NF-E2-related factor 2 promotes compensatory liver hypertrophy after portal vein branch ligation in mice

Authors

  • Keiichi Shirasaki,

    1. Departments of Medical Biochemistry, Graduate School of Medicine, Tohoku University, Sendai, Japan
    2. Gastroenterological Surgery, Graduate School of Medicine, Tohoku University, Sendai, Japan
    3. Department of Gene Expression Regulation, Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Keiko Taguchi,

    1. Departments of Medical Biochemistry, Graduate School of Medicine, Tohoku University, Sendai, Japan
    2. Department of Gene Expression Regulation, Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Michiaki Unno,

    1. Gastroenterological Surgery, Graduate School of Medicine, Tohoku University, Sendai, Japan
    Search for more papers by this author
  • Hozumi Motohashi,

    Corresponding author
    1. Department of Gene Expression Regulation, Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan
    • Address reprint requests to: Masayuki Yamamoto, M.D., Ph.D., Department of Medical Biochemistry, Graduate School of Medicine, Tohoku University, 2-1 Seiryo-cho, Aoba-ku, Sendai, Miyagi, 980-8575, Japan. E-mail: masiyamamoto@med.tohoku.ac.jp; fax: +81-22-717-8090; or Hozumi Motohashi, M.D., Ph.D., Department of Gene Expression Regulation, Institute of Development, Aging and Cancer, Tohoku University, 4-1 Seiryo-cho, Aoba-ku, Sendai, Miyagi, 980-8575, Japan. E-mail: hozumim@idac.tohoku.ac.jp; fax +81-22-717-8554.

    Search for more papers by this author
  • Masayuki Yamamoto

    Corresponding author
    1. Departments of Medical Biochemistry, Graduate School of Medicine, Tohoku University, Sendai, Japan
    • Address reprint requests to: Masayuki Yamamoto, M.D., Ph.D., Department of Medical Biochemistry, Graduate School of Medicine, Tohoku University, 2-1 Seiryo-cho, Aoba-ku, Sendai, Miyagi, 980-8575, Japan. E-mail: masiyamamoto@med.tohoku.ac.jp; fax: +81-22-717-8090; or Hozumi Motohashi, M.D., Ph.D., Department of Gene Expression Regulation, Institute of Development, Aging and Cancer, Tohoku University, 4-1 Seiryo-cho, Aoba-ku, Sendai, Miyagi, 980-8575, Japan. E-mail: hozumim@idac.tohoku.ac.jp; fax +81-22-717-8554.

    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • This work was supported by JSPS KAKENHI (grant nos.: 24249015 [to M.Y.], 24390075 [to H.M.], and 24790307 [to K.T.]), MEXT KAKENHI (grant nos. 23116002 [to H.M.] and 25117703 [to K.T.]), the Gushinkai Foundation (to K.T.), the Naito Foundation (to M.Y.), the Takeda Scientific Foundation (to H.M. and M.Y.), and the Core Research for Evolutional Science and Technology from the JST (to H.M. and M.Y.).

Abstract

Hepatectomy is a standard therapy that allows liver cancer patients to achieve long-term survival. Preceding hepatectomy, portal vein embolization (PVE) is frequently performed to increase the remnant liver size and reduce complications. Although the clinical importance of PVE is widely accepted, molecular mechanisms by which PVE leads to compensatory hypertrophy of nonembolized lobes remain elusive. We hypothesized that NF-E2-related factor 2 (Nrf2), a master regulator of cytoprotection, promotes compensatory liver hypertrophy after PVE. To address this hypothesis, we utilized three mouse lines and the portal vein branch ligation (PVBL) technique, which primarily induces the redistribution of the portal bloodstream in liver in a manner similar to PVE. PVBL was conducted in Kelch-like ECH-associated protein 1 (Keap1) conditional knockout (Keap1-CKO) mice in which Nrf2 is constitutively activated, along with Nrf2-deficient (Nrf2-KO) mice. We found that hypertrophy of nonligated lobes after PVBL was enhanced and limited in Keap1-CKO and Nrf2-KO mice, respectively, compared to wild-type mice. In Keap1-CKO mice, Nrf2 activity was increased, consistent with transient activation of the phosphoinositide 3-kinase/protein kinase B (PI3K/Akt) pathway, and reactive hepatocyte proliferation was significantly prolonged after PVBL. Importantly, Nrf2 activation by a chemical inducer was also effective for enhancement of hypertrophy after PVBL. Conclusion: Nrf2 supports compensatory liver hypertrophy after PVBL. This finding is particularly intriguing, because the primary effect of PVBL is limited to the alteration of bloodstream; this effect is much milder than changes resulting from hepatectomy, in which intrahepatic bloodstream and bile production cease. Our results suggest that premedication with an Nrf2 inducer may be a promising strategy to improve the outcome of PVE; this approach expands the indication of hepatectomy to patients with poorer liver function. (Hepatology 2014;59:2371–2382)

