The evolution of gonadotropins: some molecular data concerning a non-mammalian pituitary gonadotropin, the hormone from a teleost fish (Cyprinus carpio L.)

The amino acid and sugar compositions as well as long N-terminal sequences and the C-terminal amino acids of the two subunits of carp gonadotropin, SU I and SU II, were determined. An important homology was demonstrated between SU I and α-subunits and betwen SU II and β-subunits of mammalian gonadotropins. Moreover SU II was more closely related to the β-subunit of LH than to the β-subunit of FSH.     

Biological and immunological properties of zebra pituitary gonadotropins: comparison with horse and donkey gonadotropins

Previous studies from this laboratory have described the properties of purified luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from horse and donkey anterior pituitary glands. The present study afforded the opportunity to further characterize these previously purified hormone preparations and to compare them with enriched gonadotropin fractions from zebra pituitary glands. Although a single LH and FSH fraction was usually obtained for each pool of pituitaries, two separate zebra LH and two donkey FSH preparations were generated. Purified hormone preparations from the horse were designated eLH and eFSH. Preparations zLH-A, zLH-B, and zFSH were obtained from zebra pituitaries, and fractions dLH, dFSH-A, and dFSH-B were prepared from donkey pituitary glands. These preparations were analyzed by LH and FSH radioimmunoassays (RIAs), radioreceptor assays (RRAs), LH bioassay, and chromatofocusing. Clear immunological differences were observed between equid gonadotropins. Homologous RIAs for eLH and eFSH did not cross-react similarly, or in a parallel fashion, with gonadotropins from the donkey and zebra. In contrast, RIAs capable of assessing LH or FSH in a wide number of species showed all equid gonadotropin preparations to have considerable activity and to produce parallel dilution curves. Relative to eLH (1.00), zLH-A was found to have higher LH bioactivity:LH RIA (2.50), LH RRA:LH RIA (1.42), and LH bioactivity: LH RRA (2.21) activity ratios. The dLH and zLH-B fractions only differed from eLH in LH RRA:LH RIA activity (0.69 and 0.62, respectively). Only LH from the horse possessed clear intrinsic FSH-receptor-binding activity.(ABSTRACT TRUNCATED AT 250 WORDS)     

Investigations into the biosynthetic pathways for classical and ring B-unsaturated oestrogens in equine placental preparations and allantochorionic tissues

In on-going studies of ‘classical’ and ring B-unsaturated oestrogens in equine pregnancy, the products of metabolism of [2,2,4,6,6-²H₅]-testosterone and [16,16,17-²H₃]-5,7-androstadiene-3β,17β-diol with equine placental subcellular preparations and allantochorionic villi have been identified. Using mixtures of unlabelled and [²H]-labelled steroid substrates has allowed the unequivocal identification of metabolites by twin-ion monitoring in gas chromatography–mass spectrometry (GC–MS). Two types of incubation were used: (i) static in vitro and (ii) dynamic in vitro. The latter involved the use of the Oxycell™ cartridge (Integra Bioscience Systems, St Albans, UK) whereby the tissue preparation was continuously supplied with supporting medium plus appropriate cofactors in the presence of uniform oxygenation. [²H₅]-Testosterone was converted into [²H₄]-oestradiol-17β, [²H₄]-oestrone and [²H₃]-6-dehydro-oestradiol-17 in both placental and chorionic villi preparations, but to a greater extent in the latter, confirming the importance of the chorionic villi in oestrogen production in the horse.

On the basis of GC–MS characteristics (M⁺ m/z 477/482 (as O-methyl oxime-trimethyl silyl ether), evidence for 19-hydroxylation of testosterone was found in static incubations, while the presence of a 6-hydroxy-oestradiol-17 was recorded in dynamic incubations (twin peaks in the mass spectrum at m/z 504/507, the molecular ion M⁺). It was not possible to determine the configuration at C-6. The formation of small, but significant, quantities of [²H₄]-17β-dihydroequilin was also shown, and a biosynthetic pathway is proposed.

In static incubations of placental microsomal fractions, the 17β-dihydro forms of both equilin and equilenin were shown to be major metabolites of [²H₃]-5,7-androstadiene-3,17-diol. Using static incubations of chorionic villi, the deuterated substrate was converted into the 17β-dihydro forms of both equilin and equilenin, together with an unidentified metabolite (base peak, m/z 504/506). The isomeric 17-dihydroequilins were also obtained using the dynamic in vitro incubation of equine chorionic villi, together with the 17β-isomer of dihydroequilenin. Confirmation of the identity of 17β-dihydroequilin and 17β-dihydroequilenin was obtained by co-injection of the authentic unlabelled steroids with the phenolic fraction obtained from various incubations. Increases in the peak areas for the non-deuterated steroids (ions at m/z 414 (17β-dihydroequilin) and 412 (17β-dihydroequilenin) (both as bis-trimethyl silyl ether derivatives) were observed. Biosynthetic pathways for formation of the ring B-unsaturated oestrogens from 5,7-androstadiene-3β,17β-diol are proposed.     

