| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Human Cancer Biology |
Authors' Affiliations: Departments of 1 Microbiology and 2 Internal Medicine, 3 Center of Genomic Medicine, and 4 Graduate Institute of Clinical Medicine, National Taiwan University College of Medicine; 5 Division of Molecular and Genomic Medicine, National Health Research Institutes; 6 Hepatitis Research Center, National Taiwan University Hospital, Taipei, Taiwan
Requests for reprints: Pei-Jer Chen, Graduate Institute of Clinical Medicine, National Taiwan University College of Medicine, 7 Chung-Shan South Road, Taipei 100, Taiwan. Phone: 886-2-23123456, ext. 7072; Fax: 886-2-23317624; E-mail: peijer{at}ha.mc.ntu.edu.tw.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Experimental Design: The expression level of FAP-1 RNA in hepatocellular carcinomas was evaluated by RNase protection and quantitative PCR. Sodium bisulfite modification and subsequent single-strand conformational polymorphism and sequence analyses were used to assay the methylation of CpGs at FAP-1 promoter. Direct sequencing of the FAP-1 coding region was conducted for detecting the genetic mutations. Two common single nucleotide polymorphisms of FAP-1 were selected for evaluating their association with the hepatocellular carcinoma trait in sporadic and familial hepatocellular carcinomas. Moreover, the functional effect of FAP-1 on cellular proliferation has been evaluated by small interfering RNA approach.
Results: Around 50% of hepatocellular carcinomas showed significantly decreased expression of FAP-1 compared with the corresponding nontumorous liver tissues. In most cases, the RNA level was well correlated with the methylation status of promoter CpGs, suggesting that the promoter methylation may contribute to the down-regulation. Several genetic mutations of FAP-1 have been identified in hepatocellular carcinomas. The G/G genotype of FAP-1 cSNP6304 was significantly associated with the increased risk of multiplex familial hepatocellular carcinomas (odds ratio, 2.44; 95% confidence interval, 1.19-5.01). Finally, knockdown expression of FAP-1 was shown to enhance the cellular proliferation in PLC5 cells.
Conclusions: FAP-1 could be inactivated during hepatocarcinogenesis, mainly attributed by allelic loss and promoter methylation. The genetic mutations and polymorphisms may also confront with the higher hepatocellular carcinoma risk. These results first suggested FAP-1 as a putative TSG in hepatocarcinogenesis.
To facilitate positional cloning of the putative TSG(s) on chromosome 4q, extensive deletion mapping has been conducted aiming to define the TSG-containing region(s). Currently, three regions showing common allelic loss in hepatocellular carcinomas have been deduced, including 4q21-23, 4q25-28, and 4q32-35 (4, 5, 7–9). To focus on the region of 4q21-23, our previous detailed deletion mapping has pointed out one restricted TSG-containing region encompassing 4q22 (10). Supportively, this region overlapped with the TSG-containing regions defined by two other groups (8, 9).
In our attempt to find out the putative TSG in this region, based on two reasons, we have focused on the candidate gene of Fas-associated phosphatase-1 (FAP-1; also known as PTPN13/PTP1E/PTP-BAS/PTPL1) for further extensive genetic characterizations. First, FAP-1 functions as a protein tyrosine phosphatase (PTP), which in general counteracts with the oncogenic activity of protein tyrosine kinases in several signaling pathways (11–13). In fact, several PTPs were already proven to function as TSGs, such as PTEN (14), PTPN11 (15), and PTP
(16). In this aspect, FAP-1 was reported to function as a negative regulator of Src family kinases in the ephrinB signaling, which is activated by binding of ephrinB with its cognate EphB receptor (17). Secondly, by a comprehensive mutational analysis of 87 members of the PTP gene superfamily, Wang et al. have identified frequent somatic mutations of FAP-1 in colorectal cancers, with the mutation rate of 
10% (18). Therefore, both of the functional and genetic characteristics of FAP-1 suggested its candidacy as a plausible target TSG at this chromosome region.
