The aminoacyl tRNA synthetase complex-interacting multifunctional protein 2 (AIMP2) splice variant designated DX2 is induced by cigarette smoke carcinogens and is often detected in human lung cancer specimens. However, the function of DX2 in lung carcinogenesis is obscure. In this study, we found that DX2 expression was induced by oncogenes in human lung cancer tissues and cells. DX2 prevented oncogene-induced apoptosis and senescence and promoted drug resistance by directly binding to and inhibiting p14/ARF. Through chemical screening, we identified SLCB050, a novel compound that blocks the interaction between DX2 and p14/ARF in vitro and in vivo. SLCB050 reduced the viability of human lung cancer cells, especially small cell lung cancer cells, in a p14/ARF-dependent manner. Moreover, in a mouse model of K-Ras–driven lung tumorigenesis, ectopic expression of DX2 induced small cell and non–small cell lung cancers, both of which could be suppressed by SLCB050 treatment. Taken together, our findings show how DX2 promotes lung cancer progression and how its activity may be thwarted as a strategy to treat patients with lung cancers exhibiting elevated DX2 levels. Cancer Res; 76(16); 4791–804. ©2016 AACR.

Lung cancer is one of the most common human malignancies and is associated with extremely low survival rates (<10%; refs. 1, 2). However, the molecular mechanisms underlying the development of lung cancers remain unclear. Oncogenic mutations such as K-Ras or Her2/Neu and activation of AKT signaling by altering positive and negative regulators have been suggested to promote lung cancers (3–7). In fact, amplification of Her2/Neu or oncogenic mutation in K-Ras is found in about 30% of lung cancer (8, 9), and loss of PTEN is also suggested as one of important genetic alternation (10). Given that oncogene activation induces p53-induced apoptosis or senescence via p14/ARF (11), there may be additional factors that disrupt the functional interaction between oncogenes and p14/ARF-p53 for cancer progression.

Human lung cancers are divided into two groups, small cell lung cancer (SCLC) and non–small cell lung cancer (NSCLC), that are also divided into subgroups, adenocarcinoma, squamous cell lung carcinoma, and large-cell lung cancer. Among them, SCLC occupies 20% of lung cancer, is closely linked to smoking habit, and is very aggressive (12–14). Moreover, until now, we do not develop proper anticancer drug against SCLC.

AIMP2 is previously known as p38/JTV-1 and cofactor of aminoacyl-tRNA synthetase complex (15, 16). Differentially from predicted role, housekeeping and scaffolding protein in essential enzyme complex, AIMP2 shows diverse cellular functions such as p53 activator and substrate of Parkin (16, 17). Moreover, AIMP2 knock out mouse is neonatal lethal because of defect in lung epithelial cell differentiation (15).

Recently, exon 2 skipped alternative splicing variant of AIMP2, AIMP2-DX2 (DX2) has been reported to be highly expressed in human lung cancer (17, 18) and to be induced by benzo(a)pyrene (18), a major carcinogen associated with smoking (19). Considering previous literatures, DX2 would be important factor of lung cancer initiation or progression, and how DX2 contributes to lung carcinogenesis has not been clearly demonstrated until now. That report inspired us to examine the role of DX2 in human lung cancer, in particular SCLC because of tight relationship with smoking.

In this study, we checked the expression of DX2 in several kinds of human lung cancer cell lines and revealed that it was highly expressed in SCLC and stabilized by various oncogenic signaling including K-Ras or Her2/Neu-AKT activation. Moreover, DX2 blocked the oncogene-induced p14/ARF activation. Thus, we hypothesized that inhibition of DX2 would be one of plausible candidate for lung cancer treatment, in particular, SCLC.

Cell culture

A549, HCT116, H1299, and HEK293 cell lines were obtained from the ATCC and maintained in RPMI-1640 or DMEM containing 10% FBS and 1% antibiotics. NSCLC cell lines (NCI-H23, NCI-H322, NCI-H358, and NCI-H460) were obtained from the ATCC. SCLC cell lines (NCI-H69, NCI-H128, NCI-H209, and NCI-H146) were newly purchased from Korean Cell line Bank (KCLB) for this experiment and maintained in RPMI-1640 containing 10% FBS. MEF cells were isolated from 14.5 day embryos using a standard protocol and cultured in DMEM supplemented with 15% FBS and 1% antibiotics. Cell lines were certified by short tandem repeat analysis.

Transfection of mammalian expression plasmids and siRNA

The mammalian expression plasmids encoding GFP-p14/ARF (GFP tagged at N-terminal of full-length p14/ARF) were kindly provided by Dr. G. Peters (Cancer Research UK, London Research Institute; ref. 20). The AIMP2-related mammalian expression vectors (AIMP2 and DX2) and Ras (K-Ras, N-Ras and H-Ras) were obtained by Dr. S. Kim (Seoul National University, Seoul, Republic of Korea) and by Dr. S.G. Chi (Korea University, Seoul, Republic of Korea), respectively (18, 21). pVHL and Siah-1 vectors were obtained from Dr. Y.J. Jung (Pusan National University, Busan, Republic of Korea) and from Dr. Lashuel (Qatar Foundation), respectively. Her2/Neu WT, CA (Constitutively Active; V659E), KD (kinase dead; K753M; ref. 22), and E3 ligases [Skp2 (23), CDC4 (24)] were obtained from Addgene. For in vitro gene knockdown, siRNAs against target proteins including DX2 (18), p14/ARF (25), and control (nonsilencing; ref. 26) were generated by costume service (Cosmo Genetech). Transfection was performed for 24 hours using Jet-pei (Polyplus Transfection) reagent according to the manufacturer's protocol. In brief, cells, seeded a day before, were washed with PBS and incubated with DNA/Jet-pei mixture for 4 hours within serum-free condition.

Western blot analysis

Protein was extracted from cells in RIPA buffer (150 mmol/L NaCl, 25 mmol/L Tris-Cl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS, containing protease inhibitor cocktail). After heat inactivation with sample buffer (heated at 95°C for 7 minutes), proteins were applied to SDS-PAGE and Western blot analysis, according to general protocol (27, 28). The antibodies (Ab), used for this study, are purchased from Santa Cruz Biotechnology [HA (sc-7392), His (sc-8036), GFP (sc-9996), GST (sc-138), Actin (sc-1616), and p19/ARF (sc-32748)] or Millipore [Anti-p14/ARF (MAB3782)] or Sigma Aldrich [FLAG-M2 (F3165) and C-Myc (M5546)]. Anti-AIMP2 was kindly provided from Dr. S. Kim (Seoul National University).

Mice

All experimental procedures using laboratory animals were approved by the animal care committee of Pusan National University. DX2 (C57/BL6) and K-RasLA2 (C57/BL6) mice were obtained from Dr. S. Kim (Seoul National University) and Dr. K. Choi (Yonsei University, Seoul, Republic of Korea), respectively, and double Tg mice were generated by crossing breeding of DX2 and K-RasLA2 mice. Before experiment, all mice were maintained under temperature- and light-controlled conditions (20–23°C, 12 hour/12 hour light/dark cycle) and provided autoclaved food and water ad libitum.

Drug treatment in vivo

DK (5-month-old, N = 6) mice were administered with carrier, SLCB050 (5 mg/kg), Adriamycin (1 mg/kg), and combination of SLCB050 and Adriamycin by i.p. injection. After termination of the experiment of each group, mice were dissected and isolated lung tissues. For xenograft, 1 × 107 H446 cells were seeded in s.c. on nude mice. After 4 weeks, tumor-bearing mice were injected with Adriamycin (5 mg/kg or 1 mg/kg), SLCB050 (10 mg/kg), or combination for 6 weeks. Every week, tumor volume and body weight were measured.

Histologic analysis

After dissection of mice, tissues were fixed using 10% formalin in PBS for 24 hours and embedded in paraffin blocks according to a basic tissue processing procedure. For histologic analysis, embedded tissues were cut for 5 μm by Leica microtome and transferred onto adhesive-coated slides (Marienfeld laboratory glassware). After deparaffin and rehydration, sections were then stained with hematoxylin–eosin for routine examination.

For IHC staining, rehydrated tissue sections were incubated with antibodies to ki-67 (Abcam; ab15580), pan-keratin (Sigma; C2931), pro-surfactant C (Millipore; AB3786), NSE (DAKO; IS612), and Her2/Neu (DAKO; A0458). Antigen retrieval was performed using 10 mmol/L sodium citrate (pH 6.0) 2 times at 95°C for 10 minutes each, and endogenous peroxidase activity was blocked with 3% hydrogen peroxidase for 10 minutes. Then, the slides were dehydrated following a standard procedure and sealed with cover glass using mounting solution. Terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) reaction was done as described in the manual for In Situ Cell Death Detection Kit, POD (Hoffmann-La Roche Ltd.).

Chemical library

Personal synthetic chemicals and natural compounds library has been described previously (27, 28). Korea chemical bank also provided 8,000 chemicals for this study.

ELISA

To detected p14/ARF-DX2 binding inhibitor, we generated the screening system based on ELISA. We immobilized His-DX2 recombinant protein on 96-well plate with 0.5% paraformaldehyde. After drying and washing, we incubated GST-p14/ARF protein with 0.1 mmol/L chemicals (final concentration). After 1 hour, plates were washed using TBS-T and incubated with anti-GST antibody (1:10,000 for 30 minutes) and anti–mouse-IgG-horseradish peroxidase (HRP; 1:50,000 for 1 hour). After washing twice, plates were incubated with 3,3′,5,5′tetramethylbenzidine (TMB) solution (Calbiochem) and stop solution (1N H2SO4). Thereafter, using the ELISA reader, we determined the value at 450 nm. More detail information about this ELISA system could be obtained from our previous literature (28).

