Endothelial C/EBPβ/AEP pathway drives vascular deterioration and systemic aging

Abstract

Age-related microvascular dysfunction disrupts nutrient homeostasis and waste clearance, leading to organ failure. However, a critical gap remains in our understanding of the specific molecular drivers of vascular deterioration and how they orchestrate organism-wide aging. Here, we identify progressive activation of the CCAAT/enhancer-binding protein β (C/EBPβ)/asparagine endopeptidase (AEP) pathway in aging vascular endothelial cells contributes to vascular degeneration and lifespan reduction. Endothelial-specific C/EBPβ or AEP overexpression accelerated vascular aging and shortened lifespan in mice. Mechanistically, AEP mediates proteolytic cleavage of nicotinamide phosphoribosyltransferase (NAMPT), the rate-limiting enzyme in nicotinamide adenine dinucleotide (oxidised form, NAD⁺) biosynthesis, resulting in systemic NAD⁺ depletion and senescence that elicits both central and peripheral vascular dysfunction and ultimately systemic aging. Genetic ablation of AEP or expression of AEP-resistant NAMPT N136A mutant significantly ameliorated vascular aging and extended lifespan in endothelial-specific Tie 2-C/EBPβ transgenic mice. Pharmacologically, AEP inhibitor CP#11A or nicotinamide mononucleotide (NMN) supplementation alleviated age-related vascular decline, with CP#11A showing superior efficacy. These findings support a model in which endothelial senescence contributes to vascular dysfunction and systemic aging mediated by C/EBPβ/AEP signaling, and establish C/EBPβ/AEP as a therapeutic target to preserve vascular integrity and mitigate systemic frailty in aging populations.

INTRODUCTION

Aging is a complex biological process characterized by progressive functional decline across multiple organ systems. A key feature of aging is the deterioration of vascular networks, leading to reduced microvascular density and impaired blood flow (1). These changes contribute significantly to tissue hypoxia, metabolic dysregulation, and the progression of age-related diseases, including neurodegeneration, cardiovascular disorders, and sarcopenia (2). Despite the well-documented association between vascular dysfunction and aging, the underlying molecular mechanisms driving vascular deterioration remain poorly understood.

CCAAT/enhancer binding proteins (C/EBPs) belong to the basic-leucine zipper DNA-binding protein family and are implicated in central nervous system (CNS) inflammation (3). Notably, C/EBPβ exhibits an age-dependent upregulation in neurons, where it transcriptionally activates asparagine endopeptidase (AEP, also known as δ-secretase) (4). Neuronal C/EBPβ/AEP pathway shortens life span via selective GABAergic neuronal degeneration by forkhead box O (FOXO) repression (5). While the C/EBPβ/AEP pathway has been extensively studied in age-related neurodegeneration (6–9), its role in vascular deterioration and systemic aging remains unexplored. Given the widespread expression of C/EBPβ and AEP in central and peripheral tissues including the vasculature, this pathway likely contributes to age-related microvascular rarefaction and multiorgan dysfunction during aging.

Nicotinamide adenine dinucleotide (oxidised form, NAD⁺) is a central metabolic cofactor involved in multiple fundamental biological pathways. NAD⁺ can be synthesized from several precursors, among which nicotinamide is predominantly utilized in mammals (10). NAD⁺ dysregulation has been implicated in a wide range of pathological conditions, including neurodegenerative diseases such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), metabolic disorders, and cardiovascular diseases. For instance, NAD⁺ alleviates AD pathologies partly via the NAD⁺-EVA1C splicing axis (11). Beyond disease associations, NAD⁺ is intimately linked to organismal aging and lifespan regulation across species (12). SIRT1, a member of the NAD⁺-dependent sirtuin family, mediates beneficial effects of dietary restriction and can extend lifespan when overexpressed, while reduced NAD⁺-SIRT1 activity in endothelial cells contributes to age-associated declines in vascular function, blood flow, and endurance (13). Accordingly, NAD⁺ precursor supplementation strategies, including nicotinamide mononucleotide (NMN) and nicotinamide riboside (NR), have been extensively evaluated in clinical studies and effectively increase systemic NAD⁺ metabolite levels. However, their functional outcomes across metabolic and aging-related conditions remain variable (14). For example, NMN improves muscle insulin sensitivity and metabolic remodeling in individuals with prediabetes (15), whereas NR increases NAD⁺ metabolites without significant improvement in insulin sensitivity or mitochondrial function in insulin-resistant subjects (14, 16). These findings suggest that increasing NAD⁺ availability alone may be insufficient to fully restore NAD⁺ homeostasis during aging, highlighting an unresolved gap in understanding the upstream regulatory mechanisms governing age-associated NAD⁺ decline.

In this study, we investigated whether activation of the C/EBPβ/AEP pathway in endothelial cells represents the primary cause of age-related NAD⁺ depletion, endothelial senescence, vascular dysfunction and systemic aging. Our study establishes endothelial C/EBPβ/AEP activation as a key molecular mechanism driving NAD⁺ decline through proteolytic cleavage of NAMPT, the rate-limiting enzyme in NAD⁺ biosynthesis. This pathway triggers systemic NAD⁺ depletion and endothelial senescence, compromising vascular integrity and accelerating aging phenotypes in mice. Importantly, both genetic interventions (AEP knockout or expression of AEP-resistant NAMPT N136A mutant) and pharmacological approaches (AEP inhibitor CP#11A or NMN supplementation) effectively ameliorated these changes in Tie 2-C/EBPβ transgenic mice. Hence, this study identifies C/EBPβ/AEP as a novel therapeutic target for preserving vascular function and mitigating age-related frailty in clinical settings.

RESULTS

Endothelial C/EBPβ/AEP increases with age in human and mice

To investigate whether C/EBPβ/AEP signaling exhibits age-dependent activation in endothelial cells, we performed immunofluorescent (IF) staining analysis of human skeletal muscle and cerebral cortex. Both tissues feature rich vascular architectures and are implicated in systemic and cerebral vascular aging, respectively. Our analysis revealed a significant decline in CD31-positive endothelial cells in aged individuals compared to young controls, concomitant with an upregulation of the C/EBPβ/AEP pathway (Fig. 1, A and B). Consistent with these findings, we observed a similar upregulation in the skeletal muscle and cortex of aged wild-type (WT) C57/BL6 mice (Fig. 1, C and D). Further analysis of WT mouse muscle by immunoblotting (IB) revealed that the protein levels of C/EBPβ and AEP, along with p21, a well-established senescence marker, exhibited an age-dependent increase (Fig. 1, E and F). Quantitative RT-PCR analysis confirmed that the mRNA levels of Cebpb, Lgmn, and Cdkn1a were significantly increased with aging (Fig. 1G). Importantly, AEP enzymatic activity also showed a progressive rise in aged muscle (Fig. 1H). To establish an in vitro model of endothelial senescence, we treated human umbilical vein endothelial cells (HUVECs) with hydrogen peroxide (H₂O₂, 100 μM for 24 hours), which successfully induced cellular senescence, as evidenced by enhanced senescence-associated β-galactosidase (SA-β-gal) (Fig. 1, I and J). IF and IB analyses further confirmed the activation of the C/EBPβ/AEP pathway in H₂O₂-treated HUVECs, along with a sustained increase in p21 expression (Fig. 1, K to N). Collectively, these findings demonstrate that the C/EBPβ/AEP pathway is progressively activated in endothelial cells during aging, both in vivo and in vitro.

Fig. 1. Endothelial C/EBPβ/AEP increases with age in human and mice.

(A and B) Immunofluorescence analysis of CD31, C/EBPβ and AEP in skeletal muscle and cortex sections from young and old human (n = 9). Scale bars, 50 μm. Created in BioRender. Li, B. (2026) https://BioRender.com/eqqq63v. Additional icons were adapted from Servier Medical Art (https://smart.servier.com/), which is licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/deed.en). (C and D) Immunofluorescence analysis of CD31, C/EBPβ and AEP in skeletal muscle and cortex sections from young and old C57BL/6 mice (n = 9). Scale bars, 50 μm. Created in BioRender. Li, B. (2026) https://BioRender.com/eqqq63v. Additional icons were adapted from Servier Medical Art (https://smart.servier.com/), which is licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/deed.en). (E and F) Western blot analysis of C/EBPβ, AEP and p21 in skeletal muscle from wild type (WT) C57BL/6 mice at different ages (7, 26, 60, 180, 365, 630 days postpartum) (n = 3). (G) Quantitative PCR (qPCR) analysis of Cebpb, Lgmn, and Cdkn1a mRNA expression in skeletal muscle from C57BL/6 mice at different ages (n = 3). (H) AEP activity in skeletal muscle from C57BL/6 mice at different ages (n = 3). (I and J) SA-β-gal staining and analysis in PBS or H₂O₂-treated HUVECs (n = 3). Created in BioRender. Li, B. (2026) https://BioRender.com/8ht8xi7. (K and L) Immunofluorescence analysis of CD31, C/EBPβ and AEP in PBS or H₂O₂-treated HUVECs (n = 3). (M and N) Western blot analysis of C/EBPβ, AEP and p21 in PBS or H₂O₂-treated HUVECs (n = 3).

