C10ORF97 is a novel tumor-suppressor gene of non-small-cell lung cancer and a functional variant of this gene increases the risk of non-small-cell lung cancer

Abstract

In an earlier study we showed that C10ORF97 (chromosome-10, open reading frame-97) was expressed in almost all of the tissues and cell lines tested, and that it inhibited the growth of seven tumor cell lines, including two lung carcinoma cell lines (A549 and PG). Here, we show that C10ORF97 is downregulated in non-small-cell lung cancer (NSCLC) tissue compared with normal lung tissue. Overexpression of C10ORF97 significantly suppressed human lung carcinoma A549 cell growth (proliferation and anchorage-independent growth in soft agar) and motility (migration and adhesion). This tumor-suppressive function of C10ORF97 was also verified in vivo. We further found that C10ORF97 caused G1 arrest of A549 cells and modulated the expression level of several cell-cycle regulators (such as CDK2, cyclin-E and p27). These effects of C10ORF97 were mediated by physical association between C10ORF97 and Jun-activating domain-binding protein-1 (JAB1), and blocking of JAB1-mediated translocation of p27 from the nucleus to the cytoplasm. Together, these results indicated that C10ORF97 functions as a novel tumor suppressor by modulating several key G1/S-regulatory proteins by interacting with JAB1. These findings led us to hypothesize that a single-nucleotide polymorphism (SNP) in the C10ORF97 gene that affects its expression might be associated with susceptibility to NSCLC. SNP216 C>T (rs2297882) in the C10ORF97 Kozak sequence was identified, and allele T of SNP216 suppressed C10ORF97 expression in vitro and in vivo. Furthermore, the TT genotype of SNP216 was associated with an increased risk of NSCLC (adjusted odds ratio=1.73 (95% confidence interval: 1.33–2.25), P=4.6 × 10–5). These data indicated that C10ORF97 is a tumor suppressor of NSCLC progression and C10ORF97-SNP216 may serve as a predictor of NSCLC.

Introduction

Lung cancer is the leading cause of cancer-related death throughout the world. Global cancer statistics indicate that lung cancer is responsible for over 1.2 million deaths each year (Parkin et al., 2005; Jemal et al., 2006). Non-small-cell lung cancer (NSCLC), including squamous cell carcinoma, adenocarcinoma and large-cell carcinoma, is the predominant type of lung cancer (approximately 80–85% of all cases) (Hoffman et al., 2000; Govindan et al., 2006). In recent years, major progress has been made in understanding the molecular pathogenesis of lung cancer, leading to new strategies for early detection, diagnosis, staging and therapy (Meyerson et al., 2004; Weinstein and Joe, 2006; Sato et al., 2007). However, the proportion of patients benefiting from these achievements has remained low.

Evasion of apoptosis and the ability to proliferate uncontrollably are two of the molecular traits found in most, if not all, human cancers (Hanahan and Weinberg, 2000). Loss of cell-cycle control has been implicated in tumor development and proliferation. Cell-cycle progression is regulated by cyclins, cyclin-dependent kinases (CDKs) and CDK inhibitors (CDKIs) (Bloom and Cross, 2007; Malumbres and Barbacid, 2007). p27, a member of the Kip family of CDKIs, inhibits G1/S cell-cycle progression by inactivating cyclin-E and cyclin-A–CDK2 complexes (Sheer and Roberts, 1999). Decreased expression of p27 and increased levels of cyclin-E occur and are associated with a poor prognosis in a wide spectrum of cancers (Esposito et al., 1997; Catzavelos et al., 1999; Tsukamoto et al., 2001). The sub-cellular localization as well as the expression levels of p27 protein have an important role in cancer progression (Slingerland and Pagano, 2000; Viglietto et al., 2002; Wu et al., 2006). Jun activation domain-binding protein (JAB1), a coactivator of activator proteins (AP-1), interacts with c-jun and promotes cell growth (Claret et al., 1996). It was also shown that JAB1 can shuttle p27 from the nucleus to the cytoplasm and accelerate the degradation of p27 through the ubiquitin/proteasome pathway (Tomoda et al., 1999). Many studies have reported that JAB1 is overexpressed in different types of cancers. The overexpression of JAB1 is inversely associated with p27 levels and correlates with poor prognosis (Kouvaraki et al., 2003; Shintani et al., 2003; Osoegawa et al., 2006).

C10ORF97 (chromosome-10, open reading frame-97), localized on chromosome-10p13, is a novel gene cloned in our laboratory (Liu et al., 2002). We found that stable transfection of C10ORF97 can inhibit cell proliferation significantly in seven tumor cell lines, including two lung cancer cell lines (A549 and PG), and overexpression of C10ORF97 can induce human lung carcinoma A549 cell apoptosis (Liu et al., 2002). These data led us to hypothesize that C10ORF97 could suppress the progression of NSCLC and that elucidating the role of this novel gene (C10ORF97) in NSCLC could be useful in developing individual molecular therapeutic strategies.

Results

C10ORF97 expression was downregulated in NSCLC tissues

To examine the expression of C10ORF97 in lung cancer tissue, the microarray, Cancer Profiling Array-I, was used to compare C10ORF97 expression between matched normal tissue samples and tumor tissue samples. Upregulation was defined as C10ORF97 expression levels that were 1.5-fold higher in tumor tissues than in normal tissues, and downregulation was defined as expression levels that were 0.7-fold lower in tumor tissues than in normal tissues. Using these criteria, we found that the C10ORF97 gene was downregulated in 57% (12/21) of the lung cancer tissues and upregulated in 14% (3/21) of these tissue samples (Figure 1a). These results were verified in 420 NSCLC clinical samples using the tissue microarray. The intensity of C10ORF97 staining was scored on a gray scale as low (1+), moderate (2+) and high (3+) (Figure 1b). Low C10ORF97 staining intensity was identified in 51.7% (217/420) of the NSCLC samples and 9% (19/210) of the normal lung samples. By contrast, only 5.2% (22/420) of the NSCLC samples and 40.5% (85/210) of the normal lung samples showed high staining intensity of C10ORF97 (Table 1). These results indicated that C10ORF97 might be a tumor suppressor, at least in NSCLC.

