Minireview: Insights Into the Structural and Molecular Consequences of the TSH-β Mutation C105Vfs114X

Abstract

Naturally occurring thyrotropin (TSH) mutations are rare, which is also the case for the homologous heterodimeric glycoprotein hormones (GPHs) follitropin (FSH), lutropin (LH), and choriogonadotropin (CG). Patients with TSH-inactivating mutations present with central congenital hypothyroidism. Here, we summarize insights into the most frequent loss-of-function β-subunit of TSH mutation C105Vfs114X, which is associated with isolated TSH deficiency. This review will address the following question. What is currently known on the molecular background of this TSH variant on a protein level? It has not yet been clarified how C105Vfs114X causes early symptoms in affected patients, which are comparably severe to those observed in newborns lacking any functional thyroid tissue (athyreosis). To better understand the mechanisms of this mutant, we have summarized published reports and complemented this information with a structural perspective on GPHs. By including the ancestral TSH receptor agonist thyrostimulin and pathogenic mutations reported for FSH, LH, and choriogonadotropin in the analysis, insightful structure function and evolutionary restrictions become apparent. However, comparisons of immunogenicity and bioactivity of different GPH variants is hindered by a lack of consensus for functional analysis and the diversity of used GPH assays. Accordingly, relevant gaps of knowledge concerning details of GPH mutation-related effects are identified and highlighted in this review. These issues are of general importance as several previous and recent studies point towards the high impact of GPH variants in differential signaling regulation at GPH receptors (GPHRs), both endogenously and under diseased conditions. Further improvement in this area is of decisive importance for the development of novel targeted therapies.

Congenital hypothyroidism (CH) is the most prevalent endocrine disorder in newborns with an incidence of approximately 1:3500 (1). Most patients diagnosed with primary CH are characterized by an abnormal development of the thyroid gland (1). Secondary or central CH (CCH) is a rare disease with a reported incidence of approximately 1:30 000–1:50 000. However, these numbers may underestimate the disease rate as recent reports of a combined screening program on TSH and T₄ in Dutch neonates suggests an incidence of approximately 1:16 000 (2, 3). CCH is still considered as a rare but severe endocrine disorder (4).

Most CCH patients suffer from combined pituitary hormone deficiency syndrome either caused by a defect in pituitary gland development due to mutations in transcription factors such as pituitary-specific positive transcription factor 1 (Pit1), homeobox protein prophet of PIT-1 (Prop1), homeobox expressed in ES cells 1 (HESX1), LIM/homeobox protein Lhx3 (LHX3/4), or sex determining region Y (SRY)- box (SOX) 2/3 (SOX2/3), or by other syndromic causes such as the recently described mutation in the immunoglobulin superfamily member 1 gene (5–7). However, approximately one-third of patients with central hypothyroidism demonstrate an isolated TSH deficiency as a result of mutations in the genes encoding the TRH receptor (eg, Arg17*, Pro81Arg) (8) or the β-subunit of TSH (TSH-β) (Figure 1) (reviewed in Ref. 9). Of note, no mutation in human TRH has yet been described.

Figure 1. Schematic overview of genetic causes of CCH.

Classification of the reviewed mutation is marked in red. HESX1, homeobox expressed in ES cells 1; IGSF1, immunoglobulin superfamily member 1; LHX4, LIM/homeobox protein Lhx4; PIT1, pituitary-specific positive transcription factor 1; PROP1, homeobox protein prophet of PIT-1; (SOX)3, sex determining region Y (SRY)- box; TRHR, thyrotropin releasing hormone. receptor.

The phenotypes of patients with a mutated TSH-β gene are highly variable. Clinical signs of mild hypothyroidism include jaundice and macroglossia (10–12). In patients with severe CH, symptoms include muscle hypotonia, feeding problems, mental retardation, and a small hypoplastic thyroid gland (13–15). In the absence of thyroxine substitution at a very early stage, some patients develop a severe phenotype with irreversible mental and psychomotor impairments. Therefore, one of the biggest challenges is the early identification of such patients and the commencement of thyroid hormone replacement treatment immediately after birth. In most countries, the routine neonatal screening program for hypothyroidism is solely based on the detection of elevated levels of TSH. As a consequence, identification of newborns with CCH is often delayed compared with patients with primary CH (3, 9). Severe and irreversible damage of psychomotor development appears to be less common among such cases, because these forms of CCH only cause moderate thyroid hormone deficiency (1). However, in certain patients in whom thyroid hormone concentrations are very low, severe impairment of mental development has been observed. The reasons why some molecular defects, especially those of isolated TSH deficiency, lead to a complete loss of thyroid function is not understood, because a constitutive activity of the TSH receptor (TSHR) has been well described in vitro (16).

