| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Research Group (A.H., C.V., A.-K.W., B.B., H.R., J.H.), Department of Child and Adolescent Psychiatry, Philipps-University of Marburg, D-35039 Marburg, Germany; Department of Pharmacology and Toxicology (S.H., C.H., T.G.), Philipps-University of Marburg, D-35033 Marburg, Germany; Institute of Medical Biometry and Epidemiology (F.G., H.S.), Philipps-University of Marburg, D-35037 Marburg, Germany; Spessart Klinik (H.G.), D-63619 Bad Orb, Germany; and Obesity Treatment Centre Insula (W.S.), D-83489 Berchtesgaden, Germany
Address all correspondence and requests for reprints to: Dr. Anke Hinney, Clinical Research Group, Department of Child and Adolescent Psychiatry, Philipps-University of Marburg, Schützenstr. 49, D-35039 Marburg, Germany. E-mail: anke.hinney{at}med.uni-marburg.de.
 |
Abstract |
|---|
 Top Abstract IntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
Currently, five frameshift mutations and one nonsense mutation have been reported (2, 3, 4, 5, 9, 14 , as based on studies published before March 2003). Their combined frequency amounts to 0.59% in 1853 (extremely) obese individuals in whom the coding region of the MC4R was completely screened for mutations. None of these mutations were detected in 826 unrelated normal weight controls, who were also completely screened.
The functional implications of the previously identified 32 missense mutations (2, 3, 4, 5, 6, 9, 10, 11, 12, 13, 14, 15) are seemingly less clear: Two missense variants (V103I and I251L) occur in approximately 2–3% and 1% of both obese and nonobese subjects from European and North American populations (3, 5, 9, 14), respectively; functional studies revealed no differences between these two variants and the wild-type receptor (3, 6, 16). A third missense variant (T112M) has been shown to occur infrequently in both normal-weight controls and obese individuals (5, 6, 9, 14); compatible with these data, a functional effect of this variant was not detected in an in vitro assay system (6). Recently three novel missense variants were detected in normal-weight individuals (I102T, F202L, N240S) (14). Unfortunately, functional studies were not performed for these variants. All of the other missense mutations have been found mainly or exclusively in extremely obese index patients encompassing children, adolescents, and adults. For some of the missense mutations, functional studies (3, 6, 9, 12, 16) have shown that the variants were either only partially active (N62S, S127L, I137T, T150I, R165W, I170V, A244E, I301T, I317T) or showed no activity (P78L, G98R, C271Y, P299H), compared with wild-type receptors. Surprisingly, one mutation that was identified in an extremely obese index patient leads to a constitutively active receptor molecule (L250Q) (3). Recently it was shown (15) that the mutated receptor molecule is partially retained intracellularly.
In family studies, relatives who are carriers of the same functionally relevant MC4R mutation as the respective index patient have mostly been shown to be obese (2, 3, 4, 5, 8, 9, 10, 12, 13), thus substantiating a major gene effect. However, single carriers within these families were not overweight (3, 6, 8, 9, 10). Whereas one relative with a nonsense mutation (Y35X) (8) had been obese before a heart attack, Vaisse et al. (3) identified three lean family members with a frameshift mutation (47–48insG) ascertained via an extremely obese carrier; the respective mutation leads to a truncation in the N terminus of the protein. In the three lean relatives, obesity had never occurred; no explanation for their leanness was found. Farooqi et al. (9) detected homozygous carriers of a missense mutation (N62S) in a consanguineous family. The five homozygous mutation carriers of that family were extremely obese, whereas the heterozygous relatives were nonobese, leading the authors to conclude that this MC4R mutation confers a recessively transmitted form of obesity. Kobayashi et al. (12) reported a missense mutation (G98R) leading to a loss of function. The extremely obese Japanese girl was homozygous for the mutation; the parents and the sister, all heterozygous for the mutation, were overweight. This mutation is seemingly associated with a codominant form of obesity.
