Effect of the Melanocortin 4-Receptor Ile269Asn Mutation on Weight Loss Response to Dietary, Phentermine and Bariatric Surgery Interventions

The loss of function melanocortin 4-receptor (MC4R) Ile269Asn mutation has been proposed as one of the most important genetic contributors to obesity in the Mexican population. However, whether patients bearing this mutation respond differently to weight loss treatments is unknown. We tested the association of this mutation with obesity in 1683 Mexican adults, and compared the response of mutation carriers and non-carriers to three different weight loss interventions: dietary restriction intervention, phentermine 30 mg/day treatment, and Roux-en-Y gastric bypass (RYGB) surgery. The Ile269Asn mutation was associated with obesity [OR = 3.8, 95% CI (1.5–9.7), p = 0.005]. Regarding interventions, in the dietary restriction group only two patients were MC4R Ile269Asn mutation carriers. After 1 month of treatment, both mutation carriers lost weight: −4.0 kg (−2.9%) in patient 1, and −1.8 kg (−1.5%) in patient 2; similar to the mean weight loss observed in six non-carrier subjects (−2.9 kg; −2.8%). Phentermine treatment produced similar weight loss in six carriers (−12.7 kg; 15.5%) and 18 non-carriers (−11.3 kg; 13.6%) after 6 months of pharmacological treatment. RYGB also caused similar weight loss in seven carriers (29.9%) and 24 non-carriers (27.8%), 6 months after surgery. Our findings suggest that while the presence of a single MC4R loss of function Ile269Asn allele significantly increases obesity risk, the presence of at least one functional MC4R allele seems sufficient to allow short-term weight loss in response to dietary restriction, phentermine and RYGB. Thus, these three different interventions may be useful for the short-term treatment of obesity in MC4R Ile269Asn mutation carriers.


Introduction
Approximately 5% of severe obesity cases are associated with loss of function mutations in the hypothalamic appetite-regulating melanocortin-4 receptor (MC4R), the most common cause of monogenic non-syndromic obesity [1,2]. Ile269Asn has been identified as a loss of function MC4R mutation linked to morbid obesity and type 2 diabetes (T2D) in Latinos [3,4], by impairing both cAMP production and MC4R internalization by β-arrestin [5][6][7]. According to the Genome Aggregation Database (gnomAD), the minor allele frequency of this mutation is 1% in Latinos. Notably, a study of 640,000 multiethnic exomes found the Ile269Asn mutation present only in the Mexican population (minor allele frequency, 1%) and was associated with obesity [8]. Independent studies in this population previously reported this association in both children and adults [9].
There is currently no standardized protocol for obesity treatment in patients with MC4R mutations. Some studies have analyzed the ability of patients with MC4R mutations to lose weight with lifestyle changes, pharmacological treatment, and bariatric surgery. Studies assessing the response to lifestyle changes have reported inconsistent results. On one hand, pediatric patients with obesity and MC4R mutations other than Ile269Asn showed weight loss in response to a controlled and intensive program of restricted dieting and daily physical activity [10], while weight loss was not achieved in an independent study of European children and adolescent MC4R mutation carriers in response to a less intensive dietary program, suggesting the need for personalized treatment based on MC4R genotype [11]. Multiple studies have reported different weight loss success rates in response to bariatric surgery, showing considerable inter-individual variation [12,13]. Most studies analyzing the response to Roux-en-Y gastric bypass (RYGB) report similar short-term weight loss in patients with and without MC4R mutations [14,15], although longterm results showed weight gain in patients with certain MC4R variants [16]. Regarding pharmacological treatment, one of the first reports comparing weight loss according to MC4R genotype was conducted in a reduced number of European patients, reporting similar weight loss in patients with and without MC4R mutations treated with the MC4R agonist setmelanotide [3]. Moreover, the GLP-1 receptor agonist liraglutide was found to induce similar weight loss in obese patients with and without MC4R mutations, suggesting MC4R mutation carriers can be treated with this agonist [17]. Other widely used weight loss drugs such as phentermine have not been tested in individuals with MC4R mutations.
Because of the relatively high frequency of the Ile269Asn mutation in Mexicans with obesity, the purpose of the present study was to establish whether the response to dietary, phentermine and/or bariatric surgery interventions in Mexican patients with obesity is affected by the presence of this mutation.