FAP-1 is a large nontransmembrane PTP. The extreme NH2 terminus contains a kinase noncatalytic C-lobe (KIND) domain, which is followed by a FERM domain (band four point one, ezrin, radixin, moesin homology domain), five different PDZ domains (postsynaptic density protein-95, discs large, zonula occludens), and a COOH-terminally located tyrosine phosphatase domain (19–23). FAP-1 serves as a central scaffolding protein facilitating the assembly of a multiplicity of different proteins via the KIND domain, the PDZ domains, or the FERM domains. Alternative splicing further increases the complexity of FAP-1 in its interacting with the cellular proteins (24, 25). Interestingly, the interacting proteins of FAP-1 showed a variety of functions in regulating several tumorigenesis-related biological processes, including apoptosis, transcription, development, cytoskeleton organization, ephrinB signaling, and cytokinesis (17, 23, 24, 26–31). Aiming to examine the potential of FAP-1 as the target TSG in 4q21-22, the current study has focused on analyzing several genetic characteristics of FAP-1 in hepatocellular carcinomas, including the expression level, the promoter methylation pattern, the genetic mutations, and the association of its specific coding single nucleotide polymorphisms (cSNP) with the hepatocellular carcinoma trait. The results showed that FAP-1 was inactivated in 50% of hepatocellular carcinomas during hepatocarcinogenesis, mainly attributed by allelic loss and promoter methylation. Besides, specific genetic mutations and polymorphisms may also contribute to the risk of hepatocellular carcinoma. The current study thus provided genetic evidences suggesting FAP-1 as a putative TSG at chromosome 4q21-23 associated with hepatocarcinogenesis.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
In addition, we have included 295 control subjects and 534 hepatocellular carcinoma cases for our association analysis. The peripheral blood of these individuals was collected for the subsequent DNA extraction. The control subjects were healthy people without showing any clinical hepatic abnormalities, especially all are HBsAg negative and anti–hepatitis C virus negative. The hepatocellular carcinoma cases can be subdivided into three groups: the sporadic hepatocellular carcinomas, the HBV-related hepatocellular carcinomas showing simplex family history (having no hepatocellular carcinoma cases among first-degree relatives of the index cases), and the HBV-related hepatocellular carcinomas showing multiplex family history (having at least one hepatocellular carcinoma cases among first-degree relatives of the index cases). A detailed description of the enrollment and follow-up of the control subjects and hepatocellular carcinoma cases has been described previously (32, 33).
Besides, we have included 25 cell lines in this study, including 11 hepatocellular carcinoma cell lines (Huh7, HepG2, Hep3B, SK-Hep1, Huh6, T59, HCC36, HA22T, Mahlavu, Tong, and PLC5), 8 gastric cancer cell lines (AGS, N87, SNU16, RF1, SNU1, Kato III, RF48, and SNU5), 3 colon cancer cell lines (Lovo, Colo-205, and DLD1), 1 pancreatic cancer cell line (Paca-2), 1 kidney cancer cell line (293), and 1 transformed liver fibroblast cell line (Chang Liver). All the cells were maintained with DMEM supplemented with 10% fetal bovine serum, 2 mmol/L L-glutamine, 1 mmol/L nonessential amino acid, and penicillin/streptomycin and passaged with standard procedures described previously (34). The DNA or RNA was extracted from the peripheral blood, frozen tissues, and the cell lines by following the protocols described previously (10, 32, 34).
RNase protection analysis. The RNase protection analysis was conducted by using the RiboQuant Multiprobe RNase Protection Assay System (BD Biosciences, PharMingen, San Diego, CA). The multiprobe template set we used is hAPO-3, which contains DNA templates used for the T7 RNA polymerase-directed synthesis of high-specific activity, 
-32P-labeled antisense RNA probes. The probes can hybridize with target human mRNAs encoding FAP-1 and several apoptosis-related genes, including caspase-8, FASL, FAS, FADD, DR3, FAF, TRAIL, RNFRp55, TRADD, RIP, and two housekeeping gene products, L32 and glyceraldehyde-3-phosphate dehydrogenase. Total RNA (15 µg) from each sample was used for the RNase protection analysis by following the manufacturer's instruction. In brief, the probe set is hybridized in excess to target RNA in solution. Free probe and other ssRNA molecules were digested with RNases. The remaining RNase-protected probes were purified, resolved on denaturing polyacrylamiide gels, and imaged by autoradiography.
