No description available.

The glycoprotein hormone family comprising of Luteinizing Hormone (LH), Chorionic Gonadotropin (hCG), Follicle Stimulating Hormone (FSH) and Thyroid Stimulating Hormone (TSH) plays important role in reproduction and overall physiology of the organism. These hormones are heterodimeric molecules consisting of an identical α subunit non-covalently associated with the hormone-specific β subunit. Both subunits of all these hormones are N-glycosylated. In addition, hCGβ subunit also has four O-linked oligosaccharides located at the C-terminus of the polypeptide(1). The α and β subunits of all these hormones contain five and six disulfide bonds respectively and the crystal structures of hCG and hFSH indicate that both subunits of the hormones belong to the cystine knot family of proteins(2-4). Although the β subunits are hormone specific, there are distinct similarities in these subunits with the 12 cysteines conserved in all these subunits (1). These hormones, because of their unique structural features have proved to be important models for structure–function relationship studies of complex dimeric glycoproteins. Folding of subunits during biosynthesis, role of glycosylation in folding pathways and in vitro and in vivo bioactivity of the hormone, as well as, identification of domains important for subunit association, receptor binding and subsequent signal transduction have been topics of active investigations. The receptors of these hormones belong to the family of G-protein coupled receptors (GPCR) and have unique hormone specific exodomain not present in other members of the GPCR family and characteristic seven transmembrane domains followed by a C terminal domain(5). Primary structure analysis of Glycoprotein hormone receptors family revealed sequence conservation, maximum homology being observed in the transmembrane domain (TMD)(6). Significant homologies could be observed in the hormone specific extracellular domains (ECD) also (7). Despite these homologies, the receptors exhibit exquisite specificity with very low cross reactivity with other members of the hormone family (8). Elucidation of the molecular details of the contacts between the hormone and the receptors has not been achieved so far. Various approaches have been employed to delineate the residues or domains of both hormone and receptors involved in interaction. These include testing of chimeras or mutants of hormones or receptors for changes in activity (9-12), chemical modifications(13) and competition with peptides from either hormones (14) or receptors (15). Polyclonal and monoclonal antibodies against glycoprotein hormones and various fragments of their receptors have been used to determine the role of different domains of both in binding and response (6, 16, 17). However, till date there is no consensus on the specific mechanisms by which the glycoprotein hormone docks onto its receptor. It was proposed that the initial contact between the hormone and the receptor occurs through high affinity binding of the hormone specific β subunit to the Leucine rich regions of the ECD that results in conformational changes in both hormone, as well as, the receptor and brings hormone/ECD complex closer to the TMD of the receptor. The secondary, relatively lower affinity interactions between the hormone and the receptor then take place through common α subunit and exoloops of TMD of the receptor resulting in signal generation (18, 19). Recently a different kind of model has been proposed which suggests that the hormone does not make any direct contacts with the TMD of the receptor. The signal is transduced by the change in contacts between ECD and TMD brought about by hormone’s interaction with ECD(8, 20). The present study was initiated with an overall objective of understanding the molecular details of the hormone receptor interactions of this family, particularly hCG- LH receptor interactions. Two different approaches were employed for this purpose; the first, direct approach being structure elucidation of the members of the glycoprotein hormone family while the second approach uses antibodies against hCG as tools to probe into hormone-receptor interactions. The results obtained using these two approaches have been consolidated in the present thesis and are organized as follows. Chapter 1 is an extensive review of the literature and it builds background for the present work while the exact aim and scope of the present work have been defined in Chapter 2. Chapter 3 describes cloning, expression and purification of recombinant glycoprotein hormones hLH, hCG and single chain derivative of hCG. The Chapter 4 gives details of the molecular aspects of hCG-LH receptor interaction dissected using hCG monoclonal antibodies (MAbs). Chapter 5 discusses implications of the observations made in the present study and states the future directions envisaged. There are a number of endocrinopathies associated with abnormal levels of glycoprotein hormones and treatments of such disorders often demand large quantities of either agonists or antagonists of the hormones. The structure-function relationship studies should help in identifying domains/residues important for subunit interaction, receptor binding, and signal transduction, which would also allow engineering of agonists and antagonists of