Lactate promotes macrophage HMGB1 lactylation, acetylation, and exosomal release in polymicrobial sepsis

Abstract

High circulating levels of lactate and high mobility group box-1 (HMGB1) are associated with the severity and mortality of sepsis. However, it is unclear whether lactate could promote HMGB1 release during sepsis. The present study demonstrated a novel role of lactate in HMGB1 lactylation and acetylation in macrophages during polymicrobial sepsis. We found that macrophages can uptake extracellular lactate via monocarboxylate transporters (MCTs) to promote HMGB1 lactylation via a p300/CBP-dependent mechanism. We also observed that lactate stimulates HMGB1 acetylation by Hippo/YAP-mediated suppression of deacetylase SIRT1 and β-arrestin2-mediated recruitment of acetylases p300/CBP to the nucleus via G protein-coupled receptor 81 (GPR81). The lactylated/acetylated HMGB1 is released from macrophages via exosome secretion which increases endothelium permeability. In vivo reduction of lactate production and/or inhibition of GPR81-mediated signaling decreases circulating exosomal HMGB1 levels and improves survival outcome in polymicrobial sepsis. Our results provide the basis for targeting lactate/lactate-associated signaling to combat sepsis.

Fig. 1. Elevated lactate levels contribute to increased circulating exosomal HMGB1 levels in polymicrobial sepsis.

Lactate (0.5 g/kg body weight) was administrated through i.p. injection 6 h after CLP or sham surgery. To inhibit lactate production, sodium oxamate (OXA, 0.5 g/kg body weight) was i.p. injected 6 h before CLP or sham surgery. A Serum lactate levels were measured by Lactate Assay Kit (n = 5 for sham and CLP, n = 5 for sham + Lac, CLP + Lac, and sham + OXA, n = 4 for CLP + OXA, two-way ANOVA with Tukey’s test). B Serum HMGB1 levels among sham, CLP, sham + Lac, and CLP + Lac were assayed by western blot (n = 4 for sham, CLP and sham +Lac, n = 5 for CLP + Lac, two-way ANOVA with Tukey’s test). C The survival rate among CLP, CLP + Lac and CLP + OXA mice was compared by Kaplan–Meier test (n = 22 for CLP, n = 21 for CLP + Lac, and CLP + OXA). D Serum HMGB1 levels among sham, CLP, sham + OXA, and CLP + OXA were assayed by western blot (n = 3 for sham + OXA, n = 4 for sham, CLP, and CLP + OXA, two-way ANOVA with Tukey’s test). E HMGB1 levels in untouched serum and exosome-depleted serum of sham, CLP, Lac, and CLP + Lac were measured by ELISA (n = 4, two-way ANOVA with Tukey’s test). F Exosomes were isolated from the serum of sham, CLP, sham + Lac, and CLP + Lac mice. Exosome lysates were analyzed by western blot using antibodies against HMGB1, HSP70, and calnexin (n = 6 for each group, two-way ANOVA with Tukey’s test). G Exosomes were isolated from the serum of sham, CLP, OXA + sham, and OXA + CLP mice. Exosome lysates were analyzed by western blot using antibodies against HMGB1, HSP70 and calnexin (n = 3 for sham + OXA, n = 4 CLP + OXA, n = 6 for sham and CLP, two-way ANOVA with Tukey’s test). Values are mean ± SD. Lac lactic acid, OXA oxamate, CLP cecal ligation and puncture.

Fig. 2. Lactate induces HMGB1 cytosol accumulation and release via exosome secretion in macrophages.

