Iron regulatory protein 2 contributes to antimicrobial immunity by preserving lysosomal function in macrophages

Significance

Iron regulatory protein 2 (Irp2), an Irp in cytoplasm, is mainly responsible for cellular iron homeostasis. In this study, we show that Irp2 affects lysosome biosynthesis and acidification through hypoxia-inducible factor 2 and its target gene lactate dehydrogenase A in macrophages. Irp2 is essential for the macrophage’s immune function in clearing pathogens against infection. Lactic acid production or hypoxia-inducible factor 2 inhibitors are potential therapeutical treatments for inflammatory bowel disease with or without pyogenic liver abscesses.

Abstract

Colorectal cancer and Crohn’s disease patients develop pyogenic liver abscesses due to failures of immune cells to fight off bacterial infections. Here, we show that mice lacking iron regulatory protein 2 (Irp2), globally (Irp2^−/−) or myeloid cell lineage (Lysozyme 2 promoter-driven, LysM)-specifically (Irp2^ΔLysM), are highly susceptible to liver abscesses when the intestinal tissue was injured with dextran sodium sulfate treatment. Further studies demonstrated that Irp2 is required for lysosomal acidification and biogenesis, both of which are crucial for bacterial clearance. In Irp2-deficient liver tissue or macrophages, the nuclear location of transcription factor EB (Tfeb) was remarkably reduced, leading to the downregulation of Tfeb target genes that encode critical components for lysosomal biogenesis. Tfeb mislocalization was reversed by hypoxia-inducible factor 2 inhibitor PT2385 and, independently, through inhibition of lactic acid production. These experimental findings were confirmed clinically in patients with Crohn’s disease and through bioinformatic searches in databases from Crohn’s disease or ulcerative colitis biopsies showing loss of IRP2 and transcription factor EB (TFEB)-dependent lysosomal gene expression. Overall, our study highlights a mechanism whereby Irp2 supports nuclear translocation of Tfeb and lysosomal function, preserving macrophage antimicrobial activity and protecting the liver against invading bacteria during intestinal inflammation.

Pyogenic liver abscess (PLA) is sign of severe liver disease that results from the invasion of suppurative bacteria into the liver parenchyma and source of significant global public health concerns (1). PLA occurs under various conditions, such as biliary tract disease, colitis, pancreatitis, intravenous drug use, and trauma (2). The incidence of PLA ranges from 1.1 to 18 per 100,000 person-years, and the fatality rate can reach up to 30% due to the associated severe complications in patients (3, 4). Patients with inflammatory bowel disease (IBD) have a higher risk of PLA incidence (5), estimated at approximately 7 per 10,000 patients with IBD (6, 7). One suggestion why patients with Crohn’s disease or ulcerative colitis are at high-risk of developing this complication is that gut barrier disruption in colitis enhances microbe infiltration into the submucosae, which increases the probability that gut-derived bacteria are translocated from the gut to the liver (7–9). In some cases, the bacterial origin of the liver abscess from gut bacteria was confirmed (6, 10, 11). Therefore, proper homeostasis of the gut barrier is crucial to maintaining a healthy liver.

Macrophages are critical in maintaining the homeostasis of the gut–liver axis by preventing bacterial translocation. Macrophages reside along the entire length of the gastrointestinal tract and throughout all layers of the gut wall, strategically positioned to engulf and destroy any bacteria penetrating the epithelial barrier, preventing bacteria translocation to the portal vein (reviewed in ref. 12). Hepatic inflammation in patients with cirrhosis affects macrophage function, leading to inadequate bacterial clearance and making patients particularly prone to developing organ failure due to infection (13). Impaired clearance by intestinal macrophages also triggers a compensatory adaptive immune response, resulting in chronic inflammation that worsens Crohn’s disease (14). Recently, Galy and collaborators demonstrated that iron regulatory proteins 1 and 2 (Irp1/2) participate in the immune function of macrophages, presenting evidence that the Irp1/2 double-knockout (KO) significantly reduced its survival rate when challenged by intraperitoneal injection of Salmonella sp. (15). The authors described that the Irp1/2 KO animals retained immature neutrophils with abnormally high glycolytic activity, suggesting that neutropenia was a key underlying mechanism (16).

IRP1 and IRP2 have mostly redundant functions in regulating cellular iron homeostasis through their RNA-binding activity to iron-responsive elements (IRE) that are conserved primary mRNA sequences folding into secondary stem-loop structures (17). Under low iron conditions, IRP1 and IRP2 inhibit the translation of transcripts involved in iron storage and export of efflux by binding to IRE in the 5′-Untranslated Region (UTR) of respective mRNAs such as ferritin H&L (FTH and FTL) and ferroportin. Conversely, IRPs stabilize mRNAs to promote the expression of genes responsible for iron import and utilization by binding to IREs in the 3′-UTR of mRNA, such as divalent metal transporter 1 and transferrin receptor 1. The main difference between IRP1 and IRP2 is seen under high iron conditions, where IRP1 gains cis-aconitase activity and loses IRE binding activity by assembling an iron-sulfur cluster (4Fe-4S), whereas IRP2 undergoes iron-dependent degradation. Deleting either Irp1 or Irp2 in mice does not affect the survival of these mice, but the preclinical manifestations are different (reviewed in ref. 18). Overall, Irp2 dominates the regulation of iron homeostasis in vivo since Irp1 ablation mis-regulates iron metabolism only in the kidney and brown fat, whereas Irp2 ablation mis-regulates the expression of target proteins in all tissues (19).

Our previous studies have shown that Irp2 ablation leads to an energy-metabolism switch from oxidative phosphorylation to aerobic glycolysis in mouse embryonic fibroblasts (20), suggesting a potential role of Irp2 ablation in promoting tumor growth. To establish a model of colon cancer, we exposed both wild type (WT) and Irp2^−/− mice to azoxymethane/dextran sodium sulfate (DSS) treatment. To our surprise, we found that some Irp2^−/− mice died in the second week of treatment without the occurrence of colon cancer. The autopsy revealed that DSS-induced liver injury and abscesses are the likely cause of death. Further studies elucidated the essentiality of Irp2 in macrophage-lysosomal biogenesis and function, particularly in clearing pathogens in mice and human subjects.

Results

Liver Abscesses Occur Following DSS Treatment in Irp2^−/− and Irp2^ΔLysMMice.

