Abstract

Mutations in the melanocortin-4 receptor gene (MC4R) represent the commonest monogenic cause of human obesity. However, information regarding the precise effects of such mutations on receptor function is very limited. We examined the functional properties of 12 different mutations in human MC4R that result in severe, familial, early-onset obesity. Of the nine missense mutants studied, four were completely unable to generate cAMP in response to ligand and five were partially impaired. Four showed evidence of impaired cell surface expression and six of reduced binding affinity for ligand. One mutation in the C-terminal tail, I316S, showed reduced affinity for α-MSH but retained normal affinity for the antagonist AgRP. None of the mutations inhibited signaling through co-transfected wild-type receptors. Thus, in the most comprehensive study to date of the functional properties of naturally occurring MC4R mutations we have (1) established that defective expression on the cell surface is a common mechanism impairing receptor function, (2) identified mutations which specifically affect ligand binding affinity thus aiding the definition of receptor structure-function relationships, (3) provided evidence against the notion that these receptor mutants act as dominant-negatives, and (4) identified a potentially novel molecular mechanism of receptor dysfunction whereby a mutation alters the relative affinities of a receptor for its natural agonist versus antagonist.

INTRODUCTION

The central melanocortinergic pathways play an important role in the control of mammalian energy homeostasis, as has recently been highlighted by genetic studies in obese humans and mice. Disruption of the pro-opiomelanocortin (POMC) gene in man and mouse results in an obese phenotype (1,2), as does the transgenic overexpression of either of the melanocortin antagonists, agouti (3,4) or agouti-related protein (AgRP) (5). Additionally, murine and human obesity syndromes resulting from mutations in enzymes involved in Pro-peptide processing are associated with markedly impaired processing of POMC into the appropriate melanocortin peptides (6,7). Of the known receptors for melanocortin peptides only two, the melanocortin-3 and -4 receptors (MC3R and MC4R), are highly expressed in the central nervous system (811).

The MC4R is a seven-transmembrane G-protein coupled receptor principally expressed in the brain, including the hypothalamus (10,11). Direct evidence that the MC4R is a key regulator of appetite and body weight was provided by Huszar and colleagues, who demonstrated that mice with a targeted disruption of the Mc4r gene have increased food intake, obesity and hyperinsulinaemia (12). Animals heterozygous for the null allele showed an intermediate obesity phenotype when compared with their homozygous and wild-type littermates.

The importance of the MC4R in the regulation of human body weight became apparent in 1998 when mutations in the human MC4R were first described as a cause of human obesity. Yeo et al. (13) and Vaisse et al. (14) reported heterozygous frameshift mutations in the human MC4R that co-segregated in a dominant fashion with severe early-onset obesity. Subsequently, multiple different missense and nonsense mutations in MC4R have been reported, largely in subjects with severe obesity commencing in childhood (1518). MC4R mutations now represent the commonest known monogenic cause of non-syndromic human obesity.

A better understanding of the functional consequences of naturally occurring mutations in the human MC4R is important for a number of reasons. Firstly, not all amino acid changes thus far reported in the human MC4R are actually associated with obesity. In three such cases (V103I, I251L and T112M) examination of the signalling properties of these relatively common variant receptors have revealed them to be indistinguishable from wild-type (15,19). As MC4R sequence determination is increasingly entering the realm of clinical diagnostics it will be crucial to be able to clearly distinguish between pathogenic mutations and functionally neutral variants. Secondly, the MC4R is a highly attractive potential therapeutic target for the pharmacotherapy of obesity. Naturally occurring mutations, known a priori to be highly likely to be of functional significance because of their association with obesity, can provide valuable insights onto the structure/function relationships of this important receptor in vivo.

Having detected a large number of different MC4R mutations in a cohort of subjects with severe early-onset obesity, we now report on the functional properties of 12 of these mutants.

RESULTS

Identification of MC4R mutations

The mutant MC4Rs chosen for functional study were detected in a cohort of 350 unrelated probands with severe early-onset obesity. All of the mutations studied were considered to be highly likely to be causative, rather than incidental as (a) none were found in 108 control alleles and (b) the mutations were non-conservative and/or co-segregated with early-onset obesity in other family members (personal observations). Twelve different mutations found in 16 unrelated probands, five of which we have previously reported [‘CTCT’ deletion at codon 148 (13), N62S, R165Q, V253I and C271Y (17)]. were selected for functional characterisation. Two of the mutants, ‘A’ ins at codon 112 and ‘CTCT’ deletion at codon 148, lead to a frameshift and premature termination, Y287X is a nonsense mutation leading to a premature termination, and nine are missense mutations (see Fig. 1 and Table 1). N62S, N97D and Y287X were found as homozygotes. I316S was found in homozygous form in one proband and in heterozygous form in two other probands. The rest of the mutations were found in heterozygous form.

Functional studies of mutant receptors

The ability of all mutant receptors to generate cAMP in response to increasing concentrations of α-MSH was studied in transiently transfected HEK293 cells. The binding of radiolabelled NDP-MSH tracer to transfected HEK293 cells was also studied for all the receptors. A subset of receptors were epitope-tagged to allow studies of trafficking to the plasma membrane to be undertaken, these included mutants which showed no response in the cAMP assay (‘A’ ins at codon 112, Y287X, N97D, L106P, I125K and C271Y) and a missense mutation in the c-terminal tail (I316S), as this has previously been identified as a region important for targeting to the plasma membrane (19). Studies of receptor binding affinity for NDP-MSH and AgRP were undertaken for those mutations that show a partial response to α-MSH (N62S, R165Q, A175T, V253I and I316S) and for those mutations that we showed were present on the cell surface (N97D, L106P and I125K).

For clarity of presentation we will first describe the findings with frameshift and nonsense mutations, then the missense mutations which showed no ability to generate cAMP and finally those missense mutations which retained some, albeit impaired, ability to generate cAMP in response to ligand.

