Ketone Body Supplementation Exerts Renoprotective Effects Against Adenine‐Induced Kidney Injury via OXCT1‐Mediated Ketolysis in Mice

ABSTRACT

Ketone bodies have traditionally been recognized as glucose‐sparing energy sources, with hepatic ketogenesis and peripheral ketolysis serving pivotal functions in maintaining energy homeostasis during fasting. Although they are commonly seen as harmful due to their link with ketoacidosis, recent studies emphasize their roles in organ protection. This has sparked interest in their possible use as a treatment for chronic kidney disease (CKD). In this study, we examined both exogenous and endogenous ketone body supplementation in adenine‐induced kidney injury in mice. Supplementation with the ketone body precursor 1,3‐butanediol significantly improved adenine‐induced renal fibrosis, inflammation, and apoptotic cell death. However, genetically deleting 3‐hydroxy‐3‐methylglutaryl‐CoA synthase 2 (HMGCS2), the key enzyme for ketogenesis, in the liver, kidney, or entire body, and removing Succinyl‐CoA:3‐ketoacid‐CoA Transferase 1 (OXCT1), the enzyme for ketolysis, in the kidney alone, did not affect the severity of adenine‐induced kidney damage. In contrast, the protective effects of 1,3‐butanediol were partially diminished in mice with kidney‐specific OXCT1 deficiency, indicating that OXCT1‐mediated ketolysis is at least partly necessary for the renal protection afforded by exogenous ketone body supplementation. These findings suggest that supplementing with exogenous ketone bodies, rather than relying on endogenous hepatic or renal ketone production, protects the kidneys in adenine‐induced kidney injury in mice, implying that local ketolysis within the kidney plays a mechanistic role in this protection. Our results highlight the therapeutic potential of exogenous ketone body administration in CKD and offer insights into how renal ketone metabolism helps protect against kidney injury.

1,3‐Butanediol is a ketone body precursor that is directly converted into the ketone bodies, β‐hydroxybutyrate (BHB), in the liver. When renal ketolysis, regulated by 3‐oxoacid CoA‐transferase 1 (OXCT1), is preserved, BHB supplementation exerts renoprotective effects. In contrast, when ketone body utilization is impaired, these renoprotective effects are abolished, leading to the progression of kidney injury.

1. Introduction

Ketone bodies have long been recognized as glucose‐sparing energy sources. Hepatic ketogenesis and ATP production through ketolysis in peripheral organs are essential for maintaining organ function, particularly during fasting [1, 2]. However, in the modern era of overnutrition, their physiological roles have often been overlooked, and ketone bodies have been viewed negatively because of their association with ketoacidosis [3]. Nevertheless, accumulating evidence has recently shown that ketone bodies provide tissue‐protective effects across multiple organs [4, 5, 6, 7, 8]. The protective mechanisms include not only improving energy metabolism in damaged tissues but also performing energy‐independent functions, such as regulating intracellular signaling pathways and modulating epigenetic processes [6, 9, 10, 11]. Therefore, ketone bodies are expected to be promising targets for interventions aimed at promoting healthy aging.

Chronic kidney disease (CKD) is one of the most common conditions in modern society and remains a major global health challenge [12, 13, 14]. Although the use of renin–angiotensin–aldosterone system (RAAS) inhibitors and sodium–glucose cotransporter 2 (SGLT2) inhibitors has gradually improved renal outcomes in diabetic and non‐diabetic CKD [15, 16, 17, 18], a significant residual risk persists, highlighting the need for new treatment strategies. We previously demonstrated, in diabetic mouse models, that administering 1,3‐butanediol (1,3‐BD), a precursor of ketone bodies, reduces renal injury [19]. Moreover, additional research has shown that exogenous ketone bodies can slow disease progression in polycystic kidney disease (PKD) models and offer kidney protection in acute kidney injury (AKI) models [20, 21, 22], further emphasizing the growing interest in the therapeutic potential of ketone bodies in various kidney diseases. However, the therapeutic potential of ketone bodies in non‐diabetic CKD has not been fully understood.

Additionally, recent research has highlighted the significance of extra‐hepatic ketogenesis. For instance, in the small intestine, ketone bodies not only regulate energy metabolism during fasting but also support intestinal tissue repair in mice [23]. In the nervous system, astrocyte‐produced ketone bodies are vital for maintaining neuronal function during fasting in Drosophila [24]. Interestingly, the expression of 3‐hydroxy‐3‐methylglutaryl‐CoA synthase 2 (HMGCS2), the rate‐limiting enzyme for ketogenesis, has been observed in renal proximal tubular cells (PTCs) [25, 26]. However, the precise role of local ketone production in the kidney remains unclear. Furthermore, our previous research has demonstrated that ketolysis for ATP production is increased in injured kidneys [19]; however, it remains unclear whether this local enhancement of ketone body metabolism genuinely contributes to renoprotection. Therefore, the comprehensive role of renal ketone metabolism in the progression of kidney disease remains to be fully elucidated.

In this context, several key questions remain: Whether supplementation with exogenous ketone bodies can mitigate renal injury in non‐diabetic CKD models, whether endogenous hepatic and renal ketogenesis contributes to the progression of CKD, and whether the renoprotective effects of ketone bodies are mediated through ketolysis. To address these inquiries, the current study employed an adenine‐induced murine model of renal injury, a mouse CKD model [27], to examine the roles of exogenous ketone body supplementation with 1,3‐BD, endogenous hepatic and renal ketogenesis regulated by HMGCS2, and renal ketolysis facilitated by succinyl‐CoA:3‐oxoacid CoA transferase (OXCT1).