Endothelial AEP overexpression induces vascular dysfunction and systemic frailty in young mice

To explore whether endothelial AEP is implicated in vascular aging and systemic frailty in vivo, we generated transgenic mice that overexpressed AEP only in endothelial cells using the endothelial-cell specific promoter from the Tie 2 gene (fig. S1, A to C). To verify that AEP is expressed exclusively in endothelial cells,we first isolated endothelial cells from the muscle and brain of Tie 2-AEP Tg/Tg mice and their littermates using magnetic-activated cell sorting (MACS) (see the “Single-cell transcriptomics reveals that endothelial-specific AEP overexpression induces endothelial senescence in the muscle and brain of young mice” section). IB analysis confirmed successful overexpression of AEP in the muscle enthothlial cells of Tie 2-AEP Tg/Tg mice compared to WT controls (Fig. 2, A and B), and LE28 imaging showed AEP was strongly activated in the muscle (Fig. 2, C and D). Similarly, in the brain, IB analysis confirmed overexpression in endothelial cells (Fig. 2, E and F), and LE28 imaging showed strong activation (Fig. 2, G and H). Meanwhile, AEP expression levels in non-endothelial cells from muscle, brain, and liver of Tie 2-AEP Tg/Tg mice showed no significant difference compared to WT mice (fig. S2, A to F). Additionally, it has been reported that Tie 2 is expressed not only in endothelial cells but also in hematopoietic cells within the aorta-gonad-mesonephros region, fetal liver and adult bone marrow, as well as in several differentiated hematopoietic cells. Therefore, we isolated hematopoietic cells from Tie 2-AEP Tg/Tg mice and their littermates using flow cytometry (fig. S2G). IB analysis showed no significant difference in AEP expression levels in hematopoietic cells between Tie 2-AEP Tg/Tg and WT mice (fig. S2, H and I). These results confirmed that AEP overexpression in Tie 2-AEP transgenic mice is specifically restricted to endothelial cells rather than other cell types.

Fig. 2. Endothelial AEP overexpression induces vascular dysfunction and systemic frailty in young mice.

(A and B) Western blot of AEP in skeletal muscle endothelial cells from 14-month-old WT and Tie 2-AEP Tg/Tg mice (n = 3). (C and D) Skeletal muscle images and IVIS 100 signal quantification after LE28 injection (n = 5). (E and F) Western blot of AEP in brain endothelial cells (n = 3). (G and H) Brain images and IVIS 100 signal quantification after LE28 injection (n = 5). (I) LSCI images of muscle blood flow (MBF) pre-, during and post-ACh stimulation. (J) Quantification of ACh-induced MBF: baseline (left) and ACh-induced changes (right) (n = 3). (K) LSCI images of cerebral blood flow (CBF) pre-, during, and post-whisker stimulation. (L) Quantification of whisker-induced CBF: baseline (left) and stimulation-induced changes (right) (n = 3). (M) Evans blue dye leakage in ipsilateral cerebral hemisphere (n = 3). (N and O) TEM of skeletal muscle vessel ultrastructure (n = 3). Scale bars, 5 μm (left), 1 μm (right). (P) Kaplan-Meier survival: male WT (n = 24, mean 706 d, median 769 d), male Tg (n = 22, mean 435 d, median 518 d); female WT (n = 20, mean 710 d, median 807 d), female Tg (n = 22, mean 454 d, median 520 d). P < 0.0001 (male), P = 0.0007 (female). (Q) Total frailty index (FI) scores (n = all alive at each time point). (R and S) Frailty phenotypes in WT and Tg mice. Animal numbers below each graph. (T and U) Traces, distance and velocity in open field test (OFT) (n = 8). (V) Duration and distance until exhaustion in treadmill test (n = 8). (W) Spatial working memory assessed by Y maze (n = 8).

To determine the impact of endothelial-specific AEP activation on vascular function, we first assessed endothelial function in Tie 2-AEP mice. In the muscle, Acetylcholine (Ach)-induced vasodilation measured by laser speckle contrast imaging (LSCI) was significantly impaired compared to WT controls, indicating peripheral endothelial dysfunction (Fig. 2, I and J). We next evaluated whether similar deficits occur in the cerebral circulation. In the brain, neurovascular coupling (NVC) responses to whisker stimulation were markedly attenuated, indicating a deficit in NVC (Fig. 2, K and L). Consistently, Evans blue extravasation assays revealed increased blood-brain barrier (BBB) permeability in Tie 2-AEP mice, providing additional evidence of compromised endothelial integrity (Fig. 2M). Furthermore, transmission electron microscopy (TEM) analysis revealed that endothelial cells in the muscle of Tie 2-AEP Tg/Tg mice exhibited severe edema with membrane blurring and cytoplasmic dissolution, whereas striated muscle cells (SMCs) showed only mild morphological alterations (Fig. 2, N and O). Collectively, these findings demonstrate that endothelial-specific AEP activation critically contributes to age-related vascular decline, compromising both compromising endothelial morphology, integrity, and hemodynamic function.

To further investigate whether endothelial AEP contributes to systemic aging and lifespan, we monitored longevity of Tie 2-AEP Tg/Tg mice, which demonstrated much shorter lifespans than WT littermates for both males and females (Fig. 2P). Frailty analysis showed that Tie 2-AEP Tg/Tg mice exhibited much weaker indices than WT mice. We determined the total scores of 13 phenotypes as morbidity. As the animals aged and got closer to death, they manifested several aging phenotypes and were at their highest multi-morbidity risk (Fig. 2, Q to S, and tables S1 and S2). Muscle and brain aging are associated with declined exercise endurance and cognitive function, respectively. Exercise endurance was assessed using open field tests (OFT) and treadmill tests, while cognitive function was evaluated through Y maze and novel object recognition (NOR) tests. We initially evaluated motor and cognitive functions in young (2 months) and aged (21 months) WT mice. OFT indicated aged mice exhibited decreased locomotor activity, demonstrating slower velocity and shorter distance traveled versus young mice, suggestive of weak physical activities (fig. S3, A to C). Endurance behavioral tests revealed that old mice demonstrated much shorter duration and distance runs until exhaustion in high intensity treadmill tests than young mice (fig. S3D). Both NOR and Y maze tests indicated old mice displayed much worse memory than young mice (fig. S3, E to H). These results demonstrate that aging is associated with a decline in cognitive function and exercise endurance. We then evaluated motor and cognitive functions in the WT and Tie 2-AEP Tg/Tg mice. OFT indicated Tie 2-AEP Tg/Tg mice exhibited decreased locomotor activity, demonstrating slower velocity and shorter distance traveled versus WT mice (Fig. 2, T and U). They also revealed much shorter duration and distance runs until exhaustion in high intensity treadmill tests than WT mice (Fig. 2V). NOR and Y maze suggested that Tie 2-AEP Tg/Tg mice displayed much worse memory than WT mice (Fig. 2W and fig. S4A). Collectively, our data suggest that endothelial-specific AEP overexpression promotes vascular dysfunction, which may contribute to systemic aging in young adult mice.

Single-cell transcriptomics reveals that endothelial-specific AEP overexpression induces endothelial senescence in the muscle and brain of young mice

Endothelial senescence is widely recognized as a contributor to vascular dysfunction, which subsequently propels systemic aging (17–19). To reconcile the tissue-specific effects of AEP on endothelial senescence, we performed single-cell RNA sequencing (scRNA-seq) on the quadriceps, brain and liver from Tie 2-AEP Tg/Tg mice and their littermates (Fig. 3A). In the quadriceps, cell type composition analysis confirmed that endothelial cells represented 96.34% of the total population, while macrophages and fibroblasts constituted 1.22% and 2.44%, respectively. In the brain, endothelial cells made up 92.21% of the total population, while ependymal cells (3.53%) and microglia (1.74%) were also contributors. In the liver, the majority of cells were identified as endothelial cells (99.94%), while a small fraction represented hepatocytes (Fig. 3, B and C, and fig. S5A). Beeswarm Milo plot showing the distribution of log₂ fold change (log₂FC) in cell abundance in the quadriceps, brain and liver of WT and Tie 2-AEP Tg/Tg mice, with endothelial cells representing the predominant population (fig. S5B).

Fig. 3. Single-cell transcriptomics reveals that endothelial-specific AEP overexpression induces endothelial senescence in the muscle and brain of young mice.

(A) Schematic diagram of the single-cell RNA sequencing (scRNA-seq) study design. Created in BioRender. Li, B. (2026) https://BioRender.com/viirrv1. (B) T-distributed stochastic neighbor embedding (tSNE) plots showing cell types identified by scRNA-seq in the quadriceps, brain and liver of WT and Tie 2-AEP Tg/Tg mice. Endothelial cells were the predominant population in the quadriceps, brain and liver. (C) Normalized gene expression amounts of different marker genes across the cells isolated from the quadriceps and brain of WT and Tie 2-AEP Tg/Tg mice. (D) Log2-transformed fold change (FC) in the abundance of cellular senescence marker clusters in the quadriceps, brain, and liver of WT and Tie 2-AEP Tg/Tg mice. (E) Cellular senescence scores in quadriceps and brain tissues of WT and Tie 2-AEP Tg/Tg mice, calculated using the AddModuleScore function in Seurat v4, which computes the mean expression of the target gene set normalized by the mean expression of 100 randomly selected control genes matched by expression-level bins. (F) Gene ontology (GO) term enrichment analysis in the quadriceps and brain of WT and Tie 2-AEP Tg/Tg mice. GO terms were selected based on an FDR < 0.05. Top enriched GO terms in biological processes (BP), cellular components (CC), and molecular functions (MF) categories.

