Basic Research

GSTM1 Deletion Exaggerates Kidney Injury in Experimental Mouse Models and Confers the Protective Effect of Cruciferous Vegetables in Mice and Humans

Joseph C. Gigliotti, Adrienne Tin, Shirin Pourafshar, Sylvia Cechova, Yves T. Wang, Sun-sang J. Sung, Gabor Bodonyi-Kovacs, Janet V. Cross, Guang Yang, Nhu Nguyen, Fang Chan, Casey Rebholz, Bing Yu, Megan L. Grove, Morgan E. Grams, Anna Köttgen, Robert Scharpf, Phillip Ruiz, Eric Boerwinkle, Josef Coresh and Thu H. Le
JASN January 2020, 31 (1) 102-116; DOI: https://doi.org/10.1681/ASN.2019050449

Significance Statement

GSTM1 encodes a member of a superfamily of antioxidant enzymes, and a highly prevalent GSTM1 deletion variant is associated with kidney disease progression in two human study cohorts. In this study, the authors demonstrate that Gstm1 knockout mice exhibit increased oxidative stress, kidney injury, and inflammation in models of CKD and hypertension, and that Gstm1 loss in the parenchyma but not in bone marrow–derived cells drives renal inflammation. Importantly, consumption of broccoli powder or cruciferous vegetables was protective against kidney disease only in Gstm1 knockout mice, and was observed mainly in the human participants in the Atherosclerosis Risk in Communities Study who were homozygous for GSTM1 deletion. These findings suggest that targeting antioxidant therapy specifically in individuals carrying the GSTM1 deletion variant may be effective in delaying kidney disease progression.

Abstract

Background GSTM1 encodes glutathione S-transferase μ-1 (GSTM1), which belongs to a superfamily of phase 2 antioxidant enzymes. The highly prevalent GSTM1 deletion variant is associated with kidney disease progression in human cohorts: the African American Study of Kidney Disease and Hypertension and the Atherosclerosis Risk in Communities (ARIC) Study.

Methods We generated a Gstm1 knockout mouse line to study its role in a CKD model (involving subtotal nephrectomy) and a hypertension model (induced by angiotensin II). We examined the effect of intake of cruciferous vegetables and GSTM1 genotypes on kidney disease in mice as well as in human ARIC study participants. We also examined the importance of superoxide in the mediating pathways and of hematopoietic GSTM1 on renal inflammation.

Results Gstm1 knockout mice displayed increased oxidative stress, kidney injury, and inflammation in both models. The central mechanism for kidney injury is likely mediated by oxidative stress, because treatment with Tempol, an superoxide dismutase mimetic, rescued kidney injury in knockout mice without lowering BP. Bone marrow crosstransplantation revealed that Gstm1 deletion in the parenchyma, and not in bone marrow–derived cells, drives renal inflammation. Furthermore, supplementation with cruciferous broccoli powder rich in the precursor to antioxidant-activating sulforaphane significantly ameliorated kidney injury in Gstm1 knockout, but not wild-type mice. Similarly, among humans (ARIC study participants), high consumption of cruciferous vegetables was associated with fewer kidney failure events compared with low consumption, but this association was observed primarily in participants homozygous for the GSTM1 deletion variant.

Conclusions Our data support a role for the GSTM1 enzyme in the modulation of oxidative stress, inflammation, and protective metabolites in CKD.

CKD and its cardiovascular disease (CVD) and ESRD consequences are a significant public health burden in the United States.1 In the African American Study of Kidney Disease and Hypertension (AASK) trial, we found that those with GSTM1 deletion genotypes (0/1 or 0/0) had, respectively, a 1.7- or 2-fold increased risk for the composite outcome of a decline in GFR, commencement of dialysis or all-cause mortality, compared with those with two active alleles [GSTM1(1/1)].2 A similar effect was observed in the Atherosclerosis Risk in Communities (ARIC) study, in which GSTM1 deletion was associated with kidney failure and heart failure in both black and white participants.3

A significant portion of humans are homozygous for GSTM1 deletion [GSTM1(0/0)]4 and are therefore completely deficient of the glutathione S-transferase μ-1 (GSTM1) enzyme.5,6 The proportion of the GSTM1(0/0) genotype is population dependent and is approximately 50% in those with European ancestry and 25% in black Americans.4 The GSTM1 enzyme belongs to a superfamily of glutathione S-transferases (GSTs) and is involved in the elimination of a wide range of electrophiles, both protective and toxic, in the glutathione pathway.7,8 The noncatalytic domain of GSTM1 has also been shown to suppress the apoptosis signal-regulating kinase (ASK1).9

A growing body of evidence suggests that GSTM1 plays an important role in determining disease susceptibility. The significance of GSTM1 deletion was first recognized in cancer studies demonstrating that patients carrying the GSTM1(0/0) genotype were at increased risk for colon10 and lung cancers.11 In subsequent association studies on CVD, participants carrying GSTM1(0/0) were shown to have significantly increased risks of hypertension (HTN),12,13 coronary artery disease/atherosclerosis,14,15 and stroke.16,17 Although GSTM1 deletion is associated with CKD progression, CVD, and cancer (which could be explained by the lack of GSTM1 enzyme for the elimination of toxic electrophiles), GSTM1(0/0) genotype has also been shown to enhance the protective effect of intake of cruciferous vegetables, such as broccoli, against lung cancer in multiple populations of European and East Asian ancestries.18–22

Over the past decade, the potential benefit of isothiocyanates (ITCs) has garnered significant attention. These compounds are biologically active metabolites of glucosinolates that are abundant in cruciferous vegetables such as broccoli. In particular, the ITC sulforaphane, although not antioxidant itself, has been shown to be highly potent in inducing the phase 2 detoxification enzymes GSTs (including GSTM1), UD-glucuronosyltransferases, sulfotransferases, N-acetyltransferases, and S- and O-methyltransferases via activation of the Nrf2 pathway.23,24 Sulforaphane-rich broccoli contains high levels of glucoraphanin, which is metabolized into sulforaphane. The stronger protective effect of cruciferous vegetable intake in those with GSTM1(0/0) genotype may be due to greater total bioavailability of antioxidant-promoting compounds like sulforaphane in these individuals.25 Very few studies have assessed the effect of sulforaphane in kidney disease. Of these, two studies showed that sulforaphane induced expression of superoxide dismutase and catalase in the liver of spontaneously hypertensive rats,26 lowered their BP, and decreased renal tissue inflammation.27

Multiple studies have demonstrated an association between the loss of Gstm1 and deleterious effects in animal models. The murine GSTM1 shares approximately 80% amino acid identity with the human GSTM1 and >90% with the rat GSTM1 (basic alignment tool [NCBI BLAST], https://blast.ncbi.nlm.nih.gov/Blast.cgi). In mice, Gstm1 is a potential candidate modifier of vascular injury. A mouse strain that is more susceptible to renal vascular remodeling with pathologic features resembling human arteriolar nephrosclerosis expressed lower levels of Gstm1 than the resistant strain, and had faster proliferation and migration of vascular smooth muscle cells (VSMCs).28,29 In the rat, Gstm1 is a positional candidate gene in a quantitative trait locus for HTN.30 Gstm1 mRNA and protein levels in the kidney were found to be reduced in the stroke-prone spontaneously hypertensive rat (SHRSP), compared with the normotensive congenic and Wistar Kyoto rats,31 and were inversely correlated with kidney tissue levels of reactive oxygen species.

To further elucidate the pathways by which GSTM1 deletion affects kidney disease pathophysiology and to identify potential therapeutic targets, we generated a Gstm1 knockout (KO) mouse line to mimic the common 20 kb deletion variant in humans.6 Using these mice and wild-type (WT) controls, we studied the role of Gstm1 in a CKD model and a HTN model. We examined the importance of superoxide in the mediating pathways and of hematopoietic GSTM1 expression on renal inflammation. Further, we fed mice sulforaphane-rich broccoli powder (SRBP) to determine whether cruciferous vegetables provide a Gstm1-dependent protective effect in kidney disease, similar to that observed in lung cancer. The clinical relevance of this effect was also examined in humans, using data from the ARIC study, a community-based prospective cohort.

Methods

Additional details of the methods can be found in Supplemental Appendix 1.

Animal Models

Experiments were carried out in accordance with local and National Institutes of Health guidelines, and the animal protocol was approved by the University of Virginia and University of Rochester Institutional Animal Care and Use Committee. All mice were maintained on a 12-hour light-dark cycle with free access to standard chow and water. The mouse line carrying the Gstm1 deletion was generated by the electroporation of embryonic stem cells derived from the 129/SvEv (129S6; Taconic) strain with a targeting construct to disrupt a 20 kb segment of the Gstm1 locus (Supplemental Figure 1A). Ablation of Gstm1 was confirmed by Southern blotting and quantitative RT-PCR (Supplemental Figure 1, B and C).