2. Materials and Methods

2.1. General Information for Animal Studies

All mice were housed in a temperature‐controlled environment maintained at 23°C, with a 12‐h light and 12‐h dark cycle (20:00–08:00). The animals were cared for in facilities operated by the Research Center for Animal Life Science (RCALS) at Shiga University of Medical Science. All experimental protocols were approved by the Gene Recombination Experiment Safety Committee and the RCALS at Shiga University of Medical Science, and adhered to the ARRIVE guidelines.

2.2. 1,3‐BD Treatment in Adenine Nephropathy

Twelve‐week‐old male C57BL/6J mice (CLEA Japan Inc., Tokyo, Japan, RRID:IMSR_JAX:000664) were purchased. Adenine nephropathy was induced as previously reported with minor modifications [27]. At 13 weeks of age, the mice were divided into three groups: [1] a standard diet supplemented with 0.2% adenine, [2] a diet containing 20% 1,3‐BD (Sigma Aldrich, St. Louis, MO) supplemented with 0.2% adenine, and [3] a standard diet without adenine supplementation (control). 1,3‐BD administration was initiated simultaneously with the adenine diet. After 7 days of ad libitum feeding, renal injury was assessed.

To evaluate the renal toxicity of 1,3‐butanediol (1,3‐BD), twelve‐week‐old male C57BL/6J mice were administered 1,3‐BD or vehicle for 7 days under standard diet conditions.

2.3. Generation of Knockout Mouse Models

Whole‐body Hmgcs2‐deficient mice were generated as previously reported [19]. The Hmgcs2‐flox mouse model on the C57BL/6 background was initially established (Dr. Nakahara's Lab). The LoxP site was inserted to flank Exon 2 (455 bp) of the Hmgcs2 gene. Tamoxifen‐inducible PTC‐specific Hmgcs2‐deficient (Hmgcs2^ΔPT) or liver‐specific Hmgcs2‐deficient (Hmgcs2^ΔLiver) mice were generated by crossing Hmgcs2^f/f mice with PTC‐specific CreER (SLC34a1‐CreER) mice on the C57BL/6 background (Dr. Humphreys' Lab) [28] or with liver‐specific CreER (Albumin‐CreER) mice on the same background, which were kindly provided by Dr. Pierre Chambon (IGBMC, France) [29], respectively. To generate Hmgcs2^ΔPT mice and Hmgcs2^ΔLiver, 8‐week‐old Hmgcs2^f/f mice carrying either SLC34a1‐CreER or Albumin‐CreER were treated with tamoxifen (150 mg/kg/day) (Sigma Aldrich) for five days. The phenotypes of Hmgcs2^ΔPT and Hmgcs2^ΔLiver mice were compared to those of tamoxifen‐injected Hmgcs2^f/f mice from the same litter. The Oxct1‐flox mouse (Oxct1^f/f) model on the C57BL/6J background was previously generated and obtained from the Mutant Mouse Resource & Research Centers (Strain# MGI:6364617). Tamoxifen‐inducible PTC‐specific Oxct1‐deficient (Oxct1^ΔPT) mice were created by crossing Oxct1^f/f mice with SLC34a1‐CreER mice, followed by treating them with tamoxifen (150 mg/kg/day) for five days. The phenotypes of Oxct1^ΔPT mice were compared to those of Oxct1^f/f mice from the same litter. All mice were genotyped five weeks after birth using PCR with genomic DNA from each mouse and specific primers listed in Table 1. As mentioned above, the assessment of renal injury in adenine nephropathy was conducted by administering dietary adenine with/without 1,3‐BD to each genetically modified mouse model.

TABLE 1.

Primer sequences used in this study.

Gene name	Forward	Reverse
Genotyping (mouse)
Hmgcs2Δwhole	5′‐AATATGTGGACCAAACTGACCTG‐3′	5′‐CTTGGACTTGTCAATGATGGTCT‐3′
Cre recombinase	5′‐AGGTTCGTTCACTCATGGA‐3′	5′‐TCGACCAGTTTAGTTACCC‐3′
Hmgcs2flox	5′‐TGTCACATACCTTTAGAGTGCTTCT‐3′	5′‐ACTCTTAATTGACACCTTGATGAGC‐3′
Oxct1floxed	5′‐CCAGCATTTATAGTAGCATGGAAATC‐3′	5′‐TCAAACCTCCACAAGTGGTACAGGG‐3′
Real‐time PCR
Gapdh	5′‐CATGGCCTTCCGTGTTCCTA‐3′	5′‐GCGGCACGTCAGATCCA‐3′
Tgfb1	5′‐ATTCCTGGCGTTACCTTGG‐3′	5′‐AGCCCTGTATTCCGTCTCC‐3′
Col1a1	5′‐TTCTCCTGGCAAAGACGGAC‐3′	5′‐CGGCCACCATCTTGAGACTT‐3′
Ccl2 (MCP‐1)	5′‐TGACCCCAAGAAGGAATGGG‐3′	5′‐ACCTTAGGGCAGATGCAGTT‐3′
Adgre1 (F4/80)	5′‐TGTCTGAAGATTCTCAAAACATGGA‐3′	5′‐TGGAGCTTCATAGTTGTAAGGCA‐3′
Bax	5′‐GCGTGGTTGCCCTCTTCTAC‐3′	5′‐ACGGAGGAAGTCCAGTGTCC‐3′
Havcr1 (Kim‐1)	5′‐GCTACAGGAAGACCCACGACTATTT‐3′	5′‐GATGTTGGAGGAGTGGAGGTAGAG‐3′
Hmgcs2	5′‐CAGCTAATCCAGCCCTAGCC‐3′	5′‐AGCTCTTCGTGGGTTCTGTG‐3′
Oxct1	5′‐TGGCCAACTGGATGATACCTGG‐3′	5′‐TCCATGGTGACCACCACTTTGG‐3′