To evaluate the senescence-associated transcriptional burden, we utilized the SenMayo gene set, a curated panel of 125 genes associated with senescence and the senescence-associated secretory phenotype (SASP) (20). Among these, 47 genes were detectable in the single-cell expression matrices and were retained for downstream analysis. We then performed differential expression analysis for each of these 47 genes between Tie 2-AEP and WT groups within each tissue. The results are presented as a dot plot showing the average log₂ fold change (color) and statistical significance (−log₁₀ adjusted P value; dot size) (Fig. 3D). Notably, we observed a marked upregulation of multiple senescence-associated genes in both brain and muscle tissues, including representative factors such as Ccl5, Ccl8, Ccl4, Tgfb1, Ifi204 and Cd55, which are well-recognized components of the SASP and broader inflammatory remodeling programs (21, 22).

To further quantify cellular senescence, we calculated a senescence score at the single-cell level, and the resulting scores were compared between Tie 2-AEP and WT groups using two-sided Welch’s t-tests and visualized as violin plots (23). The results showed that endothelial cells from both muscle and brain tissues in Tie 2-AEP Tg/Tg mice exhibited significantly higher cellular senescence scores compared with those in the WT mice (P = 5.68e-76 and P = 1.29e-75, respectively) (Fig. 3E). However, alterations in senescence markers were less evident in the liver endothelial cells of Tie 2-AEP Tg/Tg mice compared to WT controls (P = 1.45e-15) (fig. S5C).

To further elucidate the tissue-specific effects of endothelial AEP overexpression, we performed gene ontology (GO) enrichment analysis comparing WT and Tie 2-AEP Tg/Tg mice across the three tissues. The pathways enriched in the brain and muscle endothelial cells were largely consistent, all including blood vessel morphogenesis, development, and angiogenesis. This consistency highlights the conserved role of AEP overexpression in promoting vascular dysfunction across both tissues. The subtle difference between the two tissues lies in the fact that pathways related to cell migration, tissue migration, and epithelial migration were more prominently enriched in the brain endothelial cells than in the muscle endothelial cells. This may be attributed to the higher intrinsic plasticity and migratory capacity of the brain endothelial cells, which are constantly engaged in neurovascular unit remodeling in response to metabolic demands. Alternatively, it may reflect region-specific vulnerability to AEP-induced vascular dysfunction, with the brain exhibiting a more pronounced compensatory response (Fig. 3F). Liver endothelial cells in Tie 2-AEP Tg/Tg mice also exhibited significant upregulation of pathways involved in blood vessel development, tube morphogenesis, and circulatory system development (fig. S5D). Together, these findings indicate that endothelial cell-specific AEP overexpression induces marked endothelial senescence in both muscle and brain tissues, which may contribute to vascular dysfunction.

Endothelial-specific C/EBPβ overexpression drives aging-related vascular dysfunction and frailty, reversible by AEP genetic deletion in mice

To explore whether endothelial C/EBPβ/AEP pathway is implicated in vascular aging and systemic frailty in vivo, we generated transgenic mice that overexpressed human C/EBPβ only in endothelial cells using the endothelial-cell specific promoter from the Tie 2 gene, and crossed them with AEP^−/− mice (fig. S1, D to G). Western blotting confirmed that overexpression of C/EBPβ elicited robust AEP activation in the quadriceps and brain tissues in Tie 2-C/EBPβ mice compared to WT mice, these biochemical events were apparently alleviated in Tie 2-C/EBPβ/AEP^−/− mice (Fig. 4, A to D). We next examined the SASP genes and found that endothelial C/EBPβ overexpression significantly increased the mRNA levels of Ccl5 and Tgfb1 in both muscle and brain tissues, which were largely rescued by AEP knockout (Fig. 4E). These findings are consistent with our single-cell RNA-sequencing data and suggest that activation of endothelial C/EBPβ is associated with the induction of senescence-related inflammatory programs in an AEP-dependent manner.

Fig. 4. Endothelial-specific C/EBPβ overexpression drives aging-related vascular dysfunction and frailty, reversed by AEP genetic deletion in mice.

(A and B) Western blot analysis of C/EBPβ and AEP in the muscle from 14-month-old WT, Tie 2-C/EBPβ Tg/Tg and Tie 2-C/EBPβ Tg/Tg/AEP^−/− mice (n = 3). (C and D) Western blot analysis of C/EBPβ and AEP in the brain (n = 3). (E) qPCR of Ccl5 and Tgfb1 in the muscle and brain (n = 3). (F) Quantification of ACh-induced MBF: baseline (left) and ACh-induced changes (right) (n = 3). (G) Quantification of whisker-induced CBF: baseline (left) and stimulation-induced changes (right) (n = 3). (H) Evans blue dye leakage in ipsilateral cerebral hemisphere (n = 3). (I and J) TEM of skeletal muscle vessel ultrastructure (n = 3). Scale bars, 5 μm (left), 1 μm (right). (K) Kaplan-Meier survival: male, WT (n = 22, mean 753 d, median 804 d), male, Tg/Tg (n = 28, mean 381 d, median 390 d), male, Tg/Tg/AEP^−/− (n = 20, mean 666 d, median 754 d); female, WT (n = 24, mean 726 d, median 774 d), female, Tg/Tg (n = 26, mean 435 d, median 489 d), female, Tg/Tg/AEP^−/− (n = 21, mean 643 d, median 728 d). P < 0.0001 (Tg/Tg vs WT, male), P < 0.0001 (Tg/Tg/AEP^−/− vs Tg/Tg, male). P < 0.0001 (Tg/Tg vs WT, female), P = 0.0012 (Tg/Tg/AEP^−/− vs Tg/Tg, female). (L) Total FI scores (n = all alive at each time point). (M and N) Frailty phenotypes in WT, Tg/Tg and Tg/Tg/AEP^−/− mice. Animal numbers below each graph. (O and P) Traces, distance and velocity in OFT (n = 8). (Q) Duration and distance until exhaustion in treadmill test (n = 8). (R) Object recognition memory assessed by NOR (n = 8).

We next assessed vascular function in mice with endothelial-specific C/EBPβ overexpression. Tie 2-C/EBPβ mice exhibited significantly reduced ACh-induced vasodilation in the muscle, attenuated NVC responses, and increased BBB permeability (Fig. 4, F to H). TEM further revealed severe edema, membrane disruption, and cytoplasmic dissolution in the muscle of Tie 2-C/EBPβ Tg/Tg mice (Fig. 4, I and J). Notably, these structural and functional impairments were largely rescued in Tie 2-C/EBPβ mice with AEP knockout, indicating that C/EBPβ-induced endothelial dysfunction is largely dependent on AEP.

We further investigated whether endothelial C/EBPβ/AEP pathway contributes to systemic aging and lifespan. Survival curves showed both male and female Tie 2-C/EBPβ Tg/Tg mice displayed much shorter lifespan than WT littermates, and depletion of AEP from these mice greatly extended life expectancy of Tie 2-C/EBPβ/AEP^−/− mice (Fig. 4K). Tie 2-C/EBPβ Tg/Tg mice demonstrated noticeable frailty indices versus WT mice, which were attenuated in Tie 2-C/EBPβ/AEP^−/− mice (Fig. 4, L to N). Behavioral tests for motor and cognition functions showed that Tie 2-C/EBPβ Tg/Tg mice revealed much poorer performance than WT mice, which were rescued when AEP was knocked out, supporting that AEP is primarily accountable for the behavioral deficits in Tie 2-C/EBPβ Tg/Tg mice (Fig. 4, O to R, and fig. S4B). Hence, activation of the endothelial C/EBPβ/AEP leads to marked impairments in motor and cognitive function, as well as systemic frailty in young adult mice. Together, these results indicate that endothelial C/EBPβ promotes endothelial senescence and contributes to vascular dysfunction and systemic aging in an AEP-dependent manner.

AEP cleaves NAMPT at N136, impairing NAD⁺ biosynthesis and promoting endothelial senescence

Consistent with previous studies showing that declining NAD⁺ levels contribute to senescence and vascular aging, we sought to determine whether the endothelial C/EBPβ/AEP axis promotes endothelial senescence by modulating NAD⁺ homeostasis (13, 21, 22, 24). In the mammalian NAD⁺ metabolism, NAMPT is the rate-limiting enzyme that catalyzes NAD⁺ biosynthesis from nicotinamide (25). To investigate if the C/EBPβ/AEP pathway directly targets NAMPT, we conducted proteolytic assay with GST-NAMPT and recombinant AEP proteins for different time points. IB analysis showed that NAMPT was time-dependently cleaved, revealed by both anti-NAMPT and anti-GST antibodies (Fig. 5A and fig. S6A). AEP is a cysteine protease with C189 as the key active residue (26). Co-transfection demonstrated that GST-NAMPT was evidently cleaved by myc-AEP, and this process was abolished in dominant-negative C189S mutant transfected cells, underscoring that AEP is accountable for cutting NAMPT into fragments (Fig. 5B and fig. S6B). AEP enzymatic activities in these cells were validated by in vitro assay (Fig. 5C). CP#11A is an inhibitor of AEP developed by our group through high-throughput screening and modification (IC₅₀ ∼ 5–10 nM), which specifically inhibits AEP activity without affecting other cysteine proteases (27, 28). To further confirm that AEP is responsible for NAMPT proteolytic truncation, we found that its specific inhibitor CP#11A abrogated NAMPT fragmentation. As expected, CP#11A strongly blocked AEP enzymatic activity and its cleavage on NAMPT (Fig. 5D and fig. S6C). AEP enzymatic activities in these cells were validated by in vitro assay (Fig. 5E). To determine the exact cutting sites on NAMPT by AEP, we purified GST-NAMPT recombinant proteins and conducted an AEP cleavage assay. LC/MS/MS study with AEP-cleaved NAMPT fragment revealed that N136 was the proteolytic cutting site (fig. S6D). NAMPT possesses numerous N residues in the polypeptide. Only mutation of N136 into A but not other locations totally blocked NAMPT cleavage by AEP, suggestive of N136 in NAMPT as a major cutting site by AEP (Fig. 5F and fig. S6E). This residue is conserved among mouse, rat and human NAMPT (fig. S6F). Next, we detected NAD⁺ concentrations with HUVECs, and found that both NAMPT full length (FL) and N136A mutant strongly mediated NAD⁺ biosynthesis as compared to N136 truncated N-terminal or C-terminal fragments (Fig. 5G). Hence, AEP cuts NAMPT at N136 residue and cripples its enzymatic activity, diminishing NAD⁺ biosynthesis.

