Skip Navigation


Carcinogenesis Advance Access originally published online on June 26, 2008
Carcinogenesis 2008 29(11):2218-2226; doi:10.1093/carcin/bgn135
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
29/11/2218    most recent
bgn135v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Takahashi, M.
Right arrow Articles by Nishikawa, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Takahashi, M.
Right arrow Articles by Nishikawa, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Cellular distributions of molecules with altered expression specific to the tumor promotion process from the early stage in a rat two-stage hepatocarcinogenesis model

Miwa Takahashi1, Makoto Shibutani1,3,*, Gye-Hyeong Woo1, Kaoru Inoue1, Hitoshi Fujimoto1, Katsuhide Igarashi2, Jun Kanno2, Masao Hirose1,4 and Akiyoshi Nishikawa1

1 Division of Pathology
2 Division of Molecular Toxicology, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-ku, Tokyo 158-8501, Japan
3 Laboratory of Veterinary Pathology, Tokyo University of Agriculture and Technology, 3-5-8 Saiwai-cho, Fuchu City, Tokyo 183-8509, Japan
4 Food Safety Commission, 2-13-10 Prudential Tower 6th Floor, Nagata-cho, Chiyoda-ku, Tokyo 100-8989, Japan

* To whom correspondence should be addressed. Tel: +81 42 367 5874; Fax: +81 42 367 5771; Email: mshibuta{at}cc.tuat.ac.jp


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
A global gene expression profiling specific to the early process of tumor promotion by fenbendazole (FB) or phenobarbital (PB) in a rat two-stage hepatocarcinogenesis model revealed 33 genes to show altered expression in common with both chemicals. The immunohistochemical distribution of transferrin receptor (Tfrc), nuclear receptor subfamily 0, group B, member 2 (Nr0b2) and minichromosome maintenance deficient 6 (MCM6), included in the altered expression profile, were therefore examined in FB- and PB-induced proliferative lesions at both early and late stages of tumor promotion. In addition, immunoexpression of transforming growth factor β receptor (TGFβR) I, TGFβRII, phosphatase and tensin homolog deleted on chromosome 10 (PTEN) and phosphorylated phosphatase and tensin homolog deleted on chromosome 10 (pPTEN) was also examined. In the early stage, most hepatocellular foci positive for glutathione S-transferase placental form (GST-P) showed co-expression of TGFβRI and lack of PTEN and pPTEN, some GST-P-positive foci co-expressing Tfrc and Nr0b2. In the late stage, selective expression of TGFβRI, but not TGFβRII, was also observed in many adenomas and carcinomas consistently expressing GST-P. Nr0b2 was variably expressed in the proliferative lesions, irrespective of the carcinogenic stage. Like the GST-P-positive foci, adenomas and carcinomas consistently lacked PTEN and pPTEN. Expression of Tfrc and MCM6 was increased in parallel with the carcinogenic stage. In conclusion, loss of PTEN and dysregulation of transforming growth factor β signaling can be considered to be involved in rat hepatocarcinogenesis from early stages. Selective expression of Tfrc in proliferative lesions suggests an involvement of changes in iron homeostasis during the process of tumor promotion/progression driven by FB or PB.

Abbreviations: CYP, cytochrome P450; DEN, N-diethylnitrosamine; FB, fenbendazole; GST-P, glutathione S-transferase placental form; MCM6, minichromosome maintenance deficient 6; Nr0b2, nuclear receptor subfamily 0, group B, member 2; PH, partial hepatectomy; PB, phenobarbital; PTEN, phosphatase and tensin homolog deleted on chromosome 10; pPTEN, phosphorylated phosphatase and tensin homolog deleted on chromosome 10; RT, reverse transcription; PCR, polymerase chain reaction; Tfrc, transferrin receptor; TGF, transforming growth factor; TGFβR, transforming growth factor β receptor


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
Molecular events in multistep carcinogenesis due to non-genotoxic carcinogens remain largely unclear. Currently, although acceptable daily intake levels have been established for non-genotoxic carcinogens such as food additives, pesticides and animal drugs contained in food, a more thorough understanding of the mechanisms of non-genotoxic carcinogenesis is needed to secure human health. The carcinogens in question usually exert tumor-promoting activity of target organs in two-stage carcinogenesis models in rodents, and signatures of carcinogenic responses can be determined by analysis of preneoplastic lesions generated at early stages of tumor promotion. This is the rationale for global gene expression profiling focusing on early stages of carcinogenesis for elucidation of putative molecular mechanisms responsible for early carcinogenic actions.

The glutathione S-transferase placental form (GST-P) has been identified as a reliable marker for preneoplastic lesions as end points in rat liver for rapid detection of carcinogenic agents (1). However, it is known that only a small proportion of such foci actually progress to liver tumors. In a recent report, it is suggested that double-positive foci for GST-P and transforming growth factor (TGF) {alpha} in the early stages of rat hepatocarcinogenesis may be most probably to develop into tumors with promotion (2). Also, dysregulation of TGFβ signaling, such as altered expression of transforming growth factor β receptors (TGFβRs), has been considered to play a critical role in the promotion and progression stages of hepatocarcinogenesis, both in humans and rodents (3–5). It has been reported that expression of TGFβ and its receptors is altered in preneoplastic lesions in rat liver during the tumor promotion stage (6,7). Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a tumor suppressor gene involved in a variety of tumor types, and down-regulation of PTEN expression may contribute to the development of human hepatocellular carcinomas (8,9). In rats, clear roles of PTEN in hepatocarcinogenesis remain to be explored (10).