Abbreviations: Adgre1, Adhesion G protein‐coupled receptor E1; Bax, Bcl‐2‐associated X protein; Ccl2, C‐C motif chemokine ligand 2; Col1a1, collagen, type I, alpha 1; Gapdh, Glyceraldehyde‐3‐phosphate dehydrogenase; Havcr1, Hepatitis A Virus Cellular Receptor 1; Hmgcs2, 3‐hydroxy‐3‐methylglutaryl‐CoA synthase 2; Oxct1, Succinyl‐CoA:3‐ketoacid CoA transferase 1; Tgfb1, Transforming growth factor 1.

2.4. Measurements for Parameters in Blood and Plasma

Glucose and ketone body concentrations (β‐hydroxybutyrate, BHB) in blood collected from the tail vein were measured using a Glutest sensor (Sanwa Kagaku Kenkyusho, Nagoya, Japan) and a Ketometer (Abbott, Chicago, IL, RRID:SCR_010477) [19]. Plasma cystatin C levels were measured using a Cystatin C measurement kit (BioVender, Asheville, NC) according to the manufacturer's instructions [19].

2.5. Urinary Albumin Measurement

Urine samples were collected over a 24‐h period utilizing metabolic cages, which allowed mice unrestricted access to food and water [19]. The collected urine samples were then centrifuged to eliminate debris. Clear supernatants obtained from the urine samples were subsequently used to measure albumin and creatinine concentrations, employing an LBIS Mouse Urinary Albumin Assay Kit (S‐type) (FUJIFILM Wako Shibayagi Corporation, Tokyo, Japan) in accordance with the manufacturer's instructions [19].

2.6. Histopathology

Histological assessment was performed as previously reported [19]. The removed renal tissues were fixed with 4% paraformaldehyde for 24 h, then dehydrated using a graded series of alcohol solutions (2 × 70%, 2 × 95%, 2 × 100%, and 2 × xylene) for two hours each, and subsequently immersed in paraffin for two sessions of six hours each. The paraffin‐embedded renal specimens were sectioned at a thickness of 4 μm and stained employing standard Hematoxylin Eosin (HE) and Masson's trichrome protocols. The stained sections were then examined and imaged utilizing an All‐in‐One Microscope (KEYENCE, Osaka, Japan). The uninjured area of the renal tissue on H&E‐stained sections was calculated by dividing the area of the normally colored cortical tubule regions by the total cortical area.

For immunostaining of HMGCS2 and F4/80 in kidney tissues, paraffin‐embedded tissue sections were employed [19]. Endogenous peroxidases were inhibited using 0.03% H₂O₂ for 20 min, and nonspecific binding was reduced with 2.5% normal horse serum for 30 min. Tissues were incubated with the primary antibody against HMGCS2 (Abcam, Boston, MA, RRID: AB_2749817) or F4/80 (BIO‐RAD Laboratories, Hercules, CA, RRID: AB_323806) for 12 h at 4°C, followed by incubation with a secondary antibody (Histofine Simple Stain MAX‐PO; Nichirei Biosciences, Tokyo, Japan) in accordance with the manufacturer's protocol. The reactions were developed using Peroxidase Stain DAB Kit (Brown Stain) (Nacalai Tesque, Kyoto, Japan) according to the manufacturer's instructions. Subsequently, the slides were counterstained with hematoxylin (Muto Pure Chemicals, Tokyo, Japan).

TUNEL staining was conducted using the TACS 2 TdT DAB Kit (Trevigen, Gaithersburg, MD) according to the manufacturer's instructions. Renal cortical regions at a magnification of 200× were imaged. After measuring the entire renal cortical area and the number of TUNEL‐positive cells, the data are expressed as the number of TUNEL‐positive cells per square millimeter.

2.7. Real‐Time PCR

Total RNA was isolated from the kidney samples, which were isolated and preserved by rapid freezing, using the RNeasy Mini Kit (QIAGEN, Hilden, Germany). cDNA synthesis was performed utilizing the PrimeScript RT Reagent Kit (Takara Bio Inc., Shiga, Japan). Real‐time PCR was conducted with the Power SYBR Green PCR Master Mix (Bio‐Rad Laboratories, Hercules, CA) on the ABI Prism 7500 Sequence Detection System (ABI Applied Biosystems, Foster City, CA). The levels of mRNA expression were quantified through the standard curve method, which involved constructing standard curves with serial dilutions of a standard template. The cycle threshold (Ct) values served as the basis for calculating mRNA expression levels from the standard curve. Data analysis included normalization to the mRNA expression levels of Glyceraldehyde‐3‐phosphate dehydrogenase (Gapdh) used as an internal control. The primers employed are listed in Table 1.