hormone action. However, structure determination of the glycoprotein hormone family using X-ray crystallography has proved to be a difficult task and it is believed that the heterogeneity in glycosylation is the primary reason for this low success rate in the process of crystallization. The first crystal structure of hCG was that of completely deglycosylated hCG but such a molecule displays antagonistic behavior(2, 3). Use of NMR spectroscopy, the alternate method commonly used for structure determination is often limited by the availability of large quantities of biologically active hormones free of any contaminants. Large quantities of LH, hCG and FSH are also required for treatment of infertile patients suffering from gonadotropin deficiency. The first goal of the present study was thus to produce and purify biologically active recombinant hCG and hLH. Owing to the inherent features of glycoprotein hormones and their potential therapeutic applications, the recombinant expression of these hormones is an important goal from both basic research, as well as, commercial point of view. Considering the above mentioned features it is clear that the expression system used for the hyperexpression of these glycoprotein hormones should also serve as a model system for investigating structure–function relationships and folding of subunits during biosynthesis, in addition to providing sufficient quantities of the hormones for clinical applications. It has been demonstrated that N-linked glycosylation during biosynthesis facilitates protein folding and conformational maturation of glycoprotein hormone subunits into an assembly-competent, biologically active form (21). Therefore, the ideal recombinant expression system should also be able to glycosylate the protein during biosynthesis. The Pichia pastoris yeast expression system was chosen for hyperexpression of glycoprotein hormones as it blends the advantages of both bacterial and mammalian expression systems. Earlier, expression of biologically active hCG and the subunits of hCG and bovine FSH using Pichia pastoris expression system has been reported from the laboratory (22, 23). Chapter 3 (section 3.3.1) of the thesis describes hyperexpression of hLH. The expression of these heterologous proteins was scaled up using fermentation procedures to fulfill the requirements of large quantities of hormones for various applications. Purification of Pichia expressed hormones turned out to be a complex task as large quantity of the hormone was secreted out in the fermentation medium (10litre volume) that was of high ionic strength. Of several different strategies attempted for concentration and partial purification of recombinant hCG, hydrophobic interaction chromatography (HIC) using Phenyl Sepharose matrix emerged as the most efficient technique as a first step of purification. Subsequently, cation exchange chromatography using SP- Sepharose matrix yielded completely purified biologically active recombinant hCG (section 3.3.2). The preliminary data also suggested that Pichia cells express a biologically active form of hCG which appeared to be less glycosylated and of lower molecular weight. Using the same protocol purification of hLH, as well as, single chain derivative of hCG, phCGαβ was achieved (section 3.3.3). These recombinant proteins were characterized extensively using various biochemical, as well as, immunological criteria and were shown to be similar to their natural counterparts with respect to their ability to bind LH receptor and to transduce signal as judged by radioreceptorassays and in vitro bioassays respectively. The hydrophobic interaction chromatography proved an important starting point for purification of all the other members of the glycoprotein hormone family expressed using Pichia pastoris expression system. With the availability of purified, biologically active recombinant hCG in large quantities it was now possible to make attempts towards structure elucidation using NMR spectroscopy. The structure determination of such complex proteins by NMR spectroscopy is made relatively easier by labeling the proteins with magnetically more active, stable isotopes of carbon and nitrogen, 13C and 15N respectively however the cost is often prohibitively high. The Pichia pastoris expression system offers simple means of labeling the proteins as the cells can be grown on simple salts of carbon and nitrogen such as 13C labeled methanol, 15N labeled ammonium chloride or ammonium sulphate. The Chapter 3 also gives a brief account of the preliminary attempts made to label the recombinant hCG with 15N and the structural studies carried out with the carbohydrate moieties of the recombinant hCG using solution NMR spectroscopy. This work was carried out in collaboration with the laboratory of Prof. J.P Kamerling of the University of Utrecht, Netherlands and the efforts are currently underway to elucidate the complete structure of the Pichia expressed hCG. The common feature of receptors and antibodies against the ligand is that both display very specific, high affinity binding towards the ligand. Hence, it is logical to speculate that the antigen binding regions of the antibodies that inhibit hormone binding and/or response, exhibit homology with distinct domains of the receptor. By identifying the epitopes recognized by such