Bdh1-Mediated βOHB Metabolism Ameliorates Diabetic Kidney Disease by Activation of Nrf2-Mediated Antioxidative Pathway

Ketogenic diet (KD) and β-Hydroxybutyrate (βOHB) has been widely reported as an effective therapy for metabolic diseases. β-hydroxybutyrate dehydrogenase 1 (Bdh1) is the rate-limiting enzyme of ketone metabolism. In this study, we investigated the Bdh1-mediated βOHB metabolic pathway in pathogenesis of diabetic kidney disease (DKD). Human renal tubule epithelial cells (HK-2 cells) induced by high glucose (HG) or palmitic acid (PA) were used to transfect with Bdh1 siRNA or plasmid-ag-Bdh1. Reactive oxygen species (ROS) levels, nuclear factor red 2-related factor 2 (Nrf2) protein expression, and βOHB-acetoacetate (AcAc)-succinate-fumarate metabolic ux were detected. Five-week-old C57 BKS db/db obese diabetic mice (db/db) and their littermate controls (+/+) were treated with KD, βOHB, and adeno-associated virus (AAV9)-Bdh1, respectively. Renal function was determined by urinary albumin/creatinine ratio (ACR), and histopathological, immunohistochemistry (IHC), TUNEL staining of kidney were also performed. The renal expression of Bdh1 was down-regulated in DKD mouse models, diabetic patients and HG or PA induced HK-2 cells. Bdh1 overexpression or βOHB treatment protected HK-2 cells from glucotoxicity and lipotoxicity by inhibiting ROS overproduction. Mechanistically, Bdh1-mediated βOHB metabolism activated Nrf2 through enhancement of metabolic ux composed of βOHB-acetoacetate-succinate-fumarate. Moreover, in vivo studies showed that AAV9-mediated Bdh1 renal expression successfully reversed the brosis, inammation and apoptosis in kidneys from C57 BKS db/db mice. Notably, either βOHB supplementation or KD feeding could elevate the renal expression of Bdh1 and reverse the progression of DKD. Our results revealed a Bdh1-mediated molecular mechanism in pathogenesis of DKD and identied Bdh1 as a potential therapeutic target for DKD.


Introduction
Diabetes mellitus (DM) is a chronic and serious metabolic disease, which has a signi cant impact on patients and their families all over the world. The latest report of the international diabetes federation predicted that the number of people with diabetes in the world will reach 700 million at the year of 2045 [1]. As the most common microvascular complications of DM, diabetic kidney disease (DKD) is the main cause of chronic kidney disease (CKD) and end-stage renal disease (ESRD) [2][3][4]. Thus, studies aimed at clarifying the pathogenesis of DKD and exploring novel therapeutic targets to treat DKD are urgently needed.
Under the physiological conditions, the production and elimination of ROS reach a dynamic balance. The abnormal increase of ROS in diabetic patients leads to oxidative stress injury and in ammatory response, ultimately promotes the occurrence and development of DKD [5][6][7]. It is well known that the transcription factor Nrf2 can maintain intracellular redox homeostasis and reduce cell damage by regulating the expression of antioxidant proteins [8][9][10]. Xiao L et al. [11] found that the protective effect of mitoQ, a mitochondria-targeted antioxidant, on high glucose-treated HK-2 cells was partially blocked by Nrf2 knockdown. Fumarate, an intermediate product of the TCA cycle, is well known to activate Nrf2-mediated antioxidant response [12][13][14]. However, the role of metabolic regulation in Nrf2-mediated anti-ROS pathway and the pathogenesis of DKD is still unclear.
In the state of prolonged fasting, strenuous exercise or disease, fatty acids will produce ketone bodies in the liver through β-oxidation. Ketone bodies will be released into blood circulation and transported to extrahepatic organs such as the brain, heart and kidneys, where they are used as metabolic fuel for the tricarboxylic acid (TCA) cycle [15]. In the heart of diabetic patients, the intake of ketone bodies is increased and utilized as an energy source partially replacing glucose [16]. As a diet causing the elevation of endogenous ketone bodies, KD was rst introduced by doctors as a therapeutic method of epilepsy [17] and then was reported to be bene cial for a variety of diseases, including diabetic cardiomyopathy and diabetic tractional retinal detachment [18][19][20][21]. In addition, KD intervention had been reported to show bene cial effect on T2DM and DKD [22][23][24]. Although an increasing number of evidence have been reported to support the relationship between KD and diseases, the underlying molecular mechanism is still unclear.
Ketone bodies is the sum of βOHB, acetoacetate (AcAc), and acetone. As the main component of ketone bodies, βOHB has been reported to be not only an alternative energy source for the body, but also mediates signal transduction in metabolic process to function in processes of antioxidant production, anti-in ammation and anti-aging [25,26]. Bdh1 is the rate-limiting enzyme of ketone metabolism and can directly catalyze the metabolism of βOHB and promote the reciprocal transformation between βOHB and AcAc [25]. It has been reported that heart speci c overexpression of Bdh1 can signi cantly ameliorate heart failure through inhibition of oxidative stress [27]. Moreover, up-regulation of Bdh1-mediated βOHB metabolism increases the concentration of fumarate, which subsequently activates Nrf2 to induce the expression of antioxidant stress response elements, and ultimately inhibit retinal degeneration under ischemic conditions [28]. However, the role of Bdh1-mediated βOHB metabolism in the pathogenesis of DKD is still unknown.
In this study, we report that Bdh1 de ciency is related to pathogenesis of DKD in vivo and glucotoxicity and lipotoxicity in vitro. We also demonstrate that Bdh1 functions as a previously unrecognized activator of Nrf2 through enhancement of metabolic ux composed of βOHB-AcAc-succinate-fumarate. Notably, the AAV-9 mediated renal expression of Bdh1 effectively relieved the progression of DKD and either βOHB supplementation or KD feeding could elevate the renal expression of Bdh1 and reverse the progression of DKD. Taken together, our ndings suggest a promising new therapy for DKD via targeting Bdh1-mediated βOHB metabolism.