Frameshift and nonsense mutations

As expected, the two frameshift and one nonsense (Y287X) mutations were unable to generate any cAMP in response to ligand (Fig. 2A) and cells transiently transfected with these mutant receptors showed binding of [125I]NDP-MSH, a high-affinity cyclic analogue of α-MSH (20), that was indistinguishable from mock-transfected cells (Fig. 2B). Both the Y287X mutant, which lacks only the C-terminal tail and the seventh transmembrane domain and the ‘A’ insert at codon 112 were N-terminally MYC-tagged, transfected into COS-7 cells and the location of transfected receptors was examined by immunofluorescent microscopy using anti-MYC antibodies. No cell surface staining was seen with either of these mutant receptors (Fig. 2C). Additionally, with the ‘A’ insert at codon 112 mutant we also observed markedly reduced immunostaining of intracellular receptor suggesting that this prematurely truncated receptor fragment may be inefficiently translated, rapidly degraded, or both.

Functionally inactive missense mutations

The C271Y mutant receptor also shows a complete inability to signal to cAMP (Fig. 2A) as well as a marked inability to bind tracer (Fig. 2B). When an N-terminally MYC-tagged version of this receptor was transfected into COS-7 cells, no cell surface immunofluorescence was seen. Thus, this mutation appears to result in a profound defect in trafficking to the plasma membrane which is likely to be the major reason for its impaired signalling (Fig. 2C).

Three further missense mutations, N97D, L106P and I125K, showed a complete inability to generate cAMP in response to ligand (Fig. 3A). Cells expressing these mutant receptors bound appreciable but markedly reduced (<20% of wild-type) amounts of tracer concentrations of [125I]NDP-MSH (Fig. 3B). COS-7 cells transfected with N-terminally MYC-tagged versions of all three receptors showed visible plasma membrane staining in all cases, but this was qualitatively reduced compared to cells transfected with wild-type receptor (Fig. 3C).

To assess the ligand binding affinity of those mutant MC4Rs that were expressed on the cell surface, we undertook competitive displacement studies using [1251]NDP-MSH. Whole HEK293 cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in the competitive binding assays. Cells were exposed to a fixed amount of [125I]NDP-MSH, and the ability of increasing concentrations of unlabelled NDP-MSH to displace radioligand binding was measured. N97D, L106P and I125K all severely affected binding of NDP-MSH (Fig. 3D and Table 2).

We also assessed the ability of AgRP(87–132) to displace [125I]NDP-MSH binding to the different MC4R mutants. AgRP(87–132) is a truncated AgRP variant, and has the same ability to bind MC4R and functionally inhibit melanocortins as the full-length molecule (21). HEK293 cells were transiently transfected with vectors expressing wild-type or mutant MC4Rs, and the ability of increasing concentrations of unlabelled AgRP(87–132) to displace [125I]NDP-MSH binding was measured. L106P and I125K severely affected the ability of AgRP(87–132) to displace NDP-MSH, while N97D exhibited a ∼57-fold decrease in binding affinity, respectively (Fig. 3E and Table 2).

Thus, although a partial defect in cell surface expression could contribute to the defective signalling capacity of these mutant receptors, it is likely that the major defect in signalling seen with N97D, L106P and I125K is likely to result from their severely impaired capacity for ligand binding. Of note, N97D, L106P and I125K are in the second transmembrane domain, first extracellular loop and third transmembrane domain respectively (see Fig. 1), regions previously shown to be critical for ligand binding (22).

Partially active missense mutations

N62S and R165Q showed a partial cAMP response to α-MSH (Fig. 4A) and cells transfected with these mutant receptors bound tracer at ∼20 and ∼30% of wild-type, respectively (Fig. 4B). Competitive displacement studies showed that N62S severely impaired affinity for NDP-MSH and reduced AgRP affinity by ∼17-fold. R165Q resulted in a receptor with a 17-fold decrease in affinity for NDP-MSH and an even more severe impairment in AgRP binding (Fig. 5B and Table 2). Thus R165Q and N62S's blunted response to ligand can be explained by a marked reduction in ligand binding capacity, although a partial defect in cell surface expression cannot be ruled out. Interestingly, N62S lies in the first transmembrane domain, a region not previously considered to be critical for ligand binding (22).

A175T and V253I both showed normal or close to normal ligand binding affinity for both NDP-MSH and AgRP (Fig. 5A and B and Table 2), despite their reduced ability to generate cAMP in response to ligand (Fig. 4A). Cells transfected with A175T and V253I bound tracer concentrations of [125I]NDP-MSH at ∼70 and ∼80%, respectively, compared with cells transfected with wild-type receptor. While the slightly reduced tracer binding seen with A175T and V253I suggests the possibility of a modest impairment in cell surface expression, it seems unlikely that this alone would be responsible for the impairment in receptor signaling.

I316S was found in three probands within our cohort of severely obese children, once in homozygous form, and twice in heterozygous form. It exhibits a partial response to α-MSH (Fig. 4A), and a 30% ability to bind tracer as compared with wild-type (Fig. 4B). However, it is unique in its differential effects on binding affinity for NDP-MSH versus AgRP. I316S resulted in a ∼14-fold decrease in binding affinity for NDP-MSH (Fig. 5A). Notably, I316S did not adversely affect the ability of the receptor to bind AgRP(87–l32), indeed, if anything, affinity appeared to be somewhat enhanced (Fig. 5B).

The effect on binding affinity resulting from I316S is particularly surprising given its location in the C-terminal tail of the receptor. Indeed this residue is localized within the putative plasma membrane targeting motif in the C-terminal tail (19). We therefore studied the cell surface expression of I316S using MYC tagged I316S. Cell surface immunostaining of this mutant seemed qualitatively similar to that seen with wild-type receptor (Fig. 4C).

