PPAR-gamma activation is associated with reduced liver ischemia-reperfusion injury and altered tissue-resident macrophages polarization in a mouse model

Abstract

Background: PPAR-gamma (γ) is highly expressed in macrophages and its activation affects their polarization. The effect of PPAR-γ activation on Kupffer cells (KCs) and liver ischemia-reperfusion injury (IRI) has not yet been evaluated. We investigated the effect of PPAR-γ activation on KC-polarization and IRI.

Materials and methods: Seventy percent (70%) liver ischemia was induced for 60mins. PPAR-γ-agonist or vehicle was administrated before reperfusion. PPAR-γ-antagonist was used to block PPAR-γ activation. Liver injury, necrosis, and apoptosis were assessed post-reperfusion. Flow-cytometry determined KC-phenotypes (pro-inflammatory Nitric Oxide +, anti-inflammatory CD206+ and anti-inflammatory IL-10+).

Results: Liver injury assessed by serum AST was significantly decreased in PPAR-γ-agonist versus control group at all time points post reperfusion (1hr: 3092±105 vs 4469±551; p = 0.042; 6hr: 7041±1160 vs 12193±1143; p = 0.015; 12hr: 5746±328 vs 8608±1259; p = 0.049). Furthermore, liver apoptosis measured by TUNEL-staining was significantly reduced in PPAR-γ-agonist versus control group post reperfusion (1hr:2.46±0.49 vs 6.90±0.85%;p = 0.001; 6hr:26.40±2.93 vs 50.13±8.29%; p = 0.048). H&E staining demonstrated less necrosis in PPAR-γ-agonist versus control group (24hr:26.66±4.78 vs 45.62±4.57%; p = 0.032). The percentage of pro-inflammatory NO+ KCs was significantly lower at all post reperfusion time points in the PPAR-γ-agonist versus control group (1hr:28.49±4.99 vs 53.54±9.15%; p = 0.040; 6hr:5.51±0.54 vs 31.12±9.58%; p = 0.009; 24hr:4.15±1.50 vs 17.10±4.77%; p = 0.043). In contrast, percentage of anti-inflammatory CD206+ KCs was significantly higher in PPAR-γ-agonist versus control group prior to IRI (8.62±0.96 vs 4.88 ±0.50%; p = 0.04). Administration of PPAR-γ-antagonist reversed the beneficial effects on AST, apoptosis, and pro-inflammatory NO+ KCs.

Conclusion: PPAR-γ activation reduces IRI and decreases the pro-inflammatory NO+ Kupffer cells. PPAR-γ activation can become an important tool to improve outcomes in liver surgery through decreasing the pro-inflammatory phenotype of KCs and IRI.

Fig 1. Hepatocellular injury after 60min of ischemia and 1, 6, 12 and 24hrs of reperfusion.

Significant lower AST levels were found in the RGZ vs CTRL group at 1hr (3092±105 vs 4469±551; p = 0.042), 6hr (7041±1169 vs 12192±1443; p = 0.015) and 12hr (5746±328 vs 8609±1259; p = 0.049) after reperfusion (A). RGZ vs control group showed a trend to a significant difference at 6hrs post-reperfusion in ALT (9185±1754 vs 13823±1465; p = 0.054) (B). Abbreviations: AST-aspartate aminotransferase, ALT-alanine aminotransferase, RGZ-Rosiglitazone, CTRL-Control. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney U test.

Fig 2. Level of apoptosis.

Representative images of TUNEL(A) and Cleaved Caspase-3(B) immunohistochemistry showing a lower grade of apoptosis in RGZ vs Control group at 1hr and 6hr post-reperfusion. Significant lower TUNEL-staining was found in the RGZ vs CTRL group at 1hr (2.46±0.49 vs 6.90±0.85%; p = 0.001) and 6hrs (26.40±2.93 vs 50.13±8.29%; p = 0.048) post-reperfusion by image analysis(C). Cleaved caspase-3 immunohistochemistry also demonstrated less staining in the RGZ group vs control group, the difference was assessed by image software analysis (D). Abbreviations: TUNEL-Terminal deoxynucleotidyl transferase(TdT) dUTP Nick-End-Labeling, RGZ-Rosiglitazone, Ctrl-Control. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney-U-test.

Fig 3. Hepatic necrosis.

Representative pictures of the left lateral lobe with H&E staining showing a less necrotic area in the RGZ vs control group (A) 24hrs after reperfusion (26.66±4.78 vs 45.62±4.57%; p = 0.032). Image software analysis demonstrated that this difference was significant (B). Abbreviations: CTRL Control, RGZ Rosiglitazone. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney U test.

