Acetaminophen-Induced Rat Hepatotoxicity Based on M1/M2-Macrophage Polarization, in Possible Relation to Damage-Associated Molecular Patterns and Autophagy

Overdose of acetaminophen (APAP), an antipyretic drug, is an important cause of liver injury. However, the mechanism in the rat model remains undetermined. We analyzed APAP-induced hepatotoxicity using rats based on M1/M2-macrophage functions in relation to damage-associated molecular patterns (DAMPs) and autophagy. Liver samples from six-week-old rats injected with APAP (1000 mg/kg BW, ip, once) after 15 h fasting were collected at hour 10, and on days 1, 2, 3, and 5. Liver lesions consisting of coagulation necrosis and inflammation were seen in the affected centrilobular area on days 1 and 2, and then, recovered with reparative fibrosis by day 5. Liver exudative enzymes increased transiently on day 1. CD68+ M1-macrophages increased significantly on days 1 and 2 with increased mRNAs of M1-related cytokines such as IFN-γ and TNF-α, whereas CD163+ M2-macrophages appeared later on days 2 and 3. Macrophages reacting to MHC class II and Iba1 showed M1-type polarization, and CD204+ macrophages tended to be polarized toward M2-type. At hour 10, interestingly, HMGB1 (representative DAMPs) and its related signals, TLR-9 and MyD88, as well as LC3B+ autophagosomes began to increase. Collectively, the pathogenesis of rat APAP hepatotoxicity, which is the first, detailed report for a rat model, might be influenced by macrophage functions of M1 type for tissue injury/inflammation and M2-type for anti-inflammatory/fibrosis; particularly, M1-type may function in relation to DAMPs and autophagy. Understanding the interplayed mechanisms would provide new insight into hepato-pathogenesis and contribute to the possible development of therapeutic strategies.


Introduction
The liver can metabolize various kinds of chemicals. Acetaminophen (APAP) is also metabolized in the liver. The chemical is used as safe and effective antipyretics at therapeutic doses; however, its overdose can be a cause of severe liver injury [1][2][3][4][5][6]. In fact, an overdose is the primary cause of acute liver failure worldwide [2]. Normally, APAP is eliminated by glucuronidation and sulfation. When ingested at large amounts, excess APAP undergoes oxidation to form the highly reactive intermediate N-acetyl-p-benzoquinone-imine (NAPQI) by cytochrome P450, particularly with CYP2E1 [7]. NAPQI is  Histopathology of APAP-induced rat hepatotoxicity. In the control (A) and at hour 10 (B), no significant changes are observed. On day 1, coagulation necrosis of hepatocytes (arrows) accompanied by a small number of inflammatory cells is seen in the affected centrilobular areas (C). On day 2, coagulation necrosis with increased infiltration of inflammatory cells (arrows) is seen in the centrilobular areas (D). On day 3, the inflammatory cells are almost decreased, and fibrosis develops (E). Myofibroblasts immunopositive for α-SMA (arrows) are seen in the fibrotic areas on days 3 (F). Hematoxylin and Eosin (A-E) and immunohistochemical stain, counterstained with hematoxylin (F). CV, central vein. Bar = 50 µm.
On day 1, coagulation necrosis of hepatocytes accompanied by a small number of inflammatory cells (mainly macrophages and neutrophils) was seen in the affected centrilobular area ( Figure 2C), of which lesion development corresponded to the significantly increased serum AST, ALT, and T. Bil. On day 2, coagulation necrosis in the centrilobular area was also seen with more infiltration of inflammatory cells ( Figure 2D). On day 3 ( Figure 2E), the inflammatory cells were gradually decreased in number, and instead of coagulation necrosis, fibrosis developed; the affected centrilobular areas almost recovered on day 5, indicative of reparative fibrosis. On days 3 and 5, α-SMA positive myofibroblasts, a fibrotic marker [11,29] was seen in the fibrous area with the greatest appearance on day 3 ( Figure 2F).