In the present study, to determine molecular mechanisms involved in the non-genotoxic hepatocarcinogenesis in rat liver, we performed a global expression profiling of mRNAs specific to the early stage of tumor promotion by fenbendazole (FB), an anti-helminthic drug (11), and phenobarbital (PB), a well-studied hepatocarcinogenesis promoter (12), in a rat two-stage hepatocarcinogenesis model using a medium-term liver bioassay (13,14). Reduction of connexin 32 in centrilobular hepatocytes as well as cytochrome P450 (CYP) 1A2 induction have been suggested to be involved in the mechanism of liver tumor promotion by FB (11). PB is also known to induce CYP enzymes, particularly CYP2B1/2 and CYP3A2 (15). It has been reported that PB inhibits cell-to-cell communication, stimulates proliferation and inhibits apoptosis of hepatocytes in preneoplastic foci (16). In addition, increased oxidative stresses due to activation of detoxifying enzymes are suggested to be responsible for hepatocarcinogenesis by PB (15). Based on the expression profiles obtained in the present study, we further examined cellular localization of molecules that showed altered expression at the early stage of hepatocarcinogenesis by FB and PB in common, such as transferrin receptor (Tfrc), nuclear receptor subfamily 0, group B, member 2 (Nr0b2) and minichromosome maintenance deficient 6 (MCM6), immunohistochemically in the FB- or PB-induced hepatocellular adenomas and carcinomas as well as in the liver cell foci. We also additionally examined changes in immunolocalization of TGFβRs, PTEN and phosphorylated phosphatase and tensin homolog deleted on chromosome 10 (pPTEN) during the course of hepatocarcinogenesis.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Supplementary material
 Funding
 References
 
Chemicals
FB (CAS No. 43210-67-9) was purchased from Sigma-Aldrich Japan K.K. (Tokyo, Japan). PB sodium (CAS No. 57-30-7) was obtained from Wako Pure Chemicals Industries (Osaka, Japan) and N-diethylnitrosamine (DEN, CAS No. 55-18-5) from Tokyo Chemical Industry, Ltd (Tokyo, Japan).

Animal experiments
Male 5-week-old F344/DuCrj rats were purchased from Charles River Japan (Kanagawa, Japan) and acclimatized on powdered basal diet (CRF-1; Oriental Yeast, Tokyo, Japan) and tap water ad libitum for 1 week. They were housed in polycarbonate cages with sterilized softwood chips as bedding in a barrier-maintained animal room conditioned at 23–25°C and 50–60% humidity on a 12-h light/dark cycle.

In the experiment 1, for gene expression profiling and immunohistochemical examination of rat livers at the early stage of tumor promotion, a two-stage hepatocarcinogenesis model was employed using a medium-term liver bioassay (13,14). A total of 120 rats were divided into 10 groups. Seven groups were initiated with a single intraperitoneal injection of DEN (200 mg/kg, dissolved in saline) and the others received a single intraperitoneal injection of saline vehicle alone. Two weeks later, rats in the DEN-initiated groups were fed diet containing FB 400, 1200 or 3600 p.p.m. (DEN + FB groups) or PB 56, 167 or 500 p.p.m. (DEN + PB groups) or basal diet (DEN-alone group). The rats without initiation were given FB 3600 p.p.m. (FB group) or PB 500 p.p.m. (PB group) or basal diet (untreated control group). The highest doses of FB or PB were set as concentrations that exhibited tumor promotion effects in rat liver in previous reports (FB, 11; PB, 12). The animals were subjected to two-thirds partial hepatectomy (PH) at week 3 except for sham surgery in the untreated control group. At week 8, the animals were killed under deep ether anesthesia by exsanguination, and their livers were immediately removed and weighed. The liver tissues were immersed in RNAlater® solution (Ambion, Austin, TX) at 4°C overnight and then stored at –80°C until use after removing the solution. In addition, liver slices were fixed in 10% phosphate-buffered formalin (pH 7.4) and processed routinely for histopathological examination.

In experiment 2, to investigate immunohistochemical distribution of selected proteins in late stages of tumor promotion, 35 animals each initiated with DEN were given diet containing FB 3600 p.p.m. or PB 500 p.p.m. from 2 weeks after initiation. They were subjected to PH at week 3 and maintained until week 59 to induce hepatocellular tumors. The livers of surviving animals were removed and the liver slices containing neoplastic lesions grossly (1 slice per a rat) were fixed in 10% phosphate-buffered formalin and prepared for histopathological examination.

The treatment groups in experiments 1 and 2 are summarized in supplementary Table S1 (available at Carcinogenesis and Online). The animal protocol was reviewed and approved by the Animal Care and Use Committee of the National Institute of Health Sciences, Japan.

RNA preparation
For microarray analysis and subsequent real-time reverse transcription (RT)–polymerase chain reaction (PCR) analysis, total RNA was isolated from livers of four rats in each group in experiment 1 using an RNeasy Mini Kit (QIAGEN K.K., Tokyo, Japan). Concentrations and quality of RNA were determined with a RiboGreen RNA Quantitation Kit (Molecular Probes, Eugene, OR) and an RNA 6000 Nano LabChip Kit (Agilent Technologies Japan, Ltd, Tokyo, Japan), respectively.

Microarray analysis
Five microgram aliquots of total RNA were subjected to amplification, consisting of RT and subsequent in vitro one-step transcription, using a MessageAmpTM II aRNA amplification Kit (Ambion) with a T7 oligo (dT) primer, according to the manufacturer's protocol. During the in vitro transcription, generated aRNAs were labeled with biotin-16-UTP and biotin-11-CTP (Enzo Biochem, Farmingdale, NY). Aliquots of 20 µg of biotinylated aRNA were fragmented, hybridized with a GeneChip® Rat Genome 230 2.0 Array (Affymetrix, Santa Clara, CA) at 45°C for 18 h, stained with streptavidin–R-phycoerythrin conjugates (Molecular Probes) and then scanned with a GeneChip® Scanner 3000 (Affymetrix).