Fig. 5. AEP cleaves NAMPT at N136 and severely impairs NAD⁺ biosynthesis.

(A) Time-dependent cleavage of GST-NAMPT by recombinant AEP in HEK293 lysates. (B) mGST-NAMPT fragmentation by AEP, but not by C189S mutant, in HEK293 cells. (C) AEP enzymatic activity (n = 3). (D) NAMPT fragmentation blocked by AEP inhibitor CP#11A. (E) AEP enzymatic activity (n = 3). (F) N136A mutant resists AEP cleavage in HEK293 lysates. (G) NAD⁺ levels (n = 3). (H and I) Western blot of C/EBPβ, AEP, NAMPT FL, NAMPT CL, NAMPT C137, SIRT1 in HUVECs (n = 3). (J and K) Immunofluorescence of CD31 and NAMPT C137 in HUVECs (n = 3). Scale bar, 100 μm. (L) NAD⁺ levels (n = 3). (M and N) Western blot of AEP, NAMPT FL, NAMPT CL, NAMPT C137 in HUVECs (n = 3). (O and P) Immunofluorescence of CD31 and NAMPT C137 in HUVECs (n = 3). Scale bar, 100 μm. (Q) NAD⁺ levels (n = 3). (R and S) Immunofluorescence of CD31 and NAMPT C137 in muscle (n = 3). Scale bar, 50 μm. (T) NAD⁺ levels in muscle (n = 3). (U) Correlation between NAD⁺ and NAMPT C137 in muscle. (V and W) Immunofluorescence of CD31 and NAMPT C137 in brain (n = 3). Scale bar, 50 μm. (X) NAD⁺ levels in brain (n = 3). (Y) Correlation between NAD⁺ and NAMPT C137 in brain. (Z and AA) Immunofluorescence of CD31 and NAMPT C137 in muscle (n = 3). Scale bar, 50 μm. (AB) NAD⁺ levels in muscle (n = 3). (AC) Correlation between NAD⁺ and NAMPT C137 in muscle. (AD and AE) Immunofluorescence of CD31 and NAMPT C137 in brain (n = 3). Scale bar, 50 μm. (AF) NAD⁺ levels in brain (n = 3). (AG) Correlation between NAD⁺ and NAMPT C137 in brain.

To verify whether AEP cleaves NAMPT at N136 and severely impairs NAD⁺ biosynthesis at the cellular level. We first generated a rabbit polyclonal antibody that specifically recognized C137 in NAMPT to see whether NAMPT 137–491 fragment is the direct product of AEP truncating NAMPT. In vitro, LV-mediated C/EBPβ overexpression in HUVECs significantly enhanced AEP expression and promoted AEP-mediated cleavage of NAMPT compared to control vector. The cleavage effect was abolished by LV-mediated AEP knockdown, demonstrating that the NAMPT 137–491 fragment is generated through AEP-dependent proteolytic processing of NAMPT FL (Fig. 5, H to K, and fig. S7, A and B). AEP enzymatic activities in these cells were validated by in vitro assay (fig. S7C). NAD⁺ precursors reverse aging partially through activating SIRT1, which mediate the benefits of exercise and DR (13). AEP-mediated cleavage of NAMPT was associated with a significant reduction in SIRT1 levels, whereas AEP knockdown robustly enhanced SIRT1 expression (Fig. 5, H and I). NAD⁺ levels were strongly reduced in C/EBPβ-overexpressing cells, which were substantially reversed when AEP was abated (Fig. 5L). As expected, inflammatory cytokines were all elevated, which were diminished when AEP expression was suppressed (fig. S7D). Mitochondrial dysfunctions, operationally defined as a decreased mitochondrial membrane potential, are hallmarks of cell senescence. We monitored the mitochondrial membrane potential by JC-1 dye. JC-1 fluorescent analysis indicated that C/EBPβ profoundly elicited mitochondrial membrane potentiation deficits, which were significantly alleviated when AEP was deleted (fig. S7, E and F). Moreover, C/EBPβ overexpression triggered substantial oxidative stress, as indicated by 4-HNE staining, which was abrogated by AEP depletion (fig. S7, G and H). SA-β-gal staining revealed that C/EBPβ overexpression incurred massive cell senescence that was greatly attenuated when AEP was knocked down (fig. S7, I and J). On the other hand, overexpression of AEP in HUVECs stimulated its activation and conspicuous NAMPT cleavage (Fig. 5, M to P). AEP enzymatic activities in these cells were validated by in vitro assay (fig. S7K). Accordingly, NAD⁺ levels declined (Fig. 5Q). As expected, inflammatory cytokines were steadily augmented (fig. S7L). AEP markedly triggered mitochondrial membrane potentiation reduction (fig. S7, M and N). SA-β-gal staining revealed much more extensive cell senescence in AEP overexpressed cells than control cells (fig. S7, O and P). These results indicate that C/EBPβ/AEP impairs NAD⁺ synthesis via proteolytic cleavage of NAMPT, thereby promoting endothelial senescence in vitro.

We further confirmed that AEP cleaves NAMPT at N136 and severely impairs NAD⁺ biosynthesis in vivo. IF staining revealed that CD31 signals were reduced in the muscle and brain of Tie 2-C/EBPβ Tg/Tg mice, accompanied by apparent AEP activation and NAMPT fragmentation. These phenomena were alleviated in Tie 2-C/EBPβ Tg/Tg/AEP^−/− mice (Fig. 5, R and S, V and W). Notably, NAD⁺ was significantly reduced in the muscle and brain of Tie 2-C/EBPβ Tg/Tg mice, which was restored in Tie 2-C/EBPβ Tg/Tg /AEP^−/− mice, inversely coupled with NAMPT C137 fragments (Fig. 5, T and U, X and Y). Similarly, IF staining revealed that CD31 signals were reduced in the muscle and brain of Tie 2-AEP Tg/Tg mice, accompanied by NAMPT fragmentation (Fig. 5, Z and AA, AD and AE). NAD⁺ was significantly reduced in the muscle and brain of Tie 2-AEP Tg/Tg mice, inversely coupled with NAMPT C137 fragments (Fig. 5, AB and AC, AF and AG). These data, along with the preceding results, support that the C/EBPβ/AEP directly mediates NAMPT proteolytic cleavage, leading to a marked disruption of NAD⁺ biosynthesis and endothelial senescence in vivo.

Blockade of NAMPT cleavage by AEP restores NAD⁺ levels, attenuates endothelial senescence, improves vascular function, and alleviates frailty in Tie 2-C/EBPβ Tg/Tg mice

To investigate the functional contribution of AEP-mediated NAMPT cleavage to endothelial senescence, vascular dysfunction, and systemic aging, we administered intravenous injections of endothelial cell-specific rAAV-Tie 2-hNAMPT and rAAV-Tie 2-hNAMPT N136A to 10-month-old Tie 2-C/EBPβ Tg/Tg mice (Fig. 6A). NAD⁺ levels were significantly elevated in the muscle and brain tissues of both NAMPT and NAMPT N136A mice compared to controls, with a more pronounced increase observed in the NAMPT N136A group (Fig. 6B). IB analysis revealed elevated levels of NAMPT FL in both NAMPT- and NAMPT N136A-overexpressing mice compared to controls across both muscle and brain tissues. While NAMPT overexpression increased C137 fragment generation, N136A mutation effectively blocked AEP-mediated proteolytic cleavage (Fig. 6, C and D, G and H). IF staining demonstrated that NAMPT overexpression enhanced vascular density, with N136A mutant exhibiting more pronounced effects (Fig. 6, E and F, I and J). NAMPT N136A mutant reduced mRNA levels of Ccl5 and Tgfb1 in both muscle and brain tissues (Fig. 6K). ACh-induced vasodilation, NVC responses, and BBB integrity were all markedly improved in NAMPT N136A-overexpressing mice, indicating that blockade of NAMPT cleavage effectively restores endothelial function (Fig. 6, L to N).

Fig. 6. Blockade of NAMPT cleavage by AEP restores NAD⁺ levels, attenuates endothelial senescence, improves vascular function, and alleviates frailty in Tie 2-C/EBPβ Tg/Tg mice.