Figure 1
figure1

The expression of C10ORF97 was downregulated in NSCLC tissues. (a) Expression of C10ORF97 in matched normal lung tissues (N) versus lung tumor tissues (T). These data were quantified using the PhosphorImager and ImageQuant softwares. Twelve of the 21 lung tissue samples (57%) showed a decrease in C10ORF97 expression in the tumor tissue versus the expression in the matched normal tissue. (b) C10ORF97 expression was assessed using a human tissue microarray of NSCLC tissue obtained from Cybrdi (Xi’an, Shanxi, China). The array contained 730 dots in total, 420 of which represented diseased tissue. The intensity of C10ORF97 staining was graded on a gray scale as low (1+), moderate (2+) or high (3+) ( × 200). As shown in Table 1, C10ORF97 was downregulated in 51.7% of the NSCLC tissues. C10ORF97, chromosome-10, open reading frame-97; NSCLC, non-small-cell lung cancer.

Table 1 The expression of C10ORF97 was downregulated in NSCLC tissues

C10ORF97 inhibited A549 cell growth and motility in vitro

To test the hypothesis that C10ORF97 functions as a tumor-suppressor gene, overexpression (using adenoviral (Ad) vector) or knockdown (by short interfering RNA (siRNA)) of C10ORF97 was performed in the human lung cancer A549 cell line, and cell proliferation, clonogenicity in soft agar, and cell adhesion and motility were evaluated. The results are shown in Figure 2.

Figure 2
figure2

C10ORF97 inhibited A549 cell growth and motility in vitro. (a) C10ORF97 inhibited A549 cell proliferation. Cells were transfected in vitro with Ad-C10ORF97 (MOI=10–20) or siRNA-C10ORF97, and cells treated with PBS, empty adenovirus vector (Ad-GFP) or siRNA-negative control were used as controls. The viability of cells was determined by MTT staining on days 1, 3, 5 and 7 after cell plating. No significant differences were found between the control groups at day 7 (n=6, P=0.95 and 0.25, respectively). (b) Ad-C10ORF97 suppressed the clonogenicity of A549 cells. Cells (1.5 × 103) from each treatment were plated onto culture dishes containing soft agar medium and were cultured for 15 days. Colonies were observed using a microscope and counted. The Ad-C10ORF97 panel shows the number of colonies formed after A549 cells were infected with Ad-C10ORF97; the Ad-GFP and PBS panels show the number of colonies formed after Ad-GFP or PBS treatment. These data represent one of three independent experiments, and are the mean±s.d.; **P<0.01. (c) C10ORF97 inhibits A549 cell migration. Cells from each treatment were trypsinized and seeded into the top chamber of the transwells. The cells were then allowed to migrate for 4–6 h through the transwell membranes. To quantitate migration, stained cells were counted in six fields from each of the three transwell filters per experiment. The data represent the mean±s.d. of three independent experiments. (d) Overexpression of C10ORF97 inhibits A549 cell adhesion. Cells were seeded at a concentration of 1 × 105 cells/well in serum-free media onto protein-coated surfaces containing matrigel and allowed to attach for 40–60 min at 37 °C. Non-adherent cells were removed and the MTT assay was performed to quantify the number of adhered cells. The experiments were performed in triplicate, and each data point represents n=6. Each bar represents the mean±s.d. Ad-C10ORF97, recombinant C10ORF97 adenovirus; C10ORF97, chromosome-10, open reading frame-97; MOI, multiplicity of infection; NSCLC, non-small-cell lung cancer; PBS, phosphate-buffered saline; siRNA, short interfering RNA.

Ad-C10ORF97 markedly suppressed A549 cell proliferation to 60% on day 5 and 32% on day 7 as compared with the controls (P<0.0001), and siRNA-C10ORF97 slightly promoted A549 cell proliferation by 45% on day 5 and 52% on day 7 compared with the controls.

To determine whether C10ORF97 inhibits the clonogenicity of A549 cells in soft agar, we pretreated A549 cells with Ad-C10ORF97 (multiplicity of infection (MOI)=10–20), control vector Ad-GFP (MOI=10–20) or phosphate-buffered saline (PBS) alone, and tested their ability to form colonies in soft agar (Figure 2b). After 15 days we found that 380 (364) colonies were formed in the Ad-GFP (PBS) group, but only 210 colonies were formed in the Ad-C10ORF97 group (P<0.001). Our results showed that Ad-C10ORF97 suppressed the colony-forming ability of A549 cells compared with the control vector (Ad-GFP) and PBS.

Next, we tested whether C10ORF97 could inhibit A549 cell adhesion to the extracellular matrix. We found that overexpression of C10ORF97 inhibited A549 cell adhesion to 51% after 45 min compared with the controls (P<0.001), and knockdown of C10ORF97 stimulated A549 cell adhesion 1.87-fold (P<0.001; Figure 2c).

We then examined the effects of C10ORF97 on A549 cell motility using transwell migration chambers. Cells transfected with Ad-C10ORF97 showed a decrease in motility of 51% compared with the controls (P<0.01), and siRNA-C10ORF97 promoted A549 cell migration 2.32-fold (P<0.001; Figure 2d).

C10ORF97 inhibited A549 cell-induced tumorigenicity and tumor growth in vivo

To examine whether the observed inhibitory effects of Ad-C10ORF97 in vitro could be extended in vivo, we tested the tumorigenic potential of Ad-C10ORF97-transfected A549 cells in nude mice (Figure 3). We first infected A549 cells in vitro with Ad-C10ORF97 or Ad-GFP at an MOI of 20 and then subcutaneously injected the infected cells into the right flank of nude mice to assess their ability to inhibit tumor growth in vivo. As shown in Figure 3a, all of the animals injected with either untreated or Ad-GFP-transfected A549 cells began to develop tumors after day 7. Once tumors were established, they grew aggressively, ultimately reaching an average volume of about 1.7–2.7 cm3 in mice that received injections of untreated cells and 0.9–2.1 cm3 in mice that received injections of Ad-GFP, within 1 month after injection. By contrast, no tumors formed in mice injected with Ad-C10ORF97-transfected A549 cells within the same time period. These results showed that Ad-C10ORF97 inhibited the tumorigenicity of A549 cells in vivo.

