Excessive Plasmin Compromises Hepatic Sinusoidal Vascular Integrity After Acetaminophen Overdose

Acetaminophen Overdose Reveals PAR4 as a Low-Expressing but Potent Receptor on the Hepatic Endothelium in Mice

BACKGROUND

The protease thrombin, which elicits multiple physiological and pathological effects on vascular endothelial cells (ECs), can signal through PARs (protease-activated receptors) 1 and 4. PAR1 is a high-affinity thrombin receptor known to signal on ECs, whereas PAR4 is a low-affinity thrombin receptor, and evidence for its expression and function on ECs is mixed. This study aims to exploit the high levels of thrombin generation and hepatic vascular dysfunction that occur during acetaminophen (APAP) overdose to determine (1) whether hepatic endothelial PAR4 is a functional receptor, and (2) the endothelial-specific functions for PAR1 and PAR4 in a high thrombin and pathological setting.

METHODS

We generated mice with conditional deletion of Par1/Par4 in ECs and overdosed them with APAP. Hepatic vascular permeability, erythrocyte accumulation in the liver, thrombin generation, and liver function were assessed following overdose. Additionally, we investigated the expression levels of endothelial PARs and how they influence transcription in APAP-overdosed liver ECs using endothelial translating ribosome affinity purification followed by next-generation sequencing.

RESULTS

We found that mice deficient in high-expressing endothelial Par1 or low-expressing Par4 had equivalent reductions in APAP-induced hepatic vascular instability, although mice deficient for both receptors had lower vascular permeability at an earlier timepoint after APAP overdose than either of the single mutants. Additionally, mice with loss of both endothelial Par1 and Par4 had reduced thrombin generation after APAP overdose, suggesting decreased hypercoagulability. Last, we found that endothelial PAR1-but not PAR4-can regulate transcription in hepatic ECs.

CONCLUSIONS

Low-expressing PAR4 can react similarly to high-expressing PAR1 in APAP-overdosed hepatic ECs, demonstrating that PAR4 is a potent thrombin receptor. Additionally, these receptors are functionally redundant but act divergently in their expression and ability to influence transcription in hepatic ECs.

Acetaminophen Overdose Reveals Protease-Activated Receptor 4 as a Low-Expressing but Potent Receptor on the Hepatic Endothelium

Background & Aims

Hepatic endothelial cell (EC) dysfunction and centrilobular hepatocyte necrosis occur with acetaminophen (APAP) overdose. The protease thrombin, which is acutely generated during APAP overdose, can signal through protease-activated receptors 1 and 4 (PAR1/PAR4). PAR1 is a high-affinity thrombin receptor that is known to signal on ECs, whereas PAR4 is a low-affinity thrombin receptor, and evidence for its expression and function on ECs is mixed. This study aims to exploit the high levels of thrombin generated during APAP overdose to determine (1) if hepatic endothelial PAR4 is a functional receptor, and (2) endothelial-specific functions for PAR1 and PAR4 in a high thrombin setting.

Methods

We generated mice with conditional deletion(s) of Par1/Par4 in ECs and overdosed them with APAP. Hepatic vascular permeability, erythrocyte congestion/bleeding, and liver function were assessed following overdose. Additionally, we investigated the expression levels of endothelial PARs and how they influence transcription in APAP-overdosed liver ECs using endothelial Translating Ribosome Affinity Purification followed by next-generation sequencing (TRAPseq).

Results

We found that mice deficient in high-expressing endothelial Par1 or low-expressing Par4 had equivalent reductions in APAP-induced hepatic vascular instability but no effect on hepatocyte necrosis. Additionally, mice with loss of endothelial Par1 and Par4 had reduced permeability at an earlier time point after APAP overdose when compared to mice singly deficient in either receptor in ECs. We also found that endothelial PAR1-but not PAR4-can regulate transcription in hepatic ECs.

