Reprogramming of the kynurenine pathway impairs NAD⁺ homeostasis and mediates doxorubicin-induced cardiotoxicity in mice

Abstract

Imbalance of Nicotinamide adenine dinucleotide (NAD⁺) homeostasis is a key contributor to various cardiac pathologies, including doxorubicin (DOX)-induced cardiomyopathy (DIC). The kynurenine pathway (KP), initiated by indoleamine 2,3-dioxygenase 1 (IDO1), serves as the primary route for de novo NAD ⁺ biosynthesis. While this pathway regulates critical biological processes such as cellular metabolism, inflammatory responses, oxidative stress, and aging, its specific role in DIC remains poorly understood. Here, we reveal a protective function of the KP in DIC by facilitating NAD⁺ synthesis. Genetic ablation of IDO1 exacerbates DOX-induced cardiac injury and structural damage in mice. In cardiomyocytes, DOX treatment upregulates α-amino-β-carboxy-muconate-semialdehyde decarboxylase (ACMSD) while downregulating quinolinate phosphoribosyl-transferase (QPRT), thereby reducing levels of the intermediate metabolite quinolinic acid (QA) and NAD⁺ levels. These effects can be pharmacologically reversed by TES-1025, an ACMSD inhibitor that enhances QPRT activity and potentiates the cardioprotective effects of the KP pathway against DIC. Mechanistically, we show that DOX modulates the STING/interferon γ/5′-AMP-activated protein kinase (p-AMPK) signaling axis to elevate ACMSD and suppress QPRT. Our findings establish a novel therapeutic potential that targets the metabolic switch ACMSD to QPRT, restoring NAD⁺ redox homeostasis and conferring protection against DIC in murine models.

Graphical abstract

Highlights

• •The de novo NAD + synthetic pathway protects cardiac injury and dysfunction induced by DOX.
• •DOX decreases QA and NAD + pool by elevating ACMSD activity.
• •DOX mediates ACMSD activation by STING/IFN-γ/AMPK axis.
• •Reprogramming of ACMSD/QPRT switch can attenuate the oxidative stress in DIC, while it does not attenuate DOX-induced cancer cell death.
• •Targeting ACMSD may as a therapeutic strategy for cardiomyopathies.

1. Introduction

Doxorubicin (DOX), an anthracycline antibiotic widely used as a chemotherapeutic agent, has significant anti-cancer effects; however, its clinical application is limited due to its cardiotoxicity [1]. DOX-induced cardiomyopathy (DIC) involves numerous pathological mechanisms, such as oxidative stress, inflammatory responses, apoptosis, calcium (Ca²⁺) dyshomeostasis, and dysregulated autophagy [2]. Currently, no approved therapies specifically prevent or treat DIC, highlighting a critical unmet need to protect cancer patients from cardiotoxicity and improve long-term survivor outcomes [3]. Nicotinamide adenine dinucleotide (NAD⁺) plays a pivotal role in mitigating DIC, and supplementation with its precursors is an effective therapeutic strategy [4]. As a vital cofactor and signaling molecule, NAD⁺ is essential for redox homeostasis, energy metabolism (as NADH), and as a substrate for sirtuins and poly-ADP-ribose-polymerases (PARPs) [5,6]. Part of this protective effect is mediated by the NAD⁺ -dependent protein deacetylase sirtuin-1 (SIRT1), a key regulator of the antioxidant response proteins, including nuclear factor erythroid 2-related factor 2 (Nrf2) and superoxide dismutase 2 (SOD2), as demonstrated in our previous work and supported by other studies [7,8].

Cellular NAD⁺ levels are sustained primarily through two biosynthetic pathways. The salvage pathway, mediated by nicotinamide phosphoribosyltransferase (NAMPT) and NMN adenylyltransferases (NMNATs), efficiently recycles nicotinamide (NAM) into NAD⁺ [9]. Alternatively, the de novo pathway synthesizes NAD⁺ from tryptophan (Trp), a process initiated by indoleamine 2,3-dioxygenase 1 (IDO1) and tryptophan-2,3-dioxygenase (TDO) to form kynurenine (Kyn) [10]. Approximately 95 % of Trp is metabolized into Kyn by this route. Kyn metabolism is a critical juncture. It is primarily processed by the enzymes kynurenine-3-monooxygenase (KMO), kynureninase (KYNU), and 3-hydroxyanthranilate 3,4-dioxygenase (3′-HAAO) to generate quinolinic acid (QA). Then QA is converted into NAD⁺ precursors by quinolinate phosphoribosyltransferase (QPRT). Alternatively, α-amino-β-carboxymuconate-ε-semialdehyde decarboxylase (ACMSD) can divert QA away from NAD⁺ synthesis and into the tricarboxylic acid (TCA) cycle. However, during inflammation, kynurenine aminotransferases (KATs) shift Kyn metabolism toward kynurenic acid (KynA), potentially reducing NAD⁺ synthesis [10] (Fig. 3A).

Fig. 3.

The kynurenine pathway links NAD + metabolism and the TCA cycle. A, Schematic diagram of the KP pathway and its metabolites. B, Western blotting results of SIRT1, as well as KP pathway enzymes including IDO1, KYNU, HAAO, ACMSD and QPRT in the WT and IDO1KO cardiomyocytes treated with DOX, and β-actin as load control (n = 3). C–H, Quantification of (B). The protein level was standardized by β-actin and the value in WT-CTL group was designated as 1. I, IF staining of IDO1, ACMSD and QPRT in cardiomyocytes with or without DOX treatment. J, NAD⁺ levels of each group (n = 5). K, ATP levels of each group (n = 5). The values are presented as means ± SD. Statistical analysis was performed by two-way ANOVA with Tukey's multiple comparisons test (C–H, J, K). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Adjusted P values were provided in the case of multiple groups. NAD⁺, nicotinamide adenine dinucleotide; TCA, tricarboxylic acid cycle; KP, kynurenine pathway; DOX, doxorubicin; KYNU kynureninase; HAAO, 3-hydroxyanthranilate 3,4-dioxygenase; ACMSD, α-amino-β-carboxy-muconate-semialdehyde decarbo-xylase; QPRT, quinolinate phosphoribosyl-transferase. SIRT1, sirtuin 1.

Given the essential role of NAD⁺ as a cofactor in vital biological processes, the maintenance of its cellular levels is critical for cardiac function. NAD⁺ depletion is a recognized feature of cardiovascular pathologies, including DIC and dilated cardiomyopathy (DCM) [5]. Consequently, elevating NAD⁺ has emerged as a promising therapeutic strategy. Three principal approaches have been identified: 1) inhibition of NAD⁺ degradating (e.g., via NAD⁺-consuming enzymes like CD38 or SARM1), 2) blocking ACMSD to promote the de novo route, and 3) supplementating with NAD⁺ precursors such as Trp, nicotinic acid (NA), nicotinamide riboside (NR), NAM, and β-nicotinamide mononucleotide (NMN) [11]. However, the precursor efficacy varies. NR and NMN typically only double NAD⁺ levels, limiting their clinical impact [12], while the NAM-related salvage pathway is constrained by NAMPT saturation [9,13]. In contrast, Trp via the KP pathway can induce superior NAD⁺ biosynthesis, especially under oxidative stress or organ injury (e.g., kidney and liver injury), as it operates independently of these limitations [14].

Normal cells maintain NAD ⁺ homeostasis through the complementary salvage and de novo biosynthesis pathways, and the dysregulation of each is linked to distinct diseases. Evidence suggests a specific role for the de novo pathway in cardiovascular health. For instance, in patients with heart failure with preserved ejection fraction (HFpEF), adverse cardiac remodeling correlates with reduced levels of de novo metabolites like 3-HAA and QA [15]. Furthermore, 3-HAA can reverse atherosclerosis induced by IDO1 inhibition [16], and Kyn administration lowers blood pressure in hypertensive models [17]. In contrast, many cancers are highly dependent on the salvage pathway while downregulating the de novo pathway, making them vulnerable to NAMPT inhibitors [18]. However, this approach systemically depletes NAD⁺ in both tumors and normal tissues like the heart, limiting its clinical utility. While supplementation with NAD⁺ precursors (NR or NMN) can boost the NAD⁺ pool and protect against DIC, it carries the risk of promoting tumor growth and chemotherapy resistance [19]. Therefore, selectively reactivating the de novo pathway could be a promising strategy to protect against DIC without stimulating tumor growth. Despite this compelling rationale, the specific role and therapeutic potential of the de novo NAD ⁺ biosynthesis pathway in protecting against DIC remains completely unexplored.