Case and Control Association
A total of 1683 adults aged 18-81 years (483 normal-weight controls and 1200 obesity cases) were genotyped to assess the association of MC4R Ile269Asn with obesity.
Control individuals mainly belonged to the GEA cohort [18]. Recruitment and inclusion criteria have been previously described, and all selected normal-weight controls were aged > 40 years [18,19]. The obesity group included 75 individuals submitted to dietary intervention, 168 treated with phentermine, and 206 submitted to bariatric surgery. These weight loss interventions were compared in individuals with and without the MC4R Ile269Asn mutation, as described in the flow diagram ( Figure 1). obesity, the purpose of the present study was to establish whether the response to dietary, phentermine and/or bariatric surgery interventions in Mexican patients with obesity is affected by the presence of this mutation.

Case and Control Association
A total of 1683 adults aged 18-81 years (483 normal-weight controls and 1200 obesity cases) were genotyped to assess the association of MC4R Ile269Asn with obesity. Control individuals mainly belonged to the GEA cohort [18]. Recruitment and inclusion criteria have been previously described, and all selected normal-weight controls were aged >40 years [18,19]. The obesity group included 75 individuals submitted to dietary intervention, 168 treated with phentermine, and 206 submitted to bariatric surgery. These weight loss interventions were compared in individuals with and without the MC4R Ile269Asn mutation, as described in the flow diagram ( Figure 1).

Figure 1.
Flow diagram: an overview of Ile269Asn carrier and matched non-carrier selection.

Dietary Intervention
A cohort of 75 Mexican mestizo adults aged 18-60 years with BMI ≥ 30 kg/m 2 were selected for dietary intervention. Inclusion and exclusion criteria were previously published [20]. Briefly, participants received menus and a thirty-day prescription with dietary indications of 750 kcal/d energy restriction based on their habitual total energy expenditure, as previously described [21]. This diet was designed with the following macronutrient distribution: 50% energy content from carbohydrates, 20-25% from protein and 25-30% from fat. Patients received a weekly food supply with 80% of the foods contained in the menu to improve compliance. Participants were advised to maintain their usual level of physical activity during the month of dietary intervention.

Phentermine Intervention
In this prospective, phase IV open-label study, 168 volunteers aged above 18 years and with a BMI ≥ 30 kg/m 2 were recruited. The purpose of the study was to evaluate the efficacy and safety of oral administration of 30 mg of phentermine in obese patients after 6 months of treatment. Inclusion and exclusion criteria were previously described [22]. Patients received medical support and were instructed to follow a 1500 Kcal/d diet (50% energy content from carbohydrates, 30% from fat and 20% from protein) and to perform 20 min/d physical activity.

Dietary Intervention
A cohort of 75 Mexican mestizo adults aged 18-60 years with BMI ≥ 30 kg/m 2 were selected for dietary intervention. Inclusion and exclusion criteria were previously published [20]. Briefly, participants received menus and a thirty-day prescription with dietary indications of 750 kcal/d energy restriction based on their habitual total energy expenditure, as previously described [21]. This diet was designed with the following macronutrient distribution: 50% energy content from carbohydrates, 20-25% from protein and 25-30% from fat. Patients received a weekly food supply with 80% of the foods contained in the menu to improve compliance. Participants were advised to maintain their usual level of physical activity during the month of dietary intervention.

