Identification of a secretory protein c19orf10 activated in hepatocellular carcinoma

The identification of genes involved in tumor growth is crucial for the development of inventive anticancer treatments. Here, we have cloned a 17‐kDa secretory protein encoded by c19orf10 from hepatocellular carcinoma (HCC) serial analysis of gene expression libraries. Gene expression analysis indicated that c19orf10 was overexpressed in approximately two‐thirds of HCC tissues compared to the adjacent noncancerous liver tissues, and its expression was significantly positively correlated with that of alpha‐fetoprotein (AFP). Overexpression of c19orf10 enhanced cell proliferation of AFP‐negative HLE cells, whereas knockdown of c19orf10 inhibited cell proliferation of AFP‐positive Hep3B and HuH7 cells along with G1 cell cycle arrest. Supplementation of recombinant c19orf10 protein in culture media enhanced cell proliferation in HLE cells, and this effect was abolished by the addition of antibodies developed against c19orf10. Intriguingly, c19orf10 could regulate cell proliferation through the activation of Akt/mitogen‐activated protein kinase pathways. Taken together, these data suggest that c19orf10 might be one of the growth factors and potential molecular targets activated in HCC.

Hepatocellular carcinoma (HCC) is one of the most common cancers with an estimated worldwide incidence of 1,000,000 cases per year. 1 Most HCCs develop as a consequence of chronic liver disease such as chronic viral hepatitis due to hepatitis C virus (HCV) or hepatitis B virus (HBV) infection. [2][3][4][5][6][7] Liver cirrhosis patients with any etiology are considered to be at an extremely high risk for HCC. [8][9][10] Indeed, $7% of liver cirrhosis patients with HCV infection develop HCC annually, 8,11 and the advancement of reliable HCC screening methods for high-risk patients is crucial for the improvement of their overall survival. 12 Currently, imaging diagnostic techniques such as ultrasonography, computed tomography, magnetic resonance image and angiography are the gold standards for the early detection of HCC. 13,14 In addition, tumor markers such as alphafetoprotein (AFP) and des-gamma carboxyl prothrombin (DCP) have been used for the screening of HCC, [15][16][17][18] although their sensitivity and specificity are not sufficiently high. Recently, a gene expression profiling approach shed new light on Glypican 3, a heparin sulfate proteoglycan anch-ored to the plasma membrane, as a potential HCC marker, and its clinical usefulness as a molecular target as well as a tumor marker is presently under investigation. 19 There are several options available for the treatment of HCC, including surgical resection, liver transplantation, radiofrequency ablation, transcatheter arterial chemoembolization and chemotherapy, while taking the HCC stage and liver function into consideration. Recently, molecular therapy targeting the Raf kinase/vascular endothelial growth factor receptor (VEGFR) kinase inhibitor sorafenib improved the survival of patients with advanced HCC, 20,21 emphasizing the importance of deciphering the molecular pathogenesis of HCC for the development of effective treatment options.
Here, we investigated the gene expression profiles of HCC by serial analysis of gene expression (SAGE) to discover a novel gene activated in HCC. [22][23][24][25] We identified a gene, c19orf10, overexpressed in HCC and determined that the encoded 17-kDa protein (c19orf10) is a secretory protein.
Murine c19orf10 was originally discovered to encode a cytokine interleukin (IL)-25/stroma-derived growth factor (SF20) in 2001. 26 The gene c19orf10 was mapped in the H2 complex region of mouse chromosome 17 between C3 and Ir5, and the hypothetical protein was predicted as globular protein. 26 However, the subsequent study failed to reproduce its proliferative effect on lymphoid cells, and the paper was retracted by the authors in 2003. 26,27 Nevertheless, independent studies revealed that c19orf10 was indeed produced by synoviocytes, macrophages and adipocytes, although the function of c19orf10 remained elusive. 28,29 In our study, we identified that c19orf10 was overexpressed in AFP-positive HCC samples. Our data imply that c19orf10 could activate the mitogen-activated protein kinase (MAPK)/Akt pathway and enhance cell proliferation in HCC cell lines, suggesting that c19orf10 may be a growth factor produced by tumor epithelial cells and/or stromal cells, and, therefore, would be a good target for the treatment of HCC.