In this study, we demonstrated that the KP pathway of de novo NAD ⁺ biosynthesis exerts a protective effect against DIC. Genetic ablation of IDO1 (IDO1KO) in mice reduced NAD⁺ levels, increased reactive oxygen species (ROS) levels in cardiomyocytes, and exacerbated cardiac fibrosis and impaired ejection fraction in DOX-treated hearts. The same cardiac damage phenotype was also consistently observed in a model of cardiac-specific IDO1 knockdown, demonstrating that KP in cardiomyocytes intrinsically protects against DIC. Furthermore, we found that DOX disrupts the KP by upregulating IDO1, KYNU, 3′-HAAO, and ACMSD while downregulating QPRT, thereby diverting KP metabolites toward the TCA cycle and depleting the NAD⁺ pool. Pharmacological inhibition of ACMSD using TES-1025 restored NAD⁺ levels and improved cardiac function in our DIC model. Importantly, TES-1025 did not compromise the anti-tumor efficacy of DOX in cancer cells. Collectively, our findings reveal, for the first time, the IDO1-derived KP pathway as a crucial endogenous protective mechanism in cardiac injury and identify ACMSD inhibition as a novel and promising therapeutic strategy for preventing cardiomyopathies.

2. Methods

2.1. Mice

C57BL/6j male mice (aged 8 weeks) were obtained from the Experimental Animal Center of Basic Medicine, Zhejiang Chinese Medical University. Age- and sex-matched IDO1 gene-knockout (Ido1^−/−) male mice with a C57BL/6j background (Ido1^tm1Alm) were procured from the Jackson Laboratory (Bar Harbor, ME, USA). All mice were maintained under a 12-h light/dark cycle in a standard pathogen-free environment with ad libitum access to food and water. Following a 1-week acclimatization period, both wild-type (WT) and Ido1^−/− mice were randomly allocated into three groups (n = 9 per group): Control group (CTL), DOX group (DOX), and DOX + TES-1025 (DOX + TES) group. For the cardiac-specific IDO1 knockdown, an additional cohort of thirty (8-week-old) male C57BL/6J mice were randomly divided into two groups (n = 15 per group) receiving either AAV9 carrying non-targeting control shRNA (AAV9-NC-Ido1) or AAV9 expressing IDO1-specific shRNA (AAV9-shRNA-Ido1) via tail vein injection at a dose of 1 × 10^∧12 viral genomic titles per mouse. After 3 weeks of viral transduction to allow maximal cardiac-specific knockdown, these mice were further subdivided into the same three treatment groups as above (n = 5 per subgroup). DOX was administered at a cumulative dose of 15 mg/kg via tail vein injection once weekly (on day 7 of each week) for 6 weeks, as previously described [7]. TES-1025 was administered intraperitoneally at a dose of 15 mg/kg. All animal experiments were conducted in compliance with the National Institutes of Health (NIH) guidelines for the care and use of laboratory animals and were approved by the Ethics Committee for Animal Experiments of Taizhou Hospital (approved ID number: TZYY

2022044 and TZYY2025106). 6 weeks after the initial DOX injection, mice were deeply anesthetized via inhalation of 5 % isoflurane and euthanized by exsanguination, which was performed by cannulating the ascending aorta and creating an incision in the right atrium (RA), followed by slow perfusion with about 10 mL of ice-cold PBS. Heart tissue samples were then collected for further analysis.

2.2. Echocardiography

Cardiac function was evaluated using transthoracic echocardiography on the 7th day following the final DOX tail vein injection. Anesthesia was induced with 3 % isoflurane and maintained with 1 % isoflurane inhalation. Echocardiographic measurements were performed using the Philips iE33 system (Philips Medical, Best, Netherlands) equipped with a 12–14 MHz linear array ultrasound transducer. Parameters including left ventricular ejection fraction (LVEF), left ventricular fractional shortening (LVFS), left ventricular internal diastolic diameter and left ventricular internal systolic diameter (LVIDd and LVIDs), end-diastolic and end-systolic left ventricular posterior wall thickness (LVPWd and LVPWs), end-diastolic and end-systolic interventricular septal thickness (IVSd and IVSs) were recorded. All images were analyzed by an investigator blinded to the experimental groups.

2.3. Histopathology

Heart tissues fixed in 10 % buffered formalin were embedded in paraffin and sectioned at a thickness of 5 μm. Following deparaffinization, dehydration, and rehydration, the sections were subjected to hematoxylin-eosin (HE) staining (Cat# G1120, Sbjbio Life Sciences, China) to evaluate changes in cardiac morphological structure and Masson trichrome staining (Cat# G1340-7, Sbjbio Life Sciences, China) to assess the degree of myocardial fibrosis. Images were captured using a light microscope (Olympus CKX41) at 200× magnification, with five random fields photographed per slide for histopathological analysis.

2.4. Cell isolation/culture and treatment

Primary cardiomyocytes were isolated from 3-day-old neonatal wild-type (WT) and IDO1 knockout (Ido1^−/−) mice. All procedures were conducted in compliance with the National Institutes of Health (NIH) guidelines for the care and use of laboratory animals. Neonatal mice were anesthetized with 5 % isoflurane until fully unconscious, sterilized with 70 % ethanol, and euthanized using sharp surgical scissors on a sterilized 10 cm plastic dish. Hearts were promptly excised, and primary cardiomyocytes were isolated using the Pierce Primary Cardiomyocyte Isolation Kit (Cat# 88281, Thermo Fisher Scientific, USA) following the manufacturer's instructions [20]. The isolated cardiomyocytes and MCF7 cells were cultured in Dulbecco's Modified Eagle Medium (DMEM, Sigma, USA) supplemented with 10 % fetal bovine serum (Gibco, USA), 1 % penicillin-streptomycin (PS), and maintained in a humidified incubator at 37 °C with 5 % CO2. Primary cardiomyocytes were treated with DOX (1 μM), TES-1025 (10 μM), NR (250 μM), NMN (1 mM), IFN-γ protein (10 ng/mL), or AICAR (500 μM), as specified. MCF7 cells were treated with DOX, TES-1025, NR, and NMN.

2.5. Cell viability assay

Cell viability was assessed using the Cell Counting Kit-8 (CCK-8) assay (Cat# HY-K0301, MedChem Express, China), as previously described [7]. Cells were seeded at a density of 1 × 10^∧4 cells per well in 96-well plates. Following 24 h of treatment with the specified drug interventions, cells were incubated with the CCK-8 reagent diluted in complete DMEM (1:10 ratio) at 37 °C for 1 h. Absorbance was measured at 450 nm using a microplate reader (Thermo Scientific™ Multiskan™ FC) to determine cell viability.

2.6. Measurement of NAD⁺

The NAD⁺ in cells and cardiac tissues was quantified using the Amplite® Colorimetric NAD/NADH Ratio Assay Kit (Cat# 15273, AAT Bioquest, Sunnyvale, CA) following the manufacturer's protocol. Briefly, cells were harvested from 6-well plates after 24 h of drug treatment, and cardiac tissues were collected. Cells and homogenized cardiac tissues were lysed in lysis buffer at 37 °C for 15 min, and the supernatant was collected as test samples. Test samples were separately incubated with NAD Extraction Solution (to measure NAD levels) at 37 °C for 15 min, followed by the addition of Neutralization Solution and Extraction Control Solution, respectively. NAD/NADH working solution was then added to the mixtures and incubated for 15 min to 2 h at room temperature in the dark. Absorbance was measured at 460 nm using a microplate reader.

2.7. Detection of ATP levels

Following drug treatment, cardiomyocytes and heart tissues were sonicated and homogenized in the lysis buffer. The supernatant from each sample was collected and incubated with 100 μL of ATP detection reagent at room temperature for 3–5 min, according to the manufacturer's protocol (Cat# S0026, Beyotime, China). ATP levels were quantified using a GloMax® 20/20 Luminometer and normalized to the total protein concentration.

2.8. Reactive oxygen species (ROS) assay

Intracellular ROS levels of cardiomyocytes were assessed using a commercial ROS assay kit (Cat# E004-1-1, Nanjing Jiancheng Bioengineering Institute, China). DOX-treated cardiomyocytes were incubated with 2′,7′-dichlorofluorescein diacetate (DCFH-DA) for 45 min in the dark, then the ROS levels were measured at 525 nm excitation using a microplate reader. For tissues ROS detection, fresh heart tissue sections were incubated with dihydroethidium (DHE, 10 μM, Invitrogen, D11347) at 37 °C for 15 min, the red fluorescence of superoxide anion-induced DHE oxidation products was visualized by fluorescence microscopy (Nikon Eclipse Ti2).

