Temporal dichotomy of neutrophil function in acute liver injury and repair

Abstract

Background & aims: Acetaminophen (APAP)-induced acute liver injury (APAP-ALI) is the leading cause of acute liver failure-induced death, with host innate immune responses driving outcomes. Neutrophils are activated and increased in APAP-ALI and reported to contribute to liver damage. However, neutrophil dysfunction in patients with acute liver failure is associated with non-survival, and recent reports highlight their importance in hepatic repair. Neutrophil-targeted therapies for APAP-ALI are hampered by this controversy and a lack of time-dependent investigation.

Methods: Hepatic neutrophils were depleted at different times in a wild-type mouse model of APAP-ALI. Fpr1 ^-/- mice, with reduced neutrophil activation, were also used. The impact of neutrophil depletion was interrogated during hepatic injury and repair after APAP-ALI, using serum biochemistry, liver and blood flow cytometry, liver histopathology, immunohistochemistry, ELISA, and NanoString analysis.

Results: Neutrophils contributed both to hepatic damage and repair after APAP-ALI. Early liver necrosis was reduced by neutrophil depletion (34% to 23%, p = 0.0018, n ≥10) and by reducing neutrophil functions (39% to 29%, p = 0.0279, n ≥11). By contrast, late neutrophil depletion resulted in markedly reduced liver repair (persistent necrosis 17% to 30%, p = 0.016, and higher serum alanine aminotransferase [1,221 to 3,725 IU/l, p = 0.0007, n ≥10]) and hepatocyte proliferation (decreased minichromosomal maintenance 2+ hepatocytes, 3% to 1%, p = 0.025, n = 10). Late neutrophil depletion reduced proliferation, growth factors, and angiogenesis transcripts (Mik6 fold change [FC] -6.322, p = 0.002; Socs2 FC -2.91, p = 0.01; vascular endothelial growth factor A FC -1.48, p = 0.01; n = 3). Similar transcript changes were identified when preventing formylated peptide receptor 1-mediated neutrophil activation, along with reduced extracellular matrix remodeling (Col12a1, FC -1.99, p = 0.0001; n ≥5). Finally, depleting neutrophils resulted in a hepatic proinflammatory monocyte/macrophage phenotype during repair stages, with increased proinflammatory-related transcripts and reduced reparative transcripts.

Conclusion: Recruited neutrophils contribute not only to hepatic damage early in APAP-ALI, but also to hepatic repair through a variety of pathways, including extracellular matrix remodeling, angiogenesis, hepatocyte proliferation, and promotion of an anti-inflammatory monocyte/macrophage phenotype.

Impact and implications: Novel therapies are required for APAP-ALI to improve patient outcomes. Neutrophil products and functions are potential targets for future therapies, but current literature controversy and a lack of time-dependent studies hinder progression. This study resolves the literature controversy, showing that neutrophils have time-dependent dichotomous roles in APAP-ALI. These insights highlight that early neutrophil-targeted interventions to reduce liver damage could be detrimental to subsequent patient recovery. Therefore, future research should aim to either elucidate isolated damaging functions or harness reparative functions of neutrophils for late-stage novel therapies for APAP-ALI.

Keywords: Acetaminophen; Extracellular matrix remodeling; Formylated-peptide-receptor 1; Hepatic; Inflammation; Macrophage; Monocyte; Paracetamol.

Graphical abstract

Fig. 1

AT7519-mediated depletion of neutrophils reduces hepatic injury and repair. (A) Model schematic. (B,C) Representative Ly6G-labeled hepatic sections, displaying AT7519-reduced neutrophils, analyzed with KS (D) p = 0.0001 and (E) p = 0.0006. (F) Early neutrophil depletion improved clinical severity (t test, p = 0.0254). (G) Clinical severity (MW, p = 0.012) and weight loss (t test, WC, p = 0.014) worsened during repair. (H) ALT and AST during injury. (J) Elevated ALT (t test p = 0.0007) and AST (t test, WC, p = 0.0001) during repair. (I) Early neutrophil depletion reduced necrosis (t test, p = 0.0018). (K) Necrosis increased during repair (MW, p = 0.016). (L,M) Representative H&E hepatic sections showing necrosis around CVs. N ≥10 in all cases; scale bars: 100 μm. ALT, alanine aminotransferase; AST, aspartate aminotransferase; CV, central veins; KS, Kolmogorov–Smirnov; WC, Welch’s correction; MW, Mann-Whitney. In all instances: ∗p <0.05, ∗∗p <0.005, ∗∗∗p <0.0005, ∗∗∗∗p <0.0001.

Fig. 2

Circulating and hepatic neutrophils are activated following APAP with prolonged survival. (A) Model schematic. (B) Blood neutrophil CD11b expression (KW, p = 0.1899). (C) Decreased blood neutrophil CD62L expression, following APAP (KW, p = 0.0456, 24 h vs. 36 h, p = 0.0344). (D,E) Representative hepatic neutrophil flow cytometry. (F) Increased neutrophil CD11b following APAP (KW, p = 0.00008; 0 vs. 16, p = 0.0015) and (G) decreased CD62L expression (KW, p = 0.00088; 0 h vs. 16 h, p = 0.0126; 0h vs. 24 h, p = 0.0086. (H) Representative Ly6G and CC3 hepatic sections at 36 h post APAP administration and in controls. (I) Magnifications showing CC3+ (pink arrows) and CC3- neutrophils (orange arrows). (J) Decreased percentage of CC3+ neutrophils (KS, p = 0.0193). Scale bars: 50 μm; n ≥5. APAP, acetaminophen (paracetamol); KS, Kolmogorov-Smirnov; KW, Kruskal–Wallis. In all instances: ∗p <0.05, ∗∗p <0.005.