SAGE and HCC samples
HCC and normal liver SAGE libraries that we had constructed were reanalyzed using SAGE 2000 software. The size of each SAGE library was normalized to 300,000 transcripts per library. Monte Carlo simulation was used to select genes whose expression levels were significantly different between the two libraries. Each SAGE tag was annotated using the gene-mapping website SAGE Genie database (http://cgap.nci.nih.gov/ SAGE/) and the SOURCE database (http://smd.stanford.edu/ cgi-bin/source/sourceSearch) as previously described. 30 An additional 15 SAGE libraries of normal and cancerous tissues from various organs were retrieved using the National Center for Biotechnology Information SAGEmap (http:// www.ncbi.nlm.nih.gov/SAGE/).
Fifteen HCC tissues (four HBV-related and 11 HCVrelated) and the corresponding noncancerous liver tissues were obtained from HCC patients who received hepatectomy. Four normal liver tissues were obtained from patients undergoing surgical resection of the liver for the treatment of metastatic colon cancer. Additionally, 36 HCC tissues (17 HBV-related and 19 HCV-related) were obtained from HCC patients undergoing hepatectomy. These samples were snap frozen in liquid nitrogen immediately after resection and used for quantitative real-time detection PCR (RTD-PCR). Total RNA was extracted using a ToTALLY RNA TM kit (Ambion, Austin, TX).
The study protocol conformed to the ethical guidelines of the Declaration of Helsinki (1975) and was approved by the institutional ethical review board committee. All patients provided written informed consent for the analysis of the specimens.

Laser capture microdissection and RNA isolation
Laser capture microdissection (LCM) was performed as previously described. 31 Briefly, 20 HCV-related surgically resected HCC tissues were frozen in OCT compound (Sakura Finetech, Torrance, CA). 32 Inflammatory cells and cancerous cells in HCC tissues were separately excised by LCM using a Laser Scissors CRI-337 (Cell Robotics, Albuquerque, NM) under a microscope. Total RNA was isolated from these cells using a microRNA isolation kit (Stratagene, La Jolla, CA) in accordance with the supplied protocol, with slight modifications. 31 Construction of C19ORF10 expression plasmid and recombinant adenovirus vector PCR was performed on a Marathon cDNA library from Huh7 cells using the following primers: sense primers: 5 0 -GACCCTAGTCCAACATGGCGGCGCCC-3 0 (the first PCR), 5 0 -ATGGCGGCGCCCAGCGGAGGGTGGAACGGC-3 0 (the nested second PCR) and antisense primers: 5 0 -CACCGGA GATGAGAAGGTGCCACCCGC-3 0 (the first PCR), 5 0 -CAG GGCTGCTGGTCACAGCTCAGTGCGCG-3 0 (the nested second PCR). The 5 0 and 3 0 ends of the cDNA were isolated using a SMART RACE cDNA Amplification kit (Clontech, Mountain View, CA) according to the manufacturer's recommendations. The PCR products were cloned into a TA vector (Invitrogen, Carlsbad, CA) to generate the pcDNA3.1-c19orf10 expression plasmid. Using this plasmid, a C-terminally FLAG-tagged construct of c19orf10 was generated and inserted in a pSI mammalian expression vector (Promega, Madison, WI), which was driven by the SV40 promoter (pSI-c19orf10).
The replication-incompetent recombinant adenovirus vector expressing FLAG-tagged c19orf10 (Ad. c19orf10-FLAG) was generated by homologous recombination using the AdMax system (Microbix, Toronto, Canada) as previously described. 33 The generated recombinant adenovirus was purified by limiting dilution, and the titer of viral aliquots was determined by the 50% tissue culture infectious dose method as previously described. 34

RTD-PCR
RTD-PCR was performed as previously described. 31 Briefly, template cDNA was synthesized from 1 lg of total RNA using SuperScript TM II RT (Invitrogen). RTD-PCR of c19orf10 (Hs. 00384077_m1), AFP (Hs00173490_m1), GPC3 (Hs01018938_m1), KRT19 (Hs00761767_s1) and the ACTB internal control (Hs99999903_m1) was performed using a TaqManV R Gene Expression Assay kit (Applied Biosystems, Foster City, CA). The expression of selected genes was measured in triplicate by DDCT method using the 7900 Sequence Detection System (Applied Biosystems).

