This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Themmen, A. P. N. Articles by Huhtaniemi, I. T. Search for Related Content PubMed PubMed Citation Articles by Themmen, A. P. N. Articles by Huhtaniemi, I. T.

Mutations of Gonadotropins and Gonadotropin Receptors: Elucidating the Physiology and Pathophysiology of Pituitary-Gonadal Function

Department of Endocrinology and Reproduction (A.P.N.T.), Erasmus University Rotterdam, 3000 DR Rotterdam, The Netherlands; and Department of Physiology (I.T.H.), University of Turku, 20520 Turku, Finland

	Abstract

Top Abstract I. Introduction II. Structure-Function... III. Normal and Pathological... IV. Mutations in Human... V. Mutations in Human... VI. Animal Models of... VII. Future Directions References

 
The recent unraveling of structures of genes for the gonadotropin subunits and gonadotropin receptors has provided reproductive endocrinologists with new tools to study normal and pathological functions of the hypothalamic-pituitary-gonadal axis. Rare inactivating mutations that produce distinctive phenotypes of isolated LH or FSH deficiency have been discovered in gonadotropin subunit genes. In addition, there is a common polymorphism in the LHß subunit gene with possible clinical significance as a contributing factor to pathologies of LH-dependent gonadal functions. Both activating and inactivating mutations have been detected in the gonadotropin receptor genes, a larger number in the LH receptor gene, but so far only a few in the gene for the FSH receptor. These mutations corroborate and extend our knowledge of clinical consequences of gonadotropin resistance and inappropriate gonadotropin action. The information obtained from human mutations has been complemented by animal models with disrupted or inappropriately activated gonadotropin ligand or receptor genes. These clinical and experimental genetic disease models form a powerful tool for exploring the physiology and pathophysiology of gonadotropin function and provide an excellent example of the power of molecular biological approaches in the study of pathogenesis of diseases.

I. Introduction

II. Structure-Function Relationships of Gonadotropins and Gonadotropin Receptors

III. Normal and Pathological Gonadotropin Function

A. During sexual differentiation

B. Mature function

IV. Mutations in Human Gonadotropin Subunit Genes

A. Common -subunit

B. Mutations of the LHß subunit

C. Genetic variants of LHß subunit

D. hCGß subunit

E. FSHß subunit

V. Mutations in Human Gonadotropin Receptor Genes

A. Activating mutations in the LH receptor

B. Inactivating mutations of the LH receptor

C. Inactivating FSH receptor mutations

D. Activating FSH receptor mutation

VI. Animal Models of Disrupted Gonadotropin Function

A. Gonadotropin overexpression

B. Targeted disruption of gonadotropin genes

C. Targeted disruption of gonadotropin receptor genes

VII. Future Directions

	I. Introduction

Top Abstract I. Introduction II. Structure-Function... III. Normal and Pathological... IV. Mutations in Human... V. Mutations in Human... VI. Animal Models of... VII. Future Directions References

 
THE INFORMATION of genetic causes of human diseases is accumulating with increasing speed and volume. This development has become possible through unraveling of structures and functions of genes present in the human genome, a task that is almost complete. Although infertility is not conventionally considered an inherited condition, a growing number of mutations specifically affecting reproductive function have been detected in humans and characterized in genetically modified animals. Among single-gene mutations affecting hypothalamic-pituitary-gonadal function (1, 2), we know today those of the orphan nuclear receptor DAX1, extracellular protein KAL (anosmin), the receptor of GnRH, the homeodomain genes HES1 and PROP1, the receptors for estrogen and androgen, the steroidogenic acute regulatory protein (StAR), a number of steroidogenic enzymes, and finally those of gonadotropins and their receptors, the topic of this review.

Mutations of genes concerned with hypothalamic-pituitary-gonadal function, due to their critical role in the development and regulation of reproductive functions, are understandably very rare and therefore not of major concern within the clinical practice of infertility treatment. However, they form today a class of diagnoses that must be taken into account upon differential diagnostics of aberrant and delayed sexual differentiation and development, as well as infertility. In addition, by displaying distinct phenotypes, these conditions have turned out to be very elucidating with regard to the main facets and certain poorly characterized details of the hormonal control of reproduction. These naturally occurring mutations are often corroborated by genetically manipulated animal models with astonishingly similar phenotypes to those of the human diseases. The same applies to the currently known human mutations of gonadotropin and gonadotropin receptor genes, as well as to their animal models.

Whereas the hormone ligand mutations that have been found to date usually represent loss-of-function mutations (homozygotes or compound heterozygotes have a phenotype), the receptor mutations can be both of the gain-of-function (also heterozygotes have a phenotype) and loss-of-function type. In addition to clear-cut disease-causing mutations, the genome of all individuals is full of small structural variations, polymorphisms, which usually are associated with repeats in noncoding regions of the genome or point mutations within the genes. They may or may not cause alterations in gene function or structure of the encoded protein, and, consequently, often have no clear-cut phenotypic expression. However, subtle changes in function of the encoded protein, with mild phenotypic expression, are also possible in these cases.

	II. Structure-Function Relationships of Gonadotropins and Gonadotropin Receptors

Top Abstract I. Introduction II. Structure-Function... III. Normal and Pathological... IV. Mutations in Human... V. Mutations in Human... VI. Animal Models of... VII. Future Directions References

 
Together with TSH, the gonadotropins LH, human CG (hCG), and FSH form the family of glycoprotein hormones. LH, FSH, and TSH are produced in the pituitary gland, while the LH homolog hCG originates from the placenta. The members of this family of relatively large proteins (molecular mass, 30–40 kDa) consist of a common -subunit and a hormone-specific ß-subunit that are associated through noncovalent interactions. The mature -subunit consists of 92 amino acid residues and is encoded by a single gene, comprising four exons, which is localized on chromosome 6q12.21 (Fig. 1). The -subunit protein contains 10 cysteines, which are involved in intrasubunit disulfide linkages and two N-linked glycosylation sites. Although the ß-subunits confer functional specificity of the hormones, they show considerable amino acid identity, ranging from 32% for the LH-TSH pair to 83% for the LH-hCG pair (excluding the nonhomologous C-terminal extension of hCG). The ß-subunit genes are located on different chromosomes: the LH/hCGß gene cluster on chromosome 19q13.32, FSHß on chromosome 11p13, and TSHß on chromosome 1p13. The LH/hCGß gene cluster consists of one LHß gene and six hCGß genes and pseudogenes (3). At least five of the hCGß genes are expressed in choriocarcinoma cells and placenta, but most of steady-state hCGß mRNAs appear to be transcribed from genes 3, 5, and 8 (4).