On the basis of these aforementioned findings, it is currently unclear to what extent mutations in MC4R indeed influence body weight. In the original two publications (2, 4), based on a single family each, the investigators had concluded that MC4R mutations are compatible with dominant inheritance of obesity. Ho and MacKenzie (16), who performed functional studies, presumed that MC4R mutations lead to haploinsufficiency of the receptor. However, individuals with large deletions in the chromosomal region (18q) that harbors the MC4R are not obese. Thus, obesity might result as a consequence of a dominant-negative effect of MC4R mutations (17). Nonsense-mediated decay inhibits the accumulation of nonsense- or frameshift-mutated mRNA and thus minimizes the synthesis of truncated proteins with potential dominant-negative effects. However, it was recently shown that MC4R mRNA is not susceptible to nonsense-mediated decay (18). Hence, it remains uncertain whether the effect of the functionally relevant mutations is due to haploinsufficiency or whether the mutated alleles exert a dominant-negative effect. Very recently Lubrano-Berthelier et al. (15) showed that approximately 80% of the mutations identified in children and adolescents cause partial or complete intracellular retention of the receptor.
The identification of lean carriers with frameshift mutations (3) suggests that other genetic and/or specific environmental factors are potentially prerequisites for the manifestation of obesity in MC4R mutation carriers. Jacobson et al. (14) have since questioned the relationship between MC4R mutations and severe obesity. A priori, obesity, which represents the right end of the body mass index (BMI) distribution, is caused by both environmental and genetic factors. Among the latter, both additive and nonadditive factors contribute to this quantitative phenotype (19). Most likely, monogenic forms of obesity are infrequent; thus, in genetic terms oligogenic or polygenic inheritance of obesity should be much more common.
In the current study, we addressed the question of whether functionally relevant mutations in the MC4R (excluding the known polymorphisms) are indeed compatible with an autosomal dominant major gene effect on the phenotype extreme obesity. For this purpose, we performed a mutation screen in 808 extremely obese children and adolescents, 231 underweight (
15th BMI percentile) and 96 normal-weight (40–60th percentile) adult controls to enable a case-control study. Functional studies were performed to determine whether newly identified variants (especially missense mutations) lead to impaired receptor function or even loss of function of the MC4R. Up to this point, we adhered to the procedure recently proposed for analyses of common diseases (20). To further substantiate our findings, we screened a total of 1040 parents of 520 of the aforementioned 808 extremely obese index patients. If indeed functionally relevant MC4R mutations represent major gene effects, we hypothesized that 1) parents should not or only infrequently harbor additional functionally relevant mutations other than those transmitted to the offspring and 2) that parents with relevant mutations are obese.
|
Study subjects |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
15th percentile; mean age 25.3 ± 3.8 yr) and 96 healthy, normal-weight students (mean BMI 21.9 ± 1.1 kg/m2, BMI between 40th and 60th percentile, mean age 24.7 ± 2.6 yr). A total of 65.5% of the extremely obese children and adolescents had an age- and gender-specific BMI percentile of 99 or greater as previously determined in a representative German population sample (21). Additionally, the 1040 parents (mothers: mean BMI 30.4 ± 6.7 kg/m2, mean age 41.5 ± 5.5 yr; fathers: mean BMI 30.4 ± 5.7 kg/m2, mean age 44.5 ± 6.2 yr) of those 520 obese children and adolescents were screened, for whom both biological parents had been ascertained before the MC4R mutation screen of the index patient, thus guaranteeing a nonbiased transmission disequilibrium test (TDT). Because we had excluded trios not compatible with Mendelian inheritance in previous single nucleotide polymorphism and microsatellite analyses, nonpaternity should be a rare event only. Written informed consent was given by all participants and in the case of minors, their parents. The study was conducted in accordance with the guidelines proposed in the Declaration of Helsinki and was approved by the Ethics Committee of the University of Marburg. The current analyses are based on a total of 2175 individuals, who were either newly screened for MC4R mutations (n = 1606) or had contributed to previous studies (5, 8). |
Phenotypic measurements |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
PCRs were performed with primers amplifying the MC4-R in two overlapping parts with the primer pairs MC4R-1 F/MC4R-1R (F1; 615 bp) and MC4R-2 F/MC4R-2R (F2, 622 bp; sequences have been described previously) (5). Subsequently, single-strand conformation polymorphism (SSCP) analyses for mutations in the MC4-R were performed at 4 C as described previously (5).