Selection of genes and normalization of expression data were performed using GeneSpring® software (version 7.2; Silicon Genetics, Redwood City, CA). To normalize chip-wide variation in intensity, per chip normalization was performed by dividing the signal strength for each gene with the level of the 50th percentile of the measurement in the chip, and each gene was divided by the average intensity in the samples of untreated control or DEN-alone groups. Genes showing expression change with differences at least 2-fold in magnitude from the untreated control or DEN-alone groups, as well as the ‘presence’ signal in >3/4 of samples in the group showing higher expression values in comparison were selected. To obtain an expression profile specific to the early stages of tumor promotion by FB or PB in the rat two-stage hepatocarcinogenesis model, genes showing altered expression in the FB-, PB- or DEN-alone group as compared with the untreated control group were selected first. Next, genes showing expression changes in the DEN + FB 3600 p.p.m. or DEN + PB 500 p.p.m. as compared with the DEN-alone group were selected. Then, genes showing altered expression specific to the promotion with FB were collected by subtracting the population of genes showing expression changes in the FB and DEN-alone groups from that showing expression changes with DEN + FB 3600 p.p.m. Similarly, genes showing altered expression specific to PB promotion were identified by subtracting the population of genes showing expression changes in the PB- and DEN-alone groups from that showing expression changes in the DEN + PB 500 p.p.m. Genes showing altered expression in common with both chemicals were also selected. Among genes selected at the highest dose of each carcinogen, those showing dose-related expression changes were further identified by analysis of expression levels with the low- and middle-dose groups for each chemical.

Real-time RT–PCR
Quantitative real-time RT–PCR was performed for confirmation of expression values obtained with microarrays using an ABI Prism 7900HT (Applied Biosystems Japan Ltd, Tokyo, Japan). The following genes were selected as targets: alcohol dehydrogenase 1 (Adh1), syntaxin 6 (Stx6), pregnancy-induced growth inhibitor (Okl38), v-maf musculoaponeurotic fibrosarcoma oncogene family, protein B (Mafb), Nr0b2, dual specificity phosphatase 1 (Dusp1), Axin2 and Igfbp1. One microgram of total RNA was applied to RT with a High-Capacity cDNA Archive Kit (Applied Biosystems Japan Ltd) in a 100-µl total reaction volume. For real-time PCR analysis, ABI Assays-on-DemandTM TaqMan probe and primer sets from Applied Biosystems (available at https://products.appliedbiosystems.com/ab/en/US/adirect/ab?cmd=catNavigate2&catID=601267/) were used except for Adh1. Real-time PCR was performed in a 50-µl reaction volume using the TaqMan probe detection system (Applied Biosystems Japan Ltd) with specific primers, the corresponding TaqMan® MGB probes (FAMTM dye labeled) and RT products. For Adh1, the primer sets were designed using Primer Express® software (version 2.0; Applied Biosystems Japan Ltd), and the sequences were TTG AAG GAA ACA ACT CCA TAT TCA TT (forward) and CAT GGC CGC TCT GCT TCT A (reverse). Amplified transcript levels were measured with the SYBR Green detection system in a 50-µl reaction mixture containing SYBR® Green PCR Master Mix (Applied Biosystems Japan Ltd), target primers and RT products. For quantification of expression data, a standard curve method was applied. Expression values were normalized to two housekeeping genes, glyceraldehyde 3-phosphate dehydrogenase and hypoxanthine–guanine phosphoribosyltransferase.

Immunohistochemistry
Formalin-fixed, paraffin-embedded liver sections were subjected to immunohistochemistry utilizing a VECTASTAIN® Elite ABC Kit (Vector Laboratories, Burlingame, CA) with 3,3'-diaminobenzidine/H2O2 as the chromogen. Rabbit polyclonal antibodies against GST-P (MBL, Nagoya, Japan; Catalog No. 311, 1:1000) were applied for all liver sections obtained at both 6 and 57 weeks of tumor promotion. Serial sections of eight rats of the untreated control, DEN-alone, DEN + FB 3600 p.p.m. and DEN + PB 500 p.p.m. groups at week 6 of promotion as well as all the hepatocellular tumors generated after promotion with FB or PB were subjected to immunohistochemistry for Tfrc (mouse monoclonal antibody, Zymed Laboratories, South San Francisco, CA; Catalog No. 13-6800, 1:200), Nr0b2 (rabbit polyclonal antibody; MBL; Catalog No. LS-A5411, 1:200), MCM6 (rabbit polyclonal antibody; GeneTex, San Antonio, TX; Catalog No. GTX24458, 1:200), TGFβRI (rabbit polyclonal antibody, Santa Cruz Biotechnology, Santa Cruz, CA; Catalog No. sc-398, 1:100), TGFβRII (rabbit polyclonal antibody, Santa Cruz Biotechnology; Catalog No. sc-220, 1:100), PTEN (rabbit polyclonal antibody; Cell Signaling Technology, MA; Catalog No. 9559, 1:100) and pPTEN (rabbit polyclonal antibody; Cell Signaling Technology; Catalog No. 9551, 1:50). pPTEN is an inactive form of PTEN and the PTEN antibody detects endogenous levels of total PTEN.

For antigen retrieval, the sections were heated in 10 mM citrate buffer by autoclaving for 10 min before incubation with the TGFβRI and MCM6 antibodies or for 20 min before incubation for PTEN and pPTEN. In the Nr0b2, Trfc and TGFβRII cases, sections were heated by microwave for 10 min before incubation.

Analysis of immunolocalization
The numbers and areas of GST-P-positive foci and the total areas of livers at week 6 of tumor promotion were measured using an Image Processor for Analytical Pathology (IPAP-WIN; Sumika Technoservice, Osaka, Japan), and then the values per unit area (cm2) of liver section were calculated. Cellular localization of immunohistochemically stained molecules was evaluated in relation with GST-P-positive foci and neoplastic lesions using serial sections without applying double-labeling experiments with GST-P. For evaluation of the immunoreactivity of Tfrc and Nr0b2 in liver cells outside GST-P-positive foci at 6 weeks of promotion, staining intensity was scored as 0 (none), 1 (slight), 2 (moderate) and 3 (strong) by observation of five randomly selected areas/rat at 200-fold magnification. For evaluation of the expression of TGFβRI, TGFβRII, PTEN, pPTEN, Tfrc and Nr0b2 in proliferative lesions (GST-P-positive foci at both 6 and 57 weeks; hepatocellular adenomas and carcinomas at 57 weeks), immunoreactivity was classified as increased, unchanged or decreased as compared with the expression levels of surrounding liver cells. Numbers of MCM6-positive cells and total number of liver cells were counted by observation of five randomly selected areas/rat at 200-fold magnification at week 6 and counted in individual GST-P-positive foci, adenomas and carcinomas at 57 weeks, and then, MCM6-positive cell index was calculated.