Animals
Five-week-old C57 BKS db/db male mice (n=20; 31.94±2.08g) and db/db littermate control (wild type, WT; n=5; 19.96±0.79g) mice were purchased from GemPharmatech Co., Ltd. (Nanjing, China). Animals were acclimatized before the experiments for at least one week, then they were randomly divided into seven groups (n=5 each). 60% high fat diet (HFD) and KD (74.2% fat, 8.9% protein, 3.2% carbohydrate) were purchased from Trophic Animal Feed High-Tech Co., Ltd. (Jiangsu, China). All animal experiments were performed under the following condition: room temperature 23±1℃ relative humidity 60% ±10%, and an alternating 12h light-dark cycle in individually ventilated cages. Animal experiments were approved by the Institutional Animals Ethics Committees of Southwest Medical University and in accordance with the National Institutes of Health (NIH) guidelines for the care and use of laboratory animals.

Human renal samples
Samples from patients who had been diagnosed with diabetic nephropathy were collected from the Department of Pathology, The A liated Hospital of Southwest Medical University. Normal samples were collected from individuals who underwent tumor nephrectomies without diabetes or renal diseases. The study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Research Ethics Committee of the A liated Hospital of Southwest Medical University after informing the patients.

Histopathological examination
The kidney tissues were xed in 4% paraformaldehyde for 24h, embedded in para n and sectioned at 4μm thickness. The sections were stained by the hematoxylin-eosin (H&E), or Masson-trichrome methods for light microscopic analysis and morphometry.

IHC staining
Brie y, 4-μm-thick para n sections were dewaxing hydration and stained with primary antibodies against Bdh1(1:100, ab193156, Abcam), IL-1β (1:100, #12242, Cell Signaling Technology). The sections were stained with biotin-labeled goat anti-rabbit IgG or biotin-labeled anti-mouse IgG and then treated with the Horseradish enzyme labeled oopaltin of Streptomyces (Beijing ZSGB Biological Technology CO., LTD. China). Each photograph of the stained sections was scanned using a light microscope.

ROS and TUNEL assay
The level of ROS in HK-2 cells were measured by DCFH-DA uorescent probe according to the ROS Assay Kit protocol (Beyotime, China). TUNEL staining for the kidney para n sections was performed according to the TUNEL Kit protocol (Roche, USA).

Human Bdh1 cDNA transfection
The human Bdh1-overexpressed plasmid (pCMV3-Bdh1-Flag) and the vector plasmid (pCMV3) were purchased from Beijing Sino Biological Inc. China and transfected into HK-2 cells with Lipofectamine 3000 (Invitrogen).  Table S1. β-actin was used as an internal reference gene to normalize target gene expression. All the samples were used in triplicates. The 2 -△△Ct method [32] was used to calculate the relative gene expression in comparison with the reference gene.

