BDH1-mediated βOHB metabolism ameliorates diabetic kidney disease by activation of NRF2-mediated antioxidative pathway

BDH1 is downregulated in diabetic kidney and high glucose- or palmitic acid-treated HK-2 cells

DKD represents a major microvascular complication in patients with diabetes and is the leading cause of chronic kidney disease and end-stage renal disease [2–4]. To gain a comprehensive understanding of potential DKD regulators, we performed RNA-seq analysis by comparing the gene expression in the kidneys of db/db and m/m mice at the age of 20 weeks when db/db mice exhibited an apparent increase in the urinary albumin-to-creatinine ratio (ACR) (Supplementary Figure 1). The volcano plot in Figure 1A shows that 165 upregulated differentially expressed genes (DEGs) and 202 downregulated DEGs were identified in the kidneys of db/db mice (detailed information on DEGs is provided in the Supplementary Table 1). Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis shows that the “synthesis and degradation of ketone bodies” pathway is significantly downregulated (Figure 1B, 1C). Given that KD has been reported as an effective treatment for diabetes [29, 30], we presumed that the downregulation of the synthesis and degradation of ketone body pathways participates in the pathogenesis of DKD. Consistently, qRT-PCR analysis confirmed that the expression of Bdh1, Oxct1, Acat1, and Hmgcs1 was altered (Figure 1D). Among these four pathway members, Bdh1 has been reported to protect the heart from heart failure in a transverse aortic constriction mouse model [27]. Thus, to identify whether Bdh1 is involved in the pathogenesis of DKD, we further identified the protein level of BDH1. As shown in Figure 1E, the protein level of BDH1 in the kidneys of db/db mice is significantly lower than that in the kidneys of m/m mice. Moreover, the decrease in BDH1 expression was confirmed using immunohistochemistry and immunofluorescence (IF) analysis (Figure 1F). Consistent with the DKD mouse model, we observed the downregulation of BDH1 in the renal tissues of diabetic patients with kidney disease using immunohistochemistry and IF staining (Figure 1G). The characteristics of the control subjects and DKD patients are summarized in Supplementary Table 2.

Figure 1. Expression of BDH1 is downregulated in diabetic kidney and HG- or PA-treated HK-2 cells. (A) Volcano plot showing differentially expressed genes (DEGs) (red, upregulated genes; green, downregulated genes) in the kidneys of m/m and db/db mice (n = 3 mice per group). (B) KEGG enrichment analysis showing the top 6 downregulated pathways. (C) Four DEGs involved in ketone body synthesis and degradation pathways. (D) qRT-PCR analysis showing the mRNA levels of Acat1, Bdh1, Oxct1, and Hmgcs1 in the kidneys of m/m and db/db mice (n = 3 mice per group). WB (E), IHC, and IF (F) showing the protein level of BDH1 in the kidneys of m/m and db/db mice. IHC and IF (G) showing the protein level of BDH1 in the kidneys of normal subjects (n = 9) and patients with DKD (n = 8). (H) qRT-PCR showing the mRNA level of Bdh1 in HK-2 cells treated with HG or PA. Representative WB images showing the protein level of BDH1 in HK-2 cells stimulated with HG (I) or PA (J). Representative IF images showing the protein level of BDH1 in HK-2 cells stimulated with HG (K) or PA (L). All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Bar: 50 and 25 μm in F, and 100 and 50 μm in G, 20 μm in K and L. Abbreviations: BDH1: β-hydroxybutyrate dehydrogenase 1; NG: normal glucose; HG: high glucose; PA: palmitic acid; MA: mannitol; DEGs: differentially expressed genes; WB: western blot; IHC: immunohistochemistry; IF: immunofluorescence; DKD: diabetic kidney disease. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; Abbreviation: ns: no significance.

Hyperglycemia and hyperlipidemia are the two most evident characteristics of type 2 diabetes [31]. Therefore, we established high glucose (HG)-induced glucotoxicity and palmitic acid (PA)-induced lipotoxicity cell models using HK-2 cells to evaluate the effects of HG or PA on BDH1 expression. As expected, both the mRNA and protein levels of BDH1 were reduced by the HG and PA treatments in HK-2 cells (Figure 1H–1J), which was confirmed using IF analysis (Figure 1K, 1L). These results indicated that decreased BDH1 expression is related to the pathogenesis of DKD.

Next, we investigated the mechanism underlying the decreased BDH1 expression in DKD mice and HG- or PA-treated HK-2 cells. Given that the mRNA expression of BDH1 could be transcriptionally regulated by MTA2-triggered R-loop in hepatocellular carcinoma stem cells [32], we hypothesized that HG- and PA-induced BDH1 reduction might be mediated by MTA2. To confirm this, we measured MTA2 expression levels in a diabetic mouse model and in glucotoxicity and lipotoxicity cell models. We identified that MTA2 expression was up-regulated in the kidneys of db/db mice (Supplementary Figure 2A) and HG- (Supplementary Figure 2B) and PA-treated HK-2 cells (Supplementary Figure 2C), indicating that the reduction in BDH1 in DKD mouse models or HK-2 cells treated with HG or PA may be related to MTA2 reduction.