Quantitative reverse transcription-PCR. We used the SuperScript cDNA System (Invitrogen, Carlsbad, CA) to reverse transcribe the RNA template into cDNA, which was used for subsequent PCR amplification. Quantitative PCR was done with the LightCycler FastStart DNA SYBR Green kit (Roche Diagnostics Applied Science, Mannheim, Germany). The forward and reverse primers designed for FAP-1 quantification are FAP-1quantF (5'-CCTACAGTGTGGGGTC-3') and FAP-1quantR (5'-GGTGAACCATCGCAGT-3'), respectively. cDNA (1 µL) was amplified in a 10 µL reaction containing 0.5 µmol/L forward and reverse primers, 2 mmol/L MgCl2, and 1 µL of 10x FastStart SYBR Master Mix. The housekeeping gene porphobilinogen deaminase (PBGD) was served as the internal control for the quantification analysis with forward and reverse primers (5'-AGAGTGATTCGCGTGGGTACC-3' and 5'-CCCTGTGGTGGACATAGCAAT-3'), respectively. The amplification condition was the same as FAP-1, except for the addition of 3 mmol/L MgCl2. The PCR conducted with LightCycler thermal cycler (Roche Diagnostics Applied Science) consisted of 10 minutes at 95°C and 45 cycles at 95°C for 0 second, 59°C for 5 seconds, 72°C for 18 seconds, and 79°C for 0 seconds to detect fluorescence. To verify the specificity of the amplification, a melting curve analysis was done at the end of amplification: 0-second denaturation at 95°C, 15-second annealing at 60°C, and continuous decrease to 40°C with 0.1°C/s beginning at 95°C. Cp values were determined by the Second Derivative Maximum methodology. The detailed procedures for the quantification were as described previously (10). The significant down-regulation was arbitrarily defined as the level less than 30% of the corresponding nontumorous tissues for each sample.
Sodium bisulfite modification and PCR amplification of FAP-1 promoter CpGs. We treated the samples with sodium bisulfite, which converts nucleotides C to T in the absence of methylation. However, once methylation occurs at the C residues, they will resist the treatment. Therefore, such treatment can be used to differentiate the methylation and unmethylation status of CpGs. In brief, 3 µg genomic DNA was digested with 15 units EcoRI restriction enzyme (New England Biolabs, Beverly, MA) for 2 hours and then purified by PCR High Pure kit (Roche Diagnostics Applied Science). The purified product was incubated with 0.2 N NaOH in a final volume of 50 µL for 10 minutes at 37°C. Sodium bisulfite NaHSO3 (520 µL; 3.5 mol/L; freshly prepared; Sigma, St. Louis, MO) and hydroquinone (30 µL; 10 mmol/L; freshly prepared; Sigma) were added to the samples and incubated at 52°C overnight. The sample was subsequently desalted by DNA Clean-Up Wizard (Promega, Madison, WI) and desulfonated by incubation with 0.3 N NaOH for 10 minutes at 37°C. The modified genomic DNA was then ethanol-precipitated and resuspended in 20 µL H2O for subsequent PCR amplification.
Bisulfite modified DNA (1 µL) was amplified with nested primer sets flanking the CpGs 5'-AGAGGTTAAGGTATTTGAAGTATTT-3' and 5'-AAAAAACTAAACCCAAAACCACTAC-3' for first PCR and 5'-TGGTTAAGTTATGAGGTTTTGAG-3' and 5'-ATTAAAATTCTCRCTRCTCTRTCAAC-3' for nested PCR (including 1 µL of 50-fold diluted first PCR product) with the Platinum Taq (Invitrogen, Carlsbad, CA). The thermal cycler conditions consisted of 5 minutes at 95°C and 10 cycles at 95°C for 45 seconds, 64°C for 45 seconds, 72°C for 30 seconds and 25 cycles at 95°C for 45 seconds, 62°C for 45 seconds, 72°C for 30 seconds, and final 72°C for 10 minutes. The PCR product was subsequently assayed by the single-strand conformational polymorphism (SSCP) and direct sequencing analysis for evaluating the methylation pattern.
SSCP analysis. For SSCP analysis, 4 µL PCR product was mixed with 4 µL freshly prepared denaturing buffer containing 100 mmol/L NaOH and 2 mmol/L EDTA, incubated at 50°C for 10 minutes, and then put on ice immediately. The DNA was analyzed with gel electrophoresis by using GeneGel Excel 12.5/24 kit and GenePhor Electrophoresis Unit (Amersham Pharmacia Biotech, Uppsala, Sweden). After electrophoresis, the gel was stained by using PlusOne DNA Silver Staining kit and the DNA bands can be visualized (Amersham Pharmacia Biotech).
