Hepatocyte-Specific Deletion of Heme Oxygenase-1 … suggests SnPP-mediated HO-1 inhibition ... icity (Eipel et al. 2007), and the implications of ... erate HO-1flox/+ mice, ...

Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 331

331

Tohoku J. Exp. Med., 2008, 216, 331-339

Received October 24, 2008; revision accepted for publication November 6, 2008.Correspondence: Masayuki Yamamoto, Department of Medical Biochemistry, Tohoku University Graduate

School of Medicine, 2-1 Seiryo-cho, Aoba-ku, Sendai 980-8575, Japan.e-mail: [email protected]

Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis in Basal and Oxidative Environments

TAKASHI MAMIYA,1 FUMIKI KATSUOKA,2 AKI HIRAYAMA,3 OSAMU NAKAJIMA,1

AKIRA KOBAYASHI,1 JONATHAN M. MAHER,2 HIROFUMI MATSUI,1 ICHINOSUKE HYODO,1 MASAYUKI YAMAMOTO

1, 2 and TOMONORI HOSOYA1

1Graduate School of Comprehensive Human Sciences and Center for Tsukuba Advanced Research Alliance, Exploratory Research for Advanced Technology-Japan Science and Technology Corpo-ration, Tsukuba, Japan

2Department of Medical Biochemistry, Tohoku University Graduate School of Medicine, Sendai, Japan

3Center for Integrative Medicine, Tsukuba University of Technology, Tsukuba, Japan

Heme oxygenase-1 (HO-1) is the rate-limiting enzyme of heme catabolism and has been assumed to be important in cellular response against oxidative stress through modification of the pro-oxidant heme into less toxic catabolites that behave as antioxidants. However, the precise mechanisms involved and the physiological significance of such activity remain to be clarified. To elucidate roles HO-1 plays in vivo, hepatocyte-specific conditional knockout (CKO) mice of HO-1 gene were generated by site-specific recombination using albumin-promoter-driven Cre-loxP system. In livers of HO-1 CKO mice HO-1 protein level decreased to approximately 30% of control mouse livers. The HO-1 CKO mice are viable, exhibit normal growth curves over six months, and show no histological and sero-logical abnormalities. We found that several cytoprotective genes, such as NAD(P)H dehydrogenase quinone 1 and glutathione S-transferase P1, showed markedly elevated expression, suggesting the increase of oxidative stress in HO-1 CKO mice even under qui-escent conditions. In vivo electron paramagnetic resonance studies demonstrated that sig-nal decay times of nitroxyl radicals were significantly longer in livers of HO-1 CKO mice than that of control mice, indicating that radical scavenging activity was significantly com-promised in the mutant liver. HO-1 CKO mice were susceptible to carbon tetrachloride hepatotoxicity. These results provide the first in vivo evidence that HO-1 acts to protect cells against the oxidative stress in both basal conditions and upon chemical insult. ──── heme oxygenase-1; oxidative stress; reactive oxygen species; in vivo EPR; carbon tetrachloride.Tohoku J. Exp. Med., 2008, 216 (4), 331-339.© 2008 Tohoku University Medical Press

T. Mamiya et al.332 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 333

Heme oxygenase (HO) is the first and rate-limiting enzyme in an enzymatic pathway that catabolizes the degradation of heme into biliver-din, iron and carbon monoxide (CO) (reviewed in Shibahara 2003; Furuyama et al. 2007). There are two isoforms of HO, i.e., HO-1 and HO-2. HO-1 is expressed in spleen, liver and many other tis-sues and expression becomes highly inducible under certain conditions. HO-2 is also constitu-tively expressed in a wide range of tissues, but do not share the same range of inducibility. In addi-tion to the primary role HO-1 plays in heme catabolism, HO-1 aids in cytoprotection from var-ious xenobiotic and endobiotic insults. In fact, HO-1 expression is induced by heme, as well as by a variety of non-heme inducers, such as hydro-gen peroxide, ultraviolet (UV) irradiation, cadmi-um (Keyse and Tyrrell 1989), endotoxin (Gemsa et al. 1974), or nitric oxide (Marquis and Demple 1998).

HO-1 has been assumed to be important in cellular response against oxidative stress through modification of the pro-oxidant heme into less toxic metabolites, such as biliverdin and bilirubin. In fact, both biliverdin and bilirubin have been reported as scavengers of peroxyl radicals (Stocker et al. 1987). Although CO is known to be toxic at high doses, the cytoprotective effects of CO has been demonstrated in conditions associated with oxidative stress-induced apoptosis (Otterbein et al. 1999). However, the precise mechanisms of how HO-1 exerts its cytoprotective effects against oxidative stress are not currently understood.

