Population and evolutionary genetics of the PAH locus to uncover overdominance and adaptive mechanisms in phenylketonuria: Results from a multiethnic study

Abstract

Background: Phenylketonuria (PKU) is the most common inborn error of amino acid metabolism in Europe. The reasons underlying the high prevalence of heterozygous carriers are not clearly understood. We aimed to look for pathogenic PAH variant enrichment according to geographical areas and patients' ethnicity using a multiethnic nationwide cohort of patients with PKU in France. We subsequently appraised the population differentiation, balancing selection and the molecular evolutionary history of the PAH locus.

Methods: The French nationwide PKU study included patients who have been referred at the national level to the University Hospital of Nancy, and for whom a molecular diagnosis of phenylketonuria was made by Sanger sequencing. We performed enrichment analyses by comparing alternative allele frequencies using Fisher's exact test with Bonferroni adjustment. We estimated the amount of genetic differentiation among populations using Wright's fixation index (Fst). To estimate the molecular evolutionary history of the PAH gene, we performed phylogenetic and evolutionary analyses using whole-genome and exome-sequencing data from healthy individuals and non-PKU patients, respectively. Finally, we used exome-wide association study to decipher potential genetic loci associated with population divergence on PAH.

Findings: The study included 696 patients and revealed 132 pathogenic PAH variants. Three geographical areas showed significant enrichment for a pathogenic PAH variant: North of France (p.Arg243Leu), North-West of France (p.Leu348Val), and Mediterranean coast (p.Ala403Val). One PAH variant (p.Glu280Gln) was significantly enriched among North-Africans (OR = 23·23; 95% CI: 9·75-55·38). PAH variants exhibiting a strong genetic differentiation were significantly enriched in the 'Biopterin_H' domain (OR = 6·45; 95% CI: 1·99-20·84), suggesting a balancing selection pressure on the biopterin function of PAH. Phylogenetic and timetree analyses were consistent with population differentiation events on European-, African-, and Asian-ancestry populations. The five PAH variants most strongly associated with a high selection pressure were phylogenetically close and were located within the biopterin domain coding region of PAH or in its vicinity. Among the non-PAH loci potentially associated with population divergence, two reached exome-wide significance: SSPO (SCO-spondin) and DBH (dopamine beta-hydroxylase), involved in neuroprotection and metabolic adaptation, respectively.

Interpretation: Our data provide evidence on the combination of evolutionary and adaptive events in populations with distinct ancestries, which may explain the overdominance of some genetic variants on PAH.

Funding: French National Institute of Health and Medical Research (INSERM) UMR_S 1256.

Keywords: Balancing selection; Metabolic adaptation; Overdominance; Phenylketonuria; Population divergence.

Fig. 1

(A) Genomic position and (B) alternative allele frequencies of the three pathogenic PAH gene variants enriched in the North (rs62508588), North-West (rs62516092) and Mediterranean coastal (rs5030857) areas of France and the variant enriched among patients with North-African ancestry (rs62508698).

Fig. 2

(A) Principal component analysis reporting patients included in the French nationwide PKU study. (B) Zoomed view of the Fig. part defined by the inset in panel A. According to PCA-defined ancestry analysis, European-, North-African, and Turkish-ancestry patients represent the three main patients’ ethnicities in the study.

Fig. 3

(A) Principal component analysis according to patients’ ethnicity. (B) The PAH gene variant rs62508698 was significantly enriched among North-African ancestry patients and significantly under-represented among patients with European-ancestry.

Fig. 4

(A) Two-dimension scatter plot reporting the pairwise population differentiation analysis using the Fst index against the overall Fst index (black stars) for PAH gene variants using whole-genome sequencing data from the 2504 subjects of the 1000 Genomes Project phase 3. The matrix at the top of the panel reports the pairwise population Fixation index. (B) Dendrogram with heat map reporting the pairwise population differentiation analysis for PAH gene variants using whole-genome sequencing data from the 1000 genomes project phase 3.

Fig. 5

Haplotype block analysis of the PAH gene using whole-genome sequencing data from the 2504 subjects of the 1000 Genomes Project phase 3. The number of haplotype blocks is reported for each super population.

Fig. 6

Phylogenetic analysis of the PAH locus using 1 kG whole-genome sequencing data.

Fig. 7

World map representation of the alternative allele frequencies of the top differentiated (African- vs East Asian-ancestry populations) and phylogenetically clustered PAH variants.

Fig. 8

(A) Two-dimension scatter plot reporting the pairwise population differentiation analysis using the Fst index against the overall Fst index for PAH gene variants evidenced in the ‘Nancy Reference center Population’. (B) Phylogenetic and timetree analyses of the PAH locus using the ‘Nancy Reference Centre Population’ exome-sequencing data on non-PKU patients.

Fig. 9

(A) Manhattan plot reporting the exome-wide association study for the association with population differentiation and balancing selection pressure (PAH gene variant rs62508698) in haplotype trend regression analysis. The two arrows indicate the top significant loci: SSPO (SCO-spondin, HGNC:21998) and DBH (dopamine beta-hydroxylase, HGNC:2689) genes. Panels (B) and (C) report the genomic context of the SSOP and SBH genes, respectively. The P-values were reported after a symmetric smoothing transformation method using a window radius value of 2, as previously described .