Subtotal Nephrectomy–Induced CKD Model

A subtotal nephrectomy (Nx) was used to induce CKD in the model, as previously described.32 Briefly, mice were anesthetized with 2%–4% isoflurane. The right kidney was removed and the cranial branch of the left renal artery was ligated, resulting in the infarction of the cranial pole of the left kidney. Mice were allowed to recover for at least 4 weeks before being subjected to any other procedures.

Angiotensin II–Induced HTN Model

The HTN model was induced by the delivery of angiotensin II (AngII; Sigma-Aldrich, St. Louis, MO) at 1000 ng/kg per minutes for 4 weeks via Alzet miniosmotic pumps (model 2004; Durect Corporation, Cupertino, CA).

Antioxidant Treatment

Tempol (4-hydroxy-Tempo; Sigma-Aldrich) was administered to mice in drinking water as a 3 mmol/L solution, simultaneous with AngII infusion. Mice treated with SRBP (Natural Sprout Company) had it mixed with powdered chow at a 1:1 ratio. SRBP contains little sulforaphane, but rather glucoraphanin, which is metabolized into sulforaphane in the gut.

BP Monitoring in vivo

BP was measured in conscious mice under unrestrained conditions by radiotelemetry (TA11PA-C10; Data Sciences International, St. Paul, MN), as previously described.33 Briefly, mice were anesthetized with 2%–4% isoflurane and had the radiotelemetry catheter implanted into the left carotid artery. A subcutaneous pouch was made along the right flank of the animal, and the transmitter was placed as caudally as possible. Mice were housed in individual cages on receivers and allowed to recover for at least 7 days after implantation of the radiotelemetry device before measurements were recorded and analyzed using Dataquest A.R.T. 20 software (Data Sciences International). Implantation was performed at least 4 weeks after Nx-CKD or before AngII-HTN initiation. Reported values are expressed as mean±SD.

Urine Measurements

Urine samples were collected over 24 hours from mice placed in individual metabolic cages for assessment of 8-isoprostane and urinary albumin and creatinine, as previously described.32 Urinary isoprostane was measured using the ELISA kit from Oxford Biomedical Research (Oxford, MI). Urinary albumin and creatinine were measured using the Albuwell M Murine ELISA kit, and Creatinine Companion kit (Exocell, Philadelphia PA).

Measurement of Tissue Superoxide

The lucigenin (9,9′-bis-N-methylacridinium nitrate, M8010; Sigma) assay was modified from previous studies as described,33 and performed on 3–5 mg of kidney cortical tissue. Luminescence counts were taken five times for 1 minute each, averaged, corrected for background, and normalized to dry tissue weight.

Histopathologic Analyses

Kidney injury was scored blinded using formalin-fixed sections stained with periodic acid-–Schiff as previously described.32 Digital quantitation of glomerular and mesangial surface areas was performed by light microscopy (Zeiss Axio Imager Z2) using Stereoinvestigator software (BMF Bioscience) as previously described.34

Podocyte Isolation and Scratch Assay

Podocytes were isolated as previously described.35 At confluence, mechanical scraping was performed with a 200 μl pipette tip, creating a rectangular-shaped wound in the podocyte layer. Images of podocytes with the scratched area were taken immediately after wound creation (0 hour) and after 14 hours, using an EVOS XL Core Cell Imaging System at ×10 magnification. The images were analyzed using ImageJ software (1.48v; National Institutes of Health) to obtain pixel counts in the scratched area void of podocytes.

Gene Expression Analysis

RNA from frozen kidney tissue was isolated by RNeasy Mini kit (Qiagen) and transcribed to complementary DNA by iScript cDNA synthesis kit (Bio-Rad). Quantitative RT-PCR analysis was performed on an iQ5 system (Bio-Rad) with normalization to Hprt as the reference gene. The sequences of all primers used in this study are listed in Supplemental Table 12.

Flow Cytometry

Single-cell suspensions were made from kidney tissue, stained with antibodies for 1 hour at 4°C, and analyzed on a FACSCalibur system (BD Biosciences, San Jose, CA) with an eight-color detector (Cytek Development, Fremont, CA). Live, singlet CD45+ cells (leukocytes) were identified and quantified on the basis of positive staining of respective markers.

BM Transplantation

BM cells were obtained from the femurs and tibias of donor mice. Recipient mice were irradiated and then 0.2 ml of BM cells (5 million cells) were injected via the tail vein. Recipient mice were maintained on antibiotics (enrofloxacin) with autoclaved water and food for 5 weeks. Eight weeks post-transplantation, HTN was induced as described above.

Statistical Analyses

Unpaired t test, one-way ANOVA with post hoc Bonferroni correction, and two-way ANOVA were used as indicated. Differences in Kaplan–Meier survival curves were tested using the log-rank test. Differences were considered statistically significant when P<0.05. All values are presented in the text and figures as mean±SD.

ARIC Study

The ARIC study was a prospective cohort study of 15,792 participants in the United States, aged 45–65 years at visit 1 (1987–1989).The overall analyzed sample in this study included 10,155 participants. The included participants provided informed consent for genetic studies. This study has been approved by the Johns Hopkins Institutional Review Board.

GSTM1 genotypes were determined using exome sequencing reads. The methods have been reported previously.3 The concordance between this algorithm of calling GSTM1 deletion using exome sequencing reads and quantitative PCR was 93% (Supplemental Figure 6). We adopted the recessive genetic model, i.e., homozygous deletion [GSTM1(0/0)] versus heterozygous deletion [GSTM1(0/1)] and no deletion [GSTM1(1/1)] combined, as our primary model on the basis of previous findings from lung cancer in humans.18–20,22 We also performed secondary analysis using the genotypic model, i.e., using the three genotypes as a categorical variable.

Cruciferous vegetable intake of the participants was assessed using a food frequency questionnaire on their dietary intake in the previous year (the Willett 131-item food frequency questionnaire).36 Among the questions in this assessment, one question was specific to broccoli intake, and another was on cabbage, cauliflower, and Brussels sprouts intake. These vegetables have commonly been considered as cruciferous vegetables containing ITC, a protective electrophile.37 Cruciferous vegetable intake was categorized into three levels: low (three or fewer times per month), medium (more than three times per month but less than once per week), and high (once or more per week).

Kidney failure was defined as ESRD ascertained using linkage to the US Renal Data System or kidney failure on the basis of International Classification of Diseases, Ninth or Tenth Revision, Clinical Modification (ICD-9-CM/ICD-10-CM) codes for hospitalization or death.38 Participants were followed from the baseline (visit 1) to December 2013.

Statistical Analyses in the ARIC Study

We compared the baseline characteristics of the participants by three levels of cruciferous vegetable intake, using t tests for nonskewed continuous variables, Wilcoxon tests for skewed continuous variables, and chi-squared tests for categorical variables. Our primary analysis evaluated the association between the three levels of cruciferous vegetable intake and kidney failure stratified by GSTM1 homozygous deletion to determine whether the protective effect of cruciferous vegetable intake was indeed stronger among participants with GSTM1(0/0). Kaplan–Meier curves were plotted by the three levels of cruciferous vegetable intake within those with and without GSTM1 homozygous deletion. Risk of kidney failure was assessed by Cox regression. Model 1 controlled for age, sex, study center, and genetic principal components. Model 2 additionally controlled for clinical risk factors of kidney failure (baseline eGFR, prevalent diabetes, HTN, coronary heart disease, smoking status, body mass index, physical activity, education levels, and total calorie intake). We evaluated these two models combining black and white participants controlling for race within each GSTM1 homozygous deletion stratum. To determine whether the association between cruciferous vegetable intake and kidney failure was stronger in those with GSTM1 homozygous deletion, we tested for the interaction between GSTM1 homozygous deletion and cruciferous vegetable intake. To determine whether the association between cruciferous vegetable intake and kidney failure within the GSTM1 homozygous deletion strata differed by the two race groups, we tested for race interaction with cruciferous vegetable intake within each GSTM1 homozygous deletion stratum.

Furthermore, we conducted three sets of secondary analyses. The first set evaluated the association between GSTM1 genotypes (0/0, 0/1, and 1/1) and kidney failure within the three levels of cruciferous vegetable intake. The second evaluated the main effect of cruciferous vegetable intake on kidney failure. The third evaluated the main effect of GSTM1 deletion genotype on kidney failure. The covariates of these secondary analyses were the same as the primary analysis.

All Cox regression analyses were conducted using Stata version 14. All other analyses were conducted using R.

Results

Gstm1 Null Mice Have Increased BP and Oxidative Stress at Baseline

Gstm1 KO mice displayed a modest but statistically significantly higher average baseline systolic BP (SBP) of 7 mm Hg compared with WT controls (Figure 1A). We previously reported through knockdown studies that GSTM1 influences oxidative stress in VSMCs in vitro.2,29 Consistent with these earlier findings, Gstm1−/− (KO) mice displayed higher baseline urinary 8-isoprostane levels (marker of oxidative stress) (Figure 1B), suggesting that the deletion directly results in increased oxidative stress in vivo. No difference in urinary albumin-to-creatinine ratio (ACR) was observed (Figure 1C), suggesting loss of GSTM1 does not influence the glomerular filtration barrier or tubular resorptive capacity in nondisease state. At 1 year of age, there were no observed differences in mortality or phenotypes between WT and Gstm1 KO mice (Supplemental Figure 1, D–F).

