Restricting intracellular Salmonella proliferation by coordinating p-TBK1 mediated mitophagy and xenophagy

ABSTRACT

Mitophagy is essential for eliminating dysfunctional mitochondria and is closely implicated in the immune evasion of several pathogens, including S. typhimurium. However, the specific mechanisms regarding the interaction between S. typhimurium and host cells in relation to mitophagy and xenophagy and their contribution to pathogen survival are unclear. Herein, using both in vitro and in vivo systems, we found that S. typhimurium escaped host innate immunity by repressing mitophagy and xenophagy to facilitate its intracellular replication. Moreover, we identified a novel xenophagy modulator, fisetin that could activate mitophagy to restrict intracellular S. typhimurium replication in RAW264.7 and bone marrow-derived macrophages, which was abolished by mitophagy inhibitor Mdivi-1. RNA-Seq transcriptome and metabolomics analysis demonstrated the effectiveness of fisetin in alleviating S. typhimurium infection. Confocal microscopy analysis revealed that fisetin-induced mitophagy promoted xenophagy, whereas inhibiting mitophagy repressed xenophagy and facilitated the survival of S. typhimurium. Our study further demonstrates that fisetin-induced mitophagy requires the recruitment of phosphorylation of TBK1 to mitochondria, which is a protein implicated in mitophagy and xenophagy. Additionally, fisetin improved the body weight loss, relative spleen, kidney, and liver weights, hepatic damage, and S. typhimurium load, all of which were abrogated by Mdivi-1 or Pink1 siRNA treatment in S. typhimurium-infected mice. Collectively, our results suggest that S. typhimurium induces mitochondrial damage whilst inhibiting mitophagy, while fisetin promotes xenophagy and restrains S. typhimurium survival by facilitating PINK1-PRKN-mediated mitophagy and p-TBK1 mitochondrial recruitment. Fisetin proves effective as a xenophagy enhancer in reducing intracellular Salmonella burden.

Abbreviations: BafA1: bafilomycin A₁; BMDM: mouse bone marrow-derived macrophage; CFU: colony-forming units; LAMP2: lysosomal-associated membrane protein 2; MAP1LC3/LC3: microtubule associated protein 1 light chain 3; LDH: lactate dehydrogenase; Mdivi-1: mitochondrial division inhibitor 1; OPTN: optineurin; PBS, phosphate-buffered saline; PINK1: PTEN induced putative kinase 1; siRNA: interfering RNA; SQSTM1/p62: sequestosome 1; S. typhimurium: Salmonella enterica serovar typhimurium; T3SS: type III secretion system 1; TBK1: TANK-binding kinase 1

Introduction

Salmonella enterica serovar typhimurium (S. typhimurium) is a foodborne bacterial pathogen that not only infects animals and animal-based foods like meat, eggs, milk, and animal offal, but as a main infective source for human beings [1]. S. typhimurium is also a well-studied intracellular pathogenic bacterium that can invade host cells via its type III secretion system 1 (T3SS1) [2]. The host cells have diverse anti-microbial strategies, among which antibacterial macroautophagy/autophagy (xenophagy) is a vital host defense [3]. Xenophagy, a crucial cell-autonomous mechanism for combating microbial infections, relies on the autophagic process’s ability to selectively entrap intracellular pathogens within autophagosomes and subsequently degrade them in lysosomes [4]. S. typhimurium has evolved several survival mechanisms and prominent strategies, including trafficking in the host, inducing vacuolar membrane damage, hijacking the V-ATPase-ATG16L1 axis that mediates the initiation of xenophagy to avoid innate immune defense mechanisms, therefore unrestricting access to the nutrient-rich cytoplasm, and facilitating replication within cells [3]. Previous studies have highlighted the tactical interplay between diverse pathogens and hosts regarding xenophagy [5,6]. Therefore, investigating the immune evasion and underlying mechanisms will contribute to the pathogenesis and development of therapeutic treatments for Salmonella infection.

Mitochondria play an important role in innate immune cell fate and function, and disruption of autophagy in these cells leads to aggregation of dysfunctional mitochondria and reduced survival [7]. Mitophagy is a selective autophagy and targets aging and dysfunctional mitochondria to maintain the quality and function of organelles and cellular homeostasis by eliminating defective mitochondria within the cell [8]. Increasing studies have demonstrated that several pathogens are able to employ mitophagy to weaken host innate immune response and facilitate continuous infection [9]. Listeria hijacks host mitophagy via mitophagy receptor NLRX1 to facilitate the interaction between microtubule-associated protein 1 light chain 3 (LC3) interacting region (LIR) and LC3, thus evading killing [10]. SARS-CoV-2 virus induces IL-18-mediated cardiopulmonary inflammation via reducing mitophagy, therefore contributing to virus invasion [11]. It is known that PTEN-induced kinase 1 (PINK1)-PRKN-mediated mitophagy is important for selectively removing damaged mitochondria [12]. Consistently, PRKN-deficient mice and flies are more susceptible to various intracellular bacterial (Mycobacterium leprae and S. typhimurium) infections, supporting that PRKN is a critical factor implicated in xenophagy to enhance host innate defense [13]. Notably. Xu et al. [3] revealed that approximately 20% of S. typhimurium were decorated with LC3 within 40 min and subsequently reduced to a basal level, suggesting the activation of autophagy by S. typhimurium is transient and S. typhimurium avoids host xenophagy-mediated capture. However, it is still unclear whether S. typhimurium inhibits host cell mitophagy, and the crosstalk between mitophagy and xenophagy and the underlying mechanisms are required for a better insight.

Host response in capturing cytosolic bacteria entails tagging them with ubiquitin [14]. The ubiquitinated bacteria are recognized by xenophagy receptors, such as SQSTM1/p62 (sequestosome 1), OPTN (optineurin), NBR1 (NBR1 autophagy cargo receptor), and CALCOCO2/NDP52 (calcium binding and coiled-coil domain 2), which subsequently bind to LC3 and initiate autophagy [15]. Of note, SQSTM1, OPTN, and CALCOCO2 are also closely implicated in mitophagy to clear defective mitochondria [16]. Previous studies revealed that these ubiquitin proteins are phosphorylated by TBK1 (TANK-binding kinase 1) and targeted to ubiquitin-tagged mitochondria or bacteria, contributing to the formation of autophagosomes during mitophagy or xenophagy [17,18]. TBK1 activation is required for the recruitment of autophagy receptor proteins CALCOCO2 and OPTN to induce mitophagy, and functions as a signal amplification during mitophagy to remove defective mitochondria [19]. Recent advance reported that OPTN provided a mitophagy niche for activating TBK1 [20]. Additionally, Song et al. [21] demonstrated that Mycobacterium bovis employed mitophagy to facilitate its proliferation within macrophages by competitively utilizing p-TBK1 to repress xenophagy. These studies suggest that TBK1 plays a critical role in xenophagy and mitophagy. Notably, unlike Mycobacterium bovis, S. typhimurium, characterized by its T3SS effectors, shares different molecular mechanisms regarding the role of TBK1 in moderating xenophagy and mitophagy, which is still unclear in response to S. typhimurium infection.

Previous studies have indicated that invasive pathogens employ mitophagy to weaken host innate immune response and facilitate continuous infection [9,10,21]. A recent study revealed that S. typhimurium prevents mitochondrial fragmentation to facilitate its intracellular proliferation, which concomitantly leads to an increase in mitochondrial reactive oxygen species, mitochondrial membrane potential depolarization, and the release of mitochondrial DNA and CYCS/cytochrome c to the cytosol [22]. These data have prompted us to investigate the following questions: 1) Does S. typhimurium inhibit mitophagy given that it inhibits xenophagy in response to pathogen invasion? 2) If it does, what role does mitophagy play in the process of xenophagy and survival of S. typhimurium? 3) What are the mechanisms by which mitophagy affects the xenophagy and proliferation of S. typhimurium? Figuring out these questions will provide new insights into the intracellular survival mechanism and treatment methods for S. typhimurium infection.

In the present study, S. typhimurium induces a loss of mitochondrial membrane potential and inhibits mitophagy to facilitate its intracellular replication. We identified and characterized a xenophagy inducer, fisetin, which restricts the intracellular survival of S. typhimurium in vitro and in vivo. Fisetin activates the PINK1-PRKN signaling pathway and recruits p-TBK1 to mitochondria, thereby promoting mitophagy, which further activates xenophagy to restrict intracellular survival of S. typhimurium. During infection, fisetin-mediated constraint of S. typhimurium is also validated by cellular RNA-Seq transcriptome analysis and LC-MS non-targeted metabolomics in a mouse infection model by using inhibitor and small interfering RNA (siRNA) transfection.