2.9. Mitochondrial respiration and glycolytic rate assay

Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were assessed to quantify mitochondrial respiration and glycolysis by using the Seahorse XF Cell Mito Stress Test and Glycolytic Rate Assay Kits (Agilent, Cat# 103015–100 and 103344–100), following the standardized protocols. In preparation for the assay, cells were seeded at a density of 1 × 10^∧4 cells/well in XF96 microplates (Agilent) 48 h before measurement. After 24 h of plating, cells were treated with either DOX, TES-1025 alone, or their combination for an additional 24-h incubation period. Then the culture medium was replaced with Seahorse XF Base Medium (pH 7.4) for 30 min in a 37 °C, non-CO2 incubator to allow temperature and pH stabilization. The mitochondrial stress test was then performed through sequential injection of pharmacological agents: oligomycin (1.5 μM) to assess ATP-linked respiration, FCCP (2 μM) to measure maximal respiratory capacity, and rotenone/antimycin A (0.5 μM each) to determine non-mitochondrial oxygen consumption. For the glycolytic rate assay, ECAR was measured after the injection of glucose (10 mM), oligomycin (2 μM), and 2-deoxy-d-glucose (2-DG, 100 mM). All measurements were conducted using the Seahorse XFe96 Analyzer and data were analyzed using wave software (Agilent Technologies, v2.6).

2.10. LC-MS/MS assay

KP pathway metabolites of interest Trp, Kyn, KA, 3HAA and QA in heart tissues were detected using a targeted liquid chromatography-tandem mass spectrometry (LC-MS/MS) instrument (AB Sciex Triple Quad™ 4500 mass spectrometer, 4500 QQQ, Applied Biosystems Inc., USA). Data were collected and analyzed on Analyst software 1.6 as previously described [21]. Briefly, l-tryptophan-d5, l-Kynurenine-d4, Kynurenic acid-d5, 3-Hydroxyanthranilic acid-d3, and Quinolinic acid-d3 (all those reagents were purchased from MCE, China) were used as reference standards. Mouse heart tissues (about 50 mg) were used for metabolite analysis. After protein precipitation, the supernatant was collected and speed vacuum dried, then reconstituted by adding 50 μL H₂O with 0.1 % formic acid. To track the metabolic flux dynamics of the kynurenine pathway following ACMSD inhibition, cardiomyocytes were incubated with ¹³C-tryptophan (MCE, China). Metabolites were extracted at designated time points (6, 12, and 24 h). Cells were washed twice with ice-cold PBS and resuspended with 800 μL of a pre-chilled methanol: water mixture (1:1, v/v) and sonicated on ice. The homogenate was centrifuged (10 min, 16,000 × rpm, 0 °C), and the supernatant was retained. The resulting pellet was subjected to a second extraction with 600 μL of methanol: water (2:1, v/v). The supernatants were combined, mixed with 600 μL of ice-cold chloroform, and centrifuged to induce phase separation. The upper polar phase was collected, evaporated to dryness under vacuum, and the metabolite pellet was reconstituted in 70 μL of methanol for LC-MS ¹³C-citrate analysis. The samples were run through a Synergi Hydro-RP column (50 mm × 2.0 mm, 2.5 μm; Phenomenex, USA) with buffer A (H₂O and 0.2 % formic acid, 5 mM ammonium formate) and buffer B (acetonitrile). The setting of buffer A was at 0.00–1.00 min, 15 % buffer B at 4.00 min, then increased buffer B to 95 % at 7.00–8.00 min, then decreased to 5 % at 8.50 min through 10.0 min.

2.11. Transmission electron microscopy (TEM) assay

Left ventricular tissues from different mouse groups were rapidly isolated and fixed in 2.5 % glutaraldehyde for over 3 h at 4 °C. After three washes with PBS, the samples were post-fixed in 1 % osmium tetroxide for 2 h, followed by dehydration through a graded acetone series and embedding in epoxy resin. The embedded samples were polymerized at 70 °C for 48 h, sectioned into 70 nm slices, and stained with uranyl acetate (UA) and lead citrate. Ultrastructural images were acquired using a transmission electron microscope (Hitachi HT7800, Tokyo, Japan).

2.12. Measurement of intracellular mitochondrial superoxide

Mitochondrial superoxide (mitoROS) levels were detected using the MitoSOX™ Red (Cat# M36008, Thermo Fisher, USA) staining kit according to the manufacturer's protocol. Briefly, after 24 h of treatment with DOX, TES-1025, or their combination, cardiomyocytes were incubated with 5 μM MitoSOX™ Red reagent for 15 min at 37 °C in the dark. Cells were washed twice with PBS, and then the images of the cells were acquired using a Zeiss LSM800 confocal microscope.

2.13. Measurement of mitochondrial membrane potential (MMP)

To evaluate the mitochondrial membrane potential (MMP), we employed JC-1 and TMRE staining and according to the manufacturer's instructions (Cat# M34152 and T669, Thermo Fisher, USA). Cardiomyocytes from each group were incubated with JC-1 (1 μM) for 10 min at 37 °C. After washing, Images of JC-1 aggregates (exhibiting red fluorescence) and monomers (exhibiting green fluorescence) were captured using a Nikon Eclipse Ti2 fluorescence microscope. The ratio (green to red) was calculated for each group to determine MMP. Cardiomyocytes were stained with 50 nM TMRE for 30 min at 37 °C in the dark, washed, and resuspended in media. MMP was assessed by measuring TMRE fluorescence via flow cytometry. Three biological replicates were performed per assay.

2.14. TUNEL staining

Terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick-end labeling (TUNEL) staining was conducted using the In Situ Cell Death Detection Kit (Cat# 11767291910, Roche Applied Science, USA) in strict accordance with the manufacturer's protocol. Fluorescence images of positive staining were acquired using a Nikon Eclipse Ti2 fluorescence microscope.

2.15. RNA preparation and analysis

RNA isolation and real-time PCR were performed as previously described [7]. Adult and neonatal mice were anesthetized with 5 % isoflurane and euthanized by exsanguination or decapitation following NIH guidelines. Tissues, including brain, heart, muscle, liver, kidney, spleen, intestine, and lung, were collected for qPCR analysis. Total RNA was extracted from various organs using TRIzol reagent (Invitrogen, USA) and reverse-transcribed into cDNA using the SuperScript First-Strand Synthesis System (Invitrogen, USA). Real-time PCR was conducted in triplicate on an Applied Biosystems 7300 Plus system (Thermo Fisher Scientific, USA) with SYBR Green Master Mix (Thermo Fisher Scientific, USA). The following primers were used for amplification: GAPDH (forward: 5′-TCTCTGCTCCTCCCTGTTCT-3′, reverse: 5′-ATCCGTTCACACCGACCTTC-3′) and ACMSD (forward: 5′-GCCTCCCACAGTTGGATAGA-3′, reverse: 5′-GCCTCAAACACAGACCCATT-3′). ACMSD expression levels were normalized to GAPDH and analyzed using the ΔΔCt method.

2.16. Western blotting analysis

Protein extraction from cells and heart tissues was performed as previously described [7]. Protein samples (20 μg) were separated on 10 % SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, USA). The membranes were blocked with 5 % skimmed milk for 1 h at room temperature and subsequently incubated overnight at 4 °C with the following primary antibodies: ACMSD (1:1000), QPRT (1:1000), IDO1 (1:1000), HAAO (1:1000), KYNU (1:1000), SIRT1 (1:1000), ACO2 (1:1000), IFN-γ (1:1000), AMPK (1:1000), p-AMPK (1:1000), Nrf2 (1:1000), SOD2 (1:2000), acSOD2 (1:1000), STING (1; 1000), p-STING (1; 1000) and β-actin (1:5000). After incubation with HRP-conjugated secondary antibody (1:5000, Proteintech, USA) for 1 h at room temperature, protein levels were detected using enhanced chemiluminescence (ECL) (Cat# 34580, Thermo Fisher, USA) and visualized on a FluorChem HD2 Imaging System.

2.17. Statistical analysis

Statistical analysis was performed using GraphPad Prism software version 10.0 (San Diego, CA, USA). All data are presented as mean ± standard deviation (SD). Comparisons between two groups were analyzed using an unpaired two-tailed Student's T-test. For comparisons involving multiple groups, one-way or two-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test was applied. A p-value <0.05 was considered statistically significant.