Figure 1.
Figure 1.

Gstm1 KO mice have higher BP and renal oxidative stress at baseline. (A) SBP measured by radiotelemetry was significantly higher in Gstm1 KO mice: WT 129.0±4.9 versus KO 135.6±2.7 mm Hg; P=0.02; n=6 each. (B) Urinary ACR was not different between groups: WT 27.3±7.2 versus KO 29.1±10.1 µg/mg; P=0.73; n=6 each. (C) Twenty four hour excretion of urinary 8-isoprostane was elevated in Gstm1 KO mice: WT 6.6±1.2 versus KO 14.2±3.1 ng/100 mg body wt/24 h; P=0.001; n=5/6. All comparisons by t test. *P<0.05; **P<0.01.

Deletion of Gstm1 Results in Poor Survival and Exaggerated HTN and Kidney Injury in the Nx-CKD Model

We previously reported that both Nrf2 and GSTM1 proteins are upregulated in the remnant kidney in WT mice after Nx,2 suggesting that this axis plays a role in adaptive response after loss of nephron mass. We subjected eight WT and Gstm1 KO mice to the Nx-CKD model. Gstm1 KO mice had significantly worse survival after Nx, with a 23% survival rate at 83 days (12 weeks) after the procedure, whereas there was no mortality in WT mice at 90 days (13 weeks, original study end point) (Figure 2A).

Figure 2.
Figure 2.

Deletion of Gstm1 in the Nx CDK model results in poor survival and exaggerated HTN and kidney injury. (A) Survival curve showing significant mortality in Gstm1 KO mice after Nx; P=0.01 versus WT; n=8 each. Time of censoring for mice that were used for terminal experiments are marked by ×. (B) SBP measured by radiotelemetry was significantly higher in Gstm1 KO mice: WT 135.0±8.4 versus KO 150.6±13.7 mm Hg; P=0.04; n=4/7. (C) Renal superoxide levels measured by lucigenin luminescence and normalized to dry tissue weight were higher in Gstm1 KO mice: WT 74.5±36.1 versus KO 250.4±153.6; P=0.04; n=6 each. (D) Urinary ACR was higher in Gstm1 KO mice: WT 235.1±60.3 versus KO 748.4±460.4 µg/mg; P=0.04; n=6 each. (E) Kidney pathology scores were greater in Gstm1 KO mice for all compartments (total: WT 8.3±2.3 versus KO 18.7±3.6; P<0.001) and in most individual compartments (glomeruli: WT 6.5±1.2 versus KO 14.5±2.7; P<0.001; tubules: WT 1.0±0.0 versus KO 1.5±0.8; P=0.20; interstitium: WT 0.8±1.2 versus KO 2.7±0.8; P=0.01), n=6 each. (F) Area fractions were greater in Gstm1 KO mice for glomerular to total noninfarcted kidney (G/T, left: WT 2.3±0.7 versus KO 3.3±0.7; P=0.03) and mesangial to glomerular (M/G, right: WT 23.7±6.1 versus KO 35.7±5.5; P=0.005), n=6 each. (G) Percent of scratched area that remained uncovered after 14 hours was reduced in Gstm1 KO mice: WT 69.3±11.5 versus KO 54.2±8.6; P=0.007; n=9 each. All comparisons by t test. *P<0.05; **P<0.01; ***P<0.001.

Because of the observed high mortality beyond 8 weeks, another cohort of mice underwent Nx and animals and tissue samples were analyzed at 56 days (8 weeks) after Nx. By radiotelemetry, Gstm1 KO mice had significantly greater SBP (Figure 2B), and elevated renal superoxide levels as measured by lucigenin luminescence (Figure 2C), consistent with the notion that GSTM1 plays a role in regulating oxidative stress. Gstm1 KO mice had higher urinary ACR (Figure 2D), and quantitation of kidney histopathology (Figure 2E, Supplemental Figure 2B) confirmed they had greater tissue injury, characterized by severe glomerulosclerosis and chronic inflammation. Digital quantitation of glomerular and mesangial surface area showed a significant increase in both measures in Gstm1 KO mice (Figure 2F), suggesting that glomerular hyperfiltration and increased glomerular remodeling39 is influenced by deletion of GSTM1.

Given the significance of the podocytes to the renal filtration apparatus and the increased albuminuria in Gstm1 KO mice after Nx, we set out to determine whether GSTM1 deficiency has a direct effect on podocyte health. Primary podocytes (synaptopodin-positive cells) were isolated from healthy, unmanipulated WT and KO kidneys, and the GSTM1 expression was found only in WT cells, as expected (Supplemental Figure 2, C–E). Migration capacity using a standard ex vivo scratch assay35 showed that Gstm1 KO podocytes migrate at a significantly higher rate than those from WT mice, as reflected by the significantly smaller noncovered area after 14 hours (Figure 2G). The motile phenotype of podocytes in vitro is generally considered to be analogous to podocyte foot effacement and injury in vivo.40,41 These data suggest that deletion of Gstm1 may exaggerate podocyte response to stress and injury, and contribute to the increased glomerular injury observed in mice with Nx-CKD.

Deletion of Gstm1 Results in Exaggerated Kidney Injury Mediated by Oxidative Stress in the AngII-HTN Model

To determine the generalizability of the effect of GSTM1 deficiency on susceptibility to kidney injury in disease states, we next determined the effect of Gstm1 deletion in a model of HTN induced by AngII infusion. Despite a modest increase in baseline SBP in Gstm1 KO mice, there was no difference in the severity of HTN (Figure 3A) or urinary ACR (Figure 3B). However, Gstm1 KO mice had approximately three times higher superoxide radicals in kidney tissue (Figure 3C), despite a lack of difference in NADPH subunits (Nox2 and Nox4, Supplemental Figure 3A)42 and SOD (cytosolic CuZn-SOD [SOD1], extracellular CuZn-SOD [SOD3], and mitochondrial Mn-SOD [SOD2], Supplemental Figure 3B). Gstm1 KO mice also had greater tissue injury compared with WT mice (Figure 3D, Supplemental Figure 3C). We next quantified indicators of inflammation given its significance in the development of AngII-HTN,43,44 and found that KO mice had a significant increase in renal expression of several genes involved in inflammation, including CXCL-1, MCP-1, and IL-6 (Figure 3E, Supplemental Figure 3D).

Figure 3.
Figure 3.

Deletion of Gstm1 in the AngII HTN model results in exaggerated oxidative stress and kidney injury that is ameliorated by Tempol. (A) SBP measured by radiotelemetry was significantly higher in Gstm1 KO mice treated with Tempol (KO+T): WT 157.8±6.7 versus KO 163.7±14.1 versus KO+T 195.6±13.2 mm Hg; P=0.001; n=5/6/4. (B) Urinary ACR was not significantly different: WT 5108.2±1302.9 versus KO 4657.9±1262.2 µg/mg; P=0.62; n=5/4. (C) Renal superoxide levels measured by lucigenin luminescence and normalized to dry tissue weight were higher in Gstm1 KO mice: WT 41.6±20.8 versus KO 128.25±30.5 versus KO+T 53.8±18.4; P<0.001; n=6/6/4. (D) Kidney pathology scores were greater in Gstm1 KO mice for all compartments (total: WT 8.8±6.1 versus KO 19.8±4.1 versus KO+T 7.3±3.8; P=0.002) and in individual compartments (glomeruli: WT 7.7±4.2 versus KO 15.2±2.7 versus KO+T 7.0±3.6; P=0.003; tubules: WT 0.8±1.3 versus KO 2.3±0.8 versus KO+T 0.3±0.5; P=0.01; interstitium: WT 0.3±0.8 versus KO 2.3±1.5 versus KO+T 0.0±0.0; P=0.006), n=6/6/4. (E) Renal mRNA levels of genes involved in inflammation normalized to Hprt in WT mice at baseline (WTBL) and with HTN (WT), n=6/5 and KO mice at baseline (KOBL), with HTN (KO) and with HTN treated with Tempol (KO+T), n=6/6/4. Expression levels were elevated in the KO mice for MCP-1 (KOBL 0.0054±0.0050 versus KO 0.048±0.016 versus KO+T 0.019±0.0068; P<0.001) and CXCL-1 (KOBL 0.0071±0.0027 versus KO 0.068±0.028 versus KO+T 0.012±0.0051; P<0.001). WT comparisons in (E) by t test. Comparisons in other panels and KO comparisons in (E) by one-way ANOVA. *P<0.05; **P<0.01; ***P<0.001; ****P<0.001.

