T-Cell Immunoglobulin and Mucin Domain-Containing Protein-4 Is Critical for Kupffer Cell Homeostatic Function in the Activation and Resolution of Liver Ischemia Reperfusion Injury

Abstract

Background and aims: Liver ischemia reperfusion injury (IRI) remains an unresolved clinical problem. This study dissected roles of liver-resident macrophage Kupffer cells (KCs), with a functional focus on efferocytosis receptor T-cell immunoglobulin and mucin domain-containing protein-4 (TIM-4), in both the activation and resolution of IRI in a murine liver partial warm ischemia model.

Approach and results: Fluorescence-activated cell sorting results showed that TIM-4 was expressed exclusively by KCs, but not infiltrating macrophages (iMФs), in IR livers. Anti-TIM-4 antibody depleted TIM-4⁺ macrophages in vivo, resulting in either alleviation or deterioration of liver IRI, which was determined by the repopulation kinetics of the KC niche with CD11b⁺ macrophages. To determine the KC-specific function of TIM-4, we reconstituted clodronate-liposome-treated mice with exogenous wild-type or TIM-4-deficient KCs at either 0 hour or 24 hours postreperfusion. TIM-4 deficiency in KCs resulted in not only increases in the severity of liver IRI (at 6 hours postreperfusion), but also impairment of the inflammation resolution (at 7 days postreperfusion). In vitro analysis revealed that TIM-4 promoted KC efferocytosis to regulate their Toll-like receptor response by up-regulating IL-10 and down-regulating TNF-α productions.

Conclusions: TIM-4 is critical for KC homeostatic function in both the activation and resolution of liver IRI by efferocytosis.

Figure 1.

Liver IR downregulates TIM-4 in KCs. B6 mice were subjected to 90m liver warm ischemia in the cephalad lobes followed by various lengths of reperfusion. Liver NPCs were isolated from the sham or ischemic livers after 6h, 3d and 7d of reperfusion. (a) Liver MФ subsets based on F4/80 and CD11b expressions (density plots). KCs, infiltrating MФs (iMФs) and neutrophils were identified by F4/80⁺CD11b⁻, F4/80⁺CD11b⁺ and F4/80⁻CD11b⁺, respectively. (b) Histograms of KC TIM-4 at different time points of reperfusion, and the TIM-4 level in KCs and iMФs at 6h and day3 post reperfusion. (c) TIM-4 levels in liver tissues measured by quantitative RT-PCR, in sham or ischemic livers at different time points post reperfusion. n=4/group, representative results of 2 independent experiments. * p<0.05, **p<0.005. (d) Representative immunofluorescence staining of CD68 and TIM-4 in human liver biopsies and their merged images. Frozen sections were stained with FITC-anti-TIM-4, PE-anti-CD68 and DAPI. Scale bar=30μM. (e) Quantitative RT-PCR and (f) Western blotting of TIM-4 in ischemic human liver tissues. Peri-tumor tissues were harvested after various lengths of ischemia (0–50min) during hepatic tumor resection in 3 series of hepatocarcinoma patients.

Figure 2.

The opposite effects of peri-operational (−2h) vs. pre-emptive (−48h) anti-TIM-4 Ab treatments on liver IRI. (a) Control or anti-TIM-4 Abs were administered at either 2h or 48h prior to the onset of liver ischemia. Liver IRI was evaluated at 6h post reperfusion. A diagram of the experimental scheme was shown. Average sALT levels and Suzuki scores of different experimental groups were plotted. Representative histological pictures of IR livers (H/E x40) were shown. Average target gene/HPRT ratios in IR livers were plotted. (b) FACS analysis of liver non-parenchymal cells (NPCs). Liver NPCs were isolated from either control Ig or anti-TIM-4 Ab-treated mice at either 2h (upper panel) or 48h (lower panel) post the treatment, as described in the materials and methods. Cells were stained with fluorochrome-labeled Abs. Representative plots of F4/80 and CD11b expressions in the myeloid cell gate, and TIM-4/Gr-1 expression in the F4/80⁺CD11b^low subset. (c) CD11b-DTR mice were treated with a single dose of DT followed by the control or anti-TIM-4 Abs. Liver ischemia was performed 48h later and IRI was evaluated at 6h post reperfusion. Average sALT levels and target gene/HPRT ratios in IR livers in different experiment groups were plotted. All results are representatives of at least two independent experiments. n=4–6 mice/group. Data were tested for normal distribution and analyzed using a one-way ANOVA with post-test. *p<0.05.

Figure 3.

