Bivariate Genome-wide Association Study Of Cognition And Blood Pressure Phenotypes

Objective:Hypertension and cognitive dysfunction are common comorbidities in the elderly,which have brought a serious burden of disease to society and become major public health problems.Studies have indicated that the two diseases have some common pathogenesis,thus it is crucial to explore their common influencing factors.And genetic factors,as an important part,have attracted more and more attention.At present,many studies have explored the genetic loci related to blood pressure and cognitive function separately,while fewer studies have focused on their shared genetic background.Studies have shown there existed a genetic correlation between cognitive function and blood pressure,besides,considering that a gene may be involved in the regulation and expression of multiple phenotypes,and the effect of most genes are weak,thus we speculate that there may be some shared genetic loci,genes,and biological pathways between the two phenotypes.This study used data from twins in Qingdao and used a bivariate genome-wide association study（GWAS）to find the shared single nucleotide polymorphisms（SNPs）,genes,and pathways between cognition-systolic blood pressure（SBP）,cognition-diastolic blood pressure（DBP）,cognition-mean arterial pressure（MAP）,and cognition-pulse pressure（PP）,and thus to provide genetic epidemiology clues to the discovery of more molecular biological mechanisms related to the incidence of hypertension and cognitive impairment.Methods:A total of 378 pairs of twins were included in this study,including 241 pairs of monozygotic twins,and 137 pairs of dizygotic twins,dizygotic twins were used for GWAS studies.Cognitive function was assessed using the Montreal Cognitive Assessment Scale（Mo CA）,blood pressure was detected using a mercury sphygmomanometer.Physical examination and blood sample collection were conducted for twins after fasting,and the distinction of monozygotic and dizygotic was identified by gender,blood,and microsatellite DNA gene scanning and typing techniques.Mx software was used to build a bivariate Cholesky decomposition model to assess the effects of genetic and environmental influences on phenotyping.137 pairs of dizygotic were genotyped using the Infinium Omni2.5 Exome-8v1.2 Bead Chip-Genotyping Chip（Illumina,San Diego,California,USA）.Impute2 software was used to impute untyped SNPs.SNPs-based analysis was used Genome-wide efficient mixed model analyses（GEMMA）.Regional plots for Top SNPs associated with cognition-blood pressure phenotypes were made using Locuszoom,functional annotations were made using Halporeg,and Regulome DB was used for functional scoring.Super Enhancer（SE）analysis was performed using the SEdb database for SNPs that were significantly associated with cognition-blood pressure phenotypes before and after imputation.VEGAS2（Versatile Gene-based Association Study-2）and PASCAL were used for gene-based analysis,and pathway enrichment analysis,respectively.Results:A total of 378 pairs of twins were included in this study,including 241 pairs of monozygotic twins and 137 pairs of dizygotic twins.The mean age of the participants was51.5±7.6 years,the mean score of cognitive function was 20.7±5.0,and the mean values of SBP,DBP,PP,and MAP were 130.7±20.1（mm Hg）,83.0±10.9（mm Hg）,47.7±14.7（mm Hg）,98.9±12.8（mm Hg）,respectively.（1）The results of the SNPs-based analysis showed that rs72815554 was significantly associated with the cognition-PP phenotype before imputation(P<5×10^-8).In addition,there were 26,33,53,and 32 SNPs associated with cognition-SBP,cognition-DBP,cognition-PP,cognition-MAP phenotypes at the suggestive significant association level(P<1×10^-5),respectively After imputation,4 and 9 SNPs were significantly associated with the cognitive-SBP and cognitive-PP phenotypes,respectively,and rs72815554 was still significantly associated with the cognition-PP phenotype(P=7.42×10^-8).In addition,there were 151,139,272,and 188 SNPs associated with cognition-SBP,cognition-DBP,cognition-PP,and cognition-MAP phenotypes at the suggestive significant association level,respectively.（2）Regional plots suggested that there existed SNPs with highly linkage disequilibrium（LD）in the upstream and downstream 400Kb regions of rs10998339,rs72815554,rs11665292,and rs10823231.Functional annotation and Regurome DB scoring found that three SNPs,including rs10998347,rs12153038,and rs10998295,were located in the transcription factor binding region,suggesting that these three loci may be involved in the transcriptional regulation process.The results of the SE analysis for 13 SNPs that significantly related to the cognition-blood pressure phenotype indicated that rs7574283and rs58113664 were located in the SE region,besides,26 and 18 SEs associated with the two SNPs were expressed in adrenal glands,arteries,atrial tissue,brain,nerves and blood cells,respectively.（3）The results of the gene-based analysis indicated that there were 1108,1154,1071,and1102 genes related to cognition-SBP,cognition-DBP,cognition-PP,and cognition-MAP phenotypes at the suggestive significant association level（P<0.05）.In addition,we found that some genes were associated with all the four cognition-blood pressure phenotypes,such as HMGN1,ELP2,TEX26,TJP3,UNC79,etc.（4）Pathway-enrichment analysis showed that there were 641,630,900,and 555 pathways related to the cognition-SBP,cognition-DBP,cognition-PP,and cognition-MAP phenotypes at the significant association level（P<0.05）.In addition,several pathways were co-associated with multiple cognition-blood pressure phenotypes,such as the signaling by ERBB4 pathway,the neuroactive ligand receptor interaction pathway,the epidermal growth factor/epidermal growth factor receptor pathway,the NF-Kappa B activation pathway,the B-cell receptor signaling pathway,the circadian clock pathway,and the peroxisome proliferation activator receptor（PPAR）αactivated gene expression pathway.Conclusions:Several SNPs,genes,and pathways associated with cognition-blood pressure phenotypes were discovered.In addition,we found several SNPs that may be involved in the transcriptional regulation process,as well as SEs.Our findings provide a basis for exploring the shared genetic basis between blood pressure and cognitive function and provide clues to the discovery of more molecular biological mechanisms related to the incidence of hypertension and cognitive impairment.