Nrf2 deficiency deteriorates diabetic kidney disease in Akita model mice

Abstract

Oxidative stress is an essential component in the progression of diabetic kidney disease (DKD), and the transcription factor NF-E2-related factor-2 (Nrf2) plays critical roles in protecting the body against oxidative stress. To clarify the roles of Nrf2 in protecting against DKD, in this study we prepared compound mutant mice with diabetes and loss of antioxidative defense. Specifically, we prepared compound Ins2^Akita/+ (Akita) and Nrf2 knockout (Akita::Nrf2^-/-) or Akita and Nrf2 induction (Akita::Keap1^FA/FA) mutant mice. Eighteen-week-old Akita::Nrf2^-/- mice showed more severe diabetic symptoms than Akita mice. In the Akita::Nrf2^-/- mouse kidneys, the glomeruli showed distended capillary loops, suggesting enhanced mesangiolysis. Distal tubules showed dilation and an increase in 8-hydroxydeoxyguanosine-positive staining. In the Akita::Nrf2^-/- mouse kidneys, the expression of glutathione (GSH) synthesis-related genes was decreased, and the actual GSH level was decreased in matrix-assisted laser desorption/ionization mass spectrometry imaging analysis. Akita::Nrf2^-/- mice exhibited severe inflammation and enhancement of infiltrated macrophages in the kidney. To further examine the progression of DKD, we compared forty-week-old Akita mouse kidney compounds with Nrf2-knockout or Nrf2 mildly induced (Akita::Keap1^FA/FA) mice. Nrf2-knockout Akita (Akita::Nrf2^-/-) mice displayed severe medullary cast formation, but the formation was ameliorated in Akita::Keap1^FA/FA mice. Moreover, in Akita::Keap1^FA/FA mice, tubule injury and inflammation-related gene expression were significantly suppressed, which was evident in Akita::Nrf2^-/- mouse kidneys. These results demonstrate that Nrf2 contributes to the protection of the kidneys against DKD by suppressing oxidative stress and inflammation.

Keywords: Diabetic kidney disease; Glutathione; Inflammation; MALDI-MSI; Nrf2; Oxidative stress.

Fig. 1

Expression of NQO1 in 18-week-old Akita::Nrf2^−/− mice. (A) Schematic presentation of WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. Eighteen-week-old male mice were used in this study. (B) Expression of Nqo1 mRNA in the kidneys of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. The expression level was normalized to Hprt, and that in WT mice was set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. (C) Immunochemistry of NQO1 in kidney sections from 18-week-old male mice. Upper images, low-magnification images including cortex and medulla. The two middle images correspond to lower- and higher-magnification images of the cortex. Lower images show the outer medulla. Cortex images are demonstrated by red squares. Bars, 500 μm (Cortex & Medulla), 200 μm (Cortex Low and Outer medulla) and 50 μm (Cortex High). ***P < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Oxidative stress and GSH levels in the kidneys of Akita::Nrf2^−/− mice. (A–C) mRNA expression levels of GSH synthesis-related genes in the kidneys of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. The expression levels of the Gclm (A), Gclc (B) and Gsr (C) genes were normalized to Hprt expression and quantified as the fold increase relative to those of WT mice, which were set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001. (D) MALDI-MSI for GSH in kidney sections from 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. GSH was conjugated with NEM to generate GSH-NEM on the kidney sections. The MS/MS signals of GHS-NEM were detected as m/z 304.1. Optical images (upper panel) and MS/MS images of GSH-NEM signals (lower panel) of kidney sections. (E) Immunohistochemical staining of the oxidative stress marker 8-OHdG in the glomeruli (upper panels) and tubules (lower panel) of kidney sections. Images from each group of 18-week-old male mice are shown. Bars, 50 μm (upper panels) and 100 μm (lower panels). *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 3