View larger version (16K): [in this window] [in a new window]   Figure 1. Schematic representation of the human gonadotropin subunit genes. In the top part of each scheme the gene structure is depicted. The open bars indicate sequences that do not encode protein (noncoding). The closed bars indicate the sequences that comprise the open reading frame. The genes are drawn to scale. In the bottom part of each scheme the protein structure is shown. The signal peptide is indicated by the shaded bar, while the mature protein is depicted by the open bar. The numbers below the protein signify the start and end of the signal peptide and the length of the mature protein product, taking the first amino acid of the mature protein as 1. Below the protein the positions (and number of amino acid) of the N-linked glycosylation sites are indicated by inverted triangles and, in the case of CGß, the O-linked glycosylation sites (circles). The connecting lines between the coding exons in the top part of the scheme and the protein structure in the bottom part serve to indicate the sections of the protein encoded by the respective exon. Note that the ß-genes consist of three exons, and that the common -subunit gene is much larger mainly because of addition of the first, noncoding exon and long intron 1. In contrast to the other ß-genes, the first exon of FSHß is noncoding and exon 3 encodes a long 3'-untranslated region (open bar). The information shown in this figure is taken mainly taken from Ref. 3 and GenBank.

The crystal structure of deglycosylated hCG (6) has revealed that the -subunit and the ß-subunits both contain a so-called cystine knot structure, similar to some remotely related signaling molecules such as transforming growth factor-ß (TGFß), nerve growth factor (NGF), and platelet-derived growth factor (PDGF). Each subunit has elongated shape with two ß-hairpin loops on one side of the central cystine knot and a long loop on the other side. The noncovalent interaction between the two subunits is stabilized by a segment of the ß-subunit that extends like a "seat-belt" around the -subunit and is "locked" by a disulfide bridge.

Just as their ligands, the receptors for the glycoprotein hormones have related structures (Fig. 2). The receptors belong to the large family of G protein-coupled receptors, whose members all have a transmembrane domain that consists of seven-membrane traversing -helices connected by three extracellular and three intracellular loops. The glycoprotein hormone receptors form a separate subgroup within this large family on the merit of their large extracellular hormone-binding domain at the N terminus. FSH and TSH bind to the FSH and TSH receptors, respectively, while LH and hCG both bind to the same LH receptor. The LH and FSH receptor genes are located on chromosome 2p21 (7) and 2p21–16, respectively (8, 9), while the TSH receptor is found on a different chromosome, 14q31 (10, 11). The relationship of the glycoprotein hormone receptors to the other G protein-coupled receptors is indicated by their sequence homology in the C-terminal half of the receptor. This domain, encoded by a single, last exon, contains the seven-transmembrane segments and the G protein-coupling domain. The extracellular domain of the glycoprotein hormone receptors is encoded by the preceding 9 or 10 exons.

View larger version (28K): [in this window] [in a new window]   Figure 2. Schematic representation of the human gonadotropin receptor genes. The structure of the genes is depicted in the top of the drawings. The open bars indicate sections of the exons that encode untranslated regions of the mRNA, while the closed bars indicate those sequences that encode the protein. Although the genes are not drawn to scale, exons that are grouped in the gene are also drawn grouped. Both genes are at least 80 kb in size. The relation between the intron/exon structure of the gene and the domains of the protein are indicated by the broken lines. The horizontally hatched part of the protein indicates the signal peptide, and the cross-hatched bars signify the seven segments of the transmembrane domain. The numbers below the protein indicate the start and end of the signal peptide and the length of the total protein product including the signal peptide. The numbered inverted triangles below the protein indicate the positions of N-linked glycosylation sites. Note that the receptor genes are very similar in structure with the exception of an additional exon 11 in the LH receptor gene (see text). Exon 1 encodes the signal peptide and a small part of the extracellular domain; the following eight or nine exons encode the rest of the extracellular domain, including the leucine-rich repeat motifs. In both receptor genes, the final exon is the largest and contains the information for the transmembrane signal transduction domain. The information shown in this figure is taken mainly from Ref. 8 and GenBank.

In the extracellular domain of the LH and FSH receptors, a number of potential N-linked glycosylation sites have been identified. There are six sites in the LH receptor: Asn 99, 174, 195, 291, 299, and 313; and four sites in the FSH receptor: Asn 191, 199, 293, and 318, although the last site is not conserved among species (Fig. 2; the amino acids in the receptor proteins are numbered by taking the first methionine of the signal peptide as 1). The role of N-linked glycosylation in receptor function is not completely elucidated, and some seemingly contradictory results have been presented. Chemical deglycosylation of the rat LH receptor, or inhibition of glycosylation by tunicamycin treatment, did not prevent correct LH receptor folding, hormone binding, and signal transduction (18, 19). Mutational analysis of the rat LH receptor revealed a decrease or even complete loss of hormone binding activity upon elimination of the potential glycosylation sites at Asn¹⁰³, Asn¹⁷⁸, and Asn¹⁹⁹ (equivalent to Asn⁹⁹, Asn¹⁷⁴, and Asn¹⁹⁵ in the human), indicating the presence of functional carbohydrate chains at these positions in the rat ovarian LH/hCG receptor (20). Additional studies of the rat LH receptor using more extensive mutational dissection showed that all potential consensus glycosylation sites are N-glycosylated, but also revealed that the deleterious effects of the mutated N-linked glycosylation sites on rat LH receptor function result from the amino acid substitutions per se, and not from absent glycosylation (19). In the case of the rat FSH receptor, two of the three glycosylation sites (Asn¹⁹¹ and Asn²⁹³) are actually glycosylated, and a carbohydrate at either residue is required for efficient and correct folding of the receptor (21).

The extracellular ligand-binding domain of the gonadotropin receptors is connected to the transmembrane signaling domain by a hinge region. It is not clear whether this structure has functions other than serving as a connecting peptide, although the part of the hinge region closest to the first transmembrane segment is well conserved between the glycoprotein hormone receptors, suggesting a special role. Interestingly, in the marmoset monkey, exon 10 of the LH receptor gene, although present in the genome, is always completely spliced out from the mature mRNA (22). Exon 10 encodes the N-terminal part of the cystine cluster that is proposed to flank the leucine-rich repeat structure at its C terminus (17). This cluster has a chemokine-like structure, indicating an important function in the LH receptor. Nevertheless, the marmoset LH receptor (22) appears to function normally, suggesting that this cystine motif is not important in this species for receptor action and may act merely as a spacer allowing correct location of the extracellular domain in relation to the transmembrane domain. However, if exon 10 is deleted from the human LH receptor, the transit of the mutated receptor to the cell membrane is hampered (23), and the same is found to occur in a naturally occurring human LH receptor mutation in which the sequence encoding exon 10 is spliced out (Ref. 24 ; see below, Section V.B).