Results

Fisetin activates autophagy flux in macrophages

To investigate the intracellular infection model and time, RAW264.7 cells were infected with S. typhimurium for 30 min to allow them to enter cells, and cells were rinsed with gentamycin-containing phosphate-buffered saline (PBS) to exclude the extracellular bacteria and subsequently cultured with gentamycin-containing DMEM medium for the indicated time points (Figure 1A). The cell morphology by confocal microscope revealed that the S. typhimurium, as indicated by the red arrow, was released from infected cells and increased in the cell culture medium as the time of infection increased. Moreover, the level of damaged cells was increased, and the cell membrane was severely disrupted at 8 h post-infection. We next determined the mitochondrial membrane potential by a flow cytometer and found that S. typhimurium treatment resulted in a dramatic loss of mitochondrial membrane potential, indicating the occurrence of mitochondrial damage (Figure 1B,C). To further investigate whether the defective mitochondria can be eliminated, we isolated mitochondria in S. typhimurium-treated RAW264.7 cells and examined the autophagic LC3B protein level in mitochondria (Figure 1D). The mitochondrial purity was validated to exclude the possibility that the mitochondrial fraction might be contaminated by the cytosolic portion via western blot. As shown, the cytosolic part (α-tubulin protein) was not detected in the mitochondrial fraction of RAW264.7 cells. Notably, S. typhimurium reduced mitochondrial LC3B protein levels at 2, 4, and 8 h post-infection when compared with uninfected cells, indicating S. typhimurium is likely to inhibit mitophagy. We further validated this result by performing mito-Keima assay to monitor mitophagy (Figure 1E,F), and found that S. typhimurium dramatically reduced mitophagy. We, therefore, hypothesized that autophagic and mitophagy inducers may contribute to the mitigation of cellular homeostasis. We identified an autophagic and mitophagy inducer, fisetin, and first determined the LC3B protein abundance in uninfected RAW264.7 cells. The result demonstrated that fisetin could increase the LC3B protein level and autophagosome levels, as evidenced by the upregulated LC3B concentration (Figure 1G) and transmission electron microscope (TEM) images (Figure 1H, I). LDH (lactate dehydrogenase) release is a biomarker that reflects cell damage, and we found that fisetin could ameliorate S. typhimurium-induced an increase in LDH release (Figure 1J), suggesting the regulative effects of fisetin in S. typhimurium-treated macrophages. Despite LC3B being the most common indicator to reflect autophagic modulation, the LC3B accumulation may result from autophagy activation or a block in the fusion of autophagosome and lysosome. We next assayed the autophagic flux by transfecting macrophages with adenoviruses to express mCherry-GFP-LC3. As shown in Figure 1K, fisetin activated autophagic flux in S. typhimurium-treated RAW264.7 macrophages. To further validate that fisetin could activate autophagy, cells were incubated with or without 100 nM bafilomycin A1 (BafA1), an inhibitor that blocks the fusion of autophagosome and lysosome. The results (Figure 1L) showed that the co-treatment (BafA1, fisetin, and S. typhimurium) further increased the LC3B protein level as compared with BafA1 and S. typhimurium treatment, demonstrating that fisetin could induce autophagy. Together, these data indicate that S. typhimurium treatment results in mitochondrial damage, but inhibits mitophagy, while fisetin can induce autophagy and show potential in regulating the survival of S. typhimurium.

Figure 1.

Fisetin activates autophagy flux in RAW264.7 macrophages.

(A) RAW264.7 cells were infected with S. typhimurium (MOI = 100) for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing phosphate buffer saline (PBS) to remove extracellular S. typhimurium, and post cultured for 2, 4, and 8 h. The presentative confocal microscopy of S. typhimurium-infected cells. Red arrow, S. typhimurium; white arrow, damaged cell. Scale bar: 10 μm.

(B-C) Flow cytometry analysis of mitochondrial membrane potential loss and its statistical analysis.

(D) Western blot for LC3B protein in mitochondrial and cytosolic fractions in S. typhimurium-infected macrophages.

(E-F) Mito-Keima assay and its statistical analysis.

(G) Western blot for LC3B protein in fisetin (20 μM)-treated RAW264.7 cells and statistical analysis.

(H-I) Transmission electron microscope (TEM) for fisetin (20 μM)-treated RAW264.7 cells and its statistical analysis. Scale bar: 2 μm.

(J) RAW264.7 cells were pretreated with fisetin (20 μM) for 3 h, co-treated with S. typhimurium (MOI = 100) for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. The lactate dehydrogenase (LDH) level of cellular supernatant.

(K) Representative confocal microscopy images for mCherry-GFP-LC3 transfected RAW264.7. Scale bar: 5 μm.

(L) RAW264.7 cells were pretreated with fisetin (Fis, 20 μM) or bafilomycin A1 (BafA1, 100 nM), or Fis + BafA1 for 3 h, co-treated with S. typhimurium (MOI = 100) for 30 min, and post cultured for 4 h. Western blot for LC3B protein in BafA1-treated RAW264.

Fisetin activates mitophagy

To further investigate whether fisetin could induce mitophagy, we isolated and validated the purity of the mitochondria from RAW264.7 cells (Figure 2A). The cytosolic fraction was not found in the mitochondrial part, and fisetin increased the LC3B protein level in the St+Fis group as compared with the St group, indicating that fisetin promotes the accumulation of LC3B. The PINK1-PRKN signaling pathway is responsible for eliminating the depolarized mitochondria, and we found that fisetin promoted the PINK1 and PRKN protein level (Figure 2B,C). We next assayed the formation of mitophagosomes via double staining the mitochondria by MitoTracker red and LC3B (Figure 2D,E), and found that fisetin ameliorated the reduced colocalization of mitochondria and LC3B by S. typhimurium treatment. Consistently, S. typhimurium inhibited the mitophagy, while fisetin promoted mitophagy in S. typhimurium-infected cells, as evidenced by the immunofluorescence double staining of LC3B and voltage-dependent anion channel 1 (VDAC1), an important protein for PINK1-PRKN-mediated mitophagy (Figure 2F,G). Likewise, the mito-Keima assay also demonstrated that the inhibition of mitophagy by S. typhimurium treatment was abrogated by fisetin treatment (Figure 2H,I). The defective mitochondria were finally eliminated by fusion with the lysosome. As shown in Figure 2J,K, S. typhimurium compromised the fusion of mitochondria and lysosomes, which was alleviated by fisetin. Additionally, we found that fisetin promoted the interaction of TOMM20, a mitochondrial marker protein, with LC3B in S. typhimurium-treated cells by co-IP (Figure 2L). To further investigate how fisetin is regulating the recruitment of autophagy receptors to mitochondria in infection conditions, we performed immunofluorescence double staining (Figure 3A–F). We found that fisetin promoted the recruitment of CALCOCO2 and SQSTM1 to mitochondria, without markedly affecting the interaction of OPTN with mitochondria. Additionally, fisetin also facilitated the mitochondrial clearance by recruiting ubiquitin (Figure 3G,H). Together, these results demonstrate that fisetin alleviated the mitophagy by S. typhimurium treatment.

Figure 2.

Fisetin activates mitophagy.

(A) RAW264.7 cells were pretreated with fisetin (Fis, 20 μM) for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. Western blot for LC3B protein in mitochondrial and cytosolic fractions in fisetin-treated cells.

(B-C) Western blot for PINK1 and PRKN proteins and their statistically analysis.

(D) Confocal microscopy analyzed the colocalization of mitochondria (red) with LC3B (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(E) Quantification of Pearson’s colocalization coefficient between mitochondria and LC3B.

(F) Confocal microscopy analyzed the colocalization of mitochondrial protein VDAC1 (red) with LC3B (green) in S. typhimurium (100 MOI) -infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(G) Quantification of Pearson’s colocalization coefficient between VDAC1 and LC3B.

(H) Mito-Keima assay analyzed the mitophagy in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(I) The statistical analysis of mito-Keima assay (550 nm/440 nm).

(J) Confocal microscopy analyzed the colocalization of mitochondria with lysosome (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(K) Quantification of Pearson’s colocalization coefficient between mitochondria and lysosome. (L) Coimmunoprecipitation between mitochondrial protein TOMM20 and LC3B in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection.

(L) Coimmunoprecipitation between mitochondrial protein TOMM20 and LC3B in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection.

Figure 3.

Fisetin facilitates the recruitment of autophagy receptors and ubiquitin to mitochondria. RAW264.7 cells were pretreated with fisetin (20 μM), or fisetin + mitophagy inhibitor Mdivi-1 (20 μM) for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and cultured for another 4 h.

(A) Confocal microscopy analyzed the colocalization of mitochondria (red) with CALCOCO2 (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(B) Quantification of Pearson’s colocalization coefficient between mitochondria and CALCOCO2.

(C) Confocal microscopy analyzed the colocalization of mitochondria (red) with SQSTM1 (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(D) Quantification of Pearson’s colocalization coefficient between mitochondria and SQSTM1.

(E) Confocal microscopy analyzed the colocalization of mitochondria (red) with OPTN (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(F) Quantification of Pearson’s colocalization coefficient between mitochondria and OPTN.

(G) Confocal microscopy analyzed the colocalization of mitochondria (red) with ubiquitin (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(H) Quantification of Pearson’s colocalization coefficient between mitochondria and ubiquitin.