Recombinant proteins, immunoprecipitation, and GST pull-down assays

The human p14/ARF fragment (full-length) was ligated into the EcoRI and HindIII sites of the pGEX-TEV vector, which is a modified vector by adding a TEV protease cleavage site to pGEX-4T1 (Invitrogen). The recombinant proteins were expressed in the Escherichia coli (E. coli) strain BL21 (DE3) as a GST-fusion proteins. The proteins were purified by glutathione affinity chromatography. To address the direct binding between two proteins, agarose bead–conjugated GST (negative control) or GST-target protein was incubated with cell lysate or recombinant protein in RIPA buffer for 1 hour at 4°C. Immunoprecipitation (IP) assay was performed with cell lysate or recombinant protein with RIPA buffer. The whole lysates were incubated sequentially with appropriated antibodies for 2 hours at 4°C and then the mixtures were added protein A/G agarose beads–conjugated secondary antibody (Invitrogen) for 2 hours. After incubation, mixtures were washed using RIPA buffer 2 times. Precipitated proteins were determined by Western blot analysis.

Immunofluorescence staining

Cells grown on coated cover glasses were fixed with 100% MeOH for 1 hour at −20°C. After washing with PBS and blocking in PBS-based buffer [containing nonrelated Goat antibody (1:500)] to eliminate nonspecific reaction, cells were incubated with primary antibodies (1:50∼1:200; overnight at 4°C) and with proper FITC or Rhodamine-conjugated secondary antibodies for 4 hours ∼ overnight at room temperature. 4, 6-diamidino-2-phenylindole (DAPI) was used to stain nuclei. After washing with PBS, cover glasses were mounted with mounting solution (Vector Laboratories; H-5501) and subjected to fluorescence microscopic analysis (Zeiss).

MTT assay

To examine cell viability or proliferation, MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay was performed. Cells were incubated in 0.5 mg/mL of MTT solution for 4 hours at 37°C. After removing solution, formazan products were dissolved in dimethyl sulfoxide (DMSO) and measured by reading absorbance on a spectrophotometer at 540 nm.

Senescence-associated β-galactosidase staining

Cells grown on coverslips were stained by using a senescence-associated β-galactosidase staining kit (9860S; Cell Signaling Technology, Inc.). In brief, cells, fixed with 4% paraformaldehyde, were incubated with staining solution for overnight at 37°C without CO2. Stained cells were observed under light microscope after mounting with VectaMount solution (Vector Laboratories).

Human lung cancer tissue samples

Eight pairs of frozen human normal and lung cancer tissue samples were provided by a biobank in Soonchunhyang University Bucheon Hospital (schbc-biobank-2011-003). Part of tissues was used for extraction of protein, total RNA, and DNA, and another part was fixed with formalin for tissue analysis. The cDNA was isolated from normal (N) or tumor (T) regions in the human lung tissues of total eight cases, and then DX2 transcripts were checked using PCR with specific primers. The primer sequences were as follows: for DX2 (forward 5′-AACGTGCACGGCAGGAGCTAC-3′, reverse 5′-CCAGCTGATAGTCTTGGCGGG-3′) and for GAPDH (forward 5′-ATCTTCCAGGAGCGAGATCCC-3′, revere 5′-AGTGAGCTTCCCGTTCAGCTC-3′). Detection of oncogenic K-Ras mutation was performed by previous literature (29).

Human patient's serum analysis

Human normal, SCLC, and NSCLC patient's sera were also obtained from a biobank in Soonchunhyang University Bucheon Hospital (schbc-biobank-2011-003). To test the autoantibodies against DX2, we attached recombinant DX2, Lamin A, and Snail onto nitrocellulose membrane (0.5 ng/well). Each membrane was incubated with patient's serum, diluted with blocking buffer (1:1,000, for 1 hour), and sequentially HRP-conjugated anti-human antibody (1:20,000) for 30 minutes. Associated Abs to proteins were visualized by ECL and X-ray film exposure. More detail procedure was available in our previous literature (28).

Analysis of tumor incidence and area

To evaluate tumor incidence, lung tissues of each mouse was fixed and embedded in paraffin. Five sections from each mouse were examined by three independent investigators who counted tumor. In addition, tumor area was calculated by tumor occupied area in total lung area using Photoshop soft wear.

Statistical analysis

To obtain the statistical significance, we performed the Student t test. In addition, to evaluate the significant enhanced effect in combination treatment with SLCB050 and conventional anticancer drug, we compared the simple added effect and real MTT value.

The expression of DX2 in lung cancer tissue and cell lines

To examine the role of DX2 in lung carcinogenesis, we first measured DX2 expression in human lung cancer tissues and found that, despite strong DX2 transcription even in adjacent noncancerous tissues, DX2 protein was detected primarily in cancer tissues with K-Ras-or Her2/Neu-activation (Supplementary Fig. S1A and S1B). This result implied that DX2 expression at transcription level did not relate to lung cancer progression. We obtained similar results in human lung cancer cell line analysis (Fig. 1A). In addition, K-Ras–mutated (A549 and H460) and PTEN silent (H1299) cell lines showed the elevated DX2 expression. Indeed, activated Her2/Neu, AKT, and oncogenic K-Ras increased exogenous (Fig. 1B and C; Supplementary Fig. S1C) and endogenous DX2 expressions without altering DX2 transcription (Fig. 1D). In addition, stabilized DX2 by oncogenes was translocated into nucleus (Fig. 1E). These results implied that EGFR-AKT-K-Ras signaling cascade regulates DX2 expression and localization at posttranscriptional level. So, we next checked the DX2 protein stability and turnover rate. Rapid turnover of DX2 within 1 hour (Fig. 1F) was extended up to 3 hours by oncogenic K-Ras (Fig. 1G) or phosphatase inhibitor (Supplementary Fig. S1D). To reveal the responsible E3 ligase, we examined the effect of several E3 ligase on AIMP2 and DX2 and found that Siah-1 eliminated DX2 (Fig. 1H). Elimination of Siah-1 could increase DX2 expression (Supplementary Fig. S1E), and Siah-1 showed the binding ability to DX2 (Supplementary Fig. S1F). Indeed, Siah-1 could promote DX2-Ubiquitinylation (Supplementary Fig. S1G). Because Siah-1 is transcriptional target of p53 (30), DX2 would be regulated by p53 pathway negatively and by oncogenic signaling positively.

Figure 1.

Oncogenic stimulations induce DX2 expression at posttranslation level. A, differential expression of DX2 in human lung cancer cell lines. H322 cells had extremely low DX2 protein expression (red arrow) despite a transcription level (red arrow) that was similar to those in the other cell lines. In contrast, A549 and H460 cells had high DX2 protein expression. AIMP2 is indicated by the black arrows. Actin and GAPDH were used as loading controls for transcription and protein expression, respectively. B, Her2/Neu increases DX2 expression. Wild-type (WT) and constitutively active (CA) Her2/Neu but not KD Her2/Neu induced DX2. In contrast, AIMP2 was not induced by Her2/Neu. HEK293 cells were transfected with the indicated vectors for 24 hours. C, active AKT increases DX2 expression. HEK293 cells were cotransfected with AIMP2 or DX2 and AKT expression vectors for 24 hours. AKT-CA (myristoylated AKT) selectively induced DX2 expression. D, oncogenes induce endogenous DX2 in H1299. Transfection with AKT-CA (myristoylated AKT), Her2/Neu-CA, and oncogenic K-Ras into H1299 for 24 hours could increase endogenous DX2. E, oncogenes promote nuclear localization of DX2. Although transfected DX2 was located in the cytoplasm in HEK293 cells, AKT-CA and Her2/Neu-CA promoted the translocation of DX2 into the nucleus. In contrast, AKT-KD and Her2/Neu-KD suppressed DX2 expression. F, rapid turnover of DX2. Compared with AIMP2, DX2 protein showed very short half-life (less than 1 hour). Transfected cells were incubated with cycloheximide (CHX; 100 μg/mL) for indicated times. G, oncogenic K-Ras extends the half-life of DX2. DX2-transfected HEK293 cells were cotransfected with the indicated Ras proteins. After 6-hour incubation with MG132 (10 μmol/L) to block protein degradation, cells were incubated with cycloheximide for the indicated times to block de novo protein synthesis. However, AIMP2 did not affect its half-life by Ras transfection. H, Siah-1 inhibits DX2 expression. Compared with other E3 ligases, Siah-1 selectively inhibited DX2 expression. However, AIMP2 was not affected by E3 ligases.

Figure 1.