3. Results

3.1. de novo NAD ⁺ pathway attenuates doxorubicin-induced cardiotoxicity in vivo

DOX has been reported to deplete NAD⁺ levels, contributing to multi-organ damage [22]. Given that NAD⁺ supplementation can mitigate DOX-induced injury, we investigated whether the de novo NAD ⁺ synthesis pathway protects against DIC by preserving NAD ⁺ levels in cardiomyocytes in vivo. To this end, wild-type (WT), IDO1 knockout (Ido1^−/−), and mice injected with either AAV9-control (AAV9-NC) or AAV9-shRNA-Ido1 were treated with DOX (2.5 mg/kg, i.v., weekly for 6 weeks, cumulative dose: 15 mg/kg) or saline [7]. Cardiac function and structural damage were assessed via echocardiography, Masson's trichrome staining, and Sirius Red staining. DOX treatment impaired cardiac function as evidenced by markedly reduced left ventricular ejection fraction (LVEF) and fractional shortening (LVFS) (Fig. 1A–C and Fig. S2A, D, E). Structural remodeling was characterized by substantial dilation of the left ventricular chambers, with significant increases in both left ventricular chambers of LV internal diastolic diameter (LVIDd) and LV internal systolic diameter (LVIDs) (Fig. S1A and B), accompanied by significant reductions in end-systolic LV posterior wall thickness (LVPWs) and end-systolic interventricular septal thickness (IVSs) (Fig. S1C and F). In contrast, the end-diastolic interventricular septal thickness (IVSd) and end-diastolic LV posterior wall thickness (LVPWd) remained unchanged (Fig. S1D and E). Notably, IDO1 deficiency exacerbated DOX-induced myocardial fibrosis, as demonstrated by enhanced collagen deposition in Masson's trichrome and Sirius Red-stained sections (Fig. 1D–G and Fig. S2B and F). Consistent with this, more pronounced histopathological damage, including disorganized myocardial fibers and interstitial edema, was observed in the absence of IDO1 by H&E staining (Fig. 1H and Fig. S2 C).

Fig. 1.

IDO1 deficiency aggravates DOX-induced myocardial dysfunction in mice. A, Representative echocardiogram images of each group in mice. Measurement of LVEF (B), LVFS (C) based on echocardiogram images (n = 9 mice per group). D, Representative images of Masson's staining, magnified views show muscle fibers (red) within the collagen-rich scar (blue). (Scale bar = 1 mm and 20 μm, respectively). E, Quantification of fibrotic area based on Masson staining (n = 3). F, Representative images of Sirus Red staining and G, quantification of (F) based on Sirus Red staining (n = 3). H, HE staining (Scale bar = 50 μm). I, DHE staining and J, quantification of DHE staining. K, NAD⁺ in each group of mice (n = 9). The values are presented as means ± SD. Statistical analysis was performed by two-way ANOVA with Tukey's multiple comparisons test (B, C, E, G, J, K). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Adjusted P values were provided in the case of multiple groups. LVEF, LV ejection fraction; LVFS, LV fractional shortening.

Fig. 2.

Effect of KP pathway on DOX-induced cardiomyocytes damage and NAD + levels in vitro. Primary cardiomyocytes are treated with DOX along or DOX with Trp, NA, NAM for 24 h: A, NAD⁺ levels decrease in the presence of different concentration DOX (n = 5). B, NAD⁺ levels decrease after DOX treatment, whereas increase after co-treatment with Trp, NR and NMN (n = 5). C and D, IDO1 protein levels of WT and Ido1^−/− cardiomyocytes (n = 3). E and F, TUNEL staining of DOX, DOX + Trp treated WT and Ido1^−/− cardiomyocytes (n = 3). G, Cell survival rate of WT and Ido1−/− cardiomyocytes exposed to 1 μM DOX at different times using CCK-8 assay. H, Representative IF staining images of SIRT1. I, Representative images of JC-1 staining of each group (Scale bar = 50 μm). J, Representative flow cytometry histograms showing TMRE fluorescence intensity in cardiomyocytes treated under different conditions. K, Quantitative analysis of TMRE fluorescence intensity across experimental groups. H, ROS levels in different groups. Data are expressed as the means ± SD. Statistical analysis was performed by one-way ANOVA with Tukey's multiple comparisons test (A, B, D) and two-way ANOVA with Tukey's multiple comparisons test (F, G, K, L). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗P < 0.0001. Adjusted P values were provided in case of multiple groups. IDO1, indoleamine 2,3-dioxygenase 1; KP, kynurenine pathway; DOX, doxorubicin; NAD⁺, nicotinamide adenine dinucleotide; Trp, tryptophan; NA, nicotinic acid; NAM, nicotinamide; ROS, reactive oxygen species. SIRT1, Sirtuin 1; WT, wild type.

Intracellular ROS and NAD⁺ levels in heart tissues were measured using dihydroethidium (DHE) and NAD⁺ assay kits. DOX treatment increased tissue ROS and decreased NAD⁺ levels in both WT and Ido1^−/− mice, but the deterioration was significantly more pronounced in Ido1^−/− mice. Furthermore, Ido1^−/− mice exhibited higher ROS levels and more severe NAD⁺ depletion compared to WT mice (Fig. 1I–K). These results revealed that the KP-mediated de novo NAD ⁺ synthesis plays a crucial role in maintaining cardiac NAD ⁺ levels and protecting against DOX-induced cardiomyocyte injury.

3.2. KP protects against DOX-induced cardiomyocyte damage by maintaining NAD⁺ levels to attenuate ROS generation in vitro

To further elucidate the protective details of KP against DIC, we established an in vitro model using primary cardiomyocytes treated with DOX alone or in combination with the de novo pathway precursor Trp or the salvage pathway precursors NR and NMN. Cellular NAD + levels were measured using a commercial NAD + assay kit. Cell viability was assessed via CCK-8 assay, and apoptosis was evaluated by TUNEL staining. Mitochondrial function was analyzed by measuring mitochondrial membrane potential (MMP) using JC-1 and TRME staining. Primary cardiomyocytes were isolated from WT and IDO1 knockout (Ido1^−/−) neonatal mice. DOX treatment triggered a dose-dependent decline in NAD⁺ levels, resulting in a 40–50 % reduction at 1 μM after 24 h (Fig. 2A). This depletion was significantly ameliorated by supple-mentation with NAD⁺ precursors Trp, NR, or NMN, which increased NAD⁺ levels by approximately one-two folds in wild-type cardiomyocytes (Fig. 2B). IDO1 knockout was confirmed by Western blot (Fig. 2C and D). Notably, IDO1 deficiency exacerbated DOX-induced cellular injury, manifesting as reduced cell viability (Fig. 2G), increased apoptosis (Fig. 2E and F), elevated ROS levels (Fig. 2L), and loss of mitochondrial membrane potential (Fig. 2I–K). Trp administration markedly attenuated these deleterious effects in wild-type cardiomyocytes, but its protective effects were significantly diminished in IDO1-deficient cells. Furthermore, Trp treatment restored SIRT1 expression, which had been suppressed by DOX exposure (Fig. 2H). These findings suggest that the KP pathway of de novo NAD ⁺ synthesis plays a critical role in maintaining NAD ⁺ levels and protecting against DOX-induced cardiomyocyte injury.

3.3. KP is a hub of regulation in the conversion between NAD⁺ and TCA circle

The kynurenine pathway functions as a pivotal metabolic hub, directing tryptophan catabolites either toward de novo NAD ⁺ synthesis through QPRT or toward energy generation via ACMSD-mediated entry into the TCA cycle, with implications for both physiological and pathological states (Fig. 3A). DOX treatment induced a dose-dependent upregulation of IDO1 (Fig. 3B, C, I and Fig. S3A-D) and increased activity of KYNU and 3-HAAO (Fig. 3B–G, H and Fig. S3E). Although these changes might be expected to enhance NAD⁺ production, we observed a paradoxical reduction in both NAD⁺ and SIRT1 levels (Fig. 3B–D, J). Mechanistically, this discrepancy was attributed to concomitant downregulation of QPRT and upregulation of ACMSD (Fig. 3B–E, F, I), which shifted the metabolic flux away from NAD⁺ synthesis. IDO1 deficiency exacerbated this dysregulation, resulting in more severe QPRT suppression and NAD⁺ depletion (Fig. 3E–J). Although ACMSD elevation might be expected to augment ATP production via enhanced TCA cycle flux, both WT and IDO1-KO cardiomyocytes exhibited reduced ATP levels (10 % and 22 % decrease versus controls, respectively; Fig. 3K). These results demonstrate that DOX not only significantly decreased ATP levels but also rewired KP metabolism through coordinated ACMSD activation and QPRT suppression, resulting in the concurrent disruption of NAD⁺ and ATP homeostasis, an effect predominantly mediated by ACMSD activation.