Mutant MC4Rs do not affect the function of co-transfected wild-type MC4Rs

Most of the MC4R mutations described here were found in heterozygous form, and co-segregate with an obese phenotype in a dominant fashion. It is plausible that such mutant receptors might exert their deleterious effects on in vivo signalling through dominant negative effects on co-expressed wild-type receptors. Thus, in order to determine if any of the mutant MC4Rs exhibit any dominant negative activity, we performed cotransfection studies with either wild-type+wild-type constructs, or wild-type+mutant constructs, such that the amount of DNA transfected each time was identical. Pilot experiments were undertaken to ensure that supramaximal concentrations of transfected DNA were being used (data not shown). The cells were then exposed to 1×10−7M α-MSH, and the generation of cAMP was assayed as described above. None of the eight mutant receptors studied appeared capable of interfering with signalling through co-transfected wild-type receptor (Fig. 6). As our studies were confined to maximally effective ligand concentrations we cannot exclude the possibility of dominant negative interactions at lower ligand concentrations. However, our data is consistent with the notion that haploinsufficiency is the most likely explanation for the clinical phenotype of heterozygous mutations.

DISCUSSION

The MC4R plays a central role in the control of energy homeostasis, with a delicate balance between α-MSH and AgRP action at this receptor exerting influences on body weight through effects on both energy intake and energy expenditure. A number of groups have used a mutagenesis approach to identifying amino acid residues critical for receptor function in both the murine and human MC4R (22,23). To date, however, very little has been reported regarding the functional consequences of naturally occurring mutations.

We report the functional properties of 12 naturally occurring human MC4R mutations that are associated with severe early-onset obesity. These 12 mutations were identified in 16 out of 350 unrelated severely early-onset obese patients. As the determination of the nucleotide sequence of the MC4R is now beginning to enter the realms of clinical diagnostics for severe obesity, it is critical that pathogenic mutations be distinguished from functionally insignificant variants (V103I, T112M and I251L). Reassuringly, all of the mutations that we detected in obese subjects which (a) resulted in a non-conservative amino acid change and (b) were absent from a control group of normal weight, showed reduced ability to generate cAMP when transfected into HEK293 cells. Conversely, all the variants that we have detected which are present in non-obese control subjects have been demonstrated either by ourselves or other workers to signal normally to cAMP generation (15,19). Thus, this appears to be a robust assay for the determination of the pathogenicity of naturally occurring variants in the MC4R.

The great majority of MC4R mutations reported to date in obese humans have been found as heterozygotes. These include nonsense, frameshift and missense mutations. This raises the possibility that such mutations might result in impaired MC4R signalling through dominant-negative mechanisms, although previous studies have not supported this notion (19). We examined the abilities of eight of our mutations to interfere with signal transduction through co-transfected wild-type receptor and also found no evidence to support the idea of dominant negativity. However, studies were only performed at maximally effective agonist concentrations and it is possible that such effects might be seen at lower levels of agonist. Nevertheless, the fact that a variety of nonsense and frameshift mutations are associated with dominantly inherited human obesity and the knowledge that mice which are heterozygous for a null Mc4r allele are significantly obese provides support for the idea that haploinsufficiency is the likely mechanism for obesity associated with heterozygous mutations in MC4R.

All of the frameshift and nonsense mutations studied, as expected, resulted in mutant receptors which were completely unable to generate cAMP in response to ligand. Of particular note in this regard is the Y287X mutation, which results in a premature truncation within the final transmembrane domain of the receptor. This mutant receptor showed normal levels of expression intracellularly but no expression on the cell surface. These findings are consistent with those of Ho and Mackenzie who demonstrated that an artificial mutant receptor lacking amino acids 306–319 within the C-terminal tail was also unable to traffic to the plasma membrane (19).

Missense mutations impaired signalling through a variety of adverse effects on trafficking, ligand binding and the coupling of ligand binding to signal transduction. One missense mutation, C271Y, resulted in the complete inability of the receptor to be expressed on the cell surface while three other functionally dead missense mutations that were studied showed qualitatively reduced cell surface staining, despite levels of total cellular expression similar to wild-type. Two further mutants showed reductions in tracer binding that were not fully explained by effects of the mutations on ligand binding affinity, suggesting a reduced level of cell surface expression. Thus, impairment of receptor expression at the plasma membrane appears to be a common mechanism whereby naturally occurring mutations in MC4R adversely affects receptor function. The mechanism of mutations causing impaired receptor expression being associated with a disease phenotype is not unique to MC4R. For example, mutations in a variety of G-protein coupled receptors affecting cell surface expression are known to cause autosomal dominant retinitis pigmentosa (24), X-linked nephrogenic diabetes insipidus (25) and Hirschsprung's disease (26).

All of the mutations occurring within the region between the first transmembrane domain (TM1) and the start of the fourth transmembrane domain (TM4; N62S, N97D, L106P and I125K; see Fig. 1) severely impaired NDP-MSH binding, AgRP binding and cAMP generation. The functional consequences of this group of mutations are largely consistent with previous mutagenesis studies using both murine and human MC4R, which have shown this region of the MC4R to be critical for both α-MSH and AgRP binding (22,23). Interestingly, the homologous mutation of L106P in murine Mc1r, L98P, results in a constitutively active MC1R. This mutation is the cause of the ‘Somber’ (Esom) phenotype of dark murine coat colouration (27). There was no evidence for constitutive activity of the L106P mutant in our hands.

Another of our mutations, I125K, is in close proximity to acidic residues D122 and D126. Yang et al. (22) showed that mutation of D122 to alanine significantly decreased the binding affinity and potency of NDP-MSH, while mutation of D126 to alanine completely abolishes ligand binding and significantly decreases AgRP(87–132) binding affinity. It has been shown that D122 is the critical residue that interacts with the core melanocortin peptide residue, Arg8 (22).

While the effects of the mutations discussed above were largely consistent with previously reported results in artificial mutants, a number of the missense mutations that we studied had functional consequences that were less expected. Thus, N62S lies in a region of the first transmembrane domain that has not previously been implicated in ligand binding, a conclusion based on the fact that mutations of either D49 or D61 to alanine had no significant effect on NDP-MSH agonist binding affinity (22). Also of note is the fact that, while we found that the alteration of arginine at 165 to glutamine (R165Q) results in a partially signalling receptor with a marked reduction in binding affinity, and Vaisse et al. (18) has reported an R165W mutation that also impairs receptor signalling, previous studies of an artificial R165A mutant showed it to have normal function (22).

