A Single Nucleotide Polymorphism in SH2B3/LNK Promotes Hypertension Development and Renal Damage

Abstract

Rationale:

SH2B3 is an adaptor protein that negatively regulates cytokine signaling and cell proliferation. A common missense single nucleotide polymorphism in SH2B3 (rs3184504) results in substitution of tryptophan (Trp) for arginine (Arg) at amino acid 262 and is a top association signal for hypertension in human genome-wide association studies. Whether this variant is causal for hypertension, and if so, the mechanism by which it impacts pathogenesis is unknown.

Objective:

Test the hypothesis that the Trp-encoding allele of rs3184504 promotes hypertension development and end-organ damage through loss of SH2B3-mediated repression of cytokine signaling to enhance T cell activation.

Methods and Results:

We used CRISPR-Cas9 technology to create mice homozygous for the major (Arg/Arg) and minor (Trp/Trp) alleles of this SH2B3 polymorphism. Trp/Trp mice exhibited 10 mmHg higher systolic blood pressure (BP) during chronic angiotensin II (Ang II) infusion compared to Arg/Arg controls. Renal injury and perivascular fibrosis were exacerbated in Trp/Trp mice compared to Arg/Arg controls following Ang II infusion. In addition, renal and ex vivo stimulated splenic CD8⁺ T cells from Ang II-infused Trp/Trp mice produced significantly more interferon gamma (IFNγ) compared to Arg/Arg controls. Interleukin-12 (IL-12)-induced IFNγ production was greater in Trp/Trp compared to Arg/Arg CD8⁺ T cells. In addition, IL-12 enhanced Stat4 phosphorylation to a greater degree in Trp/Trp compared to Arg/Arg CD8⁺ T cells, suggesting that Trp-encoding SH2B3 exhibits less negative regulation of IL-12 signaling to promote IFNγ production. Finally, we demonstrated that a multi-SNP model genetically predicting increased SH2B3 expression in lymphocytes is inversely associated with hypertension and hypertensive chronic kidney disease in humans.

Conclusions:

Taken together, these results suggest that the Trp encoding allele of rs3184504 is causal for BP elevation and renal dysfunction, in part through loss of SH2B3-mediated repression of T cell IL-12 signaling leading to enhanced IFNγ production.

Graphical Abstract

INTRODUCTION

Hypertension is a leading risk factor for death and disability-adjusted life years lost worldwide.^{1, 2} Despite current therapy, nearly half of all hypertensive individuals remain inadequately controlled. Even with adequate control of blood pressure (BP) an elevated risk of cardiovascular events remains.^3–5 Hence, there is an urgent unmet need for additional therapeutic options. Hypertension is a complex multi-system disease with genetic and environmental contributions linked to dysfunction of various organs including kidneys, vasculature, and/or central nervous system.⁶ In addition, we and others have described a key role for inflammation and innate and adaptive immune cells in the pathophysiology of hypertension and the associated end-organ damage.⁷

BP has heritability estimates of 30–60% and is primarily polygenic.⁸ Although most individual single nucleotide polymorphisms (SNPs) have modest effects, understanding the contribution and effects of individual genetic variants has the potential to yield important insights into the pathophysiology of disease. The vast majority of SNPs associated with hypertension have not been tested for causal or mechanistic roles in BP regulation.⁹ Genome-wide association studies have identified a strong association between the rs3184504 polymorphism (R262W) in SH2B adaptor protein 3 (human: SH2B3, mouse: Sh2b3) and hypertension and other cardiovascular disorders in humans.^10–13 This missense polymorphism alters amino acid 262 from encoding arginine (Arg) in the major allele to tryptophan (Trp) in the minor (risk) allele.¹⁴ The effect of this single amino acid change on SH2B3 structure and function is unclear.

SH2B3, also known as LNK, is an intracellular adaptor protein and negative regulator of signaling downstream of a variety of cytokine receptors and receptor tyrosine kinases.¹⁵ The major known role of SH2B3 is in negative regulation of inflammatory cytokine and hematopoietic growth factor signaling.^{14, 15} Consistent with this, in addition to a strong association with hypertension, a variety of genome-wide association studies have linked rs3184504 with other cardiovascular conditions including myocardial infarction, myeloproliferative disorders, and autoimmune diseases including systemic lupus erythematosus, rheumatoid arthritis, and inflammatory bowel disease.^{12, 15} Despite these associations, whether the rs3184504 polymorphism in SH2B3 plays a causal role in the development of hypertension or autoimmune disease is unknown.

We previously showed that complete loss of Sh2b3 in mice results in increased BP in response to Ang II infusion, along with increased T cell interferon gamma (IFNγ) production, renal dysfunction, and impaired vasorelaxation.¹⁶ In this study, we sought to determine the effect of the rs3184504 polymorphism on hypertension, inflammation, and renal/vascular end-organ damage using CRISPR-Cas9 technology to create mice homozygous for the major (Arg/Arg) and minor/risk (Trp/Trp) alleles. In addition, we performed in vitro and ex vivo studies to determine the role of Trp encoding SH2B3 on cytokine signaling. Our results support a causal role for the rs3184504 SNP in the pathophysiology of hypertension, in part through increased T cell IFNγ production.

METHODS

Detailed methods are available in the Supplemental Materials. Please also see the Supplemental Materials for the Major Resources Table with details of reagents.

Data availability

The data, methods, and materials that support the findings of this study are available from the corresponding author upon reasonable request. Sequencing data has been deposited in GEO under accession number GSE205031.

Animals

The Institutional Animal Care and Use Committee at Vanderbilt University Medical Center approved all animal protocols. C57BL/6J mice (Mus musculus; strain 664) were purchased from Jackson Laboratory (USA). All protocols were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the Institutional Animal Care and Use Committee at Vanderbilt University Medical Center. Mice were housed in temperature and humidity controlled rooms with 12 hour light/dark cycles and ad libitum access to water and PicoLab 5LOD chow (LabDiets; cat. 0067138). A pilot study using female Trp/Trp compared to Arg/Arg mice showed effects similar to that obtained in male mice for circulating total leukocyte counts, spleen size, and renal interferon gamma expression but protection from renal injury following 4 weeks of Ang II infusion (Figure S1). Thus, we focused on male mice for this manuscript.