2.8. Immunoblot Analysis

Kidney tissues were homogenized in an ice‐cold lysis buffer containing 20 mM Tris–HCl (pH 7.5), 140 mM NaCl, 1 mM EDTA, 1% Nonidet P‐40, and 50 mM NaF. The buffer was supplemented with PhosSTOP phosphatase inhibitor cocktail (Roche) and a complete protease inhibitor cocktail (Roche). These samples were resolved by 10% SDS‐PAGE and transferred to polyvinylidene fluoride membranes (Immobilon). The membranes were incubated with Anti‐HMGCS2 antibody (Cell Signaling Technology, Beverly, MA, RRID: AB_2798853), Anti‐β‐Hydroxybutyryllysine (Kbhb) antibody (PTM BIO, Chicago, IL, RRID: AB_3096286), Anti‐β actin antibody (Sigma‐Aldrich, St. Louis, MO, RRID: AB_476692), washed, and then incubated with secondary antibody, anti‐Rabbit IgG, HRP‐Linked Whole Ab Donkey (Cytiva, Tokyo, Japan, RRID: AB_772206) or Anti‐Mouse IgG, HRP‐Linked Whole Ab Sheep (Cytiva, Tokyo, Japan, RRID: AB_772210). After chemiluminescent detection using an enhanced chemiluminescence substrate (Cytiva, Tokyo, Japan), the membranes were imaged with a Fusion Solo 6S Edge chemiluminescence imaging system (Vilber Lourmat, Marne la Vallée, France). Image acquisition and analysis were performed using the Evolution Capt‐Edge (Vilber Lourmat, Marne la Vallée, France).

2.9. Quantification and Statistical Analysis

For comparative analysis across multiple groups, an ANOVA followed by Tukey's post hoc test was used to evaluate the effects of genotype and treatment. An unpaired Student's t‐test was used to compare the two groups. All statistical analyses were conducted using Prism 10 version 6.1 (GraphPad, RRID:SCR_021836), with a P‐value of less than 0.05 considered statistically significant. Animals were randomly divided into each group. Experimenters were not blinded to the treatment or genotypes of animals. Exclusion criteria were based on animal well‐being at the beginning of the study. No animal was excluded from this study. No power analysis was performed to determine the sample size. The sample size in each study was based on our experience with previous studies that employed knockout mice in our lab.

2.10. Declaration of AI‐Assisted Technologies

During the preparation of this work, the authors used an AI writing assistance tool, Grammarly, to improve the readability and language of the manuscript. After using this tool, the authors reviewed and edited the content as needed, taking full responsibility for the published article.

3. Results

3.1. Ketone Body Supplementation With 1,3‐BD Attenuates Adenine‐Induced Kidney Damage

Mice were allocated into three groups: A normal diet group, an adenine diet group, and an adenine diet supplemented with 1,3‐BD group (Figure 1A). 1,3‐BD is a synthetic diol that acts as a precursor to BHB and is frequently utilized to experimentally induce a ketogenic state [30]. Following a seven‐day dietary regimen, renal phenotypes were analyzed (Figure 1A). 1,3‐BD increased blood BHB levels and lowered blood glucose levels under fed conditions (8:00 a.m.) (Figure 1B,C). In contrast, after an eight‐hour fast (4:00 p.m.), 1,3‐BD showed no significant effect on either BHB or glucose levels (Figure 1B,C). The body weight loss associated with the adenine diet remained unaffected by 1,3‐BD treatment (Figure 1D).

FIGURE 1.

Ketone body supplementation with 1,3‐BD attenuates adenine‐induced kidney damage. (A) Experimental design to evaluate the effect of 1,3‐butanediol (1,3BD) on adenine‐induced nephropathy in mice. Mice were fed either a normal diet (ND; n = 12), an adenine diet (AD) without 1,3BD (n = 13), or an adenine diet supplemented with 1,3BD (n = 13). (B, C) Blood β‐hydroxybutyrate (BHB) concentrations (B) and blood glucose levels (C) at 8:00 a.m. and 4:00 p.m. in the three groups. (D) Body weight on day 7 of the dietary intervention. (E, F) Daily urinary albumin excretion (E) and plasma cystatin C levels (F) at the end of the study. (G) Hematoxylin and eosin (HE) staining of kidney sections and quantification of uninjured tubular area. Bar = 100 μm. (H) Experimental design to evaluate the effect of 1,3BD in ND‐fed wild‐type mice. Mice were fed ND without 1,3BD (n = 5) or with 1,3BD (n = 5). (I, J) Blood BHB concentrations (I) and blood glucose levels (J) in the two groups. (K) Plasma cystatin C levels at the end of the study. (L) HE staining of kidney sections. Bar = 100 μm. Data are presented as mean ± SEM. ns, not significant. *p < 0.05; **p < 0.01.

To evaluate renal phenotypes, urinary albumin excretion and plasma cystatin C levels were measured. While 1,3‐BD treatment did not influence adenine‐induced albuminuria (Figure 1E), it significantly decreased the adenine‐induced increase in plasma cystatin C levels (Figure 1F). Histological analysis revealed that adenine‐induced tubular damage, characterized by loss of brush borders and dilation of the tubular lumen in PTCs, as observed on HE staining, was significantly improved by 1,3‐BD treatment (Figure 1G).