Only one mutation studied is predicted to occur in intracellular elements of the receptor. Surprisingly, it appears to have effects on ligand binding. The I316S mutation is located within the putative cell surface targeting motif of the intracellular C-terminal domain (19). However, N-terminally tagged I316S receptors showed relatively normal levels of cell surface expression. Intriguingly, while its binding affinity for NDP-MSH was reduced, its affinity for AgRP does not appear to have been affected. Thus, this may represent the first example of a naturally occurring mutant receptor that might impair signalling by selectively affecting the relative affinity of a receptor for natural agonist versus antagonist. It is not immediately apparent how mutations affecting intracellular domains might affect the binding of ligand to the extracellular face of the receptor. However, it is conceivable that the mutation induces a sufficient alteration of the receptor tertiary structure that this affects the three dimensional structure of the ectodomain, or else that they may affect interaction with intracellular kinases or other enzymes which might post-translationally modify the receptor and modulate receptor–ligand interaction.

Two of the mutations studied, A175T and V253I, result in receptors with apparently normal ligand binding affinity but an impaired ability to signal to adenylate cyclase. To our knowledge, these are the first MC4R mutations that show this pattern of functional impairment and further detailed studies of these mutant receptors may help to shed light on the nature of the coupling mechanisms between the receptor and downstream signal transduction pathways.

In summary, we have undertaken the most comprehensive survey to date of the functional properties of naturally occurring mutations in the human MC4R. In all cases where the family and population genetics suggested that a mutation is pathogenic, we have found evidence for impaired signalling. Detailed studies of the functional properties of these mutant receptors provides support for existing models of structure–function relationships but also revealed unexpected consequences of certain mutations. In particular a mutation in the C-terminal tail, I316S, appeared to selectively reduce receptor affinity for its natural agonist while binding of its natural antagonist was preserved. These findings illustrate the ability of naturally occurring, disease-associated mutations, to provide novel insights into the structure–function relationships of the G-protein coupled receptors.

MATERIALS AND METHODS

Subjects

A total of 350 unrelated UK probands were examined for mutations at the MC4R locus. All subjects had developed severe obesity before the age of 10 years. The mean BMI (body mass index) (weight in kg/height in m2) standard deviation score (SDS) at presentation to the study was 4.17. All studies were approved by the local ethical committee and were conducted in accordance with Declaration of Helsinki principles. To facilitate better interpretation of the potential pathogenicity of any mutation found, the entire MC4R sequence was determined in 54 healthy non-obese, British Caucasian subjects.

Direct nucleotide sequencing of MC4R gene

Two primers, MC4R forward (5′-AAT AAC TGA GAC GAC TCC CTG AC-3′) and MC4R reverse (5′-CAG AAG TAC AAT ATT CAG GTAGGG-3′), were used in a PCR reaction to amplify the MC4R gene from genomic DNA isolated from whole blood. The PCR was performed using BioTaq (Bioline) and carried out under standard conditions, with 35 cycles of 95°C for 30 s, 57°C for 30 s and 72°C for 50 s. Six nested primers, MC4F1 (5′-TGA GAC GAC TCC CTG ACC CAG-3′), MC4F2(5′-CAT CAC CCT ATT AAA CAG TAC AG-3′), MC4F3 (5′-AGG CTT CAC ATT AAG AGG ATT G-3′), MC4R1 (5′-TAC AAT ATT CAG GTA GGG TAA GA-3′), MC4R2 (5′-TTG GCG GAT GGC ACC AGT GC-3′) and MC4R3 (5′-CAC TGT GAA ACT CTG TGC ATC-3′), were then used at an annealing temperature of 52°C, to sequence the resulting PCR product on both strands. Sequencing was carried out using Big Dye terminator chemistry (Perkin-Elmer) and electrophoresed on an ABI 377 automated DNA sequencer. Sequences were assembled and examined using Sequencher 4.1 software (Gene Codes).

Generation of mutant and N-terminally MYC-tagged receptors

Wild-type MC4R was cloned using primers MC4Rforward (see above) and MC4RrevEco (5′-CGC TTA AGT TAA TAT CTG CTA GAC AAG TCA C-3′). MC4Rforward flanks an endogenous EcoRI site, while MC4RrevEco contains a terminal EcoRI restriction site, thus facilitating cloning of the PCR product into the mammalian expression vector pcDNA3 (Invitrogen). PCR was performed using Expand DNA polymerase (Roche) according to manufacturer's protocols. All mutant receptors were generated from the cloned wild-type MC4R using the QuikChangeTM site directed mutagenesis kit (Stratagene) according to manufacturer's protocols. The wild-type construct was N-Terminally tagged with MYC using site directed mutagenesis as described, and this construct was then used to generate MYC-tagged mutant constructs as described. All constructs were verified by direct nucleotide sequencing as described.

Cell culture and transfection

HEK293 cells were maintained in Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% fetal bovine serum (FBS), 100 units/ml penicillin, 100 µg/ml streptomycin and 250 ng/ml Fungizone (Life Technologies). Cells were incubated at 37°C in humidified air containing 5% CO2, and transfections were carried out using Fugene reagent (Roche) according to manufacturer's protocols. Cells were generally at 70–80% confluence on the day of transfection.

Measurement of cAMP accumulation

The cAMP responsive luciferase construct (LUC) utilized herein was as previously described (28). 300 ng of wild-type or mutant MC4R expression vector were transfected with 60 ng LUC reporter construct and 20 ng of internal control plasmid, pRL-CMV (Promega), which constitutively expresses Renilla Luciferase and is used to control for transfection efficiency. Following transfection, cells were serum starved for 8 h. Varying concentrations of αMSH (Bachem) were then added to the cells and allowed to incubate for 16 h. The cells were lysed and the two different luciferases were sequentially activated and quenched using the Dual-Luciferase Reporter Assay System (Promega) and the resulting luminescence detected using an AutoLumat LB953 luminometer (EG&G Berthold) according to manufacturers' protocols. Data was analysed using OriginTM (OriginLab) and curves were fitted using the logistical equation. Data is expressed as a fold induction of luciferase activity. Each point represents the mean and standard error range of at least three independent experiments performed in quadruplicate.