Sequencing analysis. FAP-1 contains 48 exons with the coding region starts from exon 2 and ends at exon 48 (35). We designed 47 amplicons for these exons with PCR amplification procedures described previously (10). The primer sequences and PCR annealing temperatures were summarized in Table 1. The PCR products were processed to the direct sequencing analysis, which were done by ABI PRISM Big-Dye kits (Applied Biosystems, Foster City, CA) and analyzed via ABI 3100 Genetics Analyzer. In addition, the PCR products used for the SSCP analysis were also processed to the direct sequencing analysis.
|
The constructs were transformed into Saccharomyces cerevisiae Y187 yeast strain for assaying the interaction affinity. Constructs of pCL1 (encoding full-length GAL4 protein) and p53 (amino acids 1-147; K01700)-GAL4 DNA-binding domain/SV40 large T antigen (amino acids 768-881; SV4CG)-GAL4 activating domain served as positive controls for our analysis. Two assays were used for assaying the relative interaction affinity. One is assayed by PLCX-plate [Petri dishes with a thin layer of 100 mmol/L Na2PO4 (pH 7.0), 10 mmol/L KCl, 1 mmol/L MgSO4, 1 mmol/L DTT, 14 g/L agar, 0.3 to 1.0 mg/mL 5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (36)], which can evaluate the interaction affinity by directly comparing the blue color development. The other is assayed by Yeast ß-Galactosidase Assay kit (Pierce Chemical Co., Rockford, IL), which can quantitatively reflect the interaction affinity by absorbance determination (with wavelength 420 nm). In brief, the yeast cells growing for 3 days on selection plate were collected for quantification of the ß-galactosidase activity by following the manufacturer's instructions.
Small interfering RNA. The mammalian expression vector, pSUPER.retro.puro (OligoENgine, Seattle, WA) was used for expression of FAP-1 small interfering RNA in PLC-5 cells. The gene-specific insert specifies a 19-nucleotide sequence corresponding to nucleotides 653 to 671 (GCCAGGCTATTCGAGATCG) of FAP-1 (Genbank, D21209), which is separated by a 9-nucleotide noncomplementary spacer (TTCAAGAGA) from the reverse complement of the same 19-nucleotide sequence. This construct was called pSUPERsi-FAP-1. The sequence was inserted into the pSUPER.retro.puro following the manufacturer's instructions and was confirmed by sequencing analysis.
PLC-5 cells were transfected with pSUPERsi-FAP-1 and the vector control plasmids using LipofectAMINE 2000 reagent (Invitrogen, Rockville, MD) according to the manufacturer's instructions. Twenty-four hours after transfection, cells were processed for puromycin selection (2 µg/mL; Sigma); 48 hours after selection, the RNA and protein were extracted for evaluating the FAP-1 expression.
Cell proliferation assay. The cell proliferation assay was conducted by the CellTiter 96 Aqueous One Solution Cell Proliferation Assay kit (Promega), which is a colorimetric method for determining the number of viable cells. The PLC5 cells transfected with pSUPERsi-FAP-1 or the control vector were plated to the wells of a 96-well plate in culture medium (1,500 per well) without puromycin selection. The assay was conducted each day from days 0 to 4 by following the procedures of the manufacturer's instruction. CellTiter 96 Aqueous One Solution Reagent (20 µL/well) was added to each well. After 1 hour at 37°C in a humidified, 5% CO2 atmosphere, the absorbance at 490 nm was recorded using an ELISA plate reader. Each point represents the mean ± SD of four replicates and the background absorbance shown at zero cells per well was subtracted from these data.
Northern blot analysis. Total RNA was extracted with TRIzol (Invitrogen, Carlsbad, CA), and 15 µg RNA was separated by electrophoresis on a 1% formaldehyde agarose gel and transferred to a nylon membrane. The blot was probed with a DIG-labeled DNA fragment corresponding to nucleotides 1,263 to 1,857 of the FAP-1 sequence (D21209), which was generated with a PCR DIG Probe Synthesis kit (Roche Diagnostics Applied Science).
Protein extraction and Western blot analysis. Cells were lysed with the lysis buffer [20 mmol/L HEPES (pH 7.9), 0.2 mol/L NaCl, 1 mmol/L EDTA, 1 mmol/L EGTA, 0.2% Triton X-100] containing 1 mmol/L phenylmethylsulfonyl fluoride and protease inhibitor cocktail (Roche Diagnostics Applied Science) and the protein concentration was determined by BCA protein assay kit (Pierce). For Western blot analysis, 100 µg protein extract was separated by SDS-PAGE and electrophoretically transferred to nitrocellulose membrane, incubated with specific antibodies with proper dilutions at 4°C overnight (FAP-1 antibody, H300, Santa Cruz Biotechnology, Santa Cruz, CA; ß-actin antibody, Sigma-Aldrich, St. Louis, MO), incubated with horseradish peroxidase–conjugated secondary antibody for 1 hour, and then developed with the enhanced chemiluminescence kit (Pierce).
