide bonds within the AChR alpha-subunit indicate that no bonds exist between the 18-kDa fragment (containing Cys-128 and Cys-142) and the 20-kDa fragment (containing Cys-192, Cys-193, and Cys-222). These results establish that the 20-kDa fragment contributes to both the acetylcholine and the allosteric antagonist-binding sites, whereas there is no evidence that the 18-kDa fragment is part of either site.     

Potential contact sites between the protein and RNA subunit in the Bacillus subtilis RNase P holoenzyme.

We have detected by nucleotide analog interference mapping (NAIM) AMPalphaS and IMPalphaS modifications in Bacillus subtilis RNase P RNA that interfere with binding of the homologous protein subunit. Interference as well as some enhancement effects were clustered in two main areas, in P10.1a/L10.1 and P12 of the specificity domain (cluster 1, domain I) and in P2, P3, P15.1, J18/2 and J19/4 of the catalytic domain (cluster 2a, domain II). Minor interferences in P1 and P19 and a strong and weak enhancement effect in P19 represent a third area located in domain II (cluster 2b). Our results suggest that P3, P2-J18/2 and J19/4 are key elements for anchoring of the protein to the catalytic domain close to the scissile phosphodiester in enzyme-substrate complexes. Sites of interference or enhancement in clusters 1 and 2a are located at distances between 65 and 130 A from each other in the current 3D model of a full-length RNase P RNA-substrate complex. Taking into account that the RNase P protein monomer can bridge a maximum distance of about 40 A, simultaneous direct contacts to the two aforementioned potential RNA-binding areas would be incompatible with our current understanding of bacterial RNase P RNA architecture. Our findings suggest that the current 3D model has to be rearranged in order to reduce the distance between clusters 1 and 2a. Alternatively, based on the recent finding that B. subtilis RNase P forms a tetramer consisting of two protein and two RNA subunits, cluster 1 may reflect one protein contact site in domain I, and cluster 2a a separate one in domain II.     

Intrinsic and extrinsic fluorescence probes of subunit interactions in ovine lutropin.

The fluorescence of 1,8-anilinonaphthalene sulfonate (ANS) was enhanced in the presence of ovine lutropin (oLH). Fluorescence titration curves were sigmoidal with 50% saturation between 200 and 500 μm. Exclusion chromatography experiments indicated that the hormone self-associates to form dimers in the presence of excess ANS. By contrast, the isolated a and β subunits of oLH caused a much smaller enhancement of the fluorescence of ANS and did not self-associate in its presence. Dissociation of the intact hormone into its subunits was accompanied by 1) a loss in the ability to enhance ANS fluorescence, 2) the appearance of a negative differential absorption spectrum whose magnitude indicated the increased solvent-exposure of at least two tyrosines, and 3) a loss in conformational rigidity as evidenced by a decrease in polarization (P) of tyrosyl fluorescence from ~0.17 to ~0.13. Similar rates of dissociation were obtained by all three measurements and the first order rate constant at pH 3.6 and 37 °C under conditions of low ionic strength was k = 0.18 min^?1; at high ionic strength, e.g., 0.5 m KC1, dissociation was incomplete even after prolonged incubation. Acid-dissociated subunits recombined readily in 0.5 m acetate buffer, pH 5.3, and the recovery of the intrinsic absorption and fluorescence properties as well as the ability to enhance ANS fluorescence ranged between 70 and 90%. Titration of the isolated α and β subunits with acid or GdmCl had little or no effect on P, suggesting that residual secondary or tertiary structure is either absent, very stable, or its disruption does not alter the rigidity of the tyrosyl environment. The relatively high P for oLH-β (0.17) suggests a conformation which is rigid compared with oLH-α (0.13). P for both subunits decreased smoothly with increasing temperature between 20 and 70 °C. By contrast, oLH exhibited a thermal transition near 50 °C characterized by a drop in P from a value near that of β to a value near that of a as the subunits dissociated. Because α has more tyrosines with a higher average quantum yield, its fluorescence would be expected to dominate that of the hormone or of an equimolar mixture of subunits. Thus, most of the conformation changes which accompany dissociation and recombination appear to occur in the α subunit.     

Identification and characterization of two cation binding sites in the integrin beta 3 subunit.