Severe diabetes symptoms of Akita::Nrf2^−/− mice. Body weights (A) and blood glucose levels (B) of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. (C–L) Diabetes symptoms and urine osmolality of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice observed in metabolic cages. Representative appearance of 24-h urine output (C), 24-h urine volume (D), water consumption (E), food consumption (F) and urine osmolality (G), left kidney weight normalized by body weight (H), systolic blood pressure (I), fractional excretion of glucose (J), creatinine clearance (K) and urine albumin/creatinine ratio in the 24-h urine (L). The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test (A, B, D-L). *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 4

Plasma levels of uremic toxins and oxidative stress-related metabolites. Plasma levels of creatinine (A), trimethylamine N-oxide (TMAO, B), indole-3-acetic acid (IAA, C) and methionine sulfoxide (D) in 18-week-old Akita::Nrf2^−/− mice. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 5

Nrf2 deficiency sensitizes Akita mice to severe renal inflammation. (A–C) Expression of the macrophage marker F4/80 in the kidneys of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. Immunochemistry images of F4/80 (A) and quantitative analysis of the F4/80-positive area within the cortex (B) and medulla (C). (D and E) Expression of another macrophage marker, Mac-2, in the kidneys of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. Immunofluorescence of Mac-2 (D) and quantification of Mac-2-positive cells (E) were performed to assess glomerular infiltrating macrophages. Yellow arrows within the glomerular region (white dashed line) indicate Mac-2-positive cells. (F and G) mRNA expression levels of the chemokine Ccl2 (F) and the proinflammatory cytokine Il6 (G) in the kidneys of 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. The expression level was normalized to Hprt expression. Relative mRNA levels were quantified as the fold increase relative to those of WT mice, which were set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05 and ***P < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Development of fibrosis in the kidneys of Akita::Nrf2^−/− mice. (A) Quantification of cortex thickness in Masson's trichrome-stained specimens from 18-week-old male WT, Nrf2^−/−, Akita, and Akita::Nrf2^−/− mice. (B) Representative images of Masson's trichrome-stained specimens from 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. Bars, 50 μm. (C–G) Expression levels of fibrogenic Col1a1 (C), Col3a1 (D), Col4a1 (E), Tgfb1 (F) and Fn1 (G) mRNAs. Expression levels were normalized by Hprt and quantified as the fold increase relative to those of WT mice, which were set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. **P < 0.01.

Fig. 7

Glomerular and tubular abnormalities in Akita::Nrf2^−/− mouse kidneys. (A, B) PAS staining of kidney specimens from 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice (A). Bars, 20 μm. The quantified capillary area is shown in (B). (C) Images of HE-staining of WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mouse kidneys. Bars, 100 μm (upper panels) and 50 μm (lower panels). (D, E) The expression levels of Havcr1 (KIM-1, D) and Lcn2 (NGAL, E) were detected in the kidney by qPCR. mRNA levels were normalized by Hprt and expressed as the fold changes relative to those of WT mice, which were set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. **P < 0.01.

Fig. 8

Expression of metabolic Nrf2 target genes in the kidneys of 18-week-old Akita::Nrf2^−/− mice. (A) mRNA expression levels of the pentose phosphate pathway-related genes G6pdx (A) and Pgd (B), the glycogen-related gene Gbe1 (C), the gluconeogenesis-related gene G6pc (D) and the glucose reabsorption gene Slc5a2 (SGLT2, E) in the kidneys of WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice. The expression level was normalized to that of Hprt and quantified as the fold increase relative to that in the WT mice, which was set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 9

Metabolome analyses of Akita::Nrf2^−/− mice. (A–D) Plasma levels of triacylglycerol (TG, A), sphingomyelin (SM, B), cholesteryl ester (CE, C) and lysophosphatidylcholine (LysoPC, D). Metabolomics analyses were executed by using plasma samples from 18-week-old male WT, Nrf2^−/−, Akita and Akita::Nrf2^−/− mice and an MxP Quant 500 Kit. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001. (E) Principal component analyses (PCAs) of the plasma metabolome. The 261 qualified metabolites were determined, and PCA of the plasma levels of these metabolites was performed. (F and G) Volcano plots. The X-axis is the log₂ fold change (FC) of plasma levels, Nrf2^−/− compared to WT mice (G) and Akita::Nrf2^−/− mice compared to Akita mice (H). The Y-axis is the negative Log₁₀ of the two-tailed test P value. Vertical dotted lines denote a linear fold change of one. Horizonal dotted lines indicate P = 0.05.