Generation of Arg/Arg and Trp/Trp mice

Human and mouse Sh2b3 gene sequences are 78% identical, and human codon 262, the site of the rs3184504 polymorphism, corresponds to the orthologous mouse codon 234. Murine Sh2b3 encodes a Proline at codon 234 rather than Arginine or Tryptophan as in humans (Figure 1B). Hence, CRISPR-Cas9 was used to engineer codons encoding Arginine (Arg) or Tryptophan (Trp) at position 234 of murine Sh2b3 to correspond to the human rs3184504 polymorphism major and minor alleles, respectively. Protospacer (targeting) element for the guide RNA was TCTCCAGGCGTGTACATGGT. The Arg encoding donor DNA oligonucleotide was 5’-TCCTAGAGCTCAAAGCCCAAGCTCCAAGAGGCCTGTTCCAGCATCCGGGAGGTCCGAAGATGCACACGCCTGGAGATGCCTGACAACCTCTACACCTTTGTGTTGAAGGTGAGTGCCAAG and the Trp encoding donor DNA oligonucleotide sequence was 5’-CAGCCTTGAATGACCAACCCTCTTGCCCTCTCCTAGAGCTCAAAGCCCAAGCTCCAAGAGGCCTGTTCCAGCATCCGGGAGGTCCGATGGTGCACACGCCTGGAGATGCCTGACAACCTCTACACCTTTGTGTTGAAGGTGAGTGCCAAGGCTGTGGCTCCTTGGGAGCTCTGTTCTCTC. C57Bl/6J zygotes were injected with Cas9, guide RNA, and the Arg donor oligonucleotide to generate the Arg-encoding variant at codon 234 corresponding to the human major allele. Additional zygotes were injected with Cas9, guide RNA, and the Trp mutant donor oligonucleotide to generate the Trp-encoding variant corresponding to the human minor (risk) allele. Both targeting oligonucleotides also generated a silent mutation in codon 235 of murine Sh2b3 to introduce a restriction digest sequence for genotyping (Figure 1C). Founder mice were outbred to C57Bl/6J mice. F1 mice were then interbred to generate Arg/Arg and Trp/Trp mice homozygous for the major and minor alleles, respectively. These mice were demonstrated to possess the desired mutations by Sanger sequencing (Genewiz; South Plainfield, NJ). COSMID was used to identify potential off target mutation sites and PCR primers for sequence verification.¹⁷ Sequencing was performed by Genewiz on each of the top 5 most likely potential off target sites predicted by COSMID (Table S1). Sequences from Arg/Arg and Trp/Trp mice from each PCR amplified region matched expected genomic sequences using NCBI BLAST without evidence of off target mutations. Mice were subsequently interbred as Arg/Trp heterozygotes or Arg/Arg and Trp/Trp homozygous pairs to generate mice used in these experiments.

Figure 1. Creation of mice with Arg- or Trp-encoding Sh2b3 to mimic human major and minor alleles, respectively.

(A) Scores of Sorting Intolerant from Tolerant (SIFT) algorithm predictions of the effect of codon 262 amino acid substitutions on human SH2B3 function. Substitutions in red (score <0.05) are predicted to affect protein function. Letters represent abbreviations for individual amino acids. (B) Representative alignment of amino acids 253–270 of human SH2B3 and amino acids 225–243 of mouse Sh2b3 protein sequences demonstrating presence of Proline (P) at position 234 of mouse Sh2b3 and Arginine (R) at orthologous position 262 of human SH2B3 (boxed). (C) Representation of mouse Sh2b3 DNA and protein sequence in wild type (top) and engineered mouse Sh2b3 gene sequences (bottom) along with restriction enzyme digest sites. DNA sequence changes are in blue and amino acid changes in red. (D) Scores of SIFT algorithm predictions of the effect of codon 234 amino acid substitutions on mouse SH2B3 function. Substitutions in red (score <0.05) are predicted to affect protein function. Letters represent abbreviations for individual amino acids.

Mouse genotyping was performed using PCR and restriction digests. PCR was performed with a 60 degree annealing temperature for 40 cycles with the following primers: LNK-for 5’-AGCCATGACATATAGGGTTC and LNK-rev 5’-TCTGCATCTGAGGAGAGAA. Resulting PCR products were subject to restriction digest with ApaLI, HpyCH4V, and/or BsrGI (New England Biolabs). After gel electrophoresis, digested amplicon size was used to determine genotype (Figure S2).

Flow cytometry of aorta and kidneys

After euthanasia, mice were perfused via cardiac puncture with normal saline at approximately 120 mm Hg. Thoracic aortas including perivascular fat but no lymph nodes were minced with scissors and digested in collagenase A (1 mg/mL), collagenase B (1 mg/mL), and DNAseI (100 μg/mL) for 30 minutes at 37 degrees C. Kidneys were homogenized in C tubes on a gentleMACS dissociator (Miltenyi), digested at 37 degrees in collagenase D (2 mg/mL) and DNAse I (100 μg/mL) for 20 minutes, and immune cells were isolated using Percoll gradient centrifugation. Single cell suspensions were stained with Live/Dead Fixable Violet stain (Life Technologies) or Fixable Zombie NIR viability stain (Biolegend) and antibodies to extracellular antigens (Table S2). Where indicated, intracellular staining was performed after fixation and permeabilization using the Thermo Fisher Fixation and Permeabilization kit according to the manufacturer’s instructions with the following antibodies: anti-Ki67-BV711 (clone B56; BD Biosciences; cat. 563755), anti-IFNγ-PE (clone XMG1.2; BD Biosciences; cat. 562020), and anti-IL-12-APC p40/p70 (C15.6; BD Biosciences; cat. 554480). For some kidneys, after digestion and immune cell isolation by Percoll gradient centrifugation, 1×10⁶ cells were cultured in a 24-well plate in RPMI 1640 media with 10% FBS, 1% pen-strep, and 50 μM beta-mercaptoethanol for 4 hours with Cell Stimulation Cocktail containing phorbol 12-myristate 13-acetate and ionomycin (eBioscience) with Golgiplug containing Brefeldin A (BD Bioscience) according to the manufacturer’s instructions. After 4 hours, cells were harvested and single cell suspensions were stained with Live/Dead Fixable Violet stain (Life Technologies) and the following surface antibodies: anti-CD45-BV750 (clone 30-F11; Biolegend; cat. 103157), anti-CD3-PerCPCy5.5 (clone 17A2; Biolegend; cat. 100217), anti-CD8-AF488 (clone 53–6.7; Biolegend; cat. 100726), and CD4-PacOrange (clone RM4–5; Invitrogen; cat. 79–0042-82). Following cell fixation and permeabilization using the Thermo Fisher Fixation and Permeabilization kit according to the manufacturer’s instructions, the following antibodies were added: anti-IL-17A-PE-Cy7 (clone TC11–18H10.1; Biolegend; cat. 506921) and anti-IFNγ-BV421 (clone XMG1.2; Biolegend; cat. 505829). All antibodies were used at 1:100 dilution. Flow cytometry was performed on a BD FACSCanto II or Cytek Aurora cytometer and analysis was performed using FlowJo (TreeStar). FMO controls were used to set gates, and 123count eBeads counting beads (eBioscience) were added to each sample for determination of absolute cell numbers.