Oncogenic stimulations induce DX2 expression at posttranslation level. A, differential expression of DX2 in human lung cancer cell lines. H322 cells had extremely low DX2 protein expression (red arrow) despite a transcription level (red arrow) that was similar to those in the other cell lines. In contrast, A549 and H460 cells had high DX2 protein expression. AIMP2 is indicated by the black arrows. Actin and GAPDH were used as loading controls for transcription and protein expression, respectively. B, Her2/Neu increases DX2 expression. Wild-type (WT) and constitutively active (CA) Her2/Neu but not KD Her2/Neu induced DX2. In contrast, AIMP2 was not induced by Her2/Neu. HEK293 cells were transfected with the indicated vectors for 24 hours. C, active AKT increases DX2 expression. HEK293 cells were cotransfected with AIMP2 or DX2 and AKT expression vectors for 24 hours. AKT-CA (myristoylated AKT) selectively induced DX2 expression. D, oncogenes induce endogenous DX2 in H1299. Transfection with AKT-CA (myristoylated AKT), Her2/Neu-CA, and oncogenic K-Ras into H1299 for 24 hours could increase endogenous DX2. E, oncogenes promote nuclear localization of DX2. Although transfected DX2 was located in the cytoplasm in HEK293 cells, AKT-CA and Her2/Neu-CA promoted the translocation of DX2 into the nucleus. In contrast, AKT-KD and Her2/Neu-KD suppressed DX2 expression. F, rapid turnover of DX2. Compared with AIMP2, DX2 protein showed very short half-life (less than 1 hour). Transfected cells were incubated with cycloheximide (CHX; 100 μg/mL) for indicated times. G, oncogenic K-Ras extends the half-life of DX2. DX2-transfected HEK293 cells were cotransfected with the indicated Ras proteins. After 6-hour incubation with MG132 (10 μmol/L) to block protein degradation, cells were incubated with cycloheximide for the indicated times to block de novo protein synthesis. However, AIMP2 did not affect its half-life by Ras transfection. H, Siah-1 inhibits DX2 expression. Compared with other E3 ligases, Siah-1 selectively inhibited DX2 expression. However, AIMP2 was not affected by E3 ligases.

Close modal

DX2 provided drug resistance

Because DX2 was stabilized by oncogenes, we speculated that it would be related with cell growth and proliferation. However, in MEF, obtained from DX2 transgenic and K-RasLA2/DX2 double transgenic (DK) mice, DX2 did not show obvious effect on cell proliferation, comparing to oncogenic K-Ras (Supplementary Fig. S2A). Instead, DX2 expressed MEF showed the resistance to Adriamycin- or serum starvation–induced cell death and senescence (Fig. 2A and Supplementary Fig. S2B). Indeed, DK MEF cells were completely resistant to selective p53 activator (GN25; ref. 28) on K-Ras–activated cells (Fig. 2B). To get more clues for pathologic role of DX2 on cell survival, we measured cell viability after inhibition of DX2. We found that transfection of siRNA for DX2 (si-DX2) suppressed cell survival in a p14/ARF-dependent manner (Fig. 2C; refs. 31, 32). Because oncogenic stress can trigger p53 activation through p14/ARF (11), we hypothesized that the oncogenic property of DX2 is related to p14/ARF. Indeed, p14/ARF seemed to determine the localization of DX2. In p14/ARF-null cell, overexpressed DX2 was not located in nucleus (Fig. 2D), even in the same K-Ras–mutated cancer cell line, HCT116, although DX2 was able to localize in the nucleus or cytoplasm (Supplementary Fig. S2C and S2D). Moreover, DX2, but not AIMP2, could block the Her2/Neu-induced p14/ARF (Fig. 2E).

Figure 2.

DX2 provides drug resistance. A, resistance to senescence in DX2 and DK MEF. MEF cells were cultivated in serum-free condition (SF) or in low dose of Adr-treated condition for 48 hours to induce senescence. Compared with wild-type (WT) and K-Ras MEF, which were stained by SA-β-gal and showed the typical senescent morphology, DX2 and DK cells did not show the obvious response. B, resistance of DX2 and DX2/K-RasLA2 double transgenic (DK) cells to p53-activating chemical. MEF cells were incubated with 5 μmol/L of GN25, activator of p53 for 6 hours. The obvious reduction of cell viability by GN25 in K-Ras MEF was completely blocked in DK cells. C, Si-DX2 suppresses cell viability in a p14/ARF-dependent manner. Elimination of DX2 using si-DX2 (0, 1, 5, or 10 μg/mL) reduced the viability of H1299, H23, and H69 cells, which are p14/ARF positive. In contrast, p14/ARF-null cell lines were resistant to si-DX2. NULL and MT indicate homozygote deletion and mutant, respectively. D, differential localization of DX2 in lung cancer cell lines. Nuclear DX2 was detected in H1299 cells. The localization of AIMP2 did not differ between A549 and H1299 cells. E, DX2 blocks oncogene-induced p14/ARF. Increase of p14/ARF by Her2/Neu was abolished by DX2 transfection. HEK293 cells were transfected with indicated vectors for 24 hours.

Figure 2.

DX2 provides drug resistance. A, resistance to senescence in DX2 and DK MEF. MEF cells were cultivated in serum-free condition (SF) or in low dose of Adr-treated condition for 48 hours to induce senescence. Compared with wild-type (WT) and K-Ras MEF, which were stained by SA-β-gal and showed the typical senescent morphology, DX2 and DK cells did not show the obvious response. B, resistance of DX2 and DX2/K-RasLA2 double transgenic (DK) cells to p53-activating chemical. MEF cells were incubated with 5 μmol/L of GN25, activator of p53 for 6 hours. The obvious reduction of cell viability by GN25 in K-Ras MEF was completely blocked in DK cells. C, Si-DX2 suppresses cell viability in a p14/ARF-dependent manner. Elimination of DX2 using si-DX2 (0, 1, 5, or 10 μg/mL) reduced the viability of H1299, H23, and H69 cells, which are p14/ARF positive. In contrast, p14/ARF-null cell lines were resistant to si-DX2. NULL and MT indicate homozygote deletion and mutant, respectively. D, differential localization of DX2 in lung cancer cell lines. Nuclear DX2 was detected in H1299 cells. The localization of AIMP2 did not differ between A549 and H1299 cells. E, DX2 blocks oncogene-induced p14/ARF. Increase of p14/ARF by Her2/Neu was abolished by DX2 transfection. HEK293 cells were transfected with indicated vectors for 24 hours.

Close modal

DX2 contributed to SCLC occurrence

Next, we examined the in vivo oncogenic effect of DX2 using transgenic mouse model (18). Differentially from AIMP2 knockout mice, which are neonatal lethal due to defect of lung epithelial cell differentiation (15), DX2 transgenic mice are viable, suggesting that DX2 overexpression did not affect lung cell differentiation. Although DX2 transgenic mice did not produce tumor at 4 months old (Fig. 3A and B), comparing to K-RasLA2, DK mice bore larger tumor (Fig. 3B and Supplementary Fig. S3A) with low apoptotic cells (Fig. 3C). In addition, besides commonly detected typical adenocarcinomas in both K-RasLA2 and DK mice, small nuclear atypical tumors were detected between blood vessels and airways in DK mice (Fig. 3B and Supplementary Fig. S3A). To identify these tumors, we performed the IHC staining with pan-keratin (for squamous cell cancer; ref. 33), pro-surfactant protein C (for adenocarcinoma; ref. 33), and neuron-specific enolase (NSE; for SCLC; refs. 34, 35). NSE staining was visible in the small cell tumor region in DK-transgenic mice (Supplementary Fig. S3B). We also observed NSE-positive SCLC cells in tumors of DX2-transgenic mice (6-month-old), but not in K-RasLA2 mice (Fig. S3C). Small cell cancer mass was increased following age in DX2 transgenic mice (Supplementary Fig. S3D and S3E). Indeed, the tumor cells in DX2 mice had a high mitotic index and fine granular chromatin (Supplementary Fig. S3F), which are well-known cellular markers of SCLC (13). We also observed the NSE expression in primary tumor cells isolated from DX2 and DK tumors (Fig. 3D). On the basis of histologic features (Supplementary Fig. S3G), we analyzed tumor types in mouse models and found that DX2 and DK mice could promote SCLC occurrence (Fig. 3E). In fact, DX2 expression was elevated in SCLC cell lines at translation level (Fig. 3F and G). Because SCLC tissues are not available, we could not directly detect the DX2 expression in human cases. Instead, because DX2 is abnormal gene product and many DX2-positive cancer cells would be died by necrosis, we measured the DX2 autoantibody in human patient's sera. Although it is indirect method, we already tested this method for detection of cancer-specific protein. Indeed, DX2-specific autoantibody was frequently detected in SCLC sera (9/10 cases) and moderately in NSCLC (10/20 cases), whereas it was rarely produced in normal sera (1/10; Fig. 3H; Supplementary Fig. S3H; Supplementary Table S1). These results suggest that DX2 is closely related with lung cancer occurrence and progression, in particular SCLC.

Figure 3.