Statistical analysis
Data for the numbers and areas of GST-P-positive foci were assessed by one-way analysis of variance or the Kruskal–Wallis test following Bartlett's test. When statistically significant differences were indicated, the Dunnett's multiple test was employed for comparison with the DEN-alone group. The data for gene expression levels from real-time RT–PCR analysis and MCM6-positive cell indices were analyzed by the Student's or Welch's t-test following a test for equal variance. For grading immunohistochemical findings, scores of Tfrc and Nr0b2 were analyzed with the Mann–Whitney's U-test between the DEN-alone group and DEN + FB or DEN + PB groups. For the microarray data, statistical analysis was performed with GeneSpring® software, and the significance of gene expression changes was analyzed by the Student's t-test or analysis of variance between the DEN-alone and DEN + FB or DEN + PB groups.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Supplementary material
 Funding
 References
 
Microarray analysis
In all, 33 genes (15 genes up-regulated and 18 genes down-regulated) were identified as showing altered expression at week 6 of tumor promotion by both FB and PB (Table I). Examples showing dose-related expression changes numbered 18 and 32 in the DEN + FB and DEN + PB groups, respectively (supplementary Tables S2 and S3 are available at Carcinogenesis and Online). Among these, there were several whose products are involved in suppression of cell proliferation, such as Okl38 (17), protein arginine N-methyltransferase 5 (Prmt5_predicted) (18) and Dusp1 (19). In both DEN + FB and DEN + PB groups, up-regulation of Okl38 and down-regulation of Dusp1 and Mafb, were observed dose relatedly. Particularly, Mafb showed down-regulation from the lowest dose of both chemicals. The results of real-time RT–PCR for validation of microarray data are summarized in Table II. Expression levels at the highest dose of promotion in each chemical were compared between the microarray and real-time RT–PCR values. In both FB- and PB-promoted liver, many expression changes were similar between the two analysis systems, except for changes of Stx6. In addition, a low magnitude of alteration was found for Adh1 by real-time RT–PCR in the PB-promoted liver.


View this table:
[in this window]
[in a new window]

  		Table I. List of genes showing up- or down-regulation common to FB and PB promotion in the liver after 6 weeks exposure using a two-stage hepatocarcinogenesis model (≥2-fold, ≤0.5-fold)

 


View this table:
[in this window]
[in a new window]

  		Table II. Validation of microarray data by real-time RT–PCR

 
Since up-regulation of MCM6, which is involved in DNA replication and reported as a marker of proliferating cells (20), was found in the DEN + FB 3600 p.p.m. group, we selected MCM6 for immunohistochemical analysis to detect cell proliferation activity. Among genes showing altered expression specific to the early stages of tumor promotion by both FB and PB, up-regulation of Tfrc and down-regulation of Nr0b2 were observed with dose relation in the decrease of Nr0b2 by FB (Table I; supplementary Table S2 is available at Carcinogenesis Online). Tfrc is a receptor for transferrin, which plays a major role in intracellular uptake of iron (21). It has been reported that expression of Tfrc and iron homeostasis are altered in preneoplastic nodules and hepatocellular tumors in rats (22,23). Nr0b2 is an atypical orphan nuclear receptor that lacks a conventional DNA-binding domain, which acts as a co-regulator of various nuclear receptors and negatively regulates the gene expression of glucose 6-phosphatase (G6Pase), CYP7A1 and PEPCK (24,25). In the preneoplastic foci in rat liver, it is well known that loss of G6Pase is commonly observed (26). Thus, we also selected Tfrc and Nr0b2 for immunolocalization analysis in hepatocellular preneoplastic foci and tumors.

Immunolocalization at the early stage of tumor promotion
At week 6 of tumor promotion, the foci induced by FB or PB were predominantly of eosinophilic type and positive for GST-P (supplementary Table S4 is available at Carcinogenesis and Online). The numbers and areas of GST-P-positive foci were significantly increased in the DEN + FB 1200 and 3600 p.p.m. (8.04 ± 3.60 No./cm2 and 0.68 ± 0.44 mm2/cm2 at 1200 p.p.m.; 22.09 ± 11.62 No./cm2 and 5.93 ± 0.44 mm2/cm2 at 3600 p.p.m.) and DEN + PB 167 and 500 p.p.m. (6.90 ± 3.16 No./cm2 and 0.50 ± 0.24 mm2/cm2 at 167 p.p.m.; 6.20 ± 2.66 No./cm2 and 0.46 ± 0.26 mm2/cm2 at 500 p.p.m.) groups compared with the DEN-alone group (3.85 ± 2.04 No./cm2 and 0.22 ± 0.13 mm2/cm2) (supplementary Table S5 is available at Carcinogenesis Online). In the untreated controls, FB-alone and PB-alone groups, very few GST-P-positive foci were observed, and there were no intergroup differences in numbers and areas.

Tfrc showed diffuse liver cell immunoreactivity in the untreated controls and DEN-alone groups, and increased intensity was observed in liver cells on promotion with FB or PB, in parallel with the microarray results (Figure 1A; supplementary Table S4 is available at Carcinogenesis Online). In addition, a small population of GST-P-positive foci exhibited strong immunoreactivity for Tfrc.


Figure 1
View larger version (142K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Immunohistochemical distributions of Tfrc, Nr0b2 and MCM6 in liver cells after promotion by FB or PB for 6 weeks. (A) Tfrc in the DEN-alone, DEN + FB 3600 p.p.m. and DEN + PB 500 p.p.m. groups. Note fine granular Tfrc immunoreactivity in the cytoplasm of liver cells with diffuse cellular distribution in the DEN-alone group. Increased intensity of Tfrc immunoreactivity is evident after promotion with FB and PB. A liver cell focus positive for GST-P exhibits strong immunoreactivity. Bar = 100 µm. The graph shows scores (mean + standard deviation) for Tfrc staining intensity. **P < 0.01 versus DEN-alone group (Mann–Whitney's U-test). (B) Nr0b2 in the untreated control, DEN + FB 3600 p.p.m. and DEN + PB 500 p.p.m. groups. Nr0b2 was found diffusely in the cytoplasm of liver cells in the untreated control group, and decreased numbers of Nr0b2-positive cells were observed in the DEN + FB 3600 p.p.m. and DEN + PB 500 p.p.m. groups. Bar = 100 µm. **P < 0.01 versus DEN-alone group (Mann–Whitney's U-test). (C) MCM6 in the DEN-alone, DEN + FB 3600 p.p.m. and DEN + PB 500 p.p.m. groups. MCM6-positive cells with nuclear immunoreactivity were sparsely detected in the untreated control, DEN-alone and DEN + PB 500 p.p.m. groups. In the DEN + FB 3600 p.p.m. group, MCM6-positive cells were detected mainly in regenerative nodules including oval cells. Bar = 100 µm. **P < 0.01 versus DEN-alone group (Welch's t-test).