Comparative study of subunits of phenylalanyl-tRNA synthetase from Escherichia coli and Thermus thermophilus

FPLC separation of - and β-subunits of phenylalanyl-tRNA synthetases from E. coli MRE-600 and Thermus thermophilus HB8 has been carried out in the presence of urea. Native -subunits of both enzymes were primarily ₂-dimers and tended to aggregate. Most E. coli enzyme β-subunits were monomeric and only a small fraction was represented by β₂-dimers. All thermophilic β-subunits were β-dimers. It was shown that monomers and all forms of homologous subunits had no catalytic activity in tRNA^Phe aminoacylation. For the enzymes and their subunits, titration curves were obtained and isoelectric points were determined. The comparison of the relative surface charges indicated similarity of the surfaces of entire enzymes and the corresponding β-subunits. -Subunits displayed a distinctly different pH dependence of the surface charge. A spatial model of the oligomeric structure and a putative mechanism for its formation are discussed.     

Purification of equine gonadotropins and comparative study of their acid-dissociation and receptor-binding specificity

A method for the simultaneous purification of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from equine pituitaries is briefly described. Different forms of each hormone were obtained. The total yield of LH was 24.2 mg·kg^?1 with a recovery of 22% and the yield of FSH was 26 mg·kg^?1 with a recovery of 34%. The specific activities of both hormones, measured in homologous equine radio-receptor assays are equal to or higher than those of the preparations described so far. In all species studied so far the acid-dissociation curves of LH and FSH are similar; this is an agreement with the view that the binding of the common α-subunit and the specific β-subunits involves polypeptide regions which are identical in both hormones. In contrast, the acid-dissociation pK_a of equine LH was found to be considerably lower (3.9) than that of equine FSH (5.8). The equine gonadotropins exhibit a much lower specificity with receptors of a porcine testicular fraction compared with an equine fraction. Equine LH exhibited a binding activity on FSH receptors from a porcine testicular fraction equal to 20% that of equine FSH instead of only 1% for an equine binding fraction. Similarly, all the equine FSH preparations tested exhibited a five-fold higher binding-activity on porcine LH receptors than on equine LH receptors. In the porcine system, pregnant mare serum gonadotropin behaved like equine LH towards LH and FSH receptors. In contrast, on equine binding fraction, pregnant mare serum gonadotropin was only 4% as active as equine LH and was devoid of FSH activity. All the data we have obtained are consistent with the ‘negative specificity’ model we proposed recently.     

Epizoochory by large herbivores: merging data with models

The dispersal of plant seeds in the fur of large herbivores (epizoochory) is an important but complex long-distance dispersal mechanism. We developed a spatially explicit simulation model of epizoochorous seed dispersal, which was parameterized based on empirical studies of the movement and behaviour of donkeys, and the distribution, seed production, seed accessibility, seed adhesion, and seed retention on donkey fur of selected plant species in a coastal dune nature reserve in Flanders, Belgium. We compared predicted and observed seed numbers of the 14 plant species on donkey fur.

Modelled seed shadows indicate that for most species about half of all seeds dispersed by donkeys should travel a net distance of >100 m, and about 1% should travel >500 m within this more or less isodiametric 100 ha nature reserve. Seeds with longer retention times are expected to travel further than those with short retention times. Enlarging the reserve area had little impact on the forecasted dispersal distances.

Variation among plant species in the observed seed numbers found on donkey fur were well predicted by the model (R²=0.56, P=0.002), though the predictions relied on relatively crude estimates of seed production and accessibility to donkeys, indicating that more accurate estimates of these parameters are needed.

Our model confirms the important role of epizoochory in affecting long-distance seed dispersal, and provides a modelling framework for integrating the multiple components of the dispersal process.     