DX2 promotes SCLC production. A, tumor incidence in 4-month-old mice. Tumor area was calculated from five sections per sample. B, hematoxylin–eosin staining of lung tissues from the indicated transgenic mice (4-month-old). DX2/K-RasLA2 double transgenic (DK) mice had two kinds of tumors. C, DX2 induced resistance to apoptosis in contrast to numerous TUNEL-positive cells in tumor tissues of K-RasLA2 mice; tumor cells from DK mice were barely stained by TUNEL. D, immunofluorescence staining with NSE. To confirm that the tumors in DX2 and DK mice were SCLC tumors, tumor cells were isolated from lungs and cultured for 2 weeks to eliminate primary cells such as lymphocytes. Despite serial cultivation, small cells grew well. After fixation with MeOH, cells were stained with NSE. Normal lung epithelial cells were not stained by NSE and had large nuclei. In contrast, small cells from DX2 or DK mouse tumors were strongly stained by NSE. Small cells were clearly distinguishable from normal epithelial cell (white arrows). E, Ban diagram of lung tumor spectrum in each mouse model. Lung biopsy was performed with mice that showed the pathologic symptoms such as dyspnea. Thus, average age (AveAge) of each group was different and represents approximate onset time. Note that this result contained 4-month-old analysis (Fig. 4A), and some of DX2 single mice were scarified without symptoms. SCLC was always detected with other kinds of cancers. F, DX2 expression in SCLC cell lines. Two SCLC cell lines had strong DX2 expression like K-Ras–mutated H460 (red arrowhead). However, AIMP2 expression levels did not differ between NSCLC and SCLC cells. G, analysis of DX2 and AIMP2 transcription, with GAPDH as a control. H, detection of autoantibody against DX2 in sera from SCLC patients. Immobilized DX2 was recognized by antibody in 9 of 10 SCLC sera and 8 of 20 NSCLC sera (indicated in red) but not in sera of healthy individuals. Lamin A and Snail were used as positive controls.

Figure 3.

DX2 promotes SCLC production. A, tumor incidence in 4-month-old mice. Tumor area was calculated from five sections per sample. B, hematoxylin–eosin staining of lung tissues from the indicated transgenic mice (4-month-old). DX2/K-RasLA2 double transgenic (DK) mice had two kinds of tumors. C, DX2 induced resistance to apoptosis in contrast to numerous TUNEL-positive cells in tumor tissues of K-RasLA2 mice; tumor cells from DK mice were barely stained by TUNEL. D, immunofluorescence staining with NSE. To confirm that the tumors in DX2 and DK mice were SCLC tumors, tumor cells were isolated from lungs and cultured for 2 weeks to eliminate primary cells such as lymphocytes. Despite serial cultivation, small cells grew well. After fixation with MeOH, cells were stained with NSE. Normal lung epithelial cells were not stained by NSE and had large nuclei. In contrast, small cells from DX2 or DK mouse tumors were strongly stained by NSE. Small cells were clearly distinguishable from normal epithelial cell (white arrows). E, Ban diagram of lung tumor spectrum in each mouse model. Lung biopsy was performed with mice that showed the pathologic symptoms such as dyspnea. Thus, average age (AveAge) of each group was different and represents approximate onset time. Note that this result contained 4-month-old analysis (Fig. 4A), and some of DX2 single mice were scarified without symptoms. SCLC was always detected with other kinds of cancers. F, DX2 expression in SCLC cell lines. Two SCLC cell lines had strong DX2 expression like K-Ras–mutated H460 (red arrowhead). However, AIMP2 expression levels did not differ between NSCLC and SCLC cells. G, analysis of DX2 and AIMP2 transcription, with GAPDH as a control. H, detection of autoantibody against DX2 in sera from SCLC patients. Immobilized DX2 was recognized by antibody in 9 of 10 SCLC sera and 8 of 20 NSCLC sera (indicated in red) but not in sera of healthy individuals. Lamin A and Snail were used as positive controls.

Close modal

DX2 suppresses p14/ARF

To know the biological role of DX2, we tested the oncogene-induced p14/ARF. Indeed, Her2/Neu-AKT-K-Ras signaling cascade can induce p14/ARF and promote oncogene-induced apoptosis and senescence, whereas DX2 provided the resistance to senescence (Fig. 2A) and GN25-induced cell death in K-Ras–mutated MEF (Fig. 2B). Moreover, stabilized DX2 by oncogenes was located in nucleus as p14/ARF-dependent manner (Figs. 1G and 2D). First, we evaluated the expression of p19/ARF, mouse p14/ARF (36, 37), in DX2-expressed MEF and found that it was obviously reduced in DX2-expressed MEF (Fig. 4A). Similarly, elimination of DX2 could increase endogenous p14/ARF in several kinds of human cancer cell lines including H1299 (Fig. 4B and Supplementary Fig. S4A) as well as exogenous p14/ARF in A549 (Fig. 4C and Supplementary Fig. S4B). However, si-DX2 did not increase p14/ARF expression in DX2-negative cell line (H322; Fig. 4C) and nontransformed cell lines (HEK293 and L132; Supplementary Fig. S4A). Considering the fact that si-DX2 could induce exogenous p14/ARF, it would be achieved at posttranslational level. In fact, si-DX2 could obviously increase p14/ARF stability (Fig. 4D) and promote nucleoplasmic location (Fig. 4E). In addition, DX2 promoted p14/ARF ubiquitination (Fig. 4F), and MG132, proteasome inhibitor, blocked the DX2-mediated p14/ARF reduction (Fig. 4G).

Figure 4.

DX2 suppresses p14/ARF expression. A, reduction of p19/ARF in DX2 and DK MEF. p19/ARF expression was reduced to undetectable levels in DX2 and DK MEF. B, si-DX2 induces p14/ARF expression in lung cancer cell line H1299. Cells were transfected with indicated concentration of si-DX2 for 24 hours. Induction of p14/ARF was detected following dosage of si-DX2. C, si-DX2 increases ectopic expression of p14/ARF. In p14/ARF-transfected A549 and H322 cells, si-DX2 induced exogenous p14/ARF expression. D, elimination of DX2 increases p14/ARF stability. H1299, transfected with p14/ARF or DX2, was cotransfected with si-DX2 for 24 hours. Protein half-life was measured after treatment of cycloheximide (CHX). E, increase of p14/ARF in nucleoplasm by si-DX2. In GFP-p14/ARF-transfected A549, elimination of DX2 could increase p14/ARF in nucleoplasm. F, DX2 promotes the ubiquitylation of p14/ARF. HEK293 cells were transfected with the indicated vectors for 24 hours. After incubation with MG132, cell lysates were analyzed by IP with p14/ARF antibody. Ubiquitin (Ub)-conjugated p14/ARF was detected at high molecular weight (arrowheads). G, the proteasome inhibitor MG132 (10 μmol/L, 6 hours) reverses DX2-induced p14/ARF suppression and inhibits DX2.

Figure 4.

DX2 suppresses p14/ARF expression. A, reduction of p19/ARF in DX2 and DK MEF. p19/ARF expression was reduced to undetectable levels in DX2 and DK MEF. B, si-DX2 induces p14/ARF expression in lung cancer cell line H1299. Cells were transfected with indicated concentration of si-DX2 for 24 hours. Induction of p14/ARF was detected following dosage of si-DX2. C, si-DX2 increases ectopic expression of p14/ARF. In p14/ARF-transfected A549 and H322 cells, si-DX2 induced exogenous p14/ARF expression. D, elimination of DX2 increases p14/ARF stability. H1299, transfected with p14/ARF or DX2, was cotransfected with si-DX2 for 24 hours. Protein half-life was measured after treatment of cycloheximide (CHX). E, increase of p14/ARF in nucleoplasm by si-DX2. In GFP-p14/ARF-transfected A549, elimination of DX2 could increase p14/ARF in nucleoplasm. F, DX2 promotes the ubiquitylation of p14/ARF. HEK293 cells were transfected with the indicated vectors for 24 hours. After incubation with MG132, cell lysates were analyzed by IP with p14/ARF antibody. Ubiquitin (Ub)-conjugated p14/ARF was detected at high molecular weight (arrowheads). G, the proteasome inhibitor MG132 (10 μmol/L, 6 hours) reverses DX2-induced p14/ARF suppression and inhibits DX2.

Close modal

Direct interaction between DX2 and p14/ARF

Our next question is how DX2 suppresses p14/ARF. To address this, we first checked the interaction of them. Cotransfected DX2 and p14/ARF were recovered in nucleolus (Fig. 5A and Supplementary Fig. S4C). We could observe the physiologic interaction of DX2 and p14/ARF through bidirectional IP analysis (Fig. 5B and Supplementary Fig. S4D). Direct interaction between DX2 and p14/ARF was observed by in vitro GST pull-down assay using His-DX2– or AIMP2-recombinant protein (Fig. 5C). Interesting feature is that AIMP2 was coprecipitated with GST-p14 when DX2 was coexisted, indicating that DX2 could bind to AIMP2. p53-GST was used for positive control (18). To avoid the nonspecific interaction, we also checked the binding of DX2 and GST-Smad4 or von Hippel-Lindau tumor suppressor (VHL). However, these proteins did not pull down DX2 (Supplementary Fig. S4E and S4F). Instead, AIMP2 was precipitated with GST proteins, implying that AIMP2 would be sticky or multifunctional protein. Actually, AIMP2 is a cofactor of protein complex (16). These results indicate that DX2-p14/ARF interaction is specific event. To confirm the direct interaction, we re-executed GST pull-down assay using recombinant DX2 and obtained the same result (Fig. 5D). Because p14/ARF is very sticky protein (38), we did not exclude the nonspecific association of DX2 and p14/ARF. So, we re-performed IP analysis with His-AIMP2 or DX2. In this assay, we observed the selective association of DX2 and N-terminal p14/ARF (Fig. 5E and F).

Figure 5.