Fig. 3

Fpr1^-/- neutrophils are less activated with delayed hepatic recruitment. (A) Model schematic. (B) Representative Ly6G-labeled hepatic sections, showing reduced neutrophil recruitment (KS, p = 0.036). (C–E) Representative 16 h blood neutrophil flow cytometry histograms of WT (gray) and Fpr1^-/- (colored) activation markers and quantification, with reduced Fpr1^-/- neutrophil surface MPO (t test, WC, p = 0.041). (F–H) Hepatic neutrophil activation flow cytometry: (F) CD11b and (G) reduced CD62L shedding (t test, p = 0.0134), and (H) decreased surface MPO (t test, p = 0.0258). (I) Representative 24 h Ly6G, and MPO-labeled hepatic sections, showing neutrophils with (white circles) and without (blue circle) surface MPO. (J) Decreased Fpr1^-/- neutrophil MPO (t test, p = 0.0214). Scale bars: 100 μm (IHC), 50 μm (IF). FPR1, formylated peptide receptor 1; IF, immunofluorescence; IHC, immunohistochemistry; KS, Kolmogorov–Smirnov; MPO, myeloperoxidase; WC, Welch’s correction. In all instances: ∗p <0.05.

Fig. 4

Fpr1^-/- activated neutrophils contribute to both hepatic injury and repair. (A) Model schematic. (B) Representative H&E hepatic sections showing necrosis. (C) Reduced necrosis in Fpr1^-/- mice (at 16 h; MW, p = 0.0279, and 24 h; t test, p = 0.0072) (black asterisks). Resolved WT hepatic necrosis at 16–48 h (ANOVA and Tukey's test; p = 0.0164) and 24–48 h (ANOVA and Tukey's test; p = 0.0078) (pink asterisks), not present in Fpr1^-/- mice. (D) Lower clinical severity (t test, p = 0.018), ALT (MW, p = 0.0061), and GLDH (KS, p = 0.0258) in Fpr1^-/- mice during injury. (E) Lower 24-h clinical severity (MW, p = 0.0052). (F) Weight loss (MW, p = 0.2475) and GLDH (t test, p = 0.1594) at 48 h, during repair; n ≥9. Scale bars: 100 μm. ALT, alanine aminotransferase; FPR1, formylated peptide receptor 1; GLDH, glutamate dehydrogenase; KS, Kolmogorov–Smirnov; MW, Mann-Whitney; WT, wild type. In all instances: ∗p <0.05, ∗∗p <0.005.

Fig. 5

Depleting hepatic neutrophils reduces hepatic proliferation, growth factors and angiogenesis. (A) Experimental schematic. (B) ROSALIND heatmap of 81 transcripts altered between vehicle and AT7519 neutrophil-depleted livers (FC ≥1.25 or ≤-1.25). (C) Cell cycle-related genes showing upregulated cell death transcripts and downregulated proliferation transcripts. (D) Representative CC3 and Ly6G-labeled hepatic sections and magnifications showing increased CC3 following neutrophil depletion (pink arrows), quantified in (E) (MW, p = 0.015). (F) Decreased percentage of MCM2+ hepatocytes (MW, p = 0.025). (G) Representative MCM2-labeled hepatic sections; n = 10. (H) Reduced hepatic growth factor and angiogenesis-related transcripts after neutrophil depletion; n = 3. Scale bars: 50 μm. FC, fold change; MCM2, minichromosomal maintenance 2; MW, Mann-Whitney. In all instances: ∗p <0.05.

Fig. 6

Preventing FPR1 neutrophil activity reduces ECM remodeling. Graphical representations (nSolver) of hepatic NanoString mouse myeloid panel analysis from WT and Fpr1^-/- mice 24 h after APAP administration. (A) Volcano plot highlighting p <0.05 results between the two groups (purple data points). (B) Heat map of NanoString nSolver pathway analysis. (C–F) nSolver pathway scores from PCA of each covariate: (C) angiogenesis, (D) complement activation, (E) growth factor signaling, and (F) ECM remodeling. (G) Heat map of ECM-related probes: WT (orange), Fpr1^-/- (gray), and mRNA FC column z score (blue, low; orange, high). Right-hand side shows individual mouse IDs; n ≥5/group. APAP, acetaminophen (paracetamol); ECM, extracellular matrix; FPR1, formylated peptide receptor 1; PCA, principal component analysis; WT, wild type.

Fig. 7

Depleting neutrophils results in a proinflammatory monocyte phenotype during repair. Results from 36-h APAP-ALI mice with AT7519-neutrophil depletion or vehicle control. (A–E) MSD® multiplexed ELISA hepatic tissue cytokine concentrations. Neutrophil depletion (A) increased KCGRO/CXCL1 (MW, p = 0.01), and decreased (B) IFNγ (MW, p = 0.03), (C) IL1β (KS, p = 0.003), (D) IL2 (WC, p = 0.034), and (E) IL4 (MW, p = 0.062); n = 10 in all instances. (F) ROSALIND heatmap of altered hepatic monocyte/macrophage-related transcripts. (G) NanoString protein known and predicted interactions (nodes represent proteins, edges represent interactions). (H) Upregulated proinflammatory-related transcripts: Angpt2, Cd14, Plaur, and Fabp4; n = 3. (I) Downregulated anti-inflammatory-related transcripts: Socs2, Apoe, and Igf1; n = 3. APAP-ALI, acetaminophen (paracetamol)-induced acute liver injury; CXCL1, C-X-C motif chemokine ligand 1; KS, Kolmogorov–Smirnov; MSD, Meso Scale Discovery; MW, Mann-Witney. In all instances: ∗p <0.05, ∗∗p <0.005.