BLT1 signalling protects the liver against acetaminophen hepatotoxicity by preventing excessive accumulation of hepatic neutrophils

Leukotriene B4 (LTB4) is a potent chemoattractant for neutrophils. Signalling of LTB4 receptor type 1 (BLT1) has pro-inflammatory functions through neutrophil recruitment. In this study, we investigated whether BLT1 signalling plays a role in acetaminophen (APAP)-induced liver injury by affecting inflammatory responses including the accumulation of hepatic neutrophils. BLT1-knockout (BLT1−/−) mice and their wild-type (WT) counterparts were subjected to a single APAP overdose (300 mg/kg), and various parameters compared within 24 h after treatment. Compared with WT mice, BLT1−/− mice exhibited exacerbation of APAP-induced liver injury as evidenced by enhancement of alanine aminotransferase level, necrotic area, hepatic neutrophil accumulation, and expression of cytokines and chemokines. WT mice co-treated with APAP and ONO-0457, a specific antagonist for BLT1, displayed amplification of the injury, and similar results to those observed in BLT1−/− mice. Hepatic neutrophils in BLT1−/− mice during APAP hepatotoxicity showed increases in the production of reactive oxygen species and matrix metalloproteinase-9. Administration of isolated BLT1-deficient neutrophils into WT mice aggravated the liver injury elicited by APAP. These results demonstrate that BLT1 signalling dampens the progression of APAP hepatotoxicity through inhibiting an excessive accumulation of activated neutrophils. The development of a specific agonist for BLT1 could be useful for the prevention of APAP hepatotoxicity.


Results
BLT 1 signalling deficiency exacerbates APAP-induced liver injury. To explore the role of BLT 1 signalling in APAP hepatotoxicity, BLT 1 -deficient (BLT 1

−/−
) and wild-type (WT) mice were treated with 300 mg/kg APAP by intraperitoneal (i.p.) injection. We first investigated the impact of BLT 1 signalling deficiency on APAPinduced mortality in both BLT 1 −/− and WT mice, followed by monitoring the survival rate of the mice every 12 h until 96 h after drug administration. The survival rate of APAP-treated BLT 1 −/− mice was significantly lower than that of APAP-treated WT mice throughout the observation period ( Fig. 1A). At 96 h after APAP injection, all WT mice remained alive, whereas the survival rate of BLT 1 −/− mice was 32.4%. Taken together, these findings suggest that disruption of BLT 1 signalling renders mice more susceptible to APAP hepatotoxicity and mortality. Because more than half of BLT 1 −/− mice were dead within 48 h after APAP administration, we compared WT and BLT 1 −/− mice within 24 h after APAP treatment in subsequent experiments. We next examined the relative role of BLT 1 signalling in APAP hepatotoxicity by performing biochemical and histological evaluation of the liver injury. The serum level of aminotransferase (ALT) was markedly elevated in WT mice at 6 and 24 h after APAP treatment (Fig. 1B). By contrast, the ALT level at 24 h in BLT 1 −/− mice sustained high (2.3-fold increase) in comparison with that in WT mice. Histological analysis also revealed the presence of typical centrilobular necrosis in WT mice after APAP administration (Fig. 1C). The hepatic necrotic area was more extensive (23% increase) in BLT 1 −/− mice than in WT mice (Fig. 1D). These results indicate that BLT 1 −/− mice displayed sustained liver injury elicited by APAP, and that BLT 1 signalling plays an important role in APAP hepatotoxicity. In addition, substantial hepatic haemorrhagic necrosis at 24 h was observed in BLT 1 −/− mice, but not in WT mice (Fig. 1C). Indeed, quantitative analysis revealed that BLT 1 −/− mice had a 2.5-fold larger area of hepatic haemorrhage at 24 h than WT mice (Fig. 1E). These results suggest that the endogenous LTB 4 /BLT 1 signalling pathway is essential for protecting liver sinusoidal endothelial cells (LSECs) from injury in response to APAP.

BLT 1 signalling deficiency enhances the accumulation of hepatic neutrophils after APAP treatment.