Abbreviations
Akt

protein kinase B

ALB

albumin

ALT

alanine aminotransferase

AST

aspartate aminotransferase

βTrCP

β-transducin repeat-containing protein

CDDO-Im

2-cyano-3,12 dioxooleana-1,9-dien-28-imidazolide

CKO

conditional knockout

G6pd

glucose-6-phosphate dehydrogenase

Gclc

glutamate-cysteine ligase catalytic subunit

Gclm

glutamate-cysteine ligase, modifier subunit

Gsk3

glycogen synthase kinase 3

IHC

immunohistochemical

Keap1

Kelch-like ECH-associated protein 1

KO

knockout

LDH

lactate dehydrogenase

mRNA

messenger RNA

Nqo1

NAD(P)H dehydrogenase quinone 1

Nrf2

NF-E2-related factor 2

PCNA

proliferating cell nuclear antigen

PI3K

phosphoinositide 3-kinase

PVBL

portal vein branch ligation

PVE

portal vein embolization

qPCR

quantitative polymerase chain reaction

Tkt

transketolase

WT

wild-type

An increasing number of patients suffer from primary liver cancers and metastatic liver cancers in both developing and developed countries.[1] Hepatectomy is a standard curative therapy that allows for the long-term survival of liver cancer patients.[2] In spite of the technical advancements, major hepatectomy still places patients at risk for complications caused by liver failure. The liver volume that remains after surgery has been shown to be a strong predictor of postoperative complications.[3] Portal vein embolization (PVE), developed by Makuuchi et al.,[4] is a preoperative treatment designed to prevent postoperative liver failure resulting from insufficient remnant liver mass. Selective PVE produces the atrophy of the embolized lobe and the compensatory hypertrophy of the contralateral lobe, which increases the remnant liver mass and reduces the occurrence of postoperative liver failure. Although PVE is widely accepted as a beneficial procedure for patients with primary and secondary liver tumors undergoing major hepatectomy,[5] molecular mechanisms by which PVE leads to this compensatory hypertrophy of liver remain elusive.

The transcription factor, NF-E2-related factor 2 (Nrf2) is a master regulator of cytoprotective genes in response to oxidative and electrophilic stresses.[6] Under normal conditions, Nrf2 is constantly ubiquitinated by Kelch-like ECH-associated protein 1 (Keap1), resulting in the rapid degradation of Nrf2 through the proteasome pathway. Under conditions of oxidative and electrophilic stress, Keap1 is inactivated. Stabilized Nrf2 is translocated into the nucleus, binds to antioxidant/electrophile response elements, and activates the transcription of cytoprotective genes encoding detoxification enzymes (e.g., NAD(P)H dehydrogenase quinone 1 [Nqo1]), antioxidant proteins, drug transporters, and enzymes for glutathione synthesis (e.g., glutamate-cysteine ligase catalytic subunit [Gclc] and glutamate-cysteine ligase, modifier subunit [Gclm]).[7] In addition to its cytoprotective function, recent studies have demonstrated critical roles for Nrf2 in cell proliferation, particularly in cancer cells.[8] We have reported that Nrf2 stimulates the expression of metabolic genes (e.g., glucose-6-phosphate dehydrogenase [G6pd] and transketolase [Tkt]) and is activated more strongly in the presence of proliferative signals, particularly the sustained activation of the phosphoinositide 3-kinase/protein kinase B (PI3K/Akt) pathway. Consequently, Nrf2 modulates cellular metabolism and helps to achieve the metabolic reprogramming that is advantageous for cell proliferation. Other studies have also demonstrated that Nrf2 deficiency impairs liver regeneration after hepatectomy,[9, 10] suggesting an important role for Nrf2 in the promotion of cell proliferation. Thus, an emerging possibility is that Nrf2 promotes compensatory liver hypertrophy after PVE.