antibodies, it should be possible to predict contact points between the hormone and the receptor. In the present study, this hypothesis has been tested using monoclonal antibodies (MAbs) against hCG recognizing different epitopes in the hormone molecule and having different effects on hormone binding and response (Chapter 4). These MAbs were classified as α subunit specific, β subunit specific or heterodimer specific depending on their abilities to bind either subunit in addition to the hormone itself. Interestingly, it was observed that the hCGβ subunit specific MAbs, as well as, heterodimer specific MAbs inhibited hCG receptor binding and hence the response generated by hCG, while the hCGα subunit specific MAbs inhibited only response to the hormone without interfering in binding (Section 4.3.1). To dissect out these interactions further the epitopes recognized by these antibodies on hCG molecule were determined (Section 4.3.2), single chain fragment variable (ScFv) were generated from each of these antibodies and it was shown that these ScFv retain the functionality of the original antibody (Section 4.3.3). Further, the amino acid sequence of each antibody was determined (Section 4.3.4) and finally shown that the antigen binding domains of antibodies show homology to the distinct regions of the LH receptor on sequence alignments between the two using three different programs (4.3.5). The hCGβ subunit specific MAb 52/28' displayed distinct homology with the ECD of LH receptor while the α subunit specific MAb C10 showed regions homologous to TMD of the receptor and the heterodimer specific MAb E12 was found to be similar to the hinge region of the receptor. This clearly indicates that the β subunit of hCG is in close contact with ECD of the receptor while the α subunit makes contacts with the TMD of receptor. The present study thus supports the existing model of hormone receptor interactions, which states that the hormone first binds to the exodomain of the receptor mainly through its β subunit while the integrity of the α subunit is critical for signaling. (24, 25). Also, the observations made in the present study exhibit an interesting possibility of antigen antibody complexes being used as surrogate models for gaining insights into hormone receptor complex. Further, it has been reported that hCG has immunocontraceptive potential(26). Active and passive immunization studies with hCG in primates and humans have demonstrated the possibility of controlling fertility by the antibodies capable of neutralizing hCG. This forms the basis for female contraceptive vaccine that has undergone Phase II clinical trials in India. The MAb E12 characterized in the present study displayed highly specific binding to heterodimeric hCG exclusively without showing any cross reactivity with hLH (Section 4.3.1). The epitope mapping analysis revealed that this antibody recognizes a unique discontinuous epitope present only in the heterodimeric hCG and is distinct from the unique C-terminal extension of hCGβ absent in hLHβ (Section 4.3.2). The MAb, IgG or its recombinant single chain fragment variable (ScFv), inhibited response to hCG, but not to hLH (4.3.3). Thus, the epitope recognized by this MAb is an ideal candidate antigen for immunocontraception. The MAb E12 can also be used for passive immunization in case of emergency contraception. Another potential application of hCG specific antibodies is in homing and the treatment of tumors that secrete hCGβ subunit. The hCGβ subunit specific MAbs used in the present study 52/12 and 52/28' that inhibited hCG receptor binding as well as response generated by hCG can be used in treating such tumors. The functional ScFvs generated from these MAbs in the present study can be made use of on humanization. Thus, the present study has yielded some important molecules for therapeutic applications besides providing a new platform for structure-function relationship studies of the complex glycoprotein hormones.

The glycoprotein hormones, Luteinizing Hormone (LH), human Chorionic Gonadotropin (hCG), Follicle Stimulating Hormone (FSH) and Thyroid Stimulating Hormone (TSH) are heterodimeric proteins with an identical α-subunit associated non-covalently with the hormone specific β-subunit and play important roles in reproduction and overall physiology of the organism [1]. The receptors of these hormones belong to the family of G-protein coupled receptors (GPCR) and have a large extracellular domain (ECD) comprising of 9-10 leucine rich repeats (LRR) followed by a flexible hinge region, a seven helical transmembrane domain (TMD) and a C terminal cytoplasmic tail [2]. Despite significant sequence and structural homologies observed between the ECDs of the receptors and the specific β-subunits of the hormones, the hormone-receptor pairs exhibit exquisite specificity with very low cross-reactivity with other members of the family. The TSH receptor (TSHR) is an especially interesting member of this family as it not only recognizes is cognate ligand, i.e. TSH, but also binds to the non-cognate ligands such as autoantibodies. TSHR autoantibodies come in different flavors; inhibitory antibodies that compete with the hormone for receptor binding and block its action, stimulatory antibodies that activate the receptor in a hormone independent manner and neutral antibodies