Body weight and height are regulated by both genetic and environmental factors. Obesity (defined via a body mass index, [BMI] ≥ 30 kg/m²) and the eating disorder anorexia nervosa (defined by ICD-10 or DSM 5 criteria) are located at the opposite ends of the body weight distribution. Extremely obese children and adolescents have an increased body height. Children and adolescents with short normal stature (SNS) are frequently lean and thus represent a phenotype that might be regarded as opposite to extreme obesity. 

Obesity is a global health hazard with increasing prevalence. Anorexia nervosa (AN) is a mental disorder characterized by extremely low body weight, the highest mortality rate of all psychiatric disorders, and higher occurrences in females than in males. SNS is the most common reason for children or adolescents with a body height percentile (BH%) for age and sex less than 5^th to consult an endocrinologist.

Studies of genetic effects on complex diseases have initially been focused on monogenic causes. Nowadays, improvements in sequencing technologies and computational predictions provide insights into the polygenic causes of complex disorders.

In this thesis, variants in three candidate genes (polypyrimidine tract binding protein 2 gene [PTBP2] [1], melanocortin 3 receptor gene [MC3R] [2], lipocalin-2 gene [LCN2] [3]) were analyzed in children and adolescents with extreme obesity, or anorexia nervosa (AN) or short normal stature (SNS). Sanger sequencing revealed synonymous variants (SVs) in all three genes and non-synonymous variants (NSVs) in MC3R and LCN2. In-silico and in-vitro functional analyses for detected variants ensued. We detected variants that may have functional consequences on protein or mRNA structure or function. Some of them were associated with BMI or AN or body height. In detail:

(1) In PTBP2 two novel SVs, one intronic single nucleotide polymorphism (SNP), and an intronic deletion located in the highly conserved region may be functionally relevant. Analysis based on sex-stratified BMI genome-wide association study (GWAS) summary statistics implied a more pronounced effect on body weight regulation in PTBP2 in males than in females. [1]

(2) In MC3R three rare loss-of-function (LoF) variants were only found in severely obese individuals. One novel highly conserved NSV, that was predicted to increase protein stability, was detected in a single lean female. In the individuals with SNS, deviation from Hardy-Weinberg Equilibrium was detected for one variant (p.Val44Ile). We observed an increased BMI in individuals with SNS homozygous for the p.44Ile allele. The effect did not remain significant after Bonferroni correction. The detected LoF variant may suggest that dysfunction in MC3R is associated with decreased body height, obesity and delayed puberty. [2]

(3) We detected fourteen LCN2-variants (five NSVs). Two of them are located in the highly conserved signal peptide region and might induce functional consequences. The secondary structure change of lipocalin-2 due to LCN2-p.Val89Ile may decrease solubility and resulted in low lipocalin-2 level in a heterozygous female recovered from AN. Lean individuals had lower lipocalin-2 levels compared to patients with obesity. Lipocalin-2 serum levels are associated with BMI variation; in obesity levels had the highest variability. [3]

Additionally, we estimated the prevalence of putative pathogenic variants in the leptin gene using datasets extracted from gnomAD V2.1.1 in different populations. [4]

We also detected an enrichment of GWAS derived BMI SNPs on genomic loci of circRNA in comparison to non-significant variants. This effect was more pronounced in females than in males. Apart from BMI nine additional GWAS studies were analyzed; four of them (including AN and body height) showed an effect similar to the BMI analysis [5].

Our studies provided evidence that genetic factors can affect body weight and height, as well as the eating disorder AN.