Phosphoglycerate kinase 1 contributes to diabetic kidney disease through enzyme-dependent and independent manners

Abstract

Diabetic kidney disease (DKD) is characterized by abnormal metabolic profiles. Metabolomics reveals increased serum levels of 3-phosphoglycerate (3-PG) in DKD patients. The protein expression of phosphoglycerate kinase 1 (PGK1), a key rate-limiting enzyme for 3-PG synthesis, is concomitantly upregulated in DKD patients and mice. The development of DKD is significantly mitigated by renal tubular epithelial cell-specific knockout of PGK1 and robustly worsened by PGK1 overexpression. Mechanistically, PGK1-dependent enzymatic production of 3-PG facilitates DKD through inhibiting GPX1 to activate the NLRP3 inflammasome. PGK1 promotes UNC5CL-mediated inflammation by binding to aldehyde dehydrogenase-1 L1 (Aldh1l1) through its non-enzymatic activity. The transcription factor paired box protein 5 (PAX5) mediates the upregulation of PGK1 in DKD. High-throughput screening reveals that C-16 from ChemDiv, the natural product lirinidine, and the Food and Drug Administration (FDA)-approved oxantel pamoate are potent PGK1 antagonists and efficaciously prevent DKD. Overall, blocking PGK1 may be a promising avenue for DKD management.

Keywords: 3-phosphoglycerate; NLRP3; PGK1; diabetic kidney disease; drug screening; inflammation; metabolomics; oxidative stress.

Graphical abstract

Figure 1

Metabolomics revealed highly expressed 3-phosphoglycerate and the rate-limiting biosynthetic enzyme (PGK1) in diabetic kidney disease patients and mice (A) Heatmap of the serum differential metabolites in healthy control (HC) vs. DKD patients. (B) Pearson’s analysis of the correlations between the relative abundance of differential metabolites and urinary albumin creatinine ratio (UACR). (C) Circulating 3-PG contents from different groups from Affiliated Hospital of Nanjing University of Chinese Medicine (upper) and Nanjing Hospital of Chinese Medicine Affiliated to Nanjing University of Chinese Medicine (lower). (D) Representative immunohistochemical micrographs of PGK1 (n = 5). Scale bar, 100 μm. (E) The mRNA level of PGK1 of renal tubules from the GEO database (GSE30122), in which the eGFR value of DKD patients was 25–35 mL/min/1.73 m². (F) Differentially expressed proteins of kidney tissues of mouse model for T1D-DKD (n = 5) vs. control (n = 4). (G) Differentially expressed proteins of kidney tissues of mouse model for T2D-DKD (n = 5) vs. control (n = 5). (H) PGK1 protein expressions of mouse models for T1D-DKD and T2D-DKD (n = 4). (I) PGK1 activity in T1D-DKD mice (n = 6) vs. control (n = 6) and T2D-DKD mice (n = 6) vs. control (n = 6). (J) Tabula Muris database showing the distribution model of PGK1 in different cells within the kidneys. Data are represented as mean ± SD. Statistical significance was determined by unpaired Student’s t test or ANOVA.∗p < 0.05 vs. healthy controls (HCs) or control (Con). T2DM, type 2 diabetes mellitus; DKD-E, early-stage diabetic kidney disease; DKD-A, advanced-stage diabetic kidney disease. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 vs. indicated group. See also Figure S1 and Tables S1, S2, S3, S4, and S5.

Figure 2

Effects of 3-PG on the renal functions of T2D-DKD mice (A) Serum BUN levels (n = 6). (B) Scr levels (n = 6). (C) Serum cystatin C levels (n = 6). (D) The relative quantitative analysis of glomerular areas (n = 6). (E) Percentage of renal fibrosis (n = 6). (F) Representative images of H&E staining (n = 6). (G) Representative images of sirius red staining (n = 6). (H and I) Representative images and relative analysis of TUNEL staining (n = 6). (J and K) Representative images and relative analysis of dihydroethidium (DHE) staining (n = 6). (L and M) Representative images and relative analysis of F4/80 staining (n = 6). (N) 3-PG contents in kidneys (n = 6). Scale bar, 200 μm. Data are represented as mean ± SD. Statistical significance was determined by ANOVA.∗p < 0.05 vs. control (Con), †p < 0.05 vs. T2D. See also Figure S2 and Table S6.

