Gsk3β regulates the resolution of liver ischemia/reperfusion injury via MerTK

Abstract

Although glycogen synthase kinase β (Gsk3β) has been shown to regulate tissue inflammation, whether and how it regulates inflammation resolution versus inflammation activation is unclear. In a murine liver, partial warm ischemia/reperfusion injury (IRI) model, we found that Gsk3β inhibitory phosphorylation increased at both the early-activation and late-resolution stages of the disease. Myeloid Gsk3β deficiency not only alleviated liver injuries, it also facilitated the restoration of liver homeostasis. Depletion of Kupffer cells prior to the onset of liver ischemia diminished the differences between the WT and Gsk3β-KO mice in the activation of liver IRI. However, the resolution of liver IRI remained accelerated in Gsk3β-KO mice. In CD11b-DTR mice, Gsk3β-deficient BM-derived macrophages (BMMs) facilitated the resolution of liver IRI as compared with WT cells. Furthermore, Gsk3β deficiency promoted the reparative phenotype differentiation in vivo in liver-infiltrating macrophages and in vitro in BMMs. Gsk3 pharmacological inhibition promoted the resolution of liver IRI in WT, but not myeloid MerTK-deficient, mice. Thus, Gsk3β regulates liver IRI at both activation and resolution stages of the disease. Gsk3 inactivation enhances the proresolving function of liver-infiltrating macrophages in an MerTK-dependent manner.

Keywords: Hepatology; Immunology; Innate immunity; Macrophages.

Figure 1. Gsk3β N-terminal phosphorylation and its regulation of the resolution of liver IRI.

Myeloid Gsk3β WT and Gsk3β-KO mice were treated with control (Ctl) Ig or anti–IL-10 Ab 1 hour prior to the start of liver ischemia. Serum and liver tissues were harvested at various times after reperfusion, as described in Methods. (A) Western blots of total and S9-phosphorylated Gsk3β and β-actin in sham and IR livers of WT B6 mice at 6 hours, 12 hours, 24 hours, and days 3 and 7 after reperfusion. Average serum ALT (sALT) levels (B) and average Suzuki scores (C) in different experimental groups at indicated time points after reperfusion. (D) Representative liver histological images (H&E staining; original magnification, ×40; scale bar: 0.2 mm) of different experimental groups at indicated time points after reperfusion. n = 6–8 livers/group. *P < 0.05. (E) Average ratios of target gene to HPRT expression (by a removing PCR) in livers of different experimental groups at days 3 and 7 after reperfusion. Data represent mean ± SEM. Representative results from 4 livers/group. *P < 0.05 (Student’s t test).

Figure 2. Myeloid Gsk3β regulates the resolution function of iMΦs in liver IRI.

Myeloid Gsk3β WT and Gsk3β-KO mice were treated with either blank liposomes or CLs 48 hours before the onset of liver ischemia. IR livers were harvested at 6 hours, 3 days, and 7 days after reperfusion. (A) Average levels of serum ALT (sALT). (B) Average Suzuki scores of different experiment groups. (C) The Kaplan-Meier survival curves of CL-treated WT and KO mice after IR. (D) Representative liver histological images (H&E staining; original magnification, ×40; scale bar: 0.2 mm) of different experimental groups at indicated time points after reperfusion. (E) Liver gross appearance of sham or IR livers at day 7 after reperfusion of different experimental groups. (F) FACS plots of NPCs isolated from sham (48 hours after CL) or IR livers at day 7 after reperfusion of different experimental groups. Myeloid cells were first gated in the FSC/SSC plot and analyzed for F4/80 and CD11b expression. (G) Immunohistochemical staining of Ly6G⁺ cells in sham and IR livers at day 7 (left panel), and Sirius red staining of IR livers at day 7 (right panel), after reperfusion of different experimental groups (original magnification, ×40; scale bar: 0.2 mm). n = 6–8/group. *P < 0.05. (H) Average ratios of target gene to HPRT in sham and IR livers at day 7 after reperfusion of different experimental groups. Data represent mean ± SEM. Representative results from 4 livers/group. *P < 0.05 (Student’s t test). BL, blank liposome.

Figure 3. Gsk3β regulates BMMs in the resolution of liver IRI.