Additionally, all samples that revealed SSCP patterns indistinguishable from those of either one of the two known polymorphisms (V103I, I251L) were reanalyzed with denaturing HPLC by using the WAVE system (Transgenomic, Cheshire, UK). Melt temperatures for optimal separation of homo- and heteroduplexes were deduced from the melting profile of each amplicon’s DNA sequence using the WAVEmaker software (version 4.0, Transgenomic). The PCRs were performed according to standard protocols using 12 ng DNA (primers and PCR conditions as described) (5). Five microliters of unpurified denatured-renatured PCR products were loaded onto a preheated specific column (DNASep column, Transgenomic) and eluted from the column by a linear acetonitrile gradient in a 0.1-M triethylammonium acetate (TEAA) buffer, pH 7, and at a constant flow rate of 0.9 ml/min. The gradient was created by mixing eluents A (0.1 M TEAA) and B (0.1 M TEAA, 25% acetonitrile). For analyses of the MC4R F1 fragment, PCR products were eluted from the column (61 C) by increasing buffer B from 54–59% within 30 sec and from 59–68% within 4.5 min. The oven temperature for analyzing the MC4R F2-fragment was set to 59 C and 60 C, respectively. For elution of the MC4R F2 products, buffer B was increased from 54–59% within 10 sec and from 59–68% within 4.5 min. PCR amplicons with more than one peak or differences in peak appearance were sequenced. All PCR amplicons with SSCP or WAVE patterns deviant from the wild-type pattern or the patterns created by the two polymorphisms were sequenced as described previously (5). One presumed I251L polymorphism (by SSCP) turned out to be a missense mutation (G181D). Another individual turned out to harbor a frameshift mutation leading to a deletion of GA at codon 250 in addition to the I251L polymorphism.
PCR-restriction fragment length polymorphism analyses
To reconfirm all identified variants (mutations and polymorphisms), PCRs with subsequent diagnostic restriction fragment length polymorphism analyses were performed. For alleles that did not alter restriction sites, specific primers were designed to either introduce artificial restriction sites or perform allele-specific PCR reactions. The [Y35X, D37V] haplotype alleles were each detected by allele-specific PCR as described previously (8); V103I was detected as described by Gotoda et al. (1). All mutations and polymorphisms were reconfirmed by an independent mutation analysis in every carrier of a variant.
Isolation and characterization of genomic DNA encoding the MC4R
The gDNA of the human wild-type and mutant MC4R was amplified by PCR using the primer pairs EcoRI-Koz-MC4r (forward primer), 5'-GAA TTC CCC ACC ATG GTG AAC TCC ACC CAC-3' and BamHI-Stop-MC4r (reverse primer), 5'-GGA TCC TTA ATA TCT GCT AGA CAA GTC-3'. The PCR fragment was inserted into the eukaryotic expression vector pSG5 (Stratagene, La Jolla, CA). PCR products were further characterized by restriction endonuclease digestion and subsequent dideoxy sequencing (Perkin-Elmer, Boston, MA; Sequencer 310, PE Biosystems).
Transient expression of wild-type and mutant MC4Rs and functional assays
COS-7 cells were grown in DMEM supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 µg/ml streptomycin at 37 C in a humidified 5% CO2 incubator. For transient transfections COS-7 cells were split into 12-well plates (2 x 105 cells/well) and transfected with 1 µg plasmid DNA per well using Fugene reagent (Roche, Mannheim, Germany). For cAMP measurements, cells were prelabeled with [2,8-3H]adenine (2 µCi/well) (Perkin-Elmer). Assays were performed 3 d after transfection. For cAMP assay, cells were washed once in serum-free DMEM, followed by a preincubation with the same medium containing 1 mM 3-isobutyl-1-methylxanthine (Sigma, Deisenhofen, Germany) for 20 min at 37 C in a humidified 5% CO2 incubator. Subsequently cells were stimulated with appropriate concentrations of 
-MSH (Sigma) for 1 h. Reactions were terminated by aspiration of the medium and addition of 1 ml 5% trichloric acid. The cAMP content of the cell extracts was determined by anion-exchange chromatography as described (22): EC50 and Emax values were obtained from -MSH concentration-response curves (from 1 nM to 10 µM -MSH/ml), using the computer program GraphPad Prism (GraphPad Inc., San Diego, CA). Data are indicated as means ± SEM of three to six independent experiments, each carried out in duplicates.