The first TSH-β subunit mutation (G29R) that results in CCH was described in the early 1990s (17–19). Since then, 9 additional mutations were identified in the TSHB gene of patients with CCH (10–15, 20–32). In these conditions, the term TSH deficiency defines the biological inactivity of TSH and/or an interrelated absence of TSH-β subunits in patients (33). The most common mutation that causes CCH due to isolated TSH deficiency is the C105Vfs114X mutation, which is located on exon 3 of the TSHB gene and was first described in 1996 (21). This mutation primarily occurs in nonconsanguinous families, but a founder effect has been proven (25). The ensuing sections will answer the following questions. What is currently known on the molecular background of this TSH variant on a protein level? What is the exact defect caused by this sequence modification?

Insights Into the TSH-β Mutant C105Vfs114X

TSH is a member of the heterodimeric glycoprotein hormone (GPH) group (34) and has a molecular weight of 28–30 kDa. It is synthesized by the thyrotrophs of the anterior pituitary gland (35). The hormones TSH, LH (or lutropin), FSH (or follitropin), and placental choriogonadotropin (CG) share a common α-subunit that is noncovalently associated with a hormone subtype specific β-subunit. The CGA gene encodes 92 amino acids of the α-subunit and is located on chromosome 6 (36). The TSHB gene of the TSH-β subunit was first cloned in 1988 (37) and is located on chromosome 1 (1p13). It encodes a polypeptide chain of 118 amino acids in length without the signal peptide (38).

Specific glycosylations are important for the correct functioning of GPHs, especially their bioactivity (reviewed in Ref. 34). GPHs lacking N-linked oligosaccharides act as antagonists (39). Furthermore, the 2 dominant G protein-mediated intracellular signaling pathways downstream of the TSHR, cAMP formation, and inositol phosphate (IP) release have been shown to be activated by diverse TSH glycosylation variants to various degrees (40). The finding that heterodimeric GPH variants exert different effects on signaling has encouraged initiatives to design hormone subunit derivatives based on the modification of the subunit structure including fused subunits (41–44), monomeric subunits (45, 46), or tetrameric forms (46, 47). However, none of these derivatives has been successfully used to antagonize or activate TSHR.

The C105Vfs114X mutation is characterized by a T-deletion in codon 105, which results in a frameshift and premature stop codon at position 114. Consequently, the C105Vfs114X mutant lacks 5 amino acids in C-terminal length compared with the wild type. Cysteine 105 (numbering without the start sequence) is replaced by valine and the 8 consecutive C-terminal amino acids also differ to the wild type (Figure 2). In summary, this TSH mutation causes a multitude of modifications including amino acid deviations, shortening in length, and modification in the disulfide-bridge pattern. This aspect could be of importance as GPHs are structural members of the cystine knot growth factor superfamily, where disulfide bridges are of fundamental importance in their folding and function (35, 48). This spectrum of putative effects at TSH that are caused by this mutation raises questions of the reasons behind the observed molecular dysfunction and phenotype in patients. Two hypotheses can be put forward:

Figure 2. Amino acid sequence alignment of the different GPH-β-subunits.

This amino acid comparison shows similarities and differences between the β-subunits of the GPH-β-subunits (LH, FSH, thyrostimulin [thyrost], and CG), whereby TSH-β subspecies of human, bovine, and recombinant TSH-β (TR1401) were included. A color code facilitates easy comparison between the sequences. Of note, in the pathogenic TSH-β variant C105Vfs114X, the genetic modification causes substitution of cysteine to valine (red arrow at the C terminus), which simultaneously leaves the interacting cysteine in the wild type (cysteine at the N terminus, red arrow) without an interaction partner. In thyrostimulin, both cysteines are endogenously absent, whereby this ancestral hormone shows high binding and signaling capacity at the TSHR (53). Moreover, thyrostimulin is 20-amino acids shorter at the C terminus compared with human TSH. This suggests that shortening of the C terminus is not causative per se for a defect in hormone activity, as should be assumed for the pathogenic TSH variant with a modified C terminus including 5 amino acids less than the wild type. Accession numbers: bovine TSH-β, GenBank AAA30796.1; human TSH-β, GenBank AAB30828.2; human thyrostimulin-β, GenBank AAO33390.1; human CG-β, GenBank AAA96690.1; human LHβ, GenBank AAI60107.1; human FSHβ, GenBank AAB02868.1.