 
Nr0b2-positive liver cells were found diffusely in the untreated control group. Decreased numbers of Nr0b2-positive cells were observed in the DEN + FB 3600 p.p.m. and DEN + PB 500 p.p.m. groups, in parallel with the microarray data (Figure 1B; supplementary Table S4 is available at Carcinogenesis and Online). However, some liver cell foci positive for GST-P paradoxically showed increased numbers of Nr0b2-positive cells as compared with surrounding liver cells.

MCM6-positive cells were detected mainly in regenerative nodules including oval cells in the DEN + FB 3600 p.p.m. group, and their numbers were significantly increased in this group as compared with the DEN-alone group (Figure 1C).

TGFβRI and TGFβRII were immunolocalized in normal liver cells diffusely but weakly in all groups, including untreated controls (supplementary Table S4 is available at Carcinogenesis Online). Many GST-P-positive foci showed intense immunoreactivity for TGFβRI, especially in the DEN + FB 3600 p.p.m. group (Figure 2A). In contrast, very small populations of foci showed increased or decreased immunoreactivity for TGFβRII. PTEN and pPTEN immunoreactivity was diffusely but weakly observed in normal liver cells in all groups, including untreated controls. It was generally lacking in foci (Figure 2A; supplementary Table S4 is available at Carcinogenesis Online). On co-localization analysis of each molecule in association with GST-P positivity using serial sections, increased immunoreactivity of TGFβRI and decreased immunoreactivity of PTEN and pPTEN were found in most GST-P-positive foci compared with surrounding liver cells in the DEN + FB 3600 p.p.m. group. Some foci also showed co-expression of Tfrc and Nr0b2 (Figure 2B). In the DEN + PB group, although similar immunolocalization changes to FB group were observed for these molecules, rates of foci showing altered immunoreactivity were lower than those in the DEN + FB group.


Figure 2
View larger version (85K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. (A) Immunohistochemical localization of TGFβRI, PTEN, Tfrc and Nr0b2 in GST-P-positive liver cell foci after promotion by FB for 6 weeks. Increased expression of TGFβRI and decreased expression of PTEN in comparison with surrounding liver cells was found in most of GST-P-positive foci in the DEN + FB 3600 p.p.m. group, and some foci also showed co-expression of Tfrc and Nr0b2. Bar = 500 µm. (B) Immunohistochemical expression patterns of TGFβRI, TGFβRII, PTEN, pPTEN, Tfrc and Nr0b2 in GST-P-positive liver cell foci after promotion by FB or PB for 6 weeks. Expression patterns were classified into increased, unchanged and decreased in comparison with the staining intensity in the surrounding liver cells.

 
Immunolocalization at the late stage of tumor promotion
Many adenomas and carcinomas were obtained after promotion by FB or PB for 57 weeks. Most adenomas exhibited eosinophilic cytoplasm and solid growth, and carcinomas frequently showed trabecular and solid growth patterns. Foci observed at the late stage were mostly eosinophilic, but basophilic and clear cell types were also found. There were no major differences between the incidence and histological types of proliferative lesions induced by FB or PB.

All adenomas and carcinomas were positive for GST-P, although a small subset of foci, such as the basophilic type, appeared negative (Figure 3; supplementary Table S4 is available at Carcinogenesis Online). The immunoreactivity of TGFβRI was increased in the proliferative lesions in comparison with surrounding liver cells, and the percentages of lesions strongly expressing TGFβRI were increased in adenomas and carcinomas, in both FB and PB groups. Localized expression of TGFβRII in proliferative lesions was largely lacking, although a small proportion of lesions showed increased or decreased immunoreactivity. Nr0b2 was variably positive in the proliferative lesions, without relation to development from foci to carcinomas. PTEN and pPTEN were weakly positive in the liver cells outside the proliferative lesions, but were consistently lacking in FB- and PB-induced foci, adenomas and carcinomas. Immunoreactivity of Tfrc and MCM6 increased in parallel with the stage of lesion development.


Figure 3
View larger version (95K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. (A) Immunohistochemical localization of TGFβRI, PTEN, Nr0b2, Tfrc and MCM6 in hepatocellular adenomas and carcinomas generated after promotion by FB or PB for 57 weeks. (a) PB-induced adenoma and hematoxylin and eosin staining. (b) PB-induced adenoma showing intense immunoreactivity for TGFβRI. (c) FB-induced adenoma lacking PTEN expression. (d) FB-induced adenoma showing marked decrease of Nr0b2-positive cells. (e) FB-induced carcinoma and hematoxylin and eosin staining. (f) FB-induced carcinoma showing strong immunoreactivity for Tfrc. (g) FB-induced carcinoma with a high MCM6-positive cell index. Bar = 100 µm. (B) Immunohistochemical expression patterns of GST-P, TGFβRI, TGFβRII, Nr0b2, PTEN, pPTEN and Tfrc and MCM6-positive cell indices of proliferative lesions generated after promotion by FB or PB for 57 weeks. Expression patterns were classified into increased, unchanged and decreased in comparison with the surrounding liver cells. MCM6-positive indices are shown as mean + standard deviation. *P < 0.05 and **P < 0.01 versus Fo (Student's t-test). Numbers of lesions examined: No = 16, Fo = 49, Ad = 36, Ca = 52 (FB group); No = 26, Fo = 104, Ad = 30, Ca = 31 (PB group). No, normal; Fo, focus; Ad, adenoma; Ca, carcinoma.