Figure 3
figure3

C10ORF97 inhibited the tumorigenicity and tumor growth of A549 cells in vivo. (a) Ad-C10ORF97 pre-transfection suppressed A549 cell-induced tumorigenicity in nude mice. A549 cells were pretreated with Ad-C10ORF97, Ad-GFP or PBS, and injected into the right flank of nude mice (5 × 106 cells per mouse). Tumor volumes were calculated using the equation V=a × b2/2, where a is the largest diameter and b is the perpendicular diameter. No tumors were formed at any time in the course of the experiment in any of the seven mice inoculated with Ad-C10ORF97-treated cells. The difference between the tumor volumes of the Ad-control-treated mice versus that of the PBS-treated mice was not significant (t-test, P=0.325). (b) Inhibition of tumor growth in vivo by intra-tumoral injection with Ad-C10ORF97. The arrows indicate the schedule of the injections. A549 cells (5 × 106 cells per mouse) were injected into the right flank of female nude mice. When the tumors reached 5–10 mm in diameter at 10 days after tumor inoculation, the adenoviral vectors were administered as described under Materials and methods. The data represent the mean±s.d. for seven mice from each treatment group. The differences between the tumor volumes of the Ad-C10ORF97-treated mice versus that of the Ad-GFP-treated mice or the PBS-treated controls were significant (P<0.001, respectively). Ad-C10ORF97, recombinant C10ORF97 adenovirus; C10ORF97, chromosome-10, open reading frame-97; PBS, phosphate-buffered saline.

To further evaluate the therapeutic effect of Ad-C10ORF97 on tumor growth, we investigated the ability of Ad-C10ORF97 to suppress tumor growth in vivo by intra-tumoral injection of high-titer recombinant C10ORF97 adenovirus in an A549 cell-induced tumor model in mice. The growth of tumors was recorded from the first injection until 1 month after the last injection (Figure 3b). The growth of all tumors was significantly suppressed in mice treated with Ad-C10ORF97 compared with tumors in PBS-treated and Ad-GFP-treated mice (P<0.001). At the end of these experiments, the tumors in the mice treated with Ad-C10ORF97 were, on average, only about one-tenth the size of the tumors in the mice treated with PBS alone or adenovirus vector alone.

C10ORF97 overexpression caused G1 arrest in A549 cells

To further explore the tumor-suppressive pathway of C10ORF97, we tested the impact of C10ORF97 on cell-cycle kinetics in vitro by flow cytometry (fluorescence-activated cell sorting) analysis using propidium iodide staining (Figure 4a). A549 cells were treated with Ad-C10ORF97, Ad-GFP or PBS alone, and cells were harvested at different time points. As shown in Figures 4b and c, Ad-C10ORF97 infection resulted in a significant increase in the number of cells in G1 phase and a decrease in the number of cells in S phase in A549 cells 5 days after infection. The number of cells in G1 phase increased from 55% at day 1 to 72% at day 5 after infection in A549 cells, and the number of cells in S phase decreased from 36 to 17%. By contrast, no significant changes in the number of cells in G1 or S phase were observed in cells treated with Ad-GFP or PBS. This observation indicated that Ad-C10ORF97 inhibits tumor cell growth through G1 arrest.

Figure 4
figure4

C10ORF97 overexpression caused G1 arrest in A549 cells. Cells were treated with Ad-GFP, Ad-C10ORF97 or PBS, and harvested at the indicated time points. Cell-cycle distribution was analyzed by fluorescence-activated cell sorting, after propidium iodide staining. The cell distribution data (%) for each treatment group is the mean of four plates and is representative of three independent experiments. The bars represent the s.d.s of the means. (a) Effects of C10ORF97 on the cell-cycle progression of A549 cells at days 3 and 5. (b) Effects of C10ORF97 on the G1-phase populations of A549 cells. (c) Effects of C10ORF97 on the S-phase populations of A549 cells; **P<0.01 (n=4). Ad-C10ORF97, recombinant C10ORF97 adenovirus; C10ORF97, chromosome-10, open reading frame-97; PBS, phosphate-buffered saline.

C10ORF97 modulated the expression of several cell-cycle regulators

To investigate more thoroughly the role of C10ORF97 in cancer progression, the Atlas human cancer array from Clontech (Clontech Laboratories Inc., Palo Alto, CA, USA) was used to identify gene expression profiles in response to overexpression of C10ORF97 in A549 cells. Overexpression of C10ORF97 downregulated cancer progression-related genes, including cell-cycle-regulating kinases, cyclins and growth factors, whereas some tumor-suppressor genes were upregulated (data not shown).

To investigate the molecular events associated with cell-cycle regulation, western blot analysis was performed to examine the effect of Ad-C10ORF97 on the expression of cell-cycle regulators that have been implicated in the control of G1/S transition. We analyzed the expression of cyclin-E, CDK2, p21, p27 and p53 (Figure 5). As shown in Figure 5A, the level of cyclin-E protein decreased significantly after 48 h of Ad-C10ORF97 infection and a moderate decrease in the expression of CDK2 was also detected. We further found that the expression of p21 and p27 markedly elevated after infection with Ad-C10ORF97. A significant increase in the expression of p53 was also found. These results strongly supported an important role for C10ORF97 in cell-cycle progression.

Figure 5
figure5

C10ORF97 modulated the expression of several cell-cycle regulators and is associated with JAB1 in mammalian cells. (A) C10ORF97 modulated the expression of several cell-cycle regulators. Total protein was isolated from cells treated with Ad-C10ORF97 or Ad-GFP for 48 h, and the expression of cell-cycle regulators was analyzed by western blotting. As shown in the lowest panel, each sample was reprobed with actin antibody to monitor the quantity of protein loaded on each lane. (B) Interaction of Myc-tagged C10ORF97 with HA-tagged JAB1. A co-immunoprecipitation assay was performed as described under Materials and methods. a: The total amount of plasmid transfected was kept constant by inclusion of the appropriate amount of empty expression vector in each mixture. Seventy-two hours after transfection, the cells were lysed and aliquots of the lysates were used to examine the expression of HA-JAB1 or Myc-C10ORF97. b: The rest of the lysates were immunoprecipitated with a polyclonal antibody to the HA epitope. The presence of Myc-C10ORF97 in the immunoprecipitates was examined using a monoclonal antibody to the Myc epitope. (C) C10ORF97 and JAB1 colocated in A549 cells. Transfection and immunofluorescence co-location, referred to the ‘Materials and methods’ in Supplementary Information. Immunofluorescence imaging was performed on A549 cells transiently transfected with pCMV-HA-JAB1 and pCMV-Myc-C10ORF97. The location of C10ORF97 was investigated by staining for C10ORF97 (red) using an anti-Myc antibody, and for JAB1 (green) using an anti-HA antibody. The nucleus was stained with DAPI (blue). Merged images are shown, with the coincidence of red and green staining appearing as yellow/orange. Ad-C10ORF97, recombinant C10ORF97 adenovirus; C10ORF97, chromosome-10, open reading frame-97; DAPI, 4,6-diamidino-2-phenylindole; JAB1, Jun-activating domain-binding protein-1.