Phentermine Intervention
In this prospective, phase IV open-label study, 168 volunteers aged above 18 years and with a BMI ≥ 30 kg/m 2 were recruited. The purpose of the study was to evaluate the efficacy and safety of oral administration of 30 mg of phentermine in obese patients after 6 months of treatment. Inclusion and exclusion criteria were previously described [22]. Patients received medical support and were instructed to follow a 1500 Kcal/d diet (50% energy content from carbohydrates, 30% from fat and 20% from protein) and to perform 20 min/d physical activity.

Bariatric Surgery
A total of 206 female patients undergoing RYGB were recruited from the Obesity Clinic at the Hospital General de Tlahuac in Mexico City. Interventions included laparoscopic RYGB performed by three surgeons using the same operative techniques; the surgical methods have been described previously [23]. After surgery, moderate effort exercise (250-400 min per week) and moderate-to-high resistance exercise (12-15 repetitions for 3 series/large muscle group) were recommended.

Ile269Asn Carrier and Non-Carrier Matching in the Three Weight-Loss Intervention Groups
Age and BMI matching between mutation carriers and non-carriers (all female) in each intervention group was performed by selecting patients within similar baseline age and BMI ranges, seeking the closest match available. Three non-mutation carrier controls were selected for each mutation carrier. The highest difference between matched carriers and non-carriers was 7 years of age and 3 kg/m 2 BMI in the dietary intervention group, 5 years of age and 3 kg/m 2 BMI in the phentermine intervention group, and 5 years of age and 7 kg/m 2 BMI in the bariatric surgery group.

Anthropometric Measurements
Height and weight were measured following standard protocols with calibrated instruments as previously described [18,19]. BMI was calculated as body weight in kilograms divided by the square of height in meters (kg/m 2 ). Obesity was defined as a BMI ≥ 30 kg/m 2 , class I/II obesity as 30 kg/m 2 ≤ BMI < 40 kg/m 2 , and class III obesity as BMI ≥ 40 kg/m 2 . Normal weight was defined as BMI < 25 kg/m 2 and ≥18.5 kg/m 2 according to World Health Organization (WHO) criteria [24]. All measurements were recorded before and after the dietary, pharmacologic, or bariatric surgery intervention. Weight loss percentage (%WL) was calculated as (initial weight minus post-treatment)/(initial weight) × 100. The percentage of excess weight loss (%EWL) was calculated as (preoperative body weight minus follow-up body weight)/(preoperative body weight minus ideal body weight) × 100. Skeletal muscle mass and fat mass were measured using Inbody 720 multifrequency bioimpedance analysis (Biospace, Co. Cerritos, CA, USA) in the patients submitted to dietary intervention, and with an electric bioimpedance instrument (Omron HBF-514C) in those treated with phentermine.

Biochemical Measurements
Blood samples were drawn after 8-12 h of overnight fasting to determine serum levels of glucose by enzymatic assays. Serum insulin, leptin and adiponectin levels were measured by ELISA in the dietary intervention group, and by BIO-RAD BIO-PLEX Luminex System in the phentermine group. Insulin resistance was estimated using the homeostasis model assessment of insulin resistance (HOMA-IR) [25]. T2D was defined by a prior diagnosis, use of glucose-lowering medications, and/or fasting serum glucose levels ≥ 126 mg/dL [26]. All measurements were performed before the intervention and at different follow-up times.

MC4R Ile269Asn Genotyping
Genomic DNA was isolated from peripheral leukocytes using standard methods. MC4R Ile269Asn genotypes were obtained from the Multi-Ethnic Genotyping Array (MEGA, Illumina, San Diego, CA, USA) for all participants except those in the dietary intervention group, where samples were genotyped using Taqman probes. The variant did not deviate from Hardy-Weinberg equilibrium in any group. Global ancestry was estimated as previously described [19].

Local Ancestry Inference
Local ancestry along chromosome 18 was estimated in the 206 patients from the bariatric surgery group, using 50 European and 50 African individuals from the 1000 Genomes project (Phase 3) and 50 Native Americans from the MAIS cohort as reference populations [27]. Phasing was performed using SHAPEIT v2.17. Local ancestry estimation was performed using RFMix v2 with two EM iterations and a forward-backward threshold of 0.9 [28]. The mean Native American ancestry proportion of a 1Mb segment containing the MC4R Ile269Asn mutation was estimated in individuals with and without the mutation.