Western Blot analysis
Total proteins of mice renal and HK-2 cells were extracted with extraction buffer (RIPA). Nuclear proteins were extracted with Nucleoprotein Extraction Kit protocol (Shanghai Sangon Biotech, China). The protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred into a PVDF membrane (Millipore). The membranes were incubated with 5% BSA to block other contaminants, and then with primary antibodies. Immunoblotting was performed using anti-Bdh1

Statistical Analysis
Data are expressed as the means ± standard deviation (SD) from triplicate experiments. Comparisons among groups were analyzed using one-way ANOVA analysis followed by 2-tailed unpaired Student's ttest using GraphPad Prism9. Differences were evaluated using P<0.05 was considered statistically signi cant. The statistical signi cance was *p<0.05, **0.001<p<0.01; ***P<0.001.

Results
RNA-seq analysis revealed Bdh1 reduction in DKD mouse model As a major microvascular complication in diabetic patients, DKD is the leading cause of CKD and ESRD [2][3][4]. To gain a comprehensive understanding of potential DKD regulators, we performed an RNA-seq analysis comparing gene expression in the kidneys from db/db mice or WT mice after the db/db mice emerged obvious pathological features of DKD (Fig.1A). KEGG analysis showed that the "Synthesis and degradation of ketone bodies" pathway was signi cantly down-regulated ( Fig.1B-C). Given that the KD was reported as an effective treatment for diabetes [33,34], we supposed that the down regulation of "Synthesis and degradation of ketone bodies" pathway participates the pathogenesis of DKD. Consistently, qRT-PCR analysis con rmed the expression changes of Bdh1, Oxct1, Acat1, and Hmgcs1 (Fig.1D). Notably, among these four pathway members, Bdh1 has been reported to protect heart from heart failure in TCA mouse model. Thus, to identify whether Bdh1 is involved in DKD pathogenesis, we further detected the protein level of Bdh1. As shown in Fig.1E, protein level of Bdh1 in kidneys of db/db mice was signi cantly lower than that in kidneys of WT mice. Moreover, the decrease in Bdh1 expression was also con rmed by immunohistochemistry (IHC) and immuno uorescence (IF) analysis (Fig.1F). Consistent with the DKD mouse model, we also observed downregulation of Bdh1 in renal tissues of diabetic patients with kidney disease by IHC and IF staining (Fig.1G). These results indicates that the decrease of Bdh1 expression is related to the pathogenesis of DKD.
Bdh1 de ciency mediated high glucose (HG) or palmitic acid (PA)-induced ROS overproduction and in ammation As it is known, hyperglycemia and hyperlipidemia are the two most obvious characteristics of type 2 diabetes [35]. In view of this, we established HG-induced glucotoxicity and PA-induced lipidtoxicity cell model with HK-2 cells to evaluate the effect of HG or PA on the Bdh1 expression. As expected, either mRNA level or protein level of Bdh1 was obviously reduced by HG or PA treatment in HK2 cells ( Fig.2A-C), which was also identi ed by IF analysis (Fig.2D-E).
To investigate whether the Bdh1 reduction contribute to HG or PA-induced cell injury, we performed Bdh1 knockdown in HK2 cells (Fig.2F). Given that the increased ROS plays a central and prominent role in the pathogenesis of diabetic microvascular complications including DKD [36] and the overproduction of ROS was related to in ammation [37], we next detected the ROS level and observed signi cant increase of ROS in HK2 cells transfected with Bdh1 siRNA (Fig.2G). In addition, the protein level of activated proin ammatory factor, cleaved IL-1β, was also elevated by Bdh1 knockdown (Fig.2H), as well as the secretory IL-1β and IL-18 ( Fig.2I-J). Collectively, these results suggest that Bdh1 de ciency might mediated HG or PA-induced cell injury by loss of anti-ROS function.
Either Bdh1 overexpression or βOHB supplementation reversed HG or PA-induced ROS overproduction and in ammation As the Bdh1 de ciency led to increased ROS and in ammation, we next sought to determine whether HG or PA-induced Bdh1 reduction mediates HG or PA-induced cell injury. To this end, we transfected HK2 cells with ag-Bdh1 overexpression plasmid to block the HG or PA-induced Bdh1 reduction (Fig.3A). Notably, ROS assay showed that the Bdh1 overexpression signi cantly reduced the HG-induced ROS overproduction (Fig.3B, upper panels). Especially in PA-treated cells, Bdh1 overexpression nearly completely reversed the PA-induced ROS overproduction (Fig.3B, lower panels). As to in ammation, Bdh1 overexpression also reversed the HG or PA-induced activation of IL-1β (Fig.3C) and the increase of secretory IL-1β and IL-18 ( Fig.3D-E). These evidences suggests that Bdh1 may play a protective role in DKD pathogenesis and pathological hyperglycemia and hyperlipidemia-induced Bdh1 reduction might mediate cell injury.
Given that Bdh1 is a key enzyme which mainly catalyzes the rst step of βOHB metabolism, we next sought to determine whether βOHB supplementation could also exhibit protective effect on HG or PAtreated HK2 cells. In line with Bdh1 overexpression, βOHB supplementation also markedly reversed HG or PA-induced ROS overproduction (Fig.4A). Similarly, βOHB supplementation also reversed the HG or PAinduced activation of IL-1β (Fig.4B) and the increase of secretory IL-1β and IL-18 (Fig.4C-D). Taken together, these ndings suggest that Bdh1 mediated βOHB metabolism play important role in protection of HG or PA-induced cell injury.
Bdh1-mediated βOHB metabolism promoted Nrf-2 nuclear translocation through the AcAc-succinatefumarate metabolic pathway Given that Nrf2 is a well-known transcription factor that regulates transcriptional induction of AREcontaining genes encoding antioxidant enzymes in response to cellular stresses including ROS [8-10], we next sought to determine whether Bdh1 mediates anti-ROS function through activation of Nrf2. As Nrf2 is a nuclear transcription factor, we detected the protein level of Nrf2 by western blot (WB) in nuclear extracts. Of note, in HK2 cells transfected with Bdh1 siRNA, Nrf2 protein level was signi cantly lower than that in cells transfected with control siRNA (Fig.5A). Moreover, either HG or PA could induce Nrf2 reduction in nuclear, whereas Bdh1 overexpression could reversed both HG and PA-induced Nrf2 reduction (Fig.5B). Consistent with the observations made in Bdh1 overexpression, βOHB supplementation also reversed both HG and PA-induced Nrf2 reduction in nuclear extracts (Fig.5C), which was further con rmed by Nrf2 nuclear translocation assay with immunostaining (Fig.5D). These data indicate that Bdh1mediated βOHB metabolism promotes Nrf-2 nuclear translocation.
In Bdh1-mediated βOHB metabolism pathway, Bdh1 rstly metabolites βOHB into AcAc, which could enter into TCA cycle and then is metabolized into succinate and fumarate in turn (Fig.6A). As the fumarate is a well-known activator of Nrf2 signaling, we next investigated whether Bdh1 activated Nrf2 by increase of fumarate. Interestingly, we found that the concentrations of AcAc, succinate and fumarate were all decreased in HK2 cells transfected with Bdh1 siRNA (Fig.6B). Similarly, to that observed in Bdh1 siRNA transfected HK2 cells, both HG and PA treatment could reduce the levels of AcAc, succinate and fumarate, which was successfully blocked by Bdh1 overexpression (Fig.6C-D). Likewise, βOHB supplementation also reversed HG or PA-induced reduction of AcAc, succinate and fumarate (Fig.6E-F). These ndings collectively reveal a metabolic ux composed of βOHB-AcAc-succinate-fumarate, which could be regulated by Bdh1 or βOHB and affected the downstream Nrf2 signaling (Fig.6G).

