Many pathways lead to disturbed BP control according to genetic study

Surendran P, Drenos F, Young R et al.,
Nat Genet. 2016 Sep 12. doi: 10.1038/ng.3654. [Epub ahead of print]

Background

Despite its prevalence and known links to increased risk of several CV disorders, little is known about specific causal gene variants and pathways that result in high blood pressure (BP).
Genome-wide association studies (GWAS) have identified over 67 genetic regions that were found to be associated with BP or hypertension (HTN). Most of these variants are common (minor allele frequency (MAF) ≥ 0.05, are located in intronic or intragenic regions, and they do not explain more than about 2% of trait variance. These searches have revealed little insight into causal alleles and genes. Low-frequency (0.01 < MAF < 0.05) or rare (MAF ≤ 0.01) single nucleotide variants (SNVs) may have larger phenotypic effects than common SNVs and may shed light on the missing heritability.
This is the largest meta-analysis performed thus far, including ~350 000 individuals, in an attempt to identify new coding variants and loci that influence BP traits and HTN. The Exome chip was used, which contains ~240 000 mostly rare and low-frequency variants.

Main results

• Genotyping of 192763 individuals from 51 studies in a SNV discovery design revealed 51 genomic regions with genome-wide significant evidence of association with at least one of the four tested BP traits: SBP, DBP, pulse pressure (PP) and HTN.
• In an independent replication data set of 18 studies, 81 SNVs were selected for association with DBP (30), SBP (26), PP (19) or HTN (6) as the primary trait (defined as the one with the smallest association P value in a specific one of the database analyses).
• Application of formal replication criteria on these 81 SNVs yielded 17 SNV-BP associations. In addition to these 17, 13 SNV associations were identified that were genome-wide significant in the combined meta-analyses (discovery and replication).
• Four loci had genome-wide significant associations with HTN, as well as with DBP and SBP and one locus was associated with both SBP and PP.
• Out of 25 coding variants picked up in the discovery data, 6 were predicted to be damaging in one of 67 established BP loci, of which at least five are consistent with having a coding causal variant.
• Based on the identification of 30 new loci associated with BP traits, and newly discovered SNVs at 4 loci, and 5 known loci, the new percentage of the trait variance explained amounted to 3.38%, 3.41% and 2.08% respectively for SBP, DPB and PP.
• Three weighted genetic risk scores (GRSs) based on known and newly discovered SBP/DBP/PP variants were created to assess the impact of BP variants on a wide range of CV endpoints/traits. 10 mmHg increase in SBP (OR: 1.39), DBP (OR: 1.62) and PP (OR: 1.70) were all positively and significantly associated with CHD risk, as well as risk of ischemic stroke (ORs 1.93, 2,81 and 2.12 respectively).
• There was no convincing evidence pointing towards the BP-associated variants having an effect on lipid levels, BMI, waist-hip-ratio-adjusted BMI, height, eGFR or HF.

Conclusion

This large genetic study of blood pressure identified new common variants with small effects on BP traits, as well as some of the first rare coding variants of strong effect (>1.5 mmHg) that showed a robust association with BP traits in the general population. The findings suggest that BP control in the general population may occur through different pathway to monogenic disorders of BP. These data augment the trait variance explained with 1.3% for SBP, 1.2% for DBP and 0.93% for PP.
Moreover, this study has identified some rare missense variants that may be interesting candidate genes that warrant follow-up studies.  

Find this article online at Nature Genetics