To evaluate the potential renal toxicity of 1,3‐BD, mice were treated with 1,3‐BD for 7 days under normal diet conditions (Figure 1H). Under fed conditions, as observed in the adenine‐induced nephropathy model, 1,3‐BD treatment increased blood BHB levels and decreased blood glucose levels (Figure 1I,J). Plasma cystatin C levels and renal histology assessed by HE staining showed no detectable changes (Figure 1K,L), indicating no evidence of renal toxicity.

3.2. Ketone Body Supplementation With 1,3‐BD Attenuates Adenine‐Induced Renal Fibrosis, Inflammation, and Apoptosis

The effects of 1,3‐BD treatment on renal fibrosis, inflammation, and tubular cell apoptosis were evaluated. 1,3‐BD treatment significantly decreased the adenine diet‐induced increase in renal gene expression of fibrosis markers (transforming growth factor beta 1 (Tgfb1) and collagen, type I (Col1a1)), inflammatory markers (C‐C motif chemokine ligand 2 (Ccl2) and Adhesion G protein‐coupled receptor E1 (Adgre1) encoding MCP‐1 and F4/80, respectively), the apoptosis marker Bax, and proximal tubular cell damage marker Havcr1 encoding Kim‐1 (Figure 2A). Masson's trichrome staining also demonstrated that fibrosis in adenine nephropathy was ameliorated by 1,3‐BD treatment (Figure 2B). Additionally, 1,3‐BD treatment reduced the number of F4/80‐positive macrophages and TUNEL‐positive apoptotic cells (Figure 2C,D).

FIGURE 2.

Ketone body supplementation with 1,3‐BD attenuates adenine‐induced renal fibrosis, inflammation, and apoptosis. (A) Renal gene expression levels of fibrosis‐related markers (Tgfb1 and Col1a1), inflammatory markers (Ccl2 and Adgre1 encoding MCP‐1 and F4/80, respectively), apoptosis marker (Bax), and proximal tubular damage marker (Havcr1) in mice fed either a normal diet (ND), an adenine diet (AD), without 1,3‐butanediol (1,3BD), or an AD with 1,3BD. (B, C) Masson trichrome staining (B) and immunohistochemistry (IHC) for F4/80 (C) in the kidney sections in three groups. Bar = 50 μm. (D) TUNEL staining of kidney sections and quantification of TUNEL‐positive cells. Bar = 50 μm. Data are shown as mean ± SEM. ns, not significant. *p < 0.05; **p < 0.01. Adgre1, Adhesion G protein‐coupled receptor E1; Bax, Bcl‐2‐associated X protein; Ccl2, C‐C motif chemokine ligand 2; Col1a1, collagen type I alpha 1; Gapdh, Glyceraldehyde‐3‐phosphate dehydrogenase; Havcr1, Hepatitis A Virus Cellular Receptor 1; Tgfb1, Transforming growth factor beta 1.

3.3. Renal HMGCS2 Is Not Involved in the Progression of Adenine‐Induced Kidney Injury

We next investigated the role of renal HMGCS2‐mediated ketogenesis. Adenine diet increased mRNA expression levels of Hmgcs2 in the kidney, which was further enhanced by 1,3‐BD treatment (Figure 3A). In contrast, the adenine diet decreased mRNA expression levels of Oxct1, the gene responsible for ketone body utilization, while 1,3‐BD treatment had no effect (Figure 3B). The increase in renal HMGCS2 protein expression was confirmed by immunostaining (Figure 3C).

FIGURE 3.

Renal HMGCS2 is not involved in the progression of adenine‐induced kidney injury. (A, B) Renal gene expression of Hmgcs2 (A) and Oxct1 (B) in mice fed a normal diet (ND), an adenine diet (AD) without 1,3‐butanediol (1,3BD), or an AD supplemented with 1,3BD. (C) Immunohistochemistry for HMGCS2 in kidney sections of the three groups. Bar = 50 μm. (D) Study protocol: Control Hmgcs2^f/f mice (n = 11) and proximal tubule‐specific Hmgcs2 knockout (Hmgcs2^ΔPT) mice (n = 10) were fed an adenine diet (AD) for 7 days. (E, F) Blood β‐hydroxybutyrate (BHB) levels on day 1 (E) and plasma cystatin C levels (F) on day 7. (G) Hematoxylin and eosin (HE) staining of kidney sections and quantification of the uninjured tubular area. Bar = 100 μm. (H) Study protocol: Control Hmgcs2^f/f mice were fed either an AD (n = 9) or an AD supplemented with 1,3BD (n = 10); Hmgcs2^ΔPT mice were fed an AD with 1,3BD (n = 11). Western blotting for HMGCS2 and β‐actin in kidney samples from three groups. (I, J) Blood BHB levels on day 1 (I) and plasma cystatin C levels on day 7 (J). (K) HE staining of kidney sections and quantification of the uninjured tubular area. Bar = 100 μm. Data are presented as mean ± SEM. ns, not significant. *p < 0.05; **p < 0.01. Gapdh, Glyceraldehyde‐3‐phosphate dehydrogenase; Hmgcs2, 3‐hydroxy‐3‐methylglutaryl‐CoA synthase 2; Oxct1, Succinyl‐CoA:3‐ketoacid CoA transferase 1.

