Post by Admin on Nov 18, 2023 22:03:38 GMT
Material and methods
Population sample
134 unrelated individuals from Rio de Janeiro (Southeastern Brazil) with different pigmentation phenotypes of skin, hair and eyes (Table 1) were selected for MC1R genotyping. The population of study was selected non-randomly and the classification followed a qualitative analysis based on predetermined parameters of skin, hair and eye color. Collecting the biological material and the DNA extraction were described in a previous study [25]. The project was approved by the Ethics in Research Committee of Clementino Fraga Filho Hospital/UFRJ (CEP—MEMO—n.° 536/10). MC1R variants data from genetically unrelated individuals from parental populations such as Africans (GWD, LWK, MSL and YRI), Asians (CDX, CHB, CHS and JPT) and Europeans (FIN, GBR, IBS and TSI) were selected from the 1000 Genomes Project Consortium Phase 3 (for population keys see S1 Table) [26].
Table 1
Phenotypic frequency of admixed population from Rio de Janeiro.
Parameter Individuals / n = 134
Hair, n (%)
Red 14(0.10)
Blond 25(0.19)
Light brown 16(0.12)
Dark brown 33(0.25)
Black 46(0.34)
Skin, n (%)
Light 75(0.56)
Intermediate 35(0.26)
Dark 24(0.18)
Eye, n (%)
Blue 21(0.16)
Green/Hazel 30(0.22)
Dark 83(0.62)
Gender, n (%)
Female 63(0.47)
Male 71(0.53)
Age, (years)
Mean (SD) 30.87(12.24)
DNA genotyping and sequences analyses
Sanger sequencing was performed in all samples to encompass the human MC1R coding region. Amplification of the regions was performed with the GeneAmp High Fidelity enzyme of Applied Biosystems™ (30ng of DNA, 1x buffer, 1.5 mm MgCl2, 10% DMSO, 0.2mm dNTP, 1.5U of enzyme, 0.3μM of each primer—forward: 5’-GGCAGCACCATGAACTAAGCAG-3’ and reverse: 5’-CAGGGTCACACAGGAACCAGAC-3’ final volume of 50μl—cycling: 94°C 2min; 30x (94°C 20s, 63°C 20s, 72°C 1min) 72°C 7min. The amplified product was purified by PCR Cleanup kit (AxygenTM) and sequenced using Big Dye Terminator (Thermo Fisher, CA). The products were run in Applied Biosystems™ ABI 3130xl, using six primers designed by our group (5’-GAAGAACTGTGGGGACCTGGA-3’, 5’-CAGGAAGCAGAGGCTGGACAG-3’, 5’-ATGTACTGCTTCATCTGCTGC-3’, 5’-CAGGATGGTGAGGGTGACAGC-3’, 5’-TCCTGGCTATGCTGGTGCTCA-3’, 5’-ACACAATATCACCACCTCCCTCT-3’) which covered the intronless MC1R coding region and part of 3’UTR at least twice. Sequences from both strands were aligned with the SNP-contained GenBank consensus sequence format from Homo sapiens chromosome 16, reference assembly, complete sequence (GenBank version NC_000016.8 GI:51511732), using the blast tool from Geneious Pro 4.7 software (Biomatters). The region of HVI of mitochondrial DNA was sequenced based on the protocol described by another group [10] and the classification of haplogroups was achieved through the command line version of Haplogrep v.2.2.6 [27]. The correspondent geographic region for haplogroups was determined through the MITOMAP database [28]. For the parental populations analysis, all the polymorphisms from MC1R gene (89985667–89986632bp) were extracted from The 1000 Genomes database (ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/) using the VCFtools v.0.1.15 [29] and the variant annotations were obtained by SnpEff v.4.5 package [30].