PrediXcan analysis of genetically predicted gene expression and hypertension diagnoses

We applied a model of 137 SNPs, previously built from the Genotype-Tissue Expression (GTEx) database version 7 using the PrediXcan methodology,^{18, 19} to impute SH2B3 expression levels in EBV-transformed lymphocytes (r²=0.12; Table S3). GTEx version 5 was used to define a model of 21 SNPs that predict SH2B3 expression levels in whole blood (r²=0.041; Table S4). These SNP models were used to impute gene expression and were tested for association via logistic regression against five disease phenotypes encoded as phecodes (https://phewascatalog.org/phecodes) in the Vanderbilt University Medical Center BioVU electronic health record-linked biobank.²⁰ Cases were defined as phecodes for hypertension or hypertension-related diagnoses: 401 “Hypertension,” 401.1 “Essential hypertension,” 401.2 “Hypertensive heart and/or renal disease,” 401.21 “Hypertensive heart disease,” 401.22 “Hypertensive chronic kidney disease,” and 401.3 “Other hypertensive complications.” Controls for each hypertension-related phenotype were defined as individuals without phecodes 401–405.99. As a negative control, cases of gastroesophageal reflux disease (GERD) (phecode 530.11) were compared to controls without GERD-related phecodes 530–530.99 or 532–532.99. Analysis was limited to individuals of European ancestry. Logistic regression was adjusted for covariates of gender, age, genotyping array/batch, and the first 10 principal components that capture genetic ancestry. The SH2B3 expression model in lymphocytes was tested in 23,294 individuals while the whole blood model, built on an earlier version of GTEx, was tested in 9,142 individuals.

RESULTS

Trp-encoding Sh2b3 increases systolic BP in response to Ang II infusion

We first tested whether conversion of amino acid position 262 of human SH2B3 from Arginine (Arg; R) encoded by the major allele of rs3184504 to Tryptophan (Trp; W) encoded by the minor allele is expected to affect protein function. Using the SIFT algorithm,²¹ substituting Trp for Arg at this position is predicted to affect protein function (Figure 1A). To determine the effect of the rs3184504 missense polymorphism on hypertension development, we used CRISPR-Cas9 to genetically engineer the orthologous codon 234 of mouse Sh2b3 to encode Arg or Trp corresponding to the human major and minor/risk alleles, respectively (Figure 1B and 1C). Since the wild type murine Sh2b3 sequence encodes Proline (Pro; P) at codon 234, we first determined whether conversion from Pro to Arg would be expected to alter protein function. The SIFT prediction algorithm indicated that converting this position from Pro to Arg (mimicking the human major allele) would have a neutral effect on protein function (Figure 1D).²⁸ Consistent with this, comparing mice homozygous for the major allele (Arg/Arg) to C57BL/6J wild type mice, there was no difference in systolic or diastolic BP via radiotelemetric monitoring either before or during 4 weeks of angiotensin II (Ang II) infusion to induce hypertension (Figure S3). To determine the effect of the minor allele of the polymorphism on BP, we compared mice homozygous for the minor allele (Trp/Trp) to mice homozygous for the major allele (Arg/Arg) before and after Ang II infusion. There was no significant difference in systolic BP at baseline (p=0.36) or during the first two weeks of Ang II infusion (p=0.54). However, during the third and fourth week of Ang II infusion, systolic BP was significantly elevated by approximately 10 mmHg in Trp/Trp mice compared to Arg/Arg controls (Figure 2A). There was no difference in diastolic BP, mean arterial pressure, or heart rate at baseline or after Ang II infusion in Trp/Trp mice compared to Arg/Arg mice (Figure S4).

Figure 2. Trp/Trp mice exhibit increased systolic blood pressure (BP) and renal damage with Ang II infusion compared to Arg/Arg controls.

(A) Radiotelemetric systolic BP of Arg/Arg and Trp/Trp mice at baseline and during 4 weeks of Ang II infusion (140 ng/kg/min). Inset to right is magnification of systolic blood pressure (SBP) readings at 3 and 4 weeks of Ang II (n=6 Arg/Arg, 8 Trp/Trp). Two way ANOVA with repeated measures. (B) Urine albumin to creatinine ratio after 4 weeks of Ang II (n=7 Arg/Arg, 8 Trp/Trp). Mann-Whitney test. (C) Urine NGAL to creatinine ratio after 4 weeks of Ang II infusion (n=7 Arg/Arg, 8 Trp/Trp). Mann-Whitney test. (D) Masson’s trichrome staining of kidneys reveals increased perivascular fibrosis (blue) evidenced by representative images (left) and quantification (right) (n=9 Arg/Arg, 8 Trp/Trp). Unpaired t test. Scale bar=50 μm. NGAL=neutrophil gelatinase-associated lipocalin. Data are expressed as mean±SEM of biological replicates.

Trp-encoding Sh2b3 promotes renal injury

The kidney plays an important role in hypertension development, and elevated BP also contributes to renal injury.^29–31 We thus evaluated renal injury in Arg/Arg and Trp/Trp mice following 4 weeks of Ang II infusion. Interestingly, Trp/Trp mice exhibited increases in both urine albumin to creatinine ratio (marker of glomerular injury) and urine neutrophil gelatinase-associated lipocalin (NGAL) to creatinine ratio (marker of tubular injury) compared to controls (Figure 2B and 2C), suggesting a greater degree of renal injury in Trp/Trp mice. In addition, histological analysis of kidney sections revealed increases in perivascular fibrosis in Trp/Trp compared to Arg/Arg mice after 4 weeks of Ang II infusion (Figure 2D). Consistent with this, flow cytometric analysis of kidneys of Trp/Trp mice after Ang II infusion demonstrated increases in CD45⁺ leukocytes and F4/80⁺ cells compared to Arg/Arg mice (Figure 3). Further discrimination of F4/80⁺ cells revealed increases in both dendritic cells and macrophages in Trp/Trp kidneys (Figure S5A–C). A modest but significant increase in Ki67 positivity was observed in dendritic cells without increases in Ki67 positivity in macrophages (Figure S5D–E). CD4⁺ and CD8⁺ T cells did not exhibit significant differences in cell number or Ki67 positivity in kidneys of Trp/Trp compared to Arg/Arg mice (Figure 3 and Figure S6). In addition, we did not observe differences in numbers of natural killer (NK), NKT, or γδ T cells (Figure S7).

Figure 3. Increased F4/80⁺ macrophages in kidneys of Trp/Trp compared to Arg/Arg mice after 4 weeks of Ang II infusion.