Induction of mitophagy by fisetin restrains intracellular replication of S. typhimurium

We first explored the intracellular S. typhimurium load by TEM (Figure 4A) and confocal microscopy (Figure 4B), and found that fisetin reduced the proliferation of intracellular S. typhimurium. To investigate the effects of mitophagy on the intracellular survival of S. typhimurium, RAW264.7 cells were incubated with mitochondrial division inhibitor 1 (Mdivi-1, 20 μM) to inhibit mitophagy (Figure 4C). Compared with the S. typhimurium group, fisetin significantly reduced the intracellular bacterial load of S. typhimurium, which was abrogated by Mdivi-1 treatment, indicating the inactivation of mitophagy facilitated the survival of intracellular S. typhimurium. Next, RAW264.7 cells were transiently transfected with Pink1 siRNA (0.25 nM) to inhibit mitophagy (Figure 4D). As expected, compared with the S. typhimurium or negative control siRNA group (0.25 nM), fisetin significantly reduced the intracellular bacterial load, whereas knockdown of PINK1 expression promoted the survival of intracellular S. typhimurium. To further validate this result, we isolated bone marrow- derived macrophages (BMDM) and treated them with or without Mdivi-1 (20 μM) and found that fisetin resulted in a decrease in intracellular S. typhimurium load, which was reversed by Mdivi-1 treatment (Figure 4E). Moreover, the regulative effect of fisetin in lowering intracellular S. typhimurium was abolished by transfecting with Pink1 siRNA (0.25 nM, Figure 4F). These results suggest that fisetin-induced mitophagy contributes to the inhibition of the survival of intracellular S. typhimurium.

Figure 4.

Induction of mitophagy by fisetin restrains intracellular replication of S. typhimurium.

(A) RAW264.7 cells were pretreated with fisetin (20 μM) for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and cultured for another 4 h. Transmission electron microscope (TEM) of intracellular survival of S. typhimurium in RAW264.7.

(B) Confocal microscopy analyzed the colocalization of Dil with Salmonella in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(C) RAW264.7 cells were pretreated with fisetin (20 μM), or mitophagy inhibitor Mdivi-1 (20 μM), or fisetin + Mdivi-1 for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and cultured for another 4 h. The intracellular S. typhimurium burden in Mdivi-1-treated RAW264.7.

(D) siRNA (Con and Pink1 siRNA, 50 pM) treated RAW264.7 cells were pretreated with fisetin (20 μM), co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and cultured for another 4 h. The intracellular S. typhimurium burden in Pink1 siRNA-treated RAW264.7.

(E) BMDM were pretreated with fisetin (20 μM), or mitophagy inhibitor Mdivi-1 (20 μM), or fisetin + Mdivi-1 for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. The intracellular S. typhimurium burden in Mdivi-1-treated BMDM.

(F) siRNA (Con and Pink1 siRNA, 50 pM) treated BMDM were pretreated with fisetin (20 μM), co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. The intracellular S. typhimurium burden in Pink1 siRNA-treated BMDM.

Identification of the alleviative effects of fisetin in S. typhimurium-treated RAW264.7 macrophages by RNA-Seq transcriptome analysis

To further investigate the regulative effects of fisetin supplementation in S. typhimurium-treated RAW264.7 macrophages, we performed RNA-Seq transcriptome analysis of uninfected cells, infected cells, and fisetin-treated infected cells. The heatmap and volcano plot illustrated that the transcriptome profile of S. typhimurium-infected cells was dramatically distinctive with that of uninfected cells, as evidenced by upregulating 1648 differentially expressed genes (DEGs) whereas downregulating 1079 DEGs (Figure 5A,B). Moreover, fisetin supplementation displayed 302 upregulated DEGs and 694 downregulated DEGs as compared with the S. typhimurium-treated cells (Figure 5A,C).

Figure 5.

Identification of the alleviative effects of fisetin in S. typhimurium-treated RAW264.7 macrophages by RNA-Seq transcriptome analysis. RAW264.7 cells were pretreated with fisetin (20 μM), co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h, the treated cells were subjected to RNA-Seq transcriptome analysis.

(A) Heatmap illustration displaying gene expression of Con, St, and St + Fis in macrophages.

(B) Volcano plot showing differentially expressed genes (DEGs; downregulated, blue; upregulated, red) in S. typhimurium-treated cells as compared with the control.

(C) Volcano plot showing DEGs (downregulated, blue; upregulated, red) in fisetin and S. typhimurium co-treated cells as compared with the S. typhimurium-treated cells.

(D) KEGG enrichment analysis of upregulated DEGs in S. typhimurium-treated cells as compared with the control.

(E) KEGG enrichment analysis of downregulated DEGs in S. typhimurium-treated cells as compared with the control.

(F) KEGG enrichment analysis of upregulated DEGs in fisetin and S. typhimurium co-treated cells as compared with the S. typhimurium-treated cells.

(G) KEGG enrichment analysis of downregulated DEGs in fisetin and S. typhimurium co-treated cells as compared with the S. typhimurium-treated cells.

(H) GO enrichment analysis of upregulated DEGs in S. typhimurium-treated cells as compared with the control.

(I) GO enrichment analysis of downregulated DEGs in S. typhimurium-treated cells as compared with the control.

(J) GO enrichment analysis of upregulated DEGs in fisetin and S. typhimurium co-treated cells as compared with the S. typhimurium-treated cells.

(K) GO enrichment analysis of downregulated DEGs in fisetin and S. typhimurium co-treated cells as compared with the S. typhimurium-treated cells.

Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of DEGs demonstrated that genes enriched in human diseases, complement and coagulation cascades, apoptosis, cytokine-cytokine receptor interaction, TNF, NF-kappa B, and JAK-STAT signaling pathway were significantly upregulated in S. typhimurium-infected cells. Moreover, S. typhimurium also markedly reduced cell cycle, biosynthesis of unsaturated fatty acids, DNA replication, base excision repair, and mismatch repair-related genes’ expression (Figure 5D,E). Notably, fisetin supplementation significantly downregulated gene expressions related to human diseases, chemokine signaling pathway, osteoclast differentiation, cytosolic DNA-sensing pathway, NOD-like and TLR (toll like receptor) signaling pathway, intestinal IgA production, cytokine-cytokine receptor interaction, JAK-STAT signaling pathway, with environmental information processing, metabolism, cellular processes, organismal systems, and human diseases being similar to that of S. typhimurium (Figure 5F,G). Likewise, Gene Ontology (GO) analysis revealed that genes related to inflammatory response, immune response, regulation of vesicle size, and Rac protein signal transduction were upregulated, while DNA replication and cell cycle-related genes were inhibited in response to S. typhimurium treatment (Figure 5H,I). Additionally, fisetin supplementation upregulated the expression of genes regarding water homeostasis, whereas reduced gene expressions about eosinophil chemotaxis and migration, cellular response to interferon-alpha and interferon-beta, as well as immune response (Figure 5J,K). The cellular transcriptome analysis suggests that fisetin is able to improve the pathways or biological processes in S. typhimurium-infected RAW264.7 macrophages.

Fisetin ameliorates hepatic metabolic disorder in S. typhimurium-infected mice by LC-MS non-targeted metabolomics

We further verify the regulative effects of fisetin supplementation in S. typhimurium-treated mice (Figure S1A). As shown in Figure S1B, S. typhimurium treatment resulted in a dramatic loss of mouse body weight, which was ameliorated by fisetin administration. Moreover, the relative spleen, kidney, and liver weights were observed to be higher in S. typhimurium-treated mice, which was also alleviated by fisetin administration, suggesting a potential function of fisetin in mitigating S. typhimurium infection (Figure S1C-E). We next assayed the hepatic bacterial load to verify the regulative effects of fisetin on the proliferation of S. typhimurium in vivo (Figure S1F), and found that fisetin efficiently reduced the proliferation of S. typhimurium. We next validated the regulatory effects of fisetin by LC-MS non-targeted metabolomics of liver tissues in mice. The PCA score of the metabolite content identified by the positive and negative ion pattern name revealed that the hepatic metabolite in S. typhimurium-treated mice was distinctive from that of the control, which was ameliorated by fisetin administration (Figure S1G,H). Additionally, S. typhimurium challenge led to increase in uric acid, taurine, uracil, choline, creatine, L-2-aminoadipic acid, uridine, adrenic acid, linoleic acid, arachidonic acid, docosahexaenoic acid, and eicosapentaenoic acid concentrations, and decrease in galacturonic acid, L-glutamic acid, stachyose, maltotriose, and adenylosuccinic acid, which were ameliorated by fisetin administration (Figure S1I,J). Consequently, these observations suggest that fisetin treatment can partially restore the damage by S. typhimurium in both in vivo and in vitro.