HO-1 gene has been reported to be transcrip-tionally activated by the basic region-leucine zip-per (bZip) transcription factor Nrf2 (Alam et al. 1999; Ishii et al. 2000), a key transcription factor that regulates expression of a battery of antioxi-dant and xenobiotic metabolism genes (Motohashi and Yamamoto 2004; Kobayashi and Yamamoto, 2005). Upon exposure to oxidative stress, Nrf2 accumulates in the nucleus and activates a group of cytoprotective genes through antioxidant/electrophile response element (ARE/EpRE) bind-ing in regulatory regions of these genes (Itoh et al. 2004; Tong et al. 2007). Another bZip tran-scription factor, Bach1, has also been shown to

compete with Nrf2 for ARE binding; Bach1 is known to repress transcriptional activity (Sun et al. 2002).

The functional role of HO-1 has been ana-lyzed by the utilization of a variety of metal pro-toporphyrin derivatives. For instance, cobalt pro-toporphyrin induces HO-1 and protects the liver from the stresses of ischemia/reperfusion. Reac-tive oxygen species (ROS) have been implicated in ischemia/reperfusion injury, and HO-1 seem-ingly counteracts ROS insult (Kato et al. 2001). Conversely, tin protoporphyrin (SnPP) inhibits HO-1 activity and exacerbates hepatic injury upon exposure to carbon tetrachloride (CCl4) (Nakahira et al. 2003), which is metabolized into reactive intermediate trichloromethyl radicals (•CCl3) (Connor et al. 1986). However, a conflicting report suggests SnPP-mediated HO-1 inhibition protects the liver from CCl4-mediated hepatotox-icity (Eipel et al. 2007), and the implications of chemical inhibition of HO-1 activity upon oxida-tive challenge remain unclear.

Similarly, studies in mice using HO-1 gene disruption have been limited in scope due to pre-natal lethality in this mouse model. These HO-1 knockout (KO) mice were generated by Poss et al (Poss and Tonegawa 1997a) and were observed to have a high incidence of prenatal lethality, due to anemia and iron accumulation, as well as an im-plicit susceptibility to endotoxin-induced inflam-mation (Poss and Tonegawa 1997b). Embryonic fibroblasts derived from these mice were highly sensitive to oxidative stress caused by hemin and hydrogen peroxide, further supporting the notion that HO-1 aids in cytoprotective antioxidant activ-ity (Poss and Tonegawa 1997a).

To fully clarify HO-1 function in vivo, we generated hepatocyte-specific HO-1 conditional knockout (CKO) mice, as protective roles of HO-1 have been studied and suggested in the liv-ers. Livers of HO-1 CKO mice had no apparent histological abnormalities, yet stress-inducible genes were markedly elevated in the mice. In vivo electron paramagnetic resonance (EPR) test-ing revealed that radical scavenging activity was diminished in HO-1 CKO mice. Furthermore, HO-1 CKO mice were markedly susceptible to

T. Mamiya et al.332 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 333

CCl4-induced hepatotoxicity. These results dem-onstrate that HO-1 has a key role in modulating oxidative stress in hepatocytes.

METHODS

Generation of hepatocyte-specific HO-1 CKO miceThe full details of conditional targeting strategy of

HO-1 will be reported separately (Hosoya et al. in prepa-ration). Briefly, a floxed HO-1 targeting vector was con-structed with a loxP site inserted into the second intron and a floxed Neo cassette inserted downstream of the ter-minal exon. Homologous recombination in 129Sv-derived embryonic stem cells was conducted using stan-dardized procedures (Hosoya et al. 2001). Chimeric mice were crossed with C57BL/6 wild-type mice to gen-erate HO-1flox/+ mice, and the resulting mice were further crossed with Sycp1-Cre mice (Vidal et al. 1998; gener-ously supplied from the Jackson Laboratory (JAX)) to generate HO-1+/– mice. Albumin-Cre transgenic (Alb-Cre) mice (Postic et al. 1999), which specifically express Cre recombinase in hepatocytes, were also supplied from JAX and crossed with HO-1+/– mice to obtain Alb-Cre:HO-1+/– mice, which were then crossed with HO-1flox/+ mice to generate hepatocyte-specific HO-1-deficient (Alb-Cre:HO-1flox/–) mice and control mice (HO-1flox/+) on mixed C57BL/6-129Sv background. Female mice aged 6 to 9 weeks were used for all conducted experiments with the approval of University of Tsukuba Animal Experiment Committee.