Statistical analyses. Allele frequencies of two cSNP of FAP-1, cSNP4629 and cSNP6304, were generated and tested for fit to the expectations of Hardy-Weinberg equilibrium using the 
2 goodness-of-fit statistic. Logistic regression analysis using STATA statistical software (version 8.0, Stata Corp., College Station, TX) was conducted to estimate odds ratios (OR) associated with each category of cSNP6304 and cSNP4629. Estimates of the variables in the model were computed by maximum likelihood techniques, and 95% confidence intervals (95% CI) were based on the SE of the coefficients and the normal approximation.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
30% of the amount relative to their corresponding nontumorous tissues (in 45.8% of cases). The representative results of 6 cases were shown in Fig. 1B. The reduction was even found in 6 cases without showing loss of heterozygosity (LOH) at this region (as exemplified by patients 1 and 4 in Fig. 1B). In summary, FAP-1 RNA down-regulation was identified in 50% of hepatocellular carcinoma cell lines and also the primary hepatocellular carcinomas. Down-regulation of FAP-1 RNA is well correlated with the promoter CpG methylation. DNA sequence of the promoter region driving expression of the major FAP-1 RNA transcript has been identified previously (37). Interestingly, we have found a CpG-rich sequence at this promoter region just upstream of the transcription initiation site, containing 12 CpG dinucleotides (Fig. 2A). To investigate the underlying mechanism attributed to the FAP-1 down-regulation, we have tried to examine whether the methylation of promoter CpGs may influence the FAP-1 RNA expression in both hepatocellular carcinoma cell lines and primary hepatocellular carcinomas.
|
In our further attempt to analyze the FAP-1 CpG methylation patterns in primary hepatocellular carcinomas and their corresponding nontumorous tissues, we divided the samples into two groups: one group contained 12 hepatocellular carcinomas without 4q LOH and the other group contained 12 hepatocellular carcinomas with 4q LOH. The direct sequencing results were again consistent with the SSCP results and we summarized the sequencing results in Fig. 2E (for hepatocellular carcinomas without 4q LOH) and Fig. 2F (for hepatocellular carcinomas with 4q LOH). Ten of the 24 (42%) hepatocellular carcinomas showed significant methylation pattern in six or more CpGs at this promoter region. Among them, 8 cases belonged to the hepatocellular carcinomas without showing 4q allelic loss. Intriguingly, the promoter methylation started from the adjacent nontumorous tissues and progressively increased in the corresponding hepatocellular carcinomas in 3 of the 8 cases (Fig. 2E). In contrast, the promoter methylation in six or more CpGs was only detected in 2 of 12 hepatocellular carcinomas with 4q LOH (Fig. 2F). Therefore, the CpG methylation at FAP-1 promoter not only occurred in hepatocellular carcinoma cell lines but also in the primary hepatocellular carcinomas, especially in those without 4q allelic loss. The results thus indicated that either allelic loss or the promoter methylation might contribute to one of the two hits knocking down the FAP-1 RNA expression during hepatocarcinogenesis.
Genetic variations of FAP-1 occur in hepatocellular carcinomas. Another important genetic characteristic of TSG is the occurrence of genetic mutations affecting the normal gene function. Notably, the FAP-1 somatic mutations have been identified previously in 10% of colorectal cancers (18). To investigate if the frequent FAP-1 genetic mutations also occur in hepatocellular carcinomas, we have included 24 cell lines and 70 primary hepatocellular carcinomas for our direct sequencing analysis, which covered the entire FAP-1 coding region (exons 2-48). The genetic mutations identified in these DNA samples were summarized in Table 2, stratified with the exon numbers containing the genetic mutations. Moreover, we also included the SNPs of FAP-1 (available from public SNP database: dbSNP BUILD 121) and the somatic mutations identified in colorectal cancers for comparison (Table 2; ref. 18). Missense mutations were identified in 5 tumor cell lines (4 of them are hepatocellular carcinoma cell lines) and 5 primary hepatocellular carcinomas. However, the ones in the 5 primary hepatocellular carcinoma were also present in the nontumorous tissues adjacent to the corresponding hepatocellular carcinomas. Because no peripheral mononuclear cells were available for these 5 cases, we cannot conclude whether they belonged to somatic mutations, germ-line mutations, or the rare genetic polymorphisms not currently reported in the public SNP database.