CD68 Immunostaining for M1-Macrophages
A few CD68 positive macrophages were sporadically seen in the control livers ( Figure 3A). At hour 10, the number of CD68 positive cells was similar to the control level. Interestingly, CD68 positive cells were quickly increased significantly on days 1 and 2 ( Figure 3B), with a peak on day 1. Thereafter, the CD68 positive cell number on days 3 and 5 was decreased to the control level ( Figure 4A). Morphologically, CD68 positive cells were small and round in shape.

CD163 Immunostaining for M2-Macrophages
In the controls, CD163 positive cells were seen along the sinusoid, indicating that they are Kupffer cells ( Figure 3C) [30]. At hour 10 and on day 1, the appearance of CD163 positive cells was similar to that of control. CD163 positive cell number showed a significant increase on days 2 ( Figure 3D) and 3 with a peak on day 2 ( Figure 4B). The number was decreased gradually on day 5; the positive cells were large round or spindle-shaped with enlarged cytoplasm, differing from the shape of CD68 positive cells.

MHC Class II Immunostaining
In the control livers, MHC class II positive cells were infrequently seen ( Figure 5A). There was a tendency to increase in MHC class II positive cell number at hour 10. On days 1 and 2 ( Figure 5B), the positive cells were significantly increased, and thereafter, gradually decreased on days 3 and 5 ( Figure 6A). Morphologically, the positive cells showed various shapes such as round, large round, ovoid, spindle-shaped, or dendrite.

Iba1 Immunostaining
Iba1 positive cells were sporadically seen along the sinusoids in controls ( Figure 5C). At hour 10, the Iba1 positive cell numbers did not show any significant change, but they were significantly increased on days 1 and 2 ( Figure 5D), showing a peak on day 1; the appearance was gradually decreased on days 3 and 5 ( Figure 6B). Morphologically, the positive cells were similar to MHC class II positive cells, showing various configurations.  Kinetics of MHC II, Iba1, and CD204 expressing macrophages. Compared to the controls, a significantly increased number of macrophages expressing MHC class II (A) on days 1 and 2, Iba1 (B) on days 1 and 2, and CD204 (C) on days 1-3 are seen. Dunnett's test; *, significantly different from controls at p < 0.05. Bar represents mean ± SD.

CD204 Immunostaining
CD204 positive cells were seen along the sinusoids in the controls ( Figure 5E). At hour 10, no significant change was seen in the number. The positive cell number was significantly increased on days 1, 2 ( Figure 5F), and 3, showing a peak on day 2, and then, the positive cells were decreased on day 5 ( Figure 6C). The cell shapes were similar to those of MHC class II and Iba1.

HMGB1 Immunostaining
Nuclei of hepatocytes in the controls were positive for HMGB1 at basal level ( Figure 10A). At hour 10, there were some hepatocytes with fine granular reactivity in the cytoplasm in the injured area. On day 1, hepatocytes in the affected centrilobular area showed the greatest cytoplasmic positivity as fine granules and their nuclei looked negative for HMGB1 ( Figure 10B). On days 2, 3, and 5, nuclear positivity was the same as that in the controls. These findings indicated that the translocation of nuclear positivity into cytoplasm occurred at hour 10 in some hepatocytes and on day 1 in many injured hepatocytes.

Western Blotting Analysis of HMGB1 and DAMPs Receptors
HMGB1 protein expression showed significantly increased expression at hour 10 when cytoplasmic reactivity began to be seen in hepatocytes ( Figure 11A). TLR9 protein expression did not show a significant increase; however, it tended to increase at hour 10 ( Figure 11B). TLR2 and TLR4 protein expressions did not show any significant change. Protein expression of MyD88, a central adapter shared between almost all TLRs, significantly increased at hour 10 ( Figure 11C), indicating that MyD88-dependent TLR-mediated immune responses were activated at hour 10. The expression level of HMGB1 protein significantly increases at hour 10 (A). TLR9 protein expression does not show a significant increase; however, it tends to increase at hour 10 (B). MyD88 protein expression significantly increases at hour 10 (E). A representative band is shown at each examination point (A, B, and C). GAPDH was used as the loading control (lower panel). Dunnett's test; *, significantly different from controls at p < 0.05. Bar represents mean ± SD. TLR9, toll-like receptor 9; MyD88, Myeloid Differentiation 88.