that bind to the receptor but do not directly influence its functions. The inhibitory autoantibodies cause hypothyroidism and are responsible for Hashimoto’s Thyroiditis, whereas the stimulatory autoantibodies cause Graves’ thyrotoxicosis characterized by hyperthyroid condition [3]. The exact epitopes of these autoantibodies are not well delineated although it has been hypothesized that the blocking type- and the stimulatory type- autoantibodies have predominant epitopes in the TSHR ECD that overlap with hormone binding regions [4]. Insights into the mode of hormone or autoantibody binding to the receptor was primarily derived from the crystal structure of FSHR leucine rich repeat domain (LRRD) bound to single chain analog of FSH, and the crystal structures of TSHR LRRD bound to the stimulatory type human monoclonal antibody M22 [5] and the inhibitory type- monoclonal antibody K1-70 [6]. Both these crystal structures propose LRRDs as the primary ligand binding site which interacts with the hormone through its determinant loop in a hand-clasp fashion [7] while the autoantibodies mimics the hormone binding to a large extent [8] . These structures, while providing detailed understanding of the molecular interactions of the LRRs with the hormone, shed little light on the mechanism by which the signal generated at the LRRD are transduced to the downstream effector regions at the distally situated TMD. Hence, while one understands the ligand binding to a large extent, the activation process is not well understood, one of the central objective of the present study. Ligand-receptor interactions are typically studied by perturbing ligand/receptor structure by mutagenesis or by mapping conformational changes by biophysical or computational approaches. In addition to the above-mentioned approaches, the present work also uses highly specific antibodies against different domains of the receptor as molecular probes due to the ability of antibodies to distinguish between conformations likely to arise during the activation process. Use of antibodies to understand the receptor activation process is especially apt for TSHR due to the presence of physiologically relevant TSHR autoantibodies and their ability to influence hormone binding and receptor activation [9, 10]. Chapter 2 attempts to provide a comparison between the interactions of the hormone and the autoantibodies with TSHR. For this purpose, two assays were developed for identification of TSHR autoantibodies in the sera of patients suffering from autoimmune thyroid diseases (AITD), the first assay is based on the ability of TSHR autoantibodies to compete for radiolabeled hormone (The TSH binding inhibition (TBI), assay) and the second based on the capability of stimulatory antibody to produce cAMP in cells expressing TSHR (TSHR stimulatory immunoglobin (TSI) assay). A stable cell line expressing TSHR capable of recognizing both TSH and TSHR autoantibodies was thus created and used for prospective and retrospective analysis of AITD patients. Based on the TBI and TSI profiles of IgGs, purified from AITD patient's sera, it was recognized that TSHR stimulatory and TSH binding inhibitory effects of these antibodies correlated well, indicating overlap between hormone binding and IgG binding epitopes. It was also recognized that stimulatory IgGs are not affected by negative regulatory mechanism that governs TSH secretion substantiating the persistence of these antibodies in circulation. Kinetics of cAMP production by Graves’ stimulatory IgG was found to be fundamentally distinct, where the autoantibodies displayed pronounce hysteresis during the onset of the activation process when compared to the hormone. This could possibly be explained by the oligoclonality of the autoantibody population, a different mechanism of receptor activation or dissimilarity in autoantibody and hormone epitopes. To gain additional insights into the epitopes of TSHR autoantibodies and the regions that might be critical in the activation process, different overlapping fragments encompassing the entire TSH receptor ECD were cloned, expressed in E.coli as GST fusion proteins and purified: 1] the first three LRRs (TLRR 1-3, amino acid (aa) 21-127), 2] the first six LRRs (TLRR 1-6, aa 21-200), 3] the putative major hormone binding domain (TLRR 4-6, aa 128-200), and 4] the hinge region of TSH receptor along with LRR 7 to 9, (TLRR 7-HinR, aa 201-413). The receptor fragment TLRR 7-HinR was further subdivided into LRR 7-9 (TLRR 7-9, aa 201-161) and the hinge region (TSHR HinR, aa 261-413), expressed as N-terminal His-Tagged protein and purified using IMAC chromatography. Simultaneously, the full-length TSHR ECD was cloned, expressed and purified using the Pichia pastoris expression system. ELISA or immunoblot analysis of autoantibodies with the TSHR exodomain fragments suggested that Graves’ stimulatory antibody epitopes were distributed throughout the ECD with LRR 4-9 being the predominant site of binding. Interestingly, experiments involving neutralization of Graves’ IgG stimulated cAMP response by different receptor fragment indicated that fragments corresponding to the TSHR hinge region were better inhibitors of autoantibody stimulated receptor response than corresponding LRR fragments, suggesting that the hinge region might be an important component