C10ORF97 was associated with JAB1 in mammalian cells

To confirm the location of C10ORF97 in the cell signaling networks, a yeast two-hybrid system was used to screen for C10ORF97-interacting protein(s). We identified several types of clones that interacted specifically with C10ORF97. One of these clones contained a cDNA insert comprising almost the entire coding sequence (amino acids 36–335) of human JAB1, a coactivator of AP-1. To verify this result, co-immunoprecipitation experiments and immunofluorescence colocation assays were performed.

For the co-immunoprecipitation experiments, we amplified the full-length JAB1 coding sequence from a human heart cDNA library and constructed the plasmid pCMV-HA-JAB1. A549 cells were transfected with expression plasmids producing Myc-tagged C10ORF97 and HA-tagged JAB1 protein. Whole-cell extracts were also analyzed by western blotting. No Myc-tagged C10ORF97 protein was found in cells transfected with pCMV-Myc and pCMV-HA-JAB1, or in untransfected cells. Myc-tagged C10ORF97 protein was found in cells transfected with pCMV-Myc-C10ORF97 and pCMV-HA, or pCMV-Myc and pCMV-HA-JAB1, in the same quantity (Figure 5B). A lysate fraction of the transfected cells was immunoprecipitated by anti-hemagglutinin (HA) antibody and subjected to sodium dodecyl sulfate-PAGE. Immunoblotting with the anti-Myc antibody showed that C10ORF97 co-immunoprecipitated with JAB1 (Figure 5B).

A549 cells were transiently transfected with pCMV-HA-JAB1 and pCMV-Myc-C10ORF97, and treated as described under Materials and methods. As seen in Figure 5C, the C10ORF97 staining pattern (red) and the JAB1 staining pattern (green) largely overlapped (orange–yellow). These findings indicated that C10ORF97 forms a complex with JAB1.

C10ORF97 repressed AP-1 activity and enhanced p27 stability through interaction with JAB1

JAB1 was initially identified as a coactivator of c-Jun, a subunit of the COP9 signalosome. Our luciferase reporter assay showed that C10ORF97 inhibited JAB1-mediated AP-1 activation in a dose-dependent manner (Figure 6a). These data indicated that C10ORF97 represses JAB1-induced AP-1 activity.

Figure 6
figure6

C10ORF97 repressed AP-1 activity and enhanced p27 stability through interaction with JAB1. (a) C10ORF97 inhibited the JAB1-mediated activation of AP-1 reporter activity in A549 cells. A549 cells were transiently transfected with 1 μg of pCDNA3.1-JAB1, various amounts of pCDNA3.1-C10ORF97 (1, 2 and 3 μg), pRL3-luciferase (0.02 μg) and AP-1-driven luciferase reporter constructs (0.5 μg). pRL-SV40 plasmid (Promega) was co-transfected as a normalizing control. The data show that C10ORF97 inhibited JAB1-mediated AP-1 activation in a dose-dependent manner. (b) C10ORF97 blocked the p27 nuclear export mediated by JAB1. The sub-cellular localization of p27 in A549 cells transfected with pCDNA3.1-p27, pCDNA3.1-JAB1 and pCDNA3.1-C10ORF97 is indicated. The p27 images are shown in the first row, DAPI images are shown in the second row and merged images of p27 and DAPI are shown in the third row. (c) Total protein was isolated from A549 cells treated with pCDNA3.1-p27, pCDNA3.1-JAB1 or pCDNA3.1-C10ORF97 as indicated. The expression of p27 was analyzed by western blotting. AP, activator protein; C10ORF97, chromosome-10, open reading frame-97; JAB1, Jun-activating domain-binding protein-1.

It is known that p27 stability is controlled by JAB1, which induces the translocation of p27 from the nucleus to the cytoplasm. When JAB1 cDNA was co-transfected into A549 cells with p27 cDNA, p27 protein expression was reduced (Figure 6c). Additional transfection of C10ORF97 cDNA restored the JAB1-mediated suppression of p27 protein expression.

To understand the molecular mechanism of C10ORF97-mediated p27 stability, immunofluorescence microscopy was performed. As shown in Figure 6b, p27 was stably located in the nucleus (left panels) and expression of JAB1 induced translocation of p27 from the nucleus to the cytoplasm (central panels). However, C10ORF97 blocked the translocation of p27 to the cytoplasm (right panels). Taken together, these data indicated that the physical interaction between C10ORF97 and JAB1 may be the mechanism by which C10ORF97 increases p27 accumulation in the nucleus.

SNP216C>T was associated with C10ORF97 expression levels in NSCLC patients and the risk of NSCLC

Taken together, our results indicated that C10ORF97 is a novel tumor-suppressor gene of NSCLC, leading us to hypothesize that the single-nucleotide polymorphism (SNP) in the C10ORF97 gene that affects its expression might be associated with susceptibility to NSCLC.

Using HapMap data, we identified SNP216C>T (rs2297882) located in the Kozak sequence of the C10ORF97 gene, an allele with replacement of cytosine (C) by thymine (T) at position −5 from the ATG start codon (Figure 7a). At first, a PCR–ligase detection reaction assay was performed in a pilot study containing 100 individuals with NSCLC and 100 controls to determine the genotype frequency of rs2297882. A weak association was found between rs2297882 and NSCLC (data not shown).

Figure 7
figure7

Kozak sequence polymorphism in the C10ORF97 gene affected the translation efficiency of C10ORF97. (a) The C10ORF97 sequence surrounding the start codon and the location of the polymorphism. Variant rs2297882 shows replacement of cytosine (C) with thymine (T) at position −5 from the ATG start codon. (b) In vitro transcription/translation analysis. Protein production from the two forms of the C10ORF97 cDNA was evaluated in a cell-free transcription/translation system. Substantially less protein was synthesized from the −5T plasmid. This figure represents one of three independent experiments. C10ORF97, chromosome-10, open reading frame-97.