Immunofluorescence detection of cell surface expression

COS-7 cells were chosen for this study, since they were more adherent than HEK293 cells during the staining process. COS-7 cells were seeded onto eight-well chamber slides 24 h before transfection. Cells were transiently transfected with 20 ng of N-terminal MYC tagged wild-type or mutant construct as described. The next day, the cells were washed four times with Dulbecco's phosphate-buffered saline (DPBS; Sigma) and fixed for 20 min in 4% paraformaldehyde in DPBS followed by four washes with DPBS. Cells were then incubated for 5 min either with DPBS alone for non-permeabilized staining or with DPBS containing 0.2% Triton X-100 for permeabilization. Cells were incubated in blocking buffer [10% FBS and 2% bovine serum albumin (BSA) in DPBS] for 30 min, and then incubated for 1 h in blocking buffer containing 1 µg/ml of anti-c-MYC monoclonal antibodies (Sigma), which was followed by three 10 min washes with blocking buffer. Cells were then incubated for 30 min in blocking buffer containing 1:150 dilution of anti-mouse IgG fluorescein (FITC) antibody (Jackson ImmunoResearch Laboratories). After washing twice each for 10  min with DPBS, cells were mounted with coverslips using Mowiol 4-88 (Calbiochem) with 1,4-diazobicyclo-[2,2,2]-octane (DABCO) added to reduce fluorescent fading and viewed using a Bio-Rad MRC 1024 confocal microscope.

Ligand binding assays

Ligand binding assays were carried out on whole HEK293 cells, transiently transfected with wild-type or mutant MC4R constructs as described. Cells were harvested 48 h after transfection, washed once in binding buffer [minimal essential salts, 25 mM HEPES pH 7.0 (Life technologies), 0.2% BSA] and distributed in 96-well plates. Cells were incubated at room temperature for 2 h with 0.05 ml of binding buffer containing 15 000 cpm of [1251]NDP-MSH ([Nle4, D-Phe7]-α-melanocyte stimulating hormone; Amersham) and the appropriate concentration of unlabelled NDP-MSH (Bachem) or unlabelled agouti related peptide (AgRP; Phoenix Pharmaceuticals). Non-specific binding was determined in the presence of 10 µM NDP-MSH or 10 µM AgRP, respectively. The cells were then washed once with 0.1 ml ice cold binding buffer, once with 0.1 ml ice cold DPBS, and resuspended in 0.1 ml Microscint 20 (Packard). Tracer binding studies were performed as described above, but in the absence of unlabelled ligand. Radioactivity was counted using a Packard Topcount Microplate Scintillation Counter according to the manufacturer's instructions. Data was analyzed using OriginTM (OriginLab), and curves were fitted using the logistical equation. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean and standard error range of at least five independent experiments.

Co-transfection studies

HEK293 were transiently transfected as described, with either 600 ng of wild-type MC4R construct or 300 ng mutant MC4R construct+300 ng wild-type MC4R construct, 60 ng LUC and 20 ng pRL-CMV. Cells were treated as described, and exposed to 0.1 µM α-MSH. Luciferase values were measured as described, and data is expressed as a fold induction of luciferase activity. Each bar represents the mean and standard error range of at least three independent experiments performed in quadruplicate.

ACKNOWLEDGEMENTS

We thank the MRC (UK) for Programme Grant support (G.S.H.Y., E.J.L., J.K. and S.O.R.) and the Wellcome Trust (I.S.F.) and NSERC Canada (B.G.C.) for fellowship/scholarship support.

*

To whom correspondence should be addressed. Tel: +44 1223762620; Fax: +44 1223762323; Email: gyeo@hgmp.mrc.ac.uk

Figure 1. Schematic representation of the human MC4R. Amino acids are indicated as circles in single-letter code. Amino acids affected by mutations are shaded and specific mutations are then indicated with ovals. Mutations found in heterozygote form are indicated with a clear oval, and mutations found in homozygote form are indicated with a shaded oval. Codon number 1 refers to the start methionine. The seven transmembrane domains are indicated with Roman Numerals. ECL abbreviates extracellular loop, and ICL abbreviates intracellular loop.

Figure 1. Schematic representation of the human MC4R. Amino acids are indicated as circles in single-letter code. Amino acids affected by mutations are shaded and specific mutations are then indicated with ovals. Mutations found in heterozygote form are indicated with a clear oval, and mutations found in homozygote form are indicated with a shaded oval. Codon number 1 refers to the start methionine. The seven transmembrane domains are indicated with Roman Numerals. ECL abbreviates extracellular loop, and ICL abbreviates intracellular loop.

Figure 2. Frameshift and nonsense mutations, and C271Y. (A) cAMP/Luciferase reporter assay. Graphs indicate responses of wild-type, mutant and control constructs to a logarithmic increase in αMSH concentration. The control construct is empty pcDNA3 vector. Response is expressed in terms of increase in fold induction of luciferase activity over the respective unstimulated constructs. Each point represents the mean (±SE) of at least three independent experiments performed in quadruplicate. (B) Tracer binding assay. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in binding assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of transiently expressed receptors to bind radioligand was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments. (C) Cell surface immunofluorescence. Immunofluorescence studies with COS-7 cells transiently transfected with N-terminal MYC-tagged wild-type and mutant MC4Rs. Twenty-four hours after transfection, cells grown on eight chamber slides were treated with or without permeabilizing agent and stained with monoclonal anti-MYC antibodies and fluorescein (FITC)-conjugated anti-mouse IgG. Immunofluorescent signals were analysed by confocal microscopy. Vector-transfected cells showed negative staining.