Conclusions

Low-expressing PAR4 can react similarly to high-expressing PAR1 in APAP-overdosed hepatic ECs, demonstrating that PAR4 is a potent thrombin receptor. Additionally, these receptors are functionally redundant but act divergently in their expression and ability to influence transcription in hepatic ECs.

Fibrinolysis-Mediated Pathways in Acute Liver Injury

Acute liver injury (ALI), that is, the development of reduced liver function in patients without preexisting liver disease, can result from a wide range of causes, such as viral or bacterial infection, autoimmune disease, or adverse reaction to prescription and over-the-counter medications. ALI patients present with a complex coagulopathy, characterized by both hypercoagulable and hypocoagulable features. Similarly, ALI patients display a profound dysregulation of the fibrinolytic system with the vast majority of patients presenting with a hypofibrinolytic phenotype. Decades of research in experimental acute liver injury in mice suggest that fibrinolytic proteins, including plasmin(ogen), plasminogen activators, fibrinolysis inhibitors, and fibrin(ogen), can contribute to initial hepatotoxicity and/or stimulate liver repair. This review summarizes major experimental findings regarding the role of fibrinolytic factors in ALI from the last approximately 30 years and identifies unanswered questions, as well as highlighting areas for future research.

Interferon-γ+ Th1 activates intrahepatic resident memory T cells to promote HBsAg loss by inducing M1 macrophage polarization

The immune mechanism underlying hepatitis B surface antigen (HBsAg) loss, particularly type I inflammatory response, during pegylated interferon-α (PEG-IFN) therapy remains unclear. In this study, we aimed to elucidate such immune mechanisms. Overall, 82 patients with chronic hepatitis B (CHB), including 41 with HBsAg loss (cured group) and 41 uncured patients, received nucleos(t)ide analogue and PEG-IFN treatments. Blood samples from all patients, liver tissues from 14 patients with CHB, and hepatic perfusate from 8 liver donors were collected for immune analysis. Jurkat, THP-1 and HepG2.2.15 cell lines were used in cell experiments. The proportion of IFN-γ+ Th1 cells was higher in the cured group than in the uncured group, which was linearly correlated with HBsAg decline and alanine aminotransferase (ALT) levels during treatment. However, CD8+ T cells were weakly associated with HBsAg loss. Serum and intrahepatic levels of Th1 cell-associated chemokines (C-X-C motif chemokine ligand [CXCL] 9, CXCL10, CXCL11, IFN-γ) were significantly lower in the cured patients than in patients with a higher HBsAg quantification during therapy. Serum from cured patients induced more M1 (CD68+CD86+ macrophage) cells than that from uncured patients. Patients with chronic HBV infection had significantly lower proportions of CD86+ M1 and CD206+ M2 macrophages in their livers than healthy controls. M1 polarization of intrahepatic Kupffer cells promoted HBsAg loss by upregulating the effector function of tissue-resident memory T cells with increased ALT levels. IFN-γ+ Th1 activates intrahepatic resident memory T cells to promote HBsAg loss by inducing M1 macrophage polarization.

Robust coagulation activation and coagulopathy in mice with experimental acetaminophen-induced liver failure

BACKGROUND

Patients with acetaminophen (APAP)-induced acute liver failure (ALF) display both hyper- and hypocoagulable changes not necessarily recapitulated by standard hepatotoxic doses of APAP used in mice (eg, 300 mg/kg).

OBJECTIVES

We sought to examine coagulation activation in vivo and plasma coagulation potential ex vivo in experimental settings of APAP-induced hepatotoxicity and repair (300-450 mg/kg) and APAP-induced ALF (600 mg/kg) in mice.