Because BLT 1 signalling is essential for the recruitment of neutrophils into inflamed tissues, we determined the numbers of accumulated neutrophils in the liver during APAP hepatotoxicity. Immunofluorescence staining analysis with an antibody specific for the neutrophil marker, Gr-1, revealed fewer hepatic Gr-1-positive cells in both WT and BLT 1 −/− mice at 0 h (data not shown). In WT mice, hepatic neutrophils accumulated into the necrotic area over time after APAP administration ( Fig. 2A). Most Gr-1-positive cells were located in the injured lesions (Figs 2A and S1A). Unexpectedly, the accumulation of hepatic neutrophils in BLT 1 −/− mice was extensive ( Fig. 2A). Quantitative analysis revealed that the number of recruited hepatic neutrophils was 2.9-fold higher in BLT 1 −/− mice than in WT mice (Fig. 2B). The number of extra-sinusoidal Gr-1-positive cells at 24 h was higher (4.7-fold increase) in BLT 1 −/− mice than in WT mice (Figs 2C and S1B). Approximately 60% of neutrophils were situated outside of sinusoids.
Neutrophils accumulated into the livers of BLT 1 −/− mice as well as WT mice ( Fig. 2A), and we determined the levels of chemokine/chemokine receptors during APAP hepatotoxicity. RT-PCR analysis demonstrated that the hepatic level of chemokine (C-X-C motif) ligand 2 (CXCL2), but not of CXCL1, was higher in BLT 1 −/− mice than in WT mice (Fig. 2D,E). Additionally, the hepatic levels of CXC receptor 1 (CXCR1) and CXCR2 at 24 h were up-regulated in BLT 1 −/− mice in comparison with WT mice (Fig. 2F,G).
The BLT 1 receptor antagonist ONO-4057 magnifies APAP-induced liver injury. To further validate the role of BLT 1 in APAP hepatotoxicity, WT mice were treated with a BLT 1 receptor antagonist, ONO-4057, concomitantly with APAP administration. ONO-4057 exacerbated APAP-induced liver injury as evidenced by elevation of the ALT level (1.7-fold), necrotic area (18%) and haemorrhagic area (2.6-fold) in comparison with vehicle treatment (Fig. 3A-D). Compared with that in the livers of vehicle-treated mice, the number of accumulated Gr-1-positive cells in the livers of ONO-0457-treated mice was increased by 2.0-fold (Fig. 3E,F), which was associated with enhanced mRNA hepatic levels of CXCL2, CXCR1 and CXCR2 (Fig. 3G-J). These findings Scientific RepoRts | 6:29650 | DOI: 10.1038/srep29650 indicated that pharmacological inhibition of BLT 1 signalling as well as deletion of the BLT 1 gene exacerbates APAP-induced liver injury. Induction of BLT 1 and 5-LOX expression during APAP hepatotoxicity. The level of BLT 1 mRNA expression in the livers of WT mice was increased at 6 and 24 h after APAP administration by 2.6-and 2.4-fold, respectively (Fig. S2A). There were no differences in the level of BLT 2 between the groups (Fig. S2B). The hepatic  5-LOX level in WT mice was also elevated, with a peak at 6 h (an approximately 4-fold increase), and gradually reduced at 24 h (Fig. S2C). BLT 1 −/− mice exhibited similar changes to WT mice at 6 h. The 5-LOX level in the livers of BLT 1 −/− mice remained high at 24 h, but there was no difference between the phenotypes. To further examine the cellular source of LTB 4 during APAP hepatotoxicity, fluorescein immunostaining for 5-LOX was performed. Immunofluorescence staining of liver sections of WT mice with an antibody against 5-LOX labelled sinusoids and injured hepatocytes (Fig. S2D). Immunofluorescence double staining of liver sections of WT mice with antibodies against 5-LOX and CD31, CD68, Gr-1 or albumin revealed that 5-LOX in WT livers was expressed mainly in LSECs (CD31-positive cells) (Fig. S2E) and injured hepatocytes (albumin-positive cells) (Fig. S1F), and to a lesser extent in macrophages (CD68-positive cells) (Fig. S2G) and neutrophils (Gr-1-positive cells) (Fig. S2H). These results suggest that LTB 4 is derived mainly from LSECs and hepatocytes, and partly from macrophages and neutrophils.
