Type 1 Innate Lymphoid Cells Are Proinflammatory Effector Cells in Ischemia-Reperfusion Injury of Steatotic Livers

Abstract

Innate lymphoid cells (ILCs), the most recently described family of lymphoid cells, play fundamental roles in tissue homeostasis through the production of key cytokine. Group 1 ILCs, comprised of conventional natural killer cells (cNKs) and type 1 ILCs (ILC1s), have been implicated in regulating immune-mediated inflammatory diseases. However, the role of ILC1s in nonalcoholic fatty liver disease (NAFLD) and ischemia-reperfusion injury (IRI) is unclear. Here, we investigated the role of ILC1 and cNK cells in a high-fat diet (HFD) murine model of partial warm IRI. We demonstrated that hepatic steatosis results in more severe IRI compared to non-steatotic livers. We further elicited that HFD-IRI mice show a significant increase in the ILC1 population, whereas the cNK population was unchanged. Since ILC1 and cNK are major sources of IFN-γ and TNF-α, we measured the level of ex vivo cytokine expression in normal diet (ND)-IRI and HFD-IRI conditions. We found that ILC1s in HFD-IRI mice produce significantly more IFN-γ and TNF-α when compared to ND-IRI. To further assess whether ILC1s are key proinflammatory effector cells in hepatic IRI of fatty livers, we studied both Rag1^-/- mice, which possess cNK cells, and a substantial population of ILC1s versus the newly generated Rag1^-/-Tbx21^-/- double knockout (Rag1-Tbet DKO) mice, which lack type 1 ILCs, under HFD IRI conditions. Importantly, HFD Rag1-Tbet DKO mice showed significant protection from hepatic injury upon IRI when compared to Rag1^-/- mice, suggesting that T-bet-expressing ILC1s play a role, at least in part, as proinflammatory effector cells in hepatic IRI under steatotic conditions.

Keywords: T-bet; Type 1 innate lymphoid cells; fatty liver disease; innate lymphoid cells; ischemia-reperfusion injury; liver transplantation; natural killer cells; phasor fluorescence lifetime imaging.

Figure 1

ILC1s are elevated in the liver of C57B/6 wild-type HFD-IRI mice and produce IFN-γ and TNF-α. C57B/6 wild-type mice were fed either standard chow (normal diet, ND) or a lard-based high fat diet (HFD). HFD was started at five-to-seven weeks of age and maintained for 12 weeks. All mice were 17-23 weeks of age at the time of experimentation. A 45-minute partial warm ischemia time was used for IRI experiments. All analysis was performed following 24 hours of reperfusion. (A) Serum alanine aminotransferase (ALT), the indicator of liver damage, was measured (n=3-7 mice). (B) Numbers of the infiltration of inflammatory cells (i.e., CD68⁺ and Gr1⁺) in the ND and HFD mice after IRI were quantified (n=9-10). (C) Gating strategy of ILC1s and cNKs from the hepatic lymphocytes: lineage- negative (Lin^-) NK1.1⁺ cells expressing CD45 were identified as ILC1s and cNKs by expression of CD49a, CD49b, T-bet, and Eomes. Representative flow plots showing ILC1 and cNK subsets in ND-IRI and HFD-IRI. The percentages of ILC1s (D) and NKs (E) in ND and HFD mice after non-IRI or IRI (n=7-16 mice). (F) Representative flow plots showing intracellular cytokine production (IFN-γ⁺ and TNF-α⁺ cells) in ND-IRI and HFD-IRI. Fresh hepatic lymphocytes were stimulated with PMA/ionomycin for 4 hours in the presence of IL-12, IL-15, and IL-18. FACS plots gated on live, CD45⁺, Lin^-, NK1.1⁺, CD49a⁺ and CD49b^- for ILC1s and CD45⁺, Lin^-, NK1.1⁺, CD45a⁺ and CD45b⁺ for cNKs. The percentage of cytokines from ILC1 gated (G) and NK gated (H) cells in ND-IRI and HFD-IRI mice (n=5 mice per group). Significance was determined using Mann Whitney Test. *p<0.05; **p<0.01; ns, not significant.