RESULTS

APAP-induced ALF was associated with increased plasma thrombin-antithrombin complexes, decreased plasma prothrombin, and a dramatic reduction in plasma fibrinogen compared with lower APAP doses. Hepatic fibrin(ogen) deposits increased independent of APAP dose, whereas plasma fibrin(ogen) degradation products markedly increased in mice with experimental ALF. Early pharmacologic anticoagulation (+2 hours after 600 mg/kg APAP) limited coagulation activation and reduced hepatic necrosis. The marked coagulation activation evident in mice with APAP-induced ALF was associated with a coagulopathy detectable ex vivo in plasma. Specifically, prolongation of the prothrombin time and inhibition of tissue factor-initiated clot formation were evident even after restoration of physiological fibrinogen concentrations. Plasma endogenous thrombin potential was similarly reduced at all APAP doses. Interestingly, in the presence of ample fibrinogen, ∼10 times more thrombin was required to clot plasma from mice with APAP-induced ALF compared with plasma from mice with simple hepatotoxicity.

CONCLUSION

The results indicate that robust pathologic coagulation cascade activation in vivo and suppressed coagulation ex vivo are evident in mice with APAP-induced ALF. This unique experimental setting may fill an unmet need as a model to uncover mechanistic aspects of the complex coagulopathy of ALF.

Altered fibrinogen γ-chain cross-linking in mutant fibrinogen-γΔ5 mice drives acute liver injury

BACKGROUND

Hepatic deposition of cross-linked fibrin(ogen) occurs alongside platelet accumulation as a hallmark of acetaminophen (APAP)-induced liver injury.

OBJECTIVES

We sought to define the precise role of the fibrinogen γ-chain C-terminal integrin αIIbβ3 binding domain in APAP-induced liver injury.

METHODS

Mice expressing mutant fibrinogen incapable of engaging integrin αIIbβ3 due to a C-terminal fibrinogen γ-chain truncation (mutant fibrinogen-γΔ5 [FibγΔ5] mice) and wild-type mice were challenged with APAP (300 mg/kg, intraperitoneally).

RESULTS

We observed an altered pattern of fibrin(ogen) deposition in the livers of APAP-challenged FibγΔ5 mice. This led to the unexpected discovery that fibrinogen γ-chain cross-linking was altered in the livers of APAP-challenged FibγΔ5 mice compared with that in wild-type mice, including absence of γ-γ dimer and accumulation of larger molecular weight cross-linked γ-chain complexes. This finding was not unique to the injured liver because activation of coagulation did not produce γ-γ dimer in plasma from FibγΔ5 mice or purified FibγΔ5 fibrinogen. Sanger sequencing predicted that the fibrinogen-γΔ5 γ-polypeptide would terminate at lysine residue 406, but liquid chromatography tandem mass spectrometry analysis revealed that this critical lysine residue was absent in purified fibrinogen-γΔ5 protein. Interestingly, hepatic deposition of this uniquely aberrantly cross-linked fibrin(ogen) in FibγΔ5 mice was associated with exacerbated hepatic injury, an effect not recapitulated by pharmacologic inhibition of integrin αIIbβ3.

CONCLUSION

The results indicate that fibrinogen-γΔ5 lacks critical residues essential to form γ-γ dimer in response to thrombin and suggest that hepatic accumulation of abnormally cross-linked fibrin(ogen) can exacerbate hepatic injury.

Serine proteases mediate leukocyte recruitment and hepatic microvascular injury in the acute phase following extended hepatectomy

OBJECTIVE

Post-hepatectomy liver failure (PHLF) is the main limitation of extended liver resection. The molecular mechanism and the role of leukocytes in the development of PHLF remain to be unveiled. We aimed to address the impact of serine proteases (SPs) on the acute phase after liver resection by intravitally analyzing leukocyte recruitment and changes in hemodynamics and microcirculation of the liver.

METHODS

C57BL/6 mice undergoing 60% partial hepatectomy were treated with aprotinin (broad-spectrum SP inhibitor), tranexamic acid (plasmin inhibitor), or vehicle. Sham-operated animals served as controls. In vivo fluorescence microscopy was used to quantify leukocyte-endothelial interactions immediately after, as well as 120 min after partial hepatectomy in postsinusoidal venules, along with measurement of sinusoidal perfusion rate and postsinusoidal shear rate. Recruitment of leukocytes, neutrophils, T cells, and parameters of liver injury were assessed in tissue/blood samples.