The transmembrane domain with its seven membrane-spanning -helices, connected by three extracellular and three intracellular loops, is similar to the other members of the large family of G protein-coupled receptors. Evidence has been presented that the transmembrane domain is sufficient for hormone binding (25), but this report has not been substantiated by other investigators. A molecular model of the LH receptor has proposed that some parts of the extracellular loops may function as contact points for the hormone bound to the extracellular binding domain and/or to this domain itself, relaying the hormonal signal to the intracellular face of the receptor (15). Together with the cytoplasmic parts of the transmembrane -helices, the intracellular loops and the C-terminal tail of the receptor form the interaction domain with G proteins. As with other members of the G protein-coupled receptor family, the third intracellular loop and the cytoplasmic tail are most closely involved in G protein coupling and in the selectivity of coupling to specific types of G proteins (26, 27).

The LH and FSH receptors are mainly coupled to Gs, the G protein that activates the various adenylyl cyclase isoenzymes, resulting in elevation of intracellular cAMP levels. Both of these receptors, however, are also able to activate other signal transduction pathways, and in in vitro experiments with cells isolated from experimental animals (rat, mouse, porcine) or transfected with LH or FSH receptor cDNA (human, rat, mouse), increased phosphatidylinositide turnover, elevated intracellular Ca²⁺, and activation of mitogen-activated protein kinases have been found (26, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40). Coupling to Gi proteins has been demonstrated for bovine and murine LH receptor (28), while the porcine LH receptor is able to activate Gq/11 and G13 in addition to Gi (32, 41, 42). Further studies will be necessary to elucidate the identity of G proteins coupling to the FSH receptor and to unravel possible species specificity of G protein coupling of either receptor type. The alternate intracellular pathways are in most cases activated at higher hormone concentrations than the cAMP pathway and may depend on high receptor densities (33). Therefore, their physiological relevance often remains unclear, although high serum levels of hCG and LH during pregnancy and around the time of ovulation, respectively, may make use of them.

Upon hormonal stimulation, the LH and FSH receptors desensitize, i.e., the hormonal signal is relayed in a less efficient manner. This process is caused by uncoupling of the receptor from the intracellular transducing G proteins, and by internalization of the receptor, resulting in decreased density of extracellularly exposed hormone binding sites. Most experiments addressing desensitization mechanisms have been carried out with rodents, either in vitro with isolated gonadal cells or in vivo following hormonal stimulation. Incubating purified testicular Sertoli (43, 44) or Leydig cells (45, 46) with FSH or LH, respectively, leads to rapid loss of hormone binding and cAMP response caused by loss of membrane receptors through internalization of the hormone-bound receptors. Similar findings have been made with various ovarian cell models in vitro (47, 48, 49, 50). In most experimental conditions, treatment of animals with LH, hCG, or FSH results in decreased responses of their testicular (51, 52, 53) and ovarian target cells (54, 55, 56). However, these mechanisms are not operative in all tissues. In the human corpus luteum, LH receptors or their mRNA are not lost under conditions of increasing hCG levels (57) or in fetal Leydig cells in the presence of high levels of placental hCG or increased LH secretion by the fetal pituitary (58, 59, 60).

In addition to the loss of receptors through internalization, a reduction in density of gonadotropin membrane receptors can also be caused by decreased receptor synthesis. In cultured immature rat Sertoli cells, a 4-h incubation with FSH or a cAMP analog causes complete disappearance of FSH receptor mRNA, probably through a posttranscriptional process involving a change in FSH receptor mRNA stability (44). Less strong effects of FSH that did not appear to involve mRNA stability changes were observed in another study with rat Sertoli cells (61). In cultured granulosa cells from rat and porcine origin, both stimulating and inhibitory effects of FSH on FSH receptor mRNA have been observed (62, 63, 64, 65). However, these observations may be a reflection of the differentiating actions of FSH on the granulosa cells, rather than rapid sensitization or desensitization effects. In similar culture studies with porcine Leydig cells (66), mouse tumor Leydig cells (67, 68), and rat granulosa cells (69), LH/hCG caused rapid LH receptor mRNA loss. Also in the case of the down-regulation of LH receptor mRNA levels, it remains unclear whether mRNA stability is involved in all cases.

Agonist-induced phosphorylation of G protein-coupled receptors by second messenger kinases such as protein kinase A or by G protein-coupled receptor kinases (GRKs) leads to receptor uncoupling and internalization through a process that involves binding of inhibitory proteins (arrestins) to the receptor and targeting to clathrin-coated pits (70, 71). This process has been best studied for the adrenergic receptors, but it is also involved in uncoupling and internalization of the gonadotropin receptors. Both rat LH and FSH receptors are uncoupled and internalized after stimulation, and subsequent phosphorylation of the C-terminal tail [LH receptor (72, 73)] or the first and third intracellular loops [FSH receptor (74)] has been implicated to play a role in this process. In vitro coexpression of rat LH or FSH receptors with several different GRKs or ßarrestins results in increased uncoupling and internalization, demonstrating the involvement of these proteins in the regulation of gonadotropin receptor function (75, 76, 77, 78). Phosphorylation of the LH and FSH receptors is not always sufficient or necessary for the desensitization processes (42, 79, 80), as is also found for other G protein-coupled receptors (81, 82). Changes in conformation of the receptor may also be important (83).

Recently, two more distantly related members of the glycoprotein hormone receptor family, named LGR4 and HG38/LGR5, were identified from expressed sequence tag (EST) databases based on their similarities to the other family members (84, 85). The most conspicuous difference between LGR4 and HG38/LGR5 and the other family members was found to be the addition of an extra eight leucine-rich repeat in the extracellular domain. LGR4 and HG38/LGR5, whose ligands and function are as yet unknown, show a less strict tissue distribution of expression (LGR4: ovary, testis, adrenal, placenta, thymus, spinal cord, thyroid; LGR5: muscle, placenta, brain, spinal cord) than the LH, FSH (ovary, testis), and TSH (thyroid) receptors (84, 85). Several groups, however, have reported that the LH receptor gene also is expressed in several nonclassical gonadotropin target tissues, such as placenta, brain, adrenal gland, and prostate (86, 87, 88, 89) and in normal and malignant breast tissue (90). However, the physiological significance of this "ectopic" LH receptor expression still remains unclear, since the effects of the absence of LH receptor function found in patients do not indicate effects on other tissues than the gonads, although these observations are limited by the effects of the pseudohermaphroditism or amenorrhea found in such cases (see below).