Analysis of Autophagy
To evaluate the autophagy activity, immunohistochemistry and western blotting analyses were performed for LC3B. Western blotting analysis was also performed for RAGE that has been considered to promote autophagy [22].

LC3B Immunostaining for Autophagosome Marker
LC3B positive cytoplasmic fine granules in hepatocytes were sporadically detected at basal level in the control livers ( Figure 12A). Compared to expression in the controls, cytoplasmic LC3B labeled granules in hepatocytes were more clearly seen in the affected centrilobular area at hour 10 and on day 1 ( Figure 12B); particularly, larger cytoplasmic granules reacting to LC3B were occasionally seen on day 1 ( Figure 12C), presumably indicating abnormally developed autophagosomes. Furthermore, semi-quantitatively, the degree of LC3B positive cytoplasmic granules in hepatocytes showed a peak on day 1 (Table 1); hepatocytes with cytoplasmic granules were infrequently observed in the centrilobular area on days 2, 3, and 5, returning to the control levels. Figure 12. Immunohistochemistry for LC3B in APAP-induced rat hepatotoxicity. LC3B positive cytoplasmic fine granules in hepatocytes (arrows) are sporadically detected at basal level in the control liver (A). Compared to the control expression, intracytoplasmic LC3B labeled granules in hepatocytes (arrows) are increased in the centrilobular area with a peak on day 1 (B) ( Table 1). Abnormally-developed LC3B positive autophagosomes, apparently vacuolated granules, are occasionally seen in some hepatocytes (C), suggestive of abnormality of formation (arrow). Insets are higher magnifications. Immunohistochemical staining, counterstained with hematoxylin. CV, central vein; LC3B, microtubule-associated protein light chain 3. Bar = 50 µm (A-B), 15 µm (C), inset bar = 20 µm. Consistent with immunohistochemistry for LC3B, the western blotting analysis confirmed the significantly increased expression of LC3B at hour 10 and on day 1, with a peak at hour 10 ( Figure 13A). RAGE protein expression significantly increased at hour 10 as the greatest expression of LC3B was seen in immunohistochemistry ( Figure 13B). Significantly increased expression of LC3B protein is seen at hour 10 and day 1 (A) with a peak on day 10. RAGE protein expression is significantly increased at hour 10 (B). A representative band is shown at each examination point (A and B). GAPDH was used as the loading control (lower panel). Dunnett's test; *, significantly different from controls at p < 0.05. Bar represents the mean ± SD. RAGE, receptors for advanced glycation end products.