WT and Irp2 KO mice were exposed to 2.5% DSS in their drinking water for 7 d, followed by 3 d of regular water and then an additional 7 d of 2.5% DSS as shown in Fig. 1A. This exposure to DSS is well tolerated by WT but the mortality rate of Irp2^−/− mice was 33% in the second week and reached 45% at the experimental endpoint (n = 30) (Fig. 1B). Irp2^−/− mice also experienced significant weight loss (greater than 20%) at the endpoint of DSS treatment (Fig. 1C). To investigate the cause of these phenotypes, we conducted routine blood tests, biochemical examinations, and autopsy. The results showed a significantly increased number of inflammatory white blood cells (SI Appendix, Fig. S1) and increased alanine transaminase and aspartate aminotransferase activities (Fig. 1D), the former finding indicating high inflammation and the latter liver dysfunction. Upon gross examination, we identified many white or yellowish spots in the Irp2^−/− liver (Fig. 1E). Hematoxylin and Eosin (H&E) staining confirmed liver damage and massive infiltration of proinflammatory cells, a typical liver abscess feature, found exclusively in Irp2^−/− mice (Fig. 1F). To determine the presence of infection, gram-positive staining and blood plate cultures were carried out. These assays confirmed the presence of gram-positive bacteria (Fig. 1G) and significant bacterial loads (Fig. 1 H and I) in the livers of Irp2^−/−mice after DSS treatment. As DSS often induces intestinal barrier damage (21), we hypothesized that the liver abscess could be due to bacterial translocation from the intestine. To verify this hypothesis, we also examined peripheral and portal vein blood with blood plate assays. The results showed large quantities of bacteria in the Irp2^−/− portal veins after DSS treatment but barely any in peripheral blood (Fig. 1 H and I). In combination, these findings suggested that intestinal bacteria enter through the damaged intestinal barrier into the portal veins and then colonize the livers of Irp2^−/− mice.

Fig. 1.

Liver abscesses are induced by DSS treatment in Irp2 KO mice. (A) Diagram showing the timing (in days) of treatment applications in the mouse model. (B) Survival plot of the animals during the treatment period (n = 30). (C) Body weight changes during the treatment period (initially n = 30, but some animals deceased). (D) Serum parameters for liver function. (E) The appearance of the liver of WT and Irp2^−/− mice 17 d after initiation of DSS exposure. The white arrows indicated the sites of pus-filled lesions. (F) H&E staining of liver tissue. (G) Gram-staining of the liver. (H) Bacterial growth on blood agar plates from the liver, portal vein, and peripheral blood. (I) Quantification of the numbers of bacterial colonies in liver and portal vein blood after DSS exposure. n = 5/group if not specified. Data are presented as means ± SD. **P < 0.01, ***P < 0.001, ****P < 0.0001. ns: no significance. ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; ALB, Albumin. White boxes indicate the enlarged regions.

We then investigated whether intestinal macrophage function was affected by Irp2 deficiency, as these cells are the front line of defense against bacterial invasion. The intestinal macrophages are commonly considered continuous replenishers derived from myeloid monocytes circulating in the bloodstream due to constant exposure to commensal microbes and low-grade inflammation (22). To generate Irp2-ablation mice in myeloid cell lineage, we purchased Irp2^flox/flox mice from Cyagen Biosciences Inc. (Suzhou, China) and crossed them with the same genetic background (C57BL6J) LysM^cre mice (Fig. 2A). We first confirmed the KO efficiency by showing loss of Irp2 in isolated bone marrow–derived macrophages (BMDM) (Fig. 2B). The resulting Irp2^ΔLysM mice were subjected to the same DSS treatment as Irp2^−/− mice mentioned in Fig. 1. Similar results were obtained to those of global Irp2-ablation mice regarding body weight and mortality rates (Fig. 2 C and D). Additionally, only Irp2^ΔLysM mice had liver abscesses (Fig. 2E). These findings suggest that myeloid cell Irp2 deficiency can increase the risk of DSS-induced liver abscesses. The presence of bacteria in the liver was confirmed by H&E staining, gram-positive staining, and blood plate assays, which showed that the bacteria originated from the barrier-injured intestine and colonized in the liver of Irp2^ΔLysM after DSS treatment (Fig. 2 F–I). The results indicate the critical role of Irp2 in the myeloid cell lineage against bacterial infection in mice.

Fig. 2.

Irp2 deficiency in macrophages promotes DSS-induced liver abscesses. (A) Diagram showing the strategy to generate Irp2^ΔLysM mice. (B) Representative Western blotting demonstrating the Irp2 KO in BMDM isolated from Irp2^ΔLysM and Irp2^flox/flox mice. (C) Survival plot (n = 30). The experimental procedure was the same as in Fig. 1A. (D) Body weight changes post-DSS drinking (initially n = 30). (E) The appearance of the representative livers from WT and Irp2^ΔLysM mice with or without DSS treatment. White arrows indicated the sites of liver abscesses. (F) H&E staining of the liver sections. (G) Gram-positive staining with the liver sections. (H) Bacterial growth on blood agar plates from the liver, portal vein, and peripheral blood. (I) Quantification of the number of bacteria in liver tissue and portal vein blood after DSS treatment. n = 5/group if not specified. Data are means ± SD. ****P < 0.0001. White boxes indicate the enlarged regions.

Decreased Bacterial Clearance in Irp2^−/− Macrophages after Treatment with Escherichia coli.

Macrophages eliminate pathogens through phagocytosis, phagosome maturation, and pathogen clearance (23). To determine whether Irp2 depletion affected the phagocytic or clearance capacity of macrophages, the isolated BMDM or peritoneal macrophages were treated with Dextran-Fluorescein Isothiocyanate (FITC) or enhanced green fluorescent protein (eGFP)- or mCherry-expressing E. coli for 15 min to evaluate phagocytosis or for 1 h to assess pathogen clearance competence. There was no significant difference in fluorescence intensity and the number of bacterial colonies after 15 min of incubation (SI Appendix, Fig. S2 A–C), suggesting a similar phagocytosis capacity of WT and Irp2^−/− mutant macrophages. However, after 1 h of incubation, the number of bacteria and the fluorescence intensities of mCherry- and eGFP-E. coli were substantially increased in Irp2^−/− macrophages (Fig. 3 A and B and SI Appendix, Fig. S2D), suggesting a critical role of Irp2 in pathogen clearance. To further confirm the results, we incubated BMDM and peritoneal macrophages with E. coli for 15 min, removed the extracellular bacteria, and allowed an additional incubation for 60 min. The macrophages were then lysed, and the bacteria were counted on Luria-Bertani (LB)-agar plates. A higher number of bacterial colonies were derived from Irp2^−/− BMDM (371 ± 12 versus 68 ± 8, P < 0.0001) and peritoneal macrophages (498 ± 59 versus 32 ± 13, P < 0.0001) than that from the macrophages isolated from WT mice (SI Appendix, Fig. S2 E and F), supporting the crucial role of Irp2 in the immune function of macrophages for pathogen clearance.