AAV9-mediated BDH1 renal expression alleviates the progression of DKD

Based on the pronounced ability of BDH1 to inhibit ROS overproduction and inflammation, we explored the therapeutic efficacy of BDH1 in db/db mice. AAV9 is a gene delivery system that is relatively specific for renal expression (Supplementary Figure 3A), therefore, we used AAV9 as a gene delivery tool for renal BDH1 expression. The experimental strategy is illustrated in Figure 2A. Compared with vector-injected db/db mice, AAV9-Bdh1-GFP-injected db/db mice showed increased blood levels of βOHB (Supplementary Figure 3B) and unchanged plasma levels of total cholesterol (Supplementary Figure 3C) and total triglycerides (Supplementary Figure 3D). Eleven weeks after injection of the control or Bdh1-encoding virus, we performed an assay of ACR, which is the most important indicator of kidney function. Although the BDH1 renal expression did not affect the body weight, fasting blood glucose (Supplementary Figure 4A, 4B), and kidney-to-body weight ratio (Supplementary Table 3), we observed a significantly lower ACR in AAV9-Bdh1-GFP-injected mice than that in vector-injected mice (Figure 2B). To confirm whether Bdh1 was effectively expressed in the kidneys of mice based on the AAV9 vector, we measured the fluorescence intensity of GFP, which was co-expressed with Bdh1. AAV9-encoding Bdh1 was successfully delivered to the kidneys 11 weeks after injection (Figure 2C). As expected, we observed increased BDH1 expression in the kidneys of AAV9-Bdh1-GFP-injected mice compared with vector-injected mice (Figure 2D). In further histological analyses, AAV9-Bdh1-GFP-injected db/db mice showed a normal glomerular morphology while vector-injected mice presented glomerular hypertrophy (Figure 2E). In addition, DKD pathology-related fibrosis, inflammation, and apoptosis were substantially reduced by AAV9-Bdh1-GFP injection (Figure 2E–2G). Collectively, these findings strongly support the promising application of BDH1 as a therapeutic target for DKD.

Figure 2. AAV9-mediated BDH1 ectopic expression in kidneys alleviates the progression of DKD. (A) Schematic diagram illustrating the animal experimental design. During the experiment, m/m mice were fed a control diet (CD), while db/db mice were fed an HFD. Six-week-old male db/db mice were randomized to receive tail vein injection of AAV9-Bdh1-GFP (3.40E+12 vg/mL) or vector (1.90E+13 vg/mL). Mice were euthanized at 11 weeks after AAV9 injection. (B) Urinary ACR values of mice in the indicated groups (n = 5 in m/m group, n = 4 in vector, and AAV-Bdh1-GFP injected db/db group). (C) Representative renal fluorescent images of the mice 11 weeks after caudal vein delivery of AAV-Bdh1-GFP revealing GFP expression in the kidneys. (D) Representative WB image showing the protein level of BDH1 in the kidneys of indicated groups. (E) Representative photomicrographs of H&E, Masson, IHC (IL-1β), and TUNEL staining showing the pathological changes in the kidneys of indicated groups. (F) Quantification of the fibrosis area in the kidneys of indicated groups (n = 4 per group). (G) Quantification of apoptosis-positive cells in the kidneys of indicated groups (n = 4 per group). All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Bar: 100 μm in C and E. Abbreviations: CD: control diet; HFD: high fat diet; ACR: albumin-to-creatinine ratio; AAV9: adeno-associated virus 9; BDH1: β-hydroxybutyrate dehydrogenase 1; DKD: diabetic kidney disease; WB: western blot; H&E: hematoxylin and eosin; IHC: immunohistochemistry. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

BDH1-mediated βOHB metabolism reverses HG- or PA-induced ROS overproduction and inflammation

To examine whether BDH1 downregulation contributes to HG- or PA-induced cell injury, we knocked down BDH1 in HK-2 cells (Figure 3A). Given that increased ROS plays a central role in the pathogenesis of diabetic microvascular complications including DKD [33], and that ROS overproduction is related to inflammation [34], we next identified the ROS levels and observed a significant increase in ROS in HK-2 cells transfected with Bdh1 siRNA (Figure 3B). In addition, BDH1 knockdown elevated the protein level of the activated pro-inflammatory factor cleaved IL-1β (Figure 3C) and secretory IL-1β and IL-18 (Figure 3D, 3E). Collectively, these results suggest that BDH1 deficiency mediates HG- or PA-induced cell injury through the loss of anti-ROS function.