|
40%. Therefore, some genetic mutations of FAP-1 identified in primary hepatocellular carcinomas, although not being proven as somatic mutations, may affect the protein function.
|
|
Functional effect of FAP-1 on cell proliferation. We have further tried to analyze the effect of FAP-1 down-regulation on cell proliferation, one tumorigenic-related characteristic, approaching by the small interfering RNA experiments. PLC5 cell line was chosen for our study because of its highest transfection efficiency among the hepatoma cell lines with normal FAP-1 expression level (and also lacking of promoter methylation). We first confirmed the effective down-regulation of FAP-1 at both RNA and protein levels (Fig. 4A and B) and then evaluate the resulting effect on the cellular proliferation activity (Fig. 4C). The results showed that down-regulation of FAP-1 could considerably enhance the proliferation of PCL5 cells, which supports the functional role of FAP-1 as a negative regulator for cell proliferation (Fig. 4C).
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
50% of hepatocellular carcinomas in both hepatocellular carcinoma cell lines and primary hepatocellular carcinomas. In addition to the frequent allelic loss at the FAP-1-containing chromosome region, the CpGs at its promoter region was heavily methylated in most RNA down-regulated hepatocellular carcinomas, suggesting that promoter methylation may also contribute to the reduction of FAP-1 during hepatocarcinogenesis. Moreover, several missense genetic variations in FAP-1 coding region were identified in hepatocellular carcinomas and some of them were proven to affect the protein function. Finally, the G/G genotype of FAP-1 cSNP6304 was identified to be significantly associated with increased risk of multiplex familial hepatocellular carcinomas and acted in a recessive manner. Because significant decrease of FAP-1 RNA was also observed in hepatocellular carcinomas without showing allelic loss at FAP-1-containing chromosome region, alternative mechanisms were responsible for inactivating FAP-1 in these tumors. Epigenetic modifications involving DNA methylation and chromatin alteration may provide such alternative mechanisms because their involvement in transcriptionally silencing of several TSGs was well documented in many cancers (39). In fact, aberrant promoter methylation of several TSGs was frequently identified in hepatocellular carcinomas, such as SOCS-1, GSTP, APC, E-cadherin, and p15 (40). Our results showed that methylation of CpGs at the promoter region of FAP-1 was well correlated with its down-regulation. Actually, we have preliminarily tested whether demethylation could restore FAP-1 RNA expression by treating Huh7, HepG2, and Hep3B hepatocellular carcinoma cell lines with 5-aza-2'-deoxycytidine, a well-established inhibitor of DNA methyltransferase (41). By reverse transcription-PCR analysis, 5-aza-2'-deoxycytidine treatment was shown to reverse the down-regulation of FAP-1 in all three cell lines; moreover, the promoter methylation can be partially reversed (data not shown). The results supported that the CpG methylation at FAP-1 promoter might contribute to its transcriptional silencing. In addition, we have noticed that this FAP-1 promoter region is also involved in regulation of another apoptosis-related gene, JNK3, which is arranged in a head-to-head genomic organization with FAP-1 (35). Effect of the CpG methylation at this bidirectional promoter region on the expression of JNK3 warrants further investigation.
The nontumorous tissues in the current study were collected from the pathologically defined nontumorous tissues adjacent to the tumorous tissues, and most of them already showed phenotype of active hepatitis or cirrhosis. We have tried to compare the level of FAP-1 mRNA in the nontumorous tissues of our collection with normal liver tissues collected either from the liver of a 40-year-old male Caucasian (with death cause of trauma) or from the normal liver tissues adjacent to the uninfected hepatic adenoma and focal nodular hyperplasia (negative for HBsAg and anti–hepatitis C virus and free of ongoing HBV or hepatitis C virus infections). All these five normal liver tissues showed very low level of FAP-1 RNA, which can be barely detected by Northern blot analysis as reported previously (data not shown; refs. 19, 20). Determined by the sensitive quantitative PCR, the RNA level of these five samples is similar with the variation of 30% of their mean level. In contrast, the adjacent nontumorous tissues of our cases showed heterogeneous FAP-1 RNA expression and many of them showed levels with variations of >50% of the mean level of the five normal tissues. The results indicated that in hepatitis or cirrhosis tissues FAP-1 might show certain heterogeneous expression pattern. Supportively, approached by immunocytochemistry staining, Lee et al. found that FAP-1 can be detected in 80% of hepatocellular carcinoma with great variation of the protein expression level (42). Therefore, to specifically evaluate the change of FAP-1 expression during carcinogenesis, the comparison needs to be conducted in paired hepatocellular carcinoma with the corresponding nontumorous tissues for each individual as we approached in the current study. The down-regulation of FAP-1 in hepatocellular carcinomas was further supported by the microarray database established by Su et al., who constructed a molecular classification scheme for a variety of carcinomas and showed that FAP-1 was down-regulated in most primary hepatocellular carcinomas in their collection (43). In line to support this, we also identified FAP-1 to be down-regulated in 14 hepatocellular carcinomas in our analyzing of 25 pairs of hepatocellular carcinomas by using Affymetrix oligochip microarray hybridization (data not shown).