Quantitative RT-PCRQuantitative RT-PCR was performed with primers

and probes specific for glutamate-cysteine ligase catalytic subunit (Gclc), glutathione S-transferase A4 (Gsta4), glutathione S-transferase P1 (Gstp1), NAD(P)H dehy-drogenase quinone 1 (Nqo1), and thioredoxin reductase 1 (Txnrd1) mRNAs as previously described (Katsuoka et al. 2005). Ribosomal RNA was quantified in each exper-iment as an internal control.

ImmunoblottingTissue proteins were solubilized with SDS sample

buffer, separated by SDS–polyacrylamide gel electropho-resis, and then transferred onto a polyvinylidene fluoride membrane (Millipore, Billerica, MA, USA). HO-1 (SPA-895) and HO-2 antibodies (OSA-200) were pur-chased from Stressgen (Victoria, Canada).

In vivo EPR spin probe study3-carbamoyl-2,2,5,5-tetramethylpyrrolidine-1-oxyl

(Carbamoyl-PROXYL, Aldrich, Milwaukee, WI, USA) was injected via tail vein into mice (300 mM, 3 ml/kg body weight). Each mouse was then placed in a plastic holder and put into the EPR system, placing its upper abdominal area in the center, with the bladder area out-side of the resonator (see Fig. 2B). The EPR signal derived from the area is illustrated in Fig. 2C. Signal in-tensities were measured using ESR-NT software (Jeol, Tokyo, Japan). Rate constants were measured every 20 s from 4 to 12 min after the carbamoyl-PROXYL injec-tion. The peak-to-peak height of the lowest magnetic field signal in the triplet spectrum was defined as the sig-nal intensity. EPR conditions for these in vivo measure-ments were as follows: magnetic field, 37.0 ± 5.0 mT; modulation width, 0.069 mT; time constant, 0.3 s; micro-wave power, 0.1 mW; scanning time, 20 s. The spin clearance of carbamoyl-PROXYL signal intensity was semi-logarithmically plotted against time. The first order spin reduction rate constant (κ ) was estimated from the slope value of the observed clearance curve, which was obtained by best fit. The half-life was calculated using the equation: t1/2 = ln2/κ (Hirayama et al. 2003).

Carbon tetrachloride hepatotoxicityAcute liver injury was induced by CCl4 (Wako,

Osaka, Japan). CCl4 was suspended at 3 μ l/ml for low dose and 30 μ l/ml for high dose in corn oil, and 10 ml/kg of each solution was administered intraperitoneally to mice. The mice were sacrificed at the indicated time points after injection. The concentration of aspartate aminotransferase (AST), alanine aminotransferase (ALT), total bilirubin, alkaline phosphatase (ALP), and lactate dehydrogenase (LDH) were measured with an automated biochemical analyzer, DRI-CHEM 7000V (Fujifilm). Hepatic lesions were graded from 0 to 4 (0, no lesions; 1, degeneration of centrilobular hepatocytes; 2, slight cen-trilobular necrosis; 3, moderate centrilobular necrosis; 4, severe centrilobular necrosis), as previously described with minor modifications (Enomoto et al. 2001).

Histology and immunohistochemistryLivers were fixed in 10% neutral buffered formalin,

and embedded in paraffin for hematoxylin/eosin (HE) staining or immunohistochemistry using HO-1 antibody (ab5247, Abcam, Cambridge, United Kingdom). To induce HO-1 expression in hepatocytes, hemin was admi-nistered intraperitoneally (30 mg/kg body weight (bw))

T. Mamiya et al.334 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 335

24 hrs before analysis. Sections were counterstained with hematoxylin. Iron accumulation was assessed by Prussian blue staining.

RESULTSGeneration of hepatocyte-specific HO-1 CKO mice

To evaluate the antioxidative activity of HO-1 in vivo, we generated a hepatocyte-specific HO-1 CKO mice for this study. To this end, a mouse line harboring loxP sites flanking exon 3 to 5 of HO-1 gene (floxed HO-1 gene), which has been generated in this laboratory (Hosoya et al. unpublished results), was utilized. Floxed HO-1 mice were crossed with transgenic mice express-ing Cre recombinase under the control of the hepatocyte-specific albumin promoter and creat-ing the HO-1 CKO line using for the current study (METHODS). Recombination efficiency was exam-ined by Southern blotting analysis, and appears to be approximately 60% (Fig. 1A). In relative agreement, immunoblotting revealed that the

hepatic HO-1 protein expression in CKO mouse livers was approximately 30% of control mouse levels (Fig. 1B and C). No significant change in protein levels was observed for HO-2 (data not shown).