A RAW 264.7 cells were pretreated with lactate (10 mM) for 30 min before LPS (500 ng/mL) stimulation for 6 h. Cytosol and nuclear HMGB1 levels were measured by western blot (n = 3, two-way ANOVA with Tukey’s test). B RAW 264.7 cells were pretreated with oxamate (20 mM) for 2 h before LPS (500 ng/ml) stimulation for 6 h. Cytosol and nuclear HMGB1 levels were measured by western blot (n = 3, two-way ANOVA with Tukey’s test). C Representative immunofluorescent staining images of RAW 264.7 cells treated with vehicle or lactate (10 mM) for 6 h show increased accumulation of HMGB1 in the cytoplasm (indicated by white arrows) of lactate-treated cells (Scale bar, 10 µm). D 200 µg of protein lysates were precipitated with anti-HMGB1 antibody followed by immunoblotting with anti-Lamp1 antibody shows lactate (lactic acid or sodium lactate) increased the interaction between HMGB1 and Lamp1 in RAW 264.7 cells (n = 3, t test). E and F RAW 264.7 cells were stimulated with lactic acid (E) or sodium lactate (F) for 24 h and macrophage-derived exosomes were isolated from the supernatant to examine the presence of HMGB1 protein by western blot (n = 3, t test). Values are mean ± SD. OXA sodium oxamate, Lac lactic acid, LacNa, sodium lactate.

Fig. 3. Lactate directly induces HMGB1 lactylation (Klac) in macrophages.

A RAW 264.7 cells were pretreated with oxamate (20 mM) for 30 min followed by LPS (500 ng/mL) stimulation for 24 h. 200 µg of protein lysates were precipitated with anti-HMGB1 antibody followed by immunoblotting with anti-Klac antibody (n = 3, two-way ANOVA with Tukey’s test). B Sepsis was induced by CLP surgery followed by i.p. administration of lactate or vehicle. Peritoneal macrophages were prepared 24 h after CLP or sham surgery. Peritoneal macrophages in each group were isolated and cell lysates were precipitated with anti-HMGB1 antibody followed by immunoblotting with anti-Klac antibody (n = 3, two-way ANOVA with Tukey’s test). C 200 µg of protein lysates of lactic acid or sodium lactate-treated RAW 264.7 cells were precipitated with anti-HMGB1 antibody followed by immunoblotting with anti-Klac antibody showing lactate-induced Klac in HMGB1 immunocomplex (n = 3, two-way ANOVA with Tukey’s test). D RAW 264.7 cells were treated with CHC (3 mM) or DMSO for 2 h before lactate (10 mM) addition for another 24 h. HMGB1 (green) and Klac (red) co-localization was examined by confocal microscope (Scale bar, 10 µm). Nucleus was indicated by DAPI (blue) staining. Co-localization analysis was performed using Zeiss Zen microscope software. E and F RAW 264.7 cells were treated with C646 (5 µM) or vehicle for 2 h followed by lactate treatment for 24 h. Cell lysates were examined for Klac levels by western blot D or precipitated with anti-HMGB1 antibody and probed for Klac levels E (n = 3 for each group, two-way ANOVA with Tukey’s test). G CBP and p300 were silenced by transfection with specific siRNAs (40 nM) for overnight followed by lactate (10 mM) stimulation for 24 h. 200 µg of protein lysates were precipitated with anti-HMGB1 antibody followed by immunoblotting with anti-Klac antibody (n = 3, two-way ANOVA with Tukey’s test). H RAW 264.7 cells were pretreated with CHC (3 mM) or vehicle for 2 h followed by lactate (10 mM) addition for 24 h. Intracellular lactate levels were measured by Lactate Assay Kit. Not treated cells were used as control (n = 3 for each group, two-way ANOVA with Tukey’s test). I–K RAW 264.7 cells were treated with CHC (3 mM) or vehicle (DMSO) for 2 h followed by lactate treatment for 24 h. HMGB1 Klac levels were assayed by immunoprecipitation with anti-HMGB1 antibody followed by immunoblotting with anti-Klac antibody I. Interaction between HMGB1 and CD63 was assayed by immunoprecipitation with anti-HMGB1 antibody and probed for co-precipitation of CD63 in HMGB1 immunocomplex J. Exosomes were isolated from the supernatant and exosomal HMGB1 protein levels were examined by western blot K (n = 3 for each group, two-way ANOVA with Tukey’s test). Values are mean ± SD. Lac lactic acid, OXA sodium oxamate, LacNa sodium lactate, Klac lysine lactylation.

Fig. 4. Lactate induces HMGB1 acetylation.