To prove this hypothesis, we exploited the portal vein branch ligation (PVBL) technique in mice, which primarily induces the redistribution of the portal bloodstream in liver in a manner similar to PVE. We examined three lines of mice with different levels of Nrf2 activity: Keap1 conditional knockout (Keap1-CKO) mice, in which Nrf2 is constitutively activated; Nrf2-deficient (Nrf2-KO) mice; and wild-type (WT) mice. We found that Keap1-CKO mice exhibit the most prominent hypertrophy of nonligated lobes after PVBL. Importantly, pretreatment with 2-cyano-3,12 dioxooleana-1,9 dien-28-imidazolide (CDDO-Im), one of the most potent Nrf2 inducers, promotes the hypertrophy of nonligated lobes after PVBL. Therefore, Nrf2 actively supports the compensatory liver hypertrophy induced by PVBL; the primary effects of PVBL are limited to alteration of the bloodstream and are much milder than the changes in the intrahepatic bloodstream and bile production that result from hepatectomy. Our results implicate that premedication of preoperative patients with Nrf2 inducers is a promising strategy to improve the results of PVE and subsequent hepatectomy.

Materials and Methods

Animals

Keap1flox/flox::Albumin-Cre (Keap1-CKO), Nrf2–/– (Nrf2-KO), and WT mice in the C57BL6/J genetic background were used. All of the mice were treated according to the regulations of The Standards for Human Care and Use of Laboratory Animals of Tohoku University (Sendai, Japan) and Guidelines for Proper Conduct of Animal Experiments of the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

Mouse PVBL Model

The procedure of PVBL was previously described.

CDDO-Im Treatment

Mice were dosed by the oral route with vehicle or CDDO-Im (30 µmol/kg body weight, obtained from Mochida Pharmaceuticals Co., Ltd., Tokyo, Japan).

Blood and Tissue Analysis

Blood and liver samples were taken for biochemical and histological analyses, respectively.

Immunohistochemistry

Anti-Ki67 (Dako, Glostrup, Denmark) and anti-proliferating cell nuclear antigen (PCNA; Invitrogen, Carlsbad, CA) antibodies were used for assessing proliferating cells.

Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR)

Complementary DNA was synthesized from total RNA using SuperScript III Reverse Transcriptase (Invitrogen). An Applied Biosystems 7300 PCR system was used for quantitative polymerase chain reaction (qPCR) analysis (Applied Biosystems, Foster City, CA).

Immunoblot Analysis

Whole-cell and nuclear extracts were prepared from livers. Antibodies used in this analysis are described in the Supporting information.

Statistical Analysis

All differences were analyzed with the Student t test. P < 0.05 was considered statistically significant.

For materials and methods described in more detail, see the Supporting Information

Results

Liver-Specific Disruption of Keap1 Gene Enhances the Compensatory Hypertrophy of Nonligated Lobes After PVBL

To clarify the contribution of Nrf2 to the compensatory liver hypertrophy after PVBL, we exploited three mouse lines that have different levels of Nrf2 activity for the PVBL experiment: Nrf2-KO; Keap1-CKO; and WT mice. Nrf2 activity is absent in Nrf2-KO mice, inducible and transient in WT mice, and constitutively high in Keap1-CKO mice. We ligated the left branch of the portal vein and analyzed the remnant right lobes (nonligated lobes) on days 1, 3, and 7 after PVBL. Macroscopic observation of PVBL livers revealed that the size increase of nonligated lobes was more prominent in Keap1-CKO mice than in Nrf2-KO and WT mice (Fig. 1A, white circles). We observed necrotic spots in ligated lobes of Nrf2-KO and WT livers on days 1 and 3 (Fig. 1A, red arrows), but the spots were not obvious in ligated lobes of Keap1-CKO.