Statistical Analysis
Continuous variables are presented as mean ± standard error, and dichotomous variables as frequencies and percentages. Association of the Ile269Asn mutation with obesity was tested using logistic regression. This association was adjusted for sex and admixture including samples with available global ancestry estimations (482 controls and 1062 cases). The distribution of quantitative variables was evaluated using the Kolmogorov-Smirnov test; variables with non-normal distribution were log-transformed. Treatment effects between Ile269Asn mutation carriers and non-carriers were calculated using paired Student t test. Differences in the treatment effect between the groups were calculated using linear regression models. Statistical analyses were performed with the SPSS software package version 15.0 (SPSS, Chicago, IL, USA).

Case-Control Association Study
In the case-control association study, 5/483 normal weight individuals (1%) and 45/1200 individuals with obesity (3.8%) were heterozygous for the MC4R Ile269Asn mutation. Only one homozygous individual who had class III obesity was found. Table 1 shows the association of the Ile269Asn mutation with overall obesity, and stratified according to obesity class. This mutation was significantly associated with increased overall risk of obesity (p = 0.005). Because the sex ratio and mean Native American ancestry proportion were significantly different in cases and controls (Supplementary Table S1), the association was adjusted for sex and admixture remaining statistically significant (p = 0.021). Although Ile269Asn genotypes were most frequent in the obesity class III group (5.5%), the variant was significantly associated with all obesity classes (p < 0.05). Mean Native American ancestry of the chromosome 18 segment containing the MC4R Ile269Asn mutation was estimated in 206 patients with obesity, and was 58% in wildtype homozygous individuals, 79% in heterozygous patients and 100% in the Asn269Asn homozygous patient (Supplementary Table S2).

Response to Dietary Intervention
In the dietary intervention group (n = 75), only two women carried the MC4R Ile269Asn mutation (patient 1: basal BMI of 45.2 kg/m 2 and patient 2: basal BMI of 46.4 kg/m 2 with T2D). Six age and sex matched non-carriers with BMI > 40 kg/m 2 were used as controls for comparisons ( Table 2). After 1 month of dietary intervention, both MC4R mutation carriers lost weight: −4.0 kg (−2.9%) in patient 1, and −1.8 kg (−1.5%) in patient 2; while mean weight loss in the control group was −2.9 ± 0.6 kg (−2.8 ± 0.6%). Weight loss was accompanied by a decrease in fat mass and serum leptin levels in both mutation carriers and controls. Serum glucose levels increased after 1 month of dietary intervention in patient 1, from 99.5 to 105.3 mg/dL, but decreased in patient 2, from 139.6 to 117.9 mg/dL, and in controls (mean −7.6 ± 4.8 mg/dL).

