Hepatocyte HSPA12A inhibits macrophage chemotaxis and activation to attenuate liver ischemia/reperfusion injury via suppressing glycolysis-mediated HMGB1 lactylation and secretion of hepatocytes

Abstract

Rationale: Liver ischemia-reperfusion (LI/R) injury is characterized by two interconnected phases: local ischemia that causes hepatic cell damage to release damage-associated molecular pattern (DAMPs), and DAMPs that recruit immune cells to elicit inflammatory cascade for further injury of hepatocytes. High-mobility group box 1 (HMGB1) is a representative DAMP. Studies in macrophages demonstrated that HMGB1 is secreted after lactylation during sepsis. However, whether lactylation mediates HMGB1 secretion from hepatocytes after LI/R is known. Heat shock protein A12A (HSPA12A) is an atypical member of HSP70 family. Methods: Gene expression was examined by microarray analysis and immunoblotting. The hepatic injury was analyzed using released ALT and AST activities assays. Hepatic macrophage chemotaxis was evaluated by Transwell chemotaxis assays. Inflammatory mediators were evaluated by immunoblotting. HMGB1 secretion was examined in exosomes or serum. HMGB1 lactylation was determined using immunoprecipitation and immunoblotting. Results: Here, we report that LI/R decreased HSPA12A expression in hepatocytes, while hepatocyte-specific HSPA12A overexpression attenuated LI/R-induced hepatic dysfunction and mortality of mice. We also noticed that hepatocyte HSPA12A overexpression suppressed macrophage chemotaxis to LI/R-exposed livers in vivo and to hypoxia/reoxygenation (H/R)-exposed hepatocytes in vitro. The LI/R-increased serum HMGB1 levels of mice and the H/R-increased HMGB1 lactylation and secretion levels of hepatocytes were also inhibited by hepatocyte HSPA12A overexpression. By contrast, HSPA12A knockout in hepatocytes promoted not only H/R-induced HMGB1 lactylation and secretion of hepatocytes but also the effects of H/R-hepatocytes on macrophage chemotaxis and inflammatory activation, while all these deleterious effects of HSPA12A knockout were reversed following hepatocyte HMGB1 knockdown. Further molecular analyses showed that HSPA12A overexpression reduced glycolysis-generated lactate, thus decreasing HMGB1 lactylation and secretion from hepatocytes, thereby inhibiting not only macrophage chemotaxis but also the subsequent inflammatory cascade, which ultimately protecting against LI/R injury. Conclusion: Taken together, these findings suggest that hepatocyte HSPA12A is a novel regulator that protects livers from LI/R injury by suppressing glycolysis-mediated HMGB1 lactylation and secretion from hepatocytes to inhibit macrophage chemotaxis and inflammatory activation. Therefore, targeting hepatocyte HSPA12A may have therapeutic potential in the management of LI/R injury in patients.

Keywords: Chemotaxis; Heat shock protein A12A (HSPA12A); Hepatocyte; High-mobility group box 1(HMGB1); Lactylation; Liver ischemia/reperfusion (LI/R); Macrophage.

Figure 1

Increased macrophage chemotaxis was accompanied with decreased hepatocyte HSPA12A expression following LI/R. A. Hsp mRNA data set from microarray analysis. Livers from sham and LI/R mice were applied to microarray analysis. Hsp mRNA data set from microarray was heat mapped and statistically analyzed. The red arrow indicates the decrease of hspa12a mRNA expression. n = 3/group. B. HSPA12A protein expression in livers. HSPA12A protein expression was examined in mouse livers after sham or LI/R. n = 4/group. C. HSPA12A protein expression in H/R hepatocytes. HSPA12A protein expression was examined in primary hepatocytes after normoxia or H/R. n = 3/group. D. Macrophage recruitment in livers. Immunostaining for macrophage marker F4/80 was performed on cryosections that prepared from mouse livers after LI/R. DAPI was used to stain nuclei. Scale bar = 50 μm. n = 5 /group. E. Macrophage chemotaxis towards hepatocytes. H/R- or normoxia-exposed primary hepatocytes were cocultured with primary macrophages in Transwell plates for 24 h. The macrophage chemotaxis towards hepatocytes was stained with crystal violet. n = 6/group. Data are mean ± SD, ** P < 0.01 and * P < 0.05 by Student's two-tailed unpaired t test.