Figure 1.

Compensatory hypertrophy of nonligated lobes after PVBL. Keap1-CKO, Nrf2-KO, and WT mice were examined before PVBL (day 0) on days 1, 3, and 7 after PVBL. (A) Macroscopic observation of whole livers. Nonligated lobes are indicated with white circles. Necrotic foci in ligated lobes are indicated with red arrows. (B) Nonligated lobes/body-weight ratios. Data are the means ± standard deviation of 5-7 mice. *P < 0.05 and **P < 0.01, compared to WT mice on each day; #P < 0.05 and ##P < 0.01, compared with each genotype on day 0. (C) Increase in nonligated lobes/body-weight ratio per day. Average values during three periods (days 0-1, 1-3, and 3-7) were calculated. **P < 0.01.

We calculated right lobes/body-weight ratios (before PVBL) and nonligated lobes/body-weight ratios (after PVBL) for the three mouse lines (Fig. 1B). After PVBL, the ratios became significantly higher than the initial ratios at all time points, irrespective of genotypes (Fig. 1B). Whereas the initial ratio before PVBL was comparable between Nrf2-KO and WT mice, Nrf2-KO mice displayed lower ratios on days 3 and 7 than WT mice, indicating that the reactive hypertrophy of nonligated lobes is limited in Nrf2-KO mice. In Keap1-CKO mice, the initial ratio and the ratios after PVBL were all higher than in the other two mouse lines, suggesting that constitutive activation of Nrf2 is advantageous for hepatocyte proliferation during development and reactive hypertrophy of livers.

When increments of the ratios per day were compared (Fig. 1C), the values were higher in Keap1-CKO mice than WT mice during days 3-7, suggesting that the reactive hypertrophy is prolonged in Keap1-CKO livers. In Nrf2-KO mice, the values were low during days 1-3, but recovered during days 3-7, suggesting that reactive hypertrophy is delayed in Nrf2-KO livers.

We confirmed the technical adequacy of portal vein ligation by microscopic observation (Fig. 2A). Hepatic arteries (marked by A in Fig. 2A) were dilated in ligated lobes of all three mouse lines, compared with nonligated lobes. Indeed, minor diameter ratios of hepatic arteries against bile ducts were significantly higher in ligated lobes than nonligated lobes (Fig. 2B). These results indicate that the ligations of the portal branch were successful.

Figure 2.

Compensatory arterial dilatation of ligated lobes. (A) Microscopic observations of livers from Keap1-CKO, Nrf2-KO, and WT mice after PVBL stained with hematoxylin and eosin. N, nonligated lobes; L, ligated lobes; P, portal vein; A, hepatic artery; B, bile duct. (B) Minor diameter ratios of hepatic arteries against bile ducts. **P < 0.01, #P < 0.05, and ##P < 0.01, compared to sham controls of each genotype.

Damage After PVBL in Ligated Lobes Is Less Severe in Keap1-CKO Mice

We subsequently examined the severity of liver damage after PVBL. Plasma levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and lactate dehydrogenase (LDH) were measured as markers for hepatocyte injury. Whereas Nrf2-KO and WT mice displayed transient elevation of AST, ALT, and LDH on day 1, Keap1-CKO mice maintained normal levels of these markers (Fig. 3A-C). The level of plasma albumin (ALB), a marker for liver function, was within the normal range for all three mouse lines (Fig. 3D). Perioperative weight changes were not statistically significant in any of the mouse lines (data not shown). These data indicate that PVBL provokes transient liver damage in ligated lobes and that Keap1-CKO livers are resistant to the stress provoked by PVBL.

Figure 3.