Direct interaction of p14/ARF and DX2. A, colocalization of p14/ARF and DX2 in nucleolus-like structures. HEK293 cells were transfected with GFP-p14/ARF (top) or cotransfected with GFP-p14/ARF and DX2 (bottom) for 24 hours. Cells were incubated with anti-Myc antibody and Rhodamine-conjugated secondary antibody (red). GFP-p14/ARF in the nucleoplasm was reduced by DX2 transfection. B, IP analysis using His-DX2. Cells, transfected with indicated vectors, were incubated with His-DX2 recombinant protein. After IP with His Ab, coprecipitated materials were analyzed by Western blot analysis. Specific interaction between p14/ARF and DX2 was observed. Sup, supernatant after IP analysis. Despite strong expression, GFP-nucleolin did not show the interaction with DX2, indicating that DX2-p14/ARF binding was a specific event. C, specific interaction of p14/ARF with DX2 but not with AIMP2. A GST pull-down assay was performed using bead-GST-p14/ARF and His-DX2 or His-AIMP2. A pull-down assay with GST-p53 was performed as a control experiment. D,in vitro binding assay. Recombinant His-DX2 comigrated with bead-conjugated recombinant GST-p14/ARF. E, the N-terminal region of p14/ARF serves as the binding domain for DX2. Full-length p14/ARF (p14/ARF-F; AA 2–132) and the N-terminal p14-ARF fragment (p14-N; AA 2–29) were tested in a binding assay. His-DX2 was strongly associated with p14/ARF-N. F, p14/ARF binding region. A region of DX2 generated by the joining of exons 1 and 3 binds to the N-terminal region of p14/ARF.

Figure 5.

Direct interaction of p14/ARF and DX2. A, colocalization of p14/ARF and DX2 in nucleolus-like structures. HEK293 cells were transfected with GFP-p14/ARF (top) or cotransfected with GFP-p14/ARF and DX2 (bottom) for 24 hours. Cells were incubated with anti-Myc antibody and Rhodamine-conjugated secondary antibody (red). GFP-p14/ARF in the nucleoplasm was reduced by DX2 transfection. B, IP analysis using His-DX2. Cells, transfected with indicated vectors, were incubated with His-DX2 recombinant protein. After IP with His Ab, coprecipitated materials were analyzed by Western blot analysis. Specific interaction between p14/ARF and DX2 was observed. Sup, supernatant after IP analysis. Despite strong expression, GFP-nucleolin did not show the interaction with DX2, indicating that DX2-p14/ARF binding was a specific event. C, specific interaction of p14/ARF with DX2 but not with AIMP2. A GST pull-down assay was performed using bead-GST-p14/ARF and His-DX2 or His-AIMP2. A pull-down assay with GST-p53 was performed as a control experiment. D,in vitro binding assay. Recombinant His-DX2 comigrated with bead-conjugated recombinant GST-p14/ARF. E, the N-terminal region of p14/ARF serves as the binding domain for DX2. Full-length p14/ARF (p14/ARF-F; AA 2–132) and the N-terminal p14-ARF fragment (p14-N; AA 2–29) were tested in a binding assay. His-DX2 was strongly associated with p14/ARF-N. F, p14/ARF binding region. A region of DX2 generated by the joining of exons 1 and 3 binds to the N-terminal region of p14/ARF.

Close modal

Isolation of specific binding inhibitor of DX2 and p14/ARF by chemical screening

Because DX2 blocks the oncogene-induced p14/ARF induction (Fig. 2E), we speculated that specific inhibitor against DX2 and p14/ARF binding would be useful for anticancer drug. So, we performed the chemical screening (Supplementary Fig. S5A) through modified ELISA system (Supplementary Fig. S5B). After testing of ELISA system (Fig. 6A), we screened chemical library (Supplementary Fig. S5C). Because we have screened the same library using other protein binding inhibitor (Pak1-PUMA binding), we excluded the commonly isolated chemicals from candidate (Supplementary Fig. S5A and S5D). Next, we directly checked specific inhibition through GST pull-down assay using DX2-p14/ARF and p53-p14/ARF (Supplementary Fig. S5E and S5F). Finally, we obtained five chemicals that were selectively inhibited DX2-p14/ARF. Because we did not obtain enough amounts of four chemicals for further experiment instantly, we went to next step with SLCB050 (Fig. 6B) that could block the interaction of DX2 and p14/ARF more obviously than SLC24, random protein binding inhibitor (Supplementary Fig. S5D and S5G). Indeed, SLCB050 inhibited DX2-p14/ARF binding in in vitro GST pull-down assay (Fig. 6C and Supplementary Fig. S6A) as well as IP experiment (Fig. 6D) in a dose-dependent manner (Supplementary Fig. S6B). SLCB050 also blocked the interaction of DX2 and AIMP2 (Fig. 6E), but not on p53-AIMP2 or DX2 binding (Supplementary Fig. S6C) and p14/ARF (Supplementary Fig. S6D), indicating that SLCB050 would be interacted with DX2-specific region. To verify that SLCB050 is a specific inhibitor of DX2-p14/ARF, we performed the GST pull-down assay with SLCB050 and its very similar chemicals (HJH141204, HJH141206; stereoisomer) and related chemical (SLCB036: ribose ring is replaced by modified benzene ring; Supplementary Fig. S6E). We also checked the newly synthesized SLCB050 to exclude the synthetic error. Obvious inhibition effect on p14/ARF-DX2 binding was observed in SLCB050, HJH141204, and HJH141206 (Supplementary Fig. S6F). However, these chemicals did not alter the interaction of p53-p14/ARF (Supplementary Fig. S6G), and SLCB036 did not show the inhibitory effect on both binding (Supplementary Fig. S6F and S6G). These results indicate that unique chemical structure (in particular, ribose ring) is required for the binding inhibition.

Figure 6.

Isolation of SLCB050 as an inhibitor of DX2-p14/ARF binding. A, ELISA-based screening method. To confirm the validity of the ELISA system, we measured increases in ELISA values with increasing concentrations (Conc) of p14/ARF (left). One chemical, SLCB050, reduced the ELISA reaction considerably (arrow in the right panel). B, chemical structure of SLCB050. C, inhibition of the binding of p14/ARF to DX2. A GST pull-down assay showed complete blocking of DX2-p14/ARF binding by SLCB050. His-DX2 (red arrow) was detected in a higher molecular weight range than AIMP2 (black arrow) was, because His-DX2 recombinant protein was fused with thioredoxin. D, SLCB050 disrupts the interaction of p14/ARF and DX2 in cells. Cotransfected HEK293 cells were incubated with MG132 before IP analysis to prevent the reduction of both proteins. After incubation with SLCB050 (10 μmol/L, 6 hours), cells were used for IP analysis with Myc antibody (DX2). IgGH, immunoglobulin G heavy chain. E, inhibitory effect of SLCB050 on the binding between DX2 and AIMP2. In protein binding assay using His-DX2 protein, binding between DX2 and AIMP2 was reduced by SLCB050 treatment. F, dissociation of DX2 and p14/ARF. Through IP analysis, dissociation of DX2 and p14/ARF was detected. HEK293 cells were transfected with indicated vectors for 24 hours and incubated with additional 6 hours with MG132 and SLCB050. G, differences in localization of DX2 and p14/ARF after SLCB050 treatment. Colocalization of DX2 and p14/ARF (top panels) was disrupted by SLCB050 (10 μmol/L, 6 hours). DX2 was decreased and p14/ARF was increased in the nucleoplasm. H, SLCB050 treatment increases p14/ARF in the NSCLC cell line H1299 and the SCLC cell line H69. SLCB050 treatment at the indicated concentrations for 6 hours reduced DX2 and increased p14/ARF. I, dose-dependent reduction of DX2. DX2 levels decreased considerably in the p14/ARF-null H322 cells in response to SLCB050 treatment (6 hours). J, effect of SLCB050 on viability of human lung cancer cells incubated with the indicated concentrations of SLCB050 for 24 hours. Cell viability was determined by MTT assay. SCLC cells were very sensitive to SLCB050. K, soft-agar colony formation assay. H128 cells (SCLC cell line) were seeded in soft-agar plates and incubated with the indicated concentrations of SLCB050 for 48 hours. Cells were visualized by trypan blue staining after fixation. Colonies were counted (bottom).

Figure 6.