We surmise that the presence of a wild-type allele (in Southern blotting) and remaining expre-ssion of HO-1 protein (in immunoblotting) in HO-1 CKO mice is due to the expression HO-1 in Kupffer cells. HO-1 protein in Kupffer cells was observed in the livers of HO-1 CKO mice (Fig. 1D, left panels). Immunohistochemistry demon-strated that HO-1 was induced in vivo upon administration of hemin in hepatocytes of control, but not HO-1 CKO mice (Fig. 1D, right panels). We did not observe any significant differences in the number of HO-1-positive Kupffer cells between two genotypes (data not shown). These results thus demonstrate that HO-1 protein is spe-cifically knocked out in hepatocytes.

The hepatocyte-specific HO-1 CKO mice are viable and exhibit normal growth curves for up to

Fig. 1. Generation of hepatocyte-specific HO-1 CKO mice.(A) Southern blotting using genomic DNA prepared from the livers of HO-1 CKO mice. EcoRI fragments specific to floxed allele (6.9 kb, filled arrowhead) and deleted allele (4.3 kb, open arrow-head) are shown. (B and C) Immunoblot analysis for HO-1 in the livers of control and HO-1 CKO mice. Panel B data was quantified and normalized to tubulin expression in panel C. Error bars rep-resent standard deviation (S.D., n = 4). Asterisks indicate a statistically significant difference from control mice (*P < 0.05). (D) HO-1 immunohistochemistry. Liver sections from control and HO-1 CKO mice sacrificed 24 hrs after vehicle administration (left two columns) or hemin (right two col-umns) were stained with HO-1 antibody. High-magnification sections are shown in the lower pan-els. Kupffer cells from vehicle-treated mice were stained for HO-1. Note that hepatocytes were stained only in the control mice after hemin treatment. (E) HE staining of livers from control and HO-1 CKO mice. Scale bars in panels D and E correspond to 200 μm.

T. Mamiya et al.334 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 335

six months (data not shown). No gross abnormal-ity was observed upon routine histological exami-nation (Fig. 1E). While iron accumulation was reported in the liver of conventional HO-1 KO mice over 6-month old (Poss and Tonegawa 1997a), no significant iron accumulation was seen in livers at the similar age of HO-1 CKO mice (data not shown). To assess the functional capac-ity of the liver, serum ALT, AST, total bilirubin, ALP and LDH levels were measured in both gen-otypes. No significant change was observed in HO-1 CKO mice compared with control mice (data not shown). These results suggest that HO-1 is dispensable for major physiological func-tions of the liver during development and in the adult stage. Thus Alb-Cre CKO mice escape the prenatal mortality associated with whole-body HO-1 deletion.

Stress inducible genes are upregulated in livers of HO-1 CKO mice

To examine whether the stress response is defective in HO-1-deficient livers, expression of stress-inducible genes in the livers of HO-1 CKO mice was examined. Since Nrf2 plays crucial roles in stress responses, we examined the expres-sion of Nrf2-regulated genes. Of the Nrf2-target genes, Nqo1, Gstp1 and Txnrd1 were all signifi-cantly elevated in livers of HO-1 CKO mice com-pared with control mice (Fig. 2A). Gsta4 and Gclc expression was also slightly induced in HO-1 CKO mouse livers, yet was not statistically significant (Fig. 2A). These results suggest the occurrence of increased redox stress in HO-1-deficient hepatocytes. We surmise that HO-1 contributes to the maintenance of redox homeo-stasis even in unstressed hepatocytes.