	III. Normal and Pathological Gonadotropin Function

Top Abstract I. Introduction II. Structure-Function... III. Normal and Pathological... IV. Mutations in Human... V. Mutations in Human... VI. Animal Models of... VII. Future Directions References

 
A. During sexual differentiation
The advances in molecular biology techniques have allowed major progression in our knowledge of the regulation and control of sex differentiation and gonadal function. Several of the genes that function at the very basis of the gonadal development are now known (e.g., WT1, SF1), as well as the proteins that govern whether a gonad develops in the male or female direction (e.g., Sry, Sox9, Dax1) (91).

During the first phase of sex differentiation, i.e., commitment of its direction, two testicular hormones come into play, signaling the direction of gonadal development to other genital structures of the developing fetus. The Sertoli cell product anti-Müllerian hormone (AMH; also known as Müllerian inhibiting substance, MIS), causes the regression of the anlagen of the female internal genitalia, i.e., the Müllerian ducts, preventing the development of these female organs in the male. The growth and differentiation of the male internal genitalia, which develop from the Wolffian ducts and the urogenital sinus, is stimulated by the androgen testosterone, produced by Leydig cells of the fetal testis, and in some target tissues by 5-dihydrotestosterone, a metabolite of testosterone through conversion by the 5-reductase enzyme. In the human fetus, after an initial gonadotropin-independent phase, this activation of steroidogenesis, as well as Leydig cell growth and differentiation, is completely dependent on placental hCG. Although the majority of hCG is secreted to the maternal circulation, its concentration in fetal blood is high enough to stimulate fetal testicular steroidogenesis (59). The role of fetal pituitary gonadotropins in the regulation of gonadal function in utero is apparently negligible. As will be elaborated below, male development is largely dependent on correct hormone production by the fetal and postnatal testis. In contrast, the fetal ovary is hormonally silent and fetal differentiation into the female direction appears to be totally independent of gonadotropins and gonadal function. The fetal ovary apparently does not express gonadotropin receptors (92).

A short period of high gonadotropin secretion after birth in both sexes is probably responsible for the postnatal peak of testosterone measured in baby boys during the first 3–4 months of postnatal life (93), whereas nothing is known about its possible effects on the ovary. After this postnatal peak, gonadotropin secretion is suppressed to very low levels until the advent of puberty. However, pulsatile release of low levels of gonadotropins, predominately at night, is also found in prepubertal children (94). The prepubertal testes apparently have both LH and FSH receptors, since they show clear testosterone and growth responses to LH/hCG and FSH, respectively (95, 96). In contrast, no data are available on the presence of gonadotropin receptors in the prepubertal ovary. However, also prepubertal ovaries may express gonadotropin receptors, since ovarian follicles of girls with a central activation of the hypothalamus/pituitary start producing estrogens.

Puberty can be envisaged as the second phase of sexual differentiation. Reactivation of the hypothalamic-pituitary-gonadal axis results in increased secretion of LH and FSH from the pituitary, with stimulation of the cognate gonadal target cells. Again, steroid hormone production plays a central role in signaling the maturation of gonadal function to extragonadal tissues. The androgen production of Leydig cells is now activated by pituitary LH and induces the secondary sex characteristics of the adult male. Feminization and attainment of fertility of the female occur under control of estrogens and progesterone, produced by the stimulated ovarian follicles and corpus luteum through combined actions of FSH and LH.

B. Mature function
In the ovary, granulosa cells are the only target cells of FSH action, thus expressing the FSH receptor, whereas both theca, stromal, late-stage (luteinizing) granulosa, and luteal cells contain LH receptors. The known role of FSH in the ovary is to stimulate follicular maturation, including follicular estrogen production through aromatization of androgens. LH stimulates androgen production in theca cells, thus providing substrate for granulosa cell estrogen production. LH also triggers ovulation, and thereafter maintains the progesterone production of corpus luteum. In the testis, Sertoli cells are the target of FSH action and Leydig cells are the target of LH action. The specific role of FSH in testicular function is still somewhat unclear, but functions such as stimulation of Sertoli cell proliferation in the immature testis and maintenance of qualitatively and quantitatively normal spermatogenesis, through indirect effects mediated by Sertoli cells, have been proposed. The role of LH is to stimulate Leydig cell androgen production and thereby to maintain the endocrine (extratesticular) and paracrine (spermatogenic) effects of androgens.

The synthesis and secretion of gonadotropins are under positive control of the hypothalamic GnRH (GnRH), and gonadal steroid and peptide (mainly inhibin) hormones exert negative and positive feedback effects on gonadotropin synthesis and secretion, either directly at the pituitary level or indirectly via the hypothalamus, mainly by modulating GnRH secretion. For GnRH to stimulate gonadotropin secretion, it is important that it is released in pulsatile fashion from the hypothalamus to the hypophysial portal circulation. This causes pulsatile secretion of gonadotropins, which is clearer with LH, due to its shorter half-time in circulation. However, the pulsatile mode of gonadotropin action at the gonadal level is apparently not important. Recent studies on male rats, either by follow-up of endogenous LH and testosterone pulses, or by pulsatile treatment with recombinant rat LH, demonstrate that gonadal stimulation is achieved by trains of multiple LH peaks of sufficient size (97, 98). However, the pulsatility of these effects may not be critical in view of effective gonadal stimulation by tonic gonadotropin injections in experimental animals and in humans. The importance of pulsatility in FSH secretion is even less clear, due to its longer half-life in circulation.

This pituitary-gonadal function remains basically similar in the female until menopause, after which estrogen production ceases in the absence of follicles, and gonadotropin secretion increases in the absence of ovarian negative feedback effects. In the male, there is gradual suppression of testicular androgen production and reciprocal increase of gonadotropins upon aging, beyond 50–60 yr of age.

A number of diseases at the hypothalamic and pituitary levels can impair the synthesis and secretion of gonadotropins. Aberrations in the hypothalamic regulation of gonadotropin synthesis, of which the best example is the absence of GnRH neurons, can result in Kallmann’s syndrome. This syndrome is due to disturbed migration of the GnRH neurons from the olfactory placode to their final location in the hypothalamus. This migration of GnRH neurons is disturbed in the most common X-linked form of Kallmann’s syndrome (hypogonadotropic hypogonadism and anosmia) through mutation in the gene of an extracellular matrix protein, anosmin, resulting in disturbance in development of olfactory bulbs and tracts (99). The other causes of abnormally low gonadotropin secretion include craniopharyngioma and a variety of other tumors, infiltrative diseases (e.g., sarcoidosis), trauma, vascular disease, radiation therapy, pituitary infarction, metabolic diseases (e.g., hemochromatosis), and functional causes such as stress and anorexia nervosa (100, 101). Pathologically increased levels of gonadotropins are observed in connection with paraneoplastic gonadotropin secretion, central precocious puberty, and primary hypogonadism (100, 101). We can also include in this category hyperthyroidism associated with pregnancy and trophoblastic tumors, where the highly elevated levels of hCG, due to structural similarity of LH/hCG, TSH, and their cognate receptors, are able to stimulate thyroid function by binding to the TSH receptor (102). More detailed discussion of these conditions is beyond the scope of this review.