Discussion
APAP-induced hepatotoxicity has been investigated exclusively in mice and some researchers have considered that rats are not suitable for this study [2,33]. It was reported, although overall APAP metabolism was similar in both mice and rats, that mitochondrial protein adducts were lower in rats and that rats also had less oxidative stress [2,33]. It was hypothesized, therefore, that mitochondrial dysfunction is critical for the development of necrosis after APAP treatment [2]. We tried to establish APAP-induced hepatotoxicity in rats. After 15 h of fasting, APAP was injected into F344 male rats at the age of 6 weeks and the obtained centrilobular lesions were almost similar to those in the mice model. In addition to direct hepatic injury by chemicals, generally, macrophages should contribute to the modification of hepatotoxicity [8][9][10]. Since macrophages can show various functions depending on microenvironmental factors, a concept called M1-/M2-macrophage polarization has been proposed in pathological settings [12,34]. Therefore, in this study, we focused on M1-/M2-macrophage functions in relation to DAMPs and autophagy for rat APAP-induced hepatotoxicity. By pathological analyses, it was found that APAP-induced rat liver injury is very complicated.
Macrophages reacting to CD68 (for M1) showed a significant increase on days 1 and 2 following APAP injection; the peak was on day 1 when hepatocyte injury/necrosis was the most prominent, accompanied with a significant increase of AST, ALT, and T. Bil levels. mRNA expressions of IFN-γ, MCP-1, TNF-α, IL-1β, and IL-6 for M1-related factors, which are regarded as pro-inflammatory cytokines [38], also significantly increased on day 1. The appearance pattern of CD68 positive cells corresponded to mRNA expressions of these M1-related factors. These results indicated that M1 macrophages could appear in the early stages, and participate in tissue damage/inflammation via the production of pro-inflammatory factors such as IFN-γ, MCP-1, TNF-α, IL-1β, and IL-6.
Antibody of CD163 labels M2-macrophages [16,39]. In the present study, the number of CD163 positive cells increased on days 2 and 3; the appearance was later than that of CD68 positive M1-macrophages on days 1 and 2. Because M2-macrophages play an important role in reparative fibrosis after tissue injury [38,40,41], fibrotic lesions are seen on days 3 might be related to the appearance of CD163 positive M2-macrophages. Interestingly, the mRNA expression of IL-4 as an M2-macrophage-related factor was transiently increased on day 1. IL-4 is known to induce M2-macrophages [38]. IL-4 might be responsible for further activities of CD163 positive cells because M2-macrophages started to increase from day 2 onwards.
In the present hepatotoxicity, collectively, we demonstrated that M1-/M2-macrophage polarization was clearly present; CD68 positive cells appeared in the early stage as macrophages for tissue damage/inflammation promotion, and thereafter, CD163 positive cells were seen in the late stage in relation to tissue repair/reparative fibrosis. These findings of polarization are different from those seen in TAA-induced rat hepatotoxicity with the simultaneous appearance of M1-/M2-macrophages [16].
In addition to CD68 (for M1) and CD163 (for M2), in the present study, we investigated macrophages with other immunophenotypes (MHC class II, Iba1, and CD204). MHC class II molecules are expressed in antigen-presenting cells such as dendritic cells [21,42]. Iba1 is a cytoplasmic, calcium-binding, inflammation-responsive scaffold protein [36]; it is used as a marker of activated macrophages [36]. CD204 is a class A macrophage scavenger receptor, playing important roles in host defense mechanisms [37]. In the present hepatotoxicity, the number of macrophages reacting to MHC class II and Iba1 showed a significant increase on days 1 and 2 in the centrilobular lesion. The double immunofluorescence revealed that there were greater numbers of CD68 positive cells reacting simultaneously to MHC class II on days 1 to 3, in contrast to the number of MHC class II/CD163 double-positive cells. Therefore, MHC class II positive cells may have a predisposition towards M1-polarization. Double immunofluorescence using Iba1 antibody for M1-/M2-polarization could not be conducted, because Alexa 568-conjugated secondary antibody is not available. However, the expression pattern of Iba1 positive cells was similar to that of CD68 positive M1-macrophages with significantly increased levels on days 1 and 2, suggesting that Iba1 positive macrophages might play a role as M1-macrophages.
CD204 positive macrophages showed a significant increase on days 1 to 3; on days 2 and 3, the significantly increased number was corresponding to that of CD163 positive M2-macrophages. On day 3, further, although CD68, MHC class II, and Iba1 positive macrophages had already decreased, a significantly increased number of CD204 positive cells were still seen. In the double immunofluorescence, on days 1 to 3, many CD163 positive cells were reacting simultaneously to CD204, being a much greater number than that of CD204/CD68 double-positive cells. Therefore, in the present hepatotoxicity, CD204 positive cells may have a predisposition towards M2-polarization. Presumably, CD204 positive M2-macrophages that appeared on day 1 might be a source of production of M2-related factors, in particular IL-4 that showed a significant increase on day 1 and is considered to be an inducer of M2-macrophages.
M1-polarization of MHC class II positive cells and M2-polarization of CD204 positive cells have been reported in TAA-induced rat hepatic injury [16]. The analysis of M1-/M2-macrophage polarization should be useful to find out the pathogenesis of chemically-induced liver damages.