Disruption of BLT 1 signalling does not affect APAP metabolism. To rule out the possibility that exacerbated liver injury in BLT 1 −/− mice resulted from an altered capacity to metabolically activate APAP, we examined the time course of hepatic CYP2E1 mRNA expression. Hepatic expression of CYP2E1, which is involved in APAP bioactivation, did not differ between the groups (Fig. S3A). APAP overdose triggers liver injury by depletion of intracellular GSH, which in turn causes oxidative stress. We next compared the level of GSH between APAP-treated BLT 1 −/− and WT mice. The hepatic GSH concentration was significantly decreased in both BLT 1 −/− and WT mice after APAP treatment (Fig. S3B). However, there was no significant difference in the hepatic GSH concentration between the two groups of mice. The same was true for the glutathione disulphide (GSSG) level (Fig. S3C) and the ratio of GSH to GSSG (Fig. S3D). These results suggest that both groups had a similar capacity to metabolically activate APAP and that impaired BLT 1 signalling does not affect APAP metabolism.

Hepatic neutrophil activation in BLT 1
−/− mice during APAP hepatotoxicity. Substantial APAP-induced liver injury in BLT 1 −/− mice is correlated with enhanced expression of pro-inflammatory mediators and excessive accumulation of hepatic neutrophils. Neutrophil activation is essential for the aggravation of the injury. To examine the activation status of neutrophils recruited in the liver, we performed flow cytometric analysis. The number of Ly6G hi /CD11b hi cells was higher (3.3-fold) in BLT 1 −/− mice than in WT mice (Fig. 5A,B). Additionally, we examined the production of reactive oxygen species (ROS) in neutrophils (Fig. 5C). The number of ROS-producing Ly6G hi /CD11b hi cells at 24 h was higher (3.0-fold increase) in BLT 1 −/− mice than in WT mice (Fig. 5D). These results indicated that the accumulation of activated neutrophils into the liver was more extensive in BLT 1 −/− mice than in WT mice. Because neutrophil-mediated cytotoxicity is at least in part caused by the CD11b/intercellular adhesion molecule (ICAM)-1 pathway, we determined the hepatic mRNA expression of CD11b and ICAM-1. At 24 h after APAP administration, the mRNA levels of CD11b and ICAM-1 were higher in the livers of BLT 1 −/− mice than in those of WT mice (Fig. 5E,F).

Neutrophils from BLT 1 −/− mice exacerbate APAP-induced liver injury in WT mice.
To further examine the contribution of BLT 1 signalling in neutrophils to APAP-induced liver injury, we performed adoptive transfer experiments of purified BLT 1 -deficient neutrophils into WT mice treated with APAP. The isolated BLT 1 -deficient neutrophils from bone marrow (BM) were intravenously injected at 6 h after APAP administration 29 . Administration of BLT 1 -deficient neutrophils into WT mice significantly aggravated APAP-induced liver injury, as evidenced by increases in the ALT level and hepatic necrotic area in comparison with WT neutrophil-treated WT mice (Fig. 6A,B). The number of hepatic Gr-1-positive cells was higher in BLT 1 -deficient neutrophil-transferred WT mice than in WT neutrophil-transferred WT mice (Fig. 6C).