Statistical analysis
The allele frequencies and the Hardy-Weinberg equilibrium (HWE) were calculated using Adegenet v. 2.1.3 and Pegas v.0.13 packages [31, 32], respectively. The phylogenetic tree was obtained by applying the neighbor joining method (NJ function) using ape v.5.3 package [33]. To evaluate the clusters of skin color individuals from RJ and the 1000 Genomes populations, the nonmetric multidimensional scaling (NMDS) and its respective stress plot were conducted with the vegan v.2.5.6 package based on two dimensions (K), kulczynski distance and 25 as a maximum number of random starts. The phylogenetic tree and all the multidimensional scaling plots were based on the Fst genetic distance that were performed through Nei’s pairwise Fst calculation using the pairwise.neifst function and the respective confidence interval, boot.ppfst () function, both from hierfstat v.0.5.7 package [34]. All the analyzes above were performed using the R-studio v.1.3.1056. The R packages ggplot and ggtree were used to generate the graphs and trees, respectively.
MC1R mutational analysis
Mutational analyses were performed using two criteria: i. Based on the protein sequence; ii. Based on the protein three-dimensional (3D) structure. For the sequence-based predictions, the following disease-association predictors were used: PolyPhen [35], PON-P2 [36] and Meta-SNP [37], based on the Uniprot ID Q01726. For the structure-based predictions, the following predictors were used: DUET [38] and DynaMut [39]. DUET includes a consensus score and the individual scores for mCSM and SDM predictors.
MC1R model construction
The MC1R sequence was retrieved from the UniProt database (ID: Q01726) [40]. Template search with BLAST and HHBlits against the Protein Data Bank (PDB) [41] were performed using the SWISS-MODEL server [42]. The best template was the crystal structure of the human melanocortin receptor 4 (MC4R) (PDB ID: 6w25) with sequence identity of 50.18% and alignment coverage of 0.87. Since this crystal did not cover a small region (residues 29–38) of the target protein containing a mutation site on position 35, we selected an additional template, the muscarinic acetylcholine receptor M2 (PDB ID: 5zk8), with 20.21% sequence identity but a slightly higher coverage value of 0.89. Multiple template modelling of the wild-type (WT) MC1R was performed using MODELLER [43] version 9.23, using standard parameters. The best model was selected based on DOPE score [44]. Modelling of the five selected mutants was also performed through MODELLER, using as template the generated WT model.
Population sample
134 unrelated individuals from Rio de Janeiro (Southeastern Brazil) with different pigmentation phenotypes of skin, hair and eyes (Table 1) were selected for MC1R genotyping. The population of study was selected non-randomly and the classification followed a qualitative analysis based on predetermined parameters of skin, hair and eye color. Collecting the biological material and the DNA extraction were described in a previous study [25]. The project was approved by the Ethics in Research Committee of Clementino Fraga Filho Hospital/UFRJ (CEP—MEMO—n.° 536/10). MC1R variants data from genetically unrelated individuals from parental populations such as Africans (GWD, LWK, MSL and YRI), Asians (CDX, CHB, CHS and JPT) and Europeans (FIN, GBR, IBS and TSI) were selected from the 1000 Genomes Project Consortium Phase 3 (for population keys see S1 Table) [26].
Table 1
Phenotypic frequency of admixed population from Rio de Janeiro.