Fig. 3.

The increased bacterial load and decreased bacterial clearance of Irp2^−/− macrophages were associated with macrophage defective lysosomal biogenesis and acidification when infection by E. coli. (A) Confocal images of BMDM and peritoneal macrophages that were infected with eGFP-expressing (E. coli-eGFP) or mCherry-expressing E. coli (E. coli-mCherry) with MOI = 50 for 1 h. E. coli that were not phagocytosed were removed by gentamicin (200 μg/mL for 30 min). (B) Quantification of fluorescence, representing E. coli-eGFP and E. coli-mCherry in BMDM and peritoneal macrophages. (C and D) Protein levels of lysosomal markers Lamp1 and Ctsb in BMDM (C) and peritoneal macrophages (D) after incubation with E. coli with MOI = 10 for 90 min. (E) Lamp1 expression levels in BMDM and peritoneal macrophages of WT and Irp2^−/− mutant as revealed by immunofluorescence assays. (F) Quantification for E. (G) Lamp1 expression levels in BMDM and peritoneal macrophages of WT and Irp2^−/− mice post–E. coli infection, revealed by immunofluorescence assays. MOI = 50 for 1 h. (H) Quantification for G. (I) Lysosomal acidification in BMDM and peritoneal macrophages as indicated by pH-dependent pHrodo fluorescence. (J) The quantification of the fluorescence intensity of pHrodo for I. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar, 20 μm.

Macrophages and neutrophils are both critical phagocytes in the innate immune response to bacterial pathogens (24). Neutrophils, also expressing LysM and are myeloid derived, make up approximately 40 to 60% of white blood cells. Similarly, we explored the phagocytosis and bacteria-killing function of neutrophils. The results indicated no significant difference between WT and Irp2-deficient neutrophils after bacterial infection for 1 h and 4 h (SI Appendix, Fig. S2 G and H), suggesting that the bacterial-killing function of neutrophils was independent of Irp2. We therefore focused on Irp2 function in macrophages for the subsequent experiments.

Lysosomal Biogenesis and Acidification Are Defective in Irp2^−/− Macrophages.

Phagolysosome trafficking is an important innate defense mechanism that removes microbes by delivering them to lysosomes. Acidic lysosomes must fuse to the pathogen-entrapped phagosomes in macrophages, a crucial step for pathogen clearance. To detect whether Irp2 deficiency affects the fusion of lysosomes with phagosomes, we examined the colocalization of eGFP-E. coli with lysosome-associated membrane protein 1 (Lamp1), a typical lysosomal marker. As expected, bacteria were found to be associated with Lamp1-positive staining. However, no alteration of this association was observed in the absence or presence of Irp2 in both BMDM and peritoneal macrophages (SI Appendix, Fig. S3 A and B). Since more bacteria survived in Irp2^−/− BMDM and peritoneal macrophages (Fig. 3A), another explanation should exist of why Irp2 is required for the lysosomal function for bacteria clearance.

Lysosomal function relies on prior organelle biogenesis and maturation/acidification (25, 26). We thus wondered whether lysosomal biogenesis and maturation were affected by Irp2 deficiency. Lamp1 and a lysosomal cysteine protease cathepsin B (Ctsb), two lysosomal constitutive components, were determined prior to and post E. coli infection. Immunoblot results showed that Lamp1 and Ctsb protein levels decreased in Irp2^−/− macrophages compared to their WT counterparts, regardless of the bacterial infection (Fig. 3 C and D, quantified in SI Appendix, Fig. S3 C and D). Lamp1 immunofluorescence data also corroborated this lysosomal marker’s diminished presence in Irp2^−/− BMDM and peritoneal macrophages prior to infection (Fig. 3 E and F) and postinfection with E. coli (Fig. 3 G and H). The maturation of Ctsb into an active form requires lysosomal acidification. We measured the extent of lysosomal acidification using pHrodo and lysotracker, two reagents that accumulate in lysosomes through a pH-dependent mechanism. The fluorescence intensity of pHrodo was attenuated in Irp2^−/− BMDM and peritoneal macrophages (Fig. 3 I and J), and lysotracker’s fluorescence intensity also decreased in Irp2^−/− macrophages prior to and post bacterial infection (SI Appendix, Fig. S3 D–G). The results indicate that lysosomal biogenesis and acidification competence were attenuated in Irp2^−/− macrophages.

Irp2 Deletion Down-Regulates Transcript Levels and Nuclear Translocation of Transcription Factor EB (Tfeb) in Macrophages.