A rare condition of suppressed gonadotropin action is caused by disturbance of gonadotropin glycosylation. This step of gonadotropin synthesis normally occurs through action of a group of specific enzymes, which in rare cases are inactivated by mutations. Although the actions of the glycosylation enzymes are not specific for gonadotropins, their mutations nevertheless appear to influence the formation of functionally competent glycoprotein hormones. The condition is called carbohydrate-deficient glycoprotein syndrome, where several different autosomal recessive enzyme deficiencies can result in incomplete glycosylation of plasma proteins (103). The syndrome appears to cause hypergonadotropic hypogonadism in women, where the high circulating levels of immunoreactive FSH have been found to have very low bioactivity (104, 105, 106). Male patients with this syndrome virilize at puberty but display suppressed testicular volume. LH levels and LH action seem to be only marginally, or not at all, affected in the subjects. This syndrome emphasizes the importance of proper glycosylation of gonadotropins for their bioactivity. Although it remains unclear why FSH is more affected than LH, the phenotypic expression of this disease is reminiscent of genetic inactivation of FSH or its receptor (see below).

In addition to the genetic alterations of gonadotropin receptor genes, the topic of the present review, there are also other causes for the end-organ gonadotropin resistance, i.e., hypergonadotropic hypogonadism. The apparent causes include anatomical aberrations of gonadal development and structure in various forms of gonadal dysgenesis and agenesis (107). The most common form of ovarian dysgenesis is Turner’s syndrome (45, XO), The other forms include diagnoses such as pure gonadal dysgenesis, ovarian steroidogenic enzyme defects, and premature ovarian failure, but their exact pathogenesis often remains open. Testicular resistance to gonadotropins can be caused by various developmental abnormalities. The most common chromosomal aberrations are Klinefelter’s (XXY) and XX male syndromes, while the other diagnoses include various forms of idiopathic and acquired arrest of spermatogenesis, acquired immunodeficiency syndrome (AIDS), various neurological diseases, trisomy 21, effects of drugs, radiation, and environmental toxins, autoimmunity, and a number of systemic diseases (108). Gonadotropin resistance of both sexes is also possible, if there are defects in gonadal actions of other circulating hormones or of para- or autocrine effectors (109). However, distinct clinical conditions with disturbances in these functions are not yet known.

	IV. Mutations in Human Gonadotropin Subunit Genes

Top Abstract I. Introduction II. Structure-Function... III. Normal and Pathological... IV. Mutations in Human... V. Mutations in Human... VI. Animal Models of... VII. Future Directions References

 
A. Common -subunit
Although there are several reports on restriction fragment length polymorphisms (RFLP) of the human glycoprotein hormone common -subunit gene (110, 111, 112, 113), none of them appear to influence the encoded amino acid sequence. Some studies (112, 114, 115), though not all (116), report that particular common genotypes are disproportionately represented in DNA derived from trophoblastic malignancies. Paired normal and tumorous tissues from the same subject showed similar RFLP patterns, suggesting that particular common alleles predispose toward a variety of neoplasias, rather than represent somatic mutations in tumors (115). How exactly common -polymorphism is related to tumorigenesis remains obscure. It could be linked to mutation of a neighboring gene with clear causal relationship to the malignancy, although it may also represent a spurious association.

The only genetic alteration so far reported in the -subunit protein is a single Glu⁵⁶Ala amino acid substitution in subunit ectopically secreted by a human carcinoma (Table 1) (117). This mutated protein failed to associate with the ßsubunit and appeared to have significantly higher mol wt than the native -subunit. It was proposed that the detected mutation causes altered tertiary structure, self-dimerization, or altered glycosylation, which could then be responsible for the ectopic subunit’s increased size and failure to dimerize with LHß.

B. Mutations of the LHß subunit
The only true human mutation of the LHß gene causing total functional inactivation is that described by Weiss et al. (118) (Table 1). The proband was a male, who presented with delayed puberty at the age of 17 yr. He was a member of a previously identified kindred with several infertile men (129) and had low testosterone and high immunoreactive LH serum concentrations. His testosterone secretion responded normally to exogenous LH and hCG, but in an in vitro bioassay serum LH was found to be devoid of bioactivity (130). These findings, together with the occurrence of infertility in three maternal uncles (with slightly lowered testosterone and increased LH) and a family history of consanguinity, suggested that the man had an inherited defect in the structure of LH, although his mother and sister had no symptoms of reproductive inadequacy. After two years of testosterone treatment, no sign of spontaneous puberty was seen after withdrawal of the treatment. Testicular biopsy revealed arrest of spermatogenesis and absence of Leydig cells. Long-term treatment with hCG resulted in testicular enlargement, normal virilization, and onset of spermatogenesis.

Upon sequencing of the LHß subunit gene of the subject, a homozygous A-to-G missense mutation was found in codon 54, causing a Glu-to-Arg substitution (118). The subject’s mother, sister, and three uncles were found to be heterozygous for the same mutation. A gene conversion whereby sequences from the CGß gene are exchanged with or incorporated in the LHß gene was excluded, which indicated that the alteration in LHß structure represented a spontaneous germ line mutation. Coexpression of the mutated LHß gene with normal -subunit gene in CHO cells resulted in formation of immunoreactive LH /ß heterodimers, with no activity in RRA, i.e., the mutated hormone was devoid of biological activity because of inability to bind to the LH receptor. In the heterozygous family members, as expected, the bioactivity of LH was reduced in relation to immunoreactivity, since half of their LHß was encoded by the mutated gene.

This rare case clarifies some points about the developmental role of pituitary LH. Since the proband was apparently normally masculinized at birth with descended testes, pituitary LH is not needed for the stimulation of testicular testosterone production in utero. Indeed, testosterone production is initiated autonomously, but becomes subsequently dependent on placental hCG (58, 59), and fetal pituitary LH apparently plays no role in regulation of fetal testicular function. However, the endocrine function of the postnatal testes is critically dependent on pituitary LH secretion, as was demonstrated by the total absence of spontaneous puberty in this subject lacking bioactive LH. It is intriguing that the heterozygous male family members had impaired steroidogenesis and high incidence of infertility despite normal pubertal masculinization. However, since the proband’s father was an obligate heterozygote, the importance of the heterozygozity for testicular function remains open. The heterozygous women, including the proband’s mother, were apparently free of symptoms. It is curious that no other human subjects homozygous for this type of mutation have yet been detected. The female phenotype would probably resemble those with inactivating LH receptor mutation (see below). Comparison of these two conditions would elucidate the role of intrauterine LH/hCG action, if any, in ovarian development and function.