of the receptor activation process. This was in contrast to prevalent beliefs that considered the hinge region to be an inert linker connecting the LRRs to the TMD, a structural entity without any known functional significance. Mutagenesis in TSHR hinge region and agonistic antibodies against FSHR and LHR hinge regions, reported by the laboratory, recognized the importance of the hinge regions as critical for receptor activation and may not simply be a scaffold [11-13]. Unfortunately, the mechanism by which the hinge region regulates binding or response or both have not been well understood partially due to unavailability of structural information about this region. In addition poor sequence similarity within the GpHR family and within proteins of known structure, make this region difficult to model structurally. In chapter 3, effort is made to model the hinge regions of the three GpHR based on the knowledge driven and Ab initio protocols. An assembled structure comprising of the LRR domain (derived from the known structures of FSHR and TSHR LRR domains) and the modeled hinge region and transmembrane domain presents interesting differences between the three receptors, especially in the manner the hormone bound LRRD is oriented towards the TMD. These models also suggested that the α-subunit interactions in these three receptors are fundamentally different and this was verified by investigating the effects of two α-subunit specific MAbs C10/2A6 on hCG-LHR and hTSH-TSHR interactions. These two α-subunit MAbs had inverse effects on binding of hormone to the receptor. MAb C10 inhibited TSH binding to TSHR but not that of hCG, whereas MAb 2A6 inhibited binding of hCG to LHR but not of hTSH. Investigation into the accessibility of their epitopes in a preformed hormone receptor complex indicated that the α-subunit may become buried or undergo conformational change during the activation process and interaction may be different for LHR and TSHR. Fundamental differences in TSHR and LHR were further investigated in the next chapter (Chapter 4), especially with regards to the ligand independent receptor activation. Polyclonal antibodies were developed against LRR 1-6, TLRR 7-HinR and the TSHR HinR receptor fragments. The LRR 1-6 antibodies were potent inhibitor of receptor binding as well as response, similar to that observed with antibodies against the corresponding regions of LHR. Interestingly, the antibodies against the hinge region of TSHR were unable to inhibit hTSH binding, but were effective inhibitors of cAMP production suggesting that this region may be involved in a later stage of a multi-step activation process. This was also verified by studying the mechanism of inhibition of receptor response and their effect on ligand-receptor association and dissociation kinetics. Hinge region-specific antibodies immunopurified from TLRR 7-HinR antibodies behaved akin to those of the pure hinge region antibodies providing independent validation of the above results. This result was, however, in contrast to those observed with a similar antibody against LHR hinge region. As compared to the TSHR antibody, the LHR antibody inhibited both hormone binding and response. In addition, this antibody could dissociate a preformed hormone-receptor complex which was not observed for TSHR hinge region antibodies. Although unable to dissociate preformed hormone-receptor complex by itself, the TSHR HinR antibodies augmented hormone induced dissociation of the hormone-receptor complex suggesting that this region may be involved in modulation of negative cooperativity associated with TSHR. Molecular dissection of the role of hinge region of TSHR was further carried out by using monoclonal antibodies against LRR 1-3 (MAb 413.1.F7), LRR 7-9 (MAb 311.87), TSHR hinge region (MAb 311.62 and MAb PD1.37). MAb 311.62 which identifies the LRR/Cb-2 junction (aa 265-275), increased the affinity of TSHR for the hormone while concomitantly decreasing its efficacy, whereas MAb 311.87 recognizing LRR 7-9 (aa 201-259) acted as a non-competitive inhibitor of TSH binding. MAb 413.1.F7 did not affect hormone binding or response and was used as the control antibody for different experiments. Binding of MAbs was sensitive to the conformational changes caused by the activating and inactivating mutations and exhibited differential effects on hormone binding and response of these mutants. By studying the effects of these MAbs on truncation and chimeric mutants of thyroid stimulating hormone receptor (TSHR), this study confirms the tethered inverse agonistic role played by the hinge region and maps the interactions between TSHR hinge region [14] and exoloops responsible for maintenance of the receptor in its basal state. Mechanistic studies on the antibody-receptor interactions suggest that MAb 311.87 is an allosteric insurmountable antagonist and inhibits initiation of the hormone induced conformational changes in the hinge region, whereas MAb 311.62 acts as a partial agonist that recognizes a conformational epitope critical for coupling of hormone binding to receptor activation. Estimation of apparent affinities of the antibody to the receptor and the cooperativity factor suggests that epitope of MAb 311.87 (LRR 7-9) may act as a pivot involved in the initial events immediate to hormone binding at the LRRs. The