Structural diversity in the six-fold redundant set of acyl-CoA carboxyltransferases in Mycobacterium tuberculosis

Mycobacterium tuberculosis contains multiple versions of the accA and accD genes that encode the - and β-subunits of at least three distinct multi-functional acyl-CoA carboxylase complexes. Because of its proposed involvement in pathogenic M. tuberculosis survival, the high-resolution crystal structure of the β-subunit gene accD5 product has been determined and reveals a hexameric 356 kDa complex. Analysis of the active site properties of AccD5 and homology models of the other five M. tuberculosis AccD homologues reveals unexpected differences in their surface composition, providing a molecular rational key for a sorting mechanism governing correct acyl-CoA carboxylase holo complex assembly in M. tuberculosis.     

Characterization of a red blood cell antigen in donkeys and mules associated with neonatal isoerythrolysis

A red cell antigen of donkeys and mules was identified using antibodies in serum from a mare which produced a mule foal affected with neonatal isoerythrolysis (NI). Subsequently antibodies with similar activity were identified in the sera of other mares which had produced mule foals and were produced by immunization of horses with blood from donkeys. The antigen detected by these antibodies does not correspond to any recognized horse red cell alloantigen. This may be a xenoantigen since all donkeys (and mules) tested have shared this antigen and all horses tested have lacked the antigen. The results suggest that all mule pregnancies (donkey sire x horse dam) are incompatible with regard to this factor and a potential for neonatal isoerythrolysis exists in all cases.     

Zebra chorionic gonadotropin: partial purification and characterization

Six samples of pregnant zebra (z) serum from the first and second trimesters of pregnancy were analyzed by RIA and shown to have chorionic gonadotropin levels comparable to that of the mare (0.9-5.3 micrograms/ml); first trimester levels in most cases were higher than second trimester levels. A pool of the sera (10 ml) was fractionated by methods previously employed for the purification of equine (e) and donkey (d) chorionic gonadotropin to achieve a concentration of the zebra chorionic gonadotropin (zCG). A yield of 1.0 mg of glycoprotein was obtained. HPLC analysis of the material indicated the content of zCG to be about 7%. Its molecular size as judged by Ve/Vo values is smaller than eCG, greater than ovine LH, and about the same as equine LH. The zCG was tested in RIAs for LH and eCG, radioreceptor assays (RRA) for LH and FSH, and the rat testis Leydig cell assay for LH. Comparisons were made with equine and donkey chorionic gonadotropin, and equine and zebra LH. The results, preliminary because the preparation is not of high purity, showed that zCG is bioactive as an LH; immunologically similar to eCG, eLH, dCG, and zLH; and competes in RRAs for LH but not FSH receptors. It differs, therefore, from eCG and eLH--which have high levels of intrinsic FSH activity, and is more like dCG, dLH, and zLH--all of which have minimal if any FSH activity.     

Two variants in the KIT gene as candidate causative mutations for a dominant white and a white spotting phenotype in the donkey

White spotting phenotypes have been intensively studied in horses, and although similar phenotypes occur in the donkey, little is known about the molecular genetics underlying these patterns in donkeys. White spotting in donkeys can range from only a few white areas to almost complete depigmentation and is characterised by a loss of pigmentation usually progressing from a white spot in the hip area. Completely white‐born donkeys are rare, and the phenotype is characterised by the complete absence of pigment resulting in pink skin and a white coat. A dominant mode of inheritance has been demonstrated for spotting in donkeys. Although the mode of inheritance for the completely white phenotype in donkeys is not clear, the phenotype shows similarities to dominant white in horses. As variants in the KIT gene are known to cause a range of white phenotypes in the horse, we investigated the KIT gene as a potential candidate gene for two phenotypes in the donkey, white spotting and white. A mutation analysis of all 21 KIT exons identified a missense variant in exon 4 (c.662A>C; p.Tyr221Ser) present only in a white‐born donkey. A second variant affecting a splice donor site (c.1978+2T>A) was found exclusively in donkeys with white spotting. Both variants were absent in 24 solid‐coloured controls. To the authors’ knowledge, this is the first study investigating genetic mechanisms underlying white phenotypes in donkeys. Our results suggest that two independent KIT alleles are probably responsible for white spotting and white in donkeys.     