The wild-type receptor showed Emax values approximately 10-fold over wild-type basal; EC50 values were approximately 50 nM -MSH (Table 1
). We used the following criteria to define functional groups: (1) loss of function was defined via mean Emax values less than 2-fold over wild-type basal and undeterminable EC50 values; (2) reduced function was defined via EC50 values that were more than 2.5-fold increased, compared with wild-type with Emax values similar or slightly reduced (P78L) to the wild-type; (3) like wild-type was defined by Emax and EC50 values similar to wild-type MC4R. Because the distinction between loss of function and reduced function is arbitrary, we refer to the general term, impaired function, for both groups (1 and 2). Two MC4R constructs (S127L and P230L) showed constitutively active receptors, which were characterized by increased basal cAMP levels, compared with the wild-type receptor (see Table 1).
|
). When we initiated our functional analyses, two of our detected mutations (T112M and R165W) had been functionally characterized by others (Refs. 3 and 6 , respectively). Very recently the functional analyses of some more of the mutations identified in our study were published (15); these data, which are consistent with our own data, were not implemented in Table 1. The nonsense mutation leading to a stop codon at amino acid position 35 of the MC4R is compatible only with a complete loss of function (15). In vitro tests for this mutation were regarded as unnecessary. Because the missense mutation at codon 37 (D37V) is always found on a haplotype with the nonsense mutation, we did not perform functional assays pertaining to this mutation.
Statistics, confirmatory analyses
The hypothesis of the study was the association of functionally relevant mutations of the MC4R with extreme obesity. Therefore, the difference in the numbers of carriers of mutations leading to functional impairment between 808 obese and 327 nonobese individuals were investigated using a one-sided Fisher’s exact test. Subsequently TDTs were conducted in trios based on those 520 of the 808 obese index patients whose parents had also been ascertained. Here we also combined the mutations leading to functional impairment to allow for a one-sided TDT [(23); exact McNemar test] based on a larger number of informative trios. The grouping of the mutations was based on the results of functional assays as recommended by Hirschhorn and Altshuler (20).
Exploratory analyses
Further association tests for the obese and nonobese groups and TDTs were performed for other groupings of the observed mutations. Differences in age, BMI, and SD scores for BMI between carriers of mutations leading to an impaired function vs. all other probands were analyzed with two-sample t tests.
Odds ratio and genotype relative risk
Further analyses were conducted to estimate the risk to develop extreme obesity because of a mutation leading to impaired function. We estimated the odds ratio and a 95% confidence interval (CI) from the 808 obese cases and 327 nonobese controls. The genotype relative risk and a 95% exact binomial CI were estimated from those trios in which only one of the four parental alleles harbored a mutation. The frequency of the functionally relevant mutations and a 95% exact CI were estimated from the nontransmitted parental alleles in the 520 trios. This estimate is valid under the assumption of random mating and Hardy-Weinberg equilibrium.
|
Results |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
A total of 18 different nonsilent MC4R gene variants were detected in 43 of the 808 obese children and adolescents in whom the coding region of the MC4R was screened for mutations (Table 1). Based on the underlying mutation on the DNA level, these 18 variants can be classified as frameshift (n = 2), nonsense (n = 1), and missense (n = 15) mutations, respectively. The frameshift and nonsense mutations are expected to interrupt the reading frame in the domains encoding the N terminus (Y35X) and the fifth and sixth transmembrane domains [codon 211 deletion of CTCT (L211fsX216) and codon 250 deletion of GA (L250fsX284)], respectively. Nine of the missense mutations (S30F, D37V, I121T, S127L, R165W, G181D; G252S, A244E, I317T) code for nonconservative amino acid changes; six missense variants (P78L, T112M, P230L, V95I; polymorphisms V103I and I251L) were conservative. Two silent variants were detected (405A>G, 972C>T).
The two most common variants, V103I and I251L, which have previously been identified as polymorphisms without detectable functional implications (3, 6, 15), were observed in 13 (1.61%) and nine (1.11%) of the obese individuals, respectively.
The third most common mutations, Y35X and D37V, were each observed in the same five (0.62%) individuals; these two variants form a haplotype as revealed by parental transmission patterns. The missense mutations S30F, S127L, and R165W and the frameshift mutation L211fsX216 were detected in two individuals each, the remaining 10 variants, one frameshift and nine missense mutations, were observed only once in the study group (Table 1). Three index patients were compound heterozygous for two missense variants each (a: S127L and R165W; b: S30F and G252S; c: G181D and I251L), as revealed by parental transmission patterns. The haplotype S127L and V103I was identified in an index patient and her mother.
The results of the heterologous in vitro studies based on the ability of -MSH to stimulate cAMP production in addition to previous findings in the literature (3, 6) are illustrated in Table 1. Nine [the haplotype (Y35X; D37V) was counted as one mutation] of the 16 different nonsilent mutations showed an impaired function. In the applied assay, the single nonsense, two frameshift (L211fsX216, L250fsX284), and two missense mutations (V95I, G181D) each led to a loss of function. Four missense mutations (P78L, I121T, R165W, A244E) resulted in an altered receptor function, which we crudely classified as reduced in function. MC4R harboring either one of the remaining four missense mutations (S30F, T112M, G252S, I317T) or the polymorphisms I251L and V103I were functionally indistinguishable from the wild-type receptor. The missense mutations S127L and P230L showed elevated basal cAMP levels, indicating a constitutively active receptor.