Fig. 2. Antioxidant capacity of HO-1 CKO mice under basal conditions.(A) Quantitative RT-PCR was performed to examine mRNA expression of the stress response genes Nqo1, Gstp1, Txnrd1, Gsta4 and Gclc. RNA samples were prepared from control (open bar) and HO-1 CKO mice (filled bar). Error bars represent S.D. (n = 4). Asterisks indicate statistically sig-nificant differences from control (*P < 0.01). (B) Mice retained in a plastic holder for EPR spin probe analysis. Signals derived from the area illustrated as a column was detected. (C) Analysis of the carbamoyl-PROXYL decay in control and HO-1 CKO mice. Typical time-dependent signal decay plots from control (filled squares) and HO-1 CKO mice (open triangles) are shown. (D) Sig-nal half-lives in the hepatic region of control (open bar) and HO-1 CKO mice (filled bar) are shown. The half-life of carbamoyl-PROXYL decay in the upper abdominal region was significantly pro-longed in the HO-1 CKO mice compared with control mice (**P < 0.01).

T. Mamiya et al.336 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 337

Radical scavenging ability is significantly impaired in HO-1 CKO mouse liver

We next examined whether HO-1 contributes to the reduction of radical species in vivo. For this purpose, we utilized EPR spin probes to mea-sure radical-reducing activity in livers of HO-1 CKO mice. A nitroxyl radical, carbamoyl-PROX-YL, was injected into either HO-1 CKO or control mice and EPR signal decay in liver was measured over time. The half-life of carbamoyl-PROXYL was significantly longer in HO-1 CKO mice than in control mice (Fig. 2C and D). The radical scavenging activity was measured within 10 min

after the administration of the tracer, meaning that the hepatocytes were not subjected to long periods of oxidative stress. Thus basal expression of HO-1 in hepatocytes critically contributes to anti-oxidative activity of the cells.

HO-1 CKO mice are susceptible to CCl4-induced hepatotoxicity

Next, the susceptibility of HO-1 CKO mice was examined upon chemical challenge with the pro-oxidant CCl4. To this end, CCl4, a well-known chemical that induces hepatotoxicity through generation of reactive intermediate tri-

Fig. 3. HO-1 CKO mice are more susceptible to acute hepatotoxicity of CCl4.(A) Plasma ALT and AST concentrations of control (open bar) and HO-1 CKO mice (filled bar) administered with 300 μ l/kg (right two columns) and 30 μ l/kg (left two columns) of CCl4. Plasma samples were taken from animals at 12, 24, and 48 hrs after injection. Values are represented as mean ± S.D., where n = 4. Asterisks indicate a statistically significant difference from control (*P < 0.005, **P < 0.05). (B) HE staining of livers from control (upper panels) and HO-1 CKO mice (lower panels) at 12, 24, and 48 hrs after administration of 30 μ l/kg of CCl4. Arrowheads indicate necrotic regions. Scale bar corresponds to 200 μm. Hepatocellular necrotic foci were deemed slight in centrilobular regions of control mice and severe in HO-1 CKO mice.

TABLE 1. Dose effect of CCl4 on plasma AST and ALT levels of control mice.

Dose (μ l/kg) AST (U/L) ALT (U/L)

0 62 (45-90) 28 (19-40)3 120 (80-4720) 90 (40-5680)10 140 (80-7200) 155 (40-5340)30 170 (100-1410) 160 (100-2430)90 4640 (500-6010) 6720 (1300-14800)300 2280 (510-3160) 4260 (2970-4800)900 3460 (1890-5380) 4340 (3960-4580)

Median (range), n = 4. CCl4 was suspended at 0.3, 1, 3, 9, 30, and 90 μ l/ml in corn oil. 10 ml/kg of each solution or corn oil alone was administered intraperitoneally to mice.

T. Mamiya et al.336 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 337

chloromethyl radicals, was administered to both control and HO-1 CKO mice (Connor et al. 1986). A dose-response study with CCl4 was first con-ducted using control mice (Table 1). Median AST and ALT concentrations remained low at 30 μ l/kg bw administration, but the scores dramatically increased at 90 μ l/kg bw. Therefore, 30 μ l/kg bw CCl4 was used for a low dose and 300 μ l/kg bw as a high dose for subsequent studies. High dose CCl4 administration revealed no significant differ-ences between control and HO-1 CKO AST/ALT concentrations (Fig. 3A, right panels). In stark contrast, low dose administration revealed AST and ALT values that were significantly higher in CCl4-treated HO-1 CKO mice than control (Fig. 3A, left panels).

Histological examinations demonstrated that low-dose CCl4 caused significant necrosis of hepatocytes surrounding the central vein. Necro-sis was much more severe in HO-1 CKO mouse livers than in control (Fig. 3B). The pathology scores for HO-1 CKO mice were significantly higher than control (Table 2). These results dem-onstrate that the HO-1 CKO mice are susceptible to CCl4-induced hepatotoxicity.