anatgonsitic effect of MAB 311.62 on binding and response also suggested that binding of hormone is conformationally selective rather than an induced event. The hinge region, probably in close proximity with the α-subunit in the hormone-receptor complex, acts as a tunable switch between hormone binding and receptor activation. In contrast to the stimulatory nature of Cb-2 antibody such as MAb 311.62, MAb PD1.37, which identified residues aa 366–384 near Cb-3, was found to be inverse agonistic. Unlike other known inverse agonistic MAbs such as CS-17 [15] and 5C9 [16], MAb PD1.37 did not compete for TSH binding to TSHR, although it could inhibit hormone stimulated response. Moreover, unlike CS-17, MAb PD1.37 was able to decrease elevated basal cAMP of hinge region constitutively activated mutations only but not those in the extracellular loops. This is particularly important as interaction of hinge region residues with those of ECLs had been thought to be critical in maintenance of the basal level of receptor activation and are responsible for attenuating the constitutive basal activity of the mutant and wild-type receptors in the absence of the hormone. This was demonstrated by a marked increase in the basal constitutive activity of the receptor upon the complete removal of its extracellular domain, which returned to the wild-type levels upon reintroduction of the hinge region. However, careful comparison of the activities of the mutants (receptors harboring deletions and gain-of-function mutations) with maximally stimulated wild-type TSHR indicated that these mutations of the receptor resulted primarily in partial activation of the serpentine domain suggesting that only the ECD in complex with the hormone is the full agonist of the receptor. Confirmation of the above proposition has been difficult to verify primarily due to a highly transient conformational change in the tripartite interaction of the hinge region/hormone and the ECLs. The current approaches of using antibodies to probe the ECLs are difficult due to the conformational nature of the antigen as well as difficulty in obtaining a soluble protein. In chapter 5, the ligand induced conformational alterations in the hinge regions and inter-helical loops of LHR/FSHR/TSHR were mapped using the exoloop specific antibodies generated against a mini-Transmembrane domain (mini-TMD) protein. This mini-TMD protein, designed to mimic the native exoloop conformations, was created by joining the TSHR exoloops, constrained through the helical tethers and library derived linkers. The antibody against mini-TMD specifically recognized all three GpHRs and inhibited the basal and hormone stimulated cAMP production without affecting hormone binding. Interestingly, binding of the antibody to all three receptors was abolished by prior incubation of the receptors with the respective hormones suggesting that the exoloops are buried in the hormone-receptor complexes. The antibody also suppressed the high basal activities of gain-of-function mutations in the hinge regions, exoloops and TMDs such as those involved precocious puberty and thyroid toxic adenomas. Using the antibody and point/deletion/chimeric receptor mutants, dynamic changes in hinge region-exoloop interactions were mapped. The computational analysis suggests that mini-TMD antibodies act by conformationally locking the transmembrane helices by restraining the exoloops and juxta-membrane regions. This computational approach of generating synthetic TMDs bears promise in development of interesting antibodies with therapeutic potential, as well as, explains the role of exoloops during receptor activation. In conclusion (Chapter 6), the study provides a comprehensive outlook on the highly dynamic interaction of ligand and different subdomains of the TSHR (and to a certain extent of LHR and FSHR) and proposes a model of receptor activation where the receptor is in a dynamic equilibrium between the low affinities constrained state and the high affinity unconstrained state and bind to the hormone through the LRR 4-6. Upon binding the βL2 loop of the hormone contact LRR 8-10 that triggers a conformational change in the hinge region driving the α-subunit to contact the ECLs. Upon contact, the ECLs cooperatively causes helix movement in the TMH and ultimately in ICLs causing the inbuilt GTP-exchange function of a GPCR.

The glycoprotein hormones, Luteinizing Hormone (LH), human Chorionic Gonadotropin (hCG), Follicle Stimulating Hormone (FSH) and Thyroid Stimulating Hormone (TSH) are heterodimeric proteins with an identical α-subunit associated noncovalently with the hormone specific β-subunit and play important roles in reproduction and overall physiology of the organism (Pierce & Parsons, 1981). The receptors of these hormones belong to the family of G-protein coupled receptors (GPCR) and have a large extracellular domain (ECD)comprising of 9-10 leucine rich repeats (LRR) followed by a flexible hinge region, a seven helical transmembrane domain (TMD) and a C terminal cytoplasmic tail (Vassart et al, 2004). Despite significant sequence and structural homologies observed between the ECDs of the receptors and the specific β-subunits of the hormones, the hormone-receptor pairs exhibit exquisite specificity with very low cross-reactivity with other members of the family. Several