Adult underweight and normal-weight controls
Among the 231 underweight and 96 normal-weight students, a total of eight carriers of the 103I allele were identified. The second polymorphism (I251L) was detected in seven individuals, one of whom also harbored allele 103I. No frameshift or nonsense mutations were detected among these 327 students. A single conservative missense mutation was found in a normal-weight individual (T178M). The cAMP assay (Table 1) revealed that the T178M missense mutation seemingly does not affect receptor function. Finally, a single silent mutation (579C>T) was detected in an underweight proband.
Association analysis
Association analyses (Table 2) were complicated by the fact that the detected MC4R variants were rare and had different functional implications. The analyses were based on our own and previously (3, 6) generated functional data. Because of uncertainties as to their functional implications, the two mutations leading to a constitutively active receptor (S127L and P230L) were excluded from these analyses. When all mutations leading to impaired (loss of or reduced function) function were combined, the association test was significant (one sided P = 0.006). Post hoc analyses revealed that separate analyses of the frameshift plus nonsense mutations and the missense mutations show trends for association within each group (descriptive, two-sided: P = 0.114 and P = 0.124, respectively; Table 2).
|
Our subsequent analyses were based on trios recruited via those 520 of the 808 extremely obese index patients whose parents had also been ascertained. We also combined the mutations leading to functional impairment to allow for a TDT (23) based on a larger number of informative trios. A total of 11 heterozygous carriers of mutations leading to a functional impairment were detected among the 1040 parents (the two mutations leading to constitutively active receptors were again excluded from the analyses). All parents were completely screened by SSCP; non-Mendelian transmissions were not detected. The confirmatory test revealed nine transmissions and two nontransmissions (P = 0.033 exact McNemar test, one-sided; Table 3). The two nontransmitted mutations (the nonconservative missense mutation S94R and the frameshift mutation Y320fsX354) both lead to an impaired function (Table 1). The results of all post hoc TDT analyses are shown in Table 3.
|
For the combined set of mutations leading to impaired function, we estimated the odds ratio, genotype relative risk, and allele frequency. Because no functionally relevant mutation was observed in the 327 controls and 15 of the 808 obese individuals were carriers of mutations associated with an impaired function, the odds ratio estimate results in +
, the exact 95% CI ranges from 1.47 to +. Based on the nontransmitted parental alleles (which were found to result in an impaired function in the in vitro assays), the estimate for the frequency of functionally relevant mutations in the general population is 0.19% (95% CI: 0.02%–0.69%). Based on the observed nine transmissions and two nontransmissions of mutations associated with an impaired function in our 520 trios, we estimated the genotype relative risk of heterozygous carriers of functionally relevant mutations as 4.5; the 95% CI ranges from 0.9 to 42.8.
Percentile plots: extremely obese children and adolescents
Figure 1, A and B, illustrates BMI of those 21 extremely obese young index patients with frameshift, nonsense, and/or missense mutations (excluding the two polymorphisms) in relationship to both the 787 extremely obese nonmutation carriers (including carriers of one or the other polymorphism, in whom no other nonsilent mutation was present) and population-based gender-specific BMI percentiles. Three mutations (Y35X, R165W, S30F) occurred in both genders. Except for two variants (S30F and S127L), all mutations in the females led to a receptor with impaired function (Fig. 1A). BMI of the eight male mutation carriers clustered in the middle to upper range of the BMI distribution formed by all male obese subjects (Fig. 1B); four of the respective mutations were associated with an impaired receptor function. The variants detected in the other four male carriers led to receptors compatible with unaltered MC4R function; one coded for a constitutively active receptor molecule.
|
Percentile plots: parental study group
All 1040 parents of 520 of the total of 808 obese children and adolescents were also screened for MC4R mutations (Fig. 2, A and B). Fifteen different nonsilent (including the two polymorphisms) and two different silent (405C>A, 972C>T) variants were detected in these 1040 parents.