AAV9-mediated Bdh1 renal expression alleviated the progression of DKD
On the basis of the pronounced capacity of Bdh1 to inhibit ROS overproduction and in ammation, we next explored the therapeutic e cacy of Bdh1 expression in db/db mice. The experimental strategy is shown in Fig.7A. At the time point of 11 weeks after injection of the control or Bdh1-encoding virus, we performed ACR assay, which is the most important function indicator of kidney. Notably, although the Bdh1 renal expression didn't affect the body weight and fasted blood glucose (Fig.S1A-B), we observed signi cantly lower ACR in AAV9-Bdh1-injected mice than that in the AAV9-Control-injected mice (Fig.7B). To con rm whether mouse Bdh1 was effectively expressed in the kidney using AAV9 vector, we detected the uorescence intensity of GFP, which was co-expressed with Bdh1. We found that AAV9 encoding mouse Bdh1 was successfully delivered to the kidneys after 4 weeks of injection (Fig.7C). As expected, we observed increased Bdh1 expression in kidneys from AAV9-Bdh1 injected mice than that in AAV9control injected mice (Fig.7D). In further histological analysis, AAV9-Bdh1 injected db/db mice showed normal morphology of glomerulus, unlike the glomerular hypertrophy in AAV9-control injected mice (Fig.7E). In addition, the DKD pathology-related brosis, in ammation and apoptosis were also substantially reduced by AAV9-Bdh1 injection (Fig.7E-G). These ndings collectively provide strong support for the promising application of Bdh1 as a therapeutic target in DKD. βOHB supplementation alleviated the progression of DKD As βOHB supplementation showed similar effect to Bdh1 overexpression on HG or PA-induced ROS overproduction and in ammation in HK2 cells, we next sought to determine whether βOHB supplementation could ameliorate DKD. The experimental strategy is shown in Fig.8A. At the time point of 6 weeks after supplementation of βOHB by drinking water, we detected the serum level of βOHB and observed increased serum level of βOHB in db/db mice supplied with βOHB (Fig.8B). After that, we performed ACR assay. Although the βOHB supplementation didn't affect the body weight and fasted blood glucose (Fig.S1C-D), we found that there was signi cantly lower ACR in db/db mice supplied with βOHB than that with vehicle (Fig.8C). Moreover, the serum level of βOHB was negatively correlated with the value of ACR, indicating a strong ACR reduction capability of βOHB in DKD (Fig.8D). Interestingly, we also observed increased Bdh1 expression in kidneys from db/db mice supplied with βOHB (Fig.8E). In further histological analysis, db/db mice supplied with βOHB showed normal morphology of glomerulus, unlike the glomerular hypertrophy observed in db/db mice supplied with vehicle (Fig.8F). Consistent with AAV9-mediated Bdh1 renal expression, the DKD pathology-related brosis, in ammation and apoptosis were also substantially reduced by βOHB supplementation (Fig.8G-H). These results indicate that βOHB supplementation might ameliorate DKD by increasing renal expression of Bdh1, which nally promotes the βOHB metabolism.