Hypothesis 1

The modified length of TSH-β leads to impaired function such as decreased binding to the receptor or a diminished constitution of intact heterodimeric TSH.

Hypothesis 2

Modified single amino acids in the C terminus have a negative effect on TSH heterodimer formation or TSH/TSHR interaction.

In the following section, we will reflect and discuss the aforementioned hypotheses. Moreover, we aim to implement new results from a comparative perspective in addition to structural information to our analysis.

What Are the Putative Consequences of the C105Vfs114X Mutation at the Structural and Functional Level?

Hypothesis 1

Consequences of shortening the TSH-β C terminus. For the TSH C105Vfs114X mutation (and other pathogenic TSH variants), loss of the heterodimeric TSH structure has been suggested to lead to TSH deficiency (eg, Refs. 15, 29, 31), as the shortened C terminus modifies the so-called “seat-belt,” which is important for the constitution of heterodimeric GPH structure (reviewed in Ref. 34).

In 1996, Medeiros-Neto et al (21) conducted experiments on specific TSH-β variants with a C terminus truncated after the cysteine at codon 105 and the tyrosine at codon 112. The aim was to determine and compare the ratio of biological activity (at the TSHR) with the immunological activity of the wild-type and modified TSH. The authors of this previous study (21) suggested that C-terminal 13-amino acid deletion from the TSH-β subunit does not result in the observed loss of biological activity. Moreover, C-terminally truncated TSH variants are supposed to have a higher biological to immunological ratio compared with the C105Vfs114X mutation and even the wild type (21).

Importantly, only limited quantities of the unpurified C105Vfs114X-truncated mutant have been previously produced and characterized. Rigorous assessment of heterodimer concentration by mass is lacking in addition to studies on its structure, heterodimerization, heterodimer stability, N-glycosylation analysis, TSHR binding, biased signaling, and receptor specificity, which are justified by the location of the C105Vfs114X frameshift mutation and its proximity to the seat-belt region. Interestingly, C105Vfs114X was expressed in a stable HEK293 cell line and secreted high levels of free α-subunit, which is necessary in preventing intracellular degradation of the free wild-type TSH-β subunit (49). Furthermore, it may require higher levels of the α-subunit in impaired heterodimerization with TSH-β C105Vfs114X, as suggested by compensatory increases of α-subunit levels and α to TSH ratios in C105Vfs114X patients (21).

Interestingly, thyrostimulin, the second endogenous TSHR agonist and ancestral GPH (50, 51), enables further insights into this topic (Figures 2 and 3), as the C terminus of mature native thyrostimulin-β glycoprotein-beta5 (GPB5) is shorter (20 amino acids) than the C terminus of TSH-β. Anyhow, tyrostimulin is approximately 4 times more active than bovine TSH, although the capacity to form heterodimers is decreased compared with TSH (52). For thyrostimulin, the missing seat-belt structure (Figure 3) was discussed to potentially prevent a proper heterodimer constitution. Of note, significant signaling capacity for single GPH subunits, including those of thyrostimulin, has not yet been demonstrated; therefore, single subunit interactions with TSHR are incompatible with the observed high activation effects of thyrostimulin. Thyrostimulin shows receptor selectivity and does not activate eg, the follitropin receptor (FSHR), lutropin/choriogonadotropin receptor (LHCGR), or leucine-rich repeat containing receptors (LGRs) 4–8 (50, 53–55).

Figure 3. Structural comparison between human TSH and thyrostimulin.

Thyrostimulin (A) and TSH (B) are shown in a surface representation and in a putative bound state to the extracellular domain (ECD) of the TSHR. The orientation of the hormones towards the receptor is predicted according to the previously published crystal structure of the FSHR ECD/FSH complex (110). The models were designed as already described for hTSH (95) and thyrostimulin (96). Highlighted are specific differences such as the seat-belt region of hTSH, which is not present in thyrostimulin. Cysteine 105, which is modified in the TSH variant C105Vfs114X, constitutes a disulfide bridge with Cys19 (N terminus). The TSH homology model is not completed at the C terminus of the β-subunit (red), because all available GPH crystal structures serving as homology model templates (eg, FSH; pdb codes 1QFW [111], 1FL7 [112], 1XWD [113], 4AY9 [110] and CG; pdb code 1HRP [114]) terminate immediately after the last disulfide bridge. Remaining amino acids are therefore not included in the model. This is also one additional reason why a molecular homology model of the pathogenic TSH variant with modified amino acids at the TSH-β C terminus cannot be provided so far. However, the surface representation impressively demonstrates the potential impact of the seat-belt, whereby this fragment of the TSH-β subunit tightly arranges “around” parts of the α-subunit (reviewed in Ref. 48). In the bound state, it is evident that the seat-belt might be involved in the determination of hormone specificity, as previously reported (115). However, without this seat-belt region, thyrostimulin seems to be receptor-selective (50, 53–55). Moreover, this scheme suggests that the extreme C terminus, even if partially visible, is not directed towards the ECD in this monomeric TSHR constellation and may therefore not participate in direct receptor interactions. The software PyMOL (Molecular Graphics System, version 1.3; Schrödinger, LLC) was used for all structural representations.