|
, Fig. 2A) and all seven mutation carriers had BMI in the 85th centile or greater. Among the fathers, two of the nine mutation carriers had a BMI below the 85th percentile, one of these below the 50th percentile. Our in vitro assays revealed that the two variants of the normal weight fathers are seemingly functioning like the wild-type receptor. Finally, parents with mutations associated with an impaired function (n = 11) all had BMI 85th BMI percentile or greater. |
Discussion |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
To evaluate the functional relevance of the identified MC4R mutations, COS-7 cells were transiently transfected with different receptor cDNAs, and functional studies were performed. To assess receptor function at a proximal step of the cellular signaling cascade, we directly measured second-messenger production and determined agonist-induced intracellular cAMP accumulation. Based on these in vitro studies and published data (3, 6), we formed two main nonoverlapping categories (for criteria see Materials and Methods) to classify the mutations: (1) impaired function (loss of function and reduced function) and (2) like wild-type. This categorization excludes the two mutations that turned out to lead to constitutively active receptors (P230L and S127L, Table 1). Vaisse et al. (3) also described a constitutively active receptor in an obese individual; the respective receptor proteins are partially retained intracellularly (15). This result is in line with previous data indicating that expression of constitutively active receptors is often reduced (24). Additionally, such receptors have been described to be structurally unstable (25). Hence, we cannot exclude that both S127I and P230L entail an impaired function. Nevertheless, based on the uncertain functional classification of the two identified constitutively active receptor variants, we excluded them from the association and TDT analyses.
We chose the umbrella term, impaired function, for both loss of function and reduced function because our subcategorization is somewhat arbitrary. Our in vitro results are compatible with the previous data (9, 15). Because only one functional test was performed, we obviously cannot exclude the possibility that single mutations associated with a normal cAMP response could show impaired function in other assays. Indeed, recently Lubrano-Bertehlier et al. (15) showed that one of the mutations (I317T), which we had classified as like wild-type, showed impaired cell surface expression. However, because these data are not available for all of our identified mutations, we did not include this information in the biometrical analyses. Furthermore, we did not test binding of MC4R antagonists. Finally, albeit unlikely, functional alterations detected in in vitro assays do not necessarily imply that the respective variants are also functionally deviant in vivo.
The nine mutations leading to impaired function occurred in 15 of the 808 obese individuals (1.9%; 95% CI 1.0–3.0%). The frequency of mutation carriers among obese study groups ranged from 0.5% to 6.3% in previous reports (2, 3, 4, 5, 6, 9, 10, 11, 12, 13, 14, 15). All currently known mutations and polymorphisms in the coding region are seemingly randomly distributed; there is no evidence for a mutational hotspot in DNA segments coding for any of the domains of the receptor molecule.
The observation of the same haplotype formed by the nonsense mutation Y35X and the missense mutation D37V in five obese probands suggests a common founder for 0.62% of our study population. Indeed, we have previously shown that two index patients ascertained in two different German states (Bavaria and Hessia) are second-degree cousins (8). Because all other mutations occurred only once or twice, the thrifty genotype hypothesis (26) is not readily substantiated. Only a single female index patient with a BMI of 48.5 kg/m2, one of the highest in the total study group, was compound heterozygous for two functionally relevant mutations (Fig. 1A), indicating that compound heterozygotes are not detected disproportionately in our sample.
Our confirmatory analyses, which substantiate association between MC4R mutations and extreme childhood and adolescent obesity, were based on pooling of functionally relevant mutations, an approach recently recommended by Hirschhorn and Altshuler (20). The significant ratio of 15 carriers of functionally relevant mutations in 808 obese individuals to 0 in 327 nonobese individuals is suggestive of a major gene effect of the respective mutations on the phenotype extreme obesity (one-sided Fisher’s exact test: P = 0.006, lower confidence limit of the odds ratio: 1.47). Obviously, large-scaled epidemiological studies are required to obtain more exact estimates of the genetic effects and the population frequencies of the different MC4R mutations. Based on our data, we predict that they will be detectable almost only in the uppermost BMI strata. Indeed, among our index patients, carriers of mutations leading to impaired function had a higher BMI than patients carrying either wild-type alleles or alleles not associated with an impaired function.
Our approach to pool functionally relevant mutations for TDT analyses was also based on the recommendation for association studies (20). The mutations leading to an impaired function (loss of function and reduced function) had a transmission rate of 81.8% (P = 0.033, Table 3). All related post hoc TDTs (Table 3) also showed that all mutations (nonsense, frameshift, and missense) were preferentially transmitted to the obese offspring. The two nontransmissions need to be interpreted in the light of the frequent occurrence of obesity in both parents of our obese index patients. In our study group approximately 25% of the index patients have two obese (BMI 
30 kg/m2) parents. Assortative mating (27, 28) and common environmental effects as well as the sampling procedure in itself (Dempfle, A., unpublished data) need to be considered as underlying this observation. If both parents are obese, obese offspring can genetically result from a single or more than one gene transmitted from either one or both of the parents.