Figure 3. BDH1-mediated βOHB metabolism reverses HG or PA-induced ROS overproduction and inflammation. (A) Representative WB image showing the protein level of BDH1 in HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (B) DCFH-DA probe was used to identify the level of ROS in HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (C) Representative WB image showing the protein level of IL-1β in HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (D, E) ELISA showing the levels of IL-1β (D) and IL-18 (E) in the cell culture supernatant of HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (F) Representative WB image showing the protein level of BDH1 in HK-2 cells transfected with vector or Flag-BDH1. (G) DCFH-DA probe was used to identify the level of ROS in HK-2 cells based on the indicated treatment. (H) Representative WB image showing the protein level of IL-1β in HK-2 cells based on the indicated treatment. (I, J) ELISA showing the levels of IL-1β and IL-18 in the cell culture supernatant of BDH1-overexpressed HK-2 cells treated with HG (I) or PA (J). For samples used in G-J, HK-2 cells were transfected with the plasmids overexpressing Flag-BDH1 or vector for 6 h and then treated with HG or PA for 48 h. (K) DCFH-DA probe was used to identify ROS levels in HK-2 cells based on the indicated treatment. (L) Representative WB image showing the protein level of IL-1β in HK-2 cells based on the indicated treatment. (M, N) ELISA showing the levels of IL-1β and IL-18 in the cell culture supernatant of βOHB-supplemented HK-2 cells treated with HG (M) or PA (N). For samples used in K-N, HK-2 cells were treated with vehicle or βOHB and HG or PA for 48 h. All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Bar: 50 μm in B, 50 μm in top panels and 100 μm in bottom panels in G, 50 μm in K. Abbreviations: BDH1: β-hydroxybutyrate dehydrogenase 1; βOHB: β-hydroxybutyrate; HG: high glucose; PA: palmitic acid; ROS: reactive oxygen species; WB: western blot. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Since BDH1 deficiency led to increased ROS production and inflammation, we next attempted to determine whether HG- or PA-induced BDH1 reduction mediates HG- or PA-induced cell injury. Thus, we transfected HK-2 cells with a Flag-BDH1 overexpression plasmid to block HG- or PA-induced BDH1 reduction (Figure 3F). ROS assay showed that BDH1 overexpression significantly reduced HG-induced ROS overproduction (Figure 3G, top panels). In PA-treated cells, BDH1 overexpression nearly completely reversed PA-induced ROS overproduction (Figure 3G, bottom panels). Regarding inflammation, BDH1 overexpression reversed the HG- or PA-induced activation of IL-1β (Figure 3H) and the upregulation of secretory IL-1β and IL-18 (Figure 3I, 3J). This suggests that BDH1 may play a protective role in DKD pathogenesis and that pathological hyperglycemia- and hyperlipidemia-induced BDH1 reduction might mediate cell injury.

Given that BDH1 is a key enzyme that catalyzes the first step of βOHB metabolism, we next attempted to determine whether βOHB supplementation could exert a protective effect on HG- or PA-treated HK-2 cells. Consistent with the results for BDH1 overexpression, βOHB supplementation markedly reversed HG- or PA-induced ROS overproduction (Figure 3K). Similarly, βOHB supplementation reversed the HG- or PA-induced activation of IL-1β (Figure 3L) and upregulation of secretory IL-1β and IL-18 (Figure 3M, 3N). In generally, these findings suggest that BDH1-mediated βOHB metabolism plays an important role in protection against HG- or PA-induced cell injury.

BDH1-mediated βOHB metabolism promotes NRF2 nuclear translocation

Given that NRF2 is a well-known transcription factor that regulates the transcriptional induction of antioxidant response element-containing genes encoding antioxidant enzymes in response to cellular stress, including ROS [8–10], we next attempted to determine whether BDH1 mediates anti-ROS function by activating NRF2. Because NRF2 is a nuclear transcription factor, we determined the protein levels of NRF2 in nuclear extracts using western blotting. In HK-2 cells transfected with Bdh1 siRNA, the cytoplasmic NRF2 protein levels (Supplementary Figure 5A) were significantly up-regulated while the nuclear NRF2 protein levels (Figure 4A) and downstream antioxidant genes (GCLM, GPX4, and SOD2) (Figure 4B) were significantly down-regulated compared with those in cells transfected with normal control (NC) siRNA. Moreover, both HG and PA induced a reduction in NRF2 in nuclear extracts, whereas BDH1 overexpression reversed the reduction in HG- and PA-induced NRF2 (Figure 4C) and downstream antioxidative targets (GCLM, GPX4, and SOD2) (Figure 4D, 4E). In contrast, the cytoplasmic NRF2 protein level decreased in BDH1-overexpressed HK-2 cells (Supplementary Figure 5B). Consistent with the observations for BDH1 overexpression, βOHB supplementation reversed HG- and PA-induced NRF2 reductions in nuclear extracts (Figure 4F), which was further confirmed by NRF2 nuclear translocation assay with immunostaining (Figure 4G). Similarly, the expression of NRF2 downstream antioxidant genes, GCLM, GPX4, and SOD2 was significantly decreased in both HG- and PA-induced HK-2 cells and reversed by βOHB supplementation (Figure 4H, 4I), which was consistent with the NRF2 nuclear translocation results. In contrast, the cytoplasmic NRF2 protein level was decreased in HK-2 cells supplemented with βOHB (Supplementary Figure 5C) while the NRF2 mRNA level was not affected in HK-2 cells treated with HG, PA, or βOHB (Supplementary Figure 6). Additionally, AAV9-Bdh1-GFP injection promoted nuclear translocation of NRF2 (Figure 4J) and its downstream antioxidant genes (Gclm, Gpx4, and Sod2) (Figure 4K) in the kidneys of db/db mice. These data indicate that BDH1-mediated βOHB metabolism promoted NRF2 nuclear translocation.