C. Genetic variants of LHß subunit
Sequence variability of the LHß chain was observed in early reports on chemical sequencing of this protein (131), but the existence of polymorphic alleles of the LHß gene has only been recently recognized.

Upon testing the applicability of various monoclonal antibodies (Mabs) for the detection of LH, using the immuno- fluorometric assay (IFMA) principle, Pettersson and colleagues (132, 133, 134) described a healthy woman with two children, whose LH was undetectable using a Mab directed against an antigenic epitope present only in the intact LH /ß dimer (assay 1). The woman’s LH bioactivity and the ratio of bioactivity to immunoreactivity (using a subunit-specific IFMA for immunoreactive LH measurement, assay 2) were normal and in accordance with her normal fertility. Since her TSH and FSH levels were also normal, indicating no abnormalities in the common -subunit gene, the LHß gene was sequenced (119, 135). The LHß gene of the subject was found to represent a genetic variant (V) allele of the LHß gene, with two missense mutations: Trp⁸Arg (TGG -> CGG) and Ile¹⁵Thr (ATC->ACC) [Table 1 and Fig. 3 (119, 135)]. Recently, the same LHß allele was reported from Japan (120, 136) in female patients with infertility, and their LH likewise was unmeasurable with an immunometric assay kit using two Mabs. There is a complete linkage of the two mutations in all samples so far analyzed from various populations (135).

View larger version (27K): [in this window] [in a new window]   Figure 3. Structural comparison of WT LHß, the common LHß variant, and hCGß. The first 20 amino-terminal amino acids in the ß-subunits of WT and variant (VAR) LH, as well as of hCG, are depicted. The large shaded boxes indicate the positions of the two mutated amino acids in variant LHß, and the additional amino acids, differing in hCGß from LHß, are also boxed. The glycosylation signal Asn-Ala-Thr in LH-VAR-ß and hCGß is underlined.

View larger version (17K): [in this window] [in a new window]   Figure 4. Occurrence of WT and variant LHß alleles in the Finnish population. The distribution of 249 normal Finnish subjects into the normal (WT, ), low (heterozygote, •), and zero (homozygote, ) ratio groups according to the results of the ratios of LH measured by assay 1 (measuring only WT-LH) and assay 2 (measuring equally WT and V-LH). The LH level measured by assay 2 is shown on the abscissa. As no sex differences were detected, the male and female data are compiled. [Reproduced with permission from: A. M. Haavisto et al.: J Clin Endocrinol Metab 80:1257–1263, 2000 (137 ). © The Endocrine Society.]

The worldwide frequency of the V-LHß gene has been extensively studied (135, 140, 141, 142, 143, 144), and it appears to be highest in the Northern European populations (allelic frequency > 10%) and, interestingly, in Australian aboriginals (28.3%), whereas lower frequencies are found in Asian populations and American Indians (2.5–5%). In all ethnic groups with a representative number of observations, the WT and V-LHß alleles are in Hardy-Weinberg equilibrium. It is curious that such a high variability of carrier frequency as from 0 to >50% (allelic frequency from 0 to 28.3%) can be found for V-LHß. It is tempting to speculate that V-LH has, in prehistoric times, offered reproductive advantage for populations living in untoward external conditions. It also seems that its correlation with various pathologies related to pituitary-gonadal function varies between different populations, and their penetrance is apparently dependent on the genetic background (see below).

Due to its high frequency, more detailed studies on functional effects of V-LH were warranted. Both serum V-LH (137) and its recombinant form (139) are more active than WT-LH in in vitro bioassay, with lower ED₅₀ and about 20% higher maximum effect. In contrast, V-LH shows a clearly shorter half-life in circulation than WT-LH (26 vs. 48 min). As expected, the pulsatile pattern of LH secretion is not altered in carriers of the V-LHß allele (137). This leaves it somewhat open, whether the overall in vivo activity of V-LH is higher or lower than that of WT-LH. V-LH thus seems to be more active at the receptor site but the duration of its action is shorter. Most of the clinical observations indicate that V-LH represents a functionally weaker form of the hormone (see below).

To explain how a hormone with significantly shortened circulatory half-life can maintain grossly normal gonadal function, we hypothesized that its synthesis may be compensatorily enhanced. This would require alterations in the promoter function of the V-LHß gene. Indeed, when the V-LHß promoter was sequenced, a total of eight point mutations were detected within the first 650 nucleotides of its 5'-flanking sequence, and they always segregated with the two point mutations detected earlier in its coding sequence (145). The mutant promoter appeared about 50% more active than the WT promoter upon cell transfection studies, and it also displayed some qualitative differences in response to various hormonal stimuli. Hence, these findings demonstrated an intriguing evolutionary principle: if the function of a protein is altered through mutation in its gene, the change may be compensated for by additional mutations in its regulatory sequences that bring about opposite changes in synthesis of the mutated gene product.

Whether any particular phenotype(s) are related to V-LH is still somewhat unclear. It is also possible that both hetero- and homozygosity for the variant allele could give rise to different phenotypes. If we presume that V-LH represents a potent but short-acting form of LH, and WT-LH a less potent but long-acting form, a combination of both LH forms, as occurs in heterozygotes, could bring about different overall LH action than either of the forms alone. These qualitative differences between actions of WT and V-LH, including differences in their promoter function, provide the strongest evidence for possible phenotypic effects of V-LH. If the differences between the two hormone forms were only quantitative, then they could be fully compensated for by alterations in feedback regulation of LH secretion.

The first reports from Japan described V-LHß homozygosity with recurrent spontaneous abortions (136), menstrual irregularities with infertility (120), and polycystic ovarian syndrome (PCOS) (120, 146). Subsequently, various disturbances in pituitary-ovarian function have also been found in V-LHß heterozygotes (143, 147). Clear findings have not been made in Caucasian populations (137, 148, 149, 150), although in a study on predominately Jewish subjects from the Boston (Massachusetts) area, heterozygous women for V-LHß had a history of frequent use of infertility treatments (D. Cramer and I. Huhtaniemi, unpublished study). In Finland, no association of V-LH was found in women with history of recurrent miscarriages (151).