DAMPs
Damaged/dying cells may release endogenous ligands called DAMPs, and DAMPs can activate pattern recognized receptors, such as TLRs and RAGE [23]. TLRs, widely expressed on leukocytes, regulate innate and adaptive immune responses through the inflammatory cytokines produced by inflammatory cells [19,24,25]. RAGE is the receptor for advanced glycation end-products [18].
HMGB1 is a member of the high mobility group nuclear protein family and well known as DAMPs. HMGB1 protein is a nuclear DNA binding protein [43]. Under normal circumstances, HMGB1 is present in the nuclei, playing important roles in biological processes including transcription and DNA repair [43]. In response to appropriate stimuli, HMGB1 translocates from the nucleus to cytoplasm, thereafter, picked up into secretory lysosomes, and then, is secreted from the cell through exocytosis. HMGB1 released from necrotic/injured cells stimulates monocytes/macrophages through the cell-surface receptors such as TLR2, TLR4, TLR9, and RAGE [44][45][46][47]. It is reported that, in cell signals relating to HMGB1, the translocation of cytoplasmic NF-κB into the nucleus can induce an inflammatory response [44][45][46][47].
In the present hepatotoxicity, immunohistochemically, HMGB1 positivity was seen in the nuclei in the control hepatocytes and began to be seen sporadically as intracytoplasmic fine granules with greater nuclear reactivity in the centrilobular area at hour 10 when a hepatic injury could not be still detected. The peak of cytoplasmic HMGB1 positivity was on day 1 when corresponded to the greatest hepatic injury with a large number of M1-macrophages. HMGB1 protein expression revealed a significant increase at hour 10. Collectively, these findings indicated possible participation of HMGB1 in the development of APAP-induced rat liver lesions to stimulate tissue damage/inflammation which should be due to the participation of M1-macrophages.
Although TLR2/4 are reported as receptors for HMGB1 [44][45][46][47], the protein expression of TLR2 and TLR4 did not show a significant increase in the present study. Instead of these receptors, the protein of TLR9 and RAGE showed a tendency to increase at hour 10, being corresponding to the significantly increased level of HMGB1 protein and MyD88 protein. TLR9 and RAGE might be expressed as DAMPs receptors in the early stage of tissue damage/inflammation in the present hepatotoxicity.
It is interesting to note that MHC class II-expressing cells with the polarization of M1-macrophages appeared exclusively on days 1 and 2. MHC class II molecule is expressed in immune cells such as macrophages through TLRs binding to DAMPs [21]. Because MHC class II-expressing macrophages participate in complicated immune response, it may be interesting to carry out studies on the relation between the immune cells and DAMPs in this hepatotoxicity [48][49][50]. Along with HMGB1, furthermore, S100A4 and HspA1B are involved in DAMPs [49,50]. mRNAs of S100A4 and HspA1B increased on days 1 and 2, and on day 1 onwards, respectively. Various kinds of DAMPs might contribute to the present hepatotoxicity. The interaction of these DAMPs should be investigated in future studies.