The anti-TIM-4 Ab depleted KCs in livers and the repopulation kinetics of the emptied KC niche determined the outcome of liver IRI. Control Ig or anti-TIM-4 Ab were administered in separate groups of B6 mice. Liver tissues were harvested (Sham) or liver IR was performed in these treated mice (IR6h), 2h or 48h later after the Ab injection, as described in the Materials and Methods. IR-livers were harvested at 6h post reperfusion. Both sham and IR liver tissues were analyzed by immunofluorescence staining with anti-F4/80, or -CD11b, or -Clec4F Abs and DAPI. Representative images were shown. Average percentages of F4/80, CD11b and Clec4F positive cells in tissue sections of different experimental groups were plotted. All results are representatives of at least two independent experiments. n=4–6 mice/group. Data were tested for normal distribution and analyzed using a one-way ANOVA with post-test. *p<0.05.

Figure 4.

The depletion of KCs impairs the resolution of liver IRI. B6 mice were treated with either control Ig or blank liposomes (Ctl) or clodronate liposomes (CL) or anti-TIM-4 Ab 48h prior to the onset of liver ischemia. Liver IRI and inflammatory responses were evaluated at 7d post reperfusion. (a) Representative histological pictures of IR livers (H/E x40) were shown. Average Suzuki scores of different experimental groups were plotted. (b) Liver NPCs were isolated from IR livers and analyzed by FACS. Representative cell density plot based on F4/80 and CD11b expressions was shown. (c) Average target gene/HPRT ratios in IR livers were plotted. All results are representatives of at least two independent experiments. n=4–6 mice/group. Data were tested for normal distribution and analyzed using a one-way ANOVA with post-test. *p<0.05.

Figure 5.

TIM-4 regulates KC efferocytosis and innate immune activation in vitro. (a) Top panel: KCs were co-incubated with unlabeled (–) or pHrodo-SE-labeled (+) apoptotic thymocytes in the presence of control Ig or anti-TIM-4 (αTIM4) Abs for 2h. Cells were stained with fluorochrome-labeled anti-F4/80, DAPI, and visualized under fluorescence microscope. Middle panel: TIM-4⁺ KCs and TIM-4⁻ liver macrophages were isolated from livers treated with control Ig or anti-TIM-4 Ab-treated mice at −48h and measured in the efferocytosis assay. Lower panel: KCs were isolated from WT or TIM-4 KO mice and efferocytosis was measured as above. Representative images were shown and average % of F4/80⁺/pHrodo-SE⁺ cells were plotted. (b) WT (TIM-4⁺), TIM-4 KO KCs and TIM-4⁻ liver macrophages were isolated as detailed above and stimulated in vitro with LPS in the absence or presence of apoptotic thymocytes (apo-Thy) and/or anti-TIM-4 Abs. Culture supernatants were harvested at 24h post stimulation and TNF-a and IL-10 levels were measured by ELISA. All results are representatives of at least two independent experiments. n=3–5 mice/group. Data were tested for normal distribution and analyzed using a one-way ANOVA with post-test. *p<0.05.

Figure 6.

TIM-4 is critical for the homeostatic function of KCs in the inflammatory activation of liver IRI. KC-depleted mice (by CLs) were reconstituted with either WT or TIM-4 KO KCs at 0h post reperfusion, as described in the Materials and Methods. (a) The diagram of the reconstitution experiment. Liver IRI and inflammatory responses were measured at 6h post reperfusion. (b) Representative histological pictures of IR livers (H/E x40) in KC-intact (Ctl), KC-depleted (CL) and WT or TIM-4 KO KC-reconstituted groups were shown. (c) Average sALT levels and Suzuki scores in different experimental groups were plotted. (d) Average target gene/HPRT ratios in IR livers of different experimental groups were plotted. All results are representatives of at least two independent experiments. n=4–6 mice/group. Data were tested for normal distribution and analyzed using a one-way ANOVA with post-test. *p<0.05.

Figure 7.

TIM-4 is critical for the homeostatic function of KCs in the resolution of liver IRI. KC-depleted mice (by CLs) were reconstituted with either WT or TIM-4 KO KCs at 24h post reperfusion, as described in the Materials and Methods. (a) The diagram of the reconstitution experiment. The resolution of liver IRI and inflammatory responses were measured at 7d post reperfusion. (b) Representative histological pictures of IR livers (H/E x40) in KC-intact (Ctl), KC-depleted (CL) and WT or TIM-4 KO KC-reconstituted groups were shown. (c) Average Suzuki scores and (d) average target gene/HPRT ratios in IR livers of different experimental groups were plotted. All results are representatives of at least two independent experiments. n=4–6 mice/group. Data were tested for normal distribution and analyzed using a one-way ANOVA with post-test. *p<0.05.

Figure 8.

TIM-4+ KCs and liver IRI. Upper panel: Anti-TIM-4 Abs deplete KCs in vivo and the repopulation kinetics of KC niche determines the Ab-effect on liver IRI. Lower panel: TIM-4 phenotypes determine KC response/function.