Dear Editor,
Metabolic dysfunction-associated steatotic liver disease (MASLD) can progress to metabolic dysfunction-associated steatohepatitis (MASH), hepatic fibrosis, and, ultimately, cirrhosis and/or hepatocellular carcinoma [
1]. During the early stage of MASLD, fatty acid β-oxidation and tricarboxylic acid cycle activity are elevated in the mitochondria of hepatocytes, providing energy and substrates to support
de novo lipogenesis. However, with the progression to MASH, lipotoxicity-induced mitochondrial damage and dysfunction lead to an imbalance between lipid accumulation and disposal, and contribute to inflammation and fibrosis [
2]. Impaired metabolism causes the accumulation of acetyl-CoA, malonyl-CoA, and succinyl-CoA, which act as donors for protein acylation. We observed elevated lysine malonylation and succinylation in type 2 diabetes mellitus (T2DM) [
3]; however, functional studies of lysine acylation in metabolic diseases have been lagging behind.
This study aimed to explore the potential role of protein acylation in MASLD. We began by analyzing lysine malonylation and succinylation in liver tissues of mice fed with a high-fat diet (HFD), a well-established murine dietary model for MASLD [
4]. Using a tandem mass spectrometry workflow that involves the enrichment of malonylated and succinylated peptides (Fig. 1A), we identified and quantified 1123 unique malonylated peptides corresponding to 548 proteins and 3420 unique succinylated peptides mapping to 934 proteins. Compared with the normal diet (ND) group, the HFD group exhibited a significant increase in the levels of 72 malonylated and 271 succinylated peptides (≥2-fold change,
P <
0.05). Gene ontology (GO) analysis revealed that proteins showing elevated malonylation and succinylation were predominantly involved in metabolic pathways (Fig. 1B and 1C).
Since sirtuins orchestrate the elimination of protein acylation, we analyzed their expression profiles in patients diagnosed with MASH. Results shown in Fig. 1D revealed a notable decline in the expression of SIRT5 among MASH patients, and its expression was negatively correlated with the severity of steatohepatitis (Fig. S1A). Single-cell RNA sequencing (scRNA-seq) of livers from mice fed with a Western diet revealed that Sirt5 was specifically downregulated in hepatocytes (Fig. S1B). Consistently, both the Sirt5 mRNA (Fig. 1E) and the SIRT5 protein (Figs.1F and S1C) levels were decreased in the liver of HFD mice, but not in other organs (Fig. S1D). In mice fed with methionine/choline-deficient diet (MCD), hepatic SIRT5 was also reduced (Fig. 1G).
To elucidate the mechanisms that control SIRT5 expression in MASLD, we profiled the protein–protein interaction partners of SIRT5. Huh7 and HepG2 cells were transfected with Flag-tagged
SIRT5, followed by immunoprecipitation and LC–MS/MS shotgun proteomics. The E3 ubiquitin ligase tripartite motif-containing protein 21 (TRIM21) was identified as the top interactor of SIRT5 (Fig. 1H). Co-IP validated the interaction between SIRT5 and TRIM21 (Figs. 1I and 1J, and S1E). Given that TRIM21 mediates the degradation of SIRT5 [
5], we hypothesized that it might control the down-regulation of SIRT5 in MASLD.
TRIM21 expression was significantly elevated in liver specimens from MASLD patients (Fig. S1F). Consistently, TRIM21 levels were profoundly increased in the livers of HFD mice compared to ND controls (Fig. 1K). Overexpression of TRIM21 decreased SIRT5 levels in both Huh7 and HepG2 cells (Fig. 1L). To explore the molecular mechanisms underlying TRIM21 induction in MASLD, we treated cells with TNF-α and free fatty acids (FFA), factors that have been linked to MASLD progression. FFA efficiently induced lipid accumulation in hepatic cells (Fig. 1M) and, in combination with TNF-α, synergistically upregulated TRIM21 expression while further suppressing SIRT5 protein levels (Figs. 1N and S1G).
SIRT5 exhibits robust demalonylase and desuccinylase activities. Given the observed down-regulation of SIRT5 expression levels and elevated lysine malonylation and succinylation in MASLD, we hypothesized that the restoration of SIRT5 expression would mitigate the dysregulated lipid and glucose metabolism in MASLD. To test this hypothesis, we developed mice with liver-specific overexpression of SIRT5 (referred to as Liver Sirt5 OE mice) (Fig. 2A), and induced MASLD with HFD. Since patients with MASLD have a higher prevalence of T2DM, we performed glucose tolerance tests and insulin tolerance tests (ITT). In HFD-fed Liver Sirt5 OE mice, basal blood glucose levels were reduced compared with HFD-fed wild-type (WT) mice, and glucose tolerance was improved (Fig. 2B). ITT experiments indicated that Liver Sirt5 OE mice exhibited enhanced insulin sensitivity (Fig. 2C). In terms of lipid metabolism, Liver Sirt5 OE mice fed with HFD showed significantly lower levels of plasma triglycerides (TG), total cholesterol (TC) and low-density lipoprotein (LDL), compared with WT mice fed with HFD (Fig. 2D–F). Histological analysis confirmed reduced hepatic steatosis in Liver Sirt5 OE mice (Fig. 2G and 2H). Consistent with this histological improvement, the HFD‑induced increase in liver weight was significantly attenuated in Liver Sirt5 OE mice compared with WT mice (Fig. 2I). The activities of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were also decreased in Liver Sirt5 OE mice (Fig. 2J and 2K). The ultrastructure of mitochondria from Liver Sirt5 OE mice remained clear, with reduced lipid droplet accumulation (Fig. 2L). Concomitantly, we found reduced liver TG in Liver Sirt5 OE mice in response to HFD (Fig. 2M). In vitro experiments also demonstrated that SIRT5 prevents excessive lipid deposition caused by FFA stimulation in primary hepatocytes (Fig. 2N).
