Fig 2: ERRFI1 deficiency ameliorated IR-induced hepatic injury and hepatocellular apoptosis. (A) Schematic diagram showed the hepatocyte-specific ERRFI1 knockout strategy and ischemia reperfusion model. (B, C) Western blot analysis of ERRFI1 protein level in liver tissues from wild type (WT) mice and hepatocyte-specific ERRFI1-knockout (ERRFI1-HKO) mice with sham treatment or ischemia for 90 min followed by reperfusion for 6 h, and quantitative analysis is shown. (D) Liver function assessed by ALT and AST of mice with different treatment. (E) Liver pathology was determined by H&E staining (scale bar: 100 μm). (F) Suzike’s injury score was used to assess the degree of injury based on H&E staining. (G, H) TUNEL staining of apoptotic cells in liver tissues from WT mice and ERRFI1-HKO mice under different conditions (scale bar: 50 μm), and quantification showing the percentage of apoptotic cells. (I, J) Western blot analysis of Bax, Bcl-2, and cleaved caspase-3 in liver tissues from WT mice and ERRFI1-HKO mice after IR injury. For statistical analysis, one-way ANOVA was used (n = 6)

Fig 3: Loss of liver p53 exacerbates liver damage and increases ROS during CCl4-mediated liver regeneration.Representative H&E images (A) and quantification (B) of damaged area (% per field) in livers from Albumin-Cre; p53WT/WT mice (WT) and Albumin-Cre; p53FL/FL (FL) mice at indicated times (hours) following CCl4 treatment. Damaged areas outlined. Scale bars 100 μm. Higher magnification inset images highlight centrilobular liver damage including vacuolisation (black arrows). These images are taken from different H&E slides than those depicted for damaged area. Scale bars 10 μm. Images reproduced without annotation and at full size in Fig. S1 C/D. Quantifications from N = 2 untreated (0 h) mice/group, N = 6 mice/group at 24 h, N = 5 WT and N = 4 FL mice at 48 h, N = 7 mice/group at 72 h, N = 3 mice/group at 168 h. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm-Sidak’s multiple comparisons test and multiplicity-adjusted p values. ****p < 0.0001. Plasma ALT (C) and AST (D) activity (mU/mL) in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice treated as in (A). N = 3 untreated (0 h) mice/group. N = 4 mice/group/time point thereafter. Each data point represents the mean from technical duplicates per mouse. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm-Sidak’s multiple comparisons test and multiplicity-adjusted p-values. *p < 0.05, **p < 0.01, ****p < 0.0001. Staining (E) and quantification of stain area (F) of frozen sections for oil-red-O (ORO) in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice treated as in (A). Scale bars 20 μm. Quantification from N = 5 untreated (0 h), 24 h, and 48 h mice/group, N = 7 mice/group at 72 h, and N = 3 mice/group at 168 h. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm-Sidak’s multiple comparisons test and multiplicity-adjusted p values. *p < 0.05, ***p < 0.001. Data from p53FL/FL mice also used in (L) (normal water) but different representative images are shown. IHC staining (G) and quantification (H) of malondialdehyde (MDA) in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice at indicated times (hours) after CCl4 treatment. Scale bars 20 μm. Quantification from N = 6 untreated (0 h) mice/group, N = 7 WT and N = 6 FL mice at 24 h, N = 7 mice/group at 48 and 72 h, and N = 3 mice/group at 168 h. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm-Sidak’s multiple comparisons test and multiplicity-adjusted p values. **p < 0.01, ****p < 0.0001. Data from p53FL/FL mice also used in (J) (normal water) but different representative images are shown. IHC staining (I) and quantification (J) of malondialdehyde (MDA) in Albumin-Cre; p53FL/FL (FL) mice given control (water) or 30 mM N-Acetylcysteine-supplemented drinking water (NAC) for one week prior to CCl4 treatment. Images from indicated times (hours) after CCl4 treatment. Scale bars 20 μm. Quantification from N = 6 water and N = 2 NAC untreated mice (0 h), N = 6 water and N = 4 NAC mice at 24 h, N = 7 water and N = 4 NAC mice at 48-72 h. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm-Sidak’s multiple comparisons test and multiplicity-adjusted p-values. **p < 0.01, ****p < 0.0001. Data from p53FL/FL (normal water) mice also used in (H) but different representative images are shown. Staining (K) and quantification (L) of oil-red-O (ORO) in Albumin-Cre; p53FL/FL (FL) mice given control (water) or 30 mM N-Acetylcysteine-supplemented drinking water (NAC) for one week prior to CCl4 treatment. Images from indicated times (hours) after CCl4 treatment. Scale bars 20μm. Quantification from N = 5 untreated water (0 h) and N = 2 untreated NAC mice, N = 5 water and N = 4 NAC mice at 24 and 48 h, N = 7 water and N = 4 NAC mice at 72 h. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm-Sidak’s multiple comparisons test and multiplicity-adjusted p-values. *p < 0.05. Data from p53FL/FL (normal water) mice also used in (F) but different representative images are shown.