DISCUSSIONIn this study, hepatocyte-specific HO-1 CKO

mice were engineered. This conditional Cre-loxP system under the control of the albumin promoter allows for examination of hepatic HO-1 function in vivo in the adult animal. Although HO-1 inhib-itors have allowed for some insight into HO-1 liv-er function, the analysis with these chemicals has been fraught with discrepancies and complica-tions in interpretation. The present HO-1 CKO mice display no outward phenotype, but the radi-

cal-reducing capacity of the liver is compromised. Supporting this notion, compensatory expression of a set of stress-inducible genes in liver was observed in HO-1 CKO mice. When these HO-1 CKO mice were challenged with the oxidant CCl4, these mice were markedly sensitive to the increase in oxidative stress. These results provide solid line of evidence that HO-1 aids in maintain-ing redox homeostasis in the liver in vivo.

No apparent histological abnormalities were observed in HO-1 CKO mice, suggesting that HO-1 is a dispensable element in moderating the basic functions of the liver. Although HO-1 was generally thought to play a critical role in mediat-ing redox stress, basal expression HO-1 in hepa-tocytes generally remains quite low (Bauer et al. 2003). Therefore, the significance of the basal expression of HO-1 was rather unclear, while in contrast the inducible expression was thought to be important for hepatic HO-1 functions. It should be noted that in this study we observed compensatory induction of various Nrf2-target genes (Itoh et al. 1999), e.g., Nqo1, Gstp1, and Txnrd1, in HO-1 CKO mice even under basal conditions. Similar to HO-1, these genes play a critical role in mediating oxidative stress and aid-ing in cellular detoxification (Motohashi and Yamamoto 2004; Kobayashi and Yamamoto 2005). Despite marked upregulation of these genes involved in cellular protection, in vivo EPR studies revealed a diminished free-radical scav-enging capacity in HO-1 CKO mouse liver. These results demonstrate in vivo that the basal expression of HO-1 is crucial for the maintenance of liver homeostasis and that HO-2 and the other antioxidant enzymes expressed in hepatocytes are incapable of compensating for HO-1 loss. In

TABLE 2. Pathology scoring of control and HO-1 CKO mice after CCl4 treatment.

Time after treatment Control HO-1 CKO

12 hrs 0.25 ± 0.5 1.5 ± 0.624 hrs 1.75 ± 0.5 2.75 ± 0.548 hrs 1.75 ± 1.0 3.75 ± 0.5

Mean ± S.D. (n = 4). Hepatocytes damage score was significantly higher in the HO-1 CKO mice compared with the control mice at every time point (P < 0.05).

T. Mamiya et al.338 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 339

HO-1 CKO livers other stress responsive path-ways may be activated. This conclusion is shown schematically in Fig. 4.

The function of HO-1 in the liver after CCl4 exposure has remained somewhat controversial in the literature. Whereas one report presents evi-dence that HO-1 induction protects against CCl4-mediated hepatotoxicity (Nakahira et al. 2003), a second report suggests that HO-1 inhibition is the true cause of protection (Eipel et al. 2007). The current in vivo model demonstrates that hepato-cyte-specific loss of HO-1 is clearly detrimental to overall hepatotoxicity as measured by histolo-gy, pathological scoring and measurement of ALT/AST levels. We surmise that this is likely due to heme accumulation as well as a decrease in the formation of the antioxidant-scavenging and anti-inflammatory capabilities of the HO-1 cataly-sis end-products, namely biliverdin and CO. HO-1 is highly inducible after CCl4 administra-tion (Nakahira et al. 2003), and thus may serve as an important part of the antioxidant response to CCl4 bioactivation to the trichloromethyl metabo-lite by phase I enzymes.

In conclusion, the generation of hepatocyte-specific HO-1 CKO has allowed for more direct analysis of HO-1 function in both the basal condi-tion and upon chemical pro-oxidant challenge.

Although HO-1 CKO mice have strong compen-satory induction of cytoprotective genes in the basal condition, the radical scavenging capabili-ties of the liver remain impaired. Furthermore, HO-1 CKO mice are sensitive to CCl4-mediated hepatotoxicity, suggesting that HO-1 plays an important role in protection from pro-oxidants such as CCl4. Thus, the current study sheds new insight into the role of HO-1 in the liver, and implicates HO-1 as an important enzyme in main-taining redox balance in both basal conditions and upon oxidative and chemical insults.