Assembly of an adult type acetylcholine receptor in a mouse cell line transfected with rat muscle ε-subunit DNA

The mouse muscle cell line BC3H-1 expresses an acetylcholine receptor (AChR) composed of -,β-, and δ-subunits [1]. The functional characteristics of this AChR are comparable to the non-synaptic AChR subtype in mouse muscle [2,3]. To investigate the role of the ε-subunit, which is believed to replace the γ-subunit in forming the adult AChR subtype [4], BC3H-1 cells were stably transfected with cDNA encoding the rat muscle AChR ε-subunit. Expression of this cDNA was under the control of a heat shock promoter, and the plasmid carried the neomycin resistance gene for selection. Several clones were isolated that had integrated the plasmid DNA in a stable form and produced ε-subunit specific RNA after heat induction. Single-channel current recording from cells which contained abundant ε-subunit mRNA identified a novel AChR channel having a larger conductance than the native AChR in these cells. These results suggest that the rat muscle ε-subunit may assemble with mouse muscle -, β- and δ-subunits to form a mouse-rat hybrid AChR with properties similar to that of end-plate channels in the mature mammalian neuromuscular synapse. The novel AChR channel appears in the surface membrane within a few hours following the rise in ε-subunit mRNA. Thus, the notion that replacement of the γ-subunit by the ε-subunit during development is the result of the postnatal rise in the level of ε-subunit specific mRNA is further supported.     

-Subunits of Ns are released from the plasma membrane following cholera toxin activation

Cholera toxin (CT) and islet-activating protein (IAP, a Bordetella pertussis toxin) were employed to test the hypothesis that GTP-binding regulatory proteins are released from plasma membranes to a greater extent when ‘activated’ than when ‘inactivated’. CT, which activates N_s (the stimulatory GTP-binding regulatory protein of the adenylate cyclase system), catalyzed the incorporation of radioactivity from [³²P]NAD into 45 and 47.5 kDa peptides associated with rat liver plasma membranes. Following ADP-ribosylation and centrifugation at 100000 × g for 1 h, approx. 30–35% of these CT-labelled peptides were no longer associated with the plasma membranes, but were recovered from the supernatant fraction. IAP, which inactivates N_i (the inhibitory GTP-binding regulatory protein of the adenylate cyclase system) catalyzed the incorporation of radioactivity from [³²P]NAD into a 41 kDa peptide associated with the membranes. However, in contrast to the CT-labelled peptides, typically less than 5% of the lAP-labelled peptide was found in the 100000 × g supernatant fraction, but rather was almost exclusively associated with the membrane pellet. The data indicate that the -subunits of N_s are released from the plasma membrane following activation, and support the hypothesis that the βγ-subunits act to anchor the -subunits to the plasma membrane. Cholera toxin Islet-activating protein GTP-binding protein     

The F1-ATPase beta-subunit is the putative enterostatin receptor

It has been suggested that the F₁-ATPase β-subunit is the enterostatin receptor. We investigated the binding activity of the purified protein with a labeled antagonist, β-casomorphin_1–7, in the absence and presence of cold enterostatin. ¹²⁵I-β-casomorphin_1–7 weakly binds to the rat F₁-ATPase β-subunit. Binding was promoted by low concentrations of cold enterostatin but displaced by higher concentrations. To study the relationship between binding activity and feeding behavior, we examined the ability of a number of enterostatin analogs to affect β-casomorphin_1–7 binding to the F₁-ATPase β-subunit. Peptides that suppressed food intake promoted β-casomorphin_1–7 binding whereas peptides that stimulated food intake or did not affect the food intake displaced β-casomorphin_1–7 binding. Surface plasmon resonance measurements show that the β-subunit of F₁-ATPase binds immobilized enterostatin with a dissociation constant of 150 nM, where no binding could be detected for the assembled F₁-ATPase complex. Western blot analysis showed the F₁-ATPase β-subunit was present on plasma and mitochondrial membranes of rat liver and amygdala. The data provides evidence that the F₁-ATPase β-subunit is the enterostatin receptor and suggests that enterostatin and β-casomorphin_1–7 bind to distinct sites on the protein.     

Alterations in gonadal steroidogenesis in individuals expressing a common genetic variant of luteinizing hormone

Some pathologies of the pituitary-gonadal function have recently been found to be due to mutations of the gonadotropin or gonadotropin receptor genes. Although these conditions are extremely rare, they are very informative, by elucidating some less well characterized facets of normal gonadotropin function and the molecular pathogenesis of disturbances in sexual differentiation and fertility. In contrast, there is a common polymorphism in the Luteinizing Hormone (LH) β-subunit gene, where two point mutations cause two alterations in the amino acid sequence (Trp⁸→Arg and Ile¹⁵→Thr) and introduce an extra glycosylation signal to Asn¹³. The carriers of this variant gene are largely healthy, but certain mild differences in their gonadal function have been found, as reflected by alterations in gonadal steroidogenesis, pubertal development and predisposition to diseases such as infertility, polycystic ovarian syndrome, and breast and prostatic cancer. The purpose of this chapter is to review the current knowledge of the occurrence, special functional features and clinical correlates of this LH variant.     