Estimation of the genotype relative risk for heterozygous carriers of mutations leading to functional impairment indicates a 4.5-fold increased genotype relative risk (95% CI 0.9–42.8) to become obese (BMI 90th centile) for carriers of mutations leading to functional impairment, compared with noncarriers. However, the large 95% CI encompasses 1, and the upper confidence limit indicates that the relative risk might be substantially greater than 4.5. Because of the high percentage of bilinear parental obesity, the genotype relative risk may be underestimated in our study. As a result of the low frequency of functionally relevant mutations in the MC4R, these are evidently not relevant for the majority of obese individuals.
In conclusion, we tested for association and transmission disequilibrium of MC4R mutations in early-onset extreme obesity. Both approaches critically depend on the joint analyses of the functionally relevant mutations and their correct functional classification (see Ref. 20). Based on the significantly higher frequency of functionally relevant mutations in obese, compared with nonobese, individuals and the subsequent significant TDT, we conclude that MC4R mutations are compatible with a major gene effect for the development of obesity. This hypothesis is further supported by the genotype relative risk estimate of 4.5 (95% CI 0.9–42.8). In combination with the small population frequency of mutation carriers, the results indicate that MC4R mutations have a low epidemiological but a high individual relevance.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BMI, Body mass index; CI, confidence interval; SSCP, single-strand conformation polymorphism; TDT, transmission disequilibrium test; TEAA, triethylammonium acetate.
Received February 12, 2003.
Accepted May 15, 2003.
|
References |
|---|
|
TopAbstractIntroductionStudy subjectsPhenotypic measurementsMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. R. H. Buch, D. Heling, E. Damm, T. Gudermann, and A. Breit Pertussis Toxin-sensitive Signaling of Melanocortin-4 Receptors in Hypothalamic GT1-7 Cells Defines Agouti-related Protein as a Biased Agonist J. Biol. Chem., September 25, 2009; 284(39): 26411 - 26420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Calton, B. A. Ersoy, S. Zhang, J. P. Kane, M. J. Malloy, C. R. Pullinger, Y. Bromberg, L. A. Pennacchio, R. Dent, R. McPherson, et al. Association of functionally significant Melanocortin-4 but not Melanocortin-3 receptor mutations with severe adult obesity in a large North American case-control study Hum. Mol. Genet., March 15, 2009; 18(6): 1140 - 1147. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. V. Kryukov, A. Shpunt, J. A. Stamatoyannopoulos, and S. R. Sunyaev Power of deep, all-exon resequencing for discovery of human trait genes PNAS, March 10, 2009; 106(10): 3871 - 3876. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Pichler, B. Kollerits, I. M Heid, S. C Hunt, T. D Adams, P. N Hopkins, and F. Kronenberg Association of the melanocortin-4 receptor V103I polymorphism with dietary intake in severely obese persons Am. J. Clinical Nutrition, September 1, 2008; 88(3): 797 - 800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Stutzmann, K. Tan, V. Vatin, C. Dina, B. Jouret, J. Tichet, B. Balkau, N. Potoczna, F. Horber, S. O'Rahilly, et al. Prevalence of Melanocortin-4 Receptor Deficiency in Europeans and Their Age-Dependent Penetrance in Multigenerational Pedigrees Diabetes, September 1, 2008; 57(9): 2511 - 2518. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Staubert, P. Tarnow, H. Brumm, C. Pitra, T. Gudermann, A. Gruters, T. Schoneberg, H. Biebermann, and H. Rompler Evolutionary Aspects in Evaluating Mutations in the Melanocortin 4 Receptor Endocrinology, October 1, 2007; 148(10): 4642 - 4648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Hainerova, L. H. Larsen, B. Holst, M. Finkova, V. Hainer, J. Lebl, T. Hansen, and O. Pedersen Melanocortin 4 Receptor Mutations in Obese Czech Children: Studies of Prevalence, Phenotype Development, Weight Reduction Response, and Functional Analysis J. Clin. Endocrinol. Metab., September 1, 2007; 92(9): 3689 - 3696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Stutzmann, V. Vatin, S. Cauchi, A. Morandi, B. Jouret, O. Landt, P. Tounian, C. Levy-Marchal, R. Buzzetti, L. Pinelli, et al. Non-synonymous polymorphisms in melanocortin-4 receptor protect against obesity: the two facets of a Janus obesity gene Hum. Mol. Genet., August 1, 2007; 16(15): 1837 - 1844. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. M. Kublaoui and A. R. Zinn MC4R Mutations--Weight before Screening! J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1671 - 1672. [Full Text] [PDF]