Induction of mitophagy by fisetin promotes xenophagy in macrophages

Xenophagy plays a critical role in the immune defense mechanism against intracellular bacteria in macrophages [23]. Herein, we next investigated the effects of mitophagy on xenophagy by using Mdivi-1, a mitophagy inhibitor. We first validated the effects of Mdivi-1 by analyzing the colocalization of mitochondria and LC3B and found that the fisetin-induced increase in the colocalization of mitochondria and LC3B was abrogated by Mdivi-1, indicating that Mdivi-1 can inhibit mitophagy (Figure 6A,B). Likewise, Mdivi-1 abrogated fisetin-induced mitophagy as evidenced by the mito-Keima assay (Figure 6C,D). We next studied xenophagy by observing the colocalization of Salmonella and LC3B (Figure 6E,F), and found that fisetin promoted the mitophagy, while inhibiting mitophagy by Mdivi-1 reduced the level of xenophagy, suggesting that mitophagy is likely to facilitate xenophagy. To further strengthen this observation, we further performed immunofluorescence double staining of Salmonella and lysosome-associated membrane protein 2 (LAMP2, Figure 6G,H), CALCOCO2 (Figure 7A,B), SQSTM1 (Figure 7C,D), and OPTN (Figure 7E,F). Consistently, fisetin treatment promoted the elimination of S. typhimurium as evidenced by the increased colocalization of Salmonella and LAMP2, CALCOCO2, and SQSTM1, which were also abrogated by Mdivi-1. Additionally, we have also analyzed the colocalization of ubiquitin and S. typhimurium (Figure 7G,H), and found that fisetin facilitated the clearance of S. typhimurium, which was also blocked by the inhibition of mitophagy via Mdivi-1. Together, these results suggest that the induction of mitophagy by fisetin promotes xenophagy in RAW264.7 cells.

Figure 6.

Induction of mitophagy by fisetin promotes xenophagy in macrophages. RAW264.7 cells were pretreated with fisetin (20 μM), or mitophagy inhibitor Mdivi-1 (20 μM), or fisetin + Mdivi-1 for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h, the immunofluorescence staining was performed.

(A) Confocal microscopy analyzed the colocalization of mitochondria (red) with LC3B (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(B) Quantification of Pearson’s colocalization coefficient between mitochondria and LC3B.

(C) Mito-Keima assay analyzed the mitophagy in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 10 μm.

(D) The statistical analysis of mito-Keima assay (550 nm/440 nm).

(E) Confocal microscopy analyzed the colocalization of Salmonella (red) with LC3B (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(F) Quantification of Pearson’s colocalization coefficient between Salmonella and LC3B.

(G) Confocal microscopy analyzed the colocalization of Salmonella (red) with lysosome protein LAMP2 (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(H) Quantification of Pearson’s colocalization coefficient between Salmonella and LAMP2.

Figure 7.

Fisetin facilitates the autophagy receptors and ubiquitin to S. typhimurium. RAW264.7 cells were pretreated with fisetin (20 μM), or fisetin + mitophagy inhibitor Mdivi-1 (20 μM) for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h.

(A) Confocal microscopy analyzed the colocalization of S. typhimurium (red) with SQSTM1 (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(B) Quantification of Pearson’s colocalization coefficient between S. typhimurium and SQSTM1.

(C) Confocal microscopy analyzed the colocalization of S. typhimurium (red) with SQSTM1 (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(D) Quantification of Pearson’s colocalization coefficient between S. typhimurium and SQSTM1.

(E) Confocal microscopy analyzed the colocalization of S. typhimurium (red) with OPTN (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(F) Quantification of Pearson’s colocalization coefficient between S. typhimurium and OPTN.

(G) Confocal microscopy analyzed the colocalization of S. typhimurium (red) with ubiquitin (green) in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection. Scale bar: 5 μm.

(H) Quantification of Pearson’s colocalization coefficient between S. typhimurium and ubiquitin.

Fisetin promotes mitophagy by recruiting p-TBK1 to mitochondria in macrophages

Next, we aimed to investigate the underlying mechanisms that initiate mitophagy and xenophagy in response to S. typhimurium infection. Mitophagy and xenophagy belong to selective autophagy, and both require autophagic receptors, such as SQSTM1, OPTN, and CALCOCO2, to bind to LC3 and initiate mitophagy or xenophagy. Notably, TBK1 is closely implicated in mitophagy and xenophagy for the recruitment of these autophagic receptors. The activation of mitophagy by carbonyl cyanide 3-chlorophenylhydrazone (CCCP), a mitophagy activator, promoted the recruitment of p-TBK1 to mitochondria, whereas Mdivi-1 inhibited the colocalization of p-TBK1 and mitochondria, suggesting that the recruitment of p-TBK1 to mitochondria is necessary for mitophagy [21]. However, information regarding the effect of mitophagy on the p-TBK1 protein expression is scarce. Hence, we investigated the expression of p-TBK1 at different times in response to fisetin treatment (Figure S2A), and found that fisetin induced the accumulation of p-TBK1 in a time-dependent manner. To investigate how fisetin activates p-TBK1, we detected the interaction of p-TBK1/TBK1 with nucleosome assembly protein 1-like 1 (NAP1L1), a histone chaperone protein that is a TBK1 adaptor and closely implicated in the innate cellular response. We found that GSK8612, an inhibitor of TBK1, only repressed the p-TBK1 protein abundance, but not total TBK1 protein (Figure 8A,B). Moreover, we found that fisetin promoted the interaction of NAP1L1 with p-TBK1, as compared with the Salmonella infection group, which was abrogated by GSK8612 treatment. However, fisetin and GSK8612 had scarce effects on NAP1L1 protein abundance, and NAP1L1 hardly binds with total TBK1. These results indicated that fisetin activated p-TBK1 may be attributed to the enhanced interaction with NAP1L1. To further explore the effect of mitophagy on the p-TBK1 protein expression, we detected the p-TBK1 abundance by adapting the mitophagy inhibitor Mdivi-1 (Figure S2B). Intriguingly, the western blot results revealed that Mdivi-1 exerted hardly effects on the p-TBK1 protein abundance. We further validated these results by fluorescence imaging and quantification by using mitophagy inhibitor Mdivi-1 and Pink1 siRNA treatment (Figure S2C-F). Consistently, fisetin treatment dramatically promoted the expression of p-TBK1, while inhibiting mitophagy by using Mdivi-1 or Pink1 siRNA transfection did not abrogate the increased p-TBK1 level. These suggested that inhibition of mitophagy did not repress the p-TBK1 protein abundance. Notably, the failure to induce mitophagy by fisetin occurred once TBK1 was inhibited neither by using the TBK1 inhibitor GBK8612 nor by TBK1 siRNA treatment (Figure S2G-J, Figure 8C–F). Additionally, we further found that fisetin facilitated the recruitment of p-TBK1 to mitochondria, while inhibiting mitophagy by Mdivi-1 blocked the colocalization of p-TBK1 and mitochondria (Figure 8G,H). Together, our results suggest that fisetin promotes mitophagy by recruiting p-TBK1 to mitochondria, and the inhibition of mitophagy only affects the colocalization of p-TBK1 to mitochondria.

Figure 8.

Induction of mitophagy by fisetin requires the recruitment of p-TBK1 to mitochondria.

(A) Coimmunoprecipitation between mitochondrial protein p-TBK1 and NAP1L1 in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection.

(B) Coimmunoprecipitation between mitochondrial protein TBK1 and NAP1L1 in S. typhimurium (100 MOI)-infected RAW264.7 after rinsing 3 times and cultured 4 h post-infection.

(C) RAW264.7 cells were pretreated with fisetin (20 μM), or TBK1 inhibitor GBK8612 (10 μM), or fisetin + GBK8612 for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. Colocalization of mitochondria (red) with LC3B (green) analyzed by confocal microscopy. Scale bar: 5 μm.

(D) Quantification of Pearson’s colocalization coefficient between mitochondria and LC3B.

(E) siRNA (Con and TBK1 siRNA, 50 pM) treated RAW264.7 cells were pretreated with fisetin (20 μM), co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. Colocalization of mitochondria (red) with LC3B (green) analyzed by confocal microscopy. Scale bar: 5 μm.

(F) Quantification of Pearson’s colocalization coefficient between mitochondria and LC3B.

(G) RAW264.7 cells were pretreated with fisetin (20 μM), or fisetin + mitophagy inhibitor Mdivi-1 (20 μM) for 3 h, co-treated with S. typhimurium for 30 min, rinsed 3 times with gentamycin (50 μg/mL) -containing PBS, and post cultured for 4 h. Colocalization of mitochondria (red) with p-TBK1 (green) analyzed by confocal microscopy. Scale bar: 5 μm.

(H) Quantification of Pearson’s colocalization coefficient between mitochondria and p-TBK1.