Response to Phentermine Treatment after Six Months
Overall, 113/168 patients who received phentermine treatment completed the 6 month follow-up. Six of the participants (4%, all women) carried the MC4R Ile269Asn mutation. A total of 18 age-, gender-, and initial BMI-matched individuals without the Ile269Asn mutation were included as controls.
After 6 months of phentermine treatment, weight was significantly reduced in both groups (−12.7 ± 2.3 kg, p = 0.003 in MC4R mutation carriers and −11.3 ± 0.9 kg, p < 0.001 in non-carriers), with no significant difference between carriers and non-carriers (p = 0.523; Table 3). Weight loss percentage was also similar in both groups (15.5 ± 2.9% in carriers and 13.5 ± 1.1% in non-carriers, p = 0.250) (Figure 2). Moreover, fat mass percentage decline was very similar in both groups, although the difference between fat mass percentage before and 6 months after phentermine treatment was significant only in non-carriers (p < 0.001). Systolic blood pressure decreased significantly after treatment only in non-carriers (p = 0.027), with no significant difference between carriers and non-carriers (p = 0.100; Table 3).  Regarding glucose metabolism parameters, serum levels of glucose, insulin and HOMA-IR decreased in both MC4R mutation carriers and controls, without significant differences between groups. Notably, in MC4R Ile269Asn mutation carriers, serum leptin levels showed a 4-fold decline after 3 months (basal 24.7 ± 6.9 ng/mL and 6.3 ± 1.6 ng/mL after 3 months, p = 0.018) and a modest increase after 6 months of phentermine treatment, without reaching pre-treatment levels. In contrast, in non-carriers, serum leptin levels showed a moderate decline at 3 months (basal 20.2 ± 1.8 ng/mL and 18.9 ± 3.2 ng/mL after 3 months; p = 0.146), which continued to decline after 6 months (15.2 ± 3.3 ng/mL; p = 0.039). Moreover, differences in leptin levels between groups were significant after 3 months (p = 0.012, Supplementary Table S3), but not after 6 months of phentermine treatment (p = 0.484; Table 3). The most frequent phentermine-related adverse events were categorized as gastrointestinal, neurological or psychiatric. All adverse events were reported as of mild or moderate intensity in both Ile269Asn mutation carriers and non-carriers (Supplementary Table S4).

Response to RYGB Surgery
To determine whether the response to bariatric surgery differs according to the presence of the MC4R Ile269Asn mutation, we genotyped 206 female patients with obesity previously submitted to RYGB. Of these, seven (3.4%) were heterozygous and one was homozygous. Twenty-four age-, sex-and initial BMI-matched non-carriers were used as controls. Six months after the RYGB, %WL and %EWL were similar in carriers (29.9% and 66.6%, respectively) and non-carriers (27.8% and 64.9%, respectively; Table 4), and the %EWL pattern over time was similar in both groups ( Figure 3). Notably, the Asn269Asn homozygous patient had one of the highest baseline weights and BMIs (138 kg and 51.6 kg/m 2 , respectively), but showed a %WL similar to that of the heterozygous and non-carrier patients. However, the homozygous patient showed the lowest %EWL over time (51.5% after 6 months; Figure 3), likely because of the higher baseline weight. Hb1Ac levels did not differ significantly before or after RYGB in carriers and non-carriers.  Figure 3. Percentage of excess weight loss (%EWL) was similar in heterozygous carriers (n = 7) and non-carriers (n = 24) 6 months after RYGB surgery. %EWL was lower in the Asn269Asn homozygous patient (n = 1). Data are presented as mean ± SEM.

Discussion
As previously reported in other Mexican cohorts, the MC4R Ile269Asn mutation was associated with adult obesity in the present study. This mutation was formerly associated with childhood and adult obesity in Mexico [8,9], and in vitro studies showed that it results in complete loss of function of the MC4R protein [7]. Although the Ile269Asn mutation is extremely rare or absent in non-Latino populations [8], mutation carrier frequency was 3.4 and 5.4% in Mexicans with class I/II and III obesity, respectively. This mutation has been proposed as the most important genetic contributor to obesity in the Mexican population [9]. It has been suggested that the MC4R Ile269Asn mutation may have resulted from a founder event in the Native American population. In this regard, local ancestry analyses revealed that the percentage of Native American ancestry of the chromosomal segment containing Ile269Asn was higher in individuals bearing the MC4R Ile269Asn mutation than in non-carriers. Our observation is in agreement with the previous suggestion that Ile269Asn is a Native American founder mutation [9].
The loss of MC4R function may affect the response to dietary intervention through appetite regulation [11,29]. However in the present study, MC4R mutation carriers and controls showed similar weight loss, consistent with previous studies in patients with obesity and heterozygous MC4R mutations submitted to a controlled and intensive program of restricted dieting and daily physical activity for 6 weeks [10]. Together, these findings suggest that the presence of at least one functional MC4R allele is enough to achieve short-term weight loss during a restriction diet.
The present study provides evidence that Ile269Asn carriers with obesity can be successfully treated with phentermine. Specifically, 3 months of phentermine treatment (30 mg) resulted in clinically relevant weight loss, which was similar in Ile269Asn carriers (10.9 ± 3.2%) and non-carriers (8.5 ± 4.3%). Indeed, phentermine treatment led to greater weight loss than that reported with other pharmacological treatments such as liraglutide (5.7 ± 1.4% after 4 months) in patients with MC4R mutations. However it must be Figure 3. Percentage of excess weight loss (%EWL) was similar in heterozygous carriers (n = 7) and non-carriers (n = 24) 6 months after RYGB surgery. %EWL was lower in the Asn269Asn homozygous patient (n = 1). Data are presented as mean ± SEM.