Because the AngII-HTN model had fewer surgery-related complications and mortality compared with the Nx-CKD model, further studies were conducted in the AngII-HTN model. To determine whether oxidative stress was mediating the exaggerated disease phenotype in Gstm1 KO mice, Tempol, a superoxide scavenger, was administered concurrently with AngII. Although Tempol caused a mild increase in SBP (Figure 3A) and no change in ACR (Figure 3B), it did significantly decrease renal superoxide levels (Figure 3C), kidney pathology (Figures 3D, Supplemental Figure 3C), and expression of MCP-1 and CXCL-1 (Figure 3E) in KO mice. Although it is unclear why SBP increased in KO mice treated with Tempol, these findings suggest that deletion of Gstm1 enhances kidney injury through an oxidative stress pathway independent of BP.

Gstm1 KO Mice in the AngII-HTN Model Have Increased Renal Inflammation associated with the Deletion of Gstm1 in the Parenchyma

Increased expression of chemokines and cytokines in Gstm1 KO kidneys suggested increased renal inflammation. We next used flow cytometry to quantify in an unbiased manner the proportion of inflammatory cells in AngII-HTN WT and KO kidneys. The gating strategy is shown in Supplemental Figure 4A. Analysis showed more adaptive (CD4+ and CD8+ T cells) and innate (neutrophils [Ly6G, PMN] and macrophages [F4/80]) inflammatory cells in the Gstm1 KO kidney (Figure 4A) that were primarily sequestered in the renal interstitium (Supplemental Figure 4, B and C). This suggests that during AngII-HTN, loss of GSTM1 augments renal inflammation.

Figure 4.
Figure 4.

Gstm1 KO mice in the AngII HTN model have increased renal inflammation associated with the deletion of Gstm1 in the parenchyma. (A) Flow cytometry data of renal leukocytes normalized to WT counts showed Gstm1 KO mice have increased populations (P values: CD4+T=0.04; CD8+T=0.05; B=0.16; PMN=0.007; F4/80High=0.004), n=3/4. Comparisons by one-way ANOVA. *P<0.05; **P<0.01. (B) Flow cytometry data of renal leukocytes for BM chimeras generated from BM crosstransplantation (genotypes listed as donor-recipient) normalized to WT-WT showed recipient genotype was the statistically significant factor (donor P values: CD4+T=0.82; CD8+T=0.95; B=0.87; PMN=0.70; F4/80High=0.31; recipient P values: CD4+T=5.9×10−5; CD8+T=0.30; B=1.1×10−5; PMN=1.5×10−6; F4/80High=5.0×10−7), n=4/5/5/4. Comparisons by two-way ANOVA.

Because GSTM1 is abundantly expressed in hematopoietic cells and renal epithelial cells, and CXCL-1 and MCP-1 are not uniquely expressed in macrophages and neutrophils but are also expressed in renal epithelial cells, we next investigated whether deletion of Gstm1 in BM-derived cells or in the parenchyma contributes to renal inflammation and injury in the AngII-HTN model, using a BM crosstransplant approach. Analysis of the renal CD45+ population by two-way ANOVA revealed that the recipient genotype significantly influenced the degree of renal inflammation, with KO recipient mice having increased renal inflammation. There was no significant effect of donor genotype nor of the interaction between donor and recipient. Hence, our study shows that deletion of Gstm1 in the parenchyma, and not in BM-derived cells, drives renal inflammation and injury.

Supplementation with SRBP Ameliorates Kidney Injury in Gstm1 KO Mice but Not in WT Mice

We first confirmed the bioavailability and activity of synthetic sulforaphane (Supplemental Figure 5A) and showed that dietary supplementation of SRBP produced similar effects in WT mice without effect on body weight (Supplemental Figure 5, B and C). As expected, SRBP induced Gstm1 expression in the kidney in WT mice during AngII-HTN, but not in KO mice in which Gstm1 expression was not detectable (Supplemental Figure 5D). SRBP supplementation did not lower SBP in either WT or KO mice in the AngII-HTN model (Figure 5A), and in fact increased SBP in WT mice. However, SRBP significantly decreased renal levels of superoxide (Figure 5B), ACR (Figure 5C), and kidney histopathology scores in the glomeruli and interstitium in KO mice (Figure 5D, Supplemental Figure 5E). In most of these metrics, KO mice fed SRBP were similar to WT mice fed SRBP. In contrast, WT mice treated with SRBP had no significant change compared with untreated WT mice. Our result suggests that the beneficial effect of SRBP is dependent on the absence Gstm1.

Figure 5.
Figure 5.

Treatment with SRBP ameliorated kidney injury only in Gstm1 KO mice. In all panels, WT and KO mice are the same as those in Figure 3, with the data reproduced here for ease of comparison. (A) SBP measured by radiotelemetry was elevated in WT mice fed SRBP (WT+SRBP 176.9±5.9 mm Hg; P=0.001; n=6) but not in Gstm1 KO mice fed SRBP (KO+SRBP 168.9±3.5 mm Hg; P=0.24; n=6). (B) Renal superoxide levels measured by lucigenin luminescence and normalized to dry tissue weight were not affected by SRBP in WT mice (WT+SRBP 57.7±31.5 counts/min per milligram dry tissue; P=0.02; n=6) but were reduced in Gstm1 KO mice (KO+SRBP 50.4±33.9 counts/min per milligram dry tissue; post hoc versus KO P=0.004; n=5). (C) Urinary ACR was reduced by SRBP in Gstm1 KO mice (KO+SRBP 1671.5±683.6 µg/mg; post hoc versus KO P=0.05; n=5), but not WT mice (WT+SRBP 4314.4±2155.3 µg/mg; P=0.47; n=6). (D) Kidney pathology scores were unaffected by SRBP in WT mice (total: 5.4±1.5; P=0.24; glomeruli: 4.4±0.5; P=0.11; tubules: 1.0±1.0; P=0.82; interstitium: 0.0±0.0; P=0.36; n=5) but were reduced in Gstm1 KO mice (total: 4.6±1.7; post hoc versus KO P=2.0×10−6; glomeruli: 3.6±0.9; post hoc versus KO P=2.4×10−7; tubules: 1.0±1.0; post hoc versus KO P=0.10; interstitium: 0.0±0.0; P=0.004; n=5). Comparisons between WT and WT+SRBP by t test. Comparisons between WT+SRBP, KO, and KO+SRBP by one-way ANOVA. *P<0.05; **P<0.01; ***P<0.001; ****P<0.001.

In Humans, GSTM1 Deletion Modifies the Protective Effect of Cruciferous Vegetables Against Kidney Failure

To determine whether our findings of the protective effect of SRBP in GSTM1 KO mice can be translated to humans, we evaluated whether GSTM1 deletion modified the association between cruciferous vegetable intake and kidney failure in the ARIC study.

Population Characteristics

Among 10,155 participants in the ARIC study, the mean age was 54 years, 44% were men, and 27% were black. The mean baseline eGFR was 103 ml/min per 1.73 m2. The participants in the three cruciferous vegetable intake categories did not differ significantly by age, body mass index, prevalent diabetes, or baseline eGFR. By the three cruciferous vegetable intake categories, the high-intake group had significantly lower proportion of men (high intake: 32.6%, low intake: 57.0%), black participants (high intake: 17.5%, low intake: 34.7%), and current smokers (high intake: 22.3%, low intake 30.9%) (Table 1). Among the participants in this study, a subsample (n=5461) was previously studied for the association between GSTM1 deletion and kidney failure.3 Compared with the previous subsample, the additional participants in this study (n=4694) had higher intake of cruciferous vegetables (high intake: previous subsample 22.0%, additional participants 26.7%; P<0.001; Supplemental Table 1).

View this table:
Table 1.

Baseline population characteristics by cruciferous vegetable intake category (n=10, 155)

The frequencies of GSTM1 deletion genotype among black and white participants were similar to the race-specific frequencies reported in other study populations (black participants: GSTM1(0/0), 26.1%; GSTM1(0/1), 48.3%; GSTM1(1/1), 25.7%; white participants GSTM1(0/0), 52.3%; GSTM1(0/1), 39.5%; GSTM1(1/1), 8.2%; Supplemental Table 2).4

Protective Association of Higher Cruciferous Vegetable Intake was Mainly Observed among Participants with GSTM1 Homozygous Deletion

Over a median of 24.6 years, 370 kidney failure events were observed (black participants, 188; white participants, 182). In the primary analysis, the protective association of higher intake of cruciferous vegetable for kidney failure was significantly stronger among those with GSTM1(0/0) genotype than among those with GSTM1(0/1) or GSTM1(1/1) genotypes (P value for interaction between cruciferous vegetable intake and GSTM1 genotype <0.05). Among participants with GSTM1(0/0) genotype, in model 1, higher intake of cruciferous vegetables was associated with >40% reduction in risk of kidney failure (model 1 race-combined hazard ratio [HR]: medium intake, 0.58; 95% confidence interval [95% CI], 0.41 to 0.82; high intake, 0.41; 95% CI, 0.25 to 0.68; P-trend<0.001; Figure 6, Table 2). With the additional adjustment for clinical risk factors, this association remained similar (model 2 race-combined HR: medium intake, 0.68; 95% CI, 0.48 to 0.98; high intake, 0.49; 95% CI, 0.29 to 0.83; P-trend=0.005). Among those with GSTM1(0/1) or GSTM1(1/1) genotypes, cruciferous vegetable intake did not have significant dose response association with kidney failure (model 1: race-combined P-trend=0.45; model 2: race-combined P-trend=0.46; P value for interaction between cruciferous vegetable intake and GSTM1 genotype, Model 1: 0.02; Model 2: 0.03) (Table 2). Analysis with GSTM1(0/1) and GSTM1(1/1) separately yielded similar results. (Supplemental Table 3).