Fig. 10

Metabolites showing altered plasma levels in Akita::Nrf2^−/− mice. (A and B) Numbers of metabolites whose levels in plasma are significantly increased (A) or decreased (B). Venn diagram of metabolites whose levels were altered in Nrf2^−/− mice compared to WT mice (green) and in Akita::Nrf2^−/− mice compared to Akita mice (purple). (C–M) Representative metabolites whose levels are changed in 18-week-old Akita::Nrf2^−/− mice. These metabolites include TG(16:0_34:2) (C), TG(18:2_36:3) (D), α-aminoadipic acid (E), glycine (F), asparagine (G), leucine (H), lysoPC a C18:1 (I), SM C24:1 (J), SM C16:1 (K), CE(20:4) (L) and CE(18:2) (M). The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 11

Gene expression in the kidneys of 18- and 40-week-old Akita and Akita::Nrf2^−/− mice. The mRNA expression levels of the Nrf2 target gene Nqo1 (A), the inflammation-related gene Ccl2 (MCP-1, B), the fibrosis-related gene Col4a1 (C), and the kidney damage-related gene Havcr1 (KIM-1, D) in the kidneys of WT, Akita and Akita::Nrf2^−/− mice were examined. The expression level was normalized to that of Hprt and quantified as the fold increase relative to that in the WT mice, which was set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 12

Forty-week-old Akita::Keap1^FA/FA and Akita::Nrf2^−/− mice. (A) Schematic presentation of 40-week-old WT, Akita, Akita::Nrf2^−/− and Akita::Keap1^FA/FA mice. (B and C) Blood glucose levels (B) and kidney weights normalized by body weight (C) for 40-week-old male WT, Akita, Akita::Nrf2^−/− and Akita::Keap1^FA/FA mice. (D–I) mRNA expression levels of the genes Nqo1 (D), Gclc (E), Ccl2 (MCP-1, F), Col4a1 (G), Col1a1 (H) and Col3a1 (I) in the kidneys of 40-week-old male WT, Akita, Akita::Nrf2^−/− and Akita::Keap1^FA/FA mice. The expression level was normalized to Hprt expression and quantified as the fold increase relative to WT mice, which was set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 13

Tubulointerstitial changes in Akita mouse kidneys and Nrf2. (A) PAS-stained images of renal sections of 40-week-old male WT, Akita, Akita::Nrf2^−/− and Akita::Keap1^FA/FA mice. Yellow triangles indicate the abnormal tubules within the cortex (upper panel), and black triangles indicate the medullary cast (lower panel). Bars, 200 (upper) and 400 (lower) μm. (B) Numbers of quantified cast formations in each longitudinal section. (C) mRNA expression level of the Havcr1 (KIM-1) gene in the kidneys of 40-week-old male WT, Akita, Akita::Nrf2^−/−and Akita::Keap1^FA/FA mice. The expression level was normalized to Hprt expression and quantified as the fold increase relative to WT mice, which was set as 1. The results are presented as the mean ± SD. Statistical analyses were performed using ANOVA followed by Fisher's LSD post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 14

Relationship between Nrf2 levels and diabetic kidney disease symptoms in Akita diabetic model mice. Akita mice display mild oxidative stress, hyperfiltration, cast formation and modest mesangial proliferation but no obvious inflammation or fibrosis. Akita::Nrf2^−/− mice exhibit severe oxidative stress, inflammation, fibrosis, thinning renal cortex and modest mesangial proliferation. Akita::Keap1^FA/FA mice display suppressed cast formation.