Cell lines and transfection of plasmids
Human liver cancer cell lines HuH1, Huh7, Hep3B, HLE and HLF as well as HEK293 and NIH3T3 were cultured in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen) in 5% CO 2 at 37 C. Transfection of plasmids was performed using FuGENE TM 6 (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's instruction. Briefly, 5 Â 10 5 cells were seeded in a six-well plate 12 hr before transfection, and 3 lg of plasmid DNA was used for each transfection. All experiments were repeated at least twice.
Purification of c19orf10-FLAG fused protein and production of anti-c19orf10 antibody subjected to elution by competition with FLAG peptide (5 lg/ml), and each 1 ml fraction of the eluted aliquot was collected to obtain the most concentrated c19orf10-FLAG protein in accordance with the manufacturer's protocol. The anti-c19orf10 antibodies were developed by immunizing rabbits with repeated intradermal injections of purified c19orf10-FLAG. Protein concentration was measured by the Bradford method.

Silencing gene expression by short interfering RNA
The selected short interfering RNA (siRNA) targeting C19ORF10 (Si-C19ORF10; Silencer Select siRNAs s31855) and the irrelevant control sequence (Si-Control; Silencer Select siRNAs 4390843) was obtained from Applied Biosystems. Transfection of these siRNAs was performed using FuGENE TM 6 (Roche Diagnostics) as previously described. 30 Briefly, 2 Â 10 5 cells were seeded in a six-well plate 12 hr before transfection. A total of 100 pmol/l of siRNA duplex was used for each transfection. The experiments were performed at least twice.

Cell proliferation assay
Cell proliferation was evaluated in quadruplicate using a Cell Titer 96 MTS Assay kit (Promega). Briefly, 2 Â 10 3 HLE or HuH7 cells were harvested in a 96-well plate 12 hr before the transfection or addition of the recombinant proteins. Transfection of siRNAs or plasmids was performed using FuGE-NE TM 6 (Roche Diagnostics). After incubation with MTS/ PMS solution at 37 C for 2 hr, the absorbance at 450 nm was measured. The experiments were performed at least twice.

Cell cycle analysis
Cells were fixed using 80% ice-cold ethanol and incubated with propidium iodide for 10 min. DNA content was analyzed using a FACS Caliber flow cytometer (BD Biosciences, San Jose, CA) counting 10,000 stained cells. The distribution of cells in each cell cycle phase was determined using FlowJo software (Tree Star, Ashland, OR).

Western blotting
Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer, and the extracts were subsequently electrophoresed on sodium dodecyl sulfate-10% polyacrylamide gels and transferred onto protean nitrocellulose membranes. The blots were then incubated for 1 hr with an appropriate primary monoclonal antibody: phospho-PI3K (#4228), phospho-Akt , phospho-cdc2 (#9111) and b-actin (#4970) (Cell Signaling Technology, Allschwil, Switzerland) and anti-FLAG antibodies (Sigma-Aldrich, St. Louis, MO). The blots were washed and exposed to peroxidase-conjugated secondary antibodies, such as anti-mouse or rabbit IgG antibodies, and visualized using the ECL TM kit (Amersham Biosciences, Piscataway, NJ). All experiments were performed at least twice.

Statistical analyses
Unpaired t-tests and Kruskal-Wallis tests were performed on the RTD-PCR and cell proliferation data using GraphPad Prism software (www.graphpad.com).

Identification of C19ORF10 overexpression in HCC by SAGE
To comprehensively explore the candidate novel genes activated in HCC, we reanalyzed two SAGE libraries derived from HCC tissues and normal liver tissues. 30 After normalization of each SAGE library size to 300,000 tags, we compared the HCC and normal liver libraries to obtain the list of genes overexpressed in HCC. We identified 79 genes significantly overexpressed in the HCC library by more than tenfold when compared to the normal liver library (Supporting Information Table 1). Among them, we explored expressed sequence tags (ESTs) as candidates for novel HCC-related genes to identify eight unique tags corresponding to seven ESTs (Table 1). We especially focused on the EST chromosome 19 open reading frame 10 (c19orf10) because the sequence presumably encoded a secretory protein with a signal peptide sequence (Fig. 1a). When we examined the expression profiles of c19orf10 using retrieved SAGE data from various cancers and their normal counterparts, we identified that c19orf10 was abundantly expressed in human HCC (Fig. 1b). We further examined the publicly available EST profiles of c19orf10 (http:// www.ncbi.nlm.nih.gov/unigene) and confirmed its tendency to be overexpressed in HCC compared to the normal liver (data not shown). We validated the overexpression of c19orf10 in 15 independent HCC tissues and adjacent noncancerous liver tissues by RTD-PCR. Gene expression of c19orf10 was significantly higher in the HCC tissues than in the normal liver tissues and adjacent noncancerous liver tissues (p ¼ 0.014 and 0.048, respectively; Fig. 1c). C19orf10 expression was elevated in HCC tissues compared to the adjacent noncancerous liver tissues in 10 of 15 patients (66.7%; Fig. 1d).
Overexpression of C19ORF10 in AFP-positive HCC As HCC is a heterogeneous mixture of cancer epithelial cells and stromal cells, and a previous report indicated that c19orf10 is expressed in fibroblast-like synoviocytes. We, therefore, evaluated the expression of c19orf10 in tumor epithelial cells and stromal cells separately using LCM and RTD-PCR in 20 HCC tissues (Fig. 2a). Although tumor