(A) Flow cytometry representative gating for CD45⁺ immune cells, CD3⁺ T cells (with CD4⁺ helper and CD8⁺ cytotoxic T cell subsets), and F4/80⁺ macrophages in kidneys of Arg/Arg and Trp/Trp mice. Quantification of (B) CD45⁺ immune cells, (C) CD3⁺ T cells, (D) CD4⁺ helper T cells, (E) CD8⁺ cytotoxic T cells, and (F) F4/80⁺ macrophages in kidneys of Arg/Arg and Trp//Trp mice after 4 weeks of Ang II infusion (n=9 Arg/Arg, 10 Trp/Trp). Unpaired t test. Data are expressed as mean±SEM of biological replicates.

To determine whether renal injury preceded elevations in BP, mice were analyzed at baseline prior to Ang II infusion. At baseline, although urine albumin to creatinine ratio was slightly higher in the Trp/Trp mice (Figure S8A), urine NGAL to creatinine ratio was not different (Figure S8B). There was also no difference in renal perivascular fibrosis at baseline (Figure S8C). Together, these results demonstrate that homozygosity for the Trp encoding SH2B3 minor allele causes mild glomerular injury at baseline compared to controls, but a greater degree of renal injury and coincident myeloid cell infiltration after Ang II infusion.

Trp-encoding Sh2b3 does not significantly impact vascular function

In addition to renal injury, dysfunction of large arteries and resistance vessels can contribute to and be exacerbated by elevations in BP.^{32, 33} Hence, we tested endothelium-dependent and -independent vasorelaxation of the aorta and mesenteric resistance vessels to graded increases in acetylcholine and sodium nitroprusside, respectively. Vascular relaxation of thoracic aortas and mesenteric arteries to either stimulus did not differ between Arg/Arg and Trp/Trp mice after 4 weeks of Ang II infusion (Figure S9A–D). In addition, thoracic aortas of Arg/Arg and Trp/Trp mice were similar in the degree of fibrosis determined by Masson’s Trichrome staining (Figure S9E). Finally, inflammatory cell infiltration of the thoracic aorta measured by flow cytometry was not different between Trp/Trp and Arg/Arg mice after Ang II infusion (Figure S9F). These results demonstrate that vascular function is not significantly different between Arg/Arg and Trp/Trp mice and is unlikely contributing to the observed differences in Ang II-induced BP elevations.

Increased circulating leukocytes and spleen size in Trp/Trp mice

Given the important role of immune cells in hypertension,^{7, 34} and that the rs3184504 minor allele is strongly associated with increased circulating leukocyte counts in humans,³⁵ we examined circulating blood cell counts in Arg/Arg and Trp/Trp mice after 4 weeks of Ang II infusion. Increases in total leukocytes (Figure 4A), monocytes (Figure 4B), lymphocytes (Figure 4C), and neutrophils (Figure 4D) were observed in Trp/Trp mice compared to controls and splenomegaly was greater (Figure 4E). In contrast, hematocrit and circulating platelet counts did not differ between Ang II infused Trp/Trp and Arg/Arg mice (Figure S10). At baseline prior to Ang II infusion, leukocyte counts and spleen size did not differ, though platelet counts were marginally increased in Trp/Trp mice (Figure S11). These results demonstrate increased circulating numbers of multiple leukocyte lineages with Ang II-induced hypertension in mice homozygous for the Sh2b3 minor allele. Importantly, although both myeloid and T cells are increased in the circulation of Trp/Trp mice after Ang II infusion, only myeloid cells are increased in the kidneys, suggesting that increased renal immune cell content is not simply a function of increased circulating immune cell numbers.

Figure 4. Trp/Trp mice have increased circulating leukocytes and splenomegaly after Ang II infusion compared to Arg/Arg controls.

Circulating (A) white blood cell, (B) monocyte, (C) lymphocyte, and (D) neutrophil counts after 4 weeks of Ang II infusion (n=7 Arg/Arg,8 Trp/Trp). Mann Whitney test. (E) Representative images of spleen size in Trp/Trp and Arg/Arg mice (left) with quantification of spleen weight normalized to body weight (right) (n=9 Arg/Arg,12 Trp/Trp). Mann Whitney test. Scale bar=0.5 cm. Data expressed as mean±SEM of biological replicates.

RNA sequencing reveals increased T cell and IFNγ pathway activation in kidneys from Trp/Trp mice after Ang II infusion

To determine mechanisms of increased inflammation and renal damage in Trp/Trp mice, we performed bulk RNA sequencing on kidneys from Arg/Arg and Trp/Trp mice after 4 weeks of Ang II infusion. Results revealed multiple inflammation-related genes increased in Trp/Trp kidneys relative to controls including Cx3cr1, Ucp1, Il2rb, and Cd52, with the metabolism related enzyme Bcat1 decreased in Trp/Trp kidneys (Figure S12A). As an orthologous approach to determine biological pathways differing between Trp/Trp and Arg/Arg kidneys, gene set enrichment analysis was performed. Results revealed an increase in multiple gene ontology terms related to T cell activation in Trp/Trp kidneys including “adaptive immune response,” “T cell activation,” and “leukocyte migration” (Figure S12B and Table S5). Interestingly, the term “interferon gamma production” was also significantly over-represented in Trp/Trp kidneys. Together, these unbiased studies demonstrate increased inflammation-related pathways in Trp/Trp kidneys compared to controls, particularly increased T cell activation and IFNγ production.

Homozygosity for Trp-encoding Sh2b3 increases T cell IFNγ production

Given that RNA sequencing revealed gene expression changes consistent with enhanced T cell activation and IFNγ production in kidneys of Trp/Trp mice, and that renal myeloid cells were increased in Trp/Trp mice, we tested for differences in inflammatory cytokines that promote BP elevations, renal dysfunction, and enhanced renal myeloid cell content such as IFNγ^{36, 37} and IL-17A.^{22, 38} Isolated splenic T cells from 4 week Ang II infused mice were cultured and activated with anti-CD3 and anti-CD28 antibodies (Figure 5A). After 3 days of culture, significantly increased IFNγ production was observed from CD4⁺ and CD8⁺ T cells from Trp/Trp mice compared to Arg/Arg controls (Figure 5B and 5C). Production of IL-17A was modestly increased from CD4⁺ T cells but not CD8⁺ T cells from Trp/Trp mice (Figure 5D and 5E). These results suggest a major effect of Trp-encoding Sh2b3 is on increased IFNγ production. Consistent with this, after 3 days of activation with anti-CD3 and anti-CD28 antibodies, CD8⁺ T cells exhibit a selective increase in Ifnγ mRNA levels, with no difference in Il17a or Il4 and a trend for increased Tnfα mRNA levels (Figure S13). These results suggest that observed increases in IFNγ release into the culture media by Trp/Trp CD8⁺ T cells is due at least in part to increased transcription.