Since transcription factors of the MiTF/TFE family play a pivotal role in the regulation of lysosomal function and biogenesis (27), we examined the expression of the three family members, Tfeb (the basic helix–loop–helix leucine zipper transcription factor EB), Tfe3 (transcription factor binding to Immunoglobulin Heavy Constant Mu enhancer 3), Mitf (melanocyte inducing transcription factor) and a subunit of H(+)-ATPases, Atp6v0e2 (ATPase H+ transporting V0 subunit e2), an essential proton pump component. We found diminished mRNA levels in the liver of Irp2^−/− mice of Tfeb and Mitf to approximately 50% of their WT levels, not so for Tfe3 and Atp6v0e2 (Fig. 4A). All four genes expressed significantly less in the Irp2^−/− mutant than in WT following DSS treatment (Fig. 4A). Notably, the DSS treatment caused a rough 15-fold and 27-fold induction of Tfe3 and Atp6v0e2 in WT livers, whereas this response was entirely blunted in livers of Irp2^−/− mice. To examine whether a similar response was expected in macrophages, as observed in the liver, we constructed Irp2 KO RAW264.7 cells (RAW^Irp2KO, SI Appendix, Fig. S4A). In agreement with previous studies (28), Irp2-KO up-regulated FtH and hypoxia-inducible factor 2α (Hif2α) protein levels (SI Appendix, Fig. S4B). Consistent with the results in Irp2^−/−liver, Tfeb mRNA expression was also reduced in RAW^Irp2KO cells prior to and postinfection with E. coli, whereas the blunted response of Tfe3 and Atp6v0e2 postinfection was also observed (Fig. 4B), although these genes were not dramatically induced in cell culture. We then detected the protein levels of Tfeb and its downstream proteins, Lamp1 and Ctsb. Surprisingly, Tfeb protein levels were mildly higher in Irp2^−/− liver and RAW^Irp2KO cells before DSS treatment or prior to E. coli infection (Fig. 4 C and D). However, protein levels of Lamp1 and Ctsb were both diminished in mutants compared to WT prior to DSS treatment or E. coli infection (Fig. 4 C and D). Tfeb exerts its function via translocating into the nucleus, so we wondered whether the nuclear translocation was inhibited in Irp2 mutants. Indeed, the nuclear fraction of Tfeb was significantly reduced, and cytosolic Tfeb was increased considerably in RAW^Irp2KO cells (Fig. 4E). As expected, Tfeb expression increased in the liver following DSS treatment and in RAW264.7 cells after E. coli infection, whereas the changes in Lamp1 and Ctsb were found only in WT, not in the Irp2 mutants (Fig. 4 C and D, quantified in SI Appendix, Fig. S4 C and D). Consistently, the nuclear fraction of Tfeb was less in Irp2 mutant than in WT (Fig. 4 E and F). To verify the conclusion further, we observed the localization of Tfeb in RAW^Irp2KO cells by confocal microscopy. Similarly, Tfeb was more localized in the nucleus in WT than in Irp2 mutants, regardless of the infectious state (SI Appendix, Fig. S4 E and F). These results indicated that Irp2 ablation impairs nuclear localization of transcription factor Tfeb and the expression of its target genes related to lysosomal biogenesis and function.

Fig. 4.

Expression and nuclear localization of Tfeb decreases in Irp2^−/− liver tissue and macrophages. (A and B) mRNA levels of the Mitf family members, Tfeb, Tfe3, and Mitf, and their target gene Atp6v0e2 in the liver of mice treated with DSS water (A) and in macrophage-like RAW264.7 cells postinfection by E. coli (B). (C and D) Protein levels of Tfeb and its target protein Lamp1 and Ctsb in the liver of DSS-treated mice (C) and in RAW264.7 cells postinfection by E. coli (D). (E and F) Nuclear and cytosolic Tfeb in WT and RAW^Irp2KO cells before infection (E) and postinfection (F) with E. coli. MOI = 25 for 90 min. Results are expressed as means ± SD from three independent experiments. *P < 0.05, ***P < 0.001, ****P < 0.0001; ns: no significance.

Irp2-Deficiency Product Lactate Inhibits Tfeb Nuclear Localization Thereby Suppressing Lysosomal Biogenesis and Function in RAW264.7 Cells.

Our previous study demonstrated increased lactate production in Irp2^−/−murine embryonic fibroblasts (MEFs) (20) and Irp2^−/−tissues (29). Recently, it has been suggested that lactate production is coupled with the decrease in nuclear localization of Tfeb (30). We studied whether a similar phenomenon occurred in macrophages. First, we showed that lactic acid content in the medium increased in RAW^Irp2KO cells compared to WT cells and increased further after E. coli infection. In contrast, lactic acid remained constant in WT cells post E. coli infection (Fig. 5A). The increase in extracellular lactic acid could be reversed by treatment with 20 mM sodium oxamate (OXA), an inhibitor of lactic dehydrogenase A (LdhA), regardless of E. coli infection (SI Appendix, Fig. S5A).

Fig. 5.

Irp2 deficiency-induced lactate suppresses lysosomal biogenesis and function in RAW264.7 cells by inhibiting Tfeb nuclear localization. (A) Lactate contents in WT and Irp2-deficient mutant culture medium prior to and post–E. coli infection. (B) Retained E. coli-eGFP and E. coli-mCherry in WT and RAW^Irp2KO cells pretreated with LdhA-specific inhibitor OXA (20 mM) for 30 min. (C and D) Western blots showing Tfeb and its target proteins following treatment with 5 mM lactic acid (LAC) prior to C and post–E. coli infection (D). (E and F) Protein levels of Tfeb and its target proteins posttreatment with 20 mM OXA prior to E and post–E. coli infection (F). (G and H) Nuclear and cytosolic Tfeb protein levels in WT treated with 5 mM lactic acid (G) or in Irp2 KO treated with 20 mM OXA (H) followed by E. coli infection. Results are expressed as means ± SD from three independent experiments. **P < 0.01, ***P < 0.001; ns: no significance.

We then tested the effects of OXA treatment on the immune function of macrophages against E. coli infection. The bacterial load was reduced in RAW^Irp2KO after OXA treatment (Fig. 5B). Then, we measured the effects of lactic acid on the protein levels of Tfeb, Lamp1, and Ctsb. We observed an increase in Tfeb after lactic acid treatment. On the contrary, its target protein levels of Lamp1 and Ctsb decreased no matter whether the cells were infected with E. coli or not (Fig. 5 C and D; for parallel measurements of lactic acid content see SI Appendix, Fig. 5B; protein levels quantified in SI Appendix, Fig. S5 C and D). To further demonstrate the regulatory role of lactic acid on Tfeb and its target genes, we treated WT and Irp2 mutant cells with OXA to inhibit the production of lactic acid. After treatment, LdhA expression was reduced, and surprisingly, Tfeb expression decreased while Lamp1 and Ctsb were significantly increased, particularly in Irp2 mutant cells (Fig. 5 E and F, quantified in SI Appendix, Fig. S5 E and F). Again, we prepared nuclear fractions and detected Tfeb nuclear localization after lactic acid or OXA treatment. Western blot results of cell fractions showed less nuclear and more cytosolic localization of Tfeb after lactic acid treatment in WT cells (Fig. 5G, quantified in SI Appendix, Fig. S5G). In contrast, OXA treatment reversed the Tfeb localization in the Irp2 mutant (Fig. 5H, quantified in SI Appendix, Fig. S5H). The results indicated that the Irp2 ablation-mediated induction of lactic acid play a crucial role in lysosomal biogenesis and function.

Hif2α Inhibition Reverses the Bacterial Load in Irp2 Deficient Cells and Mice by Reducing Lactic Acid Production.