Our results indicated that allelic loss and promoter methylation might be the major mechanisms attributed to FAP-1 down-regulation during hepatocarcinogenesis. The somatic mutations, although not occurring frequently in hepatocellular carcinomas as that shown in colorectal cancers (18), still showed their functional effect on the protein product at least for the two located at PDZ2 domain identified in primary hepatocellular carcinoma. During hepatocarcinogenesis, aberrant methylation begins at the preneoplastic stage and the allelic loss seems to escalate sharply in dysplastic hepatocytes (44, 45). The possibility that hypermethylation may precede locus deletion has been exemplified by loci located on chromosome 16q and by the paracentromeric chromosomal translocations, both occur frequently in hepatocellular carcinomas (44, 45). It was thus proposed that epigenetic changes in gene expression that continue in hepatocellular carcinomas seem to create conditions that increase the chance of generating hepatocyte populations containing critical combinations of structurally and functionally aberrant genes. Although the potential causes of structural aberrations in genes in hepatocellular carcinomas are varied, in some instances, epigenetic changes directly precede structural alterations in the same genes. For example, reduced expression of p16INK4A is associated predominantly with promoter hypermethylation, but allelic loss and mutation also are responsible for some loss of expression of this gene (44, 45), which is similar with the case of FAP-1 reported in the currently study.
Intriguingly, the association analysis pointed out cSNP6304 to be significantly associated with the trait of hepatocellular carcinoma in cases showing multiplex family history. Hepatocellular carcinoma belongs to a complex disease, and it is well accepted that the occurrence of hepatocellular carcinoma is through a multistep carcinogenic process. In our previous study, HBV carriers with a family history of hepatocellular carcinoma have a higher risk for hepatocellular carcinoma occurrence compared with HBV carriers without the hepatocellular carcinoma family history (OR, 2.41; 95% CI, 1.47-3.95). For carriers with two or more affected relatives, the ratio increased to 5.55 (95% CI, 2.02-15.26; ref. 32). The results suggested that there existed certain hepatocellular carcinoma susceptibility genetic factor(s) predisposing the individuals toward hepatocellular carcinoma, especially in the multiplex hepatocellular carcinoma families. In the current study, we pointed out the G/G genotype of FAP-1 cSNP6304 to be significantly associated with the increased risk of hepatocellular carcinoma in multiplex families. It raised the possibility that FAP-1 belongs to one of the hepatocellular carcinoma susceptibility genetic factors in such multiplex families. Presumably, inclusion of additional hepatocellular carcinoma multiplex individuals can further consolidate the association result. However, due to the difficulty of recruiting such families even in Taiwan, we spent 7 years to collect the current 48 multiplex families. Although the current sample size is enough for revealing the significant association, we will continue recruiting the multiplex families in the future to increase the sample size and further validate this association.
The cSNP6304 locates between the PDZ5 and phosphatase domains of FAP-1, a region with no annotated function. We have tried to explore its genetic linkage and its possible protein interaction. At first, we analyzed the LD block structure where cSNP6304 locates in Han Chinese population and found that it locates in a LD block spanning 70-kb genomic region (starting from the SNP of rs10516778 and ends with the SNP of rs989902). Within this LD block, only two cSNPs changing the amino acid codons have been identified in Han Chinese population, cSNP4629 and cSNP6304, which already are included in our association analysis. Although only cSNP6304 showed significant association with the multiplex hepatocellular carcinomas, a moderately elevated point estimate OR was also noted for cSNP4629 in the multiplex hepatocellular carcinomas (OR, 1.77; 95% CI, 0.68-4.63; Table 3). The wide 95% CI indicated not enough power due to small numbers of G/G genotype in this sample. The results for the two cSNPs within this LD were consistent. The cSNP4629 locates on PDZ3 domain with known interacting protein PRK2 (protein kinase C–related kinase). By yeast two-hybrid analysis, we compared the PDZ3 domain with either G or A alleles for their relative affinity with PRK2 but did not find significant difference (data not shown). This may suggest that cSNP4629 exerts cellular functions other than this protein interaction or the cSNP6304 influences the PZD domains or other functions of FAP-1 in a not yet identified way.