The midsegment of the beta(3) subunit has been implicated in the ligand and cation binding functions of the beta(3) integrins. This region may contain a metal ion-dependent adhesion site (MIDAS) and fold into an I domain-like structure. Two recombinant fragments, beta(3)-(95-373) and beta(3)-(95-301), were expressed and found to bind fibrinogen. Whereas 0.1 mm Ca(2+) supported ligand binding to both recombinant fragments, 1.0 mm Ca(2+) suppressed binding to the longer but not the shorter fragment. These properties suggest that beta(3)-(95-373) contains both the ligand-competent (LC) and inhibitory (I) cation binding sites, and beta(3)-(95-301) lacks the I site. In equilibrium dialysis experiments, beta(3)-(95-373) contained two divalent cation binding sites, one reactive with either Mg(2+) or Ca(2+) and one Ca(2+)-specific, whereas beta(3)-(95-301) lacked the Ca(2+)-specific site. Mutant forms of beta(3)-(95-373) suggested that the LC site is a MIDAS motif involving Asp(119), Ser(121), Ser(123), Asp(217), and/or Glu(220) as coordination sites, and the I site was dependent upon residues within beta(3)-(301-323). In a molecular model of beta(3)-(95-373), a second Ca(2+) could be docked onto a flexible loop in close proximity to the MIDAS. These results indicate that the ligand competent and Ca(2+)-specific inhibitory cation binding sites are distinct and reside in beta(3)-(95-373).     

Amphibian lutropin from the bullfrog Rana catesbeiana. Complete amino acid sequence of the beta subunit.

The amino acid sequence of lutropin (LH) beta subunit of an amphibian, the bullfrog Rana catesbeiana, has been determined. The primary structure was determined by sequencing the intact protein (residues 1-44) and peptides originated by cyanogen bromide cleavage and lysyl endopeptidase digestion. 12 cysteine residues are conserved in the bullfrog and mammalian LH beta subunit. One sugar-chain-binding site at Asn-8 is also conserved in the bullfrog and in all mammals except humans. This glycoprotein is composed of 112 amino acid residues with a molecular mass of 12675 Da, considering the six cystine bridges and excepting the sugar chain. The bullfrog beta subunit has approximately 50% sequence identity with that of mammals and with the fish gonadotropin beta subunit, and about 40% with bullfrog follicle-stimulating hormone beta subunit.     

Use of protein knobs to characterize the position of conserved alpha-subunit regions in lutropin receptor complexes

Efforts to identify the manner in which human choriogonadotropin (hCG) contacts lutropin receptors (LHR) have been stymied by the complex structure of the hormone and the likelihood that it contacts the receptor at multiple sites. During studies of hCG assembly in mammalian cells, we found that addition of a cysteine to the long disordered beta-subunit COOH terminus (betaCT) enabled it to become cross-linked by a disulfide to cysteines that are substituted for residues in loop alpha2 or in the alpha-subunit COOH terminus (alphaCT). This created a "knob" on the alpha-subunit at the location of the cysteine. Knobs of various sizes and charges were useful for probing surfaces of the alpha-subunit thought previously to contact the LHR. Attachment of the betaCT to residues in loop alpha2 facing loops beta1 and beta3 reduced hormone activity only a few fold revealing that this surface does not participate in essential high affinity receptor contacts, a finding inconsistent with our earlier view of the hCG-LHR complex. In contrast, this approach showed that the opposite surface of loop alpha2 appeared to be nearer the receptor interface. Although attachment of knobs to portions of the alphaCT reduced hormone activity substantially, this finding was difficult to interpret. As discussed, this procedure should be adapted readily to other proteins and may facilitate the introduction of fluorophores, enzymes, or other reagents at specific sites on protein surfaces. It may also permit one to cross-link proteins or to obscure specific protein surfaces during the development of "Trojan Horse" therapeutics.     

Identification of protein contact sites within the glucocorticoid/progestin response element

The glucocorticoid receptor (GR) and the progestin receptor (PR) bind specifically to a variety of DNA sequences, glucocorticoid/progestin response elements (GRE/PRE), located in the proximity of responsive gene promoters. Using the isolated recombinant GR DNA-binding domain (DBD), it has recently been shown that GR interacts with the GRE/PRE, a 15-basepair partially palindromic consensus sequence, as a dimer. In this study an investigation into the GR-GRE/PRE and PR-GRE/PRE interaction has been performed using missing base contact analysis with the tyrosine aminotransferase GREII (TATII) and recombinant GR DBD as well as a fusion protein consisting of the PR DBD fused to Staph. aureus protein-A. GR and PR had identical base contact points, localized within two consecutive major grooves, binding to the same face of the DNA. Ethylation interference was also performed on the GR DBD-TATII interaction. The contact points with the backbone phosphate groups flank the contacts within the major groove for each of the two half-sites. Knowledge of the contact points within the DNA sequence together with the three-dimensional structure of the protein enables modelling of the protein-DNA interaction.     