The effects of mitophagy on the severity of S. typhimurium pathogenesis in mice by inhibitor and siRNA transfection

Our in vitro results demonstrated that the induction of mitophagy restrains the proliferation of S. typhimurium and activates xenophagy. Herein, we next investigated the effects of inhibition of mitophagy on the survival of S. typhimurium and damage in mice by treating with Mdivi-1 (intraperitoneal injection, 25 mg/kg for consecutively 9 days) or Pink1 siRNA (intraperitoneal injection on day 1, 3, 5, and 7, 5 nmol for each mouse) (Figure 9A). Consistent with our previous results (Figure S1B), S. typhimurium treatment resulted in a loss of body weight, whilst fisetin administration ameliorated the body weight loss by S. typhimurium, which was abrogated by Mdivi-1 or Pink1 siRNA treatment (Figure 9B). We next verified the efficiency of Pink1 siRNA by qPCR and found that Pink1 siRNA treatment dramatically downregulated the Pink1 mRNA expression (Figure 9C). Moreover, fisetin alleviated relative spleen, kidney, and liver weight in S. typhimurium-treated mice, and these improvement effects were also abolished by Mdivi-1 or Pink1 siRNA treatment (Figure 9D–F). To investigate the survival of S. typhimurium in the liver of mice, we enumerated the bacterial load in the liver by spread plate methods and found that fisetin administration limited the hepatic S. typhimurium load, whereas Mdivi-1 or Pink1 siRNA treatment increased the S. typhimurium abundance in the liver of mice (Figure 9G). We further assayed the hepatic damage by H&E staining (Figure 9H,I), and the results showed that fisetin administration reduced inflammatory infiltration, nuclei splitting, and karyolysis whilst enhancing the hepatic damage score in the mice infected by S. typhimurium, all of which were abrogated by Mdivi-1 or Pink1 siRNA. Additionally, we also performed immunohistochemistry staining, and the colocalization of TOMM20 and LC3B by fisetin administration could be dampened by Mdivi-1 or Pink1 siRNA treatment, further validating that they were effective in inhibiting mitophagy in mice (Figure 9J,K). Together, these results indicate that the severity of mice pathological lesions resulting from S. typhimurium infection can be alleviated by fisetin administration via activation of mitophagy, which is abrogated by mitophagy inhibitor Mdivi-1 or Pink1 siRNA treatment.

Figure 9.

The effects of mitophagy on the severity of S. typhimurium pathogenesis in mice by inhibitor and siRNA transfection. C57BL/6J mice were randomly allocated into 6 group: 1) Con (Un) group, all mice were sacrificed on day 9; 2) St group, mice were orally administered with 1 × 10⁹ CFU of S. typhimurium on day 7; 3) St+Fis group, orally administrated with 100 mg/kg fisetin for consecutive 9 days and gavaged 1 × 10⁹ CFU of S. typhimurium on day 7; 4) St+Fis+Mdivi-1 group, orally administrated with 100 mg/kg fisetin and intraperitoneal (i.P.) injection of Mdivi-1 (25 mg/kg of body weight) for consecutive 9 days and gavaged 1 × 10⁹ CFU of S. typhimurium on day 7; 5) siNC+St+Fis, orally administrated with 100 mg/kg fisetin for consecutive 9 days, gavaged 1 × 10⁹ CFU of S. typhimurium on day 7, intraperitoneal (i.P.) injection of siRNA con on days 1, 3, 5, and 7; 6) siPink1+St+Fis, orally administrated with 100 mg/kg fisetin for consecutive 9 days, gavaged 1 × 10⁹ CFU of S. typhimurium on day 7, intraperitoneal (i.P.) injection of Pink1 siRNA on days 1, 3, 5, and 7.

(A) The animal experimental design.

(B) Body weight loss.

(C) Relative mRNA expression of Pink1 in the liver of mice.

(D-F) Relative organ weight of spleen, kidney, and liver.

(G) The S. typhimurium burden in the liver.

(H) The representative images of H&E staining of liver sections. Red arrow, karyolysis; yellow arrow, nuclei splitting. Scale bar: 200 μm.

(I) The damage score of hepatic histological morphology.

(J) Quantification of Pearson’s colocalization coefficient between TOM20 and LC3B for (K).

(K) The colocalization of mitochondrial protein TOM20 (red) with LC3B (green) was analyzed by confocal microscopy. Scale bar: 5 μm.

Discussion

In the ongoing struggle between hosts and pathogens for dominance, both sides employ various strategies and countermeasures. An important method is autophagy, classified into nonselective and selective, by which hosts can target bacteria and transport them to lysosomes for degradation and restrain bacterial growth [24]. Accumulating evidence indicates that bacteria have evolved tactics to counteract autophagy [25], including inhibiting the pathways related to autophagy initiation [26,27], avoiding autophagy recognition by masking with host proteins [28,29], interacting with autophagy machinery to escape targeting [30,31], and blocking the fusion of autophagosome and lysosome [32]. In recent years, selective autophagy has been found to be employed by diverse pathogens to facilitate their survival and continuous infection as well as escape host immunity by reducing xenophagy [9]. Accumulating evidence suggests that several pathogens can inhibit host innate immune response by directly or indirectly inducing mitophagy, such as influenza A virus [33], human parainfluenza virus type 3 [34], Listeria [10], and human herpesvirus [35]. Notably, a recent study demonstrated that S. typhimurium remodeled mitochondrial dynamics to enhance intracellular proliferation in macrophages [22]. Additionally, Zhang et al [36] demonstrated that the inhibition of mitophagy may potentially contribute to facilitating the survival of S. typhimurium and escaping from host xenophagy, altering intracellular trafficking to the lysosome, and increasing susceptibility to inflammation. These evidences suggest that the mechanisms of mitophagy are complicated and may vary in accompany with various pathogens. However, information regarding the role of mitophagy and xenophagy in response to S. typhimurium, as well as how mitophagy interacts with xenophagy, is scarce.

In the present study, we found that S. typhimurium treatment led to mitochondrial damage while blocking mitophagy. Normally, mitophagy is a conserved cellular process that plays a crucial role in the autophagic clearance of damaged mitochondria, thereby ensuring the maintenance of a healthy mitochondrial population [37]. This inhibition of mitophagy may be a tactic utilized by S. typhimurium to promote its intracellular proliferation. Therefore, we supposed that mitophagy and xenophagy inducers may contribute to the maintenance of cellular homeostasis. xenophagy inducer can induce xenophagy by various mechanisms [38]. Ammanathan et al. [39] identified a xenophagy modulator, acacetin, which could clear the S. typhimurium infection in a transcription factor EB-dependent way, which belongs to a flavonoid, suggesting the potential of flavonoids in S. typhimurium infection. Here, we identified a natural flavonoid, fisetin, an autophagy inducer, which has been demonstrated to induce autophagy and inhibit inflammation via PI3K-AKT-MTOR signaling in lipopolysaccharide-induced RAW264.7 cells [40]. Moreover, fisetin could also alleviate cognitive impairment by activating mitophagy and suppressing neuroinflammation in rats with sepsis-associated encephalopathy [41]. Notably, fisetin has been described as an anti-Salmonella compound by inhibiting Salmonella T3SS regulator HilD [42]. Therefore, we hypothesized that fisetin could improve the cellular homeostasis upon S. typhimurium infection. We found that fisetin robustly enhanced the LC3B protein level and autophagic vacuole level. The mCherry-GFP-LC3 staining showed activation of autophagic flux by fisetin, and the BafA1 treatment experiment further validated that fisetin activated autophagy. Consistently, fisetin has been shown to promote autophagy, thereby mitigating hepatic injury [43], mice cognitive impairment [41], vascular smooth muscle cell senescence [44], endoplasmic reticulum stress, and mitochondrial stress [45]. These data suggest that fisetin possesses a prospective ability in ameliorating S.S. typhimurium infection.

The PINK1-PRKN dependent pathway is suggested to be the primary regulatory mechanism for mitophagy [21]. Normally, once mitochondria are damaged, like depolarization of the mitochondrial membrane, the PINK1 acts as the sensor of mitochondrial damage, and is subject to trans-autophosphorylation and thereby acts as a monomeric ubiquitin kinase and PRKN kinase, ultimately initiating mitophagy [46]. PRKN, a signal amplifier, is dispersedly distributed in the cytosol in the form of autoinhibition in healthy mitochondria [47]. PRKN is subjected to a transformation from a self‐inhibiting dormant enzyme to an active E3, enabling it to ubiquitinate a variety of outer mitochondrial membrane proteins. These ubiquitinated proteins are subsequently phosphorylated by PINK1 to facilitate the recruitment of additional PRKN to mitochondria, thereby leading to the generation of more ubiquitin chains [48]. Autophagosomes sequester ubiquitin-tagged mitochondria for lysosomal degradation. However, the ubiquitin chains can not directly combine with the autophagic membrane, instead, ubiquitinated cargoes are typically tethered to the autophagic membrane via autophagy adapters, such as SQSTM1 [49], CALCOCO2 [50], and OPTN [51], which contain a ubiquitin-binding domain and an LIR. In the present study, fisetin enhanced mitochondrial LC3B, activated PINK1-PRKN signaling pathway, and activated mitophagy by immunofluorescence staining, mito-Keima assay, as well as coimmunoprecipitation (co‑IP), and facilitated the recruitment of SQSTM1, CALCOCO2, and ubiquitin to mitochondria, suggesting that fisetin promoted mitophagy as compared with the S. typhimurium-treated cells. Consistently, time-course GFP-LC3 autophagy assay revealed that 20% WT S. typhimurium were bound with GFP-LC3 within 40 min of infection, which was subsequently reduced over time, while removal of SopF, a Salmonella T3SS effector that can block xenophagy, resulted in approximately 60% S. typhimurium being labeled with GFP-LC3 within 20 min of infection and increased to 80% at 100 min post-infection [3]. Moreover, fisetin promoted mitophagy and ameliorated the accumulation of dysfunctional mitochondria and mitochondrial reactive oxygen species [52]. Additionally, fisetin could repress neuroinflammation by activating mitophagy [41]. However, despite our results suggesting that the PINK1-PRKN signaling pathway is implicated in the mitophagy induced by fisetin, it does not exclude the possibility that a PINK1-PRKN independent autophagic pathway might be implicated in fisetin-induced mitophagy in S. typhimurium-treated cells.