Currently, the biological function of FAP-1 has not been fully characterized. Because it consists of several domains attributed to protein-protein interaction, its biological functions were thus proposed mainly mediated through its interacting proteins. The interacting proteins of FAP-1 showed functions in regulating a variety of biological events, including apoptosis, cytoskeleton organization, cytokinesis, ephrinB signaling, and transcription (17, 23, 24, 26–31). Among them, the role of FAP-1 has been studied more thoroughly in the process of apoptosis and the ephrinB signaling. Several studies pointed out the potential role of FAP-1 in inhibiting the Fas-mediated apoptosis, mainly by regulating the cell surface expression of Fas through the interaction of FAP-1 with Fas (23, 46, 47). An up-regulation of FAP-1 was found in ovarian tumors and showed a correlation between the resistance of Fas-mediated cell death and the FAP-1 expression level (48). In such a situation, FAP-1 seemed to play an oncogenic role by enhancing the cancer cell survival. However, there also existed evidence implicating that FAP-1 can exert a proapoptotic effect in 4-hydroxytamoxifen-treated human breast cancer cells, which was through an inhibition of the insulin receptor substrate-1/PtdIns 3-kinase pathway (49). Secondly, at the process of ephrinB signaling, FAP-1 plays negative regulatory role by dephosphorylaiton of ephrinB and plausibly the Src family kinases. The in vitro assay further showed that FAP-1 can specifically dephosphorylate the Tyr418 of Src and resulted in the inactivation of Src activity (17). During hepatocarcinogenesis, a variety of stimulus elicited by virus infection was reported to affect the apoptosis process, especially at the stage of chronic necroinflammatory hepatitis, which usually lasted for several decades (50). Moreover, Src activation has been shown in a significant proportion of hepatocellular carcinoma and was suggested to play critical role in the early stage of hepatocarcinogenesis (51, 52). Although we have preliminarily provided evidence that down-regulation of FAP-1 expression might enhance the cellular proliferation in PLC5 cell line, the role of FAP-1 in regulating apoptosis and antagonizing the Src activation at such early carcinogenic stage and the functional effect of FAP-1 inactivation during hepatocarcinogenesis still warrants further extensive investigations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 6/24/05; revised 11/ 1/05; accepted 12/ 2/05.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
-fetoprotein production. Gastroenterology 1996;110:184–92.[CrossRef][Medline]-fetoprotein. Hepatology 2004;40:847–54.[CrossRef][Medline] (PTP) in human breast cancer cell line MCF-7. Oncogene 2004;23:1256–62.[CrossRef][Medline]
B activation. FEBS Lett 1999;460:191–8.[CrossRef][Medline]B through interaction between the PDZ1 domain and ankyrin repeats. Biochem J 1999;337:179–84.[Medline]This article has been cited by other articles:
|
|
 |
|
  Y. Zhang, Y. Tu, J. Zhao, K. Chen, and C. Wu Reversion-induced LIM interaction with Src reveals a novel Src inactivation cycle J. Cell Biol., March 23, 2009; 184(6): 785 - 792. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Dromard, G. Bompard, M. Glondu-Lassis, C. Puech, D. Chalbos, and G. Freiss The Putative Tumor Suppressor Gene PTPN13/PTPL1 Induces Apoptosis through Insulin Receptor Substrate-1 Dephosphorylation Cancer Res., July 15, 2007; 67(14): 6806 - 6813. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, no methylation; 
, methylation/unmethylation ratio 
, methylation/unmethylation ratio < 1; 
, methylation/unmethylation ratio > 1; 
, fully methylated. D, summary of the methylation patterns of promoter CpGs in the cell lines. E, summary of the methylation patterns of promoter CpGs in 12 hepatocellular carcinoma pairs without 4q LOH. Three cases showing progressively increasing methylation patterns starting from the nontumorous part to the hepatocellular carcinomas. F, summary of the methylation patterns of CpGs in 12 hepatocellular carcinoma pairs with 4q LOH.