Autophagy
Autophagosome formation is an attempt by cells to limit the spread of subcellular damage by walling off the damaged areas [51]. Autophagy plays an important role in a wide range of diseases [7,26,52]. In drug-induced liver injury, autophagy may contribute not only to homeostasis but also to the suppression of injury [26]. LC3 (microtubule-associated protein 1A/1B light chain 3), the most commonly monitored autophagy marker, has four isoforms; out of them, LC3B is often used for analyses of autophagy function [7]. Immunohistochemically, LC3B-positive fine granules were seen in the cytoplasm of hepatocytes at hour 10, and on day 1, the positivity was the greatest in hepatocytes around the necrotic area; there were some hepatocytes with abnormally-developed granules reacting to LC3B. Furthermore, it was confirmed that LC3B protein expression level was the greatest at hour 10. These findings suggested that the autophagy already occurred functionally at hour 10 before tissue injury on days 1 and 2.
HMGB1 regulates cellular processes such as autophagy. Particularly, HMGB1 is important for oxidative stress-mediated autophagy induction [44]. RAGE, known as a promoter of inflammation via NF-κB, has functions to promote autophagy [22]. Fundamentally, it is considered that autophagy inhibits inflammation via down-regulating caspase 1-dependent inflammasomes which activate IL-1β, showing a critical regulatory function in macrophage polarization that down-regulates inflammation [53,54]; these macrophages may be regarded as M2 type. Based on the information, in the early stages at hour 10 and day 1 in the present hepatotoxicity, the increased autophagy might act as cellular protection to injury, in relation to increased HMGB1 and RAGE. However, HMGB1 released from necrotic/injured cells stimulates monocytes/macrophages through the cell-surface receptors [44,46,47]. The subtle relationship between autophagy and HMGB1 (as DMAPs) may be important in the present hepatotoxicity.

Animals
Five-week-old twenty-seven male F344/DuCrj rats (Charles River Japan, Yokohama, Japan) were used. They were housed in an animal room under controlled temperature (22 ± 3°C) and with a 12 h light-dark cycle. Animals were given a standard commercial diet (DC-8, CLEA Japan Inc., Tokyo, Japan) and tap water ad libitum. They were kept for one week to acclimatize to the environment. Twenty-three rats were fasted for 15 h and then injected intraperitoneally once by APAP (Sigma Aldrich Co., Darmstadt, Germany) dissolved in 0.5% metolose (methylcellulose; Shin-Etsu Chemical Co., Ltd, Tokyo, Japan) in distilled water at a dose of 1000 mg/kg body weight to induce liver injury. Rats were fed one day later and were sacrificed under deep isoflurane anesthesia at 10 h, and on 1, 2, 3, and 5 days (n = 4 or 5 at each point) after APAP injection. The remaining four rats served as controls were injected 0.5% metolose in distilled water in the same way, and sacrificed on day 5. The dose was determined based on data in our preliminary experiments with different doses, along with information in a previous article [2] The experimental protocols and animal housing conformed to the institutional guidelines of Osaka Prefecture University for the Care and Use of Experimental animals, were approved by the ethical committee of the University and were registered on Osaka Prefecture University Care and Use of Experimental animals register

Serum Biochemistry
Blood samples were collected from the abdominal aorta and separated sera were subjected to biochemical analyses of aspartate transaminase (AST), alanine transaminase (ALT), and total bilirubin (T. Bil) (SRS Inc., Tokyo, Japan).

Histopathology and Immunohistochemistry
Tissues from the left lateral lobe of the livers were fixed in 10% neutral buffered formalin, or periodate-lysine-paraformaldehyde (PLP) solution. Formalin solution-fixed specimens were processed routinely and embedded in paraffin wax. PLP solution-fixed specimens were embedded in paraffin with the AMeX method [16]. The fresh liver tissues were also embedded in Tissu Mount ® (Chiba Medical, Saitama, Japan) and frozen immediately at -80 • C until use.

Histopathological Examination
Sections were cut at 3 µm in thickness and were stained with hematoxylin and eosin (HE) for morphological observations.