Our previous studies demonstrated that Irp2 deficiency stabilizes Hif2α and increases lactic acid production in MEFs (20) and in the nervous system tissues of mice (29). Consistently, levels of Hif2α were elevated in the liver of Irp2^−/− mice and further increased after DSS treatment (Fig. 6A, quantified in SI Appendix, Fig. S6A). We wondered whether the expression of Hif2α was stabilized and lactic acid production increased in RAW^Irp2KO cells and showed that this is also the case (Fig. 6B, quantified in SI Appendix, Fig. S6B), supporting the idea that high production of lactic acid is derived from increased LdhA through its upregulation of Hif2α in Irp2-depleted cells (29). Then, we treated RAW^Irp2KO cells with PT2385, a compound that selectively disrupts the heterodimerization of Hif2α with Hif1β. The bacterial load was significantly lessened in Irp2-depletion macrophages after 1 h infection by E. coli with PT2385 treatment (Fig. 6C, quantified in SI Appendix, Fig. S6C). Accordingly, the lactic acid content was reversed significantly in the Irp2 mutant after PT2385 treatment regardless of infection (SI Appendix, Fig. S6D). PT2385 treatment increased the protein levels of Lamp1 and Ctsb under both infection and noninfection conditions (SI Appendix, Fig. S6 E–G). Accordingly, Tfeb nuclear localization increased significantly at the expense of Tfeb cytoplasmic localization following PT2385 treatment (SI Appendix, Fig. S6 H–J). Similarly, we isolated the BMDM and peritoneal macrophages from Irp2^flox/flox and Irp2^ΔLysM mice to measure the mRNA expression of Tfeb, Tfe3, and Atp6v0e2, the lactic content, and levels of Hif2α and lysosome-related proteins before and after bacterial infection (SI Appendix, Fig. S7 A–D). All the results were consistent with those already described from RAW^Irp2KO cells, adding evidence for the crucial role of Irp2 in the lysosomal function of macrophages through Hif2α and LdhA expression to control lactic acid production.

Fig. 6.

Irp2 restrains Hif2α expression, promoting lysosomal biogenesis and function in the macrophage’s immune responses against bacterial infection. (A and B) Hif2α and LdhA protein levels in WT and Irp2^−/− liver after DSS modeling (A) and in WT and RAW^Irp2KO cells after E. coli infection (B), revealed by immunoblotting assays. (C) Retained E. coli-eGFP and E. coli-mCherry in WT and RAW^Irp2KO cells, pretreated with Hif2 specific inhibitor PT2385. (D) Diagram of PT2385 administration during the treatment protocol. (E) Body weight changes following administration of PT2385 (initially n = 30). (F) Survival plot of the animals during the protocol’s duration (n = 30). (G) Gross appearance of representative livers from Irp2^ΔLysM mice after treatment with PT2385 in the DSS-induced IBD model. White arrows indicate the sites of abscesses. (H) H&E staining of the liver. (I) Gram-positive staining of the liver. (J) Bacterial growth from the liver, portal vein blood, and peripheral blood on blood agar plates. n = 5/group if not specified. Data are presented as means ± SD. ****P < 0.0001.

To further test these conclusions in vivo, we treated Irp2^ΔLysM mice with PT2385 every other day for 2 wk and then exposed the animals to 2.5% DSS treatment (Fig. 6D). Body weight loss was significantly reversed after PT2385 treatment in Irp2^ΔLysM(Fig. 6E), and survival rate improved considerably (Fig. 6F). Importantly, the gross pus-filled appearance of liver abscesses disappeared after PT2385 treatment in Irp2^ΔLysMmice (Fig. 6G); H&E and gram-positive staining indicated the liver abscesses were attenuated after treatment (Fig. 6 H and I). The blood agar cultures further verified the protective effects of PT2385 from pathogen infection in Irp2^ΔLysMmice (Fig. 6J, quantified in SI Appendix, Fig. S7E). Thus, the in vivo evidence also supports our conclusion that Irp2 functions in macrophage immunity as a negative regulator of Hif2α. Furthermore, pharmacologic prevention of Hif2 activation is protective against the formation of liver abscesses, and this is because lactic acid production is inhibited and Tfeb nuclear translocation permits normal lysosomal biogenesis and function.

IRP2 Expression Is Down-Regulated in the Intestine of Crohn’s Disease Patients.

We have demonstrated that the liver abscesses result from intestinal injury in Irp2-ablated mice and is accompanied by Tfeb subcellular mislocalization in macrophages. To examine the expression of IRP2 and TFEB in patients with Crohn’s disease, we collected resected tissue samples. The mRNA levels of IRP2 and TFEB were significantly reduced in lesion regions compared with nonlesion parts of the intestine from the same patients (Fig. 7A). IRP2 protein levels were reduced, while TFEB increased in lesion regions despite the loss of encoding mRNA (Fig. 7 B and C), matching well with the observations in mice and cell culture. Immunoblotting assays consistently confirmed the decrease in IRP2 and increase in TFEB protein levels, and Hif2α and LdhA were increasingly accumulating in lesion regions (Fig. 7D, quantified on the Right). To expand the relevance of these observations to a larger set of clinical samples, we searched the GEO database of ulcerative colitis and Crohn’s disease. The extracted data revealed a significant reduction of IRP2 and TFEB mRNA expression in Crohn’s disease and ulcerative colitis, compared with the normal tissues (SI Appendix, Fig. S7F). Thus, the clinical data from our collection and data from clinical databases agree with that IRP2 insufficiency might be a high-risk clinical factor in developing ulcerative colitis and Crohn’s disease, potentially also in developing PLAs in the absence of protective interventions.

Fig. 7.

IRP2 expression is reduced in clinical samples from patients with Crohn’s disease or Ulcerative colitis. (A) mRNA levels of IRP2 and TFEB in nonlesion and lesion sections of Crohn’s disease patients (n = 10 to 20). (B) Immunohistochemical staining assays of IRP2 and TFEB protein in nonlesion and lesion sections of Crohn’s patients. (C) Quantification data of B by the index of mean optical density (MOD). (D) Protein levels of IRP2, HIF2α, LDHA, and TFEB in nonlesion and lesion regions, revealed by immunoblotting assays (Left) and quantified in the Right panel. N: nonlesion; P: patient; n = 6. *an unspecific band. GAPDH and Ponceau S staining: internal controls for total proteins. Results are expressed as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Discussion

In this study, we demonstrated that Irp2 is required in macrophages to support a functional immune response (Fig. 8). Deletion of Irp2 globally or LysM-specifically caused the appearance of liver abscesses in the DSS-induced IBD model. Following our previous study, where we showed that Irp2 deficiency induces Hif2α, which turns on LdhA expression, resulting in more lactic acid production (20), we found here that the increase in lactic acid retains Tfeb in the cytosol, rather than its functional presence in the nucleus. Consequently, macrophage function is compromised, resulting in pathogen infection and liver abscesses, a lethal condition both in mice and in human patients.