Cancer Cell Biology
Sunagozaka et al.
stromal cells expressed c19orf10 at some level, the expression levels were significantly higher in tumor epithelial cells than in stromal cells (p ¼ 0.006) (Fig. 2b).
Furthermore, when we examined the expression of c19orf10 in AFP-positive (HuH1, HuH7 and Hep3B) and -negative (HLE and HLF) HCC cell lines, we identified the overexpression of c19orf10 in AFP-positive HCC cell lines (Fig. 2f). These data suggested that c19orf10 is overexpressed and may play some role in AFP-positive HCCs.

C19orf10 regulates MAPK/Akt pathways and activates cell proliferation
To explore the functional role of c19orf10 in HCC, we performed c19orf10 overexpression and knockdown studies using c19orf10-low HLE cells and c19orf10-high Hep3B and HuH7 cells, respectively. When we transfected HLE cells with pcDNA3.1 or pcDNA3.1-c19orf10 plasmids, we identified an approximately sixfold overexpression of c19orf10 when compared to the control 48 hr after transfection (p < 0.0001) (Fig. 3a). Interestingly, cell proliferation was modestly, but significantly, enhanced compared to the control 72 hr after transfection (p ¼ 0.0015) (Fig. 3b).
We also transfected siRNAs targeting an irrelevant sequence (Si-Control) or c19orf10 (Si-c19orf10) in Hep3B and

Cancer Cell Biology
Sunagozaka et al.
HuH7 cells. We observed an $50% decrease in c19orf10 expression in Hep3B cells transfected with Si-c19orf10 compared to the control 48 hr after transfection with statistical significance (p < 0.0001). In this condition, cell proliferation was suppressed to 50% compared to the control 72 hr after transfection (p < 0.0001) (Figs. 3c and 3d). When we performed cell cycle analysis of HuH7 cells transfected with Si-Control or Si-c19orf10, we identified an increase of G1-phase cells and a decrease of S-and G2-phase cells by c19orf10 knockdown, suggesting that the G1 cycle arrest was caused by the knockdown of c19orf10 (Fig. 3e).
We examined the representative MAPK/Akt pathwayassociated proteins and cell cycle regulators using Western blotting 72 hr after siRNAs transfection (Fig. 3f). Interestingly, phosphorylation of c-Raf, MEK, MAPK, PI3K and pAkt was inhibited by knockdown of c19orf10, suggesting the involvement of c19orf10 in the MAPK/Akt pathways. Furthermore, phosphorylation of Rb, CDK4 and CDK6 was also inhibited by knockdown of c19orf10, consistent with the observation of G1 cell cycle arrest by C19ORF10 knockdown. PTEN, p53 and phosphorylated CDC2 protein expression was not affected by knockdown of c19orf10.