Biochemical examination of blood plasma before and after PVBL. Plasma levels of AST (A), ALT (B), LDH (C), and ALB (D) examined before PVBL (day 0) on days 1, 3, and 7 after PVBL are shown. Data are the means ± standard deviation of 5-7 mice. #P < 0.05, compared with each genotype on day 0.

Reactive Cell Proliferation Is Prolonged in Nonligated Lobes of Keap1-CKO Mice

To understand mechanisms by which increased Nrf2 activity enhances liver hypertrophy, we examined the cell proliferation status at each stage after PVBL by utilizing immunohistochemical (IHC) detection of Ki67. We found that nonligated lobes of all mouse lines contained robustly increased Ki67-positive cells with uniform distribution throughout hepatic lobules on day 3 (Fig. 4A). Importantly, on day 7, a substantial number of Ki67-positive cells were still detectable in Keap1-CKO livers, but not in livers of the other mouse lines. We quantified the Ki67-positive cells in a section from three independent samples of nonligated and ligated lobes of each mouse line and found a significant difference in the Ki67-positive cell number between nonligated and ligated lobes of Keap1-CKO livers on day 7 (Fig. 4B). These results further support the hypothesis that Keap1-CKO hepatocytes keep proliferating for a longer time after PVBL.

Figure 4.

Proliferation markers in nonligated lobes after PBL. (A and D) IHC detection of Ki67 (A) and PCNA (D) in nonligated lobes. C, central vein; p, portal vein. Arrowheads indicate PCNA-positive cells. (B) Ki67-positive hepatocytes were counted using TissueFAXS, and quantitative data are shown. (C) Ki67 mRNA levels measured by RT-qPCR. Data are the means ± standard deviation of 3-4 mice. *P < 0.05 and **P < 0.01; #P < 0.05 and ##P < 0.01, compared to sham controls of each genotype. S, sham operated; N, nonligated lobes; L, ligated lobes.

We also examined messenger RNA (mRNA) levels of Ki67. Consistent with the level of Ki67 protein expression, Ki67 mRNA levels were higher in nonligated lobes than ligated lobes on day 3 in all three mouse lines (Fig. 4C). In nonligated lobes on day 3, Keap1-CKO mice expressed significantly higher levels of Ki67 mRNA than the other two lines, implying that Keap1-CKO hepatocytes proliferate more actively. On day 7, Ki67 mRNA levels in nonligated lobes were mostly maintained at the day 3 levels in both Keap1-CKO and WT mice, but decreased in Nrf2-KO mice. These results in aggregate demonstrate that reactive proliferation after PVBL is prolonged in Keap1-CKO livers.

To further validate this notion, we detected PCNA as a second marker of cell proliferation (Fig. 4D). On day 3, many PCNA-positive cells were detected in nonligated lobes of all mouse lines with uniform distribution. On day 7, PCNA-positive cells were still present in Keap1-CKO mice, but hardly detected in WT and Nrf2-KO mice. These results were almost the same with those of Ki67 staining, verifying the prolonged cell proliferation in Keap1-CKO livers after PVBL.

Akt and Glycogen Synthase Kinase 3 Phosphorylation Is Enhanced in Nonligated Lobes After PVBL

Nrf2 accelerates cell proliferation in the presence of proliferative signals, particularly under the sustained activation of the PI3K/Akt pathway, which enhances the nuclear accumulation of Nrf2 and increased expression of Nrf2 target genes.[8] To test whether this notion is applicable to the reactive proliferation of hepatocytes after PVBL, we examined the status of the PI3K/Akt signaling pathway after PVBL by detecting two forms of phosphorylated Akt using immunoblotting analysis. Because inactivation of glycogen synthase kinase 3 (Gsk3) by phosphorylation inhibits Keap1-independent degradation of Nrf2, enhancing nuclear accumulation of Nrf2,[12, 13] we also examined Gsk3, a substrate of activated Akt.