Hepatic neutrophils from BLT 1 −/− mice enhance the expression of MMP-9. Recruited neutrophils in the livers of BLT 1 −/− mice were correlated with APAP-induced liver injury. Because up-regulated MMP-9 can damage LSECs during APAP hepatotoxicity 30,31 , and MMP-9 released from neutrophils would be involved in sinusoidal injury during hepatitis 32 , we determined whether hepatic accumulated neutrophils during APAP hepatotoxicity enhanced the expression of MMP-9. Immunofluorescence analysis revealed that Gr-1-positive cells (neutrophils) in the liver were co-stained with MMP-9. The number of cells double-positive for Gr-1 and MMP-9 was higher in BLT 1 −/− mice than in WT mice (Fig. 7A). We further examined whether neutrophils in BLT 1 −/− mice up-regulate the expression of MMP-9 during APAP hepatotoxicity. To this end, isolated neutrophils from BM were stimulated with LTB 4 . Application of LTB 4 enhanced MMP-9 mRNA expression in neutrophils from BLT 1 −/− mice in comparison with WT mice (Fig. 7B). Treatment of isolated BLT1-deficient neutrophils with LTB 4 also increased the mRNA level of CXCR1 (Fig. 7C), but not of CXCR2 (Fig. 7D). These results suggest that activated neutrophils in BLT 1 −/− mice during APAP hepatotoxicity would damage LSECs through releasing ROS or the inflammatory mediator MMP-9.  hepatotoxicity, BLT 1 −/− mice were treated with a neutralising antibody against CXCL2. The number of Gr-1-positive cells was reduced by 35% in comparison with vehicle treatment (Fig. S4A). Concomitantly, the level of ALT in antibody-treated mice was decreased by 30% in comparison with vehicle-treated mice (Fig. S4B). We also examined the effect of the anti-CXCL2 antibody on pro-inflammatory mediators. RT-PCR analysis revealed that the hepatic levels of TNFα , IL-1β , IL-6 and MMP-9 in antibody-treated BLT 1 −/− mice were reduced by 82%, 85%, 44% and 82%, respectively, in comparison with vehicle-treated BLT 1 −/− mice (Fig. S4C-F).

Effects of an anti-CXCL2 antibody on APAP-induced liver injury in BLT
A minor role of BLT 2 in APAP hepatotoxicity. The finding that BLT 1 −/− mice displayed accelerated liver injury elicited by APAP suggests that signalling other than BLT 1 signalling is involved in the injury. Because BLT 2 is another receptor for LTB 4 , we treated BLT 1 −/− mice with the BLT 2 -prone antagonist, LY255283, to examine whether BLT 2 signalling is responsible for exacerbated APAP hepatotoxicity in BLT 1 −/− mice. LY255283 failed to reduce the level of ALT in APAP-treated BLT 1 −/− mice in comparison with vehicle administration (0.75% dimethyl sulfoxide (DMSO)) (Fig. S5). These results indicate that BLT 2 signalling is unlikely to be involved in the enhancement of APAP-induced liver injury in BLT 1 −/− mice.

Discussion
In the present study, we demonstrated that genetic deletion or pharmacological inhibition of the BLT 1 signalling pathway exacerbated APAP-induced liver injury. Additionally, BLT 1 −/− mice were susceptible to APAP hepatotoxicity and displayed lower survival rates. Herein, we report the novel observation that BLT 1 signalling unmasked potent preventive effects on liver injury elicited by APAP overdose. Specifically, the present findings reveal that, in contrast to WT mice, BLT 1 deficiency promoted the increased recruitment of activated neutrophils into the livers of APAP-treated mice. In addition, BLT 1 -deficient neutrophils exhibited an enhanced ability to produce ROS and MMP-9 in comparison with WT neutrophils. These results indicate that BLT 1 signalling protects the liver from APAP hepatotoxicity through inhibiting the excessive accumulation of neutrophils into the liver.
BLT 1 signalling exerts pro-inflammatory actions through the enhancement of neutrophil recruitment in several inflammatory diseases 19,[26][27][28] . Indeed, BLT 1 −/− mice are protected from the development of rheumatoid arthritis 26 , spinal cord injury 27 and atopic dermatitis 28 , which are associated with decreases in the accumulation of neutrophils into inflammatory sites and in the production of chemokines and cytokines. These studies also indicate that BLT 1 signalling in the infiltrated neutrophils is essential for the development of these inflammatory diseases. In liver disease models, BLT 1 signalling is responsible for liver microcirculatory dysfunction including leukocyte adhesion during endotoxemia 33 . During warm hepatic ischemia/reperfusion (I/R), BLT 1 −/− mice exhibit less accumulation of neutrophils in the liver; however, there is no significant difference in the degree of hepatic I/R injury between the phenotypes 34 . Collectively, it is suggested that BLT 1 signalling mediates neutrophil infiltration at local inflammatory sites to enhance inflammation and injury. BLT 1 signalling has potent pro-inflammatory effects in inflammatory diseases 19 . By contrast, the current study demonstrated that BLT 1 signalling suppressed the excessive infiltration of neutrophils into inflammatory sites to protect the liver from APAP hepatotoxicity. BLT 1 signalling in neutrophils has a protective action against the progression of APAP overdose-induced liver injury. Thus, BLT 1 signalling exerts anti-inflammatory actions under certain pathological conditions such as APAP hepatotoxicity, while pro-inflammatory effects of BLT 1 in other inflammatory disease models have been established.