Figure 4. BDH1-mediated βOHB metabolism promotes NRF2 nuclear translocation. (A) Representative WB image showing the protein level of NRF2 protein in the nuclear translocation of HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (B) mRNA expression of GCLM, GPX4, and SOD2 in HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (C) Representative WB images showing the protein level of NRF2 in the nuclear translocation of BDH1-overexpressed HK-2 cells treated with HG or PA. (D, E) mRNA expression of GCLM, GPX4, and SOD2 in BDH1-overexpressed HK-2 cells stimulated with HG (D) or PA (E). (F) Representative WB images showing the protein level of NRF2 in the nuclear translocation of βOHB-supplemented HK-2 cells treated with HG or PA. (G) Representative IF images showing the location of NRF2 in βOHB-supplemented HK-2 cells treated with HG or PA. (H, I) The mRNA expression of GCLM, GPX4, and SOD2 in βOHB-supplemented HK-2 cells stimulated with HG (H) or PA (I). (J) Representative IF images showing the location of NRF2 in kidneys of indicated groups. (K) mRNA levels of Gclm, Gpx4, and Sod2 in the kidneys of m/m group (n = 5), db/db + Vector (n = 4), and db/db + AAV-Bdh1-GFP group (n = 4). All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Bar: 5 μm in G, 50 μm and 10 μm in J. Abbreviations: BDH1: β-hydroxybutyrate dehydrogenase 1; βOHB: β-hydroxybutyrate; WB: western blot; HG: high glucose; PA: palmitic acid; IF: immunofluorescence. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Fumarate participates in BDH1-mediated renal protection through the AcAc-succinate-fumarate metabolic pathway

In the BDH1-mediated βOHB metabolism pathway, BDH1 first metabolizes βOHB into AcAc, which can enter the TCA cycle, and then AcAc is metabolized into succinate and fumarate in turn. Since fumarate is a well-known activator of NRF2 signaling, we examined whether BDH1 activates NRF2 by increasing in fumarate levels. We identified that AcAc, succinate, and fumarate concentrations were reduced in HK-2 cells transfected with Bdh1 siRNA (Figure 5A). Similar to what was observed in Bdh1 siRNA-transfected HK-2 cells, both HG and PA treatments reduced the levels of AcAc, succinate, and fumarate, which were successfully blocked by BDH1 overexpression (Figure 5B, 5C). Similarly, βOHB supplementation reversed the HG- or PA-induced reduction of AcAc, succinate, and fumarate (Figure 5D–5E). As the substrate of BDH1, βOHB was reduced in the kidneys of AAV9-Bdh1-GFP–injected db/db mice (Figure 5F). Consistent with the results observed in HK-2 cells, the levels of AcAc, succinate, and fumarate were reduced in the kidneys of db/db mice injected with the vector, although these changes were reversed by AAV9-Bdh1-GFP injection (Figure 5F). These findings collectively reveal a metabolic flux composed of βOHB-AcAc-succinate-fumarate that could be regulated by BDH1 or βOHB and affect downstream NRF2 signaling (Figure 5G).

Figure 5. BDH1-mediated βOHB metabolism promotes fumarate production through the AcAc–succinate–fumarate metabolic pathway. (A) Concentrations of AcAc, succinate, and fumarate in HK-2 cells transfected with NC siRNA or Bdh1 siRNA. (B, C) Concentrations of AcAc, succinate, and fumarate in BDH1-overexpressed HK-2 cells stimulated with HG (B) or PA (C). (D, E) Concentrations of AcAc, succinate, and fumarate in βOHB-supplemented HK-2 cells stimulated with HG (D) or PA (E). (F) Concentrations of βOHB, AcAc, succinate, and fumarate in the kidneys of indicated groups (n = 5 in m/m group, n = 4 in vector, and AAV-Bdh1-GFP injected db/db group). (G) Scheme showing the BDH1-regulated metabolic flux composed of βOHB–AcAc–succinate–fumarate. All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Abbreviations: BDH1: β-hydroxybutyrate dehydrogenase 1; βOHB: β-hydroxybutyrate; AcAc: acetoacetate; TCA: tricarboxylic acid; HG: high glucose; PA: palmitic acid; βOHB: β-hydroxybutyrate; AcAc: acetoacetate. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