Figure 2. Frameshift and nonsense mutations, and C271Y. (A) cAMP/Luciferase reporter assay. Graphs indicate responses of wild-type, mutant and control constructs to a logarithmic increase in αMSH concentration. The control construct is empty pcDNA3 vector. Response is expressed in terms of increase in fold induction of luciferase activity over the respective unstimulated constructs. Each point represents the mean (±SE) of at least three independent experiments performed in quadruplicate. (B) Tracer binding assay. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in binding assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of transiently expressed receptors to bind radioligand was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments. (C) Cell surface immunofluorescence. Immunofluorescence studies with COS-7 cells transiently transfected with N-terminal MYC-tagged wild-type and mutant MC4Rs. Twenty-four hours after transfection, cells grown on eight chamber slides were treated with or without permeabilizing agent and stained with monoclonal anti-MYC antibodies and fluorescein (FITC)-conjugated anti-mouse IgG. Immunofluorescent signals were analysed by confocal microscopy. Vector-transfected cells showed negative staining.

Figure 3. Missense mutations N97D, L106P and I125K. (A) cAMP/Luciferase reporter assay. As in Figure 2A. (B) Tracer binding assay. As in Figure 2B. (C) Cell surface immunofluorescence. As in Figure 2C. (D) NDP-MSH binding. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of NDP-MSH to inhibit radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments. (E) Ability of AgRP(87–132) to displace NDP-MSH. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of AgRP(87–132) to displace radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments.

Figure 3. Missense mutations N97D, L106P and I125K. (A) cAMP/Luciferase reporter assay. As in Figure 2A. (B) Tracer binding assay. As in Figure 2B. (C) Cell surface immunofluorescence. As in Figure 2C. (D) NDP-MSH binding. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of NDP-MSH to inhibit radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments. (E) Ability of AgRP(87–132) to displace NDP-MSH. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of AgRP(87–132) to displace radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments.

Figure 4. Missense mutations that respond partially to α-MSH. (A) cAMP/luciferase reporter assay. As in Figure 2A. (B) Tracer binding assay. As in Figure 2B. (C) Cell surface immunofluorescence. As in Figure 2C.

Figure 4. Missense mutations that respond partially to α-MSH. (A) cAMP/luciferase reporter assay. As in Figure 2A. (B) Tracer binding assay. As in Figure 2B. (C) Cell surface immunofluorescence. As in Figure 2C.

Figure 5. Competitive binding studies of N62S, R165Q, A175T, V253I and I316S. (A) NDP-MSH competitive binding assay. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of NDP-MSH to inhibit radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments. (B) Ability of AgRP(87–132) to displace NDP-MSH. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of AgRP(87–132) to displace radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments.

Figure 5. Competitive binding studies of N62S, R165Q, A175T, V253I and I316S. (A) NDP-MSH competitive binding assay. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of NDP-MSH to inhibit radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments. (B) Ability of AgRP(87–132) to displace NDP-MSH. Whole cells transiently transfected with vectors expressing wild-type or mutant MC4Rs were used in competitive assays. Cells were exposed to tracer amounts of [125I]NDP-MSH, and the ability of increasing concentrations of AgRP(87–132) to displace radioligand binding was measured. Data is expressed as a percentage of the maximum counts of [125I]NDP-MSH binding to wild-type MC4R. Each point represents the mean (±SE) of at least five independent experiments.

Figure 6. Mutant MC4Rs do not affect the function of wild-type MC4R. HEK 293 cells were transiently transfected with either pcDNA3 vector expressing WT MC4R or a 1:1 ratio of WT to mutant MC4Rs, such that the amount of DNA transfected each time was identical. The response of the cells to 1×10−7M α-MSH was assessed by co-transfection with a cAMP responsive reporter plasmid and measuring luciferase activity. Data is expressed as a fold induction of luciferase activity. Each bar represents the mean (±SE) of at least three independent experiments performed in quadruplicate.

Figure 6. Mutant MC4Rs do not affect the function of wild-type MC4R. HEK 293 cells were transiently transfected with either pcDNA3 vector expressing WT MC4R or a 1:1 ratio of WT to mutant MC4Rs, such that the amount of DNA transfected each time was identical. The response of the cells to 1×10−7M α-MSH was assessed by co-transfection with a cAMP responsive reporter plasmid and measuring luciferase activity. Data is expressed as a fold induction of luciferase activity. Each bar represents the mean (±SE) of at least three independent experiments performed in quadruplicate.

Table 1.

Summary of MC4R mutations

Mutation Heterozygous/homozygous in proband Present in other affected family members Conservation across speciesa Frequency in obese cohort 
‘A’ ins at codon 112 Heterozygous Yes Not applicable 
‘CTCT’ del at codon 211 Heterozygous Yes Not applicable 
Y287X Heterozygous Yes Not applicable 
N62S Heterozygous Yes Yes 
N97D Heterozygous Yes Yes 
L106P Heterozygous Not known Yes 
I125K Heterozygous Yes Yes 
R165Q Heterozygous Yes Yes 
A175T Heterozygous Yes Yes 
V253I Heterozygous Not known Yes 
C271Y Heterozygous Yes Yes 
I316S Heterozygous/Homozygous Yes Yes 1 homozygote 
    2 heterozygotes 
Mutation Heterozygous/homozygous in proband Present in other affected family members Conservation across speciesa Frequency in obese cohort 
‘A’ ins at codon 112 Heterozygous Yes Not applicable 
‘CTCT’ del at codon 211 Heterozygous Yes Not applicable 
Y287X Heterozygous Yes Not applicable 
N62S Heterozygous Yes Yes 
N97D Heterozygous Yes Yes 
L106P Heterozygous Not known Yes 
I125K Heterozygous Yes Yes 
R165Q Heterozygous Yes Yes 
A175T Heterozygous Yes Yes 
V253I Heterozygous Not known Yes 
C271Y Heterozygous Yes Yes 
I316S Heterozygous/Homozygous Yes Yes 1 homozygote 
    2 heterozygotes 

aCross-species comparison using rat (accession no. P70596), murine (accession no. NM016977), bovine (accession no. O9GLJ8), porcine (accession no. O97504) and zebrafish (accession no. AY078989) MC4R sequences.

Table 2.