Figure 5. Splenic T cells from Ang II-infused Trp/Trp mice produce more IFNγ and IL-17A compared to Arg/Arg mice.

(A) Schematic of CD4⁺ and CD8⁺ T cell isolation from spleens of Arg/Arg or Trp/Trp mice after 4 weeks of Ang II infusion. Equal cell numbers were cultured for 3 days with anti-CD3 and anti-CD28 antibodies for T cell activation followed by ELISA for inflammatory cytokines. Parts of the figure were from Servier Medical Art licensed under a Creative Commons Attribution 3.0 Unported License. (B) IFNγ from CD4⁺ T cells (n=6 Arg/Arg, 6 Trp/Trp) and (C) CD8⁺ T cells (n=8 Arg/Arg, 7 Trp/Trp) along with (D) IL-17A from CD4⁺ T cells (n=7 Arg/Arg, 6 Trp/Trp) and CD8⁺ T cells (n=7 Arg/Arg, 6 Trp/Trp). Mann-Whitney test. Data are expressed as mean±SEM of biological replicates.

To determine if similar increases in T cell IFNγ production are present in kidneys, immune cells were isolated from kidneys of Ang II-infused mice and subjected to 4 hours of phorbol myristate acetate/ionomycin stimulation followed by flow cytometry. No difference was observed in total numbers of renal CD8⁺IFNγ⁺ cells (Figure 6A), however Trp/Trp CD8⁺ T cells exhibited increased median fluorescence intensity (MFI) for IFNγ compared to controls (Figure 6B), indicating greater IFNγ production on a per cell basis in Trp/Trp T cells. Furthermore, Ifnγ mRNA levels were increased in kidneys of Trp/Trp mice after 4 weeks of Ang II infusion (Figure 6C). In contrast, there was no difference in CD8⁺IL-17A⁺ T cells or IL-17A MFI in these cells in Trp/Trp kidneys, nor in renal CD4⁺IFNγ⁺ T cells or IFNγ MFI in these cells. There was also no difference in CD4⁺IL-17A⁺ cells or IL-17A MFI in these cells (Figure S14). In addition, there was a trend for increased number of IFNγ⁺ macrophages in the kidney but no difference in the frequency of IFNγ⁺ macrophages or MFI for IFNγ in these cells (Figure S15A,C,E). There was also no difference in IFNγ⁺ number, frequency, or IFNγ MFI in dendritic cells (Figure S15B,D,F). Hence, the major effect of Trp-encoding Sh2b3 to promote IFNγ production on a per cell basis appears to be in T cells, which likely contributes to the observed increases in total renal macrophages and dendritic cells given prior reports that IFNγ can promote recruitment of myeloid cells to the kidney,³⁶

Figure 6. Increased IFNγ from renal CD8⁺ T cells of Ang II-infused Trp/Trp mice.

(A) CD8⁺IFNγ⁺ T cell numbers and (B) median fluorescence intensity (MFI) of IFNγ in CD8⁺ T cells isolated from kidneys of Arg/Arg or Trp/Trp mice after 4 weeks of Ang II infusion (n=8). (C) RT-PCR for Ifnγ from kidneys of Arg/Arg and Trp/Trp mice after 4 weeks of Ang II infusion (n=11 Arg/Arg, 8 Trp/Trp). Unpaired t test. Data are expressed as mean±SEM of biological replicates.

Interleukin-12-induced IFNγ production and Stat4 phosphorylation are enhanced in Trp/Trp T cells

To better understand the mechanism by which Trp-encoding SH2B3 results in greater T cell IFNγ production, we first focused on differences in renal myeloid cells in Trp/Trp mice, given that myeloid-derived inflammatory cytokines like interleukin-12 (IL-12) are key stimuli for T cell IFNγ production.³⁹ Bulk RNA sequencing of renal F4/80⁺ myeloid cells from Ang II-infused Trp/Trp and Arg/Arg mice revealed increases in inflammation-related gene expression in Trp/Trp renal myeloid cells. In particular, we observed enrichment of genes related to ontology terms such as “IFNγ production” and “Cytokine secretion” in renal F4/80⁺ cells from Trp/Trp mice (Figure S16). Although IFNγ itself was not enriched in the “IFNγ production” ontology term, there were multiple genes encoding receptors involved in enhanced production of IL-12 by myeloid cells such as TLR9, TLR4, IL18R1, and IL1R1 (Table S6).^40–43 We did not however observe increased production of IFNγ or IL-12 by macrophages or dendritic cells in the kidneys of Trp/Trp mice after Ang II infusion (Figures S15 and S17). These results suggest that myeloid cells may indirectly enhance T cell IFNγ production in Trp/Trp kidneys but not by producing more IFNγ or IL-12 on a per cell basis. We then tested whether T cells may exhibit enhanced responsiveness to IL-12. Indeed, IL-12 treatment of splenic CD8⁺ T cells from Trp/Trp mice revealed increased IFNγ production compared to Arg/Arg T cells (Figure 7A and Figure S18). These results suggest that Trp-encoding SH2B3 significantly enhances effects of IL-12 to promote IFNγ production compared to Arg-encoding SH2B3. Stat4 is a transcription factor activated by phosphorylation downstream of IL-12 receptor agonism to induce IFNγ transcription in T cells.^44–46 We found that IL-12-induced Stat4 phosphorylation is greater in CD8⁺ T cells from Trp/Trp compared to Arg/Arg mice (Figure 7B). These results demonstrate that Trp-encoding SH2B3 promotes greater IL-12 signaling to enhance IFNγ production in T cells. Given that SH2B3 is a negative regulator of cytokine signaling,^{15, 47} these findings suggest that Trp-encoding SH2B3 is less repressive of IL-12 signaling than Arg-encoding SH2B3.

Figure 7. Increased IFNγ production and Stat4 phosphorylation by IL-12 in Trp/Trp compared to Arg/Arg splenic CD8⁺ T cells.