To further determine whether the decrease in fumarate-mediated Bdh1 reduction-induced aggravation of glucotoxicity and lipotoxicity, we established a glucotoxicity and lipotoxicity cell model using Bdh1 siRNA-transfected HK-2 cells in vitro and a streptozotocin (STZ)-induced diabetes mouse model using Bdh1 knockout mice in vivo, and treated them with dimethyl fumarate (DM-Fumarate). The results showed that the DM-Fumarate treatment reversed the Bdh1 knockdown-induced increase in glucotoxicity (Figure 6A, 6B) and lipotoxicity (Figure 6C, 6D) in HK-2 cells. Moreover, the DM-Fumarate treatment reversed the Bdh1 knockout-induced increase in ACR (Figure 6E, 6F) and pathological injury to the kidney (Figure 6G) in STZ-induced diabetic mice. These data indicate that Bdh1 reduction aggravates glucotoxicity and lipotoxicity by down-regulating fumarate, which provides strong support for our conclusion that Bdh1 over-expression ameliorate glucotoxicity and lipotoxicity through βOHB-AcAc-succinate-fumarate metabolic flux in DKD.

Figure 6. DM-Fumarate treatment reverses the Bdh1 deficiency-induced increase of renal inflammation and pathological injury. (A–D) qRT-PCR analysis showing the IL-1β and IL-18 mRNA levels of indicated groups (5 μM DM-Fumarate for 48 h). (E) Representative WB image showing the protein level of BDH1 in the kidneys of WT and Bdh1-KO mice. (F) Urinary ACR values of mice in indicated groups (n = 3 in per group). (G) Representative photomicrographs of H&E, Masson, IHC (IL-1β), and TUNEL staining showing the pathological changes in the kidneys of indicated groups. All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Bar: 50 μm in G. Abbreviations: ACR: albumin-to-creatinine ratio; Bdh1: β-hydroxybutyrate dehydrogenase 1; DKD: diabetic kidney disease; WB: western blot; H&E: hematoxylin and eosin; IHC: immunohistochemistry; STZ: streptozotocin; DM-Fumarate: dimethyl fumarate. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; Abbreviation: ns: no significance.

βOHB supplementation alleviates the progression of DKD

As βOHB supplementation showed a similar effect as BDH1 overexpression on HG- or PA-induced ROS overproduction and inflammation in HK-2 cells, we next attempted to determine whether βOHB supplementation could ameliorate DKD. The experimental strategy is illustrated in Figure 7A. At six weeks after supplementation of βOHB (100 mM) in drinking water, we identified the blood βOHB levels and observed an increased βOHB levels in db/db mice supplemented with βOHB (Figure 7B). Subsequently, an ACR assay was performed. Although βOHB supplementation did not affect body weight, fasting blood glucose levels (Supplementary Figure 4C, 4D), and the kidney-to-body weight ratio (Supplementary Table 3), we observed that ACR was significantly lower in db/db mice supplemented with βOHB compared with mice supplemented with vehicle (Figure 7C). Moreover, the serum level of βOHB was negatively correlated with the value of ACR, indicating a strong ACR reduction capability of βOHB in DKD (Figure 7D). We observed increased BDH1 expression in the kidneys of db/db mice treated with βOHB (Figure 7E). Further histological analysis revealed that db/db mice supplemented with βOHB showed a normal glomerular morphology whereas db/db mice supplemented with vehicle showed glomerular hypertrophy (Figure 7F). Consistent with AAV9-mediated BDH1 renal expression, DKD pathology-related fibrosis, inflammation, and apoptosis were substantially reduced by βOHB supplementation (Figure 7F–7H), whereas the expression of downstream targets of NRF2 was elevated (Supplementary Figure 7A). These results indicate that βOHB supplementation might ameliorate DKD by increasing the renal expression of BDH1, which promotes βOHB metabolism.

Figure 7. βOHB supplementation alleviates the progression of DKD. (A) Schematic diagram illustrating the animal experimental design. During the experiment, m/m mice were fed with CD, while db/db mice were fed with HFD. Eight-week-old male db/db mice were randomized to receive vehicle or βOHB, which was dissolved in drinking water at a final concentration of 100 mM. Mice were euthanized at six weeks after βOHB intervention. (B) Blood level of βOHB in mice from indicated groups (n = 5 per group). (C) Urinary ACR values of mice from indicated groups (n = 5 per group). (D) Negative correlation between serum βOHB level and ACR in db/db mice with vehicle and βOHB treatment. (E) Representative WB image showing the protein level of BDH1 in the kidneys of indicated groups. (F) Representative photomicrographs of H&E, Masson, IHC (IL-1β), and TUNEL staining showing the pathological changes in the kidneys of indicated groups. (G) Quantification of the fibrosis area in the kidneys of indicated groups (n = 4 per group). (H) Quantification of apoptosis-positive cells in the kidneys of indicated groups (n = 4 per group). All results are representative of three independent experiments. In B and C, values are presented as median (interquartile range). In E, G, and H, values are presented as mean ± standard deviation. Bar: 100 μm in F. Abbreviations: CD: control diet; HFD: high fat diet; ACR: albumin-to-creatinine ratio; BDH1: β-hydroxybutyrate dehydrogenase 1; βOHB: β-hydroxybutyrate; DKD: diabetic kidney disease; WB: western blot; H&E: hematoxylin and eosin; IHC: immunohistochemistry. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Ketogenic diet alleviates the progression of DKD