(A) Timeline of CTL, NAMPT and NAMPT N136A virus administration. Icons adapted from Servier Medical Art (https://smart.servier.com/), licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/deed.en). (B) NAD⁺ levels (n = 6). (C and D) Western blot of NAMPT, C137 and AEP in muscle (n = 3). (E and F) CD31 immunofluorescence in muscle (n = 3). Scale bars, 50 μm. (G and H) Western blot of NAMPT, C137 and AEP in brain (n = 3). (I and J) CD31 immunofluorescence in brain (n = 3). Scale bars, 50 μm. (K) qPCR of Ccl5 and Tgfb1 (n = 3). (L and M) ACh-induced MBF and whisker-induced CBF (n = 3). (N) Evans blue leakage in ipsilateral cerebral hemisphere (n = 3). (O) Kaplan-Meier survival: male, CTL (n = 20, mean 527 d, median 516 d), male, NAMPT (n = 20, mean 584 d, median 647 d), male, NAMPT N136A (n = 20, mean 662 d, median 713 d); female, CTL (n = 20, mean 559 d, median 594 d), female, NAMPT (n = 20, mean 611 d, median 648 d), female, NAMPT N136A (n = 20, mean 635 d, median 686 d). P = 0.2069 (NAMPT vs CTL, male), P = 0.0070 (NAMPT N136A vs CTL, male). P = 0.0108 (NAMPT vs CTL, female), P = 0.0057 (NAMPT N136A vs CTL, female). (P) Total FI scores. (Q and R) Frailty phenotypes. (S) Treadmill duration and distance (n = 10). (T and U) OFT traces, distance and velocity (n = 10). (V) Y maze spatial working memory (n = 10). (W) NOR object recognition memory (n = 10).

Survival curves showed NAMPT or NAMPT N136A overexpression significantly elongated the lifespan of male Tie 2-C/EBPβ Tg/Tg mice compared to control virus (Fig. 6O). Frailty analysis suggested that the indices were significantly improved in both NAMPT- and NAMPT N136A-overexpressing mice versus control mice (Fig. 6, P to R). Behavioral tests for motor and cognition functions showed that NAMPT- and NAMPT N136A-overexpressing mice revealed much better performance than control mice (Fig. 6, S to W). Notably, across nearly all measures, the N136A mutant demonstrated superior functional potency compared with NAMPT. Collectively, these findings suggest that preventing AEP-mediated NAMPT cleavage confers greater improvements in NAD⁺ biosynthesis, endothelial senescence, vascular dysfunction, and systemic aging than NAMPT supplementation alone, supporting AEP-mediated NAMPT cleavage as an important regulatory mechanism in this process.

Genetic inhibition of C/EBPβ or AEP restores youthful NAD⁺ levels and rescues cognitive and motor functions in aged mice

To investigate whether downregulation of C/EBPβ or AEP could ameliorate natural aging in mice, we employed aged WT and C/EBPβ^+/− and AEP^+/− mice (figs. S8A and S9A). We found that NAMPT C137 levels were markedly reduced in the muscle of C/EBPβ^+/− and AEP^+/− mice, concomitant with elevated SIRT1 expression (figs. S8, A to E and S9, A to E). AEP enzymatic activities in these cells were validated by in vitro assay (figs. S8, F to H and S9, F to H). Consequently, NAD⁺ concentrations were increased in comparison to aged WT littermates (figs. S8I and S9I). OFT and treadmill tests supported that knockdown either C/EBPβ or AEP substantially alleviated crippled physical activities (figs. S8, J to L and S9, J to L). Y maze and NOR assays also demonstrated that cognitive functions were improved in C/EBPβ^+/− and AEP^+/− mice as compared to aged WT mice (figs. S8, M and N and S9, M and N). These results suggest that systemic knockdown of C/EBPβ and AEP restores youthful NAD⁺ levels and rescues cognitive and motor functions in aged mice.

Endothelial-specific C/EBPβ or AEP inhibition restores youthful NAD⁺ levels and rescues cognitive and motor functions in aged mice

To further validated whether conditional knockdown of C/EBPβ and AEP specifically in endothelial cells could produce the same effects, we administered intravenous injections of endothelial cell-specific rAAV-Tie 2-sh-C/EBPβ and rAAV-Tie 2-sh-AEP to 20-month-old WT mice (figs. S10A and S11A). We found that NAMPT C137 levels were markedly reduced in the muscle of mice injected with rAAV-Tie 2-sh-C/EBPβ and rAAV-Tie 2-sh-AEP, concomitant with elevated SIRT1 expression (figs. S10, B and C and S11, B and C). AEP enzymatic activities in these cells were validated by in vitro assay (figs. S10D and S11D). IF staining revealed that CD31 signals were increased in the muscle and brain of mice injected with rAAV-Tie 2-sh-C/EBPβ and rAAV-Tie 2-sh-AEP (figs. S10, E and F and S11, E and F). Consequently, NAD⁺ concentrations were increased (figs. S10G and S11G). Behavioral tests showed that motor and cognition functions were improved in mice injected with rAAV-Tie 2-sh-C/EBPβ and rAAV-Tie 2-sh-AEP, compared to aged WT mice (figs. S10, H to L and S11, H to L). Together, these observations demonstrate that endothelial-specific C/EBPβ or AEP inhibition restores youthful NAD⁺ levels and rescues cognitive and motor functions in aged mice.

AEP inhibitor CP#11A treatment rejuvenates Tie 2-C/EBPβ Tg/Tg mice without adverse effects

NAD⁺ booster NMN improves blood flow and increases endurance in aged mice by promoting SIRT1-dependent increases in capillary density (13). To evaluate the therapeutic potential of AEP inhibition against C/EBPβ-mediated vascular decline and systemic aging, we fed 10-month-old Tie 2-C/EBPβ Tg/Tg mice with vehicle, NMN or AEP inhibitor CP#11A (Fig. 7A). NAD⁺ assay showed that both NMN and CP#11A increase NAD⁺ levels in the muscle and brain of Tie 2-C/EBPβ Tg/Tg mice, with a more pronounced increase observed in the CP#11A group (Fig. 7B). IB analysis demonstrated that CP#11A strongly blocked active AEP levels and blunted NAMPT N136 cleavage in the skeletal muscle and brain of Tie 2-C/EBPβ Tg/Tg mice, SIRT1 levels were prominently escalated after NMN or CP#11A treatment (Fig. 7, C and D, G and H). Consequently, IF revealed that CP#11A treatment significantly increased vascular density (Fig. 7, E and F, I and J). Treatment with NMN and CP#11A reduced mRNA levels of Ccl5 and Tgfb1 in both muscle and brain tissues (Fig. 6K). ACh-induced vasodilation, NVC responses, and BBB integrity were all markedly improved under CP#11A treatment, indicating a robust rescue of endothelial dysfunction (Fig. 7, L to N).

Fig. 7. AEP inhibitor CP#11A treatment rejuvenates Tie 2-C/EBPβ Tg/Tg mice without adverse effects.

(A) Timeline of CTL, NMN or CP#11A treatment. Icons from Servier Medical Art (https://smart.servier.com/), licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/deed.en). (B) NAD⁺ levels (n = 6). (C and D) Western blot of C/EBPβ, AEP, C137 and SIRT1 in muscle (n = 3). (E and F) CD31 immunofluorescence in muscle (n = 3). Scale bars, 50 μm. (G and H) Western blot of C/EBPβ, AEP, C137 and SIRT1 in brain (n = 3). (I and J) CD31 immunofluorescence in brain (n = 3). Scale bars, 50 μm. (K) qPCR of Ccl5 and Tgfb1 (n = 3). (L and M) ACh-induced MBF and whisker-induced CBF (n = 3). (N) Evans blue leakage in ipsilateral cerebral hemisphere (n = 3). (O) Kaplan-Meier survival: male, CTL (n = 22, mean 519 d, median 521 d), male, NMN (n = 20, mean 620 d, median 673 d), male, CP#11A (n = 24, mean 624 d, median 718 d); female, CTL (n = 20, mean 529 d, median 530 d), female, NMN (n = 22, mean 627 d, median 668 d), female, CP#11A (n = 20, mean 652 d, median 723 d). P = 0.0175 (NMN vs CTL, male), P = 0.0088 (CP#11A vs CTL, male). P = 0.0156 (NMN vs CTL, female), P = 0.0035 (CP#11A vs CTL, female). (P) Total FI scores. (Q) Frailty phenotypes. (R and S) Tumor and disease burden. CTL (n = 42), NMN (n = 42), CP#11A (n = 44). (T) Treadmill duration and distance (n = 10). (U and V) OFT traces, distance and velocity (n = 10). (W) Y maze spatial working memory (n = 10). (X) NOR object recognition memory (n = 10).

Survival curves showed both NMN and CP#11A strongly elongated the lifespan in male Tie 2-C/EBPβ Tg/Tg mice (Fig. 7O). Frailty assay showed that both compounds significantly alleviated the aging indices, with CP#11A conferring augmented functional potency (Fig. 7, P and Q). In spite of longer lifespan, NMN and CP#11A showed no significant differences in tumor and disease burdens compared with control mice (Fig. 7, R and S). Furthermore, comprehensive hematological, biochemical, and urinalysis profiles demonstrated no evidence of hepatic or renal toxicity following CP#11A treatment, consistent with our previous safety assessments (tables S3 to S5). Behavioral tests for motor and cognition functions showed that both NMN and CP#11A rescued cognitive and motor functions in Tie 2-C/EBPβ Tg/Tg mice, with CP#11A exhibiting a more pronounced effect (Fig. 7, T to X).