Because PVBL is an invasive procedure, mice were kept in the free-feeding condition to avoid any nutritional deficits. Because Akt and Gsk3 phosphorylation is strongly influenced by the feeding, intact control mice and sham-operated mice indeed exhibited wide variations in the phosphorylation levels (Supporting Fig. 1). Moreover, there was no association between Akt and Gsk3 phosphorylation levels. Thus, we compared phosphorylation levels between ligated (L) and nonligated (N) lobes of a single mouse (N/L ratios are indicated beneath corresponding paired bands in Fig. 5A).

Figure 5.

Phosphorylation levels of Akt and Gsk3 and nuclear accumulation of Nrf2 on days 1, 3, and 7 after PVBL. Immunoblotting analysis of cytoplasmic (A) and nuclear (B) fractions of livers. Lamin B was used as loading control for nuclear proteins (B). Arrowheads indicate Gsk3-α and -β (A) and Nrf2 (B). S, sham operated; N, nonligated lobes; L, ligated lobes; N/L, ratios of normalized band intensities of nonligated versus ligated lobes.

On day 1, N/L ratios of Akt phosphorylation (S473 and/or T308) was above 1.0 in the majority of mice irrespective of genotypes (Fig. 5A). On days 3 and 7, N/L ratios of Akt S473 mostly remained above 1 (Fig. 5A). Similarly, N/L ratios of Gsk3 phosphorylation was above 1 on days 1, 3, and 7 (Fig. 5A). Phosphorylation of Akt and Gsk3 was already enhanced in nonligated lobes by 6 hours after PVBL (Supporting Fig. 2). Therefore, PVBL promotes Akt and Gsk3 phosphorylation in nonligated lobes, which continues up to day 7 after PVBL.

We then examined whether nuclear accumulation of Nrf2 was enhanced in nonligated lobes where the PI3K/Akt pathway was activated. Whereas Nrf2 protein was barely detectable in any samples of WT mice, Keap1-CKO livers accumulated detectable amounts of Nrf2 at all stages after PVBL (Fig. 5B). In Keap1-CKO mice, levels of Nrf2 were higher in nonligated lobes than ligated lobes on days 3 and 7; this effect was modest, but parallel with Gsk3 phosphorylation levels. We recently demonstrated that inhibition of Keap1-independent Nrf2 degradation by Gsk3 phosphorylation effectively increases Nrf2 abundance in the absence of Keap1.[14] Thus, it is plausible that long-lasting Gsk3 phosphorylation is responsible for the enhancement of Nrf2 accumulation up to day 7 and prolonged reactive proliferation of hepatocytes in Keap1-CKO mice.

Expressions of Nrf2 Target Genes Are Elevated in Nonligated Lobes

We next examined whether the modest increase of Nrf2 accumulation in nonligated versus ligated lobes of Keap1-CKO livers was reflected in the functional enhancement of Nrf2 by measuring expression levels of Nrf2 target genes. We also verified the Nrf2 dependency of these gene expressions by analyzing Nrf2-KO mice.

We found that expression of Nqo1 in Keap1-CKO mice was elevated in nonligated lobes, compared with ligated lobes, on days 3 and 7 (Fig. 6A). Because the expression level of Nqo1 mRNA was significantly lower in Nrf2-KO and WT livers than in Keap1-CKO livers, we expanded the relative expression levels of the former in Fig. 6A (insert). A significant elevation of Nqo1 mRNA was also observed in nonligated lobes versus ligated lobes of WT mice on days 3 and 7, which was completely absent in Nrf2-KO mice. These results indicate that Nqo1 expression in WT nonligated lobes on days 3 and 7 is dependent on Nrf2, and that Nrf2 is activated in response to PVBL in WT nonligated lobes.

Figure 6.

Expression levels of Nrf2 target genes after PVBL. Relative mRNA levels of Nrf2 target genes measured by RT-qPCR. Data are the means ± standard deviation of 4 mice. *P < 0.05; **P < 0.01; #P < 0.05 and ##P < 0.01, compared to sham controls of each genotype. S, sham operated; N, nonligated lobes; L, ligated lobes.