RESULTS

Leukocyte recruitment, sinusoidal perfusion failure rate, and shear rate were significantly increased in mice after 60% partial hepatectomy compared to sham-operated animals. The inhibition of SPs or plasmin significantly attenuated leukocyte recruitment and improved the perfusion rate in the remnant liver. ICAM-1 expression and neutrophil recruitment significantly increased after 60% partial hepatectomy and were strongly reduced by plasmin inhibition.

CONCLUSIONS

Endothelial activation and leukocyte recruitment in the liver in response to the increment of sinusoidal shear rate were hallmarks in the acute phase after liver resection. SPs mediated leukocyte recruitment and contributed to the impairment of sinusoidal perfusion in an ICAM-1-dependent manner in the acute phase after liver resection.

Plasmin-mediated cleavage of high-molecular-weight kininogen contributes to acetaminophen-induced acute liver failure

Acetaminophen (APAP)-induced liver injury is associated with activation of coagulation and fibrinolysis. In mice, both tissue factor-dependent thrombin generation and plasmin activity have been shown to promote liver injury after APAP overdose. However, the contribution of the contact and intrinsic coagulation pathways has not been investigated in this model. Mice deficient in individual factors of the contact (factor XII [FXII] and prekallikrein) or intrinsic coagulation (FXI) pathway were administered a hepatotoxic dose of 400 mg/kg of APAP. Neither FXII, FXI, nor prekallikrein deficiency mitigated coagulation activation or hepatocellular injury. Interestingly, despite the lack of significant changes to APAP-induced coagulation activation, markers of liver injury and inflammation were significantly reduced in APAP-challenged high-molecular-weight kininogen-deficient (HK-/-) mice. Protective effects of HK deficiency were not reproduced by inhibition of bradykinin-mediated signaling, whereas reconstitution of circulating levels of HK in HK-/- mice restored hepatotoxicity. Fibrinolysis activation was observed in mice after APAP administration. Western blotting, enzyme-linked immunosorbent assay, and mass spectrometry analysis showed that plasmin efficiently cleaves HK into multiple fragments in buffer or plasma. Importantly, plasminogen deficiency attenuated APAP-induced liver injury and prevented HK cleavage in the injured liver. Finally, enhanced plasmin generation and HK cleavage, in the absence of contact pathway activation, were observed in plasma of patients with acute liver failure due to APAP overdose. In summary, extrinsic but not intrinsic pathway activation drives the thromboinflammatory pathology associated with APAP-induced liver injury in mice. Furthermore, plasmin-mediated cleavage of HK contributes to hepatotoxicity in APAP-challenged mice independently of thrombin generation or bradykinin signaling.

Regulation of macrophage activation in the liver after acute injury: Role of the fibrinolytic system

The liver functions, in part, to prevent exposure of the body to potentially harmful substances ingested in the diet. While it is highly efficient at accomplishing this, it is frequently prone to liver injury due to the biotransformation of xenobiotics into toxic metabolites. To counter this injury, the liver has evolved a unique capacity to rapidly and efficiently repair itself. Successful resolution of acute liver injury relies on hepatic macrophage populations that orchestrate the reparative response. After injury, Kupffer cells, the resident macrophages of the liver, become activated and secrete proinflammatory cytokines. These cytokines recruit other immune cells, including monocyte-derived macrophages, to the liver where they contribute to the repair process. Monocyte-derived macrophages traffic into the necrotic foci where they rapidly phagocytose dead cell debris. Simultaneous with this process, these cells change phenotype from a proinflammatory macrophage to a pro-restorative macrophage that produce pro-mitogenic growth factors and anti-inflammatory cytokines. Ultimately this process triggers resolution of inflammation, and along with proliferation of other hepatic cells, restores the liver architecture and function. While the mechanisms regulating specific macrophage functions during repair remain to be elucidated, recent studies indicate a key role for the fibrinolytic system in coordinating macrophage function during repair. In this review, we will highlight the function and role of hepatic macrophages in repair after acute liver injury, and will discuss the role of the fibrinolytic enzyme, plasmin, in regulation of these various processes.