Summary of functional results

Mutation cAMP response to α-MSH Cell surface expression Percentage of maximum tracer binding (±SE) NDP-MSH binding IC50 (M) (±SE) AGRP displacement of NDP-MSH binding IC50 (M) (±SE) 
Wild-type Full Yes 100% 1.4142×10−8±4.07×10−9 6.8056×10−9±1.30×10−9 
‘A’ ins at codon 112 None No    
‘CTCT’ del at codon 148 None No    
Y287X None No    
N62S Partial — 20%±4.4% >1.000×10−6 1.2051×10−7±6.72×10−8 
N97D None Yes 10.5%±2.9% >1.000×10−6 3.9157×10−7±6.47×10−8 
L106P None Yes 18.7%±2.4% >1.000×10−6 >1.000×10−6 
I125K None Yes 13.5%±2.6% >1.000×10−6 >1.000×10−6 
R165Q Partial — 29.9%±4.8% 2.4657×10−7±1.13×10−7 >1.000×10−6 
A175T Partial — 69.4%±11.5% 2.7283×10−8±9.78×10−9 7.5202×10−9±9.23×10−10 
V253I Partial — 80.3%±14.1% 1.1607×10−8±6.83×10−9 7.1269×10−9±1.24×10−9 
C271Y None No    
I316S Partial Yes 28%±1.7% 1.9787×10−7±7.64×10−8 1.5237×10−9±4.81×10−10 
Mutation cAMP response to α-MSH Cell surface expression Percentage of maximum tracer binding (±SE) NDP-MSH binding IC50 (M) (±SE) AGRP displacement of NDP-MSH binding IC50 (M) (±SE) 
Wild-type Full Yes 100% 1.4142×10−8±4.07×10−9 6.8056×10−9±1.30×10−9 
‘A’ ins at codon 112 None No    
‘CTCT’ del at codon 148 None No    
Y287X None No    
N62S Partial — 20%±4.4% >1.000×10−6 1.2051×10−7±6.72×10−8 
N97D None Yes 10.5%±2.9% >1.000×10−6 3.9157×10−7±6.47×10−8 
L106P None Yes 18.7%±2.4% >1.000×10−6 >1.000×10−6 
I125K None Yes 13.5%±2.6% >1.000×10−6 >1.000×10−6 
R165Q Partial — 29.9%±4.8% 2.4657×10−7±1.13×10−7 >1.000×10−6 
A175T Partial — 69.4%±11.5% 2.7283×10−8±9.78×10−9 7.5202×10−9±9.23×10−10 
V253I Partial — 80.3%±14.1% 1.1607×10−8±6.83×10−9 7.1269×10−9±1.24×10−9 
C271Y None No    
I316S Partial Yes 28%±1.7% 1.9787×10−7±7.64×10−8 1.5237×10−9±4.81×10−10 