To explore the role of renal HMGCS2 overexpression in the progression of kidney injury, we created mice with PTC‐specific deletion of the Hmgcs2 gene (Hmgcs2^ΔPT). These mice, along with control Hmgcs2^f/f mice, were fed an adenine‐containing diet for seven days, after which renal function and histopathology were evaluated (Figure 3D). Because ketogenesis is not very active under ad libitum feeding, we used an intermittent fasting schedule with 8 h of fasting each day to boost endogenous ketogenesis. After 8 h of fasting, blood ketone body levels remained unchanged in Hmgcs2^ΔPT mice compared to wild‐type controls (Figure 3E). Also, there were no differences in blood cystatin C levels and histological damage between control Hmgcs2^f/f mice and Hmgcs2^ΔPT mice (Figure 3F,G).

The HMGCS2 expression in the kidney was upregulated in the 1,3‐BD‐treated group (Figure 3A,C,H), which led us to hypothesize that 1,3‐BD may exert renoprotective effects by enhancing renal ketogenesis. To test this hypothesis, we examined whether the renoprotective effect of 1,3‐BD would be abolished in Hmgcs2^ΔPT mice (Figure 3H). Renal overexpression of HMGCS2 induced by 1,3‐BD was inhibited in Hmgcs2^ΔPT mice (Figure 3H). Although 1,3‐BD treatment increased circulating BHB levels, no significant difference was observed between control mice and Hmgcs2^ΔPT mice (Figure 3I). Furthermore, while 1,3‐BD treatment significantly improved both serum cystatin C levels and histological tubular injury in the kidney, there were no significant differences in these outcomes between control mice and Hmgcs2^ΔPT mice (Figure 3J,K). These findings suggest that renal HMGCS2 and ketogenesis are not involved in the progression of kidney injury and 1,3‐BD‐mediated renoprotection in adenine‐induced kidney injury.

3.4. Systemic HMGCS2‐Mediated Ketogenesis Is Not Involved in the Progression of Adenine‐Induced Kidney Injury

Given that the elevation of circulating ketone bodies induced by 1,3‐BD has demonstrated renoprotective effects, we subsequently investigated whether the deficiency of endogenous hepatic ketogenesis would exacerbate renal injury. To this end, we generated liver‐specific Hmgcs2 knockout (Hmgcs2^ΔLiver) mice and fed them an adenine‐rich diet (Figure 4A). Hepatic deletion of HMGCS2 expression in Hmgcs2^ΔLiver mice was confirmed (Figure 4A). Although plasma BHB levels were markedly reduced in Hmgcs2^ΔLiver mice (Figure 4B), the increases in plasma cystatin C levels and tissue injury attributable to the adenine diet were not augmented (Figure 4C,D). Moreover, to exclude the possibility that HMGCS2 expression in other tissues may contribute to renoprotection, we also evaluated whole‐body Hmgcs2 knockout (Hmgcs2^Δwhole) mice on an adenine diet (Figure 4E). Hepatic and renal deletion of HMGCS2 expression in Hmgcs2^Δwhole mice was confirmed (Figure 4E). Consistent with the findings in Hmgcs2^ΔLiver mice, the increase in plasma cystatin C and tissue injury induced by the adenine diet did not deteriorate (Figure 4F–H).

FIGURE 4.

Systemic HMGCS2‐mediated ketogenesis is not involved in the progression of adenine‐induced kidney injury. (A) Study protocol: Control Hmgcs2^f/f mice (n = 8) and liver‐specific Hmgcs2 knockout (Hmgcs2^ΔLiver) mice (n = 10) were fed an adenine diet (AD) for 7 days. Western blotting for HMGCS2 and β‐actin in liver samples from two groups. (B, C) Blood ketone body levels on day 1 (B) and plasma cystatin C levels on day 7 (C). (D) Hematoxylin and eosin (HE) staining of kidney sections and quantification of the uninjured tubular area. Bar = 100 μm. (E) Study protocol: Control Hmgcs2^f/f mice (n = 10) and whole‐body Hmgcs2 knockout (Hmgcs2^ΔWhole) mice (n = 10) were fed an AD for 7 days. Western blotting for HMGCS2 and β‐actin in liver and kidney samples from two groups (F, G). Blood ketone body levels on day 1 (F) and plasma cystatin C levels on day 7 (G). (H) HE staining of kidney sections and quantification of the uninjured tubular area. Bar = 100 μm. Data are presented as mean ± SEM. ns, not significant. **p < 0.01.

3.5. Lack of OXCT1‐Mediated Ketolysis Cancels 1,3‐Butanediol‐Mediated Renoprotection in Adenine‐Induced Kidney Injury

To investigate the role of renal ketolysis in the progression of kidney injury, we generated PTC‐specific Oxct1 knockout (Oxct1^ΔPT) mice. Control Oxct1^f/f mice and Oxct1^ΔPT mice were fed an adenine diet for 7 days, and kidney injury was assessed (Figure 5A). A significant decrease in renal mRNA expression of Oxct1 in adenine‐fed Oxct1^ΔPT mice was confirmed (Figure 5B). Oxct1^ΔPT mice showed no significant increase in adenine‐induced kidney damage, nor did blood BHB levels change (Figure 5C–E).

FIGURE 5.