biochemical, immunological and molecular biological tools have been employed to elucidate the structure– function relationship of the hormones and their receptors. These studies also helped in deciphering some of the regions present in both the hormones and the receptors involved in maintaining the specificity of their interaction (Fan & Hendrickson, 2005b; Fox et al, 2001; Wu et al, 1994). However, the complete understanding of the hormone-receptor contact sites and mechanism of receptor activation are still an enigma. Understanding the molecular details of these phenomena can lead to the development of novel strategies of regulating hormone action. Binding of FSH to FSHR occurs in the large extracellular NH2-terminal domain where the participation of the LRRs (amino acids 18-259) is essential to determine the ligand selectivity (Dias & Van Roey, 2001; Fan & Hendrickson, 2005a; Szkudlinski et al, 2002). In fact, mutations in these regions lead to reduction in binding of the agonist to the receptor. It is not known how the signal from the large extracellular domain liganded complex is transmitted to the TMD (amino acids 367-695). It is envisioned that hormone binding to the LRRs leads to series of conformational changes leading activation of the TMD resulting in signal transduction. The recently reported crystal structure of the single chain form of FSH in complex with the leucine rich repeats of the FSHR (amino acids 1-268) (Fan & Hendrickson, 2005b), although provides detailed understanding of the molecular interactions of the LRRs with the hormone, fails to provide any insights into mechanism of receptor activation as the information regarding critical interaction of the hormone with TMD. This structure also did not provide any information on the role of the hinge region (amino acids 259-366) that connects the LRRs to the TMD in hormone binding and activation of the receptor. In the present study an attempt has been made to understand the role of the hinge region in hormone binding and signal transduction. The overall objective of the study is to elucidate the molecular details of the hormone receptor interactions, particularly FSH-FSHR interaction. Antibodies to glycoprotein hormones and their receptors have often provided insights into the mechanism of hormone-receptor interactions and signal transduction. While the TSH receptor antibodies and their effects on the overall physiology have been well documented (Khoo & Bahn, 2007; Rapoport & McLachlan, 2007), reports of such antibodies against FSHR or LHR and their possible effects on the reproductive functions are not available. In the present study, effects of FSHR antibodies with different specificities on FSH-FSHR interactions have been investigated. Antibodies to different regions of rat FSHR, were raised and extensively characterized and their effects of FSH-FSHR interactions and signaling were investigated. It was found that a polyclonal antibody against the hinge of the receptor (RF2 antiserum, amino acids 218-336), while having no significant effect on hormone binding and response could stimulate the receptor by itself bypassing the hormone. This stimulation of FSHR was very specific as this antiserum could not stimulate LHR or TSHR and could be blocked by preincubating the antibody with the antigen. Through competition experiments with different synthetic peptides of human FSHR, a stretch of hinge region corresponding to amino acids 296-331 was identified as the site recognized by the stimulatory antibody. This antibody did not interfere in hormone binding and could also bind to the pre-formed hormone-receptor complex suggesting that the binding site of the antibody may not participate directly in hormone binding. Subsequently the antibody was extensively characterized for its effect of hormone receptor interactions (Chapter 2). Previous studies considered the hinge region to be an inert linker connecting the LRRs to the TMD, a structural entity without any known functional significance (Vlaeminck-Guillem et al, 2002). However, the data with RF2 antibody suggested a direct role of the hinge region in signal transduction. Therefore, a systematic study to dissect the role to hinge region in hormone binding and signal transduction was conducted. Several truncations, deletions, activating and inactivating point mutations in the FSHR were generated to understand the mechanism of receptor activation. Firstly, these mutant receptors were characterized for their ability to translocate to the cell surface when transfected in the cultured mammalian cells. Secondly, affinity of all the mutant receptors for the hormone was determined in order to understand the effect of mutations on hormone binding. Finally, the cAMP response of these mutant receptors to the hormone and the stimulatory antibody was investigated to understand the effects of mutations on signal transduction. The results are described in Chapter 3. The hormone binding analysis and the affinity measurement of the mutant receptors showed that the LRRs are involved in high affinity hormone binding while the hinge region may not contribute to the process. This is in agreement with the crystal structure data which showed that the hormone was bound to the truncated receptor fragment representing only the LRRs (Fan & Hendrickson, 2005b). These binding data also