Investigating the association of the UMOD gene with primary hypertension

Fisher, Simon David (2022) Investigating the association of the UMOD gene with primary hypertension. PhD thesis, University of Glasgow.

Abstract

Hypertension, the chronic elevation of blood pressure, remains a sub-optimally managed condition for a large proportion of people and combined with its high incidence, is a large contributor to morbidity and mortality worldwide. Hypertension precision medicine aims to tailor treatment protocols to individuals in order to reduce side-effects and cost whilst improving efficacy of treatment. A hypertension GWAS by Padmanabhan et al, identified a locus on the cis-promoter of the human UMOD gene, indexed at rs13333226, as associating with reduced urinary UMOD levels, reduced systolic pressure and reduced cardiovascular disease events per copy of the ‘protective’ haplotype. UMOD, a gene expressed almost exclusively in the thick ascending limb of mammalian kidneys was subsequently identified as correlating positively with bioavailability or activity of SLC12A1 (NKCC2), a renal sodium potassium chloride co-transporter, thus associating UMOD with blood pressure through natriuresis and volume retention. In this thesis, we aimed to refine the understanding of the variants which differentially drive expression of UMOD and to elaborate on the mechanisms with associate UMOD levels with blood pressure regulation, such that our findings may inform the rationale for downstream precision medicine experimentation.

Whilst rs13333226 was the discovery variant in the initial GWAS, we hypothesised the nearby variant rs4997081 functionally drives differential expression between alleles, based on previous data. Firstly, we sought to determine the binding affinity of 50-mer oligonucleotides, centered on our target variants, for renal cell nuclear lysate, in order to infer transcriptional activity. By performing electrophoretic mobility shift assays on material derived from both HEK293 and HK2 cells, on rs13333226 and rs4997081, we showed that rs13333226 does not display differential binding affinity for nuclear lysate between genotypes, whilst the ‘risk’ (G) genotype at rs4997081 displayed significantly greater binding affinity versus the protective (C) genotype. Using mass spectrometry on purified binding complexes at rs4997081, we identified a number of potential proteins, including PARP1 as complexing at this site. The addition of TNF-α to these cells caused an abrogation in the differences in binding affinity between the risk and protective alleles at rs4997081. Together these data suggest that PARP1 differentially complexes at rs4997081 in a TNF-α dependent manner to drive differential expression of UMOD.

Following this, we aimed to characterise the association between UMOD mRNA and the renal transcriptome. We performed a low sample-size RNA-sequencing experiment (n=3/group) on human individuals stratified by expression of UMOD. We detected strong positive coexpression between UMOD and NKCC2, which we validated by qRT-PCR in 84 human samples, showing a highly significant correlation. Alongside this coexpression, we also detected significant increases in WNK1/4, KCNJ1/5 and decreases in SGK1 mRNA levels, with each of these genes implicated in renal blood pressure regulation independently in the literature, these findings may indicate additional complementary mechanisms by which UMOD associates with hypertension. By gene set enrichment and pathway analysis we showed enrichment toward NFKβ and TNF-α signaling as associating with UMOD levels, corroborating our prior findings by EMSA and mass spectrometry. Furthermore, analysis of long non-coding RNAs both using bulk RNA sequencing and independent single cell data mining indicated lnc01762 as associating with UMOD expression and thus potentially associating with blood pressure regulation. Together, these data further reinforce the canonical understanding between UMOD and blood pressure regulation through NKCC2, but also indicate a number of novel mechanisms which may be potential therapeutic targets.

Seeking to identify relationships between the UMOD promoter haplotype, UMOD expression and patient phenotype, we performed analysis of a cohort of n=84 human renal samples characterized by Sanger sequencing at both rs13333226 and rs4997081 and for expression of UMOD by qRT-PCR. Our findings suggest that protective haplotype females have significantly reduced systolic blood pressure than either risk haplotype females or risk or protective haplotype males, indicating a potential UMOD sex effect. We stress that the single-measurement blood pressure recordings collected during this sampling do not qualify as reliable, however we believe this finding requires additional research. We furthermore sought to determine relationships between antihypertensive treatments and UMOD mRNA levels and show that treatment with both bendroflumethiazide and amlodipine in this cohort associates with significant reductions in UMOD expression, independently of haplotype. When we removed individuals using these medications from our analysis, the risk genotype at rs4997081 was associated with significantly increased expression of UMOD versus the protective allele.

Aiming to expand on previous work conducted in 2014 using male, 12-week Sv129-Umod-/- mice, where it was shown that knock-out mice have lower baseline blood pressure versus wild type counterparts and blood pressure tolerance to 2% sodium chloride loading, we performed a mirrored study on Sv129-Umod-/- , Sv129-Umod-/+ and wild type Sv129 mice (n=6-8 per group). By weekly blood pressure plethysmography across a 6-week 2% sodium chloride loading period, we show no significant differences in baseline systolic or diastolic blood pressure and no differences in blood pressure tolerance in response to sodium chloride, between homozygous knock-out, heterozygous or wild type mice. By follow-up taqman qRT-PCR on mouse renal RNA, we show that UMOD mRNA expression in the homozygous knock-out animal is not fully abolished, and furthermore that Nkcc2 expression was significantly reduced in sodium chloride treated mice in a Umod gene-dose-dependent manner suggesting compensatory mechanisms were occurring to diminish differences. Based on these initial findings we emphasise the need for further validation studies with increased power.

Finally, we sought to provide novel resources toward the study of UMOD, by generating a stably transfected UMOD human cell line, as currently, no commercially available lines express UMOD. Using full length human UMOD cDNA integrated into a pTARGET plasmid, we generated a HEK293-UMOD cell line with abundant expression of both UMOD mRNA and UMOD protein, significantly elevated above the negligible level of endogenous UMOD expression. Whilst not possible within the time constraints of this thesis, we suggest a number of downstream experiments based on the protein:protein interactions of UMOD within the cell, experiments which could be performed using this novel and valuable experimental resource.

Together, these data provide novel, multi-omics insights as to the relationship between UMOD variants and UMOD mRNA levels, with further basic-science evidence toward possible blood pressure mechanisms. Using a combination of base-directed experiments, transcriptomics and proteomics, for the first time, we provide empirical evidence linking rs4997081 to transcriptional apparatus driving expression of the UMOD gene and identified PARP1 as possibly driving this, in a TNF-α dependent manner. We also show that the expression of UMOD is strongly positively associated with the expression of NKCC2, suggesting this coexpression mechanism underpins the relationship between UMOD and blood pressure which is potentially enhanced by a number of novel targets surrounding WNK and SGK based signaling. We further enrich understanding of the relationship between UMOD and blood pressure by suggesting that rs4997081, not rs13333226, should be targeted for precision medicine and that the mechanism driving the pathway between TNF-α, UMOD and NKCC2 may be a candidate for precision pharmacogenomics. Together, these data provide insights toward the mechanism of blood pressure regulation at the UMOD locus and set a basis for research with increased sample size to further validate these novel findings.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from the British Heart Foundation.
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Cardiovascular & Metabolic Health
• Funder's Name: Wellcome Trust (WELLCOTR)
• Supervisor's Name: McBride, Dr. Martin and Graham, Dr. Delyth
• Date of Award: 2022
• Depositing User: Theses Team
• Unique ID: glathesis:2022-83078
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 11 Aug 2022 15:22
• Last Modified: 11 Aug 2022 15:22
• Thesis DOI: 10.5525/gla.thesis.83078
• URI: https://theses.gla.ac.uk/id/eprint/83078

Related URLs

• Article DOI
• Article DOI