Previous studies have demonstrated that some pathogens evolve diverse strategies to favor their proliferation and persistent infection [9]. Here, we focused on the role of mitophagy in the survival of intracellular S. typhimurium, and we first found that fisetin (an autophagy inducer and natural flavonoid) decreased the survival of intracellular S. typhimurium in RAW264.7 and BMDM. Consistently, a previous study demonstrated that most of the flavonoids showed inhibitory effects on virus-induced cytopathic effect, among which fisetin, myricetin, quercetin, and genistein exhibited an inhibitory effect against herpes simplex virus type 1 [53]. Likewise, Ammanathan et al. [39] demonstrated that acacetin, a flavonoid, restrained intracellular Salmonella proliferation by promoting xenophagy. S. typhimurium, in contrast to other intracellular bacteria such as Listeria monocytogenes and Mycobacterium bovis that can induce apoptosis and mitophagy, has been shown to possess anti-apoptotic properties [54], and the ability to evade xenophagy [3] by utilizing multiple tactics. Thus, figuring out how intracellular S. typhimurium employs host mitophagy and its interaction with xenophagy may contribute to advancing the knowledge regarding Salmonella infection. In the current study, we found that fisetin reduced the survival of S. typhimurium, whilst blocking mitophagy by mitophagy inhibitor Mdivi-1 or Pink1 siRNA treatment facilitated the replication in RAW264.7 and BMDM. We further found that fisetin activated xenophagy, whereas Mdivi-1 treatment blocked the xenophagy as evidenced by the reduced colocalization of S. typhimurium with LC3B, LAMP2, CALCOCO2, SQSTM1, and ubiquitin. These data suggest that S. typhimurium infection resulted in a dramatic increase of damaged mitochondria whilst blocking mitophagy, while induction of host mitophagy by fisetin treatment is beneficial for the host to restrain intracellular S. typhimurium replication by promoting xenophagy. In consistency with our study, Zhang et al [36] showed that inhibition of mitophagy may potentially contribute to facilitating the survival of S. typhimurium and escaping from host xenophagy. Moreover, S. typhimurium infection resulted in an increase in mitochondrial reactive oxygen species, loss of mitochondrial membrane potential, the release of mitochondrial DNA and cytochrome c into the cytosol, and prevented mitochondrial fragmentation to facilitate intracellular proliferation. Mitochondrial fragmentation may promote mitophagy to remove fragmented mitochondria, and this tactic is mediated by Salmonella invasion protein A (SipA), a type III secretion system effector encoded by Salmonella pathogenicity island 1 [22], further suggesting that induction of mitophagy facilitates the clearance of intracellular S. typhimurium. We further explore the potential mechanism that may mediate the initiation of selective autophagy.

TBK1 is an essential signaling molecule for interferon-beta (IFN-beta) production, recruitment of autophagy receptors for selective autophagy as well as innate immunity [55,56]. Notably, TBK1 has been demonstrated to be implicated in mitophagy and xenophagy [57]. In the present study, we first investigated the p-TBK1 protein abundance and found that p-TBK1 was activated during fisetin-induced mitophagy. We further found that fisetin promoted the interaction between p-TBK1 and its adaptor NAP1L1, which was reversed by GSK8612 (an inhibitor of TBK1) treatment. However, fisetin or GSK8612 treatment did not affect the interaction between TBK1 and NAP1L1. These suggest that fisetin-enhanced p-TBK1 protein abundance may be related to the enhanced interaction between p-TBK1 and its adaptor NAP1L1. Moreover, inhibiting mitophagy by neither mitophagy inhibitor Mdivi-1 nor Pink1 siRNA affects the expression of p-TBK1, indicating that inhibiting mitophagy would not affect the p-TBK1 protein abundance. We next investigated the role of p-TBK1 in mitophagy by using TBK1 inhibitor GBK8612 and Tbk1 siRNA treatment, and found that the mitophagy activated by fisetin treatment was significantly dampened by GBK8612 treatment, indicating the importance of p-TBK1 in inducing mitophagy. Consistently, TBK1 could restrict bacterial growth, while knockdown of TBK1 contributed to bacterial replication and increased the number of Salmonella-containing cells [50]. Likewise, Wang et al. [58] showed that inhibiting the expression of TBK1 led to a decrease in xenophagy and facilitated the survival of Mycobacterium bovis in murine macrophages, further indicating that TBK1 is vital for the host to activate xenophagy and eliminate bacteria. Moreover, several studies revealed that TBK1 plays an important role in facilitating mitophagy [19,59,60]. We further investigated the distribution of p-TBK1 in response to mitophagy since p-TBK1 was important for mitophagy, whilst inhibiting mitophagy would not affect the p-TBK1 protein abundance. We found that fisetin facilitated the recruitment of p-TBK1 to mitochondria, which was abrogated by Mdivi-1 treatment. This suggests that fisetin-induced mitophagy required the recruitment of p-TBK1 to mitochondria. Consistently, inhibiting mitophagy markedly blocked the colocalization of p-TBK1 with mitochondria, whereas activating mitophagy promoted [21]. Intriguingly, Song et al. [21] also found that promotion of mitophagy markedly reduced the colocalization of p-TBK1 with Mycobacterium bovis, suggesting TBK1 May be competed by mitophagy and xenophagy during Mycobacterium bovis infection. Although both Mycobacterium bovis and S. typhimurium are intracellular bacteria, they employ different strategies to avoid host innate immunity, as S. typhimurium may possess more complicated tactics such as its famous type III secretion systems, its ability to prevent mitochondrial fragmentation, and its diverse virulence proteins [22,61]. Additionally, PRKN is the upstream molecule of TBK1 for the following two reasons: 1) Overexpression of TBK1 (Lv-TBK1) did not influence the PRKN expression; 2) TNF and Lv-TBK1-induced TBK1 phosphorylation was inhibited, and the co-localization of p-TBK1 and mitochondria was also weakened after using siRNA against Prkn [62]. Together, our data suggest that fisetin induced PINK1-PRKN signaling pathway-mediated mitophagy to activate xenophagy and restrict intracellular S. typhimurium, which requires the recruitment of p-TBK1 to mitochondria.

We also validated the regulative effects of fisetin in response to S. typhimurium infection by performing multiomics analysis in vitro and in vivo. The cellular RNA-Seq transcriptome analysis revealed that fisetin restored substantial gene expressions to a level that was similar to the control in S. typhimurium-infected cells, and the results in vivo study demonstrated that fisetin ameliorated the body weight loss, lowered relative organ weights (spleen, kidney, and liver), and reduced the hepatic S. typhimurium load in S. typhimurium-infected mice. Moreover, LC-MS non-targeted metabolomics analysis revealed that fisetin restored the hepatic metabolite levels in response to S. typhimurium. These data supported our hypothesis that fisetin could robustly alleviate the cytological injury induced by S. typhimurium, encouraging us to explore the potential mechanisms further. Notably, the ameliorative effects of fisetin in the animal experiment were abrogated by Mdivi-1 or Pink1 siRNA treatment in vivo. All these data suggested that fisetin could ameliorate the cytological injury by S. typhimurium infection by promoting mitophagy. Consistently, a recent study indicated that fisetin administration disrupted the interaction between the regulatory protein HilD and the promoters of its target genes, leading to a suppression in the expression of T3SS–1 effectors as well as structural proteins, thus reducing the pathology [42]. Likewise, fisetin has also been shown to interfere with the oligomerization of listeriolysin O to inhibit the virulence of Listeria monocytogenes [63,64]. Furthermore, metabolomic analysis showed that fisetin significantly modulated bacterial metabolites, and thin-layer chromatography demonstrated that fisetin could restore the activity of polymyxin E in S. typhimurium against mobilized colistin resistance-1 (a specific resistance gene endows bacteria with resistance to colistin) [65]. Additionally, S. typhimurium could prevent mitochondrial fragmentation that may inhibit mitophagy [22], while fisetin suppressed inflammation in rats with sepsis-associated encephalopathy by activating mitophagy, suggesting fisetin may be a potential material to treat S. typhimurium by induction of mitophagy. Thus, the induction of mitophagy by fisetin may offer a valuable new perspective for the development of effective therapeutic strategies against S. typhimurium infection.

In conclusion, our results suggest that S. typhimurium infection induces mitochondrial damage whilst inhibiting mitophagy, while fisetin could restrain the survival of intracellular S. typhimurium by activating PINK1-PRKN mediated mitophagy to promote xenophagy, which requires the recruitment of p-TBK1 to mitochondria. Multiomics analysis and animal experiments demonstrated the effectiveness of fisetin in treating S. typhimurium infection. Fisetin proves effective as a xenophagy enhancer in an in vivo mouse infection model, reducing intracellular Salmonella burden. These findings also provide a new perspective to advance the intracellular survival mechanism of S. typhimurium in the host.

Materials and methods

Ethics statement

The animal experiment was approved by the Institutional Animal Care and Use Committee of China Agricultural University (Approval No: AW72602202–1–5). All animal experimental procedures were performed according to the guidelines outlined in the United States National Research Council’s Guide for the Care and Use of Laboratory Animals.