Figure 2

Steatosis quantification using long lifetime species of phasor-in Rag1^-/- and Rag1-Tbet DKO hepatic IRI mouse model. HFD was started at five-to-seven weeks of age and maintained for 12 weeks. All mice were 17-19 weeks at time of evaluation. (A) Fluorescence lifetime images were color mapped for long lifetime species (LLS) present in lipid droplets, and their size distribution was calculated using a custom ImageJ script (droplet map is shown in Figure 2A , bottom). (B) The LLS signal was selected based on the phasor-FLIM map (red circle). The calculated average lipid droplet sizes (C) and the individual droplet size distributions are plotted (D). The excitation wavelengths used were 740 nm. Significance was determined using Mann Whitney Test. *p<0.05; ns, not significant. Each dot represents the value for a single mouse.

Figure 3

Absence of ILCs lead to increased liver protection to IRI. HFD was started at five-to-seven weeks of age and maintained for 12 weeks. All mice were 17-19 weeks at time of evaluation. A 45-minute partial warm ischemia time was used for IRI experiments. All analysis was performed following 24 hours of reperfusion. (A) Serum alanine aminotransferase (ALT) was measured in ND-IRI and HFD-IRI between Rag1^−/− and Rag1-Tbet DKO mice (n=5-7). The numbers of CD68⁺ (B) and Gr1⁺ (C) in the ND-IRI and HFD-IRI mice were quantified (n=9-10). (D) Representative flow plots show that the frequencies of ILC1s were nearly absent in Rag1-Tbet DKO. Data in Figure (D) are representative of ND IRI Rag1^−/− or Rag1-Tbet DKO mice. (E) Percentage (F) and the absolute number of ILC1s and cNKs in Rag1 ^−/− or Rag1-Tbet DKO mice following ND-IRI and HFD-IRI (n=5-7). Significance was determined using Mann Whitney Test. *p<0.05; **p<0.01; ns; not significant.

Figure 4

ILC1s produce higher IFN-γ in HFD-IRI. (A) Representative flow plots showing intracellular cytokine production (IFN-g⁺ and TNF-α⁺ cells) in ND-IRI and HFD-IRI Rag1 ^−/− mice. FACS plots gated on live, CD45⁺, Lin^-, NK1.1⁺, CD49a⁺ and CD49b^- for ILC1s and CD45⁺, Lin^-, NK1.1⁺, CD45a^- and CD45b⁺ for cNKs. (B) Percentage of cytokines from ILC1 gated cells in ND-IRI and HFD-IRI Rag1 ^−/− mice. (C) Percentage of cytokines from cNK gated cells in Rag1 ^−/− and Rag1-Tbet DKO mice (n=3-5 mice per group). Significance was determined using Mann Whitney Test. *p<0.05; ns, not significant.

Figure 5

Lack of T-bet leads to a decrease in mNK cells and lower expression of perforin. (A) Gating strategy of cNKs from the hepatic lymphocytes: lineage- negative (Lin^-) NK1.1⁺ cells expressing CD45 were identified as cNKs by expression of CD49a and CD49b. Immature NK (CD11b⁻ CD27⁺), Double Positive (CD11b⁺CD27⁺, DP) and mature NK (CD11b⁺CD27⁻) were identified by CD11b and CD27 expression, as shown. Representative flow plots (A) from ND-IRI Rag1^−/− compared with ND-IRI Rag1-Tbet DKO mice and bar graph (B) showing NK cell developmental stages in IRI Rag1^−/− compared with IRI Rag1-Tbet DKO mice. n=4 mice per group. (C) The frequency of CD107a staining of cNK cells from ND-IRI and HFD-IRI mice. Fresh hepatic lymphocytes were cultured for 4 h in the presence of anti-CD107a and Brefeldin (A) MFI of granzyme B (D) and perforin (E) staining in cNK cells from ND-IRI and HFD-IRI mice. n=4 mice per group. Significance was determined using Mann Whitney Test. *p<0.05. **p<0.01; ns, not significant.

Figure 6

Cytokine and chemokine gene expression array for mice liver IRI. (A) Clustergram represents unsupervised hierarchical clustering analysis of RT² qPCR profiler array for ND Rag1^−/− HFD Rag1^−/− , ND Rag1-Tbet DKO, and HFD Rag1-Tbet DKO pooled, post-IRI mice liver tissues. Comparative fold regulation analysis of post-IRI cytokines and chemokines with at least 2-fold or more up-and down-regulation between (B) ND and HFD Rag1^−/− and (C) ND and HFD Rag1-Tbet DKO mice liver tissues relative to the respective ND non-IRI control. Dotted vertical lines mark up-and down-regulation of the gene expression. All data are represented as a gene expression of at least three pooled bulk mouse liver tissues.