Figure 6.
Figure 6.

Humans with GSTM1(0/0) have reduced incidence of kidney failure with higher cruciferous vegetable intake. Kaplan–Meier plots of kidney failure by cruciferous vegetable intake category within those (A) with and (B) without GSTM1 homozygous deletion showed a significant benefit only in those with the homozygous deletion. See Table 2 for statistical analysis.

View this table:
Table 2.

Association between cruciferous vegetable intake and kidney failure stratified by GSTM1

In the analysis within each race group, the association between cruciferous vegetable intake and kidney failure by GSTM1 genotype was consistent across the two race groups (all P for interaction between race and cruciferous vegetable intake within GSTM1 strata >0.05; Supplemental Table 4). These results suggest GSTM1 deletion modifies the protective effect of cruciferous vegetable for kidney failure across populations.

Association between GSTM1 Deletion and Kidney Failure by Cruciferous Vegetable Intake Levels

In the analysis comparing GSTM1(0/0) versus GSTM1(0/1) or GSTM1(1/1), GSTM1(0/0) did not have significant association with kidney failure within each of the cruciferous intake category adjusting for demographic and clinical covariates (Supplemental Table 5). When the three GSTM1 deletion genotypes were analyzed separately, GSTM1 deletion genotypes (0/0 and 0/1) were associated with higher risk for kidney failure compared with GSTM1(1/1) among those with medium intake (P-trend=0.03; Supplemental Table 6).

Overall Association of Cruciferous Vegetable Intake and GSTM1 Deletion with Kidney Failure

Overall higher cruciferous vegetable intake was associated with lower risk of kidney failure (model 2 race-combined HR: medium intake, 0.68; 95% CI, 0.53 to 0.86; high intake, 0.70; 95% CI, 0.51 to 0.96; P-trend=0.01) (Supplemental Table 7). Similar results were observed within the race groups. In contrast, GSTM1(0/0) did not have significant overall association with kidney failure when compared with GSTM1(0/1) and GSTM1(1/1) combined (Supplemental Table 8) or by the three deletion genotypes (model 2 race-combined HR: GSTM1(0/0), 1.30; 95% CI, 0.92 to 1.82; GSTM1(0/1), 1.24; 95% CI, 0.89 to 1.72; P-trend=0.18; Supplemental Table 9), suggesting that the main effect of GSTM1 deletion can be highly dependent on environmental factors.

Discussion

We have shown that Gstm1 deletion in the mouse directly influences the severity of kidney injury in two disease models (Nx-CKD and AngII-HTN). In the Nx-CKD model, Gstm1 deletion resulted in poor survival and exaggerated HTN and kidney injury. In the AngII-HTN model, Gstm1 deletion resulted in exaggerated kidney injury associated with heightened renal inflammation. This is likely mediated by oxidative stress, which was increased in both models, as treatment with Tempol, a superoxide scavenger, in the AngII-HTN model reduced oxidative stress, expression of MCP-1 and CXCL-1, and kidney injury without lowering BP in Gstm1 KO mice. BM crosstransplantation delineated that Gstm1 in the parenchyma, and not in the hematopoietic cells, determined renal inflammation. Finally, Gstm1 deletion modified the effect of SRBP supplementation, as SRBP supplementation reduced superoxide levels, reduced ACR, and improved kidney histopathology scores in KO mice but not in WT. This modifying effect of GSTM1 deletion in mice on SRBP supplementation was also observed in humans in the ARIC study, where the protective effect of cruciferous vegetable intake on kidney failure was mainly found in those with homozygous deletion of GSTM1.

Gstm1 is a positional candidate gene for HTN in the SHRSP rats.31 Here, we found that deletion of Gstm1 in the mouse resulted in a small but significant increase in SBP at baseline and exaggerated HTN in the Nx-CKD model. This could be explained by increased oxidative stress at baseline and more severe kidney injury in the Nx-CKD model. Our result is consistent with a recent study showing that transgenic overexpression of Gstm1 in the kidney injury-prone SHRSP rat rescued renal oxidative stress and HTN.45 In the AngII-HTN model, there was no difference in SBP between WT and KO mice and we speculate that the very high dose of AngII administered potentially masked any modest BP effect of Gstm1 deletion.

In both Nx-CKD and AngII-HTN induced disease models, we demonstrated that deletion of Gstm1 directly influenced the severity of kidney injury in the mouse and that this was associated with increased renal oxidative stress. Although we cannot at this time attribute worse kidney injury to worse HTN or vice versa in the Nx-CKD model, these two phenotypes in Gstm1 KO mice correlate with those seen in human patients with CKD in whom HTN becomes more severe with worsening kidney disease.46

The mechanism(s) by which Gstm1 deletion exaggerates kidney disease is likely through oxidative stress and inflammation, as both Tempol and SRBP lowered renal superoxide levels, inflammation, and kidney pathology scores, without lowering BP, in the AngII-HTN model. We previously reported that knockdown of Gstm1 in VSMCs in vitro resulted in increased levels of superoxide.29 Our present in vivo studies found that deletion of Gstm1 also resulted in increased renal superoxide in both the Nx-CKD and AngII-HTN models. There is no evidence that GSTM1 has SOD activity to metabolize superoxide. Moreover, deletion of Gstm1 did not result in increased expression of Nox2 or Nox4 isoforms of NAD(P)H oxidase, nor altered expression level of SOD in the AngII-HTN model. These data suggest that GSTM1 might influence superoxide metabolism indirectly. Because GSTM1 enzyme has overlapping substrate specificities with other GST enzymes, the increased superoxide levels in Gstm1 KO mice also suggests that there is inadequate compensation by other GST enzyme(s) or pathway(s) that determine superoxide levels. Furthermore, enzyme kinetic assays have demonstrated that GSTM1 has activity against epoxides47 and several reactive aldehydes48–50 and, in vivo, transgenic overexpression of Gstm1 in the SHRSP rat decreases renal levels of malondialdehyde.45 The multiple substrates metabolized by GSTM1 suggest that their metabolites and downstream effects may be influenced by disease states. Identification of reactive oxygen species regulated by GSTM1 that are involved in kidney injury in these models would be informative to further elucidate the mechanisms.

In the Nx-CKD model, Gstm1 KO mice displayed increased mortality, in addition to worse HTN, oxidative stress, and kidney injury. The reason for the poor survival of these mice beyond 8 weeks after Nx is unclear. The increased ACR and glomerular and mesangial surface areas in Gstm1 KO mice may reflect increased glomerular HTN from GSTM1 deficiency, raising the question whether GSTM1 deficiency influences single nephron GFR, preglomerular myogenic tone, and postglomerular vasodilation. Furthermore, we report for the first time that GSTM1 is normally expressed in podocytes, and that GSTM1 deficiency increases podocyte migration in response to wound scratch in vitro. The “motile” phenotype of podocytes in vitro is considered analogous to podocyte injury and effacement in vivo.40,51 The potential role of GSTM1 in glomerular hyperfiltration and podocyte function in disease states will be a focus of future studies.

In the AngII-HTN model, deletion of Gstm1 resulted in worse kidney injury, but not an increase in mortality as in the Nx-CKD model. Thus, further experiments were all conducted using the AngII-HTN model. The kidney injury in the AngII-HTN model was associated with increased expression of genes involved in inflammation, particularly those expressed in macrophages, raising the possibility that GSTM1 may mediate macrophage inflammatory response in certain disease states. This is consistent with findings by Alexis et al.,52 who reported that human volunteers with the GSTM1 null genotype exposed to ozone had significantly increased expression of HLA-DR on airway macrophages and dendritic cells. Furthermore, SRBP inhibited MIF tautomerase, and decreased renal inflammation in Gstm1 KO mice. CXCL-1 and MCP-1 are not uniquely expressed in macrophages and neutrophils, but are also expressed in endothelial53 and renal epithelial cells.54 Our BM chimeras suggest that deletion of Gstm1 determines renal inflammation in AngII-HTN mechanistically through a parenchymal-driven pathway rather than via the hematopoietic cells. However, we cannot rule out that Gstm1 deletion activates resident leukocytes in the kidney using the BM crosstransplantation approach.

