Prioritization of Kidney Cell Types Highlights Myofibroblast Cells in Regulating Human Blood Pressure

Abstract

Introduction: Blood pressure (BP) is a highly heritable trait with over 2000 underlying genomic loci identified to date. Although the kidney plays a key role, little is known about specific cell types involved in the genetic regulation of BP.

Methods: Here, we applied stratified linkage disequilibrium score (LDSC) regression to connect BP genome-wide association studies (GWAS) results to specific cell types of the mature human kidney. We used the largest single-stage BP genome-wide analysis to date, including up to 1,028,980 adults of European ancestry, and single-cell transcriptomic data from 14 mature human kidneys, with mean age of 41 years.

Results: Our analyses prioritized myofibroblasts and endothelial cells, among the total of 33 annotated cell type, as specifically involved in BP regulation (P < 0.05/33, i.e., 0.001515). Enrichment of heritability for systolic BP (SBP) was observed in myofibroblast cells in mature human kidney cortex, and enrichment of heritability for diastolic BP (DBP) was observed in descending vasa recta and peritubular capillary endothelial cells as well as stromal myofibroblast cells. The new finding of myofibroblast, the significant cell type for both BP traits, was consistent in 8 replication efforts using 7 sets of independent data, including in human fetal kidney, in East-Asian (EAS) ancestry, using mouse single-cell RNA sequencing (scRNA-seq) data, and when using another prioritization method.

Conclusion: Our findings provide a solid basis for follow-up studies to further identify genes and mechanisms in myofibroblast cells that underlie the regulation of BP.

Keywords: blood pressure; cell-type; genome; kidney myofibroblast; scRNA-seq.

Graphical abstract

Figure 1

Flow diagram depicting the design of this study. Replication Step A entails replicating the discovery findings using 7 independent combinations of GWAS and single-cell RNA sequencing datasets, including mature and fetal human kidney datasets, as well as Tabula Sapiens and Muris datasets. These datasets are described in the references. Replication Step B involves replicating the discovery findings using another computational method, specifically MAGMA. EU-Meta-SBP and -DBP GWASs are the meta-analysis results of over 1 million individuals from European ancestry. EAS-SBP and EAS-DBP are GWAS results of blood pressure traits in over 145,000 East-Asian individuals. DBP, diastolic blood pressure; EAS, East-Asian; EU, European; GWAS, genome-wide association studies; MAGMA, Multi-marker Analysis of GenoMic Annotation; SBP, systolic blood pressure.

Figure 2

Discovery cell-type prioritization estimated by LDSC in systolic and diastolic BP. Cell-type prioritization results are generated using mature human kidney data. Bars represent the strength of association (−log [10] P) of cell-type expression specificity with heritability for BP phenotypes. Cell types are listed based on −log (10) P in descending order for SBP. The dashed black line indicates the Bonferroni corrected significance threshold, that is, P-value < 0.05/33. BP, blood pressure; DBP, diastolic blood pressure; LDSC, linkage disequilibrium score; SBP, systolic blood pressure.