KD has been widely used in clinical studies and reported to have antidiabetic effects; however, the underlying mechanisms have not been fully described. Because KD mainly produces βOHB and BDH1-mediated βOHB metabolism plays a protective role in DKD, we next attempted to determine whether KD could ameliorate DKD and identify whether KD functions through the BDH1-mediated βOHB metabolism pathway. As shown in Figure 8A, m/m and db/db mice were fed a standard diet (SD) or KD for nine weeks starting at eight weeks of age. Although KD feeding did not affect the body weight (Supplementary Figure 4E), the fasting blood glucose levels returned to normal levels (Supplementary Figure 4F) and the kidney-to-body weight ratio was significantly reduced in KD-fed db/db mice (Supplementary Table 3). Compared with SD-fed db/db mice, KD-fed db/db mice showed increased blood levels of βOHB (Figure 8B). The KD treatment nearly completely reversed the increase in ACR in SD-fed db/db mice (Figure 8C). We observed increased BDH1 expression in the kidneys of the KD-fed db/db mice (Figure 8D). Further histological analysis showed that the KD-fed db/db mice showed significant pathological remission in the kidneys, including fibrosis, inflammation, and apoptosis (Figure 8E–8G), whereas the expression of downstream targets of NRF2 was elevated (Supplementary Figure 7B). These results indicate that KD might ameliorate DKD by increasing blood βOHB and renal expression of BDH1, which ultimately promotes βOHB metabolism.

Figure 8. Ketogenic diet alleviates the progression of DKD. (A) Schematic diagram illustrating the animal experimental design. During the experiment, m/m mice were fed with CD, while db/db mice were fed with SD or KD starting at eight-week-old. Mice were euthanized at nine weeks after KD intervention. A schematic representation of the KD intervention strategy in db/db mice. (B) Blood level of βOHB in mice from indicated groups (n = 5 in m/m group; n = 4 in SD and KD groups). (C) Urinary ACR values of mice in indicated groups (n = 5 in m/m group; n = 4 in SD and KD groups). (D) Representative WB image showing the protein level of BDH1 in the kidneys of indicated groups. (E) Representative photomicrographs of H&E, Masson, IHC (IL-1β), and TUNEL staining showing the pathological changes in the kidneys of indicated groups. (F) Quantification of the fibrosis area in the kidneys of indicated groups (n = 4 per group). (G) Quantification of apoptosis-positive cells in the kidneys of indicated groups (n = 4 per group). All results are representative of three independent experiments. Values are presented as mean ± standard deviation. Bar: 100 μm in E. Abbreviations: ACR: albumin-to-creatinine ratio; SD: standard diet; KD: ketogenic diet; DKD: diabetic kidney disease; BDH1: β-hydroxybutyrate dehydrogenase 1; βOHB: β-hydroxybutyrate; WB: western blot; H&E: hematoxylin and eosin; IHC: immunohistochemistry. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001. Abbreviation: ns: no significance.

Animals

Bdh1 knock-out (KO) mice with a C57BL/6J background were generated by Cyagen (Cyagen Biosciences, Guangzhou, China). Eight-week-old male Bdh1-KO mice and their wild-type (WT) littermates received i.p. injections of low-dose STZ (50 mg/kg; S8050, Solarbio, Beijing, China). The STZ injection was repeated daily for five consecutive days. Five days after the final STZ injection, mice with non-fasting blood glucose >16.7 mmol/L (as measured with the Accu-Chek Advantage blood glucose monitoring system; Roche Diagnostics, Basel, Switzerland) were considered diabetic. Mice treated with sodium citrate buffer were used as controls. Two weeks after STZ injection, mice were randomized to receive DM-Fumarate (15 mg/kg body weight twice daily) or vehicle via oral gavage, as previously described [14]. Three mice were randomly assigned to each treatment group. Mice were euthanized eight weeks after STZ injection.