APAP hepatotoxicity triggers an extensive inflammatory response with cytokine formation 4,7 . The inflammatory response to APAP depends on the injury insult. Our data demonstrated that enhanced APAP-induced liver injury in BLT 1 −/− mice was associated with up-regulated expression of pro-inflammatory cytokines including TNFα , IL-1β and IL-6 ( Fig. 4A-C). Aggravated liver injury is also correlated with increases in the accumulation of neutrophils into the injured liver. The extensive hepatic centrilobular necrosis during APAP hepatotoxicity results in extensive release of damage-associated molecular patterns (DAMPs) from cells undergoing necrosis 4,35,36 . DAMPs trigger the generation of pro-inflammatory cytokines from macrophages, which leads to the recruitment of neutrophils into the injured liver 4 . As a result, APAP hepatotoxicity causes inflammatory responses including generation of pro-inflammatory mediators 7,11 . Enhanced liver injury through generation of pro-inflammatory mediators would cause further neutrophil accumulation into the injured liver. Therefore, attenuated hepatic necrosis during APAP toxicity will reduce DAMP release and consequently reduce cytokine formation and neutrophil infiltration. The levels of pro-inflammatory mediators reflect the degree of neutrophil activation and accumulation in the liver. However, it remains to be elucidated whether neutrophil recruitment is a cause or consequence of the exacerbation of APAP-induced liver injury 4,6,8,9,29 .
The role of neutrophils in the evolution of APAP-induced liver injury is controversial. Depletion of neutrophils by treatment with an anti-Gr-1 antibody and pharmacological blockage of neutrophil chemotactic receptor (CXCR2) protect mice against APAP-induced liver injury 8 . These findings suggest that neutrophils have deleterious functions in the innate immune system during APAP hepatotoxicity. By contrast, numerous interventions that prevent neutrophil cytotoxicity, such as blocking antibodies against CD18, gene deficiency of CD18 and ICAM-1, inhibitors of NADPH oxidase, and deficiency of NADPH oxidase, are ineffective in attenuating APAP hepatotoxicity 4,9 . This indicates that accumulation of neutrophils is not responsible for liver injury elicited by APAP. The results of the present study showed that BLT 1 signalling protects the liver from APAP toxicity through inhibiting the excessive accumulation of activated neutrophils into the liver, suggesting that BLT 1 signalling deficiency in neutrophils would contribute to the development of APAP hepatotoxicity.
According to the results by Liu et al. 6 , depletion of neutrophils by treatment with an anti-Gr-1 antibody protects the liver from APAP (500 mg/kg) toxicity, which is a higher dose of APAP used in the present study. This suggests that accumulated hepatic neutrophils are essential for the development and severity of liver injury elicited by even a high dose of APAP (500 mg/kg). Because the severity of liver injury elicited by APAP is dose-dependent, the neutrophil response to APAP also appears to be in a dose-dependent manner. In addition, the survival rate of BLT 1 −/− mice treated with APAP (300 mg/kg) was 48.6% by 72 h (Fig. 1A), while that of WT mice (C57Bl/6 mice) treated with APAP (500 mg/kg) was 58% by 72 h 6 . Based on these observations, it may be conceivable that the mortality of BLT 1 −/− mice treated with APAP (500 mg/kg) would be quite severe, and it would be hard to evaluate the role of neutrophils in BLT 1 −/− mice treated with higher doses of APAP. Therefore, we selected the moderate dose of APAP (300 mg/kg) in the current study.