 
MCM6-positive cell indices did not differ in relation with the immunoreactive patterns of TGFβRI or TGFβRII in each type of proliferative lesion in either DEN + FB or DEN + PB group, although carcinomas showed particular variability (Figure 4). Foci and adenomas showing different expression patterns of Nr0b2 did not show obvious variation in the MCM6-positive cell index. On the other hand, carcinomas showing decreased immunoreactivity of Nr0b2 exhibited higher MCM6-positive cell indices as compared with those showing unchanged expression of this molecule as compared with surrounding liver cells in the DEN + PB group. Although statistically non-significant, a similar tendency was also observed in the DEN + FB group. In both DEN + FB and DEN + PB groups, MCM6-positive cell indices in adenomas or carcinomas lacking PTEN and pPTEN immunoreactivity were lower than in the corresponding lesions showing unchanged expression of these antigens as compared with surrounding liver cells (DEN + PB group: significantly different for pPTEN in adenomas; DEN + FB group: significantly different for PTEN and pPTEN in carcinomas). There was an increasing tendency for MCM6-positive cell indices in adenomas and carcinomas that showed increased immunoreactivity of Tfrc as compared with those in the corresponding lesions showing unchanged expression in both DEN + FB and DEN + PB groups, but it was statistically non-significant.


Figure 4
View larger version (33K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. MCM6-positive cell indices sorted by expression patterns for each molecule in proliferative lesions induced after promotion by FB and PB for 57 weeks. The values are mean + standard deviation. *P < 0.05 and **P < 0.01 versus corresponding lesions showing unchanged expression in comparison with surrounding liver cells (Student's t-test). Numbers of lesions examined: Fo = 49, Ad = 36, Ca = 52 (FB group); Fo = 104, Ad = 30, Ca = 31 (PB group). No, normal; Fo, focus; Ad, adenoma; Ca, carcinoma.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Supplementary material
 Funding
 References
 
Both FB and PB are known to induce CYPs, and it has been suggested that effects such as inhibition of cell-to-cell communication and oxidative stress are involved in their mechanisms of carcinogenic action (11,15,16). In the present study, we subtracted gene clusters simply responding to FB and PB alone as well as those responding to initiation and PH from those showing altered expression after initiation plus PH and following promoter treatment, and therefore, the obtained genes can be considered to play some role in the promotion process by non-genotoxic hepatocarcinogens. As a result, chemically inducible genes related to drug detoxification, inhibition of cell-to-cell communication and oxidative stress responses were excluded from the expression profile. Several genes, such as Okl38, Prmt5 and Dusp1, are known to be related to suppression of cell proliferation (17–19), suggesting that expression changes of genes toward growth suppression observed in the early stages of tumor promotion was probably a reflection of gene expression in mitotically inactive uninitiated cells that comprise the predominant cell population in the liver at this time point.

Immunohistochemical examination revealed that a subset of GST-P-positive foci exhibit increased immunolocalization of Tfrc, Nr0b2 and TGFβRI and decreased immunoexpression of PTEN and pPTEN after promotion by FB or PB. In contrast to the promotion with FB, only a small proportion of foci showed altered expression by PB at the same early stage. This difference between FB and PB might reflect differences in the action or potential for promotion of these chemicals. However, since similar immunoexpression patterns were observed in the FB- and PB-induced tumors at the late stage, it is presumed that the role of the molecules examined here may be similar in the process of hepatocarcinogenesis. Simply, slower growth of GST-P-positive foci in the PB-promoted livers as compared with the FB-promoted ones might have lowered the chance of acquiring or losing phenotypes within the foci. It should be noted that the size of foci with PB was rather small as compared with the FB case.

Increased expression of Tfrc was observed here in a subset of GST-P-positive foci in the early stage of tumor promotion by FB or PB, and levels increased in proportion to the stage of lesion development. Iron plays an important role in many essential cell functions including the synthesis of enzymes necessary for cellular growth and metabolism, and expression of Tfrc is closely linked to the proliferation status of the cell (27). It has been reported that expression of Tfrc and iron homeostasis is altered in preneoplastic nodules and hepatocellular tumors in rats (22,23). A correlation between Tfrc expression and cell proliferation has been also shown in human breast carcinomas (28). Therefore, it is suggested that change in iron homeostasis was linked to the processes of tumor promotion and progression by FB and PB in the present study.

Nr0b2 negatively regulates transcription of G6Pase (25), and loss of G6Pase is commonly observed in preneoplastic foci in the rat liver (26). Our present finding that some liver cell foci showed positive immunoreactivity of Nr0b2 as compared with surrounding liver cells in the early stage of tumor promotion might have caused decreased expression of G6Pase. In the late stage, Nr0b2 was variably expressed in the proliferative lesions without relation to lesion development. Interestingly, carcinomas showing decreased expression of Nr0b2 exhibited higher MCM6-positive cell indices in both PB- and FB-promoted livers, suggesting that the loss of Nr0b2 may be necessary for acquisition of a malignant phenotype, different from the suggested role for formation of G6Pase (–) foci at the early stage of tumor promotion.

TGFβ is a potent growth inhibitor in liver cells, and thus, it has been considered that down-regulation of TGFβ signaling is critical for hepatocarcinogenesis (5). There have been many studies examining expression levels of its receptors in human hepatocellular tumors, and overall, most have revealed reduction (5). In rats, decrease of TGFβRs was reported in PB-promoted hepatocellular tumors, and it is considered that down-regulation of TGFβRs might provide a selective growth advantage to tumor cells by allowing them to escape the growth inhibitory effects of TGFβ (3,29). Despite growth inhibitory effects of TGFβ signaling on liver cells, immunoreactivity of TGFβRI and TGFβRII observed in the present study was increased or unchanged, respectively, during tumor development. Although TGFβ can also act as a promoter of tumor progression during the late stage of tumorigenesis and induce tumor invasion and metastasis (30), biological effects of increased TGFβRI expression in foci from the early stage of tumor promotion cannot be unequivocally concluded. Since MCM6-positive cell indices did not differ with the levels of TGFβRI and TGFβRII, the TGFβ signaling pathway is possibly impaired in these proliferative lesions. It has been proposed that the ratio between TGFβRI and TGFβRII influences the cellular fate by controlling signaling (31). Therefore, differences in expression patterns between TGFβRI and TGFβRII in GST-P-positive lesions might lead to dysregulation of TGFβ signaling. Dysregulation of downstream effectors of TGFβ, such as smad2 and smad4, has been reported to contribute to the tumor progression during chemical carcinogenesis in rats (32). In addition, increased expression of smad7, an inhibitory downstream molecule of the TGFβ signaling, has been found in many of advanced human hepatocellular carcinomas, suggestive of its role for acquisition of resistance to TGFβ, especially in hepatocellular tumors without reduction of TGFβRs (33). Thus, further investigations of factors downstream of TGFβ signaling are needed to clarify the reason for the unexpected cellular distribution of TGFβRI and TGFβRII in relation to hepatocarcinogenesis.