(A) Splenic CD8⁺ T cells were cultured with anti-CD3 and CD28 antibodies along with vehicle or IL-12 (10 ng/mL) for three days with detection of IFNγ by ELISA in the culture supernatants. Data shown are technical replicates (n=3 Arg/Arg IL-12, n=4 in other groups) of cells from one mouse per genotype which are representative of three experiments from separate biological replicates. Results of the other 2 biological replicate experiments are shown in Figure S18. Statistics based on align-and-rank nonparametric factorial ANOVA with the Holm’s post hoc test. (B) Representative images of percent phospho-Stat4⁺ cells (left) from flow cytometric analysis of CD8⁺ T cells stimulated for three days with anti-CD3 and CD28 antibodies followed by treatment for 15 minutes with vehicle or IL-12 (10 ng/mL). Quantitation of the percentage of phospho-Stat4 positive cells from live singlets (right). Data shown are three experiments, each with CD8⁺ T cells from a separate mouse of each genotype, with four technical replicates of cells in each experiment. Data are expressed as mean±SEM. Statistics based on align-and-rank nonparametric factorial ANOVA with Holm’s post hoc test incorporating experiment and mouse cluster. pStat4=phosphorylated Stat4. (C) Working model of the effect of the tryptophan (Trp) encoding minor allele of the rs3184504 polymorphism in SH2B3 on hypertension and end-organ damage. Trp-encoding SH2B3 results in less inhibition of Stat4 phosphorylation with resultant enhanced T cell IFNγ production to promote increased blood pressure and renal injury in mice. IL-12=interleukin-12; IL-12R=interleukin-12 receptor. Created with Biorender.com.

Genetically predicted SH2B3 expression in lymphocytes is inversely associated with hypertension and related end-organ damage in humans

Our findings suggest that the rs3184504 minor allele encoding Trp reduces SH2B3 function in lymphocytes leading to increased BP and end-organ damage in mice. In order to better determine the relevance of these findings to human hypertension, we utilized the PrediXcan methodology¹⁸ to test whether genetically-predicted increased expression of SH2B3 in lymphocytes is inversely associated with human hypertension and end-organ damage. PrediXcan utilizes expression quantitative trait loci (eQTLs) to build SNP-based predictive models capturing genetically-determined gene expression in a tissue-specific manner. These models can then be used to impute gene expression from genotypes, and the imputed gene expression can be tested for association with phenotypes of interest using electronic health records linked to genetic biobanks like Vanderbilt’s BioVU.^{19, 20} We used a model of 137 SNPs (excluding rs3184504) built from the GTEx database to impute SH2B3 expression levels in lymphocytes and tested for association with phenotypes related to hypertension and end-organ damage in 23,294 de-identified patients of European ancestry in BioVU (r²=0.116). Their demographics are listed in Table S7. Of the six hypertension-related phenotypes tested, all six including “Essential Hypertension,” “Hypertension,” “Hypertensive heart and/or renal disease,” “Hypertensive chronic kidney disease,” “Other hypertensive complications,” and “Hypertensive heart disease” exhibited significant associations with genetically predicted SH2B3 levels (Table 1). The odds ratio less than 1 for these associations demonstrates that increased SH2B3 levels are associated with decreased risk for these hypertension-related diagnoses. Furthermore, we show that the “Essential Hypertension” phecode remains significant when adjusted for the presence of type 2 diabetes mellitus, hyperlipidemia, and obesity (Table S8). As a negative control, we tested whether this genetic model of SH2B3 expression in lymphocytes is associated with gastroesophageal reflux, another common and chronic condition but unrelated to hypertension. There was no significant association observed (cases 4408, controls 13499; p=0.22; odds ratio 0.98). As an additional control to test the specificity of the association in lymphocytes, we tested whether genetically predicted levels of SH2B3 in whole blood also associates with hypertension diagnoses. Interestingly, only one hypertension-related phenotype (“Hypertensive heart and/or renal disease”) was significantly associated with genetically predicted SH2B3 levels in whole blood (Table S9). This suggests a stronger association of SH2B3 levels in lymphocytes compared to other immune cells with hypertension diagnoses. In summary, these results reveal that increased expression levels of SH2B3 in lymphocytes is associated with decreased risk of hypertension and end-organ damage, consistent with findings in our mouse model of decreased function of Trp-encoding Sh2b3 leading to increased hypertension and renal damage.

Table 1:

Association between genetically predicted SH2B3 expression levels in lymphocytes and hypertension-related diagnoses

Phecode	Trait	Cases	Controls	OR	p-value
401.1	Essential hypertension	12930	7812	0.94	0.00003
401	Hypertension	13238	7812	0.94	0.00004
401.2	Hypertensive heart and/or renal disease	3283	7812	0.91	0.00008
401.22	Hypertensive chronic kidney disease	1934	7812	0.91	0.00080
401.3	Other hypertensive complications	917	7812	0.88	0.00084
401.21	Hypertensive heart disease	1520	7812	0.92	0.00514

SNP model of 137 SNPs predicting increased SH2B3 expression in EBV-transformed lymphocytes is detailed in Table S3. OR=odds ratio. Bonferroni-corrected p-value <0.0083 considered significant.

DISCUSSION

Our results of multiple functional studies provide novel evidence that the minor allele of rs3184504, a polymorphism in SH2B3 associated with hypertension in humans, plays a causal role in promoting the development of hypertension and renal damage in mice. The effect of this polymorphism was directly tested using CRISPR-Cas9-mediated creation of Arg/Arg mice homozygous for the major allele and Trp/Trp mice homozygous for the minor allele. Trp/Trp mice exhibited greater elevations in systolic BP in response to Ang II compared to Arg/Arg mice. While there was no difference in resistance artery vascular reactivity or aortic fibrosis, renal damage was greater in Trp/Trp mice as evidenced by increased albuminuria and urinary NGAL, renal perivascular fibrosis, and renal macrophage content. Interestingly, both renal and splenic CD8⁺ T cells from Trp/Trp mice exhibited selectively greater production of the inflammatory cytokine IFNγ compared to Arg/Arg mice. CD8⁺ T cells from Trp/Trp mice also exhibited increased IL-12-induced IFNγ production, along with enhanced IL-12 signaling evidenced by increased Stat4 phosphorylation. Finally, genetically imputed increased levels of SH2B3 were associated with lower incidence of hypertension and related end-organ damage in humans. Taken together, these results demonstrate that the rs3184504 minor allele is a causal variant promoting increases in BP and renal damage and findings suggest that this occurs through enhanced CD8⁺ T cell IL-12 signaling to promote increased IFNγ production (Figure 7C).