3.4. Switch of ACMSD to QPRT promotes de novo NAD ⁺ biosynthesis and alleviates the mitochondrial damage caused by DOX in vitro

Given the reciprocal regulation of ACMSD (upregulated) and QPRT (downregulated) during DOX treatment, we hypothesized that shifting this metabolic balance toward QPRT activation could stimulate de novo NAD ⁺ synthesis in cardiomyocytes. While TES-1025 has been established as a selective ACMSD inhibitor in renal tissue [14], its cardiac effects remained unexplored. To explore this potential, we verified ACMSD expression levels among different tissues in adult and neonatal mice (Fig. 4A and Fig. S4A) by using real-time PCR. The ACMSD mRNA level in the adult heart is expressed 5 times higher than brain, while the kidney expresses the highest in 1000 times of brain (Fig. 4A), which was similar as reported [23]. TES-1025 effectively inhibited ACMSD activity in cardiomyocytes, and a 10 μM concentration of TES-1025 was selected for further experiments (Fig. 4B and C). To test our hypothesis, primary cardiomyocytes were divided into four groups: control, TES-1025 (TES), DOX, and DOX + TES. We first examined how metabolic reprogramming through ACMSD/QPRT switch affects ATP production. Both DOX and TES-1025 alone decreased cellular ATP levels compared to control. Notably, co-treatment with TES-1025 and DOX further reduced ATP by 16 % compared to DOX alone (Fig. 4D). To investigate the metabolic basis for the ATP reduction, we traced the fate of ¹³C-labeled citrate derived from ¹³C-labeled tryptophan (Fig. 4E). We found that DOX treatment increased the flux of 3-HAA into the TCA cycle, whereas TES-1025 co-treatment limited this flux. This restriction of substrate entry was accompanied by a normalization of TCA cycle activity, as evidenced by reduced expression of aconitase 2 (ACO2) (Fig. 4F, G and Fig. S6A and B). Since ACO2 catalyzes the conversion of citrate to isocitrate, a critical step for ATP generation [24], its downregulation provides a mechanistic explanation for the observed reduction in ATP. We also found that co-treatment of TES-1025+DOX decreased DOX-induced ACMSD, while increasing the expression of QPRT and SIRT1 compare DOX alone (Fig. 4F, G). Furthermore, TES-1025 treatment enhanced Nrf2 activation and reduced SOD2 acetylation (Fig. S5A and B). To determine whether SIRT1 mediated these observed effects, we performed siRNA-mediated knockdown of SIRT1 in cardiomyocytes. Notably, these beneficial effects were abolished upon SIRT1 silencing (Fig. S5A and B).

Fig. 4.

Blocking ACMSD increases NAD⁺ and protects mitochondria from DOX toxicity in vitro. A, mRNA levels of ACMD in adult brain, heart, muscle, liver, kidney, spleen, intestine and lung tissues by using real-time PCR (n = 3). B and C, Western blotting and quantification of ACMSD in cardiomyocytes treated with different concentrations of TES1025 (n = 3). D, ATP levels of each group (n = 5). E, Quantification of citrate levels in cardiomyocytes following treatment with DOX alone or in combination with TES-1025 at indicated time points. F, Western blot of ACMSD, QPRT, ACO2, and SIRT1 in cardiomyocytes with or without ACMSD inhibitor treatment after DOX exposure. β-actin as load control (n = 3). G, Quantification of protein levels of (F), the expression level of each protein was standardized by β-actin and the value in WT-CTL group was designated as 1. H, The levels of NAD⁺ in different groups (n = 5). I, Representative images of mitochondria marked by MitoTracker Green and mitochondrial superoxide radicals visualized by MitoSOX Red staining in different groups (Scale bar = 50 μm). J, Quantification of MitoSOX intensity based on MitoSOX Red staining (n = 3), the value in WT-CTL group was designated as 1. K and M, Functional profiling of mitochondrial respiration (OCR) and glycolysis (ECAR) in response to treatment with TES-1025, DOX, or their combination. I and N, Quantitative analysis of (K) and (M). O, Representative images of JC-1 staining of each group (Scale bar = 50 μm). P, Representative flow cytometry histograms showing TMRE fluorescence intensity in cardiomyocytes treated under different conditions. Q, Quantification of (P). The values are presented as mean ± SD. Statistical analysis was performed by one-way ANOVA with Tukey's multiple comparisons test (A, C, D, J, H, Q) and two-way ANOVA with Tukey's multiple comparisons test (L, N). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Adjusted P values were provided in the case of multiple groups. ATP, adenosine triphosphate production; NAD⁺, nicotinamide adenine dinucleotide; DOX, doxorubicin; ACMSD, α-amino-β-carboxy-muconate-semialdehyde decarboxylase; QPRT, quinolinate phosphoribosyl-transferase; ACO2, aconitase 2; SIRT1, sirtuin 1.

We next examined how ACMSD/QPRT switch affects NAD⁺ levels and mitochondrial ROS. TES-1025 treatment effectively restored de novo NAD ⁺ biosynthesis in DOX-treated cardiomyocytes, resulting in a 1.5-fold increase in NAD⁺ compared to the DOX group (Fig. 4H). This was accompanied by a 24 % reduction in mitochondrial ROS (Fig. 4I, J and Fig. S6C and D) and stabilization of mitochondrial membrane potential (Fig. 4O–Q). To further investigate the protective effects of TES-1025 against DOX-induced mitochondrial dysfunction, a comprehensive analysis of mitochondrial respiratory and glycolytic rate parameters was performed using the Seahorse XF Analyzer (Fig. 4K–N). Our results demonstrate that treatment with 1 μM DOX markedly impaired mitochondrial respiration, significantly reducing the basal respiration. Co-treatment with TES-1025 partially restored basal respiration (Fig. 4K and L). DOX severely com-promised the maximal respiratory capacity, which was significantly improved by TES-1025 (Fig. 4K and L). Furthermore, DOX-treated cardiomyocytes exhibited a significant increase in extracellular acidification rate (ECAR) (Fig. 4M and N), enhancing glycolytic capacity and glycolytic reserve, all of which were reversed by TES-1025 co-treatment. These data demonstrated that TES-1025 reduces the DOX-induced glycolytic surge, effectively ameliorates DOX-induced mitochondrial respiratory chain impairment, and enhances overall mitochondrial function. Finally, since TES-1025 elevates NAD⁺ in cardiomyocytes, we investigated whether it would also protect cancer cells from DOX-induced death. We treated MCF-7 cells with DOX alone or in combination with NR, NMN, or TES-1025. Strikingly, while both NR and NMN significantly attenuated DOX-induced cell death, and TES-1025 did not impair DOX's cytotoxic efficacy (Fig. S7A–D).

Collectively, our findings strongly suggest that TES-1025-mediated ACMSD inhibition enhances QPRT activity to elevate NAD⁺ levels, improving cellular antioxidant capacity and mitigating DOX-induced mitochondrial damage in cardiomyocytes. The improved mitochondrial respiration despite reduced ATP production suggests a metabolic reprogramming that prioritizes redox homeostasis over energy generation, ultimately promoting cardiomyocyte survival under DOX-induced stress.

3.5. TES-1025 protects myocardial injury from DOX in mice

To evaluate the therapeutic potential of ACMSD inhibition in vivo, we administered TES-1025 alongside DOX treatment in mice. This intervention revealed a dual capacity to redirect KP metabolites toward NAD⁺ synthesis while preserving cardiac structure and function. TES-1025 co-treatment significantly attenuated DOX-induced cardiac dysfunction, as demonstrated by improved LVEF (13 % increase vs DOX alone) and LVFS (16 % increase), effectively reversing the initial declines of 20 % and 25 %, respectively (Fig. 5A–C). This cardioprotective effect was further evidenced by a substantial reduced myocardial fibrosis, as shown by Masson's trichrome and Sirius red staining (Fig. 5D–H and Fig. S8), and a marked decrease in apoptosis (Fig. S9). Ultrastructural analysis demonstrated that TES-1025 conferred notable mitochondrial protection, preventing DOX-induced damage, including cristae disruption, cytoplasmic vacuolization (V), and myofibril disintegration. These pathological alterations were more exacerbated in IDO1KO mice (Fig. S8H). These findings collectively establish that pharmacological ACMSD inhibition via TES-1025 rebalances the ACMSD-QPRT axis to favor NAD⁺ biosynthesis over TCA cycle flux, thereby preserving mitochondrial integrity and maintaining cardiac function under DOX treatment.

Fig. 5.