Heterozygous women for the V-LHß allele have higher levels of serum testosterone, estradiol, and sex-hormone-binding globulin (148), which indicates differences in ovarian LH action between WT- and V-LH. In a multicenter study from Finland, the United Kingdom, The Netherlands, and the United States, with a total of 1,466 subjects, of whom 363 had PCOS and 79 polycystic ovaries without other symptoms of PCOS, it was found that the V-LHß frequency was 5- to 7-fold lower in obese PCOS subjects compared with that in other groups, i.e., lean PCOS subjects and lean and obese controls (2–4.5% vs. 10.3–33.3%, P < 0.05) (149). Thus, V-LH may protect obese women from developing symptomatic PCOS, which indicates that the more powerful WT-LH induces the pathological ovarian responses. However, this finding was not corroborated in the UK population, despite similar diagnostic criteria (148, 149). This is in keeping with the multifactorial pathogenesis of PCOS (152) and emphasizes that its pathogenesis in different populations may vary according to genetic background. Even though V-LH may not be directly related to PCOS, determination of V-LH may improve the prediction of risk of PCOS, especially in obese women. Its high frequency in various populations must be kept in mind because many widely used immunoassay reagents do not detect this LH form. One diagnostic criterion for PCOS, the elevated LH/FSH ratio, may remain undetected if such an LH assay is used, as has been emphasized recently (153). Other phenotypic associations with V-LH include the delayed tempo of pubertal progression in boys heterozygous for V-LHß (154), and in elderly men, it was more common in those with low testosterone and high LH concentrations (150). However, larger numbers of observations from various ethnic groups are needed to resolve the role of V-LH in pathologies of gonadal function and infertility.

Hence, there is some evidence for association of V-LH, as a protective or predisposing factor, with various pathologies of LH action. Since findings in various ethnic groups do not always agree, the overall genetic background of the population may be important for the phenotypic expression and penetrance of this relatively mild polymorphic alteration in structure and function of the LH molecule. V-LH may thus be an example of influence of genetic heterogeneity on reproductive functions. Additional polymorphisms affecting reproductive endocrine functions are likely to be found. Interestingly, another polymorphism of LHß, detected recently (Ser¹⁰²Gly) in Singapore, has also been implicated in female infertility (155).

D. hCGß subunit
Several polymorphisms have been detected in the hCGß/LHß gene complex by RFLP analysis (156, 157), but whether they result in sequence differences in LH or hCG has not been studied in detail. Layman et al. (157) were unable to detect large deletions or duplications of the hCGß/LHß gene complex by genomic Southern blotting in patients with suspected disorders of hCG production, such as recurrent abortion, primary unexplained infertility, and gestational trophoblastic neoplasia. A very recent study (122) showed that of the six hCGß genes present in the human genome, the one most highly expressed, number 5 (158), is highly conserved. Altogether six polymorphisms were detected in this gene in a random population, and they were, with the exception of one, either silent or located in introns. An A-to-G transition in exon 3 of hCG gene 5 was found to alter the amino acid: Val⁷⁹ Met (Table 1). When the mutated ß-subunit was coexpressed with the common -subunit gene in CHO cells, the assembly of the two subunits was found to be inefficient, although those dimers that did form had normal bioactivity. This mutation was found in 4.2% of randomly chosen healthy subjects, but only in heterozygous form. A limited search in subjects with infertility (n = 41) yielded one additional silent mutation, but the above mentioned amino acid change was not found in this population. This may be due to the limited sample size or to the possibility that the mutation is embryonic-lethal because of insufficient production of biologically active hCG. Whether the intronic mutations detected were truly silent, or affected the rate of transcription or mRNA splicing, remains to be studied.

E. FSHß subunit
A total of five subjects (three women and two men) with different inactivating mutations of the FSHß gene have so far been described in the literature (Table 1). The first mutation reported was a homozygous 2-bp deletion in codon 61 (Val⁶¹) of the FSHß gene in a woman suffering from primary amenorrhea and infertility (123). The mutation gave rise to a completely altered amino acid sequence between codons 61 and 86 of the FSHß chain, which was followed by a premature stop codon, and lack of translation of amino acids 87–111. Consequently, the translated FSHß protein was truncated and unable to associate with common -subunit to form bioactive or immunoreactive /ß-dimers. The affected woman had apparently had normal adrenarche, but no menarche or telarche. Treatment of the patient with exogenous FSH resulted in follicular maturation, ovulation, and successful pregnancy. Her mother, heterozygous for the mutation, had suffered from menstrual irregularity and infertility, but these symptoms were unlikely related to the mutation, since the heterozygous relatives of other similar patients have been reported to be free of symptoms (see below).

The second case of inactivating FSHß mutation was also a female with similar phenotype, i.e., primary amenorrhea and poorly developed secondary sex characteristics (126). She had undetectable serum FSH and estradiol, high LH, and absent FSH response to GnRH stimulation test. Upon DNA sequencing, she appeared to be a compound heterozygote for two mutations in the FSHß subunit gene. One was the same as the mutation described by Matthews et al. (123), and the other was a missense T-to-G mutation, causing a Cys⁵¹to-Gly transition in the mature FSHß protein. Cells transfected with the FSHß gene carrying this mutation failed to produce immunoreactive FSH, apparently because of the loss of a cysteine critical for formation of proper disulfide bonds, as well as for synthesis and secretion of the hormone. No symptoms were found in the relatives heterozygous for either of the two FSHß mutations of the proband, suggesting that one intact FSHß gene is sufficient to maintain functionally adequate FSH secretion.

In a third female patient with isolated FSH deficiency, reported earlier (159), the cause was originally suggested to be due to circulating FSH antibodies (160). However, the molecular pathogenesis of this case was recently "re-revisited," and it was found to be due to the same homozygous 2-bp deletion as that of the first detected FSHß mutation (124). The FSH antibodies apparently developed in response to treatment with urinary gonadotropins (161), which were recognized as foreign protein by the patient’s immune system.

The female cases with FSHß inactivation are in good agreement and demonstrate that FSH is necessary for normal follicular development, ovulation, and fertility. Likewise, pubertal development is hampered in the absence of sufficient numbers of later stage follicles to harbor the granulosa cells needed for adequate estrogen production. As will be described below, this phenotype is practically identical to that caused by inactivating FSH receptor mutation.