AcknowledgmentsWe thank Drs. Ken Itoh, Souichiro Murata and

Keiko Taguchi for their generous help and Dr. Hozumi Motohashi for her critical reading of the manuscript. We also thank Ms. Naomi Kaneko for her help of histology and our laboratory members for useful discussions. This work was supported in part by grants from ERATO-JST (MY), the Ministry of Education, Science, Sports and Culture (FK, AK and MY). Present address of AK is Department of Medical Life Systems, Doshisha University, Kyota-nabe, Kyoto 610-0321, ON is Research Laboratory for Molecular Genetics, Yamagata University, Yamagata 990-9585, Japan, and TH is Department of Cell and Developmental Biology, University of Michigan Medical School , Ann Arbor, MI 48109-2200, USA.

Fig. 4. Schematic model for decreased antioxidant capacity in HO-1 deficient hepatocytes.Under basal conditions, HO-1 aids in maintaining redox homeostasis, e.g. endogenous stress gener-ated by a variety of endogenous metabolic pathways, in tandem with other antioxidant enzymes. When HO-1 is lacking, an increase in antioxidant and detoxification genes is observed, yet the total antioxidant capacity is diminished overall.

T. Mamiya et al.338 Hepatocyte-Specific Deletion of Heme Oxygenase-1 Disrupts Redox Homeostasis 339

ReferencesAlam, J., Stewart, D., Touchard, C., Boinapally, S., Choi, A.M.

& Cook, J.L. (1999) Nrf2, a Cap’n’Collar transcription factor, regulates induction of the heme oxygenase-1 gene. J. Biol. Chem., 274, 26071-26078.

Bauer, I., Wanner, G.A., Rensing, H., Alte, C., Miescher, E.A., Wolf, B., Pannen, B.H., Clemens, M.G. & Bauer, C. (2003) Expression pattern of heme oxygenase isoenzymes 1 and 2 in normal and stress-exposed rat liver. Hepatology, 27, 829-838.

Connor, H.D., Thurman, R.G., Galizi, M.D. & Mason, R.P. (1986) The formation of a novel free radical metabolite from CCl4 in the perfused rat liver and in vivo. J. Biol. Chem., 261, 4542-4548.

Eipel, C., Eisold, M., Schuett, H. & Vollmar, B. (2007) Inhibi-tion of heme oxygenase-1 protects against tissue injury in carbon tetrachloride exposed livers. J. Surg. Res., 139, 113-120.

Enomoto, A., Itoh, K., Nagayoshi, E., Haruta, J., Kimura, T., O’Connor, T., Harada, T. & Yamamoto, M. (2001) High sensitivity of Nrf2 knockout mice to acetaminophen hepa-totoxicity associated with decreased expression of ARE-regulated drug metabolizing enzymes and antioxidant genes. Toxicol. Sci., 59, 169-177.

Furuyama, K., Kaneko, K. & Vargas, P.D. (2007) Heme as a magnificent molecule with multiple missions: heme deter-mines its own fate and governs cellular homeostasis. Tohoku J. Exp. Med., 213, 1-16.

Gemsa, D., Woo, C.H., Fudenberg, H.H. & Schmid, R. (1974) Stimulation of heme oxygenase in macrophages and liver by endotoxin. J. Clin. Invest., 53, 647-651.

Hirayama, A., Yoh, K., Nagase, S., Ueda, A., Itoh, K., Morito, N., Hirayama, K., Takahashi, S., Yamamoto, M. & Koyama, A. (2003) EPR imaging of reducing activity in Nrf2 transcrip-tional factor-deficient mice. Free Radic. Biol. Med., 34, 1236-1242.

Hosoya, T., Oda, Y., Takahashi, S., Morita, M., Kawauchi, S., Ema, M., Yamamoto, M. & Fujii-Kuriyama, Y. (2001) Defective development of secretory neurones in the hypo-thalamus of Arnt2-knockout mice. Genes Cells, 6, 361-374.

Ishii, T., Itoh, K., Takahashi, S., Sato, H., Yanagawa, T., Katoh, Y., Bannai, S. & Yamamoto, M. (2000) Transcription fac-tor Nrf2 coordinately regulates a group of oxidative stress-inducible genes in macrophages. J. Biol. Chem., 275, 16023-16029.

Itoh, K., Wakabayashi, N., Katoh, Y., Ishii, T., Igarashi, K., Engel, J.D. & Yamamoto, M. (1999) Keap1 represses nuclear activation of antioxidant responsive elements by Nrf2 through binding to the N-terminal Neh2 domain. Genes Dev., 13, 76-86.