To examine whether mitochondrial respiration was affected by SIRT5, we measured the oxygen consumption rate (OCR) of primary hepatocytes from Liver Sirt5 OE and WT mice, using glucose as substrate. As indicated in Fig. 2O, SIRT5 elevated the maximal OCR capacity, suggesting that it enhances mitochondrial respiratory function. SIRT5 also increased the capacity for fatty acid β-oxidation (Fig. 2P), suggesting that SIRT5 promotes glucose and lipid metabolism by boosting mitochondrial bioenergetic capacity.
We conducted proteomic analysis to compare the levels of malonylated and succinylated peptides in Liver Sirt5 OE and WT mice fed either ND or HFD. Malonylation increased in 404 out of 1137 peptides (WT/Liver Sirt5 OE ratio ≥ 2, P < 0.05), defining Sirt5-regulated demalonylation targets. Succinylation increased in 772 out of 3872 peptides under the same thresholds, defining Sirt5-regulated desuccinylation targets. GO enrichment showed these proteins are enriched in glucose and lipid metabolism, indicating that Liver Sirt5 OE drives demalonylation and desuccinylation of metabolic proteins (Fig. 2Q and 2R). Notably, we found that 35 malonylated peptides in WT hepatocytes due to HFD were demalonylated in SIRT5 high-expressing hepatocytes, and 52 peptides showed a similar trend for succinylation. Pathway enrichment analysis showed these overlapped proteins were mainly enriched in metabolism pathways, such as fatty acid β-oxidation, and tricarboxylic acid cycle (Fig. S2A and S2B).
Consistent with a previous report [
6], we found multiple succinylation sites on hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunit α (HADHA), a rate-limiting enzyme involved in the fatty acid oxidation pathway, were regulated by SIRT5 (Fig. S2C and S2D). In addition to HADHA, citrate synthase (CS), a rate-limiting enzyme of the tricarboxylic acid cycle, was also succinylated and regulated by SIRT5 (Fig. S2E). These results support our hypothesis that HFD may induce post-translational modifications (PTM) on key metabolic proteins and lead to dysfunction, while hepatic SIRT5 may modify such PTMs and restore the function of these proteins.
Using immunoprecipitation and LC–MS/MS shotgun proteomics, we detected and validated the interaction between SIRT5 and TANK-binding kinase 1 (TBK1) (Fig. 2S). TBK1 has recently been implicated as a potential link between hepatic inflammation and energy metabolism [
7]. Although no malonylated or succinylated TBK1 was detected in liver tissue of mice fed with HFD or ND diet (Fig. S2A and S2B), we predicted that TBK1 is a potential substrate of lysine malonylation and succinylation with a toolkit we developed [
8]. SIRT5 has been reported as a modulator of TBK1. SIRT5 desuccinylates TBK1 at K38, K154, and K692 and potentiates the inflammatory response [
9]. On the other hand, SIRT5 desuccinylates TBK1 at K137 and suppresses the downstream inflammatory pathways [
10]. The inconsistency of the role of SIRT5 in TBK1 regulation might be due to the complicated signaling pathways mediated by TBK1. The combination of succinylation at different K residues might interfere with either the protein complex formation or the kinase activity, and thus shows diverse downstream effects. We observed elevated TBK1 phosphorylation in the liver of HFD mice (Fig. 2T), indicating the chronic inflammation in MASLD models. Notably, overexpression of SIRT5 suppressed TBK1 phosphorylation (Figs. 2U and S2F) and downregulated the expression of proinflammatory cytokines (Figs. 2V and S2G ), suggesting that SIRT5 may function as a negative regulator of chronic hepatic inflammation.
Collectively, our study demonstrated a protective role of SIRT5 on the metabolic abnormality during the progression of MASLD (Fig. 2W). FFAs and proinflammatory factors upregulated the E3 ubiquitin ligase TRIM21, which in turn mediates the downregulation of SIRT5. Down-regulated SIRT5 expression led to increased lysine succinylation and malonylation in metabolism-related proteins, resulting in the dysregulation of glucose and lipid metabolism. Hepatic overexpression of SIRT5 eliminated the hyper-succinylation and hyper-malonylation, and restored the metabolic flux. Of note, TBK1, a key mediator linking metabolism and inflammation, is also a potential target of SIRT5. SIRT5 attenuated TBK1-mediated inflammation by fine-tuning the crosstalk between lysine acylation and serine phosphorylation. The results broaden the current understanding of SIRT5 as a key regulator in PTM and metabolic diseases.
Research limitations
The mouse model of HFD and MCD diet-induced MASLD used in this study may not fully recapitulate the complexity of human MASLD. The precise regulatory effects of TBK1 activation by SIRT5 during the progression of MASLD remain to be further elucidated. Additionally, the study did not explore the long-term effects or potential side effects of SIRT5, which are critical considerations for assessing its translational and clinical applicability.
The Author(s) 2026. Published by Oxford University Press on behalf of Higher Education Press.
PDF
(2448KB)