The priming or activation of neutrophils, as evidenced by reactive oxygen formation, is required for neutrophil-mediated liver injury 4 . Activated neutrophils in inflammation exert their cytotoxic effects through the generation of ROS. However, it is indicated that hepatic neutrophils are not activated in terms of ROS production during APAP hepatotoxicity 4,9 , suggesting that neutrophils recruited into the injured liver play a minor role in the aggravation of liver injury elicited by APAP. By contrast, the present study revealed that BLT 1 -deficient neutrophils (Ly6G hi /CD11b hi ) recruited into the injured liver at 24 h after APAP treatment displayed higher levels of ROS production than WT neutrophils (Fig. 5C,D), indicating that neutrophils from BLT 1 −/− mice are activated at 24 h post-APAP administration. Activated BLT 1 -deficient neutrophils in the liver were associated with enhanced liver injury elicited by APAP, indicating that BLT 1 signalling contributes to prevention of APAP hepatotoxicity through the inactivation of neutrophils recruited into the injured liver. Of interest, adoptive transfer of BLT 1 -deficient neutrophils into WT mice treated with APAP exacerbated liver injury (Fig. 6A,B), indicating that BLT 1 signalling in neutrophils has a protective action against liver injury elicited by APAP overdose. The current study also revealed that extensive accumulation of neutrophils in the livers of BLT 1 −/− mice treated with APAP was associated with up-regulated hepatic expression of CD11b and ICAM-1 (Fig. 5E,F). ICAM-1 is up-regulated in LSECs and is induced in hepatocytes in areas of parenchymal injury, where ICAM-1 induction correlates with the degree of liver damage 37 . Up-regulation of ICAM-1 is essential for neutrophil extravasation into the extra-sinusoidal space to attack hepatocytes 38 . However, a deficiency in CD18 on the surface of neutrophils or in ICAM-1 has no impact on APAP-induced liver injury in mice 4 .
Taken together, hepatic neutrophil recruitment is a characteristic feature of immune responses in APAP hepatotoxicity; however, diverging results concerning the role of neutrophils in APAP-induced liver injury need to be resolved. Investigating the association between neutrophils and other cell types of the immune system, such as Scientific RepoRts | 6:29650 | DOI: 10.1038/srep29650 resident or infiltrating macrophages and monocytes, might provide new insights into the contribution to innate immunity during APAP hepatotoxicity 12 .
Haemorrhagic hepatic necrosis is a characteristic of liver injury elicited by APAP overdose 39 . In BLT 1 −/− mice and ONO-0457-treated WT mice, recruitment of neutrophils was correlated with haemorrhagic necrosis, which could lead to animal death and a poor animal survival rate due to severe impairment in hepatic microcirculation 11,39 . Thus, it is suggested that BLT 1 signalling is critical for suppressing inflammatory responses including excessive accumulation of neutrophils and for protecting LSECs from injury in response to APAP. During APAP hepatotoxicity, LSECs are injured as evidenced by gap formation in their cytoplasm. The formation of gaps results in the penetration of erythrocytes, severe haemorrhage and liver microcirculatory dysfunction 11,39 . Although the mechanisms by which administration of APAP damages LSECs are not entirely clear 31 , we found that the production of ROS and MMP-9 by recruited neutrophils in BLT 1 −/− mice was enhanced in comparison with those in WT mice 40 . Isolated neutrophils from BLT 1 −/− mice expressed a higher level of MMP-9 in response to LTB 4 in cell culture (Fig. 7B). Excessive ROS and MMPs could be responsible for injury to sinusoids through proteolytic cleavage of the LSEC membrane [30][31][32] . It is also suggested that MMP-9 facilitates the migration of leukocytes into inflamed livers 41,42 . Disruption of BLT 1 signalling induces more severe APAP hepatotoxicity with excessive activated neutrophil recruitment, probably due to the impairment of neutrophil function maintained by BLT 1 receptors. However, the mechanisms by which BLT 1 signalling regulates the release of ROS and MMP-9 remain to be elucidated. We and others previously reported that VEGFR1 signalling is essential for maintaining the integrity of LSECs in terms of their function and structure 11,43,44 . The current study also found that VEGFR1 expression was reduced in the livers of BLT 1 −/− mice during APAP hepatotoxicity, suggesting that hepatic haemorrhage partly results from down-regulated expression of VEGFR1 in the livers of