Reagents and antibodies

Chemical reagents used in the current study like fisetin (HY-N0182), bafilomycin A1 (HY-100558), Mdivi-1 (HY-15886), GBK8612 (HY-111941) were purchased from MedChemExpress. The primary antibody used in the present study including mouse monoclonal anti-LC3 antibody (Cell Signaling Technology, 12741S), rabbit monoclonal anti-LC3 antibody (Cell Signaling Technology, 83506S), rabbit monoclonal anti-phospho-TBK1 antibody (5483S), mouse monoclonal anti-TBK1 antibody (51872) were purchased from Cell Signaling Technology; HRP-conjugated alpha tubulin monoclonal antibody (HRP-66031), rabbit polyclonal anti-VDAC1 antibody (55259–1-AP), rabbit polyclonal anti-PINK1 antibody (23274–1-AP), rabbit polyclonal anti-PRKN/PARK2/Parkin antibody (14060–1-AP), rabbit polyclonal anti-TOMM20 antibody (11802–1-AP), mouse monoclonal anti-LAMP2 antibody (66301–1-Ig), rabbit polyclonal anti-CALCOCO2/NDP52 antibody (12229–1-AP), rabbit polyclonal anti-SQSTM1/p62 antibody (18420–1-AP), rabbit polyclonal anti-OPTN antibody (10837–1-AP), rabbit polyclonal anti-ubiquitin antibody (10201–2-AP), rabbit polyclonal anti-NAP1L1 antibody (14898–1-AP) were purchased from Proteintech; rabbit polyclonal anti-Salmonella antibody (ab35156) and mouse monoclonal anti-Salmonella antibody (ab8274) were purchased from Abcam. HRP-conjugated Mouse anti-Rabbit lgG Light chain (As061) was purchased from ABclonal Technology. Fetal bovine serum (FBS 16,140,071) and penicillin-streptomycin (15140163) were purchased from Gibco.CSF1 (HY-P7085), gentamycin (HY-A0276), CCK8 assay kit (HY-K0301), Protein A/G Magnetic Beads (HY-K0202) were purchased from MedChemExpress. Mitochondrial membrane potential assay kit with JC-1 (C1071), Mitochondrial isolation kit (C3601), LDH kit (C0016), MitoTracker Red CMXRos (C1035), LysoTracker Green (C1047S) were purchased from Beyotime Biotechnology. mito-Keima adenoviruses (HBAD-1022) and mCherry-GFP-LC3 (HB-AP210 001) were purchased from HanBio. TRIzol reagent (RN65), reverse transcription reagent (PC58), and SYBR Premix (PC60) were purchased from Aidlab Biotechnology. Salmonella-Shigella agar (LA0075), and hematoxylin and eosin (G1120) were purchased from Solarbio. Enhanced chemiluminescence kit (HX1868), Dylight488-goat anti-mouse IgG (H+L) (HX2082), Dylight488-donkey anti-rabbit IgG (H+L) (HX2084), Dylight594-goat anti-mouse IgG (H+L) (HX2075), Dylight594-goat anti-rabbit IgG (H+L) (HX2074), Dylight405-goat anti-rabbit (HX2091) were purchased from Huaxingbio.

Cells culture

RAW264.7 cells were cultured in DMEM-F12 medium containing 10% FBS and 1% penicillin-streptomycin, and cultured in a 5% CO₂ incubator at 37°C.

Bone marrow derived-macrophages isolation and culture

BMDM cells were isolated according to the method in previous studies [66,67]. Briefly, bone marrow, harvested from femurs and tibias of 6–8 weeks old C57BL/6J mice, was flushed with prechilled DMEM medium. The cell pellets were filtered via a 70 μm cell strainer and lysed with RBC lysis buffer (eyotime Biotechnology, PF00014). After centrifuging at 500 g for 5 min, the cells were suspended in DMEM medium with 10 ng/mL of macrophage colony-stimulating factor, 10% FBS, and 1% penicillin-streptomycin. The DMEM medium was replaced on days 3 and 5. The BMDM cells were fully differentiated and prepared for use on day 7.

Bacterial culture and infection

The bacterial culture and infection were according to a previous method by Xu et al. [3] with some modifications. Briefly, S. typhimurium was grown overnight in LB medium with shaking at 220 rpm in a 37°C incubator and then sub-cultured at a ratio of 1:33 for 3–4 h.

For cell infection, RAW264.7 cells were cultured with penicillin-streptomycin-free DMEM medium at 37°C in a 5% CO₂ incubator. The cells were pre-treated with fisetin for 3 h and co-treated with fisetin and S. typhimurium throughout the infection. The MOI was 100 for RAW264.7 cell infection. The RAW264.7 cells were incubated with S. typhimurium at 37°C for 30 min, after which the cells were washed with PBS (containing 50 μg/mL gentamycin) three times to remove the extracellular S. typhimurium and cultured in fresh DMEM medium that contains 50 μg/mL gentamycin for 4 h.

Intracellular S. typhimurium proliferation

The intracellular S. typhimurium proliferation was determined with a previous method [3]. Briefly, cells in 6-well plates were infected with S. typhimurium for 30 min. After washing three times, cells were cultured in a gentamycin-containing DMEM medium for 4 h. Then, cells were lysed in prechilled PBS solution supplemented with 0.1% Triton X-100 (Beyotime Biotechnology, ST1723), and colony-forming units (CFU) were assayed via serial-dilution plating on Salmonella-Shigella agar plates.

Flow cytometry analysis of mitochondrial membrane potential

Cells co-treated with fisetin and S. typhimurium were harvested and incubated with a mitochondrial membrane potential assay kit with JC-1 (Beyotime Biotechnology, C1071) for 20 min at 37°C. Subsequently, cells were analyzed by a flow cytometer (Beckman Coulter, Miami, FL, USA).

Mitochondria isolation

Mitochondria was isolated via a commercial kit (Beyotime Biotechnology, C3601) according to the instructions of the manufacturer.

Western blot analysis

Cell samples were lysed in RIPA lysis buffer (Beyotime Biotechnology, P0013D), and the concentration of protein was evaluated by the bicinchoninic acid method. Proteins were subsequently denatured and subjected to SDS-PAGE, transferred onto a PVDF membrane (Millipore, IPVH00010) (Millipore, Billerica, USA), and incubated with antibodies. Protein bands were bonded with an enhanced chemiluminescence kit, followed by visualizing by using the ImageQuant LAS 4000 mini system (GE Healthcare, Piscataway, NJ, USA). Protein intensity was measured using ImageJ software (National Institutes of Health, Bethesda, Maryland, USA, https://imagej.nih.gov/ij/). Results were expressed relative to the control group.

Cell viability assay

The cell viability was determined by using a CCK8 assay kit (MedChemExpress, HY-K0301). Briefly, cells were treated with differential doses of fisetin for 3 h, Mdivi-1, and GSK8616 for 8 h. The absorbance was assayed at 450 nm with a microplate reader (SpectraMax M3, Sunnyvale, CA, USA). Results are exhibited as percentages relative to the control group.

Mito-Keima assay

RAW 264.7 cells were transfected with mito-Keima adenoviruses (HanBio, HBAD-1022) according to the instructions of the manufacturer. Briefly, cells were seeded on glass-bottom 24-well plates. Cells at 60–70% confluence were treated with fresh basal DMEM medium with adenovirus and 2% FBS for 8 h, after which cells were incubated with complete DMEM medium (containing 10% FBS and 1% penicillin-streptomycin) for 24 h. Images were captured by using a confocal microscope (TCS SPE, Leica, Germany).

LDH analysis

LDH was assayed with a commercial kit (Beyotime Biotechnology, C0016) according to the instructions of the manufacturer.

Transmission electron microscopy

Cells co-treated with fisetin and S. typhimurium were harvested and fixed with 2.5% glutaraldehyde, followed by post-fixed in 1% osmium tetroxide. Fixed cells were subjected to dehydration with a gradient of alcohol and being embedded in epoxy blocks. Sections (1 μm) were stained with lead citrate and uranyl acetate, which were visualized by a transmission electron microscope (JEM-1400PLUS, Japan).

Adenovirus transfection and confocal fluorescence microscopy

RAW 264.7cells were transfected with adenoviruses to express mCherry-GFP-LC3 (HanBio, HB-AP210 001) according to the instructions of the manufacturer. Briefly, cells were seeded on glass-bottom 24-well plates. Cells at 60–70% confluence were treated with fresh basal DMEM medium with adenovirus and 2% FBS for 8 h, after which cells were incubated with complete DMEM medium (containing 10% FBS and 1% penicillin-streptomycin) for 24 h. Cells treated with fisetin were captured by using a confocal microscope (TCS SPE, Leica, Germany).

Immunofluorescence staining

Cells were fixed with 4% formaldehyde for 20 min at ambient temperature and treated with 0.1% Triton X-100 for 5 min. Cells were blocked with a 3% bovine serum albumin solution containing 0.01% Triton X-100 for 1 h, and incubated with a diluted primary antibody overnight at 4°C. Cells were then incubated with a secondary antibody for 2 h at ambient temperature. Subsequently, cells were incubated with a different species of primary antibody overnight at 4°C and the corresponding secondary antibody for 2 h at ambient temperature. After staining with Hoechst 33,342 for 3 min, cells were visualized under a confocal microscope.