In human hepatocellular carcinomas, it has been reported that expression of PTEN is reduced or absent in almost half of cases and reduced PTEN expression may be involved in pathogenesis and tumor progression (8,34). In rats, there have been only few reports of PTEN expression in the liver tumors. Silins et al. (10) described a subset of preneoplastic foci exhibiting a PTEN-positive phenotype and suggested there to be targeted for sphingolipid-induced apoptosis. In the present study, the levels of both PTEN and pPTEN were decreased consistently in the foci, adenomas and carcinomas, suggesting that reduction of PTEN function may be involved in rat hepatocarcinogenesis from an early stage. Involvement of the PI3 kinase pathway, including PTEN and AKT, has recently been shown in TGFβ-induced invasion during the tumor progression process (35,36), implying a possible contribution to late-stage tumor promotion. Expression of the PTEN gene can be suppressed by TGFβ (37). Since increased expression of TGFβRI and decreased expression of PTEN and pPTEN were often found concomitantly in the present study, further investigation of this relationship appears warranted. Although the relation to loss of tumor suppressor function is unclear, neoplastic lesions lacking PTEN/pPTEN expression in the present study exhibited lower MCM6-positive cell indices than those retaining these antigens. Some functions related to tumor progression other than proliferation might be activated in tumors lacking PTEN/pPTEN.

In conclusion, we here found 33 genes showing altered expression specific to the early stages of tumor promotion by both FB and PB in the rat two-stage hepatocarcinogenesis model using microarray technique. Immunohistochemical analysis indicated that changes of iron homeostasis following increased expression of Tfrc, specifically in proliferative lesions, contribute to tumor promotion and progression by FB or PB. Loss of Nr0b2 might possibly play a role in acquisition of the malignant phenotype in hepatocellular tumors. In addition, loss of PTEN and dysregulation of TGFβ signaling may be considered to be involved in rat hepatocarcinogenesis from early stages. Thus, our approaches applied here utilizing microarray analysis and following immunohistochemical screening may help searching early biomarkers of carcinogenesis. Although further studies should address roles in the processes of hepatocarcinogenesis, obtained molecules in the present study may be beneficial for detection and evaluation of non-genotoxic carcinogens that we are at risk, as well as for understanding of the mechanism of non-genotoxic carcinogenesis to secure human health.


   Supplementary material
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
Funding
 References
 
Supplementary Tables S1–S5 can be found at http://carcin.oxfordjournals.org/


   Funding
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
References
 
Health and Labour Sciences Research Grants (Research on Food Safety) from the Ministry of Health, Labour and Welfare of Japan.


   Acknowledgments
 
We thank Miss Tomomi Morikawa and Ayako Kaneko for their technical assistance in conducting the animal study.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 