Fig 4. Serum cytokines.

Anti-inflammatory cytokine IL-10 was found with higher levels in RGZ vs CTRL group 12hrs (119.98±46.22 vs 76.66±19.73 pg/ml; p = 297) following IRI difference became significant at 24hr time-point (172.64±34.18 vs 34.18±6.89 pg/ml; p = 0.034) (A). TNF-alpha showed lower levels since 1hr after reperfusion difference that became significant 24hr (4.15±0.95 vs 18.08±2.35pg/ml; p = 0.013) after IRI (B). Serum IL-6 levels were similar for both groups at all time-points (C). Abbreviations: TNF-alpha Tumor necrosis factor, IL Interleukin, CTRL Control, RGZ Rosiglitazone. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney-U test.

Fig 5. Hepatic macrophages.

Immunohistochemistry technique showing how the total percentage of F4/80 staining decreases progressively after the IRI event. No difference was found when RGZ treated group was compared against the control group. Abbreviations: RGZ Rosiglitazone, CTRL Control. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney U test.

Fig 6. Kupffer cell populations.

Flow-cytometry analysis showed a significant decrease on pro-inflammatory NO+ KCs population in RGZ vs CTRL group since 1hr (28.49±4.99 vs 53.54±9.15%; p = 0.040), this difference further increase at 6hr (5.51±0.54 vs 31.12±9.58%; p = 0.009) and 24hr (4.15±1.50 vs 17.10±4.77%; p = 0.043) following reperfusion (A). Analysis of anti-inflammatory CD206+ KCs population showed a significant increase in RGZ vs CTRL group prior to IRI (8.62±0.96 vs 4.88±0.50%; p = 0.040) (B). No difference in the percentage of anti-inflammatory IL10+ KCs was found prior or after reperfusion (C). Abbreviations: RGZ-Rosiglitazone, CTRL-Control. Five experiments (n = 5) per group per time point were performed. Results are shown as mean±SEM, Mann-Whitney-U-test.

Fig 7. Pro-inflammatory-NO+/ anti-inflammatory-CD206+ Kupffer cells ratio prior and after reperfusion.

Pro-inflammatory-NO+/ anti-inflammatory-CD206+ KCs ratio was lower in the RGZ treated group at all the studied time points. Significant lower ratio was found in the RGZ vs CTRL group since 1hr after reperfusion group (3.45±0.32 vs 5.68±0.37; p = 0.001), this difference remained significant at 6hr (0.81±0.11 vs 4.63±0.79; 0.008) and 24hr (0.49±0.05 vs 2.66±0.46; p = 0.018) post-reperfusion. Abbreviations: CTRL control, RGZ Rosiglitazone. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney U test.

Fig 8. PPAR-γ antagonist-6hrs after reperfusion.

Flow-cytometry shows how the effect on pro-inflammatory NO+ KC polarization was significantly blocked by the use of PPAR-γ antagonist (GW9662) (RGZ: 3.93±3.61%, RGZ+GW9662: 35.86±21.85%, Control: 31.12±21.42%, p = 0.009) (A). Serum AST levels were also significantly reversed when antagonist was included as an intervention (B). The grade of apoptosis was significantly increased with the use of antagonist in combination with RGZ when compared with RGZ alone (C). Abbreviations: DAF-FM 4-amino-5-methylamino-2’,7’-Difluoroflurescein, TUNEL-Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling, RGZ-Rosiglitazone, CTRL-Control. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney-U-test.

Fig 9. Hepatic injury and apoptosis in RGZ treated mice following reperfusion.

Analysis of hepatic injury showed no significant difference in the RGZ vs Control group for AST (14451±3518 vs 18231±3688U/L, p = 0.482), Inhibition of PPAR-γ with GW9662 increased further the levels of AST when compared with control group (A). Representative images of TUNEL immunohistochemistry showing a similar grade of apoptosis in RGZ, CTRL and RGZ+GW-9662 groups at 6hr post-reperfusion (B). TUNEL positive staining was found similar in RGZ vs control group at 6hr post-reperfusion (42±6 vs 48±3%, p = 0.442) (C). GW-9662 showed a reversal of the slight beneficial effect on apoptosis found in the RGZ group (C). Note: Staining was assessed with image software analysis. Abbreviations: AST-aspartate aminotransferase, RGZ-Rosiglitazone, CTRL-Control, TUNEL-Terminal deoxynucleotidyl transferase(TdT) dUTP Nick-End-Labeling. Five experiments (n = 5) per group per time point were performed. Results are shown as mean ± SEM, Mann-Whitney U test.