Although our previous research shows that CP#11A specifically inhibits AEP enzyme activity without affecting other enzymes and biochemical effects, to further investigate whether CP#11A possesses any off-target effects, we fed Tie 2-C/EBPβ Tg/Tg/AEP^−/− mice with CP#11A. The concentration of CP#11A in serum is significantly elevated in the CP#11A group compared to the control (fig. S12A). However, there was no significant difference in the relative NAD⁺ levels between the control and CP#11A-treated groups (fig. S12B). OFT, treadmill tests, NOR and Y maze tests revealed no significant differences between the two groups (fig. S12, C to G). These results indicate that CP#11A treatment had no off-target effects, does not further restore NAD⁺ levels or rescue cognitive and motor function in Tie 2-C/EBPβTg/Tg/AEP^−/− mice. Together, these results support the notion that inhibition of AEP-mediated NAMPT cleavage restores NAD⁺ levels, attenuates endothelial senescence, and improves vascular function, with concomitant mitigation of systemic aging, thereby highlighting AEP as a potential therapeutic target.

DISCUSSION

Our previous studies have established that C/EBPβ/AEP pathway, which escalates age-dependently across multiple organs, plays a vital role in mediating aging-related pathologies, including AD, PD, atherosclerosis, diabetes, and cancers (6–9). Given that aging is a systemic process affecting multiple organs synchronously, and that capillary endothelial cells are ubiquitous across tissues with broad expression of C/EBPβ/AEP, we hypothesized that endothelial C/EBPβ/AEP activation drives vascular deterioration and systemic aging. Our findings demonstrates that progressive upregulation of endothelial C/EBPβ/AEP signaling impairs NAD⁺ biosynthesis and promotes endothelial senescence, which may contribute to vascular dysfunction and systemic aging. Thus, this study extends our previous work by revealing a conserved mechanism linking pathological and physiological aging.

C/EBPβ, activated by inflammation or oxidative stress, drives the expression of inflammatory cytokines (e.g., IL-1β and IL-6) and upregulates key effectors such as AEP and ApoE4 in neurodegenerative diseases (29–32). Endothelial-specific overexpression of C/EBPβ in mice markedly increases AEP levels, leading to NAMPT cleavage, NAD⁺ depletion, and a reduction in SIRT1, likely through indirect mechanisms (Figs. 4 and 5). These mice exhibit increased cellular senescence, vascular degeneration, and extensive frailty (Fig. 4 and fig. S4). The strategy that we used to construct Tie 2-C/EBPβ and Tie 2-AEP transgenic mice differ from traditional Cre/flox systems, as they overexpress C/EBPβ and AEP from birth. However, this approach is advantageous for observing the effects of elevated C/EBPβ and AEP on lifespan and aging in a relatively short period, as certain genetic mutations or augmentation causing premature aging begin in early life. However, this construction strategy does not fully mimic the progressive escalation in C/EBPβ and AEP levels during human aging. Accordingly, endothelial knockdown of either C/EBPβ or AEP ameliorates these aging-associated phenotypes compared to age-matched WT littermates (figs. S10 and S11), and C/EBPβ⁺/⁻ mice display an extended lifespan (5).

In the endothelial cell-specific AEP overexpression mouse model, single-cell RNA sequencing of the muscle and brain tissues revealed that, although Cdkn2a and Cdkn2b did not exhibit significant changes, their absence does not preclude the presence of cellular senescence, as senescence is a highly heterogeneous process. Indeed, senescent cells can engage distinct transcriptional programs without necessarily showing robust upregulation of canonical markers such as Cdkn2a or Cdkn2b (33, 34). Another possible explanation is that, in Tie 2-AEP transgenic model, AEP overexpression induces an early or intermediate stage of senescence, in which Cdkn2a and Cdkn2b have not yet accumulated to detectable levels, while other senescence-associated programs are already activated. Notably, we observed a marked upregulation of multiple senescence-associated genes in both brain and muscle tissues. Representative factors such as Ccl5 and Tgfb1 are well-recognized components of the SASP and broader inflammatory remodeling programs, which are regulated by NAD⁺ metabolism and NAD⁺-dependent signaling pathways (21, 22). This observation also provides a mechanistic explanation for how AEP-mediated cleavage of NAMPT, leading to reduced NAD⁺ biosynthesis, may be associated with the induction of senescence.

Our observations are consistent with previous reports that aging is associated with increased endothelial apoptosis and reduced neovascularization, leading to vascular loss and subsequent sarcopenia and functional decline (35–39). NAD⁺-related signaling pathways, including endothelial NAD⁺-H₂S axis disruption, have been implicated in vascular aging (13), yet the upstream causes of age-associated NAD⁺ decline remain incompletely understood. Although NAD⁺ precursor supplementation effectively increases systemic NAD⁺ levels in humans, its functional benefits remain inconsistent across aging-related conditions, suggesting incomplete restoration of tissue NAD⁺ homeostasis (14). In this context, our study reveals that NAMPT, a rate-limiting enzyme for NAD⁺ production, is cleaved by AEP during aging, resulting in impaired NAD⁺ biosynthetic capacity and disruption of NAD⁺ homeostasis. Moreover, CP#11A shows comparable or even superior therapeutic effects to NMN in improving aging phenotypes in premature aging Tie 2-C/EBPβ transgenic mice (Fig. 7), suggesting that pharmacological inhibition of AEP may represent a promising strategy to complement NAD⁺ precursor supplementation in aging and age-associated diseases.

Our previous studies have demonstrated that CP#11A is a highly selective AEP inhibitor that shows no activity against related proteases such as caspases, with chronic administration exhibiting excellent safety profiles (27, 28). The compound possesses favorable pharmacokinetic properties, including good oral bioavailability and blood-brain barrier penetration (40). To further investigate whether CP#11A displays any off-target effects, we feed Tie 2-C/EBPβ Tg/Tg/AEP^−/− mice with CP#11A in this study. Importantly, CP#11A treatment showed no significant effects on NAD⁺ levels or behavioral performance compared to vehicle controls (fig. S12), demonstrating its remarkable target specificity. These results confirm that CP#11A’s therapeutic effects are mediated exclusively through AEP inhibition, as it produces no additional benefits in AEP-deficient models. Together with its established pharmacological profile, these findings strongly support CP#11A as a promising clinical candidate for targeted AEP inhibition therapies.

In summary, our study demonstrates that endothelial C/EBPβ/AEP activation acts as a pivotal molecular mechanism for age-related NAD⁺ decline by mediating the proteolytic cleavage of NAMPT, thereby promoting endothelial senescence, which impairs vascular integrity and accelerates systemic aging in mice. These findings provide mechanistic insight into the vascular aging while establishing C/EBPβ/AEP signaling as a novel therapeutic target for interventions aimed at preserving endothelial function and mitigating age-related vascular deterioration. The demonstration of this previously unrecognized pathway opens new avenues for developing targeted therapies against vascular aging and its associated comorbidities.

MATERIALS AND METHODS

Mice

Wild-type C57BL/6 J mice were ordered from the Jackson Laboratory (000664). C/EBPβ knockout mice have been described (41). Since the homozygous mutation is lethal on pure strain backgrounds, Cebpb mice were maintained as heterozygotes on two separate strain backgrounds (C57BL/6 and 129Sv). The AEP knockout mice on a mixed 129/Ola and C57BL/6 background were generated as reported (42). Tie 2-AEP mice and Tie 2-C/EBPβ mice on a C57BL/6 background were generated by Cyagen Company. The homology between mouse and human C/EBPβ genes is around 70%. Therefore, constructing a human C/EBPβ and AEP transgenic animal model is advantageous for better studying the role of the human C/EBPβ/AEP pathway in aging. Tie 2-C/EBPβ mice were crossed with AEP^−/− mice to generate Tie 2-C/EBPβ/AEP^−/− mice. Age-matched WT littermates with the same genetic background were used as controls. All the mice were bred in specific pathogen-free facilities at the Shenzhen Institutes of Advanced Technology (SIAT), Chinese Academy of Sciences (CAS), maintained on a 12-hour light/dark cycle with unrestricted access to water and food. All animal experiments followed guidelines of the Institutional Animal Care and Use Committee of Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (Approval No. LLSQ2111110006).

Experimental design and timeline

In Tie 2-AEP Tg/Tg, Tie 2-C/EBPβ Tg/Tg, and Tie 2-C/EBPβ Tg/Tg/AEP^−/− mouse models and their WT littermates, assessments were conducted according to a defined timeline. Behavioral tests including OFT, treadmill test, Y-maze, and NOR test, were performed at 12 months of age. At 13 months, in vivo assessments were carried out, including measurements of MBF and CBF, as well as AEP activity detection. Evaluation of BBB integrity along with qPCR, IB, IF and TEM analyses was conducted at 14 months of age. FI assessments were performed longitudinally from 3 to 21 months of age, and survival was monitored continuously until the endpoint.

For heterozygous intervention, C/EBPβ^+/− AEP^+/− mice and their WT littermates underwent OFT, treadmill test, Y-maze, and NOR test at 22 months of age, followed by in vivo AEP activity detection at 23 months of age. At 24 months of age, mice were sacrificed for qPCR, IB, and IF analyses.

For AAV-mediated endothelial-specific intervention, 20-month-old WT mice received intravenous injections of rAAV-Tie 2-sh-C/EBPβ and rAAV-Tie 2-sh-AEP. They underwent OFT, treadmill test, Y-maze, and NOR test at 23 months of age. At 24 months of age, mice were sacrificed for qPCR, IB, and IF analyses.

For NAMPT mutant and drug intervention studies, Tie 2-C/EBPβ Tg/Tg mice received intravenous administration of rAAV-Tie 2-control, rAAV-Tie 2-NAMPT, or rAAV-Tie 2-NAMPT N136A at 10 months of age. In parallel, pharmacological treatments (CP#11A or NMN) were initiated at 10 months of age. NMN (500 mg/kg) was administered via the drinking water, while CP#11A (7.5 mg/kg) was delivered through the diet. Behavioral tests began at 12 months of age. At 13 months, in vivo vascular assessments, including MBF and CBF measurements, were performed. Evaluation of BBB integrity along with qPCR, IB, and IF analyses was conducted at 14 months of age. FI was assessed longitudinally from 12 to 21 months of age, and survival was monitored continuously until the endpoint.