Discussion
As previously reported in other Mexican cohorts, the MC4R Ile269Asn mutation was associated with adult obesity in the present study. This mutation was formerly associated with childhood and adult obesity in Mexico [8,9], and in vitro studies showed that it results in complete loss of function of the MC4R protein [7]. Although the Ile269Asn mutation is extremely rare or absent in non-Latino populations [8], mutation carrier frequency was 3.4 and 5.4% in Mexicans with class I/II and III obesity, respectively. This mutation has been proposed as the most important genetic contributor to obesity in the Mexican population [9]. It has been suggested that the MC4R Ile269Asn mutation may have resulted from a founder event in the Native American population. In this regard, local ancestry analyses revealed that the percentage of Native American ancestry of the chromosomal segment containing Ile269Asn was higher in individuals bearing the MC4R Ile269Asn mutation than in noncarriers. Our observation is in agreement with the previous suggestion that Ile269Asn is a Native American founder mutation [9].
The loss of MC4R function may affect the response to dietary intervention through appetite regulation [11,29]. However in the present study, MC4R mutation carriers and controls showed similar weight loss, consistent with previous studies in patients with obesity and heterozygous MC4R mutations submitted to a controlled and intensive program of restricted dieting and daily physical activity for 6 weeks [10]. Together, these findings suggest that the presence of at least one functional MC4R allele is enough to achieve short-term weight loss during a restriction diet.
The present study provides evidence that Ile269Asn carriers with obesity can be successfully treated with phentermine. Specifically, 3 months of phentermine treatment (30 mg) resulted in clinically relevant weight loss, which was similar in Ile269Asn carriers (10.9 ± 3.2%) and non-carriers (8.5 ± 4.3%). Indeed, phentermine treatment led to greater weight loss than that reported with other pharmacological treatments such as liraglutide (5.7 ± 1.4% after 4 months) in patients with MC4R mutations. However it must be considered that the latter cohort did not receive lifestyle counseling, diet or exercise in-terventions [17]. Additional studies are required to assess whether phentermine treatment induces effective weight loss in subjects with other MC4R mutations.
In addition to weight loss, patients showed decreased fat mass, lower fasting serum glucose levels and improved HOMA-IR measurements in both groups (Ile269Asn carriers and non-carriers). This is important considering that this mutation has been associated with an increased risk of type 2 diabetes in the Latino population [30,31]. Moreover, decreased plasma leptin and increased adiponectin levels may be explained by the weight loss and fat percentage decreases observed in phentermine-treated patients. Notably, the decrease in plasma leptin levels was of significantly higher magnitude in Ile269Asn carriers 3 months after the intervention. However, it was transient, as no differences between carriers and non-carriers were observed 6 months after treatment. Further studies are required to assess whether this is mediated by an effect of phentermine on individuals with impaired MC4R signaling.
The mechanism of action of phentermine is not fully understood; however, it is known that phentermine induces norepinephrine (NE) and dopamine release by inhibiting its recapture in the hypothalamus [32,33]. In the rat model, NE injection into the hypothalamus was found to reduce food consumption, while NE depletion caused by ascending ventral NE axon lesions led to overeating, thus suggesting that NE promotes satiety [34,35]. However, other studies suggest an opposite effect of exogenous NE in the hypothalamus, decreasing satiety in rats [36,37]. These effects could be modulated by the energy homeostasis-regulatory proopiomelanocortin [38][39][40], which in turn regulates MC4R, promoting satiety [41][42][43]. Our finding of similar weight loss in heterozygous carriers and non-carriers suggests that weight loss can be achieved by phentermine treatment in the presence of at least one functional MC4R allele. Conversely, alternative MC4R signaling mechanisms may be involved in the phentermine-induced weight loss effect, such as dopamine signaling in the pre-frontal cortex known to inhibit the appetite.