References

1
Krude, H., Biebermann, H., Luck, W., Horn, R., Brabant, G. and Gruters, A. (
1998
) Severe early-onset obesity, adrenal insufficiency and red hair pigmentation caused by POMC mutations in humans.
Nat. Genet.
 ,
19
,
155
–157.
2
Yaswen, L., Diehl, N., Brennan, M.B. and Hochgeschwender, U. (
1999
) Obesity in the mouse model of pro-opiomelanocortin deficiency responds to peripheral melanocortin.
Nat. Med.
 ,
5
, l
066
–1070.
3
Miller, M.W., Duhl, D.M., Vrieling, H., Cordes, S.P., Ollmann, M.M., Winkes, B.M. and Barsh, G.S. (
1993
) Cloning of the mouse agouti gene predicts a secreted protein ubiquitously expressed in mice carrying the lethal yellow mutation.
Genes. Dev.
 ,
7
,
454
–467.
4
Michaud, E.J., Bultman, S.J., Klebig, M.L., van Vugt, M.J., Stubbs, L.J., Russell, L.B. and Woychik, R.P. (
1994
) A molecular model for the genetic and phenotypic characteristics of the mouse lethal yellow (Ay) mutation.
Proc. Natl Acad. Sci. USA
 ,
91
,
2562
–2566.
5
Ollmann, M.M., Wilson, B.D., Yang, Y.K., Kerns, J.A., Chen, Y., Gantz, I. and Barsh, G.S. (
1997
) Antagonism of central melanocortin receptors in vitro and in vivo by agouti-related protein.
Science
 ,
278
,
135
–138.
6
Naggert, J.K., Fricker, L.D., Varlamov, O., Nishina, P.M., Rouille, Y., Steiner, D.F., Carroll, R.J., Paigen, B.J. and Leiter, E.H. (
1995
) Hyperproinsulinaemia in obese fat/fat mice associated with a carboxypeptidase E mutation which reduces enzyme activity.
Nat. Genet.
 ,
10
,
135
–142.
7
Jackson, R.S., Creemers, J.W., Ohagi, S., Raffin-Sanson, M.L., Sanders, L., Montague, C.T., Hutton, J.C. and O'Rahilly, S. (
1997
) Obesity and impaired prohormone processing associated with mutations in the human prohormone convertase 1 gene.
Nat. Genet.
 ,
16
,
303
–306.
8
Gantz, I., Konda, Y., Tashiro, T., Shimoto, Y., Miwa, H., Munzert, G., Watson, S.J., DelValle, J. and Yamada, T. (
1993
) Molecular cloning of a novel melanocortin receptor.
J. Biol. Chem.
 ,
268
,
8246
–8250.
9
Roselli-Rehfuss, L., Mountjoy, K.G., Robbins, L.S., Mortrud, M.T., Low, M.J., Tatro, J.B., Entwistle, M.L., Simerly, R.B. and Cone, R.D. (
1993
) Identification of a receptor for gamma melanotropin and other proopiomelanocortin peptides in the hypothalamus and limbic system.
Proc. Natl Acad. Sci USA
 ,
90
,
8856
–8860.
10
Gantz, I., Miwa, H., Konda, V., Shimoto, V., Tashiro, T., Watson, S.J., DelValle, J. and Yamada, T. (
1993
) Molecular cloning, expression, and gene localization of a fourth melanocortin receptor.
J. Biol. chem.
 ,
268
,
15174
–15179.
11
Mountjoy, K.G., Mortrud, M.T., Low, M.J., Simerly, R.B. and Cone, R.D. (
1994
) Localization of the melanocortin-4 receptor (MC4-R) in neuroendocrine and autonomic control circuits in the brain.
Mol. Endocrinol.
 ,
8
,
1298
–1308.
12
Huszar, D., Lynch, C.A., Fairchild-Huntress, V., Dunmore, J.H., Fang, Q., Berkemeier, L.R., Gu, W., Kesterson, R.A., Boston, B.A., Cone, R.D. et al. (
1997
) Targeted disruption of the melanocortin-4 receptor results in obesity in mice.
Cell
 ,
88
,
131
–141.
13
Yeo, G.S., Farooqi, I.S., Aminian, S., Halsall, D.J., Stanhope, R.G. and O'Rahilly, S. (
1998
) A frameshift mutation in MC4R associated with dominantly inherited human obesity.
Nat. Genet.
 ,
20
,
111
–112.
14
Vaisse, C., Clement, K., Guy-Grand, B. and Froguel, P. (
1998
) A frameshift mutation in human MC4R is associated with a dominant form of obesity.
Nat. Genet.
 ,
20
,
113
–114.
15
Gu, W., Tu, Z., Kleyn, P.W., Kissebah, A., Duprat, L., Lee, J., Chin, W., Maruti, S., Deng, N., Fisher, S.L. et al. (
1999
) Identification and functional analysis of novel human melanocortin-4 receptor variants.
Diabetes
 ,
48
,
635
–639.
16
Hinney, A., Schmidt, A., Nottebom, K., Heibult, O., Becker, I., Ziegler, A., Gerber, G., Sina, M., Gorg, T., Mayer, H. et al. (
1999
) Several mutations in the melanocortin-4 receptor gene including a nonsense and a frameshift mutation associated with dominantly inherited obesity in humans.
J. Clin. Invest.
 ,
84
,
1483
–1486.
17
Farooqi, I.S., Yeo, G.S., Keogh, J.M., Aminian, S., Jebb, S.A., Butler, G., Cheetham, T. and O'Rahilly, S. (
2000
) Dominant and recessive inheritance of morbid obesity associated with melanocortin-4 receptor deficiency.
J. Clin. Invest.
 ,
106
,
271
–279.
18
Vaisse, C., Clement, K., Durand, E., Hercberg, S., Guy-Grand, B. and Froguel, P. (
2000
) Melanocortin-4 receptor mutations are a frequent and heterogeneous cause of morbid obesity.
J. Clin. Invest.
 ,
106
,
253
–262.
19
Ho, G. and MacKenzie, R.G. (
1999
) Functional characterization of mutations in melanocortin-4 receptor associated with human obesity.
J. Biol. Chem.
 ,
274
,
35816
–35822.
20
Hruby, V.J., Lu, D., Sharma, S.D., Castrucci, A.L., Kesterson, R.A., al-Obeidi, F.A., Hadley, M.E. and Cone, R.D. (
1995
) Cyclic lactam alpha-melanotropin analogues of Ac-Nle4-cyclo[Asp5, D-Phe7,Lys10] alpha-melanocyte-stimulating hormone-(4-10)-NH2 with bulky aromatic amino acids at position 7 show high antagonist potency and selectivity at specific melanocortin receptors.
J. Med. Chem.
 ,
38
,
3454
–3461.
21
Yang, Y.K., Thompson, D.A., Dickinson, C.J., Wilken, J., Barsh, G.S., Kent, S.B. and Gantz, I. (
1999
) Characterization of Agouti-related protein binding to melanocortin receptors.
Mol. Endocrinol.
 ,
13
,
148
–155.
22
Yang, Y.K., Fong, T.M., Dickinson, C.J., Mao, C., Li, J.Y., Tota, M.R., Mosley, R., Van Der Ploeg, L.H. and Gantz, I. (
2000
) Molecular determinants of ligand binding to the human melanocortin-4 receptor.
Biochemistry
 ,
39
,
14900
–144911.
23
Haskell-Luevano, C., Cone, R.D., Monck, E.K. and Wan, Y.P. (
2001
) Structure activity studies of the melanocortin-4 receptor by in vitro mutagenesis: identification of agouti-related protein (AgRP), melanocortin agonist and synthetic peptide antagonist interaction determinants.
Biochemistry
 ,
40
,
6164
–6179.
24
Bunge, S., Wedemann, H., David, D., Terwilliger, D.J., van den Born, L.I., Aulehla-Scholz, C., Samanns, C., Horn, M., Ott, J., Schwinger, E. et al. (
1993
) Molecular analysis and genetic mapping of the rhodopsin gene in families with autosomal dominant retinitis pigmentosa.
Genomics
 ,
17
,
230
–233.
25
Oksche, A. and Rosenthal, W. (
1998
) The molecular basis of nephrogenic diabetes insipidus.
J. Mol. Med.
 ,
76
,
326
–337.
26
Tanaka, H., Moroi, K., Iwai, J., Takahashi, H., Ohnuma, N., Hori, S., Takimoto, M., Nishiyama, M., Masaki, T., Yanagisawa, M. et al. (
1998
) Novel mutations of the endothelin B receptor gene in patients with Hirschsprung's disease and their characterization.
J. Biol. Chem.
 ,
273
,
11378
–11383.
27
Robbins, L.S., Nadeau, J.H., Johnson, K.R., Kelly, M.A., Roselli-Rehfuss, L., Baack, E., Mountjoy, K.G. and Cone, R.D. (
1993
) Pigmentation phenotypes of variant extension locus alleles result from point mutations that alter MSH receptor function.
Cell
 ,
72
,
827
–834.
28
Persani, L., Tonacchera, M., Beck-Peccoz, P., Vitti, P., Mammoli, C., Chiovato, L., Elisei, R., Faglia, G., Ludgate, M., Vassart, G. et al. (
1993
) Measurement of cAMP accumulation in Chinese hamster ovary cells transfected with the recombinant human TSH receptor (CHO-R): a new bioassay for human thyrotropin.
J. Endocrinol. Invest.
 ,
16
,
511
–519.