Isolation of SLCB050 as an inhibitor of DX2-p14/ARF binding. A, ELISA-based screening method. To confirm the validity of the ELISA system, we measured increases in ELISA values with increasing concentrations (Conc) of p14/ARF (left). One chemical, SLCB050, reduced the ELISA reaction considerably (arrow in the right panel). B, chemical structure of SLCB050. C, inhibition of the binding of p14/ARF to DX2. A GST pull-down assay showed complete blocking of DX2-p14/ARF binding by SLCB050. His-DX2 (red arrow) was detected in a higher molecular weight range than AIMP2 (black arrow) was, because His-DX2 recombinant protein was fused with thioredoxin. D, SLCB050 disrupts the interaction of p14/ARF and DX2 in cells. Cotransfected HEK293 cells were incubated with MG132 before IP analysis to prevent the reduction of both proteins. After incubation with SLCB050 (10 μmol/L, 6 hours), cells were used for IP analysis with Myc antibody (DX2). IgGH, immunoglobulin G heavy chain. E, inhibitory effect of SLCB050 on the binding between DX2 and AIMP2. In protein binding assay using His-DX2 protein, binding between DX2 and AIMP2 was reduced by SLCB050 treatment. F, dissociation of DX2 and p14/ARF. Through IP analysis, dissociation of DX2 and p14/ARF was detected. HEK293 cells were transfected with indicated vectors for 24 hours and incubated with additional 6 hours with MG132 and SLCB050. G, differences in localization of DX2 and p14/ARF after SLCB050 treatment. Colocalization of DX2 and p14/ARF (top panels) was disrupted by SLCB050 (10 μmol/L, 6 hours). DX2 was decreased and p14/ARF was increased in the nucleoplasm. H, SLCB050 treatment increases p14/ARF in the NSCLC cell line H1299 and the SCLC cell line H69. SLCB050 treatment at the indicated concentrations for 6 hours reduced DX2 and increased p14/ARF. I, dose-dependent reduction of DX2. DX2 levels decreased considerably in the p14/ARF-null H322 cells in response to SLCB050 treatment (6 hours). J, effect of SLCB050 on viability of human lung cancer cells incubated with the indicated concentrations of SLCB050 for 24 hours. Cell viability was determined by MTT assay. SCLC cells were very sensitive to SLCB050. K, soft-agar colony formation assay. H128 cells (SCLC cell line) were seeded in soft-agar plates and incubated with the indicated concentrations of SLCB050 for 48 hours. Cells were visualized by trypan blue staining after fixation. Colonies were counted (bottom).

Close modal

Antitumor effect of the chemical in small-cell lung cancer cell lines

So, we next examined the effect of SLCB050 in cell system. Treatment of SLCB050 could block the interaction (Fig. 6F) and colocalization of DX2 and p14/ARF in nucleus (Fig. 6G). Indeed, this chemical could suppress DX2 expression itself through rapid degradation (Supplementary Fig. S7A) and relocalization from nucleolus to cytosol (Supplementary Fig. S7B). Thus, we could observe the increase of p14/ARF in SLCB050-treated H1299, H69 (Fig. 6H and Supplementary Fig. S7C), as well as reduction of DX2 in p14/ARF-deficient cell lines (H322, H460, and A549; Fig. 6I; Supplementary Fig. S7C). We next measured the cell viability in several lung cancer cell lines and found that p14/ARF-deficient cell lines were resistant to SLCB050, whereas SCLC cell lines were sensitive to it (Fig. 6J). In fact, SLCB050 completely suppressed the H128 cell growth (Fig. 6K and Supplementary Fig. S7D). Moreover, SLCB050 derivatives, HJH141204 and 1206, but not SLCB036, also suppressed cell viability (Supplementary Fig. S7E), indicating that tumor cell growth suppression is achieved by chemical-DX2 binding.

The combinational effect of SLCB050 with antitumor drugs

We previously showed that DX2 provided the drug resistance (Fig. 2B; Supplementary Fig. S2A and S2B), and others reported that p14/ARF is critical for drug sensitivity (39). So, we tested that SLCB050 could restore the sensitivity to anticancer drug. Resistance to GN25 in DX2 and DK MEF was abolished by cotreatment of SLCB050 (Fig. 7A). We also observed the significant enhanced effect of SLCB050 with Adr in DX2 and DK MEF (Fig. 7B) and SCLC cell line H69 that was partially responded to SLCB050 (Fig. 7C). However, p14/ARF-deficient H322 did not show enhanced response to combinational treatment (Fig. 7C). To extend this, we generate tumor xenograft model using H446 that was resistant to Adr and partially responded to SLCB050 (Supplementary Fig. S7F). Exponentially increased tumor volume was moderately suppressed by SLCB050 injection (10 mg/kg, 3 times/week; Fig. 7D and Supplementary Fig. S7G). Although we also observed the tumor repression effect by injection of 5 mg/kg Adr (3 times/week), it evoked rapid weight loss and death (Fig. 7D and Supplementary Fig. S7H). In contrast, combinational treatment showed more obvious antitumor effect despite low dosage of Adr (Fig. 7D). Indeed, combinational treatment of Adr and SLCB050 could obviously induce p53 expression in primary tumor cells obtained from DX2 or DK mouse (Fig. 7E). Next, we tested the effect of SLCB050 on nontransformed lung cell (L132). However, this chemical did not alter the cell viability or drug sensitivity on these cell lines (Supplementary Fig. S7I). Next, we checked the effect of combination treatment of SLCB050 with other commercial anticancer drugs such as Taxol and Carboplatin. However, these chemicals did not show significant enhanced effect with SLCB050 (Supplementary Fig. S7J). These results indicate that inhibition of DX2 could enhance drug sensitivity through reactivation of p14/ARF in DX2-overexpressed cancer cells.

Figure 7.

Antitumoral effect of SLCB050. A, significant enhanced effect of SLCB050 with GN25, p53 activator in DX2 and DK MEFs. Resistance to GN25 in DX2 and DK cells was abolished by SLCB050 treatment. Cells were incubated with 5 μmol/L of GN25 and SLCB050 for 6 hours. The viability was determined by MTT assay. ##, significant enhanced effect by combination treatment. B, significant enhanced effect of Adr and SLCB050 on viability of primary tumor cells from DX2 and DK mice. Cell viability was determined by MTT assay after the cells were incubated with 0.2 μg/mL of Adr and/or 10 μmol/L SLCB050 for 24 hours. ##, significant enhanced effect by combination treatment. C, sensitivity of SCLC cell lines to combination treatment with Adr and SLCB050, compared with the sensitivity of the p14/ARF-null H322 cells. Only H69 showed the significant enhanced effect by combination treatment with SLCB050 and Adr (##). D, tumor-suppressive effect of SLCB050. In xenograft model using SCLC H446, 10 mg/kg of SLCB050 suppressed tumor growth. Although Adr (5 mg/kg) blocked the tumor growth, within 4 weeks, all of them were ceased by high toxicity (*). In contrast, low dose of Adr (1 mg/kg) with SLCB050 obviously suppressed tumor growth without severe toxicity. Chemicals were delivered via i.p injection 3 times per week. E, in primary tumor cells, combinational treatment could induce p53 synergistically. p53 expression was determined by Western blotting. F, hematoxylin–eosin staining of lung tumor tissues from SLCB050-treated DK mice. Combination treatment with SLCB050 and Adr induced tumor regression. G, reduction of tumor volume by combination treatment. Compared with tumor areas in age-matched untreated (DMSO) or single chemical-treated DK mice (SLCB050: 5 mg/kg and Adr: 1 mg/kg), tumor areas in mice treated with the combination of SLCB050 and Adr (Adr+SLCB050) were considerably smaller. H, reduction of DX2 expression in mouse lung tumor tissues in response to SLCB050 treatment. I, TUNEL staining in combined treated DK mice lung tissues. Specific staining on tumor cells was detected in Adr+SLCB050-treated mice.

Figure 7.

Antitumoral effect of SLCB050. A, significant enhanced effect of SLCB050 with GN25, p53 activator in DX2 and DK MEFs. Resistance to GN25 in DX2 and DK cells was abolished by SLCB050 treatment. Cells were incubated with 5 μmol/L of GN25 and SLCB050 for 6 hours. The viability was determined by MTT assay. ##, significant enhanced effect by combination treatment. B, significant enhanced effect of Adr and SLCB050 on viability of primary tumor cells from DX2 and DK mice. Cell viability was determined by MTT assay after the cells were incubated with 0.2 μg/mL of Adr and/or 10 μmol/L SLCB050 for 24 hours. ##, significant enhanced effect by combination treatment. C, sensitivity of SCLC cell lines to combination treatment with Adr and SLCB050, compared with the sensitivity of the p14/ARF-null H322 cells. Only H69 showed the significant enhanced effect by combination treatment with SLCB050 and Adr (##). D, tumor-suppressive effect of SLCB050. In xenograft model using SCLC H446, 10 mg/kg of SLCB050 suppressed tumor growth. Although Adr (5 mg/kg) blocked the tumor growth, within 4 weeks, all of them were ceased by high toxicity (*). In contrast, low dose of Adr (1 mg/kg) with SLCB050 obviously suppressed tumor growth without severe toxicity. Chemicals were delivered via i.p injection 3 times per week. E, in primary tumor cells, combinational treatment could induce p53 synergistically. p53 expression was determined by Western blotting. F, hematoxylin–eosin staining of lung tumor tissues from SLCB050-treated DK mice. Combination treatment with SLCB050 and Adr induced tumor regression. G, reduction of tumor volume by combination treatment. Compared with tumor areas in age-matched untreated (DMSO) or single chemical-treated DK mice (SLCB050: 5 mg/kg and Adr: 1 mg/kg), tumor areas in mice treated with the combination of SLCB050 and Adr (Adr+SLCB050) were considerably smaller. H, reduction of DX2 expression in mouse lung tumor tissues in response to SLCB050 treatment. I, TUNEL staining in combined treated DK mice lung tissues. Specific staining on tumor cells was detected in Adr+SLCB050-treated mice.