We next investigated whether the −5C to T change in the Kozak sequence affected the translation efficiency of −5C and −5T cDNAs. An in vitro transcription/translation assay was performed. We found that significantly less protein was produced from the −5T cDNA than from the −5C form (Figure 7b), which strongly suggested that the polymorphism affected the efficiency of translation.

To test whether the expression level of the C10ORF97 gene in tumor tissues was associated with SNP216, diseased tissues were acquired from 182 individuals and immunochemistry was performed on these specimens. The intensity of C10ORF97 staining was graded on a gray scale as low (+), moderate (2+) or high (3+). A strong correlation was found between the C10ORF97 expression level and the SNP216 genotype (P=0.003) (Table 2). This finding confirmed the results of the reporter assay.

Table 2 C10ORF97 expression level in NSCLC patients correlated with the rs2297882 genotype

The functional significance of this variant prompted us to validate the association between SNP216 and lung cancer in a larger case-controlled study. We randomly selected 418 Chinese NSCLC patients with histologically confirmed NSCLC and 743 age- and sex-matched controls. The distribution of SNP216 is shown in Table 3, and conformed to Hardy–Weinberg equilibrium for both the cases and the controls. The estimated NSCLC risk was significantly higher in subjects carrying SNP216TT than in those carrying SNP216TC or SNP216CC, indicating a recessive model for the T-allele. Thus, the recessive model was applied to analyze the variant's association with NSCLC. The frequency of the SNP216 TT genotype was significantly higher in the patients than in the controls (69.9 versus 56.8% (crude odds ratio=1.76, 95% confidence interval: 1.37–2.27), P=1.3 × 10–5), after adjustment for age, sex and tobacco smoking, using unconditional logistic regression analysis, and the TT genotype resulted in a significantly increased risk of NSCLC (odds ratio=1.73, 95% confidence interval: 1.33–2.25, P=4.6 × 10–5).

Table 3 The SNP (rs2297882) in the C10ORF97 Kozak sequence was associated with NSCLC

Discussion

The C10ORF97 gene was first cloned in our laboratory in 2000 and we reported the suppressive activity of C10ORF97 on tumor cell line growth in 2002 (Liu et al., 2002). In the current study, we have shown that (a) C10ORF97 was downregulated in cancer tissues; (b) C10ORF97 functioned as a tumor suppressor both in vitro and in vivo; (c) C10ORF97 might achieve its tumor-suppressive function through inhibiting cyclin-E/CDK2 and promoting p27/p21/p53, causing G1 arrest; (d) C10ORF97 repressed AP-1 activity and increased p27 accumulation in the nucleus by binding to JAB1; and (e) SNP216 (rs2297882) in the C10ORF97 Kozak sequence was associated with susceptibility to lung cancer, and affected C10ORF97 translation and C10ORF97 expression levels in the tumor tissues of NSCLC patients.

In an earlier study, we found that the C10ORF97 gene was expressed in almost all tissues and cell lines tested, although its expression varied markedly (Liu et al., 2002). This ubiquitous expression of C10ORF97 suggested that it may have an important role in the regulation of cell progression. In this study, we found significantly decreased expression of C10ORF97 in lung cancer tissues compared with normal tissues. Our results suggested that decreased expression of C10ORF97 might be associated with tumorigenesis or be important for tumor progression and maintenance, and that the C10ORF97 protein might have a role in growth inhibition or differentiation in normal tissues. The mechanism behind the decrease in C10ORF97 expression observed in cancer tissues remains unknown, but may reflect both genetic and epigenetic events, and needs further exploration.

To validate our hypothesis that the C10ORF97 gene functions as a growth suppressor, a recombinant C10ORF97 adenovirus (Ad-C10ORF97) was constructed and we showed that Ad-C10ORF97 leads to inhibited tumor cell growth and tumorigenicity both in vitro and in vivo. The growth rate of tumor cell lines was significantly suppressed and the number of soft agar colonies dramatically decreased after infection with Ad-C10ORF97. The reduced growth potential of Ad-C10ORF97-transfected A549 cells was also reflected by the elimination of tumorigenicity. The intra-tumoral injection of high-titer Ad-C10ORF97 also resulted in markedly suppressed A549-induced tumor growth. Taken together, these findings strongly supported the hypothesis that the C10ORF97 gene functions as a growth suppressor.

Through cell-cycle analysis we found that Ad-C10ORF97 caused G1 cell-cycle arrest, indicating that C10ORF97 induced a blockage in cell-cycle progression from G1 to S phase. Previous studies have shown that cyclins, CDKs and CDKIs are key regulators of eukaryotic cell-cycle progression (Sherr, 2000; Kastan and Bartek, 2004; Massague, 2004). Different cyclin–CDK complexes are positive regulators involved in different cell-cycle transitions, and the activities of CDK complexes are negatively regulated by CDKIs. Negative regulation of cell growth is often lost in malignant cells, whose cell growth proceeds unchecked, eventually leading to the development of a tumor. p27, belonging to the CIP/KIP family of CDKIs, fulfills an essential role in the control of cell proliferation by directly controlling the G1/S transition. Absence of or reduced expression of p27 has been observed in many human cancers, including breast, lung, prostate, colon, skin and ovarian cancers (Lloyd et al., 1999; Blain et al., 2003; Masciullo et al., 2003). We found a significant reduction in the expression of cyclin-E and CDK2, and a remarkable increase in the protein level of p27, p21 and p53 after Ad-C10ORF97 transfection. These results indicated that the increased p27/p21/p53 expression levels may be responsible for the decreased expression of cyclin-E and CDK2, which potentially induced G1 arrest and blockage of the G1/S transition.

