Genetic variants in genes involved in creatine biosynthesis in patients with severe obesity or anorexia nervosa

Abstract

Increased thermogenesis in brown adipose tissue might have an obesity-reducing effect in humans. In transgenic mice, depletion of genes involved in creatine metabolism results in disrupted thermogenic capacity and altered effects of high-fat feeding on body weight. Data analyses of a sex-stratified genome-wide association study (GWAS) for body mass index (BMI) within the genomic regions of genes of this pathway (CKB, CKMT1B, and GATM) revealed one sex-dimorphic BMI-associated SNP in CKB (rs1136165). The effect size was larger in females than in males. A mutation screen of the coding regions of these three candidate genes in a screening group (192 children and adolescents with severe obesity, 192 female patients with anorexia nervosa, and 192 healthy-lean controls) identified five variants in each, CKB and GATM, and nine variants in the coding sequence of CKMT1B. Non-synonymous variants identified in CKB and CKMT1B were genotyped in an independent confirmation study group (781 families with severe obesity (trios), 320 children and adolescents with severe obesity, and 253 healthy-lean controls). In silico tools predicted mainly benign yet protein-destabilizing potentials. A transmission disequilibrium test in trios with severe obesity indicated an obesity-protective effect of the infrequent allele at rs149544188 located in CKMT1B. Subsequent correlation analyses in 1,479 individuals of the Leipzig Obesity BioBank revealed distinct correlations of CKB with the other two genes in omental visceral adipose tissue (VAT) and abdominal subcutaneous adipose tissue (SAT). Furthermore, between-subject comparisons of gene expression levels showed generally higher expressions of all three genes of interest in VAT than in SAT. Future in vitro analyses are needed to assess the functional implications of these findings.

Keywords: BAT; GWAS; TDT; creatine metabolism; in silico.

FIGURE 1

Experimental procedure of the genetic study. The genetic study comprised sex-specific analyses for CKB, CKMT1B and GATM as well as a mutation screen in children and adolescents with severe obesity, healthy-lean controls, and patients with anorexia nervosa. In silico analyses were used to predict the functional implications of the detected variants.

FIGURE 2

p-values of SNPs analyzed in GWAS for BMI or AN located within the genes of interest and their adjacent regions. Here, the negative logarithmic p-values extracted from the sex-stratified BMI GWAS [Pulit et al., 2019; (A, B)] and the GWAS for AN (Watson et al., 2019; (C, D) are plotted against the genomic positions of the SNPs located within CKB (A, C); dark gray), CKMT1B [(B, D); dark gray] and GATM [(B, D); black] as well as their 500 kb adjacent regions (light gray; GRCh37; see Table 1; Supplementary Table S2). For the data regarding the sex-stratified BMI GWAS, the negative logarithmic p-values are presented for the female (dot), male (square), and combined (triangle) analyses. The threshold for genome-wide significance is indicated by the dotted lines (-log₁₀(p) > 7.3).

FIGURE 3

Z-scores for BMI-associated SNPs located in CKB and the 500 kb adjacent region. Here, the absolute z-scores were plotted against the chromosomal position of each SNP (GRCh37). After the extraction of the sex-stratified BMI GWAS data (Pulit et al., 2019), ambiguous SNPs were removed. Subsequently, the effect directions were harmonized, and the z-scores were calculated for each genome-wide and thus BMI-associated SNP within the region of CKB (black) and its 500 kb surrounding region (gray). A SNP is considered sex-dimorphic if |z-score| ≥ 3.

FIGURE 4

Correlations of CKB, CKMT1B, and GATM with clinical parameters. The presented data are based on RNA-sequencing data of omental visceral (A) and abdominal subcutaneous (B) adipose tissues from 1,479 individuals of the LOBB. The correlations were calculated with Pearson’s correlation coefficient. p-values were adjusted for multiple comparisons using Hommel’s method (Hommel, 1988). Positive correlations are shown in blue, while negative correlations are represented in red. The size of the dot refers to the degree of correlation. Significance; p_adj. < 0.001 (***), p_adj. < 0.01 (**), and p_adj. < 0.05 (*). BMI: body mass index. CrP: c-reactive protein. FPI: fasting plasma insulin. FPG: fasting plasma glucose. HbA1C: hemoglobin A1C. HDL: high-density lipoprotein. HOMA-IR: homeostatic model assessment for insulin resistance. LDL: low-density lipoprotein. NEFA: non-esterified fatty acids. WHR: waist-to-hip ratio.

FIGURE 5

Gene expression of CKB, CKMT1B and GATM in probands with and without obesity. Expression data of CKB (A), CKMT1B (B) and GATM (C) are based on 1,479 probands with and without obesity from the LOBB. As the data were not normally distributed in all instances, non-parametric Kruskal–Wallis one-way ANOVAs were performed to check whether the location parameters of the distributions were the same in each group. The corresponding effect size was calculated as epsilon square, while pairwise comparisons were performed with Dunn’s test. p-values were adjusted based on Hommel’s method (Hommel, 1988). SAT: subcutaneous adipose tissue. VAT: visceral adipose tissue.