Chemical, biological and immunological properties of pituitary gonadotropins from the donkey (Equus asinus): comparison with the horse (Equus caballus)

Donkey gonadotropins (donkey luteinizing hormone, dLH; donkey follicle-stimulating hormone, dFSH) have been isolated in purified form from 191 donkey pituitaries using essentially the same procedures previously employed for the purification of equine gonadotropins. Chemically, dLH and dFSH were observed to be similar to equine LH (eLH) and FSH (eFSH) in fractionation behavior and glycoprotein nature. Two forms of the dFSH molecule were observed, as is the case for eFSH. Donkey LH had significantly less total carbohydrate (13.5%) and sialic acid (1.9%) than eLH (26.7% and 5.8%, respectively). Carbohydrate (17-21%) and sialic acid (2.4%) content of the two dFSH preparations closely resembled that of eFSH. A slightly higher tyrosine content in the donkey gonadotropins was noted in a comparison of amino acid compositions. Immunologically, in a heterologous FSH radioimmunoassay (RIA), dFSH preparations were equal to or twice as active as eFSH preparations. However, in homologous RIAs for equine chorionic gonadotropin (eCG), eFSH and eLH, both the dLH and dFSH preparations were considerably less active than the equine gonadotropins, and their inhibition curves were all nonparallel. Biologically, in the Steelman-Pohley assay both dFSH preparations were equipotent and as potent as eFSH (approximately 40 times NIH-FSH-S12). In the Sertoli cell assay for cAMP (FSH assay) and the Leydig cell assay for testosterone (LH assay), both dFSH and dLH were 2- or 6-fold more active than eFSH and eLH, respectively. In rat and equine testis FSH homologous radioreceptor assays, dFSH preparations were as active and up to 6-fold more active than eFSH. In contrast, dLH was 10-fold less active than eLH in the equine LH homologous radioreceptor assay. Unlike eLH, dLH was found to possess little intrinsic FSH activity or FSH inhibitory activity, and the small amount of FSH activity observed was most likely due to FSH contamination. Therefore, eLH behaves much like eCG (pregnant mare's serum gonadotropin, PMSG) which also possesses both LH and FSH activity. In contrast, dLH behaves more like donkey chorionic gonadotropin (dCG) which possesses only a low degree of FSH activity.     

''Cross-talk'' between phospholipase C and adenylyl cyclase involves regulation of G-protein levels in GH3 rat pituitary cells.

We have investigated the possibility that adenylyl cyclase (AC) activity and membrane protein levels of the -subunits of the stimulatory and inhibitory G-proteins of AC (G_s and G_i−2) in cultured prolactin-producing rat pituitary adenoma cells (GH₃ cells) are modulated by phospholipase C (PLC)-generated second messengers. Pretreatment of cells (6–48 h) with ionomycin (1 μM) or 1-oleoyl-2-acetylglycerol (OAG; 1μM) showed that ionomycin regulated G_s levels in a time-dependent, biphasic manner; a two-fold increase followed a 40% initial reduction, while OAG lowered G_s levels by more than 50% at all time-points. G_i−2 levels remained unchanged by both pretreatments. OAG, but not ionomycin, increased basal AC activity without increasing enzyme protein levels. Alterations in AC responsiveness to peptide hormones (e.g. thyroliberin and vasoactive intestinal peptide) correlated to membrane G_s protein -subunit content. These results demonstrate the involvement of G-protein translation regulation as one mechanism of ‘cross-talk’ between the PLC- and AC-dependent signalling pathways.     

The binding of eosin-labeled subunit δ to the isolated chloroplast ATPase, CF1, as revealed by rotational diffusion in solution

We investigated the binding of subunit δ to solubilized chloroplast ATPase. Purified δ was covalently labeled with eosin 5-isothiocyanate and its rotational correlation time was determined by a photoselection technique as a function of added CF₁ (containing δ) and of CF₁(−δ) (lacking δ). In aqueous buffer the rotational correlation time of labeled δ was 33 ns. This is compatible with a rather elongated shape with the dimensions 2b = 100 Å/2a = 28 Å. Binding of δ to CF₁ decreased the rotational correlation time about 10-fold. The result was a biphasic decay of the laser flash-induced absorption anisotropy which was analyzed to yield the proportion of δ (bound to CF₁) relative to δ (free). CF₁(−δ), which completely lacked the δ-subunit, bound one δ (mol/mol) with high affinity (K_d ≈ 100 nM) and at least another δ with about 20-fold lower affinity. The δ-containing CF₁, revealed only the low-affinity site(s) for δ. This was compatible with a 1:1 stoichiometry of δ in isolated CF₁.