Fig. 8.

A schematic model presenting IRP2 functions in macrophages leading to prevention from developing liver abscesses in the context of intestinal barrier injury. Irp2 preserves lysosomal biogenesis and function by regulating Tfeb nuclear localization in macrophages. Irp2 deficiency up-regulates Hif2a expression, which induces lactic acid production by targeting LdhA genes. The increased lactic acid mislocalizes Tfeb, a key transcription factor to regulate a series of lysosome genes. Thus, macrophage lysosomal function is compromised, giving the pathogens a chance to route into the portal vein from the injured intestinal barrier. The constant pathogen invasion and translocation into the liver will promote the development of liver abscesses. (The scheme is created in BioRender.com).

Decades ago, a few case reports revealed that PLAs are connected to various colonic diseases, such as IBDs (31–33), and more recently, additional cases have been reported (6, 10). It is believed that the mucosal barrier separating enteric bacteria from the circulation allows a route for bacteria invasion into the portal system, with subsequent hematogenous spread to the liver (11). Located in the subepithelial lamina propria close to vast numbers of luminal bacteria and antigenic stimuli, gut macrophages are part of the first-line defense mechanisms in protecting the mucosa against harmful pathogens (34). The liver functions as a second firewall through a slow-flowing vascular system that could harbor antigens without translocation toward lymph nodes (35). The Kupffer cells in the liver sinusoids are located in an optimal position to clear bacteria, while increased gut leakage exhausts the scavenging capacity of Kupffer cells (36). In this case, the number of Kupffer cells decreases, and the recruitment and infiltration of monocytes to the liver increase, suggesting the vital role of monocyte-derived macrophages in the gut–liver axis (37, 38). In our study, the DSS-induced liver abscess in both Irp2^−/− and Irp2^ΔLysM indicates the crucial role of Irp2 in monocyte-derived macrophages. We have not excluded the role of Irp2 in other cell types (including hepatocytes, Kupffer cells, or intestinal epithelial cells). Although mice with hepatocyte-specific Irp2 deficiency display normal red blood cell and plasma iron parameters (39), the responses of the hepatocytes to inflammation or pathogen infection are still unknown. In the scenario of this study, mice with DSS treatment or patients with IBD bear intestinal epithelial injury, and the intestinal barrier integrity is compromised, which is considered to contribute to IBD (40).

Lysosome-mediated clearance of pathogens by macrophages is critically defective in IBD (reviewed in refs. 12 and 41). TFEB, one of the MiT/TFE family members harboring the coordinated lysosomal expression and regulation network motif, regulates lysosomal biogenesis and function (42–44). It has been demonstrated that TFEB is linked to an increased risk of developing colitis (reviewed in ref. 45). In this study, we showed that the lysosomal function and competence of bacterial killing were compromised mainly due to the defective translocation of Tfeb and associated impaired lysosomal acidification in Irp2-deficient macrophages. Disruption of lysosomal acidity has been previously shown to trigger functional iron deficiency, resulting in Hif1/2 upregulation and mitochondrial dysfunction (46, 47). This mutual interaction suggests the strong physiologic coupling between lysosomal acidity and iron homeostasis, while Irp2 function reestablishes iron homeostasis ensuring lysosomal biogenesis through Tfeb activation (nuclear translocation). Although Irp2 deficiency causes liver iron accumulation (19, 20, 28, 39) (SI Appendix, Fig. S8 A and B), ferric iron is found in ferritin causing a functional iron deficiency, as demonstrated in fibroblast/macrophages (19, 20, 28, 39) (SI Appendix, Fig. S8 C and D). Thus, Hif2a is up-regulated and induces LdhA causing the production of lactic acid. Inhibition of Hif2 reverses all phenotypes. Additionally, to see whether iron addition on its own would reverse macrophage lysosomal biogenesis and function, we measured the related proteins, Lamp1 and Ctsb, and showed no improvement (SI Appendix, Fig. S8 E and F). This result also makes sense as Hif2a responds to both iron and oxygen. Irp2 is required for this dual response and regulates iron homeostasis and lysosome function in parallel.

As expected from prior work on Irp2 deficient mice (20, 29), lactic acid accumulated in Irp2-ablated macrophages, resulting from Hif2α (not Hif1α) mediated LdhA upregulation (Fig. 6 and SI Appendix, Fig. S9 A and B). Lactic acid has been found to block the interaction between Rheb-Guanosine triphosphate (GTP) and GTPase activating Tuberous sclerosis complex 1 and 2 (TSC1 and TSC2) to preserve the Rheb-GTP levels and activate mTORC1 signaling (48). mTORC1 directly phosphorylates TFEB to keep TFEB cytosolic localization (49) by physical interaction (50). Inactivation of mTORC1 makes TFEB unphosphorylated and lets TFEB enter the nucleus for lysosomal gene expression (49). In agreement with these findings, we propose that the observed loss of nuclear Tfeb localization results from Irp2 deficiency–induced lactate production. Indeed, blockage of lactate production increased Tfeb nuclear localization in Irp2-ablated macrophages. Inhibition of LDHA in macrophages was also shown to down-regulate proinflammatory cytokines and, therefore, exert anti-inflammatory effects (51), consistent with our observations.

A few studies have demonstrated that IRP1 down-regulates Hif2α through IRP–IRE binding within the 5′-UTR of the Hif2α mRNA (52, 53). However, Irp1 expression is either mild or compensatory elevated in Irp2 KO mice (29). Whether Irp1 or Irp2-dependent Hif2α regulation is cell-type specific is unknown. Due to the significant increase of Hif2α in Irp2 deficient macrophages and livers, we treated Irp2^−/− mice with PT2385 before IBD modeling. This treatment mitigated the weight loss and surprisingly rescued all the Irp2 ablation mice from liver abscess, supporting that Irp2 deficient-induced Hif2α might also be through the IRP–IRE binding pathway (54). Clinically, the IBD patient’s immunocompromised state likely contributes to the development of liver abscesses, and a gut–liver link was reported many times (6, 10, 32, 33). Infliximab, a chimeric anti-tumour-necrosis factor-α monoclonal antibody, is a commonly used drug for the treatment of IBD. However, infliximab may enhance liver damage that promotes bacterial growth, as it causes hepatotoxicity, autoimmune hepatocellular injury, and transaminitis in patients with ulcerative colitis (55). Hif2α inhibitors were reported beneficial for treating Crohn’s disease (56), supporting our findings here and suggesting their pharmaceutical use in IBD.