TES-1025 protects the heart from DOX damage in vivo. A, Representative echocardiogram images of control group, DOX (15 mg/kg) and TES group (15 mg/kg) (n = 9 mice per group). Measurement of LVEF (B) and LVFS (C) based on echocardiogram images (n = 9 mice per group). D, Representative image of Masson's staining in each group (Scale bar = 20 μm, n = 9). E, Quantification of fibrotic area based on Masson's staining (n = 3). F, Representative images of Sirus Red staining and G, quantification of (F) based on Sirus Red staining (Scale bar = 50 μm, n = 3). H, Representative images of HE staining in each group (Scale bar = 50 μm). I, Representative transmission electron microscopy (TEM) images of cardiac tissues in each group. Red arrows indicate vacuolization. Yellow (mt) indicates mitochondria and blue (Mf) indicates myofibrils (Scale bar = 2 μm and 1 μm, respectively, n = 5). Images are from the following group: CTL, control group, DOX, doxorubicin treated group, DOX + TES, DOX + TES-1025 treated group. The values are presented as means ± SD. Statistical analysis was performed by one-way ANOVA with Tukey's multiple comparisons test (B, C, E, G). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Adjusted P values were provided in the case of multiple groups. mt, mitochondria; Mf, myofibrils; V, vacuolization.

3.6. TES-1025 protects DOX-induced heart injury by changing the metabolites toward de novo NAD ⁺ synthesis in vivo

TES-1025 regulates the switch of ACMSD to QPRT and protects against DOX-induced heart injury. To determine whether the cardioprotective effects stem from a metabolic shift linking NAD⁺ homeostasis and the TCA cycle activity in vivo, we performed LS-MS/MS metabolomic profiling alongside measurements of ROS, NAD⁺, and ATP levels in heart tissues. Western blot and immunohistochemistry (IHC) were used to assess protein levels of ACMSD, QPRT, ACO2, and SIRT1. Consistent with our in vitro findings, TES-1025 co-treatment with DOX has cardioprotective effects by inhibiting ACMSD while upregulating QPRT and SIRT1 (Fig. 6A, B, Fig. S10A, B and Fig. S11). This was accompanied by a 35 % increase in cardiac NAD⁺ (Fig. 6E), a 50 % reduction in ROS (Fig. 6C, D), and a further exacerbation of ATP depletion (Fig. 6F). Meanwhile, LC-MS/MS metabolomic profiling revealed that TES-1025 co-treatment effectively counteracted DOX-induced metabolic perturbations. including: (1) restored QA levels (1.5-fold elevation vs DOX alone; Fig. 6I), which had been reduced by 45 % in the DOX group. (2) increased flux toward NAD⁺ synthesis, evidenced by elevated QA/Kyn (2.6-fold increase was observed, showing a clear visual trend that did not reach statistical significance) and QA/3-HAA (3.0-fold) ratios (Fig. 6M, N), and a 64 % reduction in the Kyn/Trp ratio (Fig. 6L). (3) significant reduction in DOX-induced Kyn accumulation alongside increased kynurenic acid (KA) levels (Fig. 6H–K), suggesting that TES-1025 diverts metabolic flux toward the KA branch. Notably, TES-1025 retained the same efficacy in IDO1KO mice, sustaining ACMSD inhibition and QPRT/SIRT1 upregulation (Fig. S10A, B), and increasing QA levels (Fig. S10E). In addition, a modest elevation in the QA/Kyn and QA/3-HAA ratios was observed, though these changes did not reach statistical significance (Fig. S10I and J). This effect occurred despite atypical metabolic responses, including a stable Trp level (Fig. S10C) and mild Kyn accumulation (Fig. S10D). This suggests the occurrence of limited cardiomyocyte Kyn uptake.

Fig. 6.

ACMSD inhibition increases NAD + levels in mice. A, Western blotting result of ACMSD, QPRT, SIRT1 and ACO2 in different groups of cardiac tissues from mice. B, Quantification of A (n = 3). The level of each protein was standardized by β-actin and the value in WT-CTL group was designated as 1. C, Representative images of DHE staining and D, Quantification of DHE staining (n = 3). E, levels of NAD⁺ in mouse heart tissue of each group (n = 9). F, levels of ATP in mouse heart tissue of each group (n = 9). G-K, Representative concentrations of different metabolite measured by LS/MS/MS in mouse cardiac tissues treated with vehicle, DOX and DOX + TES-1025: Trp (G), Kyn (H), QA (I), 3-HAA (J) and KA (K) (n = 9). L, levels of Kyn/Trp ratio in each group. M, levels of QA/Kyn ratio in each group. N, levels of QA/3-HAA ratio in each group. O, levels of KA/Kyn ratio in each group. The values are presented as means ± SD. Statistical analysis was performed by one-way ANOVA with Tukey's multiple comparisons test (B, D-O). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Adjusted P values were provided in the case of multiple groups. ACMSD, α-amino-β-carboxy-muconate-semialdehyde decarboxylase; QPRT, quinolinate phosphoribosyl-transferase; Trp, tryptophan; Kyn, kynurenine; QA, quinolinic acid; KA, kynurenic acid; 3-HAA, 3-hydroxyanthranilic acid.

Together, TES-1025 counteracted DOX-induced TCA cycle hyperactivation by rebalancing the ACMSD-QPRT axis, redirecting metabolism flux toward NAD⁺ production, and enhancing NAD⁺ biosynthesis. These results reveal that TES-1025 provides cardioprotection through both KP-dependent and -independent mechanisms.

3.7. DOX induces ACMSD activation via IFN-gamma/AMPK axis

IDO1, the first and rate-limiting enzyme of KP, is activated by various factors, including interferon-gamma (IFN-γ) [25]. IFN-γ administration is known to reduce Trp while increasing Kyn levels in vivo [26]. The STING pathway activates IFN-γ and is itself activated by DOX. Separately, 5′-AMP-activated protein kinase (AMPK), a central metabolic stress sensor, has been shown to negatively regulate ACMSD while positively regulating QPRT expression in hepatic cells [27]. Interestingly, DOX treat-ment stimulates IFN-γ production, which contributes to its anti-tumor efficacy [28]. However, IFN-γ is also known to suppress AMPK activity in cardiomyocytes and other cell types [29]. Therefore, we hypothesize that in cardiomyocytes, DOX activates the STING pathway, which in turn induces IFN-γ expression. This DOX-induced IFN-γ activation modulates the ACMSD/QPRT switch via the IFN-γ/AMPK axis. To test this, we first assessed the STING/IFN-γ/AMPK axis in primary cardiomyocytes treated with DOX. DOX increased the p-STING/STING ratio (2.2-fold), as well as the expression of IFN-γ (1.4-fold) and ACMSD (1.6-fold), while decreasing p-AMPK (35 % reduction) and QPRT (52 % reduction) compared to controls (Fig. 7A and B). In contrast, the STING inhibitor H-151 significantly attenuated the DOX-induced upregulation of IFN-γ (Fig. 7C and D). Next, we evaluated the specific effects of IFN-γ and AMPK agonist AICAR on cardiomyocytes under both control (NC) and DOX-treated (DOX) conditions (Fig. 7E). IFN-γ significantly increased ACMSD expression in both groups (2.1-fold in NC &1.4-fold in DOX, respectively, p < 0.05) and significantly decreased phosphorylated AMPK(p-AMPK) in DOX-treated cells (43 % reduction), though the reduction in NC did not reach statistical significance (Fig. 7E–G). IFN-γ also reduced QPRT levels in both NC and DOX (25 % reduction), but these reductions were not statistically significant (Fig. 7E–I). In contrast, AICAR treatment markedly reduced ACMSD expression by 46 % and increased p-AMPK by 1.9-fold in both NC and DOX groups (Fig. 7E–G). Furthermore, co-treatment with IFN-γ and AICAR abolished the IFN-γ-induced upregulation of ACMSD and suppression of QPRT (Fig. 7E–H, I).

Fig. 7.

DOX increases ACMSD and decreases QPRT through the STING/IFN-γ/AMPK pathway. A, Western blotting result of STING, p-STING, IFN-γ, AMPK, p-AMPK, ACMSD and QPRT in the cardiomyocytes treated with DOX. B, Quantification of (A) (n = 3). The level of each protein was standardized by β-actin and the value in WT-CTL group was designated as 1. C and D, Western blotting and quantification of STING, p-STING and IFN-γ protein levels after treatment with DOX alone or in combination with H-151. E, Western blotting result of IFN-γ, AMPK, p-AMPK, ACMSD and QPRT in the NC and DOX treated cardiomyocytes with or without IFN-γ or AICAR administration. F–I, Quantification of (E) (n = 3). The values are presented as means ± SD. Statistical analysis was performed by one-way ANOVA with Tukey's multiple comparisons test (B, D) and two-way ANOVA with Tukey's multiple comparisons test (F, G, H, I). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Adjusted P values were provided in case of multiple groups. IFN-γ, interferon-gamma; AMPK, AMP-activated protein kinase; ACMSD, α-amino-β-carboxy-muconate-semialdehyde decarboxylase; QPRT, quinolinate phosphoribosyl-transferase.