Parameter Individuals / n = 134
Hair, n (%)
Red 14(0.10)
Blond 25(0.19)
Light brown 16(0.12)
Dark brown 33(0.25)
Black 46(0.34)
Skin, n (%)
Light 75(0.56)
Intermediate 35(0.26)
Dark 24(0.18)
Eye, n (%)
Blue 21(0.16)
Green/Hazel 30(0.22)
Dark 83(0.62)
Gender, n (%)
Female 63(0.47)
Male 71(0.53)
Age, (years)
Mean (SD) 30.87(12.24)
DNA genotyping and sequences analyses
Sanger sequencing was performed in all samples to encompass the human MC1R coding region. Amplification of the regions was performed with the GeneAmp High Fidelity enzyme of Applied Biosystems™ (30ng of DNA, 1x buffer, 1.5 mm MgCl2, 10% DMSO, 0.2mm dNTP, 1.5U of enzyme, 0.3μM of each primer—forward: 5’-GGCAGCACCATGAACTAAGCAG-3’ and reverse: 5’-CAGGGTCACACAGGAACCAGAC-3’ final volume of 50μl—cycling: 94°C 2min; 30x (94°C 20s, 63°C 20s, 72°C 1min) 72°C 7min. The amplified product was purified by PCR Cleanup kit (AxygenTM) and sequenced using Big Dye Terminator (Thermo Fisher, CA). The products were run in Applied Biosystems™ ABI 3130xl, using six primers designed by our group (5’-GAAGAACTGTGGGGACCTGGA-3’, 5’-CAGGAAGCAGAGGCTGGACAG-3’, 5’-ATGTACTGCTTCATCTGCTGC-3’, 5’-CAGGATGGTGAGGGTGACAGC-3’, 5’-TCCTGGCTATGCTGGTGCTCA-3’, 5’-ACACAATATCACCACCTCCCTCT-3’) which covered the intronless MC1R coding region and part of 3’UTR at least twice. Sequences from both strands were aligned with the SNP-contained GenBank consensus sequence format from Homo sapiens chromosome 16, reference assembly, complete sequence (GenBank version NC_000016.8 GI:51511732), using the blast tool from Geneious Pro 4.7 software (Biomatters). The region of HVI of mitochondrial DNA was sequenced based on the protocol described by another group [10] and the classification of haplogroups was achieved through the command line version of Haplogrep v.2.2.6 [27]. The correspondent geographic region for haplogroups was determined through the MITOMAP database [28]. For the parental populations analysis, all the polymorphisms from MC1R gene (89985667–89986632bp) were extracted from The 1000 Genomes database (ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/) using the VCFtools v.0.1.15 [29] and the variant annotations were obtained by SnpEff v.4.5 package [30].
Statistical analysis
The allele frequencies and the Hardy-Weinberg equilibrium (HWE) were calculated using Adegenet v. 2.1.3 and Pegas v.0.13 packages [31, 32], respectively. The phylogenetic tree was obtained by applying the neighbor joining method (NJ function) using ape v.5.3 package [33]. To evaluate the clusters of skin color individuals from RJ and the 1000 Genomes populations, the nonmetric multidimensional scaling (NMDS) and its respective stress plot were conducted with the vegan v.2.5.6 package based on two dimensions (K), kulczynski distance and 25 as a maximum number of random starts. The phylogenetic tree and all the multidimensional scaling plots were based on the Fst genetic distance that were performed through Nei’s pairwise Fst calculation using the pairwise.neifst function and the respective confidence interval, boot.ppfst () function, both from hierfstat v.0.5.7 package [34]. All the analyzes above were performed using the R-studio v.1.3.1056. The R packages ggplot and ggtree were used to generate the graphs and trees, respectively.
MC1R mutational analysis
Mutational analyses were performed using two criteria: i. Based on the protein sequence; ii. Based on the protein three-dimensional (3D) structure. For the sequence-based predictions, the following disease-association predictors were used: PolyPhen [35], PON-P2 [36] and Meta-SNP [37], based on the Uniprot ID Q01726. For the structure-based predictions, the following predictors were used: DUET [38] and DynaMut [39]. DUET includes a consensus score and the individual scores for mCSM and SDM predictors.
MC1R model construction
The MC1R sequence was retrieved from the UniProt database (ID: Q01726) [40]. Template search with BLAST and HHBlits against the Protein Data Bank (PDB) [41] were performed using the SWISS-MODEL server [42]. The best template was the crystal structure of the human melanocortin receptor 4 (MC4R) (PDB ID: 6w25) with sequence identity of 50.18% and alignment coverage of 0.87. Since this crystal did not cover a small region (residues 29–38) of the target protein containing a mutation site on position 35, we selected an additional template, the muscarinic acetylcholine receptor M2 (PDB ID: 5zk8), with 20.21% sequence identity but a slightly higher coverage value of 0.89. Multiple template modelling of the wild-type (WT) MC1R was performed using MODELLER [43] version 9.23, using standard parameters. The best model was selected based on DOPE score [44]. Modelling of the five selected mutants was also performed through MODELLER, using as template the generated WT model.