The location of a single N-linked carbohydrate chain differs in TSH-β and thyrostimulin (GPB5). In TSH-β, the N-linked glycosylation sequence NTT (residues 23–25) is positioned close to Cys 19. In thyrostimulin (GPB5), the NET sequence is located within the β-3 strand (residues 88–90), which may help explain the variable degree of heterodimer instability among GPH variants missing seat-belt structure and “latching.” In conclusion, the occurrence of thyrostimulin heterodimerization, also in the absence of a seat-belt structure and stabilizing C-terminal disulfide bridge, suggests that the formation of a biologically active heterodimeric hormone (Figure 3) depends on different features. The hypothesis that subunit assembly is presumably not drastically interrupted by shortening of the GPH-β-subunit is further supported by studies on CG deletion constructs (56).

Hypothesis 2

Modifications of amino acid composition at the TSH-β C terminus. To date, no reports have focused on the functional impact of the 8 modified amino acids at the C terminus of the pathogenic TSH-β variant consecutively after the substituted cysteine 105 (Figure 2). However, there are data available on the Cys105Val mutation and frameshift deletion. The Cys105Val mutation has been suggested to significantly reduce bioactivity and expression levels (21).

A simple 113–118 deletion was not observed to affect the bioactivity of recombinant TSH, suggesting that the last 6 amino acid C-terminal residues are not important in hormone biosynthesis and function (57). Furthermore, detailed studies on the disulfide bridges at the human CG-α (58) and β-subunits (59) were conducted in the early 1990s. Mutations at cysteines were shown to affect human CG β-subunits folding. In detail, the intracellular folding pathway of hCG-β is characterized by a rapid sequence of events and intermediate conformational states, which are each highly dependent on the order of cysteine linkage. Both α- and β-subunits are folded into nonglobular cystine knot structures, with 3 loops extending from the core motif containing 3 knotting disulfide bridges (34, 35, 48, 60). Prevention of specific intracellular cysteine linkage by site-directed mutations has been found to interrupt the proper folding process and was accompanied by intracellular localization and partially folded protein structures (59). Mutations in CG-β at the Cys26-Cys110 disulfide bridge, which corresponds to the Cys19-Cys105 interaction in TSH-β, were proposed to inhibit the conversion of folding intermediates to a final state, whereby disulfide bond formation usually occurs at a late stage.

Xing et al (61) reported on the assembly of hCG subunits to heterodimers before stabilization of the seat-belt structure by the C-terminal disulfide bridge. These findings suggest a significant role of this specific disulfide bridge between the N and the C termini and have also been confirmed in detailed studies concerning seat-belt function on hCG (61–63). Finally, heterodimer constitution and participating determinants and mechanisms have been described in detail for hCG, but differences between the diverse GPHs may exist (64).

The comparison of TSH with thyrostimulin delivers a further important hint on the putative molecular mechanism of the TSH-β mutant C105Vfs114X, especially in relation to the aforementioned functions of disulfide bridges during subunit folding and assembly. The pathogenic genetic modification causes a substitution of cysteine 105 with valine, which simultaneously would leave an unpaired cysteine at the N-terminal Cys19 (Figures 2 and 3). In thyrostimulin, both corresponding cysteines are endogenously absent (also the cysteine corresponding to Cys19 in TSH) (Figure 2), whereby this ancestral hormone has demonstrated high binding and signaling capacity in in vitro studies (53). It can be assumed that replacement of one cysteine side chain, such as for the TSH-β mutant C105Vfs114X (and interruption of the disulfide bridge), results in the ability of the remaining cysteine to interact with a different cysteine of the β-subunit (false positive disulfide bridge) in the TSH variant, which may dramatically impact the local structure or entire protein due to an altered disulfide-bridge pattern (59). This conclusion would be in accordance to previously reported effects of the mutant Cys105Val (21). The Cys105Val substitution has been suggested to cause a defect in TSH structure, presumably by modifying the folding process, as observed for CG (59).