Given that TES-1025 alters ATP production and glycolytic capacity in DOX-treated cardiomyocytes, we hypothesized it might activate AMPK. Confirming this, we found that TES-1025 significantly increased p-AMPK levels and concurrently decreased the ATP/ADP ratio (Fig. S12). In the hearts of DOX-treated mice, TES-1025 co-treatment did not affect the upstream STING/IFN-γ pathway but significantly increased p-AMPK levels (Fig. S13). The activation of AMPK was subsequently associated with decreased ACMSD and increased QPRT expression (Fig. 6A and B). These results confirm our hypothesis that DOX regulates the ACMSD/QPRT metabolic switch via the STING/IFN-γ/AMPK signaling axis in cardiomyocytes.

4. Discussion

This work elucidates the dual role of the IDO1-derived kynurenine pathway in doxorubicin-induced cardiotoxicity, functioning as both a metabolic regulator and a potential therapeutic target. Genetic ablation and cardiac-specific knockdown of IDO1 exacerbated DOX-induced cardiomyopathy through a cascade of metabolic disturbances: (1) impaired NAD⁺ regeneration (Figs. 1K and 3J), (2) collapse of redox homeostasis (ROS increased by 20 %, Figs. 2L and 4J and Fig. S6 D), (3) mitochondrial ultrastructural damage (evidenced by TEM showing cristae disruption, Fig. 5 and Fig. S8 H), and (4) activation of apoptotic pathways (Fig. 2E and Fig. S9 A, B). We demonstrate that KP activation maintains cardiac homeostasis by dynamically partitioning metabolites between energy production (TCA cycle) and NAD⁺ biosynthesis. DOX pathologically reprograms this pathway by coordinately activating ACMSD while suppressing QPRT, thereby diverting metabolites toward the TCA cycle instead of NAD⁺ production (Fig. 4D, E, H). Mechanistically, this reprogramming is driven by DOX-induced STING pathway activation and subsequent IFN-γ secretion, which suppresses AMPK phosphorylation (35 % reduction in p-AMPK/AMPK ratio; Fig. 7A and B). This signaling cascade drives ACMSD upregulation (1.4-fold versus control, Fig. 7A and B) and downregulation of QPRT (Fig. 7A and B). This shunt redirects QA away from NAD⁺ synthesis and creates a bioenergetic crisis. Therapeutic intervention with the ACMSD inhibitor TES-1025 (15 mg/kg) restored myocardial NAD⁺ levels (1.9-fold versus DOX alone, Fig. 6E) and improved cardiac ejection fraction (Fig. 5B and C), underscoring the clinical potential of targeting this metabolic pathway to mitigate doxorubicin-induced cardiotoxicity.

NAD⁺ serves as an essential cofactor for mitochondrial respiration and a key substrate for metabolic regulators that maintain redox homeostasis and energy metabolism [5]. Its established role in activating SIRT1 and promoting longevity [30] has positioned NAD⁺ as a promising therapeutic target for cardiovascular diseases [31,32]. While NAD⁺ levels can be increased through precursor supplementation or biosynthetic pathway activation. NR and NMN supplementation attenuate the development of heart failure in mice and patients [33,34], yet salvage pathway is constrained by its rate-limiting enzyme NAMPT. This enzyme is frequently overexpressed in malignancies but deficient in cardiac and metabolic disorders, highlighting the therapeutic potential of alternative NAD ⁺ biosynthesis routes. The de novo NAD ⁺ biosynthesis pathway represents such an alternative. Trp, the KP pathway's initial substrate, demonstrates superior NAD⁺ -boosting capacity compared to nicotinic acid or nicotinamide in mammalian models [9,13]. Notably, the dynamic regulation between NAD⁺ pathways appear context-dependent, as evidenced by studies in THP-1 human monocytes and peripheral blood mononuclear cells (PBMC), where the system shifts from NAMPT-dependent salvage during mild inflammation to IDO1-mediated de novo synthesis in severe sepsis [35]. This transition expands nuclear NAD⁺ pools and sustains SIRT1 activation, while IDO1 Inhibition predominantly decreased nuclear NAD⁺ and SIRT1 levels [35]. Accumulating evidence reveals the cardioprotective potential of the KP across various cardiovascular pathologies. Beneficial effects have been documented in heart failure [15], atherosclerosis and vascular inflammation [16,36], hypertension and pulmonary hypertension (PH) [17,37], and mitochondrial quality control in myocardial infarction (MI) [38]. The de novo pathway is crucial for cardiac endothelial cells [39], CD8⁺ T cells [40], Macrophage [41] and preventing congenital NAD⁺ deficiency disorders (CNDD) [42], as demonstrated by humans and mice with HAAO/KYNU mutations that exhibit NAD⁺ depletion and developmental defects [42,43], and by QPRT inhibition causing pellagra-like NAD⁺ deficiency [44]. Furthermore, QPRT derived QA evaluation rescues the compromised salvage pathway for NAD⁺ production in myxomatous mitral valve disease (MMVD) [45]. These established findings provide crucial context for our observation that IDO1 deletion induces cardiac NAD⁺ depletion, collectively confirming the KP pathway's cardioprotective role.

The heart is highly dependent on mitochondrial oxidative phosphorylation, which generates approximately 95 % of its ATP, with fatty acids contributing 40–60 % of this energy under physiological conditions. During doxorubicin-induced stress, impaired fatty acid oxidation shifts metabolism toward glycolysis and alternative substrates, including tryptophan-derived metabolites. This adaptation increases pyruvate and AMS flux into the TCA cycle, elevating TCA enzyme expression [46]. Although this metabolic rewiring initially preserves ATP production by utilizing NADH for oxidative phosphorylation, it concurrently activates ACMSD, which diverts QA away from NAD⁺ synthesis and into the TCA cycle. However, DOX induces mitochondrial damage and severely depletes overall cellular ATP. Thus, the increase in ATP from ACMSD activation appears insufficient to fully counteract the deficit under DOX stress. Notably, IDO1^−/− mice exhibit more severe reductions in both ATP and NAD⁺ levels than WT mice during DOX treatment (Figs. 1K and 3J, K), yet still maintain the ACMSD/QPRT metabolic switch, suggesting limited cardiomyocyte uptake of kynurenine pathway intermediates (Kyn, 3-HAA) may constrain NAD⁺⁺ biosynthesisin the absence of IDO1.

ACMSD orchestrates a metabolic switch between energy production and NAD + biosynthesis, diverting the kynurenine pathway intermediate ACMS toward TCA cycle entry at the expense of QA-dependent NAD⁺ synthesis. This energy pathway is conserved from bacteria to mammals [47,48]. While ACMSD is known to be highly expressed in kidney, liver, and brain [23], our study reveals its functional significance in the heart. We found that ACMSD expression levels in the adult mouse heart is approximately 5 times higher than in the brain (Fig. 4A). Previous research has demonstrated that ACMSD inhibition elevates NAD⁺ production and confers protection against AKI and liver injury [14,49]. ACMSD mutation causes the elevation of QA in Parkinson's disease and Major Depressive Disorder patient [50,51]. In a transgenic mouse model called ANDY (acquired niacin dependency) that the human ACMSD (hACMSD) gene was overexpressed in mouse, causing the remarkable decrease in NAD⁺ level [52]. Recently, a report also showed that TES-1025 promoted NAD ⁺ de novo synthesis and SIRT1 activity, enhanced AMPK expression and significantly reduced hair cell death [53]. Here, our study provides the first direct evidence that metabolic reprogramming of the ACMSD-to-QPRT switch plays a cardioprotective role in doxorubicin-induced cardiotoxicity by restoring NAD⁺ biosynthesis. However, given that DOX is primarily used as anticancer agent, the potential impact of ACMSD inhibition on oncologic outcomes warrants discussion. While ACMSD inhibition could theoretically increase NAD⁺ content in cancer cells and possibly influence therapeutic efficacy, it is important to note that many cancer types, particularly neuroblastoma and other malignancies, exhibit low basal expression of ACMSD [54]. This inherently low expression level likely restricts the potential for NAD⁺ modulation through ACMSD inhibition within tumor tissues. In MCF-7 cells, co-treatment with NR or NMN significantly attenuated DOX-induced cell death, whereas TES-1025 did not (Fig. S7A–D). This suggests that ACMSD inhibition is unlikely to compromise the anti-cancer efficacy of doxorubicin, a distinct advantage over broad NAD⁺ precursor NR and NMN supplementation. Nevertheless, the therapeutic implications of ACMSD modulation in oncology warrant further investigation.