Ketogenic diet alleviated the progression of DKD
The KD has been widely used in clinical studies and reported to have an anti-diabetic effect, while the underlying mechanisms have not been fully demonstrated. Given that the major production of KD is βOHB and Bdh1-mediated βOHB metabolism plays protective role in DKD, we next sought to determine whether KD could ameliorate DKD and whether it functions through Bdh1-mediated βOHB metabolism pathway. As shown in Fig.9A, WT or db/db mice were subjected to a standard diet (SD) or KD for 9 weeks, starting at the age of 8 weeks. Although the KD feeding didn't affect the body weight (Fig.S1E), the fasted glucose was reversed into normal level in KD-fed db/db mice (Fig.S1F). Compared with SD-fed db/db mice, db/db mice fed with KD showed increased blood level of βOHB (Fig.9B). Notably, KD treatment nearly completely reversed the increase of ACR in SD-fed db/db mice (Fig.9C). Interestingly, we observed increased Bdh1 expression again in kidneys from db/db mice fed with KD (Fig.9D). In further histological analysis, db/db mice fed with KD showed signi cantly pathological remission in kidneys, including brosis, in ammation and apoptosis (Fig.9E-G). These results indicate that feeding KD might ameliorate DKD by increasing blood βOHB and renal expression of Bdh1, which nally promotes the βOHB metabolism.