We also examined the expressions of four additional target genes, G6pd, Tkt, Gclm, and Gclc, and found that on day 3 after PBL, the mRNAs corresponding to these four genes were higher in nonligated lobes than ligated lobes and/or sham controls of Keap1-CKO livers (Fig. 6B-E). A similar tendency was observed on day 7.

Expression levels of two indifferent genes, peroxisome proliferator-activated receptor gamma and sterol regulatory element-binding transcription factor 1, and two genes involved in liver regeneration, hepatocyte growth factor and cyclin D1, did not show apparent differences between nonligated lobes and ligated lobes/sham controls of Keap1-CKO mice (Supporting Fig. 3), excluding the possibility that elevation of the aforementioned Nrf2-target genes is rather nonspecifically induced by PVBL. These results suggest that increased activity of Nrf2 in Keap1-CKO livers is further potentiated in nonligated lobes and contributes to prolonged proliferation of hepatocytes in nonligated lobes.

CDDO-Im Accelerates Liver Hypertrophy of Nonligated Lobes

We finally tested whether chemically induced activation of Nrf2 is equally effective for enhancing hypertrophy of nonligated lobes after PVBL. To this end, we administered CDDO-Im, which stabilizes Nrf2 by inactivating Keap1,[15] every other day from 2 days before surgery until day 6 after PVBL (Fig. 7A). Mice were sacrificed for analysis on day 7 after PVBL.

Figure 7.

Effects of CDDO-Im on the compensatory hypertrophy of nonligated lobes in Nrf2-KO and WT mice. (A) Protocol of CDDO-Im administration. (B) Macroscopic observation on day 7 after PVBL. Nonligated lobes are indicated with white circles. (C) Nonligated lobes/body-weight ratios in Nrf2-KO and WT mice on day 7 after sham operation or PVBL. Right lobes were measured instead of nonligated lobes in sham controls. (D) Expression of Nqo1 measured by RT-qPCR. Data are the means ± standard deviation of 5-6 mice. *P < 0.05; **P < 0.01; #P < 0.05 and ##P < 0.01, compared to each group of WT mice. (E and F) Immunoblotting analysis of nuclear (E) and cytoplasmic (F) fractions of livers. Arrowheads indicate Nrf2 (E) and Gsk3-α/β and α-tubulin (F). S, sham operated; N, nonligated lobes; L, ligated lobes; N/L, ratios of normalized band intensities of nonligated versus ligated lobes.

The efficacy of CDDO-Im as an Nrf2 inducer in liver was confirmed by the increased accumulation of Nrf2 and robust elevation of Nqo1 expression after treatment with CDDO-Im (Fig. 7D,E, WT). Nqo1 expression levels were much lower in Nrf2-KO mice than in WT mice, and no increase was observed after CDDO-Im treatment (Fig. 7D,E, Nrf2-KO). CDDO-Im treatment did not change the plasma levels of ALT, AST, LDH, and ALB in Nrf2-KO mice or WT mice (Supporting Fig. 4), indicating that CDDO-Im did not show adverse effects at the dose used in this experiment.

In WT mice, CDDO-Im treatment caused slight hypertrophy in sham-operated mice by 118%. In PVBL-operated mice, CDDO-Im treatment enhanced the hypertrophy of nonligated lobes by 131% (Fig. 7B,C, WT). The fold changes in the nonligated lobes/body-weight ratios by PVBL were 242% and 269% in vehicle-treated and CDDO-Im-treated mice, respectively. Thus, CDDO-Im treatment promotes liver enlargement. To prove the Nrf2-dependency of this effect, we conducted similar experiments using Nrf2-KO mice. The increased liver enlargement resulting from CDDO-Im treatment that was observed in WT mice was completely canceled in Nrf2-KO mice (Fig. 7B,C, Nrf2-KO). Therefore, CDDO-Im promotes liver hypertrophy by activating Nrf2.