Human samples

Postmortem human muscle and brain samples for IF staining were obtained from the Chinese Brain Bank Center (CBBC). Written informed consent was acquired from all participants or their legal representatives prior to sample collection. This study was approved by the Ethics Committee of the Chinese Brain Bank Center (Approval No. 2024-scuec-046). All procedures involving human tissues were conducted in strict compliance with the ethical guidelines of the Institutional Review Board (IRB) of Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences. Subject baseline characteristics for the samples are listed in table S6, Supporting Information. All tissues were fixed with 4% paraformaldehyde.

Cells

HEK293 cells were grown in DMEM supplemented with 10% FBS and 1% Penicillin/Streptomycin (Pen/Strep) at 37°C with 5% CO₂. HUVECs were cultured in endothelial cell culture medium supplemented with 10% FBS and 1% Pen/Strep at 37°C with 5% CO₂.

Survival

Survival experiments were performed as previously described (43). The principle endpoint of lifespan study was natural death. We recorded the age at which mice were found dead or selected for euthanasia (a procedure for mice deemed unlikely to survive for the next 48 h and are in enormous discomfort). The criteria for euthanasia were based on an independent assessment by a veterinarian, according to AAALAC guidelines and only where the condition of the animal was considered (44). Severe lethargy, rapid weight loss (over two weeks >20%), severe distended abdomen and body condition score with signs of pain (grimace), inability to move despite the stimuli, severe ulcer or bleeding tumor, severe temperature loss with abnormal breathing rate. Animals found dead or euthanized were necropsied for pathology score.

Frailty index

Frailty Index was performed as previously described (45). For the subjective properties of the assessments, all measurements were completely blinded. These assessments indicate age-associated deterioration of health and include evaluation of the animal musculoskeletal system, the vestibulocochlear/auditory systems, the ocular and nasal systems, the digestive system, the urogenital system, the respiratory system, signs of discomfort, body weight and body surface temperature. 0 is assigned if no sign of frailty is observed and the animal is healthy for that phenotype. A moderate phenotype and a severe phenotype will be scored 0.5 and 1 respectively.

Y maze test

The Y maze task was conducted as previously described (46). Briefly, during the training phase, the mice were placed into the start arm facing the wall and were allowed to explore the start and trained arm for 5 min, while the entry to the 3rd arm (novel arm) was blocked. The maze was cleaned between each mouse to remove odour cues, and the trained arm was alternated between mice. After training, the mouse was returned to its home cage. After 45 min, the mouse was returned to the start arm and was allowed to explore all three arms for 5 min. The number of entries and the time spent in each arm was quantified using the Smart Video Tracking Software (Panlab; Harvard Apparatus). The percentage of entries in each arm was defined as the number of entries in each arm divided by the total number of entries in all arms during the first minute of the task. The discrimination index was quantified by (novel arm - trained arm)/(novel arm + trained arm). Mice that did not perform three entries during the first minute of testing were excluded.

Novel object recognition (NOR) test

Mice were presented with two identical objects during the first session, and then one of the two objects was replaced by a novel object during a second session. On day 1, a habituation phase in an empty arena (for 5 min), was followed 24 h later by the training phase, which allows for a 5 min exploration in the habituated arena in which two identical objects are placed in opposite quadrants. The testing phase followed a gap of 20 min. For testing, one object was replaced with a novel object followed by 5 min of exploration. Data were collected using the ANY-maze Video Tracking System v.3.3 (Stoelting).

Open field test (OFT)

The OFT measures hyperactivity through locomotion and anxious behavior. The open field box consisted of a square black box (40 cm x 40 cm x 40 cm). Each animal was placed in the box for ten min. The amount of time and distance traveled in the box (measured with videotrack) was measured.

Treadmill test

Mice were acclimated to a motorized treadmill for 3 days prior to testing. During the exercise test, mice ran at an initial speed of 8 m/min with incremental increases of 2 m/min every 2 min at a 5° incline until exhaustion. Exhaustion was defined as the inability to continue running despite repeated mechanical stimulation. Running distance and time were recorded.

Magnetic-activated cell sorting (MACS) of CD31⁺ endothelial cells

Cells were centrifuged at 400 g for 5 min, and the pellet was resuspended in 80 μL of buffer (PBS with 0.5% BSA and 2 mM EDTA), followed by addition of 20 μL of CD45 microbeads. After a 15-minute incubation at 4°C in the dark, cells were centrifuged at 300 g for 10 min. The cell suspension was loaded onto an LS column pre-rinsed with 1 mL of buffer. The CD45-negative flow-through was collected, and the column was washed with 2 mL of buffer. The CD45-negative fraction was centrifuged at 400 g for 5 min, resuspended in 90 μL of buffer with 10 μL of CD31 microbeads, and incubated for 15 min at 4°C in the dark. After centrifugation at 300 g for 10 min, cells were resuspended in 1 mL of buffer and applied to a fresh LS column. The column was washed with 2 mL of buffer, then removed from the magnet. CD31-positive endothelial cells were eluted by flushing with 4 mL of buffer using the plunger. The eluted cells were centrifuged at 400 g for 5 min and resuspended in RPMI-1640 for downstream applications.

Single-cell RNA sequencing

The scRNA-seq was performed using the 10X Genomics Chromium platform Chi-Biotech (Shenzhen, China). Cell suspensions were prepared by following the 10X Genomics Cell Preparation Guide for washing, counting, and concentrating cells from both abundant and limited cell suspensions. A total of fewer than 100,000 cells were loaded onto Chromium microfluidic chips with 3′ v3.1 chemistry and barcoded using the 10X Chromium Controller. RNA from the barcoded cells was reverse-transcribed, and sequencing libraries were constructed using the Chromium Single Cell 3′ v3.1 reagent kit (10X Genomics). The sequencing was carried out on the Illumina NovaSeq 6000 platform according to the manufacturer’s instructions. For data analysis, FASTQ files were processed with Cell Ranger, performing alignment, filtering, barcode counting, and unique molecular identifier (UMI) counting. The Cell Ranger count pipeline generates a feature-barcode matrix, which is used for downstream analysis, including dimensionality reduction, clustering, and gene expression analysis. Secondary filtering was conducted using Seurat, removing genes expressed in fewer than three cells and ensuring each cell had at least 200 expressed genes. Cell Ranger aggr was employed to normalize data from multiple sequencing runs and combine feature-barcode matrices from different GEM wells into a unified dataset. Following secondary analysis using Seurat, gene expression analysis was performed to identify marker genes, with differential expression analysis between samples carried out using the edgeR package. Enrichment analysis of marker genes was performed using the clusterProfiler R package for Gene Ontology (GO) terms and the KEGG database. Reactome pathway analysis and Protein-Protein Interaction (PPI) network analysis were implemented using the ReactomePA and STRINGdb R packages, respectively.

Hematopoietic cells isolation

Mouse tibia and femur were isolated under sterile conditions and placed in a sterile cell culture dish. The muscle and connective tissue were carefully removed from the bones using forceps and small scissors. Mouse bone marrow cells were isolated by removing tibia and femur and flushing the bone marrow with 5–7 mL RPMI complete medium. The flushing process was repeated until the bones turned white, indicating all marrow had been flushed out. The bone marrow was filtered through a 70 μm cell strainer and centrifuged at 600 g for 5 min at 4°C. Red blood cell lysis was performed with red blood cell lysis buffer, followed by washing with 1% BSA in 1× PBS. The cells were centrifuged again at 600 g for 5 min at 4°C, and the supernatant was discarded. For cell surface staining, cells were resuspended in 100 μL 1% BSA in 1× PBS, incubated with fluorescence-conjugated antibodies at 2–8°C for 45 min, washed twice with 1% BSA in 1× PBS, and then resuspended in 500 μL 1% BSA in 1× PBS. Flow cytometry was performed with gating on Lin-Sca-1 + c-Kit+CD135-CD34- to sort the hematopoietic stem cells. Data were analyzed accordingly.

Neurovascular coupling (NVC) measurement

NVC was assessed to evaluate the functional integrity of the cerebrovascular response to neuronal activity. Briefly, mice were anesthetized and placed on a stereotaxic frame while maintaining body temperature. A cranial window was prepared over the somatosensory cortex. CBF responses were measured using LSCI during contralateral whisker stimulation. Whiskers were stimulated at a frequency of 5–10 Hz for defined intervals, and relative changes in CBF were recorded and analyzed. The magnitude and kinetics of the CBF response were used as indicators of NVC efficiency. Reduced CBF responses were interpreted as impaired neurovascular coupling.

Blood-brain barrier (BBB) permeability assessment

BBB integrity was evaluated using Evans Blue dye extravasation. Evans Blue (2% in saline) was administered via intravenous injection and allowed to circulate for 2 hours). Mice were then transcardially perfused with PBS to remove intravascular dye. Brain tissues were harvested, weighed, and homogenized in formamide, followed by incubation at 55–60°C to extract the dye. After centrifugation, the supernatant was collected, and Evans Blue concentration was quantified by measuring absorbance at 620 nm using a spectrophotometer. Evans blue content was calculated from a standard curve and expressed as micrograms of dye per gram of brain tissue (μg/g). Increased dye extravasation indicated compromised BBB integrity.