Very recently, two men with FSHß mutations have been described (125, 127, 162). The report from Sweden by Lindstedt et al. (127, 162) described a 32-yr-old man of Serbian origin with azoospermia and normal puberty, but with selective absence of FSH. The LH-testosterone axis of the patient was apparently normal. Genetic analysis demonstrated a homozygous T-to-C mutation, predicting a Cys⁸²-to-Arg substitution in the FSHß protein. The second male, described from Israel (125), was an 18-yr-old man with slightly delayed puberty, small testes, azoospermia, and plasma FSH concentration below 0.5 IU/liter. Conspicuously, his testosterone level was low (4.5 nmol/liter) and LH high (24.5 IU/liter). Upon DNA sequencing, the same homozygous 2-bp deletion in codon 61 was found as reported before with the female patients (123, 124, 126). It was postulated, on the basis of studies on hCG biosynthesis (6, 163), that in the Cys⁸²Arg mutation, elimination of cysteine would result in inability to form the first intramolecular disulfide bond of FSHß. This would then result in abnormal tertiary structure during FSHß synthesis, with extensive intracellular degradation of the products, inability to associate with common -subunit, defective glycosylation, and finally inability to form biologically active hormone.

As will be elaborated below, the phenotypes of women with inactivating FSHß and FSH receptor mutations, as well as the female knockout mice for FSH ligand and receptor, are in perfect agreement. However, there is an apparent discrepancy between phenotypes of men with inactivating FSHß and FSH receptor mutations, which still leaves open the final word about the role of FSH in testicular function. Since the number of males with FSHß mutation so far reported is only two, and that of men with FSH receptor mutation five (164), the information about the role of FSH in the male that can be obtained from earlier descriptions of men with idiopathic isolated FSH deficiency can be valuable. These studies were done at the time when genetic diagnostics was not yet available. Some of the subjects with "isolated FSH deficiency" had associated disorders, such as cryptorchidism, hypospadias, omphalocele, deafness, the olfactory-genital dysplasia syndrome, chromosomal alterations, autoimmunity, or short stature (160, 165, 166, 167, 168), and therefore such patients may not be representative of truly isolated FSH deficiency. However, many of them seem to fulfill the diagnostic criteria. Those reported in full-length articles are summarized in Table 2, together with the two males with genetic proof of FSHß mutation. In addition, one abstract exists on isolated FSH deficiency in a male (169).

On the other hand, the two men with documented FSHß mutations (see above) were possibly only detected because of their azoospermia and/or delayed puberty. The Swedish patient, unlike some other FSH-deficient men (173), was resistant to FSH treatments for periods of 120 and 210 days (162), which may indicate additional contributing factor(s) to his azoospermia. The Israeli patient, in addition, had Leydig cell hypofunction, not demonstrated in any of the other FSH-deficient subjects, or those with FSH receptor defect (164), indicating the likelihood of an additional FSH-independent pathology of his testicular function.

In summary, the majority of information available indicates that FSH action per se is not mandatory for the pubertal initiation of spermatogenesis and fertility. However, qualitatively and quantitatively fully normal spermatogenesis apparently needs FSH action. The phenotype of men with defective FSH action varies from severe to mild impairment of spermatogenesis, in the face of apparently normal Leydig cell function. The azoospermia found in some of the men may be due to additional contributing factors, and not to truly isolated FSH deficiency. However, it is apparent that additional cases of genetically proven FSH deficiency are needed, before the existing discrepancy between phenotypes of the ligand and receptor deficiency, as well as the animal models with disrupted FSHß and FSH receptor genes (see below), can be resolved. At the moment, it may be warranted to state that treatment of men with idiopathic oligozoospermia and normal to elevated FSH concentration with FSH has no scientific basis, and that prospects of a male contraceptive method based on inhibition of FSH secretion or action are not promising.

	V. Mutations in Human Gonadotropin Receptor Genes

Top Abstract I. Introduction II. Structure-Function... III. Normal and Pathological... IV. Mutations in Human... V. Mutations in Human... VI. Animal Models of... VII. Future Directions References

 
In both LH and FSH receptor genes, activating and inactivating mutations have been identified with very different phenotypic effects. In the case of loss of function, it can be expected that the inactivating gene mutations range from missense changes of single amino acid residues, small deletions or insertions, frameshift mutations, and nonsense mutations that cause receptor truncation, to deletions of large parts of the receptor gene. Inactivating receptor gene mutations are found in homozygous or compound heterozygous states and the syndromes that are caused by these gene alterations follow a recessive pattern of inheritance. In contrast, activating mutations are much more limited in their character or position in the receptor gene. Actually, all activating receptor mutations in the gonadotropin receptors have been identified in the exon that encodes a small extracellular extension, the complete transmembrane domain of the receptor, and its intracellular C terminal tail (exon 11 in the LH receptor and exon 10 in the FSH receptor; see Fig. 2). Maps of the currently known LH and FSH receptor mutations are presented in Figs. 5 and 6.

View larger version (59K): [in this window] [in a new window]   Figure 5. Mutations in the LH receptor protein. Schematic structure of the LH receptor protein and localization of the inactivating (open squares) and activating (filled circles) mutations currently known in the human LH receptor. The short lines across the amino acid chain separate the 11 exons. For references, see Table 3.

View larger version (48K): [in this window] [in a new window]   Figure 6. Mutations in the FSH receptor protein. Schematic structure of the FSH receptor protein and localization of the inactivating (open squares), and one activating (filled circle), mutations currently known in the human FSH receptor. The short lines across the amino acid chain separate the 10 exons. For references, see Table 3.

View larger version (16K): [in this window] [in a new window]   Figure 7. Mutant Met571Ile LH receptor increases basal cAMP-responsive reporter activity. HEK293 cells were cotransfected with a cAMP-responsive luciferase reporter plasmid (202 ), a ß-galactosidase reporter plasmid driven by a constitutive promoter together with the expression plasmid pSG5 containing the human WT LH receptor cDNA (open circles) or a human LH receptor cDNA containing the Met571Ile mutation (closed circles). Cells were incubated with the indicated concentrations of hCG for 4 h. The luciferase activity in the cell lysates is presented after normalization for ß-galactosidase activity as a measure of transfection efficiency. Expression with the WT LH receptor-expression construct results in a clear response to hCG with low levels of reporter activity in the absence of hormone. The Met571Ile LH receptor shows similar response to hCG but with a substantial increase in basal cAMP-responsive reporter activity. [Courtesy of Dr. M.Verhoef-Post (unpublished).]

Comparison of the hCG dose-response activity of the receptor mutants and the WT-LH receptor reveals that, in many cases, the mutant LH receptor molecules display a lower response to a maximal stimulatory dose of hCG than the WT receptor (see e.g., Refs. 204, 206, 216). A decreased number of plasma membrane binding sites in cells expressing the mutant receptor molecules may be the reason for such a decrease (27, 190, 205, 210, 211, 213). A low number of cell surface expression may be caused by mutational effects on posttranslational modification and transport of the LH receptor protein (232) or increased internalization of the activated receptors (83, 233, 234, 235). However, a clear correlation between number of binding sites and maximal response is not always present. Comparing WT and two mutant LH receptors, Yano et al. (211) found that although both mutant receptors showed si