CD11b-DTR mice were treated with diphtheria toxin at 24 hours prior to the onset of liver ischemia. BMMs derived either from myeloid Gsk3β WT orGsk3β-KO mice were injected at 24 hours after reperfusion, as described in Methods. Sham and IR livers harvested at days 3 and 7 after reperfusion were analyzed. (A) Liver histological images (H&E staining; original magnification, ×40; scale bar: 0.2 mm). (B) Average Suzuki scores of different experiment groups at indicated time points after reperfusion. n = 6×8/group. *P < 0.05. (C) Average ratio of target gene to HPRT in sham and IR livers at days 3 and 7 after reperfusion of different experimental groups. Data represent mean ± SEM. Representative results from 4 livers/group. *P < 0.05 (Student’s t test). Ctl, control; DT, diphtheria toxin.

Figure 4. Myeloid Gsk3β deficiency enhances proresolving gene expressions in macrophages.

NPCs were isolated from sham and IR livers at days 3 and 7 after reperfusion in CL-treated WT or myeloid Gsk3β-KO mice, as described in Methods. (A) Average ratios of target gene to HPRT. (B) Western blot of MerTK levels in liver NPCs of different experimental groups at indicated time points after reperfusion. (C and D) BMMs derived from WT or myeloid Gsk3β-KO mice were stimulated in vitro for 24 hours with LPS (C) or the LXR agonist DMHCA (D). Average ratios of target gene to HPRT of different experiment groups are plotted. KCs isolated from WT sham livers were used as the control in D. (E) Western blot of MerTK in LXR-stimulated BMMs. Gene expression was measured by quantitative reverse-transcription PCR. Data represent mean ± SEM. Representative results from 3–4 livers/group. *P < 0.05 (Student’s t test). Ctl, control.

Figure 5. Myeloid Gsk3β deficiency enhances proresolution functions in BMMs in vitro.

BMMs derived from WT or myeloid Gsk3β-KO mice were stimulated in vitro for 24 hours with a control (Ctl) or the LXR agonist DMHCA and tested in an in vitro efferocytosis assay by incubating with pHrodo-labeled apoptotic thymocytes, as described in Methods. Cells were stained with FITC-labeled anti–F4/80, and efferocytosis was quantitated under a confocal microscope. (A) Representative fluorescence images of F4/80-stained BMMs. (B) Average percentage of efferocytes in total macrophage population. (C) Average TNF-α and IL-10 levels in the culture supernatants of WT and Gsk3β KO BMMs stimulated with LPS for 24 hours in the absence or presence of apoptotic thymocytes. Cytokine levels were quantitated by ELISA. Data represent mean ± SEM. Representative results from 4 livers/group. *P < 0.05 (Student’s t test). AC, apoptotic cells.

Figure 6. Pharmacological inhibition of Gsk3 facilitates the resolution of liver IRI.

WT B6 mice were subjected to 90 minutes of liver ischemia, as described in Methods. At 24 hours after reperfusion, they were divided into 2 groups and treated with vehicle control (Ctl) or SB216763. IR livers were harvested at days 3, 5, and 7 after reperfusion and analyzed by histological evaluation. Average serum ALT (sALT) levels (A) and average Suzuki scores (B) of the 2 experimental groups at indicated time points after reperfusion. (C) Representative liver histological images (H&E staining) of different experiments at indicated time points after reperfusion. Data represent mean ± SEM. Representative results from 6–8 livers/group. *P < 0.05 (Student’s t test).

Figure 7. MerTK is critical for the proresolving therapeutic effect of pharmacological Gsk3 inhibition.

WT and myeloid MerTK-KO mice were treated with CLs 48 hours prior to the onset of liver ischemia. Vehicle control (Ctl) and SB216763 were administered at 24 hours and on days 3 and 5 after reperfusion, as described Methods. IR livers were harvested on day 7 after reperfusion. (A) Liver gross appearance and histological images of different experiment groups. (B) Average Suzuki scores of IR livers of different experimental groups at day 7 after reperfusion. (C) Average ratios of target gene to HPRT in IR livers of different experimental groups at day 7 after reperfusion. Liver gene expression was determined by quantitative reverse-transcription PCR. Data represent mean ± SEM. Representative results from 4–6 livers/group. *P < 0.05 (Student’s t test).

Figure 8. Gsk3β as the converging point in the differentiation and the execution of pre-resolving functions in macrophages.

Gsk3β is able to regulate (1) TLR-mediated sterile inflammation (DAMPs from necrotic cells); (2) Notch mediated signaling (to upregulate LXR); as well as (3) MerTK-mediated resolution functions (binds PtdSer on apoptotic cells via Gas6/ProS) and LXR-mediated resolution functions, downstream or independent of PI3K/Akt, are potentially regulated by Gsk3β. Gsk3β inactivation downregulates the proinflammatory, but upregulate the immunoregulatory or reparative, gene programs, including IL-10, MerTK, and LXR-α, and possibly specific proresolving mediators (SPMs).