We further analyzed Akt and Gsk3 phosphorylation levels in these samples (Fig. 7F). N/L ratios of Akt and Gsk3 phosphorylation were mostly above 1.0 on day 7 after PVBL irrespective of genotypes or CDDO-Im treatment. N/L ratios of Nqo1 expression were also above 1.0 for WT mice, suggesting that Nrf2 activity was augmented in nonligated lobes. Thus, not only genetic, but also pharmacological inactivation of Keap1 cooperates with PI3K/Akt pathway activation to enhance Nrf2 activity. In aggregate, these results demonstrate that pharmacological induction of Nrf2 is effective for the promotion of compensatory liver hypertrophy after PVBL.

Discussion

The roles that Nrf2 plays in regulation of cell proliferation have been attracting wide-ranging attention. Aberrant accumulation of Nrf2 in cancer cells strongly accelerates their proliferation,[8] but Nrf2 does not inherently drive uncontrolled cell division. This multifaceted function of Nrf2 is nicely demonstrated in Keap1-null mice, which exhibit excessive proliferation of cells in specific lineages at specific differentiation stages. In forestomach of Keap1-null mice, despite the similar accumulation level of Nrf2, basal layer cells proliferate vigorously in an Nrf2-dependent manner, whereas spinous layer cells, which are more differentiated than basal layer cells, do not proliferate as vigorously.[8] Similarly, Keap1-null megakaryocytes exhibit a transient acceleration of cell-cycle progression in the course of differentiation in an Nrf2-dependent manner.[16] In both cases, Nrf2 does not efficiently enhance cell proliferation if the proliferative signals diminish. Consistent with this notion, our current study demonstrates that constitutively stabilized/activated Nrf2 promotes hepatocyte proliferation transiently after PVBL. Consistent with these observations, our long-term inspection indicates that Keap1 knockdown (i.e., Nrf2-constitutively activated) mice do not initiate carcinogenesis per se.[17] Based on these observations, we propose that Nrf2 acts as a context-dependent or facultative accelerator of cell proliferation.

We found that Gsk3 phosphorylation, Nrf2 accumulation, Nrf2 target gene up-regulation, and hepatocyte proliferation, which are observed on days 3 and 7 after PVBL, are all augmented in nonligated versus ligated lobes of Keap1-CKO mice. It has been shown that Nrf2 phosphorylated by Gsk3 is subjected to the β-transducin repeat-containing protein (βTrCP)/Cul1-dependent ubiquitination, which is a Keap1-independent mechanism of Nrf2 degradation.[12, 13] In nonligated lobes of Keap1-CKO mice, where Gsk3 is phosphorylated and inactivated, βTrCP-dependent degradation of Nrf2 ceases, resulting in augmentation of Nrf2 accumulation and target gene up-regulation. Simultaneous inactivation of Keap1-dependent and -independent degradation systems of Nrf2 appears to underlie the enhancement of Nrf2 activity and Nrf2-mediated cell proliferation in nonligated lobes of Keap1-CKO mice.

One of the salient results of this study is that oral administration of an Nrf2 inducer, CDDO-Im, is effective to enhance compensatory liver hypertrophy after PVBL. The straightforward extension of our result is the clinical application of Nrf2 inducers for liver cancer patients undergoing PVE and subsequent hepatectomy. A larger liver remnant size will be expected, complications will be reduced, and indications for hepatectomy will be expanded to patients with poorer liver function. Our study has raised the possibility of a new therapeutic procedure that achieves a higher level of safety for hepatectomy.

When Nrf2 inducers are given to liver cancer patients before hepatectomy, one possible drawback would be exacerbation of cancers because the Nrf2 inducer might promote proliferation of cancer cells and confer resistance against chemo- and radiotherapy. On the other hand, cancer cells have often already established high levels of Nrf2 through multiple pathways[18] and additional chemical induction of Nrf2 may be negligible. In addition, Nrf2 inducers have been reported to strengthen anticancer immunity.[19] Therefore, we surmise that optimization of the medication period and dose of Nrf2 inducers will maximize the beneficial effect on compensatory liver hypertrophy and minimize adverse effects on cancer proliferation.

Acknowledgment

The authors thank Ms. Eriko Naganuma for her great contribution to the histological analyses and the Biomedical Research Core of Tohoku University Graduate School of Medicine for technical support.

Ancillary