Double Immunofluorescence
Fresh frozen sections (10 µm in thickness) were used for double immunofluorescence staining. Sections were fixed in acetone:methanol mixed solution (1:1) at 4 • C for 15 min except for double immunofluorescence with CD204/CD68 or CD204/CD163. For the CD204/CD68 or CD204/CD163 staining, sections were fixed in PLP solution at 4 • C for 15 min. All sections for double immunofluorescence were dried for 30 min at room temperature, followed by blocking with 10% normal goat serum for 30 min at room temperature. Thereafter, they were incubated with the primary antibodies at 4 • C overnight; after washing in PBS, they were incubated with the secondary antibody for 45 min at room temperature: Alexa 568/anti-mouse IgG1 (Invitrogen Co., CA, USA; ×500). For double immunofluorescence with CD204/CD68 or CCR2/CD68, sections were incubated with the labeled-CD68 antibody (Mouse anti-rat CD68: Alexa Fluor ® 488, AbD Serotec, Oxford, UK). The labeled-MHC class II antibody (Mouse anti-rat MHC Class II H-2I-Ak/s: Alexa Fluor ® 488, AbD Serotec, Oxford, UK) was used for the staining with MHC class II/CD68 or MHC class II/CD163. Sections stained with CD204/CD163 or CCR2/CD163 were incubated with the labeled-CD163 antibody (mouse anti-rat CD163: FITC, AbD Serotec, Oxford, UK). All labeled antibodies were reacted for 45 min at room temperature. After washing in PBS, sections were mounted with SlowFade ® Gold antifade reagent with 4', 6-diamino-2-phenylindole (DAPI; Vectashield ® ; Vector Laboratories, CA, USA) for nuclear stain. Signals were detected by VS120 Virtual Slide System (Olympus, Tokyo, Japan).

Real-Time Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)
To evaluate the expression of M1-or M2-macrophage-related factors, and chemokines such as CCR2 and CCL7, real-time RT-PCR was performed. Liver tissues from the left intermediate lobe were immediately soaked in RNA later ® (Qiagen GmbH, Hilden, Germany) and stored at -80 • C. Total RNA was extracted by using an SV Total RNA Isolation System (Promega, WI, USA). Total RNA was reverse-transcribed to cDNA using Super Script ® VILO TM cDNA Synthesis Kit (Invitrogen Co., CA, USA).
For quantification of IL-6, IFN-γ, MCP-1, TNF-α, IL-1β, IL-4, IL-10, TGF-β1, CCR2, and CCL7, Thunderbird ® Probe qPCR Mix (Toyobo, Co., Ltd., Osaka, Japan) were used with TaqMan Gene Expression Assays (Life Technologies, Massachusetts, MA, USA) ( Table 3). The amplification program consisted of 1 cycle at 95 • C with a 1 min hold followed by 40 cycles at 95 • C of denaturing temperature with a 15 s hold, 60 • C of extension temperature with a 30 s hold, and then 20 • C of cooling with a 10 s hold. The expression values of target genes were normalized by the expression values of 18s rRNA.

Cell Counts
The number of CD68, CD163, Iba1, MHC class II, and CD204 positive cells in centrilobular areas of the hepatic lobule was counted using WinROOF software (Mitani Corp., Fukui, Japan) and expressed as the number of positive cells per unit area (cells/mm 2 ) [16].

Statistical Analysis
Data are represented as mean ± standard deviation (SD). Statistical analyses were performed using Dunnett's test. Significance was accepted at p < 0.05.

Conclusions
To shed some light on the pathogenesis of APAP-induced rat liver injury, we focused on the M1-/M2-macrophage polarization concept in relation to the effects of DAMPs and autophagy. M1-macrophages and its related factors increased in the early stages, whereas M2-macrophages subsequently appeared in the late stage. Particularly, HMGB1 worked as one of the DAMPs before tissue injury. HMGB1 binding to TLR9 might initiate inflammation, which should be closely related to the polarization of M1-macrophages. Autophagosomes, demonstrable with LC3B immunohistochemistry and western blot analysis, might take part in the early events via RAGE expression, of which functions might have cytoprotection to injury. Collectively, it was found that the association of M1-/M2-macrophage polarization with DAMPs and autophagy might have been responsible for the pathogenesis of APAP-induced rat liver injury. Understanding the mechanisms of DAMP, autophagy, and M1-/M2-macrophages at molecular levels would provide deeper insight into the hepato-pathogenesis. In addition, to investigate the detailed mechanisms of APAP-induced rat hepatotoxicity at the molecular level, this rat model would be useful for the development of a protective compound for hepatotoxicity in therapeutic strategies. This study is the first report of APAP-induced rat hepatotoxicity with detailed analyses focusing on M1-/M2-macrophage polarization.