Reactive oxygen species (ROS) function as crucial signaling molecules at physiological levels but become cytotoxic when overproduced, triggering mitochondrial dysfunction and cell death [55]. Chronic mitochondrial dysfunction ultimately leads to energy depletion and heart failure. In heart failure management, maintaining the delicate balance between redox homeostasis and energy metabolism is important, however current evidence indicates that mitigating oxidative stress represents the paramount priority for effective cardioprotection [56]. Caloric restriction (CR) has demonstrated significant potential for modulating NAD⁺ biosynthesis pathways. Studies in CR-AKI mice show that CR preferentially regulates genes involved in de novo NAD ⁺ synthesis (KYNU, HAAO, ACMSD, QPRT) while leaving the salvage pathway largely unaffected [57]. This targeted metabolic reprogramming suggests that interventions in the kynurenine pathway may offer novel therapeutic strategies for maintaining cardiac redox balance while avoiding the oncogenic risks associated with salvage pathway activation.

5. Limitations, translational implications and future directions

Our study demonstrates that metabolic reprogramming through the ACMSD-to-QPRT switch enhances myocardial antioxidant capacity and confers resistance to oxidative stress in DIC mouse models. However, several limitations should be noted: (1) we use TES-1025 pharmacological treatment in vitro and in vivo and we found it on-target ACMSD inhibition, resulting in a comprehensive ACMSD/QPRT switch and metabolic profiling. However, while we did not use cardiac-specific knockout ACMSD mice, therefore, systemic effects beyond cardiomyocyte-specific mechanisms cannot be excluded. (2) Our study utilized 8-week-old male mice, a model that is frequently used in DOX-induced cardiotoxicity research. According to Jackson Laboratory guidelines, mice aged 3–6 months correspond to 20–30 years young adults in human. Future studies incorporating older mice and both sexes would improve pathological relevance. especially for modeling age-related or sex-specific susceptibility in cardiomyopathy. (3) We use primary neonatal cardiomyocytes for our in vitro experiments. The metabolic phenotype of these cells differs from that of adult cardiomyocytes in the human heart.

While neonatal cardiomyocytes rely predominantly on glycolysis for energy production, and adult cardiomyocytes generate most of their ATP through mitochondrial fatty acid oxidation. Since DOX directly impairs mitochondrial function and lipid metabolism, our in vitro observations may not fully recapitulate the pathogenic mechanisms occurring in the adult heart.

NAD⁺ is essential for all cells, but cancer cells are particularly dependent on the salvage pathway for its production, making the enzyme NAMPT a key therapeutic target [19,58,59]. A major clinical challenge for NAMPT inhibitors is their toxicity to normal cells from systemic NAD⁺ depletion, which is similar to that caused by chemotherapy drugs DOX-induced cardiotoxicity. While NAD⁺ precursors like NMN or NR can protect the heart [33,34], they can also rescue cancer cells by replenishing NAD⁺ via the salvage pathway, thereby reducing anticancer efficacy [60,61]. This contrast highlights the potential of targeting the de novo NAD ⁺ biosynthesis pathway. Cardiomyocytes can efficiently use this pathway, while many cancers have impaired de novo synthesis due to downregulation of key enzymes like ACMSD and QPRT [18,54,62]. Therefore, targeting the ACMSD-QPRT axis may offer a strategy to selectively protect the heart without promoting tumor survival. While TES-1025 demonstrates promising pharmacological effects, these observations remain a pharmacological proof-of-concept until genetic validation can be obtained. Although cancers primarily rely on salvage pathways, they can adapt. For instance, a recent study showed that triple-negative breast cancer can activate de novo synthesis by upregulating QPRT [63], indicating that the safety of modulating this pathway in cancer patients needs further investigation. The development of a cardiac-targeted delivery system represents a promising strategy to enhance specificity and minimize systemic effects. For instance, ACMSD inhibitors such as TES-1025 could be delivered via liposomal encapsulation or nanoparticle-based carriers to reduce off-tumor exposure.

6. Conclusion

This study establishes the kynurenine pathway as a critical regulator of NAD⁺ homeostasis in doxorubicin-induced cardiomyopathy, identifying the metabolic switch from ACMSD to QPRT as a pivotal control point for de novo NAD ⁺ biosynthesis. By demonstrating that pharmacological modulation of this switch (via TES-1025-mediated ACMSD inhibition) restores cardiac NAD⁺ levels and mitigates oxidative damage, our work provides both mechanistic insight and a promising therapeutic potential for chemotherapy-induced cardiotoxicity. These findings may extend to other forms of cardiomyopathy characterized by NAD⁺ depletion, suggesting new avenues for metabolic intervention in heart failure.

Clinical trial number

Not applicable.

Ethics approval and consent to participate

All animal experiments were conducted according to the guidelines of National Institutes of Health (NIH) for the care and use of laboratory animals and approved by the Ethics Committee for Animal Experiments of Taizhou Hospital (approved ID number: TZYY2022044 and TZYY2025106).

Consent for publication

Not applicable.

Funding

This work was supported in part by the National Natural Science Foundation of China (NSFC81974026)

CRediT authorship contribution statement

Danlei Li: Conceptualization, Data curation, Formal analysis, Project administration, Visualization, Writing – original draft. Yang Zhang: Data curation, Methodology, Project administration. Yuanyuan Kuang: Formal analysis, Methodology, Project administration. Zhong Lin: Methodology, Project administration. Ping Wang: Methodology. Jianjun Jiang: Funding acquisition, Resources. Wenhu Pi: Conceptualization, Funding acquisition, Project administration, Supervision, Validation, Writing – review & editing. Qilin Ma: Funding acquisition, Resources, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

None declared.

Nonstandard Abbreviations and Acronyms

ACMS

2-amino-3-carboxymuconic-6-semialdehyde

ACMSD

Aminocarboxymuconate semialdehyde decarboxylase

ACO2

Aconitase 2

AMPK

AMP-activated protein kinase

AMS

Aminomuconic semialdehyde

ATP

Adenosine Triphosphate

AKI

Acute kidney injury

ANDY

Acquired niacin dependency

cGAS-STING

cyclic GMP/AMP synthase–stimulator of interferon genes

CD38

Cluster of differentiation 38

DOX

Doxorubicin

DIC

Doxorubicin - induced cardiomyopathy

FA

Fatty acids

3′HAAO

3-Hydroxyanthranilate 3,4-Dioxygenase

HE

Hematoxylin-eosin

HFpEF

Heart failure with preserved ejection fraction

3HK

3-hydroxyKynurenine

3HAA

3-hydroxyanthrani-lic acid

IDO1

Indole-2,3-dioxygenase 1

IFN-γ

Interferon-gamma

KAT

Kynurenine amino transferase

KMO

Kynurenine 3-monooxygenase

KP

Kynurenine pathway

Kyn

Kynurenine

KYNA

Kynurenic Acid

KYNU

Kynureninase

LVEF

Left ventricular ejection fraction

LVFS

Left ventricular fractional shortening

LVIDd

Left ventricular internal diastolic diameter

LVIDs

Left ventricular internal systolic diameter

LVPWd

End-diastolic left ventricular posterior wall thickness

LVPWs

End-systolic left ventricular posterior wall thickness

IVSd

End-diastolic interventricular septal thickness

IVSs

End-systolic interventricular septal thickness

MMVD

Myxomatous mitral vaslve diseae

NA

Nicotinic acid

NAD⁺

Nicotinamide adenine dinucleotide

NAM

Nicotinamide

NAMPT

Nicotinamide phosphoribosyl transferase

NMN

Nicotinamide mononucleotide

NMNATs

NMN adenylyl transferases

NR

Nicotinamide riboside

OXPHOS

Oxidative phosphorylation

PARPs

Poly-ADP-ribose-polymerases

QA

Quinolinic acid

QPRT

Quinolinate Phosphoribosyltransferase

TDO

Tryptophan 2,3-dioxygenase

ROS

Reactive oxygen species

SARM1

Sterile alpha and TIR motif containing 1

SESN2

Sestrin2

SIRT1

Sirtuin 1

TCA

Tricarboxylic Acid

Trp

Tryptophan

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.