Discussion
In this study, we identi ed Bdh1 in renal cells as a potential therapeutic target for DKD. Our studies showed that expression of Bdh1 was reduced in DKD and HG or PA-treated HK-2 cells. Bdh1 overexpression or βOHB treatment can protect HK-2 cells from glucotoxicity and lipotoxicity.
Mechanistically, we found that Bdh1-mediated βOHB metabolism inhibits oxidative stress by activation of Nrf2 through upregulation of fumarate production. Of note, AAV9-mediated Bdh1 renal expression, βOHB supplementation or KD feeding can respectively reversed the brosis, in ammation and apoptosis in DKD. Thus, Bdh1-mediated βOHB metabolism in kidney lights a new way for DKD treatment.
In kidney, proximal renal tubular epithelial cells have reabsorption function and play an important role in the pathogenesis of DKD. However, the molecular mechanism by which tubular epithelial cells contribute to DKD is unclear. Of note, we found that the expression level of Bdh1 in HK-2 cells was much higher than that in Mpc5 cells (podocyte cell line) and SV40-MES cells (glomerular mesangial cell line) (date not shown). Our studies demonstrate a renal tubular Bdh1-mediated mechanism which participates in progression of DKD.
In the pathophysiology of diabetic kidney disease, increased oxidant species have been identi ed as the single unifying upstream event which holds a central and prominent role in progression of DKD [38]. Nrf2, a master positive regulator for genes related to antioxidant effects, is negatively regulated by kelch like epichlorohydrin associated protein 1 (Keap1) [39]. Thus, clinical trials of Keap1 inhibitors or Nrf2 inducers were conducted for treating DKD [40]. In addition, fumarate is well known to inhibit the binding of Keap1 to Nrf2 by modifying the cysteine residues [12,26]. In this study, we identi ed Bdh1 as another Nrf2 activator in progression of DKD. We showed that Bdh1 overexpression upregulated the fumarate level by promoting metabolic ux composed of βOHB-AcAc-succinate-fumarate (Fig. 6). Moreover, Bdh1 also has been reported to play protective role in pressure overload-induced heart failure, whereas the underlying mechanism is unknown [27]. Our study indicates that the molecular mechanism by which Bdh1 functions protective effect in DKD might also contribute to the protective effect of Bdh1 in heart.
In recent years, gene therapy has emerged as a novel therapeutic modality that has the potential to cure substantial disease. As one of the gene deliveries vectors, AAV has achieved preclinical and clinical success in the treatment of human diseases by gene replacement, gene silencing and gene editing, which has been identi ed as a safe, well tolerated and effective therapeutic vector [41]. The US Food and Drug Administration recently approved AAV-based gene therapy for infant GM1 ganglioside storage disease and frontotemporal lobe dementia caused by granule protein mutation [42]. However, there is no evidence on AAV-based gene therapy of DKD. In this study, we successfully observed the renal expression of GFP by AAV9-GFP intravenous injection, indicating that AAV9 is an effective gene delivery tool for kidney target in mice (Fig. 7). Moreover, AAV9 mediated Bdh1 renal expression signi cantly inhibited the progression of DKD in db/db mice. Although the AAV9 is not a kidney-speci c vector and the renal delivery is not so e cient as heart and liver [43], the effective amelioration of DKD which was observed in AAV9-Bdh1-injected db/db mice strongly suggests the AAV9 mediated gene targeting therapy as a promising treatment.
Ketone bodies metabolism refers to the processes of production in the liver (ketogenesis) and utilization in extrahepatic organs (ketolysis) [25]. KD is a diet with high fat and low carbohydrate, which can simulate the fasting metabolism and make the body in a state of ketogenesis [34]. However, the roles of ketone bodies or KD in various disease are not so much clear and even contradictory. In one hand, βOHB and KD are clinically bene cial in several common human neurodegenerative diseases [26] and are reported to ameliorate hyperglycemia, mitochondrial dysfunction and cardiomyopathy in db/db mice [18]. KD was reported to reverse diabetic kidney disease in T1DM and T2DM mouse model as early as 2011 [24], which is consistent with our data (Fig. 9), but the underlying mechanism is still unknown. In the other hand, increased ketone bodies were reported to promote DKD progression in diabetic patients in another clinical study [44]. KD feeding promotes liver brosis and leads to liver damage in a T2DM mouse model [45]. In addition, KD-induced high protein intake may accelerate the progression of individual's kidney disease [46]. In this study, we showed that either βOHB supplementation or KD feeding could alleviate the progression of renal brosis, in ammation and apoptosis in db/db mice (Fig. 9). Of note, either βOHB supplementation or KD feeding elevated the renal expression of Bdh1, which functions as a strong antioxidant by activation of Nrf2. Our study further con rmed the protective effect of βOHB and KD on DKD and a Bdh1-mediated mechanism. Collectively, as coins have both sides, βOHB and KD might have unknown side effects when they protecting individuals from disease, which further suggests the Bdh1targeted therapy as a more appropriate selection.
Taken together, our results provide evidence for the proposed mechanism depicted in Fig. 10. Hyperglycemia and hyperlipidemia in db/db mice lead to down-regulated expression of Bdh1, which subsequently down-regulates the fumarate level by metabolic ux composed of βOHB-AcAc-succinatefumarate. The fumarate reduction decreases nuclear translocalization of Nrf2, a key inhibitor of ROS. The overproduction of ROS nally activates the DKD-related in ammation, brosis and apoptosis in kidney. Our ndings also highlight the possibility of Bdh1 renal expression, βOHB supplementation and KD in attenuating DKD.   Values are presented as mean ± SD. *P<0.05; **P<0.01; ***P<0.001.