Figure 3

3-PG downregulated GPX1 to activate NLRP3 inflammasome in DKD (A) Heatmap showing the differentially expressed proteins in HK-2 cells (n = 3). ∗p adjust < 0.05; ∗∗p adjust < 0.01. (B) The enriched signaling pathways by analyzing differentially expressed proteins. (C) Representative blots and quantitative analysis of GPX1 in HK-2 cells and kidneys from T1D-DKD and T2D-DKD mice (n = 4). (D) Representative blots and quantitative analysis of GPX1, NLRP3, ASC, caspase-1, and IL-1β (n = 3). (E) Overexpression of GPX1 reversed HG-induced NLRP3 inflammasome activation (n = 3). Data are represented as mean ± SD. Statistical significance was determined by unpaired Student’s t test or ANOVA.∗p < 0.05 vs. NG or control (Con), †p < 0.05 vs. vehicle (Veh). See also Figure S3 and Table S7.

Figure 4

Effects of specific knockout of PGK1 in renal tubular epithelial cells on the renal functions of T2D-DKD mice (A) Serum BUN levels (n = 6). (B) Scr levels (n = 6). (C) Serum cystatin C levels (n = 6). (D) The relative quantitative analysis of glomerular areas (n = 6). (E) Percentage of renal fibrosis (n = 6). (F) Representative images of H&E staining (n = 6). (G) Representative images of sirius red staining (n = 6). (H and I) Representative images and relative analysis of TUNEL staining (n = 6). (J and K) Representative images and relative analysis of dihydroethidium (DHE) staining (n = 6). (L and M) Representative images and relative analysis of F4/80 staining (n = 6). (N) Renal 3-PG contents (n = 6). (O) Representative blots and quantitative analysis of PGK1 (n = 4). Scale bar, 200 μm. Data are represented as mean ± SD. Statistical significance was determined by ANOVA.∗p < 0.05 vs. control (Con), †p < 0.05 vs. PGK1^flox/flox. See also Figure S4 and Table S6.

Figure 5

PGK1 activated UNC5CL-mediated inflammatory response in DKD (A) Heatmap showing the differentially expressed proteins in HK-2 cells (n = 3). (B) Volcano plot showing the differentially expressed proteins in HK-2 cells. (C and D) Representative blots and quantitative analysis of GPX1 (n = 4). (E) Effects of silencing UNC5CL on the protein expression of collagen I, 12-Lox, and IL-1β (n = 3). (F) Effects of silencing PGK1 on the protein expression of UNC5CL, IκBα, P-IκBα, P-NF-κB, and P-JNK (n = 3). (G) Effects of silencing PGK1 on the mRNA levels of CXCL1, CXCL2, CXCL3, IL8, CCL20, TNFAIP3, PLA2G4C, and TNFRSF9 (n = 4). Data are represented as mean ± SD. Statistical significance was determined by unpaired Student’s t test or ANOVA. ∗p < 0.05 vs. NG, control (Con) or HG + Con small interfering RNA (siRNA), †p < 0.05 vs. Control (Con) siRNA. See also Figure S5.

Figure 6

Aldh1l1 bridges the link between PGK1 and UNC5CL in the development of DKD (A) Aldh1l1 was identified to interact with PGK1. (B) Representative blots and quantitative analysis of Aldh1l1 (n = 4). (C–E) CoIP results showing the complex formation of PGK1/UNC5CL/Aldh1l1 in NG- or HG-incubated HK-2 cells (n = 4). (F–H) CoIP results showing the complex formation of PGK1/UNC5CL/Aldh1l1 in kidneys from control and T1D-DKD mice (n = 4). (I–K) CoIP results showing the complex formation of PGK1/UNC5CL/Aldh1l1 in kidneys from control and T2D-DKD mice (n = 4). (L) Effects of Aldh1l1 small interfering RNA (siRNA) on the complex formation of PGK1/UNC5CL (n = 4). (M and N) Effects of Aldh1l1 siRNA on the impact of PGK1 overexpression (OE) on the UNC5CL-mediated inflammation (n = 4). Data are represented as mean ± SD. Statistical significance was determined by unpaired Student’s t test.∗p < 0.05 vs. NG or control (Con). See also Figures S5 and S6.

Figure 7

Screening of potential PGK1 inhibitors that ameliorate DKD (A–C) Molecular dynamics simulation and molecular docking of C-16, lirinidine, oxantel pamoate, and PGK1. (D) Representative images of H&E staining (n = 6). (E) Representative images of sirius red staining (n = 6). (F and I) Representative images and relative analysis of TUNEL staining (n = 6). (G and J) Representative images and relative analysis of dihydroethidium (DHE) staining (n = 6). (H and K) Representative images and relative analysis of F4/80 staining (n = 6). (L) Diagram interpreting the function and mechanism of the PGK1/3-PG axis in DKD. Scale bar, 200 or 100 μm. Data are represented as mean ± SD. Statistical significance was determined by ANOVA.∗p < 0.05 vs. control (Con), †p < 0.05 vs. T2D. See also Figure S7 and Table S6.