Anyhow, it remains unclear how the modified C terminus with different amino acids impacts the TSH structure or TSH/TSHR interplay. This may be examined by directed mutagenesis, whereby a careful examination and analysis of biological activity and immunoreactivity of any TSH-β mutant is demanded. Failure to bind a specific monoclonal antibody for the detection of TSH was shown to be a direct consequence of amino acid substitution (Arg55Gly) (65).

Because of this information-gap on TSH-β mutants, we analyzed current knowledge regarding the impact of pathogenic naturally occurring mutations from cases reported for the homologous GPH, FSH, LH, and CG β-subunits. Naturally occurring mutations in the gonadotropic hormone β-subunits are quite rare (66, 67). One of the reasons for this is the induced infertility of affected individuals in gonadotropic hormone variants. Such mutations have been found in both males and females (68, 69). In males, sexual differentiation is normal; however, induction of puberty and spermatogenesis is restrained. Affected women are infertile and puberty proceeds in a normal manner. It must be noted that there is limited information concerning the molecular mechanisms on a structural-functional level, with only few exceptions, eg, for the human CG β-subunit (70). To date, most interpretations for the pathogenic molecular mechanisms were indirectly extrapolated from the very few in vitro studies available (eg, Ref. 59). Hypogonadism caused by a single amino acid substitution in the β-subunit of hLH has been previously reported and potentially leads to a loss of binding capacity to the receptor (eg, Qln54Arg) (71). Hypogonadotropic hypogonadism in a patient with 2 mutations in the LH-β gene expressed in a compound heterozygous state was recently identified (72). This finding is most likely due to a different molecular mechanism by preventing heterodimerization of hormone subunits, or as a result of missfolded protein structure.

In females, FSH is generally required for follicular growth, estrogen production, and oocyte maturation. Absence of FSH results in delayed puberty and infertility. In 1993, a homozygous frameshift mutation in FSH-β was reported in a female and led to primary amenorrhea and infertility (73). Interestingly, this mutation caused shortening of the C terminus and an altered amino acid constitution in FSH, which is principally comparable with the consequences of the TSH-β C105Vfs114X mutation. Based on previous insights on gonadotropic hormones, the authors of this study conclude that associations between α- and β-subunits are abolished in this pathogenic FSH variant. This hypothesis was also proposed for the pathogenic Cys54Gly variant (74). An example of a FSH-β mutation in males has been reported (75) and the patient showed normal puberty and virilization with underdeveloped testicles and azoospermia. Interestingly, a cysteine substitution (Cys82Arg) was also found in this patient, and the disruption of heterodimerization between the hormone subunits and/or modulation of the tertiary protein structure was favored as the molecular mechanism of dysfunction. Anyhow, this hypothesis awaits experimental validation.

Previous studies on known LH and FSH mutations have helped in better understanding the physiological impact of these variants and in complementation to respective receptor knock-out and knock-in mice (eg, Refs. 66, 67, 76–79) or other animal models (80–82). However, as has been determined for TSH mutations, very little is known on the molecular-structural effects caused by these mutations. It is therefore of importance to critically investigate the mechanisms, decipher putative links between hormone-induced or interrupted signaling, and their physiological effects (eg, Refs. 83–85). Moreover, detailed determination of the GPHR/GPH interplay under pathogenic conditions should assist in elucidating the natural mechanisms causing the observed disease symptoms. This may also aid in planning directed interventions by specifically designed hormone derivatives.

What Can Be Hypothesized Concerning the Interaction Between TSH Variants and the TSHR?