Photoaffinity labeling of the indole sites on the Escherichia coli tryptophan synthase alpha-subunit

The alpha subunit of the Escherichia coli tryptophan synthase catalyzes the reversible aldolytic reaction: Indole-3-glycerol phosphate in equilibrium indole + glyceraldehyde 3-phosphate. The use of 5-azidoindole as a photoaffinity label has made the generation of a number of enzyme-substrate complexes possible, each with a given degree of saturation of the two postulated indole sites. When assayed in the reverse reaction (indole-3-glycerol phosphate synthesis), samples of alpha subunit treated at concentrations of 5-azidoindole less than or equal to 2 mM show a progressive 30-40% activation. A gradual inactivation occurs only in samples irradiated at concentrations in excess of 2 mM 5-azidoindole, and this inactivation is complete at 8-10 mM. A quantitatively similar activation occurs in the forward reaction (indole synthesis), however inactivation in this case is incomplete, with complexes treated at 8-12 mM 5-azidoindole retaining 30-40% relative activity in this reaction. When treated alpha subunits were assayed for their abilities to complement the beta 2-subunit in the reactions indole + L-serine leads to L-tryptophan + H2O and indole-3-glycerol phosphate + L-serine leads to L-tryptophan + glyceraldehyde 3-phosphate, quantitatively lesser amounts of activation followed by total inactivation are observed over a similar range of 5-azidoindole concentrations.     

Amphibian lutropin and follitropin from the bullfrog Rana catesbeiana. Complete amino acid sequence of the alpha subunit.

The amino acid sequence of the alpha-subunit of the gonadotropins, lutropin and follitropin from bullfrog, Rana catesbeiana, has been determined. The alpha-subunit was identified in both hormones by the amino acid composition and ovulation activity of lutropin in the Xenopus ovary, by means of reconstituted hormones in various combinations. The amino acid sequences of two identical alpha-subunits from lutropin and follitropin were determined or deduced by different strategies. The alpha-subunit of those gonadotropins have 97 amino acid residues, the longest among the known alpha-subunits of gonadotropins, and one arginine insertion at position 29. Ten cysteine residues and two sugar-chain binding sites at Asn57 and Asn83 are completely conserved among the species. The molecular mass of this subunit is 11,026 Da not including the sugar chains. The bullfrog alpha-subunit has approximately 70% sequence identity with mammalian alpha-subunits.     

Identification of the sites phosphorylated by cyclic AMP-dependent protein kinase on the beta 2 subunit of L-type voltage-dependent calcium channels.

Voltage-dependent L-type calcium (Ca) channels are heteromultimeric proteins that are regulated through phosphorylation by cAMP-dependent protein kinase (PKA). We demonstrated that the beta 2 subunit was a substrate for PKA in intact cardiac myocytes through back-phosphorylation experiments. In addition, a heterologously expressed rat beta 2a subunit was phosphorylated at two sites in vitro by purified PKA. This beta 2a subunit contains two potential consensus sites for PKA-mediated phosphorylation at Thr164 and Ser591. However, upon mutation of both of these residues to alanines, the beta 2a subunit remained a good substrate for PKA. The actual sites of phosphorylation on the beta 2a subunit were identified by phosphopeptide mapping and microsequencing. Phosphopeptide maps of a bacterially expressed beta 2a subunit demonstrated that this subunit was phosphorylated similarly to the beta 2 subunit isolated from heart tissue and that the phosphorylation sites were contained in the unique C-terminal region. Microsequencing identified three serine residues, each of which conformed to loose consensus sites for PKA-mediated phosphorylation. Mutation of these residues to alanines resulted in the loss of the PKA-mediated phosphorylation of the beta 2a subunit. The results suggest that phosphorylation of the beta 2a subunit by PKA occurs at three loose consensus sites for PKA in the C-terminus and not at either of the two strong consensus sites for PKA. The results also highlight the danger of assuming that consensus sites represent actual sites of phosphorylation. The actual sites of PKA-mediated phosphorylation are conserved in most beta 2 subunit isoforms and thus represent potential sites for regulation of channel activity. The sites phosphorylated by PKA are not substrates for protein kinase C (PKC), as the mutated beta 2 subunits lacking PKA sites remained good substrates for PKC.