Close modal

In vivo antitumor effect of SLCB050

To explore the effect of SLCB050 in mouse model, we injected it into DK mice following our experimental schedule (Supplementary Fig. S8A). Combinational treatment of SLCB050 and low-dose Adr strikingly suppressed tumor progression (Fig. 7F and G) without significant weight loss (Supplementary Fig. S8B). However, nontoxic dose of Adr did not show antitumor effect on this mouse model (Fig. 7F and G; Supplementary Fig. S8B). More detailed histologic analysis showed that SCLC region was more obviously erased by combinational treatment (Supplementary Fig. S8C). In DX2-reduced condition (Fig. 7H), apoptotic tumor cells were obviously increased (Fig. 7I and Supplementary Fig. S8D). Our results indicate that DX2, produced by aberrant splicing of AIMP2, promotes tumor progression, in particular small cell lung cancer, via direct interaction and inhibition of p14/ARF (supplementary Fig. S9). Thus, under DX2-expressed conditions, oncogene-induced tumorigenesis would be easily progressed because oncogene-induced cell death or senescence is abolished.

Although DX2 is elevated in human lung cancer (16, 18), physiopathological role of DX2 has not been revealed. In this study, we found that DX2 is stabilized by oncogenic stresses such as oncogenic K-Ras or Her2/Neu-AKT signaling activation (Fig. 1). Concerning this, siah-1 is speculated as responsible E3-ligase (Fig. 1H and Supplementary Fig. S1F and S1G). However, we did not exclude the involvement of other mechanism on DX2 destabilization, because despite Siah-1 knockdown, AKT can induce DX2 expression (Supplementary Fig. S1E) and phosphatase inhibitor can extend DX2 half-life (Supplementary Fig. S1D). Thus, more intensive investigation on DX2 regulation should be performed, including how DX2 is transcriptionally induced.

Based on differential sensitivity of DX2 transgenic MEF to p53 activators and cellular distribution, we assumed that oncogenic role of DX2 would be related to nuclear proteins (Fig. 2D) and drug resistance (Fig. 2A and B). Moreover, the effect of DX2 knockdown is fully dependent on p14/ARF status but not on p53 status (Fig. 2C). Indeed, si-DX2 can suppress the viability of H1299, p53-null lung cancer cell line but not that of A549, despite intact p53 (Fig. 2C). In addition, DX2 blocks the oncogene-induced p14/ARF expression (Fig. 2D). These results suggest that oncogenic property of DX2 is tightly related with p14/ARF. In fact, DX2 suppresses p14/ARF expression (Fig. 4) through direct interaction (Fig. 5).

In general, oncogenic stimulations activate p53 pathway via p14/ARF and resulted in senescence or apoptosis. Thus, cancer cells should escape from oncogene-induced cell death or senescence. In previous, we revealed that induction of Snail in response to oncogenic K-Ras suppresses p53 in pancreatic cancer (27). Similarly, Her2/Neu-AKT pathway inhibits activation of p14/ARF-p53 network through DX2 induction. However, elimination of DX2 via siRNA or SLCB050 could suppress cell viability in p53-mutated cells (Figs. 2C and 6J), suggesting that p14/ARF executes additional tumor-suppressive functions. Indeed, it has been reported that p14/ARF regulates cell proliferation through ribosomal RNA processing (40). Thus, we are now investigating the effect of DX2 on this.

More interesting feature is that DX2 transgenic mice, including DK mice, generate SCLC (Fig. 3 and Supplementary Fig. S3). In previous, SCLC mouse model is p53/pRb double knockout system (41). Although p53 and pRB are most frequently deleted genes in SCLC, they are also important for most types of cancers, including NSCLC, colon cancer, pancreatic cancer, and even in sarcoma (42). Thus, this model has some limitation for study of SCLC-specific carcinogenesis. Although our double transgenic model (DK mice) can induce SCLC with NSCLC, it would be useful for investigation of early SCLC carcinogenesis.

For the human study, we were looking for human specimen. However, unfortunately, we did not obtain the fresh SCLC tissues, because standard protocol of SCLC does not include surgery. Instead, we obtained the human serum of SCLC. In our previous study, some kinds of proteins that are elevated in human cancer could work as autoantigen (27). Thus, we also check the DX2 autoantibody in this study. Indeed, we could detect DX2 autoantibody from SCLC patient's sera (Fig. 3H). Although it is a prototype at present, more development would provide a new rapid detection method for SCLC through DX2 autoantibody. Concerning this, we are examining the physiopathological relevance of SCLC and DX2 expression. Because p14/ARF is inactivated in SCLC without genetic mutation (43), DX2 may contribute to this.

Because we provide the evidence about interaction of DX2 and p14/ARF, our next step is screening of binding inhibitor that is required for proof of our concept. Indeed, though ELISA-based chemical screening, we found small chemical inhibitors against p14/ARF-DX2 binding. This chemical (Fig. 6B) and its resembling derivatives (Supplementary Fig. S6E) could suppress the interaction of DX2 and p14/ARF (Fig. 6C–G; Supplementary Fig. S6) and cell viability in a p14/ARF-dependent manner (Fig. 6J). Moreover, these chemicals can induce p14/ARF expression and suppress DX2 (Fig. 6H and I).

In fact, SLCB050 show enough anticancer effect in small cell lung cancer cell lines as p14/ARF-dependent manner (Fig. 6J). We also observed the significant enhanced effect with GN25, activator of p53 through inhibition of Snail-p53 binding (Fig. 7A) or Adriamycin (Fig. 7B–E). However, SLCB050 did not show even additive effect with EGFR inhibitors in NSCLC (data not shown) or TAXOL in SCLC (Fig. S7J). These results suggest that SLCB050 would be related with p53-p14/ARF-mediated anticancer signaling. In fact, combinational treatment of SLCB050 and low dosage of Adr could obviously suppress tumor progression and induce cell death in our mouse model. This result indicates that inhibition of DX2 is very attractive target for therapy for SCLC as well as NSCLC. However, until now, we do not completely understand how SLCB050 can induce cell death in p53-deficient SCLC (Fig. 7C). To reveal this, we are now checking several possibilities including p73 activation or blocking of ribosomal RNA processing.

In conclusion, our results have demonstrated that DX2 promotes SCLC progression by inhibiting p14/ARF and that SLCB050 can suppress tumor progression by blocking DX2. SLCB050 may therefore be effective in treatment of lung cancers.

No potential conflicts of interest were disclosed.

Conception and design: A.-Y. Oh, J. Kim, J.-H. Lee, G.-Y. Song, B.-J. Park

Development of methodology: J. Kim, J.-H. Lee, G.-Y. Song, B.-J. Park

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.-Y. Oh, Y.S. Jung, J.-H. Lee, J.-H. Cho, S. Park, N.-C. Ha, J.S. Park, C.-S. Park, G.-Y. Song, B.-J. Park

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.-Y. Oh, H. Park, S. Lim, G.-Y. Song, B.-J. Park

Writing, review, and/or revision of the manuscript: A.-Y. Oh, J. Kim, G.-Y. Song, B.-J. Park

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.-Y. Oh, J.-H. Lee, H.-Y. Chun, G.-Y. Song

Study supervision: B.-J. Park

Other (designed and synthesized specific inhibitor against DX2 and p14/ARF binding such as SLCB050): G.-Y. Song

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (NRF-2013R1A1A2010008; B.-J. Park), Global Frontier Project (2012M3A6A4054952; B.-J. Park), and by the Ministry of Education (NRF-2013R1A1A2062764; G.-Y. Song).

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.