  1. Tatematsu M, et al. Relative merits of immunohistochemical demonstrations of placental, A, B and C forms of glutathione S-transferase and histochemical demonstration of gamma-glutamyl transferase as markers of altered foci during liver carcinogenesis in rats. Carcinogenesis (1985) 6:1621–1626.[Abstract/Free Full Text]
  2. Kitano M, et al. Possible tumor development from double positive foci for TGF-{alpha} and GST-P observed in early stages on rat hepatocarcinogenesis. Cancer Sci. (2006) 97:478–483.[CrossRef][Medline]
  3. Reisenbichler H, et al. Transforming growth factor-beta receptors type I, II and III in phenobarbital-promoted rat liver tumors. Carcinogenesis (1994) 15:2763–2767.[Abstract/Free Full Text]
  4. Ueno T, et al. Relation of type II transforming growth factor-beta receptor to hepatic fibrosis and hepatocellular carcinoma. Int. J. Oncol. (2001) 18:49–55.[Web of Science][Medline]
  5. Breuhahn K, et al. Dysregulation of growth factor signaling in human hepatocellular carcinoma. Oncogene (2006) 25:3787–3800.[CrossRef][Web of Science][Medline]
  6. Lim IK, et al. Regulation of selection of liver nodules initiated with N-nitrosodiethylamine and promoted with nodularin injections in Fischer 344 male rats by reciprocal expression of transforming growth factor-β1 and its receptors. Mol. Carcinog. (1999) 26:83–92.[CrossRef][Web of Science][Medline]
  7. Park DY, et al. Expression of transforming growth factor (TGF)-beta1 and TGF-beta type II receptor in preneoplastic lesions during chemical hepatocarcinogenesis of rats. Toxicol. Pathol. (2001) 29:541–549.[CrossRef][Web of Science][Medline]
  8. Fabregat I, et al. Survival and apoptosis: a dysregulated balance in liver cancer. Liver Int. (2007) 27:155–162.[CrossRef][Web of Science][Medline]
  9. Wang L, et al. Epigenetic and genetic alterations of PTEN in hepatocellular carcinoma. Hepatol. Res. (2007) 37:389–396.[CrossRef][Web of Science][Medline]
  10. Silins I, et al. Dietary sphingolipids suppress a subset of preneoplastic rat liver lesions exhibiting high PTEN, low phospho-Akt and high levels of ceramide species. Food Chem. Toxicol. (2006) 44:1552–1561.[CrossRef][Web of Science][Medline]
  11. Shoda T, et al. Liver tumor promoting effects of fenbendazole in rats. Toxicol. Pathol (1999) 27:553–562.[Abstract/Free Full Text]
  12. Tatematsu M, et al. Reciprocal relationship between development of glutathione S-transferase positive liver foci and proliferation of surrounding hepatocytes in rats. Carcinogenesis (1988) 9:221–225.[Abstract/Free Full Text]
  13. Shirai T. A medium-term rat liver bioassay as a rapid in vivo test for carcinogenic potential, a historical review of model development and summary of results from 291 tests. Toxicol. Pathol. (1997) 25:453–460.[Abstract/Free Full Text]
  14. Ito N, et al. Early detection of carcinogenic substances and modifiers in rats. Mutat. Res. (2000) 462:209–217.[CrossRef][Web of Science][Medline]
  15. Kinoshita A, et al. Phenobarbital at low dose exerts hormesis in rat hepatocarcinogenesis by reducing oxidative DNA damage, altering cell proliferation, apoptosis and gene expression. Carcinogenesis (2003) 24:1389–1399.[Abstract/Free Full Text]
  16. Whysner J, et al. Phenobarbital mechanistic data and risk assessment: enzyme induction, enhanced cell proliferation, and tumor promotion. Pharmacol. Ther. (1996) 71:153–191.[CrossRef][Web of Science][Medline]
  17. Ong CK, et al. Genomic structure of human OKL38 gene and its differential expression in kidney carcinogenesis. J. Biol. Chem. (2004) 279:743–754.[Abstract/Free Full Text]
  18. Gilbreth M, et al. Negative regulation of mitosis in fission yeast by the shk1 interacting protein skb1 and its human homolog, Skb1Hs. Proc. Natl Acad. Sci. USA. (1998) 95:14781–14786.[Abstract/Free Full Text]
  19. Brondello JM, et al. Reduced MAP kinase phosphatase-1 degradation after p42/p44MAPK-dependent phosphorylation. Science (1999) 286:2514–2517.[Abstract/Free Full Text]
  20. Schrader C, et al. Minichromosome maintenance protein 6, a proliferation marker superior to Ki-67 and independent predictor of survival in patients with mantle cell lymphoma. Br. J. Cancer (2005) 93:939–945.[CrossRef][Web of Science][Medline]
  21. Ponka P, et al. The transferrin receptor: role in health and disease. Int. J. Biochem. Cell Biol. (1999) 31:1111–1137.[CrossRef][Web of Science][Medline]
  22. Pascale RM, et al. Transferrin and transferrin receptor gene expression and iron uptake in hepatocellular carcinoma in the rat. Hepatology (1998) 27:452–461.[CrossRef][Web of Science][Medline]
  23. Holmström P, et al. Expression of iron regulatory genes in a rat model of hepatocellular carcinoma. Liver Int. (2006) 26:976–985.[CrossRef][Web of Science][Medline]
  24. Seol W, et al. An orphan nuclear hormone receptor that lacks a DNA binding domain and heterodimerizes with other receptors. Science (1996) 272:1336–1339.[Abstract]
  25. Kim JY, et al. Orphan nuclear receptor small heterodimer partner represses hepatocyte nuclear factor 3/Foxa transactivation via inhibition of its DNA binding. Mol. Endocrinol. (2004) 18:2880–2894.[Abstract/Free Full Text]
  26. Sawaki M, et al. Phenotype of preneoplastic and neoplastic liver lesions during spontaneous liver carcinogenesis of LEC rats. Carcinogenesis (1990) 11:1857–1861.[Abstract/Free Full Text]
  27. Bomford AB, et al. Transferrin and its receptor: their roles in cell function. Hepatology (1985) 5:870–875.[CrossRef][Web of Science][Medline]
  28. Wrba F, et al. Ki-67 immunoreactivity in breast carcinomas in relation to transferrin receptor expression, estrogen receptor status and morphological criteria. An immunohistochemical study. Oncology (1989) 46:255–259.[Web of Science][Medline]
  29. Mansbach JM, et al. Phenobarbital selectively promotes initiated cells with reduced TGF beta receptor levels. Carcinogenesis (1996) 17:171–174.[Abstract/Free Full Text]
  30. Akhurst RJ, et al. TGF-beta signaling in cancer—a double-edged sword. Trends Cell Biol. (2001) 11:S44–S51.[Web of Science][Medline]
  31. Huang SS, et al. TGF-beta control of cell proliferation. J. Cell Biochem. (2005) 96:447–462.[CrossRef][Web of Science][Medline]
  32. Park DY, et al. Expression and localization of the transforming growth factor-beta type I receptor and Smads in preneoplastic lesions during chemical hepatocarcinogenesis in rats. J. Korean Med. Sci. (2003) 18:510–519.[Web of Science][Medline]
  33. Park YN, et al. Expression of Smad7 in hepatocellular carcinoma and dysplastic nodules: resistance mechanism to transforming growth factor-beta. Hepatogastroenterology (2004) 51:396–400.[Medline]
  34. Hu TH, et al. Expression and prognostic role of tumor suppressor gene PTEN/MMAC1/TEP1 in hepatocellular carcinoma. Cancer (2003) 97:1929–1940.[CrossRef][Web of Science][Medline]
  35. Shigeoka Y, et al. Sulindac sulfide and caffeic acid phenethyl ester suppress the motility of lung adenocarcinoma cells promoted by transforming growth factor-beta through Akt inhibition. J. Cancer Res. Clin. Oncol. (2004) 130:146–152.[CrossRef][Web of Science][Medline]
  36. Hjelmeland AB, et al. Loss of phosphatase and tensin homologue increases transforming growth factor beta-mediated invasion with enhanced SMAD3 transcriptional activity. Cancer Res. (2005) 65:11276–11281.[Abstract/Free Full Text]
  37. Wu H, et al. PTEN signaling pathways in melanoma. Oncogene (2003) 22:3113–3122.[CrossRef][Web of Science][Medline]
Received February 23, 2008; revised May 17, 2008; accepted May 25, 2008.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
29/11/2218    most recent
bgn135v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Takahashi, M.
Right arrow Articles by Nishikawa, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Takahashi, M.
Right arrow Articles by Nishikawa, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?