Proinflammatory Mediators, IL (Interleukin)-1β, TNF (Tumor Necrosis Factor) α, and Thrombin Directly Induce Capillary Tube Regression

OBJECTIVE

In this work, we examine the molecular basis for capillary tube regression and identify key proregressive factors, signaling pathways, and pharmacological antagonists of this process. Approach and Results: We demonstrate that the proinflammatory mediators, IL (interleukin)-1β, TNF (tumor necrosis factor) α, and thrombin, singly and in combination, are potent regulators of capillary tube regression in vitro. These proregressive factors, when added to endothelial cell-pericyte cocultures, led to selective loss of endothelial cell-lined tube networks, with retention and proliferation of pericytes despite the marked destruction of adjacent capillary tubes. Moreover, treatment of macrophages with the TLR (toll-like receptor) agonists Pam3CSK4 and lipopolysaccharide generates conditioned media with marked proregressive activity, that is completely blocked by a combination of neutralizing antibodies directed to IL-1β and TNFα but not to other factors. The same combination of blocking antibodies, as well as the anti-inflammatory cytokine IL-10, interfere with macrophage-dependent hyaloid vasculature regression in mice suggesting that proinflammatory cytokine signaling regulates capillary regression in vivo. In addition, we identified a capillary regression signaling signature in endothelial cells downstream of these proregressive agents that is characterized by increased levels of ICAM-1 (intercellular adhesion molecule-1), phospho-p38, and phospho-MLC2 (myosin light chain-2) and decreased levels of phospho-Pak2, acetylated tubulin, phospho-cofilin, and pro-caspase3. Finally, we identified combinations of pharmacological agents (ie, FIST and FISTSB) that markedly rescue the proregressive activities of IL-1β, TNFα, and thrombin, individually and in combination.

CONCLUSIONS

Overall, these new studies demonstrate that the major proinflammatory mediators, IL-1β, TNFα, and thrombin, are key regulators of capillary tube regression-a critical pathological process regulating human disease.

Dichotomous Role of Plasmin in Regulation of Macrophage Function after Acetaminophen Overdose

Kupffer cells and monocyte-derived macrophages are critical for liver repair after acetaminophen (APAP) overdose. These cells produce promitogenic cytokines and growth factors, and they phagocytose dead cell debris, a process that is critical for resolution of inflammation. The factors that regulate these dynamic functions of macrophages after APAP overdose, however, are not fully understood. We tested the hypothesis that the fibrinolytic enzyme, plasmin, is a key regulator of macrophage function after APAP-induced liver injury. In these studies, inhibition of plasmin in mice with tranexamic acid delayed up-regulation of proinflammatory cytokines after APAP overdose. In culture, plasmin directly, and in synergy with high-mobility group B1, stimulated Kupffer cells and bone marrow-derived macrophages to produce cytokines by a mechanism that required NF-κB. Inhibition of plasmin in vivo also prevented trafficking of monocyte-derived macrophages into necrotic lesions after APAP overdose. This prevented phagocytic removal of dead cells, prevented maturation of monocyte-derived macrophages into F4/80-expressing macrophages, and prevented termination of proinflammatory cytokine production. Our studies reveal further that phagocytosis is an important stimulus for cessation of proinflammatory cytokine production as treatment of proinflammatory, monocyte-derived macrophages, isolated from APAP-treated mice, with necrotic hepatocytes decreased expression of proinflammatory cytokines. Collectively, these studies demonstrate that plasmin is an important regulator of macrophage function after APAP overdose.