Hypertension is a polygenic disease, with an overall heritability of BP of approximately 30–60% that results largely from the contribution of many common SNPs with small individual effect sizes. Genome-wide association studies have identified a multitude of SNPs associated with BP. Determining which of these variants are causal for hypertension, however, has been challenging. Experimental approaches have included in vitro evaluation of isolated cells from individuals with particular SNPs as well as overexpression of polymorphic expression constructs.^47–50 These approaches are complicated, however, by the presence of multiple genetic differences in the isolated cells and/or an inability to test effects of polymorphisms on complex physiological processes such as BP regulation. In vivo approaches include generation of mice with genomic excision of large portions of a gene coding sequence as in classical “knockout” mice,^51–56 however such studies likely do not recapitulate effects of single nucleotide polymorphisms. Indeed, as observed previously for studies on Sh2b3 in mice and rats,^{16, 57} genomic deletions of different portions of the coding sequence can yield contrasting results. Transgenic overexpression studies of polymorphic gene variants have been performed in mice, however exogenous overexpression likely does not recapitulate effects of gene polymorphisms under endogenous regulation or expression levels.^{58, 59} Hence, we have taken the approach of using CRISPR-Cas9 technology to specifically engineer the major and minor alleles of the rs3184504 polymorphism into the endogenous Sh2b3 locus in mice in order to directly test the effect of the polymorphism on hypertension development in vivo. This approach has enabled us to demonstrate that the minor allele of this polymorphism is causal for elevations in BP and end-organ damage and to determine mechanisms for this effect. In addition to testing roles in hypertension, our unique animal model also provides a valuable tool for the study of additional human diseases associated with the rs3184504 polymorphism including type 1 diabetes mellitus,⁶⁰ rheumatoid arthritis,⁶¹ celiac disease,⁶² and myeloproliferative neoplasms.⁶³

Findings from these studies suggest a mechanism of increased BP in Trp/Trp mice via reduced SH2B3 function leading to enhanced IFNγ production in T cells. Our results demonstrate that splenic and renal CD8⁺ T cells from Trp/Trp mice produce more IFNγ than from Arg/Arg mice. This is consistent with prior evidence from our group that mice with genetic deficiency of Sh2b3 resulting from deleting exons 3 through 8 exhibit increased splenic T cell IFNγ production.¹⁶ In addition, humans with heart failure homozygous for the rs3184504 minor allele exhibit increased left ventricular IFNγ levels.⁶⁴ Our results suggest that the mechanism for increased T cell IFNγ production involves at least partial loss of SH2B3 function on signaling downstream of the IL-12 receptor. SH2B3 has been shown to function as a negative regulator of cytokine signaling including downstream of IL-12.⁶⁵ Hence, our findings that T cells from mice homozygous for Trp-encoding Sh2b3 exhibit enhanced IL-12-induced Stat4 phosphorylation and IFNγ production suggest that Trp-encoding Sh2b3 mediates less repression of IL-12 signaling. IFNγ has been demonstrated to play a functional role in mediating both increased BP and renal dysfunction in hypertension. For example, mice with deficiency of IFNγ signaling have reduced BP elevations and improved renal function following hypertension induction.^{16, 36, 37} IFNγ is thought to mediate renal injury in hypertension at least in part through recruitment of macrophages with resultant enhanced cytokine production leading to glomerular and tubular damage and fibrosis.^{66, 67} Indeed, loss of IFNγ signaling in a mouse model of hypertension reduces renal macrophage content,³⁶ and macrophage depletion reduces BP and renal injury in hypertension.⁶⁸ Our findings of increased macrophages and CD8⁺ T cell IFNγ production in the kidneys of Trp/Trp mice after Ang II infusion are consistent with IFNγ promoting renal injury in hypertension through recruitment of macrophages.

Our findings suggest the potential for a precision medicine approach targeting inflammatory cytokines such as IFNγ to lower BP and reduce end-organ damage with particular benefit in individuals harboring the rs3184504 minor allele. Such a genetically based approach to treatment could have a significant public health impact, particularly in individuals of European ancestry where the minor allele frequency is 46–50%. Of note, there is significant variation in the minor allele frequency across ancestry groups as Africans, East Asians, and South Asians have minor allele frequencies for rs3184504 less than 10%.⁶⁹ These ancestral differences may reflect varying selection pressures related to the degree of T cell IFNγ production. This SNP seems to have arisen over 1500 years ago in European and nearby populations as part of a linkage disequilibrium block that may have provided an enhanced immune response and thus protection from infections such as the plague at the expense of increased susceptibility to autoimmune and cardiovascular disease.^{47, 70} Moreover, increased IFNγ production enhances cerebral pathology in malaria infection,⁷¹ and thus may have driven down the allele frequency in populations where malaria is prevalent. This suggests that therapeutic approaches targeting the impact of the rs3184504 minor allele based on genetic ancestry may be particularly beneficial.

In addition to an association with hypertension, the rs3184504 minor allele is associated with increased circulating blood cell counts and multiple autoimmune diseases.^{14, 15} Our findings of increased circulating leukocytes including lymphocytes, monocytes, and neutrophils in Trp/Trp mice provide evidence for a causal role for this polymorphism in enhancing proliferation and/or differentiation of multiple leukocyte lineages. These findings are consistent with a role for SH2B3 in negatively regulating early hematopoiesis as demonstrated in Sh2b3-deficient mice.^{47, 72–74} Interestingly, these findings may also relate to recent observations of a link between cardiovascular disease and two processes of enhanced expansion of circulating leukocytes: clonal hematopoiesis and lymphocytosis.^{75, 76} Clonal hematopoiesis of indeterminant significance (CHIP) is an age-related process of somatic mutation of hematopoiesis-related genes, including SH2B3, leading to selective clonal proliferation of hematopoietic cells.⁷⁷ CHIP is strongly associated with cardiovascular disease^78–80 and hypertension.⁸¹ Siedlinski M et al. recently demonstrated an association between lymphocytosis and hypertension in humans.⁷⁶ Interestingly, in this study the rs3184504 polymorphism in SH2B3 nearly fully mediated the relationship between lymphocytosis and hypertension.⁷⁶ Mechanisms by which CHIP and lymphocytosis promote hypertension remain unclear, however it is interesting to speculate that differences in cytokine production may mediate the increased hypertension risk. Based on our results, increased IFNγ production from partial loss of SH2B3 function via the rs3184504 minor allele, occurring either as a germline or somatic mutation, may represent a common mechanism for increased BP and cardiovascular events associated with lymphocytosis and clonal hematopoiesis. Future studies testing circulating and tissue IFNγ levels and relationship to lymphocyte counts and CHIP in individuals with the rs3184504 minor allele could help address this hypothesis.