C19orf10 encodes the secretory protein and stimulates cell proliferation
As the sequence of c19orf10 suggested that it encodes a secretory protein, we transfected pSI-c19orf10-FLAG in NIH3T3 cells and examined the culture supernatant. Immunoprecipitation of the collected culture supernatant 48 hr after transfection using anti-FLAG antibodies indicated the existence of a 17-kDa protein (c19orf10), compatible with the molecular weight of the 142 amino acids protein encoded by c19orf10 (Fig. 4a). We purified c19orf10-FLAG protein from the supernatant of HEK293 cells infected with Ad. c19orf10-FLAG using an anti-FLAG column. Supplementation of purified c19orf10-FLAG into the culture media for 72 hr enhanced the proliferation of HLE cells in a dose-dependent manner with statistical significance, whereas control FLAG peptides and BSA had no effects on cell proliferation (Fig. 4b). Western blot analysis of HLE cells cultured with purified c19orf10-FLAG (40 ng/ml) or BSA control (40 ng/ml) indicated the immediate strong phospholyration of Akt peaked 5 min after supplementation (Fig. 4c). The modest phospholyration of GSK3b (Ser9) and p44/42 MAPK also followed and peaked 60 min after c19orf10 supplementation. These data suggest that Akt pathway might be directly involved in the c19orf10-mediated cell proliferation signaling with the subsequent activation of MAPK pathway. Furthermore, addition of antibodies against c19orf10 to the culture media abolished the cell proliferation induced by c19orf10, whereas control IgG had no effects (Fig. 4d). Taken together, these data suggest that c19orf10 may be a growth factor overexpressed in AFP-positive HCCs and activates the Akt/MAPK pathways, potentially through the activation of an unidentified c19orf10 receptor.

Discussion
SAGE facilitates the measurement of transcripts from normal and malignant tissues in a nonbiased and highly accurate, quantitative manner. Indeed, SAGE produces a comprehensive gene expression profile without a priori gene sequence information, leading to the identification of novel transcripts potentially involved in the pathogenesis of human cancer. 19 In our study, we identified seven SAGE tags potentially corresponding to novel genes activated in HCC. Among them, we identified the secretory protein c19orf10 activated in a subset of HCCs.
Several serum markers including AFP, DCP and Glypican 3 are currently used for the detection and/or the evaluation of the treatment for HCCs in the clinic. [15][16][17][18]35 These markers are known as oncofetal proteins, that is, expressed in the fetus, transcriptionally suppressed in the adult organ and reactivated in the tumor. We identified that the expression of c19orf10 positively correlated with AFP expression but did not correlate with the expression of GPC3 or the biliary marker KRT19. As c19orf10 was rarely detected in the normal liver, it is possible that c19orf10 is also an oncofetal protein activated in HCC. We are currently developing a system to detect serum c19orf10 in HCC patients, and the significance of the serum c19orf10 value as an HCC marker should be clarified.
Recent advancement in molecular biology has revealed the considerable diversity of transcription initiation and/or termination of genes altered in the process of carcinogenesis.
Indeed, using 5 0 SAGE approach, we recently discovered the novel intronic transcripts activated in HCC. 36 Interestingly, when we investigated the transcription initiation of c19orf10 using the 5 0 SAGE database, we identified a potential 5 0 splice variant initiated from the second exon of c19orf10 (data not shown). Although we have not yet validated the presence of 5 0 splice variants in c19orf10 by PCR, examination of 5 0 EST database also suggested the presence of the similar splice variants (GenBank Accession Number CR980295, BQ680744, BQ648461, etc.). Alteration of transcription initiation/termination in c19orf10 might affect the abundance or function of c19orf10 protein, and the details of 5 0 splice variants in c19orf10 should be clarified in future studies.
Molecular targeting therapy has rapidly emerged for solid tumors as well as for leukemia. [37][38][39] Sorafenib is a multikinase inhibitor targeting Raf kinase in the MAPK pathway as well as VEGFR and the platelet-derived growth factor receptor. 40,41 In our study, we identified that c19orf10 activates the MAPK and Akt/PI3K pathways and contributes to the proliferation of HCC cell lines, although we still could not discover the potential receptor of c19orf10. Development of a neutralizing c19orf10 antibody may provide novel therapeutic options for HCC patients to inhibit these signaling pathways, and its efficacy should be evaluated in the future.
Recently, c19orf10 was found to be expressed in fibroblast-like synoviocytes in the synovium using a proteomics approach. 29 In addition, a recent article indicated that c19orf10 was expressed in preadipocyte cells and involved in adipogenesis using two-dimensional electrophoresis mass spectrometry analysis. 28 Thus, c19orf10 may have pleiotropic effects on various lineages of normal organs in various developmental stages, and the clarification of its distribution and biological properties in the whole body may provide more detailed information about the function of c19orf10.
In conclusion, we have identified the protein c19orf10 that regulates the Akt/MAPK pathways and cell cycle through an unidentified mechanism in HCC. Although further studies should be conducted to detect the potential c19orf10 receptor or signaling molecules binding to c19orf10, our study suggests that c19orf10 may be a novel growth factor, a potential tumor marker and also a potential target molecule for HCC treatment.