Although deletion of Gstm1 increased kidney injury and inflammation, dietary administration of SRBP in the mice protected against renal injury and inflammation only in Gstm1 KO mice. Similarly, the protective effect of cruciferous vegetable consumption in lowering the risk of kidney failure in ARIC participants was observed largely in those with GSTM1(0/0) genotype. This modifying effect of GSTM1 is consistent with that observed in lung cancer, illustrating the pleiotropic effect of GSTM1. We have shown that GSTM1 deletion modifies the effect of the renal risk variants in the apoL1 (APOL1) gene,55 suggesting that the modifying effects of this deletion may extend even further. It is worth noting that APOL1 genotype in the donor kidney, and not in the recipient, determines outcomes in kidney transplantation.56 This raises the question how APOL1 and GSTM1 interact in the kidney to affect susceptibility to kidney injury and kidney disease progression.

Taken together, these studies suggest that GSTM1 deletion may be a modifier of a broad range of disease phenotypes. The lack of phenotype in KO mice at 1 year of age is consistent with the notion that GSTM1 deletion is a modifier of stressors. We speculate that loss of GSTM1 enzyme can increase the accumulation of toxic metabolites and reactive oxygen species, thereby worsening disease progression; however, GSTM1 deficiency can also increase the level of protective metabolites from dietary intake in disease states, such as with sulforaphane, which activates the Nrf2 signaling pathway, thereby inducing a wide variety of phase 2 detoxification enzymes.23,24 In line with this notion, in the Sulforaphane in Treating Patients with Recurrent Prostate Cancer Trial, those with the GSMT1 null genotype have a longer median half-life (t1/2) of sulforaphane in the blood (2.6, range [2.0–5.5] hours), compared with those with the active genotype (2.1, range [1.8–2.9] hours) (https://clinicaltrials.gov/ct2/show/results/NCT01228084?sect=Xedc0156#outcome8). Another study reported that bioavailability of free sulforaphane and sulforaphane metabolites were increased in GSTM1 null subjects compared with those with the active allele.25 Moreover, we show in the mouse that SRBP can increase the expression level of GSTM1 in WT mice, which may in turn further decrease the bioavailability of sulforaphane, thereby limiting its beneficial effect of activating other antioxidants in kidney disease in WT mice. Furthermore, upregulation of phase 2 antioxidants by sulforaphane could explain its effectiveness in KO mice because there is increased ROS to scavenge even at baseline (Figure 1B). Delineation of the substrates/molecules in the sulforaphane-GSTM1 axis in kidney injury will be a focus of future studies.

It should be noted that in this study, the directions of association of GSTM1 deletion with kidney failure were consistent with the previous smaller study3 in ARIC, whereas the current effect sizes were smaller and not statistically significant. The association between GSTM1 deletion and kidney failure in humans may depend on the levels of protective and toxic intracellular electrophiles that are ligands of GSTM1. The levels of these electrophiles vary by environment and study populations and thus the main effect of GSTM1 deletion may vary across populations.

Our data support a dual role of GSTM enzyme in the modulation of oxidative stress, inflammation, and protective metabolites in kidney disease. Given the high prevalence of the GSTM1 deletion, its population-attributable risk fractions on kidney function decline and ESRD was estimated to be sizeable at 38.4% in hypertensive black Americans.55 Moreover, the risk is not race-specific and has similar effects in white Americans.3 In the context of personalized and precision medicine, increased consumption of cruciferous vegetables and/or sulforaphane may be protective in those lacking GSTM1 who are genetically most at risk.

Disclosures

Dr. Coresh reports grants from National Institutes of Health, grants from National Kidney Foundation, during the conduct of the study. Dr. Grams reports grants from National Institute of Diabetes and Digestive and Kidney Diseases, National Kidney Foundation, and received travel support for an invited speakership to the DCI annual directors meeting. All other authors have nothing to disclose.

Funding

This work was supported by the National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases grants R01 DK094907 (to Dr. Le), R21 DK112087 (to Dr. Tin), and T32 DK072922 (to Dr. Pourafshar and Bodonyi-Kovacs), the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) grant KO 3598/3-1, and Coordinated Research Center grant 1140 (to Dr. Köttgen). The Atherosclerosis Risk in Communities Study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C).

A generous gift of Gstm1 knockout mouse line was funded by the National Institutes of Health grant R21HL098941 (to Dr. Koller).

Acknowledgments

Dr. Gigliotti, Dr. Tin, and Dr. Le, designed the study and wrote the paper, Dr. Gigliotti, Dr. Tin, Dr. Pourafshar, Dr. Wang, Dr. Cechova, Dr. Sung, Dr. Bodonyi-Kovacs, Dr. Cross, Dr. Yang, Ms. Chan, and Miss Nguyen conducted the experiments and manuscript revision. Ruiz performed analysis of histopathology in a blinded manner. All other authors contributed to the Atherosclerosis Risk in Communities (ARIC) study and the review and editing of the manuscript.

We thank Dr. Beverly Koller from the University of North Carolina for the generous gift of Gstm1 knockout mouse line. We also thank Ms. Angela Rose and Ms. Denise Gerrity for assistance with the experiments confirming MIF tautomerase inhibition in mouse tissues following treatment with sulforaphane or feeding with broccoli sprout powder. The stereology data described here was obtained using the “MBF Bioscience and Zeiss microscope system for stereology and tissue morphology” funded by National Institutes of Health grant 1S10RR026799-01. The authors thank the staff and participants of the ARIC study for their important contributions.

Some of the data reported here have been supplied by the US Renal Data System. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the US Government.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019050449/-/DCSupplemental.

Supplemental Appendix 1. Supplemental methods and references.

Supplemental Figures 1–6 and legends.

Supplemental Tables 1–12.

Footnotes

  • J.C.G., A.T., and S.P. contributed equally to this work.

  • Published online ahead of print. Publication date available at www.jasn.org.