AP-1 is a central component of many signal transduction pathways in a variety of cell types and is critical for cell proliferation, differentiation, apoptosis and oncogene-induced transformation (Shaulian and Karin, 2002). JAB1 was initially identified as a coactivator of AP-1 (Claret et al., 1996). Recent studies have shown that JAB1 interacts with a variety of proteins and controls their functions, having roles in diverse cell signaling pathways, including regulation of gene transcription and cell-cycle progression by phosphorylation and degradation of target proteins (Chamovitz and Segal, 2001). JAB1 was also identified as a direct negative regulator of p27 that operates by transferring p27 from the nucleus to the cytoplasm to accelerate its degradation (Tomoda et al., 1999, 2002). In this study, we showed that C10ORF97 physically interacted with JAB1, as shown by a yeast two-hybrid system, co-immunoprecipitation and immunofluorescence colocation assay. Through this interaction, C10ORF97 inhibited JAB1-mediated AP-1 activation in a dose-dependent manner, and blocked the JAB1-mediated translation of p27. These findings might, at least in part, explain how C10ORF97 increases the expression level of p27. Recently, elevated JAB1 expression was reported in human pituitary tumors, epithelial ovarian cancer, breast cancer and some types of lymphoma (Sui et al., 2001; Korbonits et al., 2002; Kouvaraki et al., 2003; Rassidakis et al., 2003). In these cases, the JAB1 levels were inversely associated with the p27 levels and positively correlated with disease progression. Thus, JAB1 levels are thought to be a novel indicator of aggressiveness and high-grade tumor behavior, implying that control of JAB1 could be a novel target for the development of new cancer therapies. Based on this, we speculated that C10ORF97, directly interacting with JAB1, could modulate the progression of tumors showing high-level JAB1 expression.

C10ORF97 was previously described as a nuclear protein in 293 cells (Liu et al., 2002). In the present study, we found that C10ORF97 was localized both in the cytoplasm and the nucleus in NSCLC tissues and cell lines. Similar phenomena are observed with several tumor suppressors. For example, breast cancer metastasis suppressor-1 localization was predominantly nuclear, but 60–70% of cancers also showed cytoplasmic immunostaining, and 11% of cancers showed lower nuclear than cytoplasmic breast cancer metastasis suppressor-1 expression (Frolova et al., 2009). As C10ORF97 increased p27 accumulation in the nucleus by physical interaction with JAB1, we speculated that localization of C10ORF97 in the cytoplasm may attenuate the transfer of p27 to the nucleus by interfering with the physical interaction between C10ORF97 and JAB1, because JAB1 was mainly localized in the nucleus, thereby promoting cell proliferation, the characteristic feature of cancer cells.

In the present study, SNP216C>T (rs2297882) was shown to cause a decrease in C10ORF97 translation. More importantly, genotype–phenotype analysis indicated that the SNP216 TT genotype, which led to lower expression of C10ORF97, was associated with an increased risk of developing lung cancer in the Chinese population. This genotype–phenotype correlation might be explained by the tumor-suppressive function of C10ORF97. Despite the relatively small sample size, these results showed significantly increased odds ratios with small P-values, and were therefore unlikely to be attributable to selection bias or unknown confounding factors.

In summary, our study indicated that C10ORF97 functions as a tumor suppressor by modulating several key G1/S-regulatory proteins by interacting with JAB1, and suggested the potential application of Ad-C10ORF97 in human cancer gene therapy. In addition, decreased levels of C10ORF97 might be a new target for the early detection of lung cancer.

Materials and methods

Materials

The antibodies against CDK2, cyclin-E, p27, p21, p53, β-actin, myc, HA, Flag and JAB1 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA) or Sigma (St Louis, MO, USA). Nu/nu mice (BALB/c, 4- to 6-week-old females) were purchased from the Laboratory Animal Center, Chinese Academy of Medical Sciences (Beijing, China). The human tissue collection protocol was approved by the Fuwai Hospital Ethics Committee. Informed written consent was obtained from patients themselves or their legal representatives. Animal experiments conformed to the guiding principles of Chinese National Law for Animal Use in Medical Research and were approved by the Fuwai Hospital Committee for Animal Care and Use.

Cell lines

The human lung cancer A549 cell lines and the 293 cell line were obtained from the Institute of Cell Biology, Academic Sinica (Beijing, China). Human lung cancer A549 cell lines were cultured at 37 °C and 5% CO2 in F12 medium (Gibco BRL, Gaithersburg, MD, USA) containing 10% fetal bovine serum (HyClone, Logan, UT, USA). The 293 cell line was used in the construction, amplification and titration of adenoviral vectors, and was maintained in Dulbecco's modified Eagle's medium (Gibco BRL) containing 10% fetal bovine serum.

Cancer profiling array analysis

The Cancer Profiling Array-I was purchased from Clontech (cat. no. 7481-1) and hybridized according to the manufacturer's protocol.

Production and purification of the C10ORF97 monoclonal antibody

Production and purification of the C10ORF97 monoclonal antibody are described in the Supplementary material.

Tissue microarray analysis

The Human Tissue Microarray of NSCLC tissue was obtained from Cybrdi (Xi’an, Shanxi, China). The array contained 630 spots in total, 420 of which represented diseased tissue from one individual specimen that was selected and pathologically confirmed, and the remaining spots represented normal lung tissues. The arrays were fixed with formalin, embedded in paraffin and immunostained with a mouse monoclonal anti-C10ORF97 antibody (1:900 dilution) using the avidin–biotin peroxidase complex method. The intensity of C10ORF97 staining was scored on a gray scale as low (1+), moderate (2+) or high (3+).

Recombinant adenovirus production, stable transfection, stealth siRNA treatment, real-time RT–PCR, cell proliferation assay, cell migration assay, cell adhesion assay and soft agar colony-forming assay

The above methods are described in the Supplementary material.

In vivo studies

Nu/nu mice (BALB/c, 4- to 6-week-old females) were obtained from CAMS (the Chinese Academy of Medical Sciences) and seven mice were used for each treatment. A549 cells were pre-infected with Ad-C10ORF97 or Ad-GFP at an MOI of 50, or with PBS alone for 48 h. The treated cells (5 × 106 cells per mouse) were subcutaneously injected into the right flank of the nude mice. Tumor dimensions were measured every 2–3 days using a linear caliper. The tumor volume was calculated using the equation V=a × b2/2, where a is the largest dimension and b is the perpendicular diameter.

To further investigate the tumor-suppressive effect of C10ORF97, mice were first subcutaneously injected with A549 cells (5 × 106 cells per mouse), leading to tumor formation. Once tumors had grown to 5–10 mm in diameter (at about 10 days after injection), the mice received intra-tumoral injections of Ad-C10ORF97, Ad-GFP or PBS four times, on days 1, 3, 5 and 7, at a total dose of 3 × 1010 plaque-forming units (p.f.u.) per tumor. The tumor volume was measured and calculated as described above.