Itoh, K., Tong, K.I. & Yamamoto, M. (2004) Molecular mecha-nism activating Nrf2-Keap1 pathway in regulation of adap-tive response to electrophiles. Free Radic. Biol. Med., 36, 1208-1213.

Kato, H., Amersi, F., Buelow, R., Melinek, J., Coito, A.J., Ke, B., Busuttil, R.W. & Kupiec-Weglinski, J.W. (2001) Heme oxygenase-1 overexpression protects rat livers from ischemia/reperfusion injury with extended cold preserva-

tion. Am. J. Transplant., 1, 121-128.Katsuoka, F., Motohashi, H., Ishii, T., Aburatani, H., Engel, J.D.

& Yamamoto, M. (2005) Genetic evidence that small maf proteins are essential for the activation of antioxidant response element-dependent genes. Mol. Cell. Biol., 25, 8044-8051.

Kobayashi, M. & Yamamoto, M. (2005) Molecular mecha-nisms activating Nrf2-Keap1 pathway in regulation of anti-oxidant genes. Antioxidants Redox Signaling, 7, 384-395.

Keyse, S.M. & Tyrrell, R.M. (1989) Heme oxygenase is the major 32-kDa stress protein induced in human skin fibro-blasts by UVA radiation, hydrogen peroxide, and sodium arsenite. Proc. Natl. Acad. Sci. USA, 86, 99-103.

Marquis, J.C. & Demple, B. (1998) Complex genetic response of human cells to sublethal levels of pure nitric oxide. Cancer Res., 58, 3435-3440.

Motohashi, H. & Yamamoto, M. (2004) Nrf2-Keap1 defines a physiologically important stress response mechanism. Trends Mol. Med., 10, 549-557.

Nakahira, K., Takahashi, T., Shimizu, H., Maeshima, K., Uehara, K., Fujii, H., Nakatsuka, H., Yokoyama, M., Akagi, R. & Morita, K. (2003) Protective role of heme oxygen-ase-1 induction in carbon tetrachloride-induced hepatotox-icity. Biochem. Pharmacol., 66, 1091-1105.

Otterbein, L.E., Mantell, L.L. & Choi, A.M. (1999) Carbon monoxide provides protection against hyperoxic lung inju-ry. Am. J. Physiol., 276, L688-694.

Poss, K.D. & Tonegawa, S. (1997a) Heme oxygenase 1 is required for mammalian iron reutilization. Proc. Natl. Acad. Sci. USA, 94, 10919-10924.

Poss, K.D. & Tonegawa, S. (1997b) Reduced stress defense in heme oxygenase 1-deficient cells. Proc. Natl. Acad. Sci. USA, 94, 10925-10930.

Postic, C., Shiota, M., Niswender, K.D., Jetton, T.L., Chen, Y., Moates, J.M., Shelton, K.D., Lindner, J., Cherrington, A.D., & Magnuson, M.A. (1999) Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J. Biol. Chem., 274, 305-315.

Shibahara, S. (2003) The heme oxygenase dilemma in cellular homeostasis: new insights for the feedback regulation of heme catabolism. Tohoku J. Exp. Med., 200, 167-186.

Sun, J., Hoshino, H., Takaku, K., Nakajima, O., Muto, A., Suzuki, H., Tashiro, S., Takahashi, S., Shibahara, S., Alam, J., Taketo, M.M., Yamamoto, M. & Igarashi, K. (2002) Hemoprotein Bach1 regulates enhancer availability of heme oxygenase-1 gene. EMBO J., 21, 5216-5224.

Stocker, R., Yamamoto, Y., McDonagh, A.F., Glazer, A.N. & Ames, B.N. (1987) Bilirubin is an antioxidant of possible physiological importance. Science, 235, 1043-1046.

Tong, K.I., Padmanabhan, B., Kobayashi, A., Shang, C., Hirotsu, Y., Yokoyama, S. & Yamamoto, M. (2007) Differ-ent electrostatic potentials define ETGE and DLG motifs as hinge and latch in oxidative stress response. Mol. Cell. Biol., 27, 7511-7521.

Vidal, F., Sage, J., Cuzin, F. & Rassoulzadegan, M. (1998) Cre expression in primary spermatocytes: a tool for genetic engineering of the germ line. Mol. Reprod. Dev., 51, 274-280.