1.
Punturieri
A
,
Szabo
E
,
Croxton
TL
,
Shapiro
SD
,
Dubinett
SM
. 
Lung cancer and chronic obstructive pulmonary disease: needs and opportunities for integrated research
.
J Natl Cancer Inst
2009
;
101
:
554
9
.
2.
Siegel
R
,
Naishadham
D
,
Jemal
A
. 
Cancer statistics, 2012
.
CA Cancer J Clin
2012
;
62
:
10
29
.
3.
Schneider
PM
,
Hung
MC
,
Chiocca
SM
,
Manning
J
,
Zhao
XY
,
Fang
K
, et al
Differential expression of the c-erbB-2 gene in human small cell and non-small cell lung cancer
.
Cancer Res
1989
;
49
:
4968
71
.
4.
Slebos
RJ
,
Kibbelaar
RE
,
Dalesio
O
,
Kooistra
A
,
Stam
J
,
Meijer
CJ
, et al
K-ras oncogene activation as a prognostic marker in adenocarcinoma of the lung
.
N Engl J Med
1990
;
323
:
561
5
.
5.
Salgia
R
,
Skarin
AT
. 
Molecular abnormalities in lung cancer
.
J Clin Oncol
1998
;
16
:
1207
17
.
6.
Vivanco
I
,
Sawyers
CL
. 
The phosphatidylinositol 3-Kinase AKT pathway in human cancer
.
Nat Rev Cancer
2002
;
2
:
489
501
.
7.
David
O
,
Jett
J
,
LeBeau
H
,
Dy
G
,
Hughes
J
,
Friedman
M
, et al
Phospho-Akt overexpression in non-small cell lung cancer confers significant stage-independent survival disadvantage
.
Clin Cancer Res
2004
;
10
:
6865
71
.
8.
Sharma
SV
,
Bell
DW
,
Settleman
J
,
Haber
DA
. 
Epidermal growth factor receptor mutations in lung cancer
.
Nat Rev Cancer
2007
;
7
:
169
81
.
9.
Karachaliou
N
,
Mayo
C
,
Costa
C
,
Magrí
I
,
Gimenez-Capitan
A
,
Molina-Vila
MA
, et al
KRAS mutations in lung cancer
.
Clin Lung Cancer
2013
;
14
:
205
14
.
10.
Hollander
MC
,
Blumenthal
GM
,
Dennis
PA
. 
PTEN loss in the continuum of common cancers, rare syndromes and mouse models
.
Nat Rev Cancer
2011
;
11
:
289
301
.
11.
Pomerantz
J
,
Schreiber-Agus
N
,
Liégeois
NJ
,
Silverman
A
,
Alland
L
,
Chin
L
, et al
The Ink4a tumor suppressor gene product, p19Arf, interacts with MDM2 and neutralizes MDM2's inhibition of p53
.
Cell
1998
;
92
:
713
23
.
12.
Jackman
DM
,
Johnson
BE
. 
Small-cell lung cancer
.
Lancet
2005
;
366
:
1385
96
.
13.
Pleasance
ED
,
Stephens
PJ
,
O'Meara
S
,
McBride
DJ
,
Meynert
A
,
Jones
D
, et al
A small-cell lung cancer genome with complex signatures of tobacco exposure
.
Nature
2010
;
463
:
184
90
.
14.
Travis
WD
. 
Update on small cell carcinoma and its differentiation from squamous cell carcinoma and other non-small cell carcinomas
.
Mod Pathol
2012
;
25
:
S18
S30
.
15.
Kim
MJ
,
Park
BJ
,
Kang
YS
,
Kim
HJ
,
Park
JH
,
Kang
JW
, et al
Downregulation of FUSE-binding protein and c-myc by tRNA synthetase cofactor p38 is required for lung cell differentiation
.
Nat Genet
2003
;
34
:
330
36
.
16.
Kim
S
,
You
S
,
Hwang
D
. 
Aminoacyl-tRNA synthetases and tumorigenesis: More than housekeeping
.
Nat Rev Cancer
2011
;
11
:
708
18
.
17.
Lee
Y
,
Karuppagounder
SS
,
Shin
JH
,
Lee
YI
,
Ko
HS
,
Swing
D
, et al
Parthanatos mediates AIMP2-activated age-dependent dopaminergic neuronal loss
.
Nat Neurosci
2013
;
16
:
1392
400
.
18.
Choi
JW
,
Kim
DG
,
Lee
AE
,
Kim
HR
,
Lee
JY
,
Kwon
NH
, et al
Cancer-associated splicing variant of tumor suppressor AIMP2/p38: pathological implication in tumorigenesis
.
PLoS Genet
2011
;
7
:
e1001351
.
19.
Denissenko
MF
,
Pao
A
,
Tang
M
,
Pfeifer
GP
. 
Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in p53
.
Science
1996
;
274
:
430
2
.
20.
Llanos
S
,
Clark
PA
,
Rowe
J
,
Peters
G
. 
Stabilization of p53 by p14ARF without relocation of MDM2 to the nucleolus
.
Nat Cell Biol
2001
;
3
:
445
52
.
21.
Park
JI
,
Lee
MG
,
Cho
K
,
Park
BJ
,
Chae
KS
,
Byun
DS
, et al
Transforming growth factor-bold italic beta1 activates interleukin-6 expression in prostate cancer cells through the synergistic collaboration of the Smad2, p38-NF-kappaB, JNK, and Ras signaling pathways
.
Oncogene
2003
;
22
:
4314
32
.
22.
Li
YM
,
Pan
Y
,
Wei
Y
,
Cheng
X
,
Zhou
BP
,
Tan
M
, et al
Upregulation of CXCR4 is essential for HER2-mediated tumor metastasis
.
Cancer Cell
2004
;
6
:
459
69
.
23.
Ohta
T
,
Xiong
Y
. 
Phosphorylation- and Skp1-independent in vitro ubiquitination of E2F1 by multiple ROC-cullin ligases
.
Cancer Res
2004
;
61
:
1347
53
.
24.
Rajagopalan
H
,
Jallepalli
PV
,
Rago
C
,
Velculescu
VE
,
Kinzler
KW
,
Vogelstein
B
, et al
Inactivation of hCDC4 can cause chromosomal instability
.
Nature
2004
;
428
:
77
81
.
25.
Eymin
B
,
Claverie
P
,
Salon
C
,
Leduc
C
,
Col
E
,
Brambilla
E
, et al
p14ARF activates a Tip60-dependent and p53-independent ATM/ATR/CHK pathway in response to genotoxic stress
.
Mol Cell Biol
2006
;
11
:
4339
50
.
26.
Saleem
M
,
Kweon
MH
,
Johnson
JJ
,
Adhami
VM
,
Elcheva
I
,
Khan
N
, et al
S100A4 accelerates tumorigenesis and invasion of human prostate cancer through the transcriptional regulation of matrix metalloproteinase 9
.
Proc Natl Acad Sci U S A
2006
;
103
:
14825
30
.
27.
Lee
SH
,
Lee
SJ
,
Chung
JY
,
Jung
YS
,
Choi
SY
,
Hwang
SH
, et al
p53, secreted by K-Ras-Snail pathway, is endocytosed by K-Ras-mutated cells; implication of target-specific drug delivery and early diagnostic marker
.
Oncogene
2009
;
28
:
2005
14
.
28.
Lee
SH
,
Shen
GN
,
Jung
YS
,
Lee
SJ
,
Chung
JY
,
Kim
HS
, et al
Antitumor effect of novel small chemical inhibitors of Snail-p53 binding in K-Ras-mutated cancer cells
.
Oncogene
2010
;
29
:
4576
87
.
29.
Jiang
W
,
Kahn
SM
,
Guillem
JG
,
Lu
SH
,
Weinstein
IB
. 
Rapid detection of ras oncogenes in human tumors: applications to colon, esophageal, and gastric cancer
.
Oncogene
. 
1989
;
4
:
923
8
.
30.
Fiucci
G
,
Beaucourt
S
,
Duflaut
D
,
Lespagnol
A
,
Stumptner-Cuvelette
P
,
Géant
A
, et al
Siah-1b is a direct transcriptional target of p53: Identification of the functional p53 responsive element in the siah-1b promoter
.
Proc Natl Acad Sci U S A
2004
;
101
:
3510
5
.
31.
Nicholson
SA
,
Okby
NT
,
Khan
MA
,
Welsh
JA
,
McMenamin
MG
,
Travis
WD
, et al
Alterations of p14ARF, p53, and p73 genes involved in the E2F-1-mediated apoptotic pathways in non-small cell lung carcinoma
.
Cancer Res
2001
;
61
:
5636
43
.
32.
Sato
M
,
Horio
Y
,
Sekido
Y
,
Minna
JD
,
Shimokata
K
,
Hasegawa
Y
. 
The expression of DNA methyltransferases and methyl-CpG-binding proteins is not associated with the methylation status of p14(ARF), p16(INK4a) and RASSF1A in human lung cancer cell lines
.
Oncogene
2002
;
21
:
4822
9
.
33.
Ji
H
,
Ramsey
MR
,
Hayes
DN
,
Fan
C
,
McNamara
K
,
Kozlowski
P
, et al
LKB1 modulates lung cancer differentiation and metastasis
.
Nature
2007
;
448
:
807
10
.
34.
Gazdar
AF
,
Bunn
PA
 Jr
,
Minna
JD
,
Baylin
SB
. 
Origin of human small cell lung cancer
.
Science
1985
;
229
:
679
80
.
35.
Park
KS
,
Liang
MC
,
Raiser
DM
,
Zamponi
R
,
Roach
RR
,
Curtis
SJ
, et al
Characterization of the cell of origin for small cell lung cancer
.
Cell Cycle
2011
;
10
:
2806
15
.
36.
Kamijo
T
,
Zindy
F
,
Roussel
MF
,
Quelle
DE
,
Downing
JR
,
Ashmun
RA
, et al
Tumor suppression at the mouse INK4a locus mediated by the alternative reading frame product p19ARF
.
Cell
1997
;
91
:
649
59
.
37.
Sherr
CJ
. 
Divorcing ARF and p53: An unsettled case
.
Nat Rev Cancer
2006
;
6
:
663
73
.
38.
Ayrault
O
,
Andrique
L
,
Larsen
CJ
,
Seite
P
. 
Human Arf tumor suppressor specifically interacts with chromatin containing the promoter of rRNA genes
.
Oncogene
2004
;
23
:
8097
104
.
39.
Feldser
DM
,
Kostova
KK
,
Winslow
MM
,
Taylor
SE
,
Cashman
C
,
Whittaker
CA
, et al
Stage-specific sensitivity to p53 restoration during lung cancer progression
.
Nature
2010
;
468
:
572
5
.
40.
Sugimoto
M
,
Kuo
ML
,
Roussel
MF
,
Sherr
CJ
. 
Nucleolar Arf tumor suppressor inhibits ribosomal RNA processing
.
Mol Cell
2003
;
11
:
415
24
.
41.
Minna
JD
,
Kurie
JM
,
Jacks
T
. 
A big step in the study of small cell lung cancer
.
Cancer Cell
2003
;
4
:
163
6
.
42.
Sherr
CJ
,
McCormick
F
. 
The RB and p53 pathways in cancer
.
Cancer Cell
2002
;
2
:
103
12
.
43.
Gazzeri
S
,
Della Valle
V
,
Chaussade
L
,
Brambilla
C
,
Larsen
CJ
,
Brambilla
E
. 
The human p19ARF protein encoded by the beta transcript of the p16INK4a gene is frequently lost in small cell lung cancer
.
Cancer Res
1998
;
58
:
3926
31
.