Irp2 deficiency in mice causes microcytic anemia and neurodegeneration due to functional cellular iron depletion (57–59), and Hif2 inhibition alleviates both manifestations (28, 29). This study expands the role of Irp2-Hif2 into the innate immunity relevant to lysosomal function, showing that Irp2 affects lysosomal biosynthesis and acidification of macrophages in defense against pathogen infection. Thus, Hif2a inhibitors likely have therapeutic applications in the treatment of IBD and its complications.

Materials and Methods

Participants.

This study included 20 patients with bowel resection for Crohn’s disease, aged 16 to 65. Participants were recruited from the Department of General Surgery, Jinling Hospital, Affiliated Hospital of Nanjing University Medical School in 2023 (SI Appendix, Table S1). Intestinal samples containing the injured segments and the resection margin were collected and separated into two parts: one snap-frozen and stored at −80 °C for further use and another fixed in 4% phosphate-buffered formalin and embedded in paraffin for histological analysis. This study complies with the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of Nanjing University Medical School. The IRB ensures that informed consent was properly obtained and all patients provided written informed consent. The study is conducted in accordance with ethical guidelines and regulations.

Animals and Genotyping.

All experiments with rodents were performed according to the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Animal Experimentation Administration of Nanjing University. The mice were fed a standard rodent diet, maintained in a constant 12-h light/dark cycle, and had access to food and water ad libitum. C57BL/6 Irp2^+/− mice were purchased from MMRRC at UC Davis, United States, Cat. No. 030490-MU, and Irp2^−/− mice were acquired by breeding Irp2^+/− heterozygous mice. To obtain Irp2^ΔLysM mice, mating between LysM-Cre mice and control Irp2^flox/flox mice (Cyagen Biosciences, Suzhou, China) was carried out. Mouse DNA was extracted following instructions (Mouse Tissue Lysis Kit, APExBIO). The primers for PCR genotyping are as follows: forward 5′-TTTACCATTTGGCCCCTGACAAC-3′; reverse 5′-TACAGCCAAAACATAGTCTAGAGGT-3′.

Reagents and Antibodies.

The information on the primary antibodies used in this work is as follows: We obtained anti-Cathepsin B (CTSB) (cat#31718) from Cell Signaling Technology Inc. (Shanghai, China). We obtained anti-CTSB (cat#31718), anti-LAMP1 (cat#67300-1-Ig), and anti-TFEB (cat#13372-1-AP) from Proteintech Group Inc. (Wuhan, China). We obtained anti-Actin Beta (cat#66009-1-Ig) and anti-hif2a (cat#ab109616) from Abcam. We obtained anti-ferritin and anti-IRP2, which were self-made and validated in previous studies (20, 60). Secondary antibodies purchased from Life Technologies included anti-mouse or anti-rabbit conjugated to horseradish peroxidase for Western blots and anti-rabbit conjugated to Alexa Fluor™ 594 and anti-mouse conjugated to Alexa Fluor™ 488 for confocal microscopy.

Other reagents used in this work include lactic acid from Sinopharm (Shanghai, China), sodium oxamate from MACKLIN, HIF2 inhibitor PT2385 from MedChemExpress (Shanghai, China), Lysotracker from Beyotime, and pHrodo from Life Technology.

DSS-Induced Liver Abscess Modeling and Drug Administration.

For the liver abscess model, 7- to 9-wk-old mice were treated with 2.5% DSS (Yeasen Biotech Co., Ltd., Shanghai, China) for a week, followed by a 3-d interval with water and an additional week with 2.5% DSS. After 17 d, mice were killed by avertin for about 5 min. Tissues and blood were collected immediately thereafter.

For drug intervention, mice were divided into four groups: Irp2^ΔLysM mice with or without DSS plus vehicle DMSO (Irp2^ΔLysM ±DSS+DMSO) and Irp2^ΔLysM mice with or without DSS plus PT2385 (Irp2^ΔLysM±DSS+PT2385, 0.4 mg/kg). PT2385 was dissolved in DMSO, diluted with normal saline, and injected intraperitoneally into 7 to 9-wk-old male mice for 2 wk and then treated with DSS (2.5%), same as for the DSS-induced liver abscess model. The vehicle DMSO or PT2385 was given every other day.

Cell Lines and Cell Culture.

The murine macrophage cell line RAW264.7 was obtained from the American Type Culture Collection and cultured in Dulbecco’s modified Eagle’s Medium (DMEM) (Biosharp) supplemented with 10% fetal bovine serum (FBS) and penicillin (100 U/mL), and streptomycin (100 mg/mL) at 37 °C in 5% CO₂ incubator.

Histological Assays and Gram-Staining.

H&E staining was performed as previously described (61). Tissue sections were treated with dewaxing and hydrating steps, followed by staining with hematoxylin for 5 min, then Eosin Y for 5 min. Sections of human intestine tissue were used to assess the protein levels of IRP2 and TFEB by immunohistochemical assays. The tissues were probed with primary antibodies followed by horseradish peroxidase-labeled secondary antibodies and staining with 3,3’-diaminobenzidine enhancement. The Gram-positive staining was operated according to the Gram-staining kit (Servicebio Technology Co. Ltd.). Images were captured under a light microscope (Olympus, Tokyo, Japan). The positive signal intensity was analyzed using NIH Image J software.

Western Blots.

Cells or tissues were harvested and washed with ice-cold phosphate-buffered saline (PBS) and lysed in lysis buffer as previously described (61). The protein concentrations of lysates were measured using a standard Branford assay. For Western blots, 15 to 60 µg of protein from the whole cell lysate was electrophoresed by sodium dodecyl sulfate polyacrylamide gel electrophoresis. The separated proteins were transferred from the acrylamide gel onto the nitrocellulose membrane probed with primary antibodies followed by horseradish peroxidase-labeled secondary antibodies according to the standard methodologies (61). The proteins were visualized using enhanced chemiluminescence and imaged by the digital Western blot imaging system (Tanon). Band intensity was quantified by NIH ImageJ software.

RT-qPCR.