Ach-induced vasodilation assay

Endothelium-dependent vasodilation in skeletal muscle was evaluated using ACh-induced blood flow responses measured by LSCI. Briefly, mice were anesthetized and the skeletal muscle was surgically exposed and maintained under physiological conditions with continuous superfusion of warm saline. A laser speckle imaging system was positioned above the muscle to record baseline vascular blood flow. Following baseline acquisition, ACh was locally applied to the muscle surface at defined concentrations. Changes in perfusion were continuously recorded, and relative blood flow was quantified as a percentage change from baseline levels. Regions of interest (ROIs) were selected to analyze perfusion signals. Attenuated increases in blood flow in response to ACh were interpreted as impaired endothelium-dependent vasodilation.

Virus production and delivery

rAAV-Tie 2-hNAMPT, rAAV-Tie 2-hNAMPT N136A virus were purchased from the BrainVTA company. Tie 2-C/EBPβ mice were injected with rAAV-Tie 2-vectors, rAAV-Tie 2-hNAMPT and rAAV-Tie 2-hNAMPT N136A virus in a volume of 100 μL via tail veins. The titer of the virus was 2 × 10¹³ genomes per ml. LV-C/EBPβ and LV-shAEP virus were purchase from OBio company.

In vitro NAMPT cleavage assay

To assess the cleavage of NAMPT by AEP in vitro, HEK293 cells were transfected with GST-NAMPT plasmids by PEI. Forty-eight hours after transfection, the cells were collected, washed once in PBS, lysed in lysis buffer (50 mM sodium citrate, 5 mM dithiothreitol (DTT), 0.1% CHAPS and 0.5% Triton X-100, pH 7.4), and centrifuged for 10 min at 14,000 g at 4°C. To measure the cleavage of purified NAMPT fragments by AEP, GST- NAMPT were purified with glutathione beads. The purified NAMPT proteins were incubated with recombinant AEP (5 mg/mL) in AEP buffer (50 mM sodium citrate, 5 mM DTT, 0.1% CHAPS and 0.5% Triton X-100, pH 6.0) for 5 to 15 min. The samples were then boiled in SDS loading buffer and analyzed by immunoblotting.

Generation of AEP-derived NAMPT fragment antibodies

Generation of antibodies that specifically recognize the AEP-generated NAMPT fragments. The anti-NAMPT C137 antibodies were generated by immunizing rabbits with the peptide TDPECYWLTNWIETI. The antiserum was pooled and the titers against the immunizing peptide were determined by ELISA. The maximal dilution giving a positive response with the chromogenic substrate for horseradish peroxidase was 1:50000. The immunoreactivity of the antiserum was further confirmed by Western blotting and immunohistochemistry.

Mass spectrometry analysis

Protein samples were in-gel digested with trypsin. Peptide samples were resuspended in loading buffer (0.1% formic acid, 0.03% trifluoroacetic acid and 1% acetonitrile) and loaded onto a 20-cm nano-high- performance liquid chromatography column (internal diameter 100 mm) packed with Reprosil-Pur 120 C18-AQ 1.9 mm beads and eluted over a 2 h 4–80% buffer B reverse phase gradient (buffer A: 0.1% formic acid and 1% acetonitrile in water; buffer B: 0.1% formic acid in acetonitrile) generated by a NanoAcquity UPLC system (Waters Corporation). Peptides were ionized with 2.0 kV electrospray ionization voltage from a nano-ESI source (Thermo) on a hybrid LTQ XL Orbitrap mass spectrometer (Thermo). Data-dependent acquisition of MS spectra at 120,000 resolution (full width at half maximum) and tandem mass spectrometry (MS/MS) spectra were obtained in the Orbitrap after electron-transfer dissociation with supplemental activation with high energy (EThcD) for peptide masses. To identify AEP-cleavage sites in human NAMPT, Proteome Discoverer 2.0 (PD) was used to search and match MS/MS spectra to a complete human proteome database (NCBI reference sequence revision 62, with 68,746 entries) with a ± 10-ppm mass accuracy threshold and allowable cleavages at glutamates and asparagines. A percolator was used to filter the peptide spectral matches to a false discovery rate of <1%. All MS/MS spectra for putative AEP-generated NAMPT cleavage sites were manually inspected.

Genotyping PCR analysis

Small piece of tail or tissue obtained from Tie 2-AEP mice, Tie 2-C/EBPβ, Tie 2-C/EBPβ/AEP^−/− and WT was incubated in 50 μL of alkaline lysis reagent (25 mM NaOH, 0.2 mM EDTA, pH 12) and incubated at 100°C for 1 hr. After cooling, 50 μL of neutralizing reagent (40 mM Tris-HCl, pH 5) was added and 2 μL of the supernatant was used for PCR to detect the excision of C/EBPβ or LGMN gene using the primers.

RNA analysis

Total mRNA was isolated from cells and tissues using TRIzol. cDNAs were synthesized from 1 mg of total RNA using iScript Reverse Transcription Supermix. qPCR was performed with LightCycler 48 SYBR Green I Mastermix using the LightCycler 480 System according to the manufacturer’s instructions. Relative mRNA expression levels were calculated using the ΔΔC_t method. Quantification of relative mRNA expression was normalized to the expression of Gapdh.

Western immunoblotting

Cells and brain tissue were washed with ice-cold PBS and lysed in 50 mM Tris-HCl, pH 7.4, 40 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, 1.5 mM Na₃VO₄, 50 mM NaF, 10 mM sodium pyrophosphate and 10 mM sodium β-glycerophosphate, supplemented with protease inhibitor cocktail at 4°C for 0.5 h, and centrifuged for 25 min at 15,000 rpm. The supernatant was boiled in SDS loading buffer. After SDS-PAGE, the samples were transferred to a nitrocellulose membrane. The membrane was blocked with TBS containing 5% non-fat milk and 0.1% Tween 20 (TBST) at room temperature for 2 h, followed by incubation with primary antibodies at 4°C overnight, and with secondary antibodies at room temperature for 2 h. Details of the antibodies are provided in table S7. After washing with TBST, the membrane was developed using the enhanced chemiluminescent detection system.

Immunofluorescence

We used free-floating 12 μm sections in immunostaining. For immunofluorescence staining, the sections were incubated overnight at 4°C with primary antibodies. Details of the antibodies are provided in table S7. After washing with PBST, the sections were incubated with a mixture of Alexa Fluor 488-, 555- and 647-coupled secondary antibodies for detection. DAPI (1 μg/mL) was used for staining nuclei. Lastly, coverslips were mounted on glass slides and imaged using a confocal microscope (LSM 980, Zeiss).

SA-β-gal assay

SA-β-gal assay was performed as previously described (47). Both cells and tissue explants were washed twice with PBS, then fixed for 5 min with 2% formaldehyde and 0.2% glutaraldehyde in PBS at room temperature, washed with PBS and then incubated overnight at 37°C in staining solution with 40 mM citric acid NA phosphate, 5 mM K₄Fe(CN)₆, 5 mM K₃Fe(CN)₆ (Fluka analytical, 34272), 150 mM NaCl, 2 mM MgCl₂ and 1 mg/ml X-Gal (Roche, R0404) in water. The samples were washed twice with PBS before imaging by microscopy (Zeiss Axio Vert.A1).

Assessing AEP activity in vivo (LE28 signal detection)

The Cy5-tagged legumain activity-based probe LE28 (2 mg/kg) was injected into mice via tail vein. Mice were anesthetized with isoflurane and placed in the chamber of IVIS 100 system at 4 h after injection. AEP activity was monitored by imaging Cy5 fluorescence produced by LE28.

AEP activity assay

Tissue homogenates or cell lysates (10 μg) were incubated in 200 μl assay buffer (20 mM citric acid, 60 mM Na₂HPO₄, 1 mM EDTA, 0.1% CHAPS and 1 mM dithiothreitol, pH 6.0) containing 20 μM AEP substrate, Z-Ala-Ala-Asn-AMC (Bachem). AMC released by substrate cleavage was quantified by measuring at 460 nm using a fluorescence plate reader at 37°C in kinetic mode.

NAD⁺/NADH assay

Levels of NAD⁺ in HUVECs and muscle homogenates were measured using commercially available kit. Alternatively, NAD⁺ levels in muscle were measured by assay in-house developed method (48). In brief, muscle samples were homogenized in extraction buffer (10 mM Tris-HCl, 0.5% Triton X-100, 10 mM Nicotinamide, pH 7.4) and then centrifuged (12,000 x g for 5 min at 4°C), after which an aliquot of supernatant was taken for protein quantification. After phenol:chloroform:isoamylalcohol (25:24:1) and chloroform extractions the supernatant was separated in two aliquots. One was used to measure total NAD. The other aliquot was acidified with HCl, and then neutralized with NaOH on ice to quantify NAD⁺. Samples were mixed in a 96-well plate samples were mixed with alcohol dehydrogenase at room temperature. Total NADH and NAD⁺ were quantified using a plate reader.

Statistical analysis

The statistical analyses were performed with GraphPad Prism 8.0. software. Unless indicated, statistical significance was calculated by unpaired, two-tailed Student’s t-tests, ordinary ANOVA test for three or more groups (Bonferroni’s multiple comparisons test) and Mann-Whitney U-tests (when a Gaussian distribution was not assumed). For the Kaplan-Meier analysis a log-rank test was performed. Statistical significance was defined as P < 0.05.

Supplementary Materials

This PDF file includes:

Figs. S1 to S12

Tables S1 to S7