Five-week-old C57 BKS db/db male mice (n = 20; 31.94 ± 2.08 g) and db/db littermate controls (m/m; n = 5; 19.96 ± 0.79 g) were purchased from GemPharmatech Co., Ltd., (Nanjing, China). Animals were acclimatized for at least one week before the experiments and then randomly divided into five groups (m/m, db/db+AAV9-Vector, db/db+AAV9-Bdh1-GFP, db/db+standard diet (SD), and db/db+KD; n = 5 in each group). A 60% high-fat diet was purchased from Research Diets, Inc. (D12492, USA), SD (TP201442) and KD (74.2% fat, 8.9% protein, and 3.2% carbohydrates; TP201414) were purchased from Trophic Animal Feed High-Tech Co., Ltd., (Jiangsu, China). For the βOHB supplementation experiment, 5-week-old C57 BKS db/db male mice (n = 10; 31.08 ± 1.51 g) and db/db littermate controls (m/m; n = 5; 18.7 ± 1.32 g) were purchased from GemPharmatech Co., Ltd., (Nanjing, China). Animals were acclimatized before the experiments for at least one week and then randomly divided into three groups (m/m, db/db+Vehicle, and db/db+βOHB; n = 5 in each group). All animal experiments were performed under the following conditions: room temperature 23°C ± 1°C, relative humidity 60% ± 10%, and alternating 12 h light–dark cycles in individually ventilated cages. The animal experiments were approved by the Institutional Animal Ethics Committee of Southwest Medical University and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Human renal samples

Samples from patients who had been diagnosed with DKD were collected at the Department of Pathology of the Affiliated Hospital of Southwest Medical University. Normal samples were collected from individuals without diabetes or renal disease who had undergone tumor nephrectomy. The study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Research Ethics Committee of The Affiliated Hospital of Southwest Medical University. Informed consent was obtained from all patients.

Animal experiments

For AAV9-mediated Bdh1 renal expression, 100 μL of pAAV-ITR-CAG-Bdh1-IRES-EGFP-WPRE-Sv40polyA-ITR (AAV9-Bdh1-GFP, 3.40E+12 vg/mL, Beijing Syngentech Co., Ltd., Beijing, China) and vector (1.90E+13 vg/mL) was injected into db/db or m/m mice via the caudal vein. For βOHB supplementation, to reduce the risk of accidental death of mice, sodium β-hydroxybutyrate (CAS:150-83-4; Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) was dissolved in drinking water at a final concentration of 100 mM as previously described [45] and supplied to the db/db mice. Body weights and blood glucose levels were recorded weekly.

HK-2 cell culture

HK-2 cells (ATCC, USA) were cultured in Dulbecco’s Modified Eagle Medium (HyClone, Logan, UT, USA) containing 10% fetal bovine serum (ScienCell, Carlsbad, CA, USA) and supplemented with 1% penicillin–streptomycin (Beyotime, Shanghai, China). HK-2 cells were cultured at 37°C with 5% CO₂ until 60–70% confluence [46]. Then, the cells were treated with normal glucose (NG; 5.5 mM), mannitol (MA; 24.5 mM), high glucose (HG; 30 mM), vehicle (40 μM BSA), palmitic acid (PA; 200 μM, Sigma-Aldrich, Saint Louis, MO, USA), βOHB (5 mM), and DM-Fumarate (5 μM) for 48 h. PA [47], βOHB [48], and DMF [49] were prepared as previously described.

RNA-seq study and analysis

Total RNA from renal tissues (n = 3 each for the m/m and db/db groups) was extracted separately using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s procedure. All purified libraries were sequenced using an Illumina NovaSeq 6000 (LC Sciences, CA, USA) at Shanghai Majorbio Bio-Pharm Technology Co., Ltd., (Shanghai, China). Gene expression was normalized to the number method of fragments per kilobase per million reads. The resulting P-values were adjusted following the Benjamini and Hochberg method [50] to regulate the false discovery rate. DEGs were identified based on an adjusted P-values of < 0.05 and | log2 (fold change) | ≧1.5. To evaluate significantly enriched metabolic or signal transduction pathways, pathway enrichment analysis was performed using the KEGG database (version 2017.08) [51, 52]. Categories with false discovery rates (FDRs) <0.05 were considered significant KEGG pathways. Raw sequencing data are available from the NCBI BioProject SRA database (PRJNA876094).

Histopathological examination

Kidney tissues were fixed in 4% paraformaldehyde for 24 h, embedded in paraffin, and sectioned to a thickness of 4-μm. The sections were stained with hematoxylin and eosin (H&E) or Masson’s trichrome for light microscopy and morphometry.

Immunohistochemistry staining

Briefly, 4-μm-thick paraffin sections were dewaxed, hydrated, and stained with primary antibodies against BDH1 (1:100, ab193156, Abcam, UK) and IL-1β (1:100, #12242, Cell Signaling Technology, USA). The sections were stained with biotin-labeled goat anti-rabbit IgG or biotin-labeled anti-mouse IgG and then treated with horseradish enzyme-labeled oopaltin from Streptomyces (Beijing ZSGB Biological Technology Co., Ltd., Beijing, China). The stained sections were scanned under a light microscope.