Total RNA was extracted by TRIzol (Vazyme Biotech Co. Ltd., Nanjing, China) and then reverse-transcribed into complementary DNA by HiScript III RT SuperMix for qPCR (Vazyme Biotech Co. Ltd.). mRNA expression of Mitf/Tfeb family genes and downstream genes were measured by quantitative PCR with ChamQ SYBR qPCR Master Mix (Vazyme Biotech Co. Ltd.) using the primers shown in SI Appendix, Table S2. 18S rRNA was used as an internal control. The relative mRNA expression levels were calculated by the delta–delta C_T method.

Isolation of Primary Macrophages and Neutrophils.

BMDMs and peritoneal macrophages were isolated and cultured as described previously (62). Briefly, 7 to 9-wk-old mice were euthanized. Legs were dissected, and bone marrow was extracted from the tibia and femur bones after removing surrounding muscle under an aseptic condition. Bone marrow cells were collected and filtered using a 70 µM cell strainer. The cell suspension was centrifuged at 400× g for 5 min at room temperature, and the cell pellet was incubated in Ammonium-Chloride-Potassium lysing buffer (NH₄Cl 150 mM, KHCO₃ 10 mM, Ethylenediaminetetraacetic acid 0.1 mM pH7.4) for 5 min to remove red blood cells. To elicit the differentiation of bone marrow progenitor cells into BMDM, the cells were cultured in DMEM with colony-stimulating factor (Peprotech, Cranbury, NJ). Cells were cultured for 7 d for further experiments.

For peritoneal macrophage isolation, the mice were intraperitoneally injected with 4% starch broth (NaCl 0.5 g, beef extract 0.3 g, peptone 1.0 g, and starch 3.0 g in 100 mL of distilled H₂O) 3 d before the mice were killed. After anesthesia, the abdominal skin was carefully cut to 1 cm, and 5 to 8 mL PBS with 3% FBS was injected into the enterocele. After 10 min massage, the extract was centrifuged (400× g, 5 min). The sediment was then plated into plates for attachment or cryopreserved for further assays.

The mouse neutrophils from bone marrow were prepared by the methods described previously (63). Briefly, the mice were executed and bone marrow cells were collected. Cells suspension was centrifuged and then treated on a 78%, 69%, and 52% Percoll gradient. The neutrophils were obtained for further assays.

Gene KO by CRISPR/Cas9 in Cells.

The Irp2 targeting 20 bp guide RNA (gRNA) sequence were as follows: Forward: 5′-CACCGAGTTACTCTTACTTTACCAG-3′; Reverse: 5′-AAACCTGGTAAAGTAAGAGTAACTC-3′.

The gRNA sequences targeting the Irp2 gene were cloned into the vector expressing Cas9 nuclease to construct the CRISPR/Cas9 KO plasmid and transfected into RAW264.7 cells using a liposome transfection kit (Yeasen Biotech Co., Ltd.). After transfection, single clones (RAW^Irp2KO) are selected with 0.4 mg/mL puromycin, and the genomic DNA is extracted for PCR and sequencing to confirm the KO. Total RNA and proteins were extracted to validate the KO by qPCR and Western blot. The impact of Irp2 KO on the expression of iron metabolism-related genes was examined for functional validation.

Immunofluorescence Assays.

Cells were seeded on coverslips until 70% confluence. Bacteria-treated cells were then fixed with 4% formaldehyde solution for 10 min at room temperature. After fixation, cells were washed twice with PBS and incubated with 0.2% Triton X-100 for 10 min to destroy the cell membrane. Cells were then incubated with blocking buffer (1% Bovine Serum Albumin in PBS) for 60 min and subsequently incubated with primary antibody overnight. Afterward, slips were incubated with a secondary antibody for 3 h at room temperature. Images were captured by a confocal laser microscope (Olympus FV3000, Tokyo, Japan) and analyzed using NIH ImageJ software.

Macrophage and Neutrophil Phagocytosis Assays.

Macrophages were plated in 24-well plates. FITC-dextran (250 µg/mL) or E. coli expressing eGFP/mCherry were added to each well at a multiplicity of infection (MOI) of 10 to 50 and incubated at 37 °C for 15 min for phagocytosis assays and 60 min for clearance assays. Neutrophils were plated in 24-well plates and infected with E. coli expressing eGFP/mCherry at a MOI of 50 and incubated at 37 °C for 1 h for phagocytosis assays and 4 h for clearance assays. Subsequently, cells were washed 3 times with PBS and treated with gentamycin (100 to 200 μg/mL) at 37 °C for 30 min to remove extracellular bacteria if needed. The fluorescence intensity was measured by confocal laser microscopy. Bacterial colonies were counted after incubation at 37 °C for 24 h on LB plates.

For bacterial load assays, macrophages were lysed by adding 0.02% Triton X-100 for 10 min, and the lysate was smeared onto LB-agar. Bacterial colonies were counted after incubation at 37 °C for 24 h.

Subcellular Localization and Lysosomal Acidification assays.

Macrophages were stained with Lysotracker Red for 30 min, infected with E. coli-eGFP (MOI = 50) for 1 h, and then washed with PBS three times. Alternatively, they were stained with pHrodo-labeled E. coli for 1 h. The cells were imaged with laser scanning confocal microscopy.

For subcellular fractionation, the cytosolic and nuclear fractions were separated using a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Biotech.). Briefly, one million cells seeded in 60 mm dishes for 24 h were rinsed twice with ice-cold PBS, immediately incubated with cytoplasmic extraction buffer on ice for 15 min, and then harvested using a cell scraper. The samples were vortexed for 30 s, followed by centrifugation at 12,000× g for 10 min at 4 °C, and the supernatant containing cytosolic fraction was transferred to a new tube. The pellet was resuspended in a nuclear extraction buffer on ice for 30 min, followed by centrifugation at 16,000× g for 10 min at 4 °C. The resulting supernatant was collected as the nuclear fraction.

For the lysosomal acidification assay, macrophages were stained with Lysotracker for 1 h or pHrodo-labeled E. coli for 1 h, and the cells were imaged with a laser scanning confocal microscope.

Determination of Lactate Concentration.

Lactate concentration was measured using a lactate assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions.

Statistics.

Data were presented as the mean value ± SD from the mean. All the experiments were repeated more than three times independently. Student’s t test or one-way ANOVA was performed using GraphPad Prism 8. Significance was considered at P < 0.05.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

The RNA-seq data were cited from the Gene Expression Omnibus database under accession codes GSE9452 and GSE1710 (64, 65). Other study data are included in the article and/or SI Appendix.