|
|
|
|
|
|
A. Hinney, T. Bettecken, P. Tarnow, H. Brumm, K. Reichwald, P. Lichtner, A. Scherag, T. T. Nguyen, P. Schlumberger, W. Rief, et al. Prevalence, Spectrum, and Functional Characterization of Melanocortin-4 Receptor Gene Mutations in a Representative Population-Based Sample and Obese Adults from Germany J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1761 - 1769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Lubrano-Berthelier, B. Dubern, J.-M. Lacorte, F. Picard, A. Shapiro, S. Zhang, S. Bertrais, S. Hercberg, A. Basdevant, K. Clement, et al. Melanocortin 4 Receptor Mutations in a Large Cohort of Severely Obese Adults: Prevalence, Functional Classification, Genotype-Phenotype Relationship, and Lack of Association with Binge Eating J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1811 - 1818. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Herbert, N. P. Gerry, M. B. McQueen, I. M. Heid, A. Pfeufer, T. Illig, H.-E. Wichmann, T. Meitinger, D. Hunter, F. B. Hu, et al. A common genetic variant is associated with adult and childhood obesity. Science, April 14, 2006; 312(5771): 279 - 283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Buono, F. Pasanisi, C. Nardelli, L. Ieno, S. Capone, R. Liguori, C. Finelli, G. Oriani, F. Contaldo, and L. Sacchetti Six Novel Mutations in the Proopiomelanocortin and Melanocortin Receptor 4 Genes in Severely Obese Adults Living in Southern Italy Clin. Chem., August 1, 2005; 51(8): 1358 - 1364. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-K. Wermter, K. Reichwald, T. Buch, F. Geller, C. Platzer, K. Huse, C. Hess, H. Remschmidt, T. Gudermann, G. Preibisch, et al. Mutation analysis of the MCHR1 gene in human obesity Eur. J. Endocrinol., June 1, 2005; 152(6): 851 - 862. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I M Heid, C Vollmert, A Hinney, A Doring, F Geller, H Lowel, H-E Wichmann, T Illig, J Hebebrand, F Kronenberg, et al. Association of the 103I MC4R allele with decreased body mass in 7937 participants of two population based surveys J. Med. Genet., April 1, 2005; 42(4): e21 - e21. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. H. Larsen, S. M. Echwald, T. I. A. Sorensen, T. Andersen, B. S. Wulff, and O. Pedersen Prevalence of Mutations and Functional Analyses of Melanocortin 4 Receptor Variants Identified among 750 Men with Juvenile-Onset Obesity J. Clin. Endocrinol. Metab., January 1, 2005; 90(1): 219 - 224. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Ma, P. A. Tataranni, C. Bogardus, and L. J. Baier Melanocortin 4 Receptor Gene Variation Is Associated With Severe Obesity in Pima Indians Diabetes, October 1, 2004; 53(10): 2696 - 2699. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A Dempfle, A Hinney, M Heinzel-Gutenbrunner, M Raab, F Geller, T Gudermann, H Schafer, and J Hebebrand Large quantitative effect of melanocortin-4 receptor gene mutations on body mass index J. Med. Genet., October 1, 2004; 41(10): 795 - 800. [Full Text] [PDF]
|
|
|
|
|
|
C. Lubrano-Berthelier, C. Le Stunff, P. Bougneres, and C. Vaisse A Homozygous Null Mutation Delineates the Role of the Melanocortin-4 Receptor in Humans J. Clin. Endocrinol. Metab., May 1, 2004; 89(5): 2028 - 2032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Valli-Jaakola, M. Lipsanen-Nyman, L. Oksanen, A. N. Hollenberg, K. Kontula, C. Bjorbaek, and C. Schalin-Jantti Identification and Characterization of Melanocortin-4 Receptor Gene Mutations in Morbidly Obese Finnish Children and Adults J. Clin. Endocrinol. Metab., February 1, 2004; 89(2): 940 - 945. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