Immunofluorescence staining

IF staining of kidney paraffin sections and HK-2 cells was performed using anti-BDH1 antibody (1:100) and NRF2 (1:50, sc518033, Santa Cruz, USA). The membranes were then incubated with Cy3/FITC fluorescent dye-conjugated secondary antibodies (1:200 dilution, Biosynthesis Biotech, USA) for 60 min at room temperature in the dark. Nuclei were labeled with DAPI, and images were obtained using a fluorescence microscope (Leica, Wetzlar, Germany).

ROS and TUNEL assay

ROS levels in HK-2 cells were measured using a DCFH-DA fluorescent probe according to the manufacturer protocol (Beyotime, Shanghai, China). TUNEL staining of the kidney paraffin sections was performed according to the TUNEL Kit protocol (Roche, Indianapolis, IN, USA).

Human Bdh1 cDNA transfection

Human Bdh1-overexpressed plasmid (pCMV3-BDH1-Flag) and vector plasmid (pCMV3) were purchased from Beijing Sino Biological, Inc., (Beijing, China) and transfected into HK-2 cells using Lipofectamine 3000 (Invitrogen).

SiRNA transfection

The siRNAs for BDH1 (stB0005692B) and NC (siN0000001-1-5) were purchased from RiboBio (Guangzhou, China). The siRNA sequence of the human BDH1 gene was 5′-GCCTAAACAGTGACCGATT-3′. The BDH1 and NC siRNA were transfected into HK-2 cells with riboFECT™ CP Reagent and riboFECT™ CP Buffer (RiboBio, Guangzhou, China).

Measurement of βOHB, AcAc, succinate, and fumarate

The βOHB content was measured using mouse βOHB ELISA kits (F9242-A, Shanghai Fankel Industrial Co., Ltd., Shanghai, China). AcAc content was measured using human acetoacetate ELISA kits (JL15388, Shanghai J&L Biological, Shanghai, China) and mouse acetoacetate ELISA kits (2M-KMLJM220724m, Nanjing Camilo Bioengineering Co., Ltd., Nanjing, China). The succinate and fumarate contents were measured using Succinate Assay kits (MAK335, Sigma-Aldrich, Saint Louis, MO, USA) and Fumarate Assay kits (MAK060, Sigma-Aldrich, Saint Louis, MO, USA) according to the manufacturer’s instructions.

qRT-PCR analysis

Total RNA from renal tissue and HK-2 cells was extracted using TRIzol reagent (Invitrogen, USA). ReverTra Ace qPCR RT Master Mix (FSQ-201, Toyobo, Japan) was used for reverse transcription, and a QuantiNova SYBR Green PCR Kit (QIAGEN, Germany) was used for qRT-PCR. qRT-PCR was performed with Analytikjena qTOWER 3G real-time PCR system (Analytik Jena, Jena, Germany) according to the manufacturer’s instructions. All primers used in this study are listed in Supplementary Table 4. β-actin was used as an internal reference gene to normalize target gene expression. All samples were measured in triplicate. The 2^−ΔΔCt method [53] was used to calculate gene expression relative to that of the reference gene.

Western blot analysis

Total proteins from kidneys of mice and HK-2 cells were extracted using an extraction buffer (RIPA). Nuclear proteins were extracted using a Nucleoprotein Extraction Kit (Shanghai Sangon Biotech, Shanghai, China). Protein samples were separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred onto a PVDF membrane (Millipore, Burlington, MA, USA). The membranes were incubated with 5% BSA to block other contaminants and then with primary antibodies. Immunoblotting was performed using anti-BDH1 antibody (1:1000 dilution), anti-NRF2 antibody (1:500 dilution), anti-IL-1β antibody (1:1000 dilution), anti-β-actin antibody (1:4000 dilution, AF0003, Beyotime, China), anti-Tubulin antibody (1:4000 dilution, AF0001, Beyotime, China), anti-Histone H3 antibody (1:1000 dilution, AF0009, Beyotime, China), and anti-Flag antibody (1:1000 dilution, AF5051, Beyotime, China).

Statistical analysis

All in vitro experiments were repeated three times. ImageJ software (version 1.80; National Institutes of Health, USA) was used for western blot analysis. Data were expressed as the mean ± standard deviation or median (interquartile range). Statistical analyses and figure generation were performed using GraphPad Prism 9. Continuous variables were analyzed using the two-tailed unpaired Student’s t-test or the Mann-Whitney U test, followed by normality tests for data from two groups. One-way ANOVA tests were performed for multiple comparisons of values, followed by Tukey’s test. Differences at P < 0.05 were considered statistically significant. Statistical significance was set at ^*P < 0.05, ^**P < 0.001, P < 0.01, and ^***P < 0.001.

Data availability

Data supporting the findings of this study are available from the corresponding author upon reasonable request.