MitoTracker red staining

MitoTracker Red was stained according to the instructions of the manufacturer. Briefly, cells co-treated with fisetin and S. typhimurium were incubated with MitoTracker Red CMXRos (Beyotime Biotechnology, C1035), (50 nM) for 20 min. Then cells were rinsed with basal DMEM medium three times and visualized under a confocal microscope.

LysoTracker Green staining

LysoTracker Green was stained according to the instructions of the manufacturer. Briefly, cells were incubated with LysoTracker Green (Beyotime Biotechnology, C1047S), (50 nM) for 30 min and rinsed with basal DMEM medium three times, which was visualized under a confocal microscope.

co‑IP assay

Treated cells were lysed with RIPA buffer for western blot and co-IP, and the concentration of protein was evaluated by the bicinchoninic acid method. Protein extracts were incubated with anti-LC3 antibody or anti-mouse secondary antibody on a rotator overnight at 4°C, and incubated with precleared Protein A/G Magnetic Beads (MedChemExpress, HYK0202) according to the instruction of the manufacturer. The protein beads were incubated with SDS-PAGE buffer for 10 min at 99°C. The supernatant was subsequently subjected to western blot, with the whole-cell lysates being the input control.

Quantitative PCR (q-PCR)

RNA was extracted by using TRIzol reagent (Aidlab Biotechnology, RN65) and transcribed into cDNA using a commercial kit (Aidlab Biotechnology, PC58) according to the instructions of the manufacturer. qRT-PCR was performed on an Archimed X4 system using a SYBR Premix (Aidlab Biotechnology, PC60). Gene expression was calculated using the 2^−ΔΔCt method, and Gapdh was used as an internal reference gene. The primers used in the present study were as follows: Gapdh, 5’-AAGCCCATCACCATCTTCCA-3’; Gapdh, 5’-CACCAGTAGACTCCACGACA-3’; Pink1, 5’-ATCTGGTTCAGCAGGGCATT-3’; Pink1, 5’- AGGGACAGCCATCTGAGTCC-3’; Tbk1, 5’-GTACGGTGGCTGGTTGAACT-3’; Tbk1, 5’-TCTTGATGTGCCCAGGTGTC-3’.

RNA-Seq transcriptome analysis

RAW264.7 cells were seeded in a 10 cm cell culture dish and treated with fisetin and S. typhimurium for the indicated times. Six replicates of cell samples for each group were harvested, and total RNA was extracted with TRIzol reagent. After validating the RNA quality by 1.5% agarose gel electrophoresis, RNA sequencing was conducted via Shanghai Majorbio Bio-Phamtech Corporation. All bioinformatics analysis was performed by using the Majorbio Cloud Platform (https://www.majorbio.com).

Animal model of infection

Animal experiment 1: 6- to 8-week-old male C57BL/6J mice (n = 32), purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China), were raised at a temperature of 25 ± 2°C with a 12:12 h light/dark cycle, and had ad libitum access to water and feed. Mice were randomly assigned to 4 groups: the control group (Con), the fisetin treatment group (Fis), the S. typhimurium group (St), and the S. typhimurium and fisetin co-treatment group (St + Fis). Mice in the Fis and St+Fis groups were orally gavaged with 100 mg/kg of fisetin for 9 consecutive days, while mice in the St and St+Fis groups were gavaged with an equal volume of 0.5% CMC-Na (Sangon Biotech, A501427) sterile water. On day 7, mice in the ST and St+Fis groups were gavaged with 1 × 10⁹ CFU of S. typhimurium, while mice in the CON and Fis groups were given 200 μL of sterile water. All mice were euthanized by carbon dioxide asphyxiation and subsequently dissected. Spleen, kidney, and liver were subsequently separated and weighed, and organ index was expressed as the percentage of organ weight relative to body weight.

Animal experiment 2: 6- to 8-week-old male C57BL/6J mice (n = 36), purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China), were raised at a temperature of 25 ± 2°C with a 12:12 h light/dark cycle, and had ad libitum access to water and feed. Mice were randomly assigned to 6 groups: the control group (CON), the S. typhimurium group (St), the S. typhimurium and fisetin co-treatment group (St + Fis), the S. typhimurium + fisetin + Mdivi-1 co-treatment group (St + Fis + Mdivi-1), the S. typhimurium + fisetin + negative control siRNA (siNC+St+Fis), and the S. typhimurium + fisetin + Pink1 siRNA (siPink1+St+Fis). S. typhimurium treatment and fisetin treatment were according to the methods in animal experiment 1. Mice were intraperitoneal (i.p) injected with Mdivi-1 (25 mg/kg of body weight) for consecutive 9 days. Mice were knocked down by i.p injection with negative control siRNA and Pink1 siRNA (5 nM per mouse) on days 1, 3, 5, and 7. All mice were euthanized by carbon dioxide asphyxiation and subsequently dissected. Spleen, kidney, and liver were subsequently separated and weighed, and organ index was expressed as the percentage of organ weight relative to body weight. Liver samples were fixed in 4% paraformaldehyde for H&E and immunohistochemistry staining.

Enumeration of viable bacteria in liver

Approximately 0.1 g of liver sample was harvested, weighed, and mixed with 1 mL sterile saline, which was immediately subjected to homogenization by a Biohomogenizer (Bead Ruptor, Omni International, USA). The liver homogenate was diluted into a series of differential concentration gradients and was assayed via serial-dilution plating on Salmonella-Shigella agar plates.

Hepatic LC-MS non-targeted metabolomics

The metabolomics of liver tissue was analyzed by Shanghai Majorbio Bio-Pharm Technology Co., Ltd. in accordance with a previous method by Yin et al. [68]. All bioinformatics analysis was performed by using the Majorbio Cloud Platform (https://www.majorbio.com).

Transient transfection assay

Mouse Pink1- and Tbk1-targeting siRNA oligonucleotides were purchased from Suzhou Genepharma Co., Ltd (Suzhou, China). Pink1, 5’-GGAUUAUCUGAUAGGGCAATT-3’; Pink1, 5’-UUGCCCUAUCAGAUAAUCCTT-3’; Tbk1, 5’-GCUCCUGUCUGAUAUCCUATT-3’; Tbk1, 5’-UAGGAUAUCAGACAGGAGCTT-3’.

Cells were seeded in 24-well cell culture plates (1 × 10⁵ cells each well) overnight. Pink1 or TBK1, and negative control siRNA were transfected into Raw264.7 and BMDM cells at 50 pM final concentration using Lipo8000™ Transfection Reagent (Beyotime biotechnology, C0533) in accordance with the manufacturer’s instructions. The transfection medium was replaced with a fresh complete DMEM medium after 8 h to reduce cytotoxicity. The cells were transfected for 24 h before the next experiment, and the efficiency of siRNA silencing was determined via qRT-PCR.

Histopathological analysis

Liver samples were fixed in 4% formaldehyde, embedded in paraffin after gradient dehydration steps, sectioned, and stained with hematoxylin and eosin. At least six slides from each mouse were randomly visualized for pathological examination by a blinded observer. The histological score criteria are presented in Table S1.

Immunohistochemistry staining

Liver samples were fixed in 4% formaldehyde, embedded in paraffin after gradient dehydration steps, and sectioned. The 5-μm liver sections were subjected to antigen retrieval by using a commercial reagent (Beyotime Biotechnology, P0083) according to the manufacturer’s instructions. Then, the sections were blocked with 3% bovine serum albumin for 1 h at ambient temperature, incubated with a diluted primary antibody overnight at 4°C, next incubated with a secondary antibody for 2 h at room temperature, incubated with a different species primary antibody overnight at 4°C and corresponding secondary antibody for 2 h at ambient temperature. After staining with Hoechst 33,342 for 3 min, sections were visualized under a confocal microscope.

Statistical analysis

All cell experiments were conducted on three separate occasions. Data were analyzed by using an Unpaired t-test (two-tailed) and one-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparison test with GraphPad Prism 8 (GraphPad Software, CA). Results were exhibited as means ± SEM, and *p < 0.05, **p < 0.01, ***p < 0.001.

Correction Statement

This article was originally published with errors, which have now been corrected in the online version. Please see Correction https://doi.org/10.1080/15548627.2025.2609439

Funding Statement

This work was supported by the National Outstanding Youth Science Fund Project of National Natural Science Foundation of China [No. 31625025], the National Key R&D Program of China [2022YFF1100102, 2022YFC2105005], the National Natural Science Foundation of China [No.32172749].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author.

Author contribution

Jun Li: Conceptualization, Investigation, Writing – original draft; Yang Yang: Conceptualization, Investigation; Yao Ge: Methodology, Investigation; Xinyu Zhang: Methodology, Investigation; Haozhen Liu: Investigation, Methodology; Yinfeng Chen: Data curation; Ying Yang: Writing – review &editing; Zhenlong Wu: Writing – review &editing, Supervision, Funding acquisition

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15548627.2025.2534298