Figure 2

H-Ki mice displayed attenuation of LI/R injury. A. Experimental protocol. LI/R was induced in mice. B. Mouse survival. Mice mortality was recorded after LI/R. **P < 0.01 by log-rank test, n = 68 in h-Ki group and n = 38 in WT group. C. Serum ALT and AST activities. Serum was collected from mice after sham or LI/R for measuring ALT and AST activities. Data are mean ± SD, ** P < 0.01 by two-way ANOVA followed by Tukey's test. n = 7/group. D. Histological examination. After sham or LI/R, liver tissues were collected, paraffin-embedded sectioned, and HE stained. Suzuki's scoring was performed and hepatic necrosis areas were measured to indicate histological injury. Scale bar = 200 μm. Data are mean ± SD, ** P < 0.01 by two-way ANOVA followed by Tukey's test. n = 8 for sham-WT group and n = 10 for all the other groups.

Figure 3

Hepatocyte HSPA12A inhibited macrophage chemotaxis after LI/R. A-B. Mouse experiments. LI/R was induced in mice (A). After LI/R, liver tissues were collected, cryosectioned, and immunestained for F4/80. DAPI was used to indicate nuclei (B). Scale bar = 50 μm. n = 5/group. C-E. Cell experiments. Primary hepatocytes with HSPA12A overexpression (Ki) or knockout (Ko) were isolated from livers of h-Ki or HSPA12A Ko mice. Hepatocytes from WT mice served as controls. HSPA12A expression was examined using immunoblotting (C). After exposed to H/R- or normoxia, hepatocytes were cocultured with primary macrophages in Transwell plates (D). After 24 h's coculture, macrophages chemotaxis toward hepatocytes was stained with crystal violet (E). Scale bar = 50 μm. n = 4/group. Data are mean ± SD, ** P < 0.01 and * P < 0.05 by two-way ANOVA followed by Tukey's test.

Figure 4

Hepatocyte HSPA12A paracrinally inhibited macrophage inflammatory activation. A-B. Mouse experiments. LI/R was induced in mice (A). Liver tissues were then collected to examine the indicated gene expression using immunoblotting (B). n = 4/group. C-D. Cell experiments. (C). After exposed to H/R or normoxia, culture medium of primary hepatocytes was collected as hepatocyte conditioned medium (CM) and then added to macrophages (D). After treated with hepatocyte CM for 24 h, macrophages were collected to examine the indicated gene expression using immunoblotting. n = 4/group. Data are mean ± SD, ** P < 0.01 and * P < 0.05 by two-way ANOVA followed by Tukey's test.

Figure 5

Hepatocyte HSPA12A inhibited HMGB1 secretion from hepatocytes. A-C. Mouse experiments. LI/R was induced in mice (A). Mouse serum was collected to examine HMGB1 protein levels using immunoblotting, and equal volume of serum separated on SDS-PAGE and stained with Ponceau S solution served as protein loading control (B). Also, liver tissues were collected, cytosolic fraction prepared, and immunoblotted for HMGB1 (C). n = 4/group (B) and n = 6/group (C). D-F. Hepatocyte experiments. H/R was induced in primary hepatocytes (D). After then, culture medium was collected, exosome isolated, and immunoblotted for HMGB1 (E). Blotting for HSP70 and Calnexin served as positive and negative markers, respectively (E). Also, primary hepatocytes were immunestained with HMGB1, and DAPI was used to stain nuclei (F). Scale bar = 5 μm. n = 3/group (E) and n = 4/group (F). Data are mean ± SD, ** P < 0.01 by two-way ANOVA followed by Tukey's test.

Figure 6

Hepatocyte HSPA12A modulated HMGB1 secretion to paracrinally regulate macrophage chemotaxis and activation. A. HMGB1 knockdown decreased HMGB1 exosomal secretion of hepatocytes. HMGB1 was knockdown in primary hepatocytes (Si-Hmgb1) and scramble RNA treatment served control. After exposed to H/R, culture medium was collected, exosome isolated, and immunoblotted for HMGB1. Blotting for HSP70 and Calnexin served as positive and negative markers, respectively. n = 3/group. B. Hepatocyte HMGB1 knockdown inhibited macrophage chemotaxis. After exposed to H/R, primary hepatocytes were cocultured with primary macrophages in Transwell plate. After coculture for 24 h, macrophage chemotaxis towards hepatocytes was stained with crystal violet. Scale bar = 50 μm. n = 4/group. C. Hepatocyte HMGB1 knockdown inhibited macrophage activation. After exposed to H/R, culture medium of primary hepatocytes was collected as hepatocyte conditioned medium (CM). The CM was then added to macrophages. After treated with hepatocyte CM for 24 h, macrophages were collected to examine the indicated gene expression using immunoblotting. n = 4/group. D. Hepatocyte HMGB1 knockdown decreased ALT and AST leakage when coculture with macrophages. Primary hepatocytes were exposed to H/R, followed by coculture with primary macrophages in Transwell plate. After coculture for 24 h, medium was collected for ALT and AST activity examination. n = 5/group. Data are mean ± SD, * P < 0.05 and ** P < 0.01 by two-way ANOVA followed by Tukey's test. ns, no significance.