A Cell lysates of lactate-treated RAW 264.7 macrophages were immunoprecipitated with anti-HMGB1 antibody and probed for acetylation (Kac) levels by western blot (n = 3 for each group, t test). B Western blot shows that lactate-induced HMGB1 acetylation at lysine 12 (K12) and 29 (K29) residues (n = 3 for each group, t test). C Both lactic acid and sodium lactate at 10 mM induced HMGB1 acetylation (n = 3 for each group, t test). D Oxamate treatment decreased HMGB1 acetylation levels in a dose-dependent manner in RAW 264.7 cells (n = 3 for each group, one-way ANOVA with Tukey’s test). E–G BMDMs and peritoneal macrophages were stimulated with lactate for 6 h and HMGB1 acetylation levels were assayed. Lactate-induced HMGB1 acetylation in BMDMs (E, indicated by white arrows). Immunofluorescent staining F and western blot analysis G show increased HMGB1 acetylation in lactate-treated peritoneal macrophages. H Sepsis was induced by CLP surgery followed by i.p. administration of lactate or vehicle. Peritoneal macrophages were isolated 24 h after CLP or sham surgery. Western blot was performed to detect HMGB1 acetylation levels in isolated peritoneal macrophages (n = 3, two-way ANOVA with Turkey’s test). Scale bar, 10 µm. Lac lactate, LacNa sodium lactate, OXA sodium oxamate.

Fig. 5. Lactate promotes HMGB1 acetylation through YAP-mediated suppression of deacetylase SIRT1 in macrophages.

A Decreased expression of deacetylase SIRT1 in both cytosol and nuclear fractions of RAW 264.7 cells stimulated with lactate for 6 h (n = 3, t test). B Suppression of SIRT1 deacetylase activity by EX527 (10 μM) increased HMGB1 acetylation in 264.7 cells (n = 3, t test). C Activation of SIRT1 deacetylase by its activator SRT2183 (10 μM) decreased acetylated-HMGB1 expression in RAW 264.7 cells (n = 3, t test). D RAW 264.7 cells were transduced with Ad-GFP or Ad-SIRT1 overnight. Expression of SIRT1 and acetylated-HMGB1 (K12) were assessed by western blot (n = 3, t test). E RAW 264.7 cells were stimulated with lactate (10 mM) for 6 h and the cytosol expression of YAP, p-YAP (Ser127), p-LATS1 (Thr1079), LATS1, and nuclear expression of YAP were assayed by western blot (n = 3 for each group, t test). F Western blot analysis of YAP, SIRT1 and acetylated-HMGB1 (K12) expressions in wild type (WT) and YAP knockout (YAP^−/−) peritoneal macrophages (n = 3, t test). G CiiiDER predicts a putative TEAD binding site locating on the promoter region of mouse SIRT1. H ChIP assay of the relative enrichment of TEAD4 on the promoter region of mouse SIRT1 (n = 3, two-way ANOVA with Tukey’s test). Values are mean ± SD. Lac lactic acid.

Fig. 6. Lactate promotes HMGB1 acetylation via β-arrestin2-dependent recruitment of acetylases p300/CBP in macrophages.

A Lactate upregulated p300 mRNA level in RAW 264.7 cells (n = 3, t test). B Western blot shows that lactate increased the expression of protein CBP and p300 in RAW 264.7 cells (n = 3 for each group, t test). C Representative images of expression and localization of CBP (red) and p300 (green) in lactate-treated RAW 264. 7 cells. The nucleus was stained with DAPI (blue). Co-localization analysis was performed using Zeiss Zen microscope software. (Scale bar, 10 µm). D Inhibited activity of acetylases p300/CBP by C646 suppressed lactate-induced HMGB1 acetylation (n = 3, two-way ANOVA with Turkey’s test). E Western blot shows that lactate-increased β-arrestin2 (β-arr2), but not β-arrestin1 (β-arr1), nuclear expression in the nucleus of RAW 264.7 cells (n = 3 for each group, t test). F RAW 264.7 cells were transfected with control siRNA (siR-Con) or β-arrestin2 specific siRNA (siR-β-arr2) overnight before lactate treatment. Expression of cytosol acetylated-HMGB1 (k12) and acetylated-HMGB1 (K29), and expression of nuclear β-arrestin2, p300 and CBP were examined by western blot (n = 4 for K12, n = 3 for K29, t test). Values are mean ± SD. Lac lactic acid, β-arr β-arrestin.