Fig 4: Loss of liver p53 leads to hepatic hypertrophy, chronic ROS and DNA damage, and impaired liver function after CCl4-mediated chronic regeneration.A Quantification of relative hepatocyte size in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice at 7 days after completion of 10-week CCl4 chronic regeneration regime. N = 9 mice/genotype. Data presented as mean ± SEM and analysed using an unpaired two-tailed t-test with Welch’s correction. **p < 0.01. B H&E and IHC staining for alpha-smooth muscle actin (aSMA), and picrosirius red (PSR) staining in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice at 7 days after completion of 10-week CCl4 chronic regeneration regime. Scale bars 10 μm. Images representative of N = 10 WT and N = 8 FL mice for H&E, N = 4 WT and N = 5 FL mice for aSMA, and N = 7 mice/group for PSR. C Quantification of IHC staining for aSMA in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice from (B) at either 7 days after completion of 10-week CCl4 chronic regeneration regime (chronic) or in untreated age-matched mice (control). N = 5 control mice/genotype, N = 4 WT and N = 5 FL chronic mice. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm–Sidak’s multiple comparisons test and multiplicity-adjusted p values. **p < 0.01. D Quantification of staining for picrosirius red (PSR) in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice as in (C). N = 4 control mice/genotype and N = 7 chronic CCl4 mice/genotype. Data presented as mean ± SEM and analysed using two-way ANOVA with Holm–Sidak’s multiple comparisons test and multiplicity-adjusted p values. **p < 0.01, ***p < 0.001, ****p < 0.0001. Images (E) and quantification (F) of IHC staining MDA, CYP2A5, and gH2AX in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice at 7 days after completion of 10-week CCl4 chronic regeneration regime as in (C). Scale bars 20 μm. N = 4 mice/group for MDA, N = 7 mice/group for CYP2A5, and N = 9 mice/group for gH2AX. Data presented as mean ± SEM and analysed using two-tailed t-tests and the Sidak–Bonferroni method to account for multiplicity of tests. **p < 0.01, ****p < 0.0001. G Plasma ALT and AST activity (mU/mL) in Albumin-Cre; p53WT/WT (WT) and Albumin-Cre; p53FL/FL (FL) mice at 7 days after completion of 10-week CCl4 chronic regeneration regime. N = 7 mice/group. Each data point represents the mean from technical duplicates per mouse. Data presented as mean ± SEM and analysed using two-tailed t-tests and the Sidak–Bonferroni method to account for multiplicity of tests. ***p < 0.001.

Fig 5: MDEVs alleviate SALI by promoting Tregs differentiation and inhibiting M1 polarization synchronously(A and B) HE stains (A, scale bars: 100 μm) and Suzuki’s score (B) in mice liver during SALI with/without WT/Roquin-1san/san MEDVs injection (n = 6).(C) serum ALT and AST in SALI mice with/without WT/Roquin-1san/san MEDVs MDEVs injection (n = 6).(D and E) Flow cytometry and quantitation of M1 møs and Tregs in mice liver during SALI with/without MEDVs injection (n = 6).(F–I) Serum TNF-α (f)/TGF-β (g)/IL-6 (h)/IL-10 (i) in SALI mice with/without MDEVs injection (n = 6). p < 0.01∗, p < 0.01 ∗∗, p < 0.001 ∗∗∗, p < 0.0001 ∗∗∗∗. The data are expressed as the mean ± SD. The Student’s t test was used for comparison between the two groups.