References

  1. ↵
    1. United States Renal Data System
    : 2017 USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States, Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 2017
  2. ↵
    1. Chang J,
    2. Ma JZ,
    3. Zeng Q,
    4. Cechova S,
    5. Gantz A,
    6. Nievergelt C, et al
    .: Loss of GSTM1, a NRF2 target, is associated with accelerated progression of hypertensive kidney disease in the African American Study of Kidney Disease (AASK). Am J Physiol Renal Physiol 304: F348–F355, 2013pmid:23220723
    OpenUrlCrossRefPubMed
  3. ↵
    1. Tin A,
    2. Scharpf R,
    3. Estrella MM,
    4. Yu B,
    5. Grove ML,
    6. Chang PP, et al
    .: The loss of GSTM1 associates with kidney failure and heart failure. J Am Soc Nephrol 28: 3345–3352, 2017pmid:28720685
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Garte S,
    2. Gaspari L,
    3. Alexandrie AK,
    4. Ambrosone C,
    5. Autrup H,
    6. Autrup JL, et al
    .: Metabolic gene polymorphism frequencies in control populations. Cancer Epidemiol Biomarkers Prev 10: 1239–1248, 2001pmid:11751440
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Sprenger R,
    2. Schlagenhaufer R,
    3. Kerb R,
    4. Bruhn C,
    5. Brockmöller J,
    6. Roots I, et al
    .: Characterization of the glutathione S-transferase GSTT1 deletion: Discrimination of all genotypes by polymerase chain reaction indicates a trimodular genotype-phenotype correlation. Pharmacogenetics 10: 557–565, 2000pmid:10975610
    OpenUrlCrossRefPubMed
  6. ↵
    1. Xu S,
    2. Wang Y,
    3. Roe B,
    4. Pearson WR
    : Characterization of the human class Mu glutathione S-transferase gene cluster and the GSTM1 deletion. J Biol Chem 273: 3517–3527, 1998pmid:9452477
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Mannervik B,
    2. Board PG,
    3. Hayes JD,
    4. Listowsky I,
    5. Pearson WR
    : Nomenclature for mammalian soluble glutathione transferases. Methods Enzymol 401: 1–8, 2005pmid:16399376
    OpenUrlCrossRefPubMed
  8. ↵
    1. Hayes JD,
    2. Flanagan JU,
    3. Jowsey IR
    : Glutathione transferases. Annu Rev Pharmacol Toxicol 45: 51–88, 2005pmid:15822171
    OpenUrlCrossRefPubMed
  9. ↵
    1. Cho SG,
    2. Lee YH,
    3. Park HS,
    4. Ryoo K,
    5. Kang KW,
    6. Park J, et al
    .: Glutathione S-transferase mu modulates the stress-activated signals by suppressing apoptosis signal-regulating kinase 1. J Biol Chem 276: 12749–12755, 2001pmid:11278289
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Cotton SC,
    2. Sharp L,
    3. Little J,
    4. Brockton N
    : Glutathione S-transferase polymorphisms and colorectal cancer: A HuGE review. Am J Epidemiol 151: 7–32, 2000pmid:10625170
    OpenUrlCrossRefPubMed
  11. ↵
    1. Hou S-M,
    2. Ryberg D,
    3. Fält S,
    4. Deverill A,
    5. Tefre T,
    6. Børresen A-L, et al
    .: GSTM1 and NAT2 polymorphisms in operable and non-operable lung cancer patients. Carcinogenesis 21: 49–54, 2000pmid:10607733
    OpenUrlCrossRefPubMed
  12. ↵
    1. Capoluongo E,
    2. Onder G,
    3. Concolino P,
    4. Russo A,
    5. Santonocito C,
    6. Bernabei R, et al
    .: GSTM1-null polymorphism as possible risk marker for hypertension: Results from the aging and longevity study in the Sirente Geographic Area (ilSIRENTE study). Clin Chim Acta 399: 92–96, 2009pmid:18840420
    OpenUrlCrossRefPubMed
  13. ↵
    1. Cruz-Gonzalez I,
    2. Corral E,
    3. Sanchez-Ledesma M,
    4. Sanchez-Rodriguez A,
    5. Martin-Luengo C,
    6. Gonzalez-Sarmiento R
    : An association between resistant hypertension and the null GSTM1 genotype. J Hum Hypertens 23: 556–558, 2009pmid:19279659
    OpenUrlCrossRefPubMed
  14. ↵
    1. Wang XL,
    2. Greco M,
    3. Sim AS,
    4. Duarte N,
    5. Wang J,
    6. Wilcken DE
    : Glutathione S-transferase mu1 deficiency, cigarette smoking and coronary artery disease. J Cardiovasc Risk 9: 25–31, 2002pmid:11984214
    OpenUrlCrossRefPubMed
  15. ↵
    1. Manfredi S,
    2. Federici C,
    3. Picano E,
    4. Botto N,
    5. Rizza A,
    6. Andreassi MG
    : GSTM1, GSTT1 and CYP1A1 detoxification gene polymorphisms and susceptibility to smoking-related coronary artery disease: A case-only study. Mutat Res 621: 106–112, 2007pmid:17408703
    OpenUrlCrossRefPubMed
  16. ↵
    1. Moon KS,
    2. Lee HJ,
    3. Hong SH,
    4. Kim HM,
    5. Um JY
    : CYP1A1 and GSTM1/T1 genetic variation in predicting risk for cerebral infarction. J Mol Neurosci 32: 155–159, 2007pmid:17873299
    OpenUrlCrossRefPubMed
  17. ↵
    1. Türkanoğlu A,
    2. Can Demirdöğen B,
    3. Demirkaya S,
    4. Bek S,
    5. Adali O
    : Association analysis of GSTT1, GSTM1 genotype polymorphisms and serum total GST activity with ischemic stroke risk. Neurol Sci 31: 727–734, 2010pmid:20517701
    OpenUrlPubMed
  18. ↵
    1. Carpenter CL,
    2. Yu MC,
    3. London SJ
    : Dietary isothiocyanates, glutathione S-transferase M1 (GSTM1), and lung cancer risk in African Americans and Caucasians from Los Angeles County, California. Nutr Cancer 61: 492–499, 2009pmid:19838921
    OpenUrlCrossRefPubMed
    1. Wang LI,
    2. Giovannucci EL,
    3. Hunter D,
    4. Neuberg D,
    5. Su L,
    6. Christiani DC
    : Dietary intake of cruciferous vegetables, glutathione S-transferase (GST) polymorphisms and lung cancer risk in a Caucasian population. Cancer Causes Control 15: 977–985, 2004pmid:15801482
    OpenUrlCrossRefPubMed
  20. ↵
    1. Zhao B,
    2. Seow A,
    3. Lee EJ,
    4. Poh WT,
    5. Teh M,
    6. Eng P, et al
    .: Dietary isothiocyanates, glutathione S-transferase -M1, -T1 polymorphisms and lung cancer risk among Chinese women in Singapore. Cancer Epidemiol Biomarkers Prev 10: 1063–1067, 2001pmid:11588132
    OpenUrlAbstract/FREE Full Text
    1. London SJ,
    2. Yuan JM,
    3. Chung FL,
    4. Gao YT,
    5. Coetzee GA,
    6. Ross RK, et al
    .: Isothiocyanates, glutathione S-transferase M1 and T1 polymorphisms, and lung-cancer risk: A prospective study of men in Shanghai, China. Lancet 356: 724–729, 2000pmid:11085692
    OpenUrlCrossRefPubMed
  22. ↵
    1. Brennan P,
    2. Hsu CC,
    3. Moullan N,
    4. Szeszenia-Dabrowska N,
    5. Lissowska J,
    6. Zaridze D, et al
    .: Effect of cruciferous vegetables on lung cancer in patients stratified by genetic status: A mendelian randomisation approach. Lancet 366: 1558–1560, 2005pmid:16257343
    OpenUrlCrossRefPubMed
  23. ↵
    1. James D,
    2. Devaraj S,
    3. Bellur P,
    4. Lakkanna S,
    5. Vicini J,
    6. Boddupalli S
    : Novel concepts of broccoli sulforaphanes and disease: Induction of phase II antioxidant and detoxification enzymes by enhanced-glucoraphanin broccoli. Nutr Rev 70: 654–665, 2012pmid:23110644
    OpenUrlCrossRefPubMed
  24. ↵
    1. Boddupalli S,
    2. Mein JR,
    3. Lakkanna S,
    4. James DR
    : Induction of phase 2 antioxidant enzymes by broccoli sulforaphane: Perspectives in maintaining the antioxidant activity of vitamins A, C, and E. Front Genet 3: 7, 2012pmid:22303412
    OpenUrlPubMed
  25. ↵
    1. Gasper AV,
    2. Al-Janobi A,
    3. Smith JA,
    4. Bacon JR,
    5. Fortun P,
    6. Atherton C, et al
    .: Glutathione S-transferase M1 polymorphism and metabolism of sulforaphane from standard and high-glucosinolate broccoli. Am J Clin Nutr 82: 1283–1291, 2005pmid:16332662
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Amin A,
    2. CanGongora M,
    3. Elbarbry F
    : Dietary doses of sulforaphane affect hepatic drug metabolizing enzymes in spontaneously hypertensive rats. Phytother Res 29:1412–1420, 2015
    OpenUrl
  27. ↵
    1. Elbarbry F,
    2. Vermehren-Schmaedick A,
    3. Balkowiec A
    : Modulation of arachidonic acid metabolism in the rat kidney by sulforaphane: Implications for regulation of blood pressure. ISRN Pharmacol 2014: 683508, 2014pmid:24734194
    OpenUrlPubMed
  28. ↵
    1. Le TH,
    2. Fogo AB,
    3. Salzler HR,
    4. Vinogradova T,
    5. Oliverio MI,
    6. Marchuk DA, et al
    .: Modifier locus on mouse chromosome 3 for renal vascular pathology in AT1A receptor-deficiency. Hypertension 43: 445–451, 2004pmid:14718357
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Yang Y,
    2. Parsons KK,
    3. Chi L,
    4. Malakauskas SM,
    5. Le TH
    : Glutathione S-transferase-micro1 regulates vascular smooth muscle cell proliferation, migration, and oxidative stress. Hypertension 54: 1360–1368, 2009pmid:19822795
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. McBride MW,
    2. Carr FJ,
    3. Graham D,
    4. Anderson NH,
    5. Clark JS,
    6. Lee WK, et al