Fig. 7. Lactate-induced HMGB1 acetylation is mediated by GPR81 signaling.

A–D RAW 264.7 cells were treated with 3-OBA (5 mM) for 2 h followed by lactate administration for 6 h. Blockage of GPR81 by 3-OBA attenuated lactate-increased nuclear expression of CBP (n = 5), p300 (n = 3) and β-arrestin2 (n = 3) A. Blockage of GPR81 by 3-OBA attenuated lactate-induced phosphorylation of YAP (n = 3) and LATS1 (n = 3), and lactate-suppressed nuclear YAP expression (n = 3) and SIRT1 expression (n = 3) B. Lactate-promoted acetylation of HMGB1 was attenuated by 3-OBA pretreatment C. Immunofluorescent staining shows that blockage of GPR81 by 3-OBA diminished lactate-induced HMGB1 acetylation (green) and its cytoplasmic accumulation (indicated by white arrows) D. E RAW 264.7 cells were pretreated with 3-OBA (5 mM) for 2 h followed by lactate treatment for 24 h. Exosomes were isolated from the culture medium and exosomal HMGB1 protein levels were examined by western blot. (n = 3) F 3-OBA (0.5 g/kg body weight) was i.p. injected 6 h prior to CLP or sham surgery. Serum was collected 24 h following CLP for exosome isolation. Protein levels of HMGB1 in serum exosomes of sham, CLP, 3-OBA + sham and 3-OBA + CLP mice were examined by western blot (n = 4). Values are mean ± SD. Two-way ANOVA with Turkey’s test was performed. Scale bar, 10 µm. Lac lactate.

Fig. 8. Macrophage-derived exosomal HMGB1 induces endothelial dysfunction.

A RAW 264.7 cells were stimulated with lactate (10 mM) for 24 h and exosomes were isolated from the supernatant. HUVECs were incubated with 2.5 µg/mL exosomes, derived from either control macrophages or lactate-treated macrophages, for 6 h. Western blot shows that exosomes derived from lactate-treated macrophages reduced the expression of VE-cadherin and Claudin5, while increased ICAM1 expression, in HUVECs, as compared with HUVECs incubated with exosomes derived from control cells (n = 3 for each group, t test). B HUVECs were seeded into the upper chambers of a transwell system. HUVECs monolayers were then treated with 2.5 µg/mL exosomes, derived from either control macrophages or lactate-treated macrophages, for 6 h followed by addition of FITC-dextran. Fluorescence was quantified in the lower chamber 5 min after administration of FITC-dextran. Exosomes derived from lactate-treated macrophages increased the endothelial permeability (n = 4 for each group, t test). C RAW 264.7 cells were transfected with HMGB1 siRNA or control siRNA for 24 h. HMGB1 expression was examined by western blot (n = 3, t test). D RAW 264.7 cells were transfected with HMGB1 siRNA or control siRNA for 24 h followed by lactate stimulation. Exosomes were collected from supernatants and added to endothelial cell culture. Expression of endothelial VE-cadherin, Claudin 5, and ICAM1 was assayed by western blot (n = 4 for VE-cadherin, n = 3 for claudin 5 and ICAM1, Two-way ANOVA with Turkey’s test). E Scheme of lactate-induced HMGB1 lactylation/acetylation and exosomal release during sepsis. During sepsis, macrophages can uptake lactate through monocarboxylate transporters (MCTs), which leads to HMGB1 lactylation in a p300/CBP-dependent mechanism (1a). In addition, extracellular lactate increases β-arrestin2-recruited acetylase p300/CBP nuclear translocation (1b) and suppresses LATS/YAP-mediated deactivation of deacetylation SIRT1 (1c), via GPR81-dependent signaling, resulting in increased acetylation of HMGB1. Lactylated/acetylated HMGB1 in turn, is translocated into lysosomes in the cytoplasm of macrophages (2) and released via exosome secretion (3) from macrophages. Secreted exosomal HMGB1 further disrupts endothelium barrier function (4) by decreasing VE-cadherin and claudin 5 expressions and increasing ICAM1 expression in endothelial cells. Values are mean ± SD. EXO exosome, Mɸ macrophage, Lac lactate.