The activation mechanism is likely similar for all GPHRs with few specificities for each specific receptor/ligand pair (79, 86–90). The 2 distinguishable extracellular parts of the receptor, the leucine-rich repeat domain (LRRD) and the hinge region, harbor determinants for hormone binding (reviewed in Refs. 86, 87). The hinge region structurally links the LRRD with the serpentine domain (Figure 4) (48). Two disulfide bridges between the C-terminal LRRD and the C terminus of the hinge region are obligatory for correct receptor function. Hormones bind to the LRRD and hinge region (binding sites 1 and 2) (Figure 4), which triggers conformational changes at a convergent center of the LRRD and hinge region. An inhibitory function (91, 92) of the extracellular part is potentially reversed by this interaction (93). Furthermore, an “intramolecular agonistic unit” or “internal agonist” close to the transmembrane domain 1 is activated (94–96). This signal is conveyed via structural rearrangements of the transmembrane spanning-helices towards the intracellular site (97). The largest spatial movement affects transmembrane helix 6 (TMH6), involving a combination of horizontal and rotational movements around a pivotal helix-kink at the highly conserved Pro639 (97, 98). An “active” receptor conformation opens a spatial crevice for a fit-in of intracellular interaction partners and is stabilized by interaction with a G protein (99) or arrestin molecule. In summary, a highly ordered sequence of structural rearrangements and interaction events between different proteins (hormone/receptor/effector) defines the activation process. Any modification, eg, the substitution of interacting amino acids or slight alterations in receptor structure, may affect interaction and signaling, which is supported by a huge number of pathogenic and designed inactivating or activating receptor mutants (100–103).

Figure 4. Human TSH bound to the TSHR/G-protein complex.

The individual structural homology models in this complex are based on structural templates from different family A GPCR crystal structures (serpentine domain, cartoon, and translucent surface) (reviewed in Ref. 87). A detailed description of the TSHR/TSH complex modeling is provided in Brüser et al (94). Furthermore, the Gs protein was implemented in this current TSHR/TSH model as suggested by the structural template used for the transmembrane domain (β-adrenergic receptor 2, pdb code 3SN6) (116). Structural templates for GPHR modeling approaches are complexes between the hormone FSH and the FSHR ectodomain (110, 113, 117) or the partial TSHR ectodomain bound with an activating or inactivating antibody (118, 119). Specific emphasis points to the bound TSH (cartoon) with highlighted disulfide bridges (spheres). Cysteine 105 interacts with cysteine 19 (numbering without signal sequence). In the mutation Cys105Val, this disulfide bridge will be interrupted and results in so far unknown structural modifications, either by a new artificial disulfide bridge or modification of the local structure in close proximity of position 105.

These structure-function relationships also suggest that mutated hormones, that still bind to the receptor, should mediate different functional effects (61). Importantly, the capacity for binding and signal induction are distinguishable parameters. This is shown by comparing activating and inactivating TSH antibodies. Inactivating antibodies either block TSH binding or are associated with a receptor in a constraining mode, which leads to inverse agonistic effects (reviewed in Refs. 104–109). For TSH variants, such varying effects might be feasible due to their specific modifications. Mutations at binding-sensitive regions would interfere with binding at the receptor; however, in maintained binding, a blockade of receptor activation or basal signaling activity may occur, which is relevant for TSHR (16) and LHCGR/FSHR functions (79).

The 3-dimensional receptor/ligand complex model (Figure 4) revealed that despite many remaining uncertainties regarding the exact receptor complex arrangement, specific contacts between the hormone and receptor are responsible for direct receptor activation, most likely at several different regions (reviewed in Ref. 96). Assuming an altered TSH protein structure caused by mutation, the mode of binding should differ from the wild type and may cause differences in the resulting receptor signaling output. From a structural perspective the complex nature of TSHR/TSH or GPHR/GPH interactions and interrelations also implicates a multitude of potential mechanisms that may account for signaling properties in dependence of ligand structure. “Inactivation” caused by GPH variants may result from binding and inhibition of the constitutive basal signaling capacity of the receptor, whereby these hormone variants may act as inverse agonists.

Conclusions

The present review summarized and discussed the molecular mechanisms of naturally occurring GPH variants and extrapolated current knowledge on the TSH-β variant C105Vfs114X. Despite the availability of several insights on the structure-function relationship of this variant and other GPH mutants, it has not yet been characterized how such mutations and frameshift deletions modify protein structure or impact the signaling spectrum mediated by GPHRs. Nevertheless, the mutant Cys105Val appears to negatively influence TSH structure and biological activity. Specific shortening of the C-terminal TSH-β alone does not lead to a dramatic decrease in activity. Moreover, it remains unresolved how different TSH-β amino acids at the C terminus modify receptor signaling or TSH activity and how the Cys105Val mutation impacts TSH properties. Focused investigation of these issues will support ideas for GPH variant designs with directed properties for the diagnosis and treatment of hypo- or hyperthyroidism with higher precision and in a more personalized manner.

Funding Statement

This work was supported by the Deutsche Forschungsgemeinschaft Projects KL 2334/2–2, BI893/5–2, and BI 893/6–3.