Hypertension is a systemic disease with pathogenic contributions from a variety of dysfunctional organ systems including the vasculature and kidney. Emerging evidence also suggests an important contribution of immune cells and inflammation, which likely manifests through promoting dysfunction of the organ systems described above.⁸² Although elevated BPs in Trp/Trp mice were not accompanied by evidence of dysfunction in large or resistance arteries in our study, renal damage was present in Trp/Trp mice. Mild renal damage was evident at baseline in Trp/Trp mice prior to Ang II infusion when BP was similar to controls. After 4 weeks of Ang II infusion, renal damage in Trp/Trp compared to Arg/Arg mice was more severe, as evidenced by increased perivascular fibrosis and urinary NGAL levels. This increase in NGAL levels occurred in the presence of mild (10 mmHg) increases in BP in these mice, suggesting that the rs3184504 polymorphism may contribute to renal damage independently of BP elevation, perhaps through enhanced renal inflammation and CD8⁺ T cell IFNγ production. In keeping with this, prior human genetic studies demonstrate that amongst multiple hypertension-associated SNPs, the SH2B3 rs3184504 polymorphism is uniquely associated with reduced kidney function.⁸³ In addition to direct effects of the Sh2b3 polymorphism on kidney function, the observed renal damage in Trp/Trp mice may have also contributed to the elevated systolic blood pressure in these mice after Ang II.⁸⁴

Mice homozygous for the minor Trp-encoding allele exhibit a similar phenotype to mice with genetic deficiency of the Sh2b3 gene, suggesting that Trp-encoding Sh2b3 represents at least a partial loss of Sh2b3 function. Our findings of increased blood pressure, renal damage, and T cell IFNγ production in Trp/Trp mice parallel those seen in mice homozygous for Sh2b3 deficiency.¹⁶ In general however, the magnitude of the increase in blood pressure and renal damage was greater in Sh2b3-deficient mice, suggesting that Trp-encoding Sh2b3 represents a partial loss of function. In addition, Sh2b3-deficient mice demonstrated enhanced aortic inflammation and decreased vascular relaxation with Ang II-induced hypertension induction which were not observed in Trp/Trp mice compared to controls. These findings suggest that Sh2b3, or at least the R262W polymorphism, may have greater effects to limit renal compared to vascular inflammation and dysfunction in hypertension. Prior studies of Sh2b3 deficiency have also demonstrated roles for Sh2b3 to enhance hematopoietic cell proliferation.^{47, 73} We also observe increased circulating leukocyte counts including monocytes, lymphocytes, and neutrophils. However, at least in the kidney, we observed only a modest increase in dendritic cell proliferation and no change in macrophage proliferation, suggesting that the primary effect of Sh2b3 to limit hematopoiesis and proliferation is in the bone marrow and/or circulation rather than in target organs. Alternatively, SH2B3 may limit cell proliferation primarily through the SH2 rather than the PH domain, so there could be domain-dependent effects of alterations in the protein.^85–87 Finally, although others have observed increased IFNγ production by myeloid cells with genetic Sh2b3 deficiency in vitro,⁶⁵ we did not observe increased IFNγ production by myeloid cells in the kidney. Our findings of increased T cell signaling and IFNγ production in response to IL-12 suggests T cell-intrinsic roles of Sh2b3 to enhance IFNγ production, although we cannot rule out additional contributions of Sh2b3 in myeloid cells to the observed increases in blood pressure and renal damage.

Taken together, the results of multiple functional studies presented in this investigation provide evidence that the minor allele of rs3184504 promotes systolic BP elevations and renal damage. Our results are consistent with prior genome-wide association studies demonstrating a strong association between this polymorphism in SH2B3 and human hypertension, as well as network analysis identifying SH2B3 as a key driver of hypertension.^10–12 Our findings suggest the therapeutic potential of enhancing SH2B3 function and/or inhibiting IFNγ in hypertension, particularly in those carrying the rs3184504 minor allele. The “reverse translational” approach used here has provided new insights into the genetic basis of the pathogenesis of hypertension and has important implications for development of precision medicine approaches to treat this disease.

Novelty and Significance.

What Is Known?

• Hypertension is a polygenic condition with high heritability.

• Multiple single nucleotide polymorphisms (SNPs) are associated with hypertension, however determining which SNPs are causal in blood pressure elevations and related end-organ damage has been challenging.

• rs3184504 is a missense SNP in SH2B3 that is strongly associated with hypertension in multiple genome-wide association studies, and systems biology-based network analysis suggests that SH2B3 is a key driver of hypertension.

What new information does this article contribute?

• By creating mice with the rs3184504 SNP encoding arginine or tryptophan we reveal that the minor allele encoding tryptophan is causal for blood pressure (BP) elevations and renal damage.

• The SNP minor allele encoding tryptophan also causes increased renal T cell IFNγ production in vivo.

• IL-12-induced signaling and downstream IFNγ production are enhanced in T cells encoding the minor allele.

Hypertension is a polygenic disease with many associated SNPs, however whether individual SNPs play causal roles has remained unclear. A common missense SNP in SH2B3 (rs3184504) results in substitution of tryptophan (Trp) for arginine (Arg) at amino acid 262 and is strongly associated with hypertension. We used CRISPR-Cas9 to create mice with the minor and major alleles. We found that mice homozygous for the Trp (minor) allele have higher BP and greater renal damage, consistent with the Trp allele being a causal variant for hypertension and end-organ damage. Splenic and renal T cells from mice homozygous for the Trp allele also exhibit selectively increased IFNγ production, and IL-12 signaling was enhanced in splenic CD8⁺ T cells. These findings suggest that the minor allele results in partial loss of SH2B3 function leading to reduced negative regulation of IL-12 signaling and increased IFNγ production. Consistent with this, increased genetically imputed levels of SH2B3 in lymphocytes are associated with decreased incidence of hypertension and related end-organ damage in humans. Taken together, our results demonstrate that the Trp allele of rs3184504 in SH2B3 is a causal variant leading to BP elevations and renal damage, likely through enhanced IFNγ production by T cells.

Non-Standard Abbreviations and Acronyms

Ang II

Angiotensin II

Arg

Arginine

BP

Blood pressure

DC

Dendritic cell

ELISA

Enzyme-linked immunoabsorbent assay

eQTL

Expression quantitative trait loci

FMO

Fluorescence minus one

GERD

Gastroesophageal reflux disease

GTEx

Genotype-Tissue Expression

HLD

Hyperlipidemia

IFNγ

Interferon gamma

IL-17A

Interleukin-17A

IL-12

Interleukin-12

IL-12R

Interleukin-12 receptor

LNK

Lymphocyte adaptor protein

MFI

Median fluorescence intensity

Pro

Proline

NGAL

Neutrophil gelatinase-associated lipocalin

NK

natural killer

PH

Pleckstrin homology

RT-PCR

Reverse transcriptase polymerase chain reaction

SH2B3

SH2B Adaptor Protein 3

SIFT

Sorting Intolerant from Tolerant

SNP

Single nucleotide polymorphism

STAT4

Signal transducer and activator of transcription 4

T2DM

Type 2 diabetes mellitus

TCR

T cell receptor

TNFα

Tumor necrosis factor alpha

Trp

Tryptophan