Figure 7

HSPA12A modulated hepatocyte HMGB1 lactylation and exosomal secretion to paracrinally regulate macrophage chemotaxis. A. Klac-HMGB1 levels in hepatocytes. After exposed to H/R or normoxia, primary hepatocytes were immunoprecipitated with anti-HMGB1 antibody followed by immunoblotting for Klac and HMGB1. n = 3/group. B. Colocalization of Klac and HMGB1 in cytoplasm of hepatocytes. After exposed to H/R or normoxia, primary hepatocytes were immunestained with Klac and HMGB1. DAPI was used to counter stain nuclei. Note that the H/R-induced Klac-HMGB1 colocalization was promoted in cytoplasm of HSPA12A Ko hepatocytes. Scale bar = 10 μm. n = 3/group. C. C646 decreased HMGB1 lactylation levels. Primary hepatocytes were exposed to H/R in the presence or absence of C646. After then, hepatocytes were immunoprecipitated with anti-HMGB1 antibody followed by immunoblotting for Klac and HMGB1. n = 3/group. D. C646 decreased exosomal HMGB1 secretion. Primary hepatocytes were exposed to H/R in the presence or absence of C646. After then, culture medium was collected, exosome isolated, and immunoblotted for HMGB1. Blotting for HSP70 and Calnexin served as positive and negative markers, respectively. n = 3/group. E. C646 abolished the enhanced chemotactic effects of Ko hepatocytes on macrophage chemotaxis. Primary hepatocytes were exposed to H/R in the presence or absence of C646, followed by coculture with primary macrophages in Transwell plate. After coculture for 24 h, macrophage chemotaxis was crystal violet stained and quantified. Scale bar = 50 μm. n = 4/group. Data are mean ± SD, ** P < 0.01 by two-way ANOVA followed by Tukey's test. ns, no significance.

Figure 8

HSPA12A inhibited glycolysis-derived lactate to suppress HMGB1 lactylation and exosomal secretion of H/R-exposed hepatocytes. A. Extracellular lactate. After exposed to normoxia or H/R, culture medium of primary hepatocytes was collected for measuring lactate levels. n = 6/group. B. Expression of glycolysis-related genes. After exposed to normoxia or H/R, primary hepatocytes was collected for measuring the indicated gene expression using immunoblotting. n = 4/group. C. Brief scheme of glycolytic process and inhibition. Glucose is uptaken into cells, converted to pyruvate by enzymes (HK II, PFK and PKM2) and followed by converting to lactate by LDHA, and finally lactate is exported to extracellular space. For inhibiting lactate production, Oxamate was used to inhibit LDHA activity. D. Oxamate inhibited lactate production. Primary hepatocytes were exposed to H/R in the presence or absence of Oxamate. After then, culture medium was collected for measuring lactate levels. n = 6/group. E. Oxamate inhibited HMGB1 lactylation. Primary hepatocytes were exposed to H/R in the presence or absence of Oxamate. After then, hepatocytes were immunoprecipitated with anti-HMGB1 antibody followed by immunoblotting for Klac and HMGB1. n = 3/group. F. Oxamate inhibited exosomal HMGB1 secretion. Primary hepatocytes were exposed to H/R in the presence or absence of Oxamate. After then, exosomes were isolated from culture medium and immunoblotted for HMGB1. Blotting for HSP70 and Calnexin served as positive and negative markers, respectively. n = 3/group. G. Mechanistic scheme. By suppressing glycolysis-derived lactate production in hepatocytes, HSPA12A inhibits HMGB1 lactylation and exosomal secretion of hepatocytes, thereby disrupts the “hepatocyte damage - macrophage chemotaxis/activation - hepatocyte damage” inflammatory toxic cycle, and ultimately leads to protection against LI/R injury. Data are mean ± SD, ** P < 0.01 by two-way ANOVA followed by Tukey's test. ns, no significance.