BLT 1 −/− mice. The present study demonstrates that BLT 1 deficiency and a BLT 1 antagonist exacerbated APAP-induced liver injury, as indicated by increases in the ALT level, hepatic centrilobular necrotic area, haemorrhagic area and expression of pro-inflammatory mediators (Fig. 3). BLT 1 −/− mice were susceptible to APAP hepatotoxicity and displayed lower survival rates (Fig. 1A). Thus, BLT 1 signalling negatively regulates severe liver injury elicited by APAP overdose. The current study also indicates that hepatic neutrophil activation and recruitment are essential for the development of severe liver injury in BLT 1 −/− mice after APAP administration. A key issue that remained to be addressed is how BLT 1 deficiency promotes increased neutrophil infiltration into the liver at 24 h after APAP treatment. To address this question, we determined the levels of chemokines (CXCL1 and CXCL2) and their receptors (CXCR1 and CXCR2) because chemokine gradients of CXCL1 and CXCL2 guide neutrophils through CXCR1 and CXCR2 into necrotic areas 45 . The hepatic CXCL2 level was significantly higher in BLT 1 −/− mice than in WT mice at 24 h post-APAP treatment (Fig. 2E). In addition, treatment of BLT 1 −/− mice with an anti-CXCL2 antibody attenuated inflammatory responses to APAP, including reductions in the levels of ALT, neutrophil recruitment and inflammatory mediators (Fig. S4). Although we have not examined the sources of CXCL2 during APAP hepatotoxicity, hepatocytes and Kupffer cells can generate CXC chemokines 45,46 . In addition, the mechanisms by which the absence of BLT 1 signalling enhances CXCL2 in the liver after APAP administration are not clear.
A lack of BLT 1 signalling amplified APAP-induced liver injury, suggesting that signalling other than BLT 1 signalling is involved in the enhanced liver injury. Because BLT 2 , as well as BLT 1 , is a receptor for LTB 4 , we examined whether BLT 2 contributes to the exacerbated APAP hepatotoxicity in BLT 1 −/− mice. However, the BLT 2 -prone antagonist failed to affect the exacerbated APAP hepatotoxicity in BLT 1 −/− mice (Fig. S5). Although BLT 1 signalling appears to be essential for protection against APAP hepatotoxicity, the mediator of exacerbated APAP hepatotoxicity in BLT 1 −/− mice remains to be clarified. A recent report 47 showed that a lack of the 5-LOX gene and pharmacological inhibition of 5-LOX synthase reduce APAP-induced liver injury, as demonstrated by decreases in the ALT level and hepatic necrosis. The protective effect of 5-LOX inhibition on APAP hepatotoxicity is attributed to attenuation of APAP bioactivation and oxidative stress. In their study, the effect of 5-LOX on neutrophil recruitment in response to APAP was not addressed. Because LTB 4 is synthetised through 5-LOX, our results appear to contradict their results. However, 5-LOX is high upstream in the LTB 4 /BLT 1 signalling pathway. 5-LOX produces not only LTB 4 , but also other LTs such as LTC 4 , LTD 4 and LTE 4 , and these LTs act through different receptors. Additionally, 5-LOX metabolises arachidonic acid to 5-hydroperoxyeicosatetraenoic acid, which is the precursor for LTs and a bioactive product. Thus, it is conceivable that the effects of selective inhibition of a 5-LOX-derived lipid mediator receptor on APAP-induced liver injury are different from those of the broad inhibition of 5-LOX-derived lipid mediators. Although BLT 1 inhibition is a selective approach, inhibition of BLT 1 signalling has adverse effects and is not suitable for use as a therapeutic tool to treat APAP hepatotoxicity.
In conclusion, we clarified the role of BLT 1 signalling in APAP hepatotoxicity in mice. BLT 1 signalling dampens the progression of APAP-induced liver injury through inhibiting an excessive accumulation of activated neutrophils. The present study suggests that the development of a specific agonist for BLT 1 signalling in neutrophils could be useful for the prevention of APAP hepatotoxicity.

Methods
Animals. Male C57Bl/6 WT mice (8 weeks old) were obtained from Crea Japan (Tokyo, Japan). Male BLT 1knockout (BLT 1 −/− ) mice (8 weeks old) were developed previously 48 . Mice were maintained at constant humidity (60 ± 5%) and temperature (25 ± 1 °C) on a 12 h light/dark cycle. All animals were provided food and water ad libitum. All experimental procedures were approved by the Animal Experimentation and Ethics Committee of the Kitasato University School of Medicine, and were performed in accordance with the guidelines for animal experiments set down by the Kitasato University School of Medicine.