Lack of OXCT1‐mediated ketolysis cancels 1,3‐butanediol‐mediated renoprotection in adenine‐induced kidney injury. (A) Study protocol: Control Oxct1^f/f mice (n = 9) and proximal tubule‐specific Oxct1 knockout (Oxct1^ΔPT) mice (n = 12) were fed an adenine diet (AD) for 7 days. (B) mRNA of Oxct1, normalized to mRNA of Gapdh, in kidney samples from two groups. (C, D) Blood β‐hydroxybutyrate (BHB) concentrations on day 1 (C) and plasma cystatin C levels on day 7 (D). (E) Hematoxylin and eosin (HE) staining of kidney sections and quantification of the uninjured tubular area. Bar = 100 μm. (F) Study protocol: Control Oxct1^f/f mice were fed either an AD (n = 9) or an AD supplemented with 1,3‐butanediol (1,3BD) (n = 9); Oxct1^ΔPT mice were fed an AD supplemented with 1,3BD (n = 12). (G, H) Blood glucose levels (G) and blood BHB concentrations (H) at 8:00 a.m. and 4:00 p.m. in the three groups. (I) Plasma cystatin C levels at the end of the study. (J) HE staining of kidney sections, along with quantification of uninjured tubular area. Bar = 50 μm. Data are presented as mean ± SEM. ns, not significant. *p < 0.05; **p < 0.01. Gapdh, Glyceraldehyde‐3‐phosphate dehydrogenase; Oxct1, Succinyl‐CoA:3‐ketoacid CoA transferase 1.

Finally, to examine whether OXCT1‐mediated ketolysis in PTCs contributes to the renoprotective effects of 1,3‐BD, Oxct1^f/f mice and Oxct1^ΔPT mice were fed an adenine diet containing 1,3‐BD. (Figure 5F). While blood glucose levels at 8 a.m. were lower in Oxct1^ΔPT mice treated with 1,3‐BD compared to non‐treated Oxct1^f/f mice, there was no difference in levels between Oxct1^f/f and Oxct1^ΔPT mice under 1,3‐BD treatment (Figure 5G). PTC‐specific OXCT1 deficiency did not affect the elevation of blood BHB levels induced by 1,3‐BD (Figure 5H). In Oxct1^ΔPT mice, the 1,3‐BD‐mediated improvement in plasma cystatin C levels tended to be abolished (Figure 5I). Additionally, the beneficial effect of 1,3‐BD on tubular injury assessed by HE staining was significantly reduced in Oxct1^ΔPT mice.

3.6. Lack of OXCT1‐Mediated Ketolysis Cancels 1,3‐Butanediol‐Mediated Anti‐Fibrotic, Anti‐Inflammatory, and Anti‐Apoptotic Effects in Adenine‐Induced Kidney Injury

Furthermore, adenine diet‐induced renal fibrosis (Masson's trichrome staining), inflammation (F4/80 immunohistochemistry), and apoptosis (TUNEL staining) were significantly ameliorated by 1,3‐BD treatment, and these effects were attenuated by PTC‐specific OXCT1 deficiency (Figure 6A–C). These findings suggest that enhanced ketolysis via OXCT1 in PTCs contributes to the renoprotective effects of 1,3‐BD. Additionally, the 1,3‐BD‐mediated increase in renal Kbhb was not enhanced in Oxct1^ΔPT mice (Figure 6D), suggesting that the post‐translational protein modification, Kbhb, does not appear to be involved in the cancellation of 1,3‐BD–mediated renoprotection.

FIGURE 6.

Lack of OXCT1‐mediated ketolysis cancels 1,3‐butanediol‐mediated anti‐fibrotic, anti‐inflammatory, and anti‐apoptotic effects in adenine‐induced kidney injury. (A, B) Representative images of Masson trichrome staining (A) and immunohistochemistry (IHC) for F4/80 (B) in control Oxct1^f/f mice were fed either an adenine diet (AD) or an AD supplemented with 1,3‐butanediol (1,3BD); Oxct1^ΔPT mice were fed an AD supplemented with 1,3BD. Bar = 100 μm. (C) TUNEL staining of kidney sections and quantification of TUNEL‐positive cells. Bar = 100 μm. (D) Representative Western blot images showing pan‐lysine β‐hydroxybutyrylation (Kbhb) levels in kidney samples from three groups of mice. Data are presented as mean ± SEM. ns, not significant. *p < 0.05; **p < 0.01.

4. Discussion

This study demonstrates that exogenous ketone body supplementation with 1,3‐BD could be a potential therapeutic strategy for CKD. Utilizing various genetically modified mice with altered ketone metabolism in the whole body, liver, or kidney, the findings indicate that endogenous ketogenesis contributes less to the severity of renal injury. On the other hand, increasing circulating ketone levels through exogenous supplementation provides renal protection. Mechanistically, this protective effect is partly due to OXCT1‐dependent ketolysis in PTCs.

Recent research suggests that targeting elevated circulating ketone body levels holds therapeutic potential for kidney disease. Improvements in renal injury have been observed in various animal models, including ischemic AKI [21], cisplatin‐induced renal injury [22], DKD [19, 31], and PKD [20]. This study offers additional evidence emphasizing the importance of ketone body supplementation in CKD. Clinically, the application of ketone body supplementation in PKD is currently under investigation, and the safety profile of ketone esters has been documented [32]. Future research on the effectiveness of ketone body supplementation in managing human CKD could lead to the development of a new therapeutic approach.