Several studies have reported that weight loss after RYGB is similar in MC4R mutation carriers and non-carriers [14,15]. In accordance, weight loss response after RYGB was similar in heterozygous Ile269Asn and control patients. Notably, in a heterozygous MC4R murine knockout model, although energy balance was disrupted, causing obesity, these mice responded well to RYGB surgery, suggesting that the presence of a single functional MC4R copy does not impair RYGB-induced weight loss [15]. Altogether, this suggests that heterozygous MC4R Ile269Asn patients are appropriate candidates for RYGB. A small number of studies assessing the effect of bariatric surgery in homozygous MC4R mutation patients have been published, with inconsistent results for different mutations [44]. In the present study, the only Asn269Asn homozygous patient showed a %WL similar to that of heterozygous and non-carrier patients, but the lowest %EWL after 6 months, most likely due to her higher baseline weight. This suggests that Asn269Asn homozygosity does not impair RYGB-induced weight loss at least in the short term. It must be considered that despite the key role of MC4R in obesity, single mutations in MC4R might not be sufficient to significantly impact weight loss outcomes because of the contribution of other variants involved in polygenic obesity [45]. Some of these variants may contribute to obesity through other mechanisms implicated in body weight loss, requiring further study.
Some study limitations must be acknowledged. Firstly, only a small number of patients with the Ile269An mutation were included in each intervention group (n = 2 for diet, 6 for phentermine and 7 for bariatric surgery), and only one homozygous MC4R Asn269Asn carrier was identified in the bariatric surgery group. A larger number of heterozygous and homozygous patients are required to obtain adequate statistical power to confirm our findings. Moreover, it would be important to test a dosage effect in different intervention groups; however, homozygous patients are very rare, even in the Mexican population (0.0001%) [8]. In addition, we cannot rule out the presence of other mutations in the MC4R gene in study participants, because we did not sequence this gene. The phentermine group had a high dropout rate (32.8%). However, this rate is consistent with that of other pharmacotherapy trials for obesity [46,47]. Unfortunately, in the intervention groups, only female Ile269Asn patients were found. Thus, studies including male carriers are necessary because of the known sex differences in obesity and weight loss [48]. Finally, our follow-up after interventions was short (1 to 6 months), and because long-term studies have reported that patients with other MC4R mutations gain weight in the long term, longer follow-up studies are necessary for Ile269Asn carriers.

Conclusions
In conclusion, this is the first study to provide evidence that dietary restriction, pharmacological treatment with phentermine and bariatric surgery induced significant weight loss in Mexican patients with obesity in the short-term, even in those bearing the Ile269Asn mutation. Thus, these three different interventions may be useful for the short-term treatment of obesity in MC4R Ile269Asn mutation carriers.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/genes13122267/s1, Table S1. Comparison of basal anthropometric parameters in participants with obesity and normal weight; Table S2. Local ancestry percentage of a 1Mb segment containing the MC4R locus according to genotype; Table S3. Comparison of anthropometric and biochemical parameters after 3 months of phentermine treatment (30 mg/d) in Ile269Asn carriers and non-carriers; Table S4. Phentermine-related adverse events reported during the study in Ile269Asn carriers and non-carriers.