corroborated the earlier data indicating that the antibodies against the hinge region do not interfere in hormone-receptor interactions. Further, the analysis of different N-terminally truncated receptor mutants provided strong evidence indicating that the constraining intramolecular interactions between the extracellular and the transmembrane domains are required to maintain the FSHR in an inactive conformation in the absence of an agonist. The analysis of the constitutive basal activity of the mutant receptors in absence of hormone suggested that certain regions of the extracellular domain had an attenuating effect over the TMDs that prevented constitutive activation of the receptor. This was demonstrated by a marked increase in the basal constitutive activity of the receptor upon the complete removal of its extracellular domain. Detailed analysis of the mutants suggested that LRR portion does not contribute to this attenuating effect, but it is the hinge region that perhaps interacts with the TMDs and dampens its basal constitutive activity. This attenuating effect was further narrowed down to a small stretch of 35 amino acids (296-331) within the hinge region. It was striking that the similar stretch was identified as the binding site of the stimulatory receptor antibody. In pharmacology, an ‘inverse agonist’ is an agent which binds to the receptor and reverses the constitutive activity of receptors. Thus the hinge region of the receptor could be termed as a ‘tethered inverse agonist’ of the TMD, since it is covalently associated with the TMD and their interactions dampen the basal constitutive activity of the receptor. However, careful comparison of the activities of the mutants (receptors harboring deletions and gain-of-function mutations) with maximally stimulated wild-type FSHR indicated that these mutations of the receptor resulted only in partial activation of the serpentine domain suggesting that only the ECD in complex with the hormone is the full agonist of the receptor. Moreover, the hinge region stabilizes the TMD in an inactive conformation and the activating mutations disengage the inhibitory ECD–TMD interactions bringing about partial activation of the receptor. Most interestingly, the deletion of amino acids 296-331 from hFSHR resulted in no further response to the hormone indicating that this part of the receptor is also critical for hormonal activation, perhaps playing a dual role in the attenuation of the basal activity and a direct involvement in the hormonal activation of the receptor. Progressive sequential deletions of ten amino acids from 290 to 329 yielded similar results (high basal cAMP production with concomitant loss of hormone and antibody response) clearly demonstrating that the integrity of this region is absolutely essential for hormonal activation. In conclusion, the study provides a conclusive evidence to show that the hinge region of FSHR, although not involved in primary high affinity hormone binding, plays a critical role in the modulation of the receptor activity in absence, as well as, presence of the hormone. A large array of reproductive abnormalities is associated with malfunctioning of FSHR. To explore the possibility of using the stimulatory antibodies for therapeutic purpose, three inactivating mutations of hFSHR were analyzed. In corroboration with the earlier reports (Doherty et al, 2002; Touraine et al, 1999), the mutants A419T and L601V are incapable of transducing the signal, despite having adequate cell surface expression and wild type affinities for the hormone, mainly because of defective TMD. The RF2 antibody failed to elicit any response from these mutants suggesting that its ability to activate the receptor depends on the status of the TMD. Interestingly, the activating mutant D576G, which showed very high basal cAMP production, could be stimulated by both antibody and the hormone to the nearly wild type levels suggesting that in this mutant the interactions between the hinge region and TMD are similar to that of wild type and higher basal cAMP production could be due to different interactions of the TMD with the G-Proteins. Structure-function studies of glycoprotein hormones and their receptors have been hampered due to low levels of expression of the properly folded proteins in heterologous systems (Chazenbalk & Rapoport, 1995; Hong et al, 1999b; Peterson et al, 2000; Sharma & Catterall, 1995; Thomas & Segaloff, 1994). Previous studies from the laboratory have shown that the Pichiapastoris,which blends the advantages of both bacterial and mammalian expression systems, can be used to hyper-express biologically active hormones (Blanchard et al, 2008; Gadkari et al, 2003; Samaddar et al, 1997). In addition, the same expression system has been used to produce single chain hormone analogs (Roy et al, 2007; Setlur & Dighe, 2007). Further, methodologies for Pichiafermentation and purification of recombinant hormones from the fermentation media have been wellestablished in the laboratory. Chapter 4 describes the work carried out to express, purify and characterize a fully functional hFSHR extracellular domain. Thus a stage is now set to attempt structural studies with the receptor. The results are discussed at the end of each of these chapters and future directions have been discussed at the end of this thesis.