The question remains as to why exogenous administration of ketone bodies confers a renoprotective effect, whereas the deficiency of endogenous systemic ketogenesis in Hmgcs2^ΔWhole and Hmgcs2^ΔLiver does not worsen adenine‐induced renal injury. We hypothesize that one contributing factor is the variance in ketone body concentrations. During active intake of a diet containing 1,3‐BD, blood BHB levels in mice increase to approximately 1.5 mM. In mice fed an adenine diet, blood BHB levels only rise to 0.7–0.8 mM. In the adenine nephropathy model, body weight decreases after adenine administration, leading to a lack of fat reserves that serve as substrates for ketogenesis. As a result, this model may not be ideal for understanding the physiological importance of endogenous ketogenesis in kidney diseases. Further research using various experimental models is needed to determine whether endogenous HMGCS2‐mediated ketogenesis has a protective role in kidney health.

In the field of ketone body research, increasing attention has been directed toward extrahepatic ketogenesis and its organ‐specific functions. In the adenine nephropathy model, we observed increased HMGCS2 expression in PTCs, suggesting that local renal ketogenesis may contribute to renoprotection. However, PTC‐specific Hmgcs2 deletion did not exacerbate renal injury, suggesting that renal HMGCS2 induction is not sufficient to confer protection in this model. Alternatively, this finding may reflect a model‐specific limitation of adenine nephropathy, in which early and pronounced weight loss depletes adipose tissue, thereby limiting the substrates required for effective ketogenesis. In addition, because the present study focused on short‐term kidney injury, it remains possible that PTC‐specific HMGCS2 deficiency may exert more substantial effects over longer disease durations. Therefore, further validation using alternative CKD models and/or longer observation periods in the adenine nephropathy model will be necessary to clarify the precise role of renal HMGCS2 expression in CKD progression.

While multiple mechanisms may contribute to the renoprotective effects of ketone bodies, our findings specifically highlight the importance of the enhanced ketolysis via OXCT1 in 1,3‐BD‐mediated renoprotection. However, it is also notable that the reduction in the therapeutic effect of 1,3‐BD in Oxct1^ΔPT mice was significant but only partial. Two factors may explain this. First, OXCT1 is present not only in PTCs but also in other nephron segments, suggesting that OXCT1‐dependent improvements in energy metabolism across various renal cell types could contribute to partial protection. Second, energy‐independent mechanisms might also be involved. Recent studies indicate that ketone bodies, especially BHB, serve not only as ATP sources but also as regulators of transcription and signaling via BHB‐mediated post‐translational protein modification [9, 11]. Indeed, Kbhb was increased in the kidneys of 1,3‐BD–treated mice, and its involvement in renoprotection cannot be ruled out. However, no additional changes in Kbhb were detected in Oxct1^ΔPT mice, suggesting that this modification is unlikely to account for the reduced protective effects of 1,3‐BD in Oxct1^ΔPT mice.

Several limitations exist within this study. As previously mentioned, validation was performed exclusively using the adenine nephropathy model; it is crucial to test other kidney disease models, particularly those not associated with body weight loss. Additionally, the study was limited to male mice, leaving it unclear whether females would show similar results. The observation that PTC‐specific OXCT1 deletion weakened the renoprotective effect of 1,3‐BD suggests that ketolysis‐dependent ATP production makes a substantial contribution to renoprotection. However, to establish this mechanism more definitively, additional approaches such as real‐time in vivo imaging and ¹³C‐BHB tracing will be required. Such studies would enable a more thorough understanding of the specific mechanisms by which ketone bodies exert renoprotective effects and would greatly facilitate the development of future kidney disease therapeutics targeting ketone body metabolism.

In conclusion, our findings indicate that external ketone body supplementation protects the kidneys in CKD models. They also offer new mechanistic insights into how renal ketone metabolism contributes to organ protection, indicating that ketone body–based treatments could be a promising therapeutic strategy for CKD.

Author Contributions

S. Omachi, S. Sugahara, and S. Kume designed the study. S. Omachi and J. Horiguchi performed the experiments. K. Kufukihara and J. Nakahara generated the Hmgcs2‐flox mouse model. T. Kusaba and B.D. Humphreys established the SLC34a1‐CreER mouse model. S. Omachi, S. Kume, S. Sugahara, M. Yasuda‐Yamahara, K. Yamahara, S. Ida , S. Kuwagata, Y. Tanaka‐Sasaki, and M. Chin‐Kanasaki analyzed and discussed the data. S. Omachi, S. Sugahara, M. Chin‐Kanasaki, and S. Kume drafted and revised the manuscript. All authors reviewed and approved the final version.

Funding

Grants‐in‐Aid for Scientific Research (KAKENHI) from the Japan Society for the Promotion of Science [21H03353, 23 K21619, 22689028, 25 K03027]. JST [Moonshot R& D; Grant No. JPMJMS2023]. Naito Foundation. Astellas Foundation for Research on Metabolic Disorders. Terumo Foundation for Life Sciences and Arts. This work was supported by MEXT | Japan Society for the Promotion of Science (JSPS), grants 21H03353, 23K21619, 22689028, 25K03027. MEXT | Japan Science and Technology Agency (JST), Moonshot R& D; JPMJMS2023. Naito Foundation (x5185; x85e4; x8A18; x5ff5; x79D1; x5b66; x632F; x8208; xxx8CA1; x56E3). Astellas Foundation for Research on Metabolic Disorders. Terumo Foundation for Life Sciences and Arts.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

All data generated or analyzed during this study are included in this published article.