    .: Microarray analysis of rat chromosome 2 congenic strains. Hypertension 41: 847–853, 2003pmid:12624007
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. McBride MW,
    2. Brosnan MJ,
    3. Mathers J,
    4. McLellan LI,
    5. Miller WH,
    6. Graham D, et al
    .: Reduction of Gstm1 expression in the stroke-prone spontaneously hypertension rat contributes to increased oxidative stress. Hypertension 45: 786–792, 2005pmid:15699453
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Salzler HR,
    2. Griffiths R,
    3. Ruiz P,
    4. Chi L,
    5. Frey C,
    6. Marchuk DA, et al
    .: Hypertension and albuminuria in chronic kidney disease mapped to a mouse chromosome 11 locus. Kidney Int 72: 1226–1232, 2007pmid:17851470
    OpenUrlCrossRefPubMed
  33. ↵
    1. Cechova S,
    2. Zeng Q,
    3. Billaud M,
    4. Mutchler S,
    5. Rudy CK,
    6. Straub AC, et al
    .: Loss of collectrin, an angiotensin-converting enzyme 2 homolog, uncouples endothelial nitric oxide synthase and causes hypertension and vascular dysfunction. Circulation 128: 1770–1780, 2013pmid:24048198
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Gigliotti JC,
    2. Huang L,
    3. Ye H,
    4. Bajwa A,
    5. Chattrabhuti K,
    6. Lee S, et al
    .: Ultrasound prevents renal ischemia-reperfusion injury by stimulating the splenic cholinergic anti-inflammatory pathway. J Am Soc Nephrol 24: 1451–1460, 2013pmid:23907510
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Cechova S,
    2. Dong F,
    3. Chan F,
    4. Kelley MJ,
    5. Ruiz P,
    6. Le TH
    : MYH9 E1841K mutation augments proteinuria and podocyte injury and migration. J Am Soc Nephrol 29: 155–167, 2018pmid:28993503
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Fawzi WW,
    2. Rifas-Shiman SL,
    3. Rich-Edwards JW,
    4. Willett WC,
    5. Gillman MW
    : Calibration of a semi-quantitative food frequency questionnaire in early pregnancy. Ann Epidemiol 14: 754–762, 2004pmid:15519898
    OpenUrlCrossRefPubMed
  37. ↵
    1. Tang L,
    2. Paonessa JD,
    3. Zhang Y,
    4. Ambrosone CB,
    5. McCann SE
    : Total isothiocyanate yield from raw cruciferous vegetables commonly consumed in the United States. J Funct Foods 5: 1996–2001, 2013pmid:24443655
    OpenUrlCrossRefPubMed
  38. ↵
    1. Rebholz CM,
    2. Coresh J,
    3. Ballew SH,
    4. McMahon B,
    5. Whelton SP,
    6. Selvin E, et al
    .: Kidney failure and ESRD in the Atherosclerosis Risk in Communities (ARIC) study: Comparing ascertainment of treated and untreated kidney failure in a cohort study. Am J Kidney Dis 66: 231–239, 2015pmid:25773483
    OpenUrlCrossRefPubMed
  39. ↵
    1. Schlöndorff D,
    2. Banas B
    : The mesangial cell revisited: No cell is an island. J Am Soc Nephrol 20: 1179–1187, 2009pmid:19470685
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Kriz W,
    2. Shirato I,
    3. Nagata M,
    4. LeHir M,
    5. Lemley KV
    : The podocyte’s response to stress: The enigma of foot process effacement. Am J Physiol Renal Physiol 304: F333–F347, 2013pmid:23235479
    OpenUrlCrossRefPubMed
  41. ↵
    1. Schordan S,
    2. Schordan E,
    3. Endlich K,
    4. Endlich N
    : AlphaV-integrins mediate the mechanoprotective action of osteopontin in podocytes. Am J Physiol Renal Physiol 300: F119–F132, 2011pmid:21048023
    OpenUrlCrossRefPubMed
  42. ↵
    1. Taniyama Y,
    2. Griendling KK
    : Reactive oxygen species in the vasculature: Molecular and cellular mechanisms. Hypertension 42: 1075–1081, 2003pmid:14581295
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Just A,
    2. Olson AJ,
    3. Whitten CL,
    4. Arendshorst WJ
    : Superoxide mediates acute renal vasoconstriction produced by angiotensin II and catecholamines by a mechanism independent of nitric oxide. Am J Physiol Heart Circ Physiol 292: H83–H92, 2007pmid:16951043
    OpenUrlCrossRefPubMed
  44. ↵
    1. Trott DW,
    2. Harrison DG
    : The immune system in hypertension. Adv Physiol Educ 38: 20–24, 2014pmid:24585465
    OpenUrlCrossRefPubMed
  45. ↵
    1. Olson E,
    2. Pravenec M,
    3. Landa V,
    4. Koh-Tan HHC,
    5. Dominiczak AF,
    6. McBride MW, et al
    .: Transgenic overexpression of glutathione S-transferase μ-type 1 reduces hypertension and oxidative stress in the stroke-prone spontaneously hypertensive rat. J Hypertens 37: 985–996, 2019pmid:30308595
    OpenUrlCrossRefPubMed
  46. ↵
    1. Whaley-Connell AT,
    2. Sowers JR,
    3. Stevens LA,
    4. McFarlane SI,
    5. Shlipak MG,
    6. Norris KC, et al. Kidney Early Evaluation Program Investigators
    : CKD in the United States: Kidney Early Evaluation Program (KEEP) and National Health and Nutrition Examination Survey (NHANES) 1999-2004. Am J Kidney Dis 51[Suppl 2]: S13–S20, 2008pmid:18359403
    OpenUrlCrossRefPubMed
  47. ↵
    1. Bernardini S,
    2. Hirvonen A,
    3. Järventaus H,
    4. Norppa H
    : Trans-stilbene oxide-induced sister chromatid exchange in cultured human lymphocytes: Influence of GSTM1 and GSTT1 genotypes. Mutagenesis 16: 277–281, 2001pmid:11320155
    OpenUrlCrossRefPubMed
  48. ↵
    1. Berhane K,
    2. Widersten M,
    3. Engström A,
    4. Kozarich JW,
    5. Mannervik B
    : Detoxication of base propenals and other alpha, beta-unsaturated aldehyde products of radical reactions and lipid peroxidation by human glutathione transferases. Proc Natl Acad Sci U S A 91: 1480–1484, 1994pmid:8108434
    OpenUrlAbstract/FREE Full Text
    1. Hubatsch I,
    2. Ridderström M,
    3. Mannervik B
    : Human glutathione transferase A4-4: An alpha class enzyme with high catalytic efficiency in the conjugation of 4-hydroxynonenal and other genotoxic products of lipid peroxidation. Biochem J 330: 175–179, 1998pmid:9461507
    OpenUrlAbstract/FREE Full Text
  50. ↵
    1. Paumi CM,
    2. Smitherman PK,
    3. Townsend AJ,
    4. Morrow CS
    : Glutathione S-transferases (GSTs) inhibit transcriptional activation by the peroxisomal proliferator-activated receptor gamma (PPAR gamma) ligand, 15-deoxy-delta 12,14prostaglandin J2 (15-d-PGJ2). Biochemistry 43: 2345–2352, 2004pmid:14979731
    OpenUrlCrossRefPubMed
  51. ↵
    1. Kistler AD,
    2. Altintas MM,
    3. Reiser J
    : Podocyte GTPases regulate kidney filter dynamics. Kidney Int 81: 1053–1055, 2012pmid:22584591
    OpenUrlCrossRefPubMed
  52. ↵
    1. Alexis NE,
    2. Zhou H,
    3. Lay JC,
    4. Harris B,
    5. Hernandez ML,
    6. Lu TS, et al
    . The glutathione-S-transferase Mu 1 null genotype modulates ozone-induced airway inflammation in human subjects. J Allergy Clin Immunol 124: 1222–1228.e5, 2009
    OpenUrlCrossRefPubMed
  53. ↵
    1. Miyake M,
    2. Goodison S,
    3. Urquidi V,
    4. Gomes Giacoia E,
    5. Rosser CJ
    : Expression of CXCL1 in human endothelial cells induces angiogenesis through the CXCR2 receptor and the ERK1/2 and EGF pathways. Lab Invest 93: 768–778, 2013pmid:23732813
    OpenUrlCrossRefPubMed
  54. ↵
    1. Hung CC,
    2. Chang CT,
    3. Chen KH,
    4. Tian YC,
    5. Wu MS,
    6. Pan MJ, et al
    .: Upregulation of chemokine CXCL1/KC by leptospiral membrane lipoprotein preparation in renal tubule epithelial cells. Kidney Int 69: 1814–1822, 2006pmid:16625148
    OpenUrlCrossRefPubMed
  55. ↵
    1. Bodonyi-Kovacs G,
    2. Ma JZ,
    3. Chang J,
    4. Lipkowitz MS,
    5. Kopp JB,
    6. Winkler CA, et al
    .: Combined effects of GSTM1 null allele and APOL1 renal risk alleles in CKD progression in the African American Study of Kidney Disease and Hypertension trial. J Am Soc Nephrol 27: 3140–3152, 2016pmid:26940095
    OpenUrlAbstract/FREE Full Text
  56. ↵
    1. Reeves-Daniel AM,
    2. DePalma JA,
    3. Bleyer AJ,
    4. Rocco MV,
    5. Murea M,
    6. Adams PL, et al
    .: The APOL1 gene and allograft survival after kidney transplantation. Am J Transplant 11: 1025–1030, 2011pmid:21486385
    OpenUrlCrossRefPubMed
Back to top

In this issue

View Selected Citations (0)
Print
Download PDF
Email Article
Citation Tools
Request Permissions
GSTM1 Deletion Exaggerates Kidney Injury in Experimental Mouse Models and Confers the Protective Effect of Cruciferous Vegetables in Mice and Humans
Joseph C. Gigliotti, Adrienne Tin, Shirin Pourafshar, Sylvia Cechova, Yves T. Wang, Sun-sang J. Sung, Gabor Bodonyi-Kovacs, Janet V. Cross, Guang Yang, Nhu Nguyen, Fang Chan, Casey Rebholz, Bing Yu, Megan L. Grove, Morgan E. Grams, Anna Köttgen, Robert Scharpf, Phillip Ruiz, Eric Boerwinkle, Josef Coresh, Thu H. Le
JASN Jan 2020, 31 (1) 102-116; DOI: 10.1681/ASN.2019050449
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Keywords