Cell-cycle analysis, microarray analysis, western blot analysis, plasmids for the yeast two-hybrid screening, yeast two-hybrid screening, co-transfection of A549 cells, co-immunoprecipitation analysis, immunofluorescence microscopy, immunofluorescence colocation, transient transfection of the reporter gene and plasmids for in vitro transcription/translation assay

The above methods are described in the Supplementary material.

In vitro transcription/translation assay

The efficiency of C10ORF97 protein synthesis was assessed using the TNT quick-coupled transcription/translation system (Promega Corp., Madison, WI, USA). The two forms of the C10ORF97 cDNA carrying either the C (pCMV6-XL5-C) or T (pCMV6-XL5-T) allele were used as templates, for which transcription was driven from the T7 promoter. Reactions were performed and the translation products were detected using Transcend non-radioactive translation detection systems (Promega Corp.) according to the manufacturer's instructions.

Study subjects

We randomly selected 418 Chinese NSCLC patients with histologically confirmed NSCLC between 2004 and 2006 in the Beijing Cancer Hospital and 743 age- and sex-matched controls. The age and gender distribution of the two groups was similar in both the case (mean±s.d. age: 58.1±10.9 years; 70.8% male) and the control groups (mean±s.d. age: 61.6±9.3 years; 69.4% male). All subjects were of Han ethnicity and all patients and controls provided written informed consent for the genetic studies, which were approved by the ethical committee of Beijing Cancer Hospital, China.

Selection and genotyping of rs2297882

SNP216C>T (rs2297882) was selected based on its relevance to gene expression, being located in the Kozak sequence of the 5′-untranslated region of the C10ORF97 gene. Genomic DNA was isolated from peripheral blood using the RelaxGene Blood DNA System (TianGen, Beijing, China) according to the manufacturer's protocol. Variant rs2297882 was genotyped by a ligase detection reaction by the Shanghai Biowing Applied Biotechnology Co. Ltd (Shanghai, China). The details of this reaction are provided in the Supplementary material.

Statistical analysis

Both in vitro and in vivo results were expressed as the mean±s.d. or the mean±s.e. The Student's two-sided t-test was used to compare the values of the test and the control samples. A χ2-test was used in testing categorical variables, the Hardy–Weinberg equilibrium of a polymorphism and allele frequencies. The association between C10ORF97 polymorphism and the risk of NSCLC was calculated by multivariate logistic regression adjusted by age, sex and smoking status. The relationship between the rs2297882 genotype and C10ORF97 expression was analyzed by the Mann–Whitney U-test. All statistical analyses were performed using the SPSS 13.0 software and a value of P<0.05 was considered significant.

References

  1. Blain SW, Scher HI, Cordon-Cardo C, Koff A . (2003). p27 as a target for cancer therapeutics. Cancer Cell 3: 111–115.

    CAS  Article  Google Scholar 

  2. Bloom J, Cross FR . (2007). Multiple levels of cyclin specificity in cell-cycle control. Nat Rev Mol Cell Biol 8: 149–160.

    CAS  Article  Google Scholar 

  3. Catzavelos C, Tsao MS, DeBoer G, Bhattacharya N, Shepherd FA, Slingerland JM . (1999). Reduced expression of the cell cycle inhibitor p27Kip1 in non-small cell lung carcinoma: a prognostic factor independent of Ras. Cancer Res 59: 684–688.

    CAS  Google Scholar 

  4. Chamovitz DA, Segal D . (2001). Jab1/CSN5 and the COP9 signalosome. A complex situation. EMBO Rep 2: 96–101.

    CAS  Article  Google Scholar 

  5. Claret FX, Hibi M, Dhut S, Toda T, Karin M . (1996). A new group of conserved coactivators that increase the specificity of AP-1 transcription factors. Nature 383: 453–457.

    CAS  Article  Google Scholar 

  6. Esposito V, Baldi A, De Luca A, Groger AM, Loda M, Giordano GG et al. (1997). Prognostic role of the cyclin-dependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res 57: 3381–3385.

    CAS  PubMed  Google Scholar 

  7. Frolova N, Edmonds MD, Bodenstine TM, Seitz R, Johnson MR, Feng R et al. (2009). A shift from nuclear to cytoplasmic breast cancer metastasis suppressor 1 expression is associated with highly proliferative estrogen receptor-negative breast cancers. Tumour Biol 30: 148–159.

    CAS  Article  Google Scholar 

  8. Govindan R, Page N, Morgensztern D, Read W, Tierney R, Vlahiotis A et al. (2006). Changing epidemiology of small-cell lung cancer in the United States over the last 30 years: analysis of the surveillance, epidemiologic, and end results database. J Clin Oncol 24: 4539–4544.

    Article  Google Scholar 

  9. Hanahan D, Weinberg RA . (2000). The hallmarks of cancer. Cell 100: 57–70.

    CAS  Article  Google Scholar 

  10. Hoffman PC, Mauer AM, Vokes EE . (2000). Lung cancer. Lancet 355: 479–485.

    CAS  Article  Google Scholar 

  11. Jemal A, Siegel R, Ward E, Murray T, Xu J, Smigal C et al. (2006). Cancer statistics, 2006. CA Cancer J Clin 56: 106–130.

    Article  Google Scholar 

  12. Kastan MB, Bartek J . (2004). Cell-cycle checkpoints and cancer. Nature 432: 316–323.

    CAS  Article  Google Scholar 

  13. Korbonits M, Chahal HS, Kaltsas G, Jordan S, Urmanova Y, Khalimova Z et al. (2002). Expression of phosphorylated p27(Kip1) protein and Jun activation domain-binding protein 1 in human pituitary tumors. J Clin Endocrinol Metab 87: 2635–2643.

    CAS  Article  Google Scholar 

  14. Kouvaraki MA, Rassidakis GZ, Tian L, Kumar R, Kittas C, Claret FX . (2003). Jun activation domain-binding protein 1 expression in breast cancer inversely correlates with the cell cycle inhibitor p27(Kip1). Cancer Res 63: 2977–2981.

    CAS  PubMed  PubMed Central  Google Scholar 

  15. Liu B, Liu Y, Chen J, Wei Z, Yu H, Zhen Y et al. (2002). CARP is a novel caspase recruitment domain containing proapoptotic protein. Biochem Biophys Res Commun 293: 1396–1404.

    CAS  Article  Google Scholar 

  16. Lloyd RV, Erickson LA, Jin L, Kulig E, Qian X, Cheville JC et al. (1999). p27kip1: a multifunctional cyclin-dependent kinase inhibitor with prognostic significance in human cancers. Am J Pathol 154: 313–323.

    CAS  Article  Google Scholar 

  17. Malumbres M, Barbacid M . (2007). Cell cycle kinases in cancer. Curr Opin Genet Dev 17: 60–65.

    CAS  Article  Google Scholar 

  18. Masciullo V, Susini T, Zamparelli A, Bovicelli A, Minimo C, Massi D et al. (2003). Frequent loss of expression of the cyclin-dependent kinase inhibitor p27(Kip1) in estrogen-related endometrial adenocarcinomas. Clin Cancer Res 9: 5332–5338.

    CAS  PubMed  Google Scholar 

  19. Massague J . (2004). G1 cell-cycle control and cancer. Nature 432: 298–306.

    CAS  Article  Google Scholar 

  20. Meyerson M, Franklin WA, Kelley MJ . (2004). Molecular classification and molecular genetics of human lung cancers. Semin Oncol 31: 4–19.

    CAS  Article  Google Scholar 

  21. Osoegawa A, Yoshino I, Kometani T, Yamaguchi M, Kameyama T, Yohena T et al. (2006). Overexpression of Jun activation domain-binding protein 1 in nonsmall cell lung cancer and its significance in p27 expression and clinical features. Cancer 107: 154–161.

    CAS  Article  Google Scholar 

  22. Parkin DM, Bray F, Ferlay J, Pisani P . (2005). Global cancer statistics, 2002. CA Cancer J Clin 55: 74–108.

    Article  Google Scholar 

  23. Rassidakis GZ, Claret FX, Lai R, Zhang Q, Sarris AH, McDonnell TJ et al. (2003). Expression of p27(Kip1) and c-Jun activation binding protein 1 are inversely correlated in systemic anaplastic large cell lymphoma. Clin Cancer Res 9: 1121–1128.

    CAS  PubMed  PubMed Central  Google Scholar 

  24. Sato M, Shames DS, Gazdar AF, Minna JD . (2007). A translational view of the molecular pathogenesis of lung cancer. J Thorac Oncol 2: 327–343.

    Article  Google Scholar 

  25. Shaulian E, Karin M . (2002). AP-1 as a regulator of cell life and death. Nat Cell Biol 4: E131–E136.

    CAS  Article  Google Scholar 

  26. Sherr CJ . (2000). The Pezcoller lecture: cancer cell cycles revisited. Cancer Res 60: 3689–3695.

    CAS  PubMed  PubMed Central  Google Scholar 

  27. Sheer CJ, Roberts JM . (1999). CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev 13: 1501–1512.

    Article  Google Scholar 

  28. Shintani S, Li C, Mihara M, Hino S, Nakashiro K, Hamakawa H . (2003). Skp2 and Jab1 expression are associated with inverse expression of p27(KIP1) and poor prognosis in oral squamous cell carcinomas. Oncology 65: 355–362.

    CAS  Article  Google Scholar 

  29. Slingerland J, Pagano M . (2000). Regulation of the cdk inhibitor p27 and its deregulation in cancer. J Cell Physiol 183: 10–17.

    CAS  Article  Google Scholar 

  30. Sui L, Dong Y, Ohno M, Watanabe Y, Sugimoto K, Tai Y et al. (2001). Jab1 expression is associated with inverse expression of p27(kip1) and poor prognosis in epithelial ovarian tumors. Clin Cancer Res 7: 4130–4135.

    CAS  PubMed  PubMed Central  Google Scholar 

  31. Tomoda K, Kubota Y, Arata Y, Mori S, Maeda M, Tanaka T et al. (2002). The cytoplasmic shuttling and subsequent degradation of p27Kip1 mediated by Jab1/CSN5 and the COP9 signalosome complex. J Biol Chem 277: 2302–2310.

    CAS  Article  Google Scholar 

  32. Tomoda K, Kubota Y, Kato J . (1999). Degradation of the cyclin-dependent-kinase inhibitor p27Kip1 is instigated by Jab1. Nature 398: 160–165.

    CAS  Article  Google Scholar 

  33. Tsukamoto S, Sugio K, Sakada T, Ushijima C, Yamazaki K, Sugimachi K . (2001). Reduced expression of cell-cycle regulator p27(Kip1) correlates with a shortened survival in non-small cell lung cancer. Lung Cancer 34: 83–90.

    CAS  Article  Google Scholar 

  34. Weinstein IB, Joe AK . (2006). Mechanisms of disease: oncogene addiction—a rationale for molecular targeting in cancer therapy. Nat Clin Pract Oncol 3: 448–457.

    CAS  Article  Google Scholar 

  35. Wu FY, Wang SE, Sanders ME, Shin I, Rojo F, Baselga J et al. (2006). Reduction of cytosolic p27(Kip1) inhibits cancer cell motility, survival, and tumorigenicity. Cancer Res 66: 2162–2172.

    CAS  Article  Google Scholar 

  36. Viglietto G, Motti ML, Fusco A . (2002). Understanding p27(kip1) deregulation in cancer: downregulation or mislocalization. Cell Cycle 1: 394–400.

    CAS  Article  Google Scholar 

Download references

Acknowledgements

We thank Professor Gu Xiaocheng and Zheng Xiaofeng (Department of Biochemistry and Molecular Biology, College of Life Sciences, Peking University, Beijing, China) for preparation and purification of the C10ORF97 protein. Sources of support: This study was supported by the National High Technology Research and Development Program of China (863 Program, No. 2006AA02Z477 to RH), the National Natural Science Foundation of China (No. 30500199 to JC and No. 30900809 to YS) and the Foundation of Beijing Municipal Committee of Science and Technology (No. D0905001040631 to JJ).

Author information

Affiliations

Authors

Corresponding authors

Correspondence to J Ji or R Hui.

Ethics declarations

Competing interests

The authors declare no conflicts of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Shi, Y., Chen, J., Li, Z. et al. C10ORF97 is a novel tumor-suppressor gene of non-small-cell lung cancer and a functional variant of this gene increases the risk of non-small-cell lung cancer. Oncogene 30, 4107–4117 (2011). https://doi.org/10.1038/onc.2011.116

Download citation

Keywords

  • C10ORF97
  • tumor suppressor
  • non-small-cell lung cancer
  • genetic variant

Further reading

Search

Quick links