SUMMARY
Despite often leading to a platinum resistance, platinum-based chemotherapy continues to be the standard treatment for non-small cell lung cancer (NSCLC) and ovarian cancer. In this study, we used array-CGH and qRT-PCR methodologies to identify and validate cytogenetic alterations that arise after cisplatin treatment in four paired cisplatin-sensitive/resistant cell lines. We found the presence of a common deletion of the gene ITF2 in cisplatin-resistant cell lines. Our translational approach included the expression analysis in a total of 35 lung and ovarian primary tumors, showing a frequent downregulation of ITF2. Whole transcriptome sequencing (RNA-seq) and Wnt-pathway analysis in a subgroup of NSCLC patients identified four genes with a potential role in the development of this malignancy. Further functional assays overexpressing ITF2 in NSCLC cisplatin-resistant cells suggest the regulation of HOXD9 by the Wnt pathway and its implication on NSCLC progression.
One-sentence summary Our study identifies Wnt-signaling pathway as a molecular mechanism behind the development of platinum-resistance through ITF2 action; as well as to present two potential therapeutic targets for further studies, ITF2 and HOXD9.
- Abbreviations used in the manuscript
- HULP
- Hospital Universitario La Paz
- TCF4
- Transcription Factor 4
- CDDP
- cisplatin
- qRT-PCR
- quantitative real-time PCR
- qMSP
- quantitative methylation-specific PCR
- NSCLC
- non-small cell lung cancer
- ATT
- adjacent tumor tissues
- Abbreviations
- ATT
- Adjacent Tumor Tissues
- CGH
- Comparative genomic hybridization
- NSCLC
- Non-Small Cell Lung Cancer
- NLM
- Normal lung mean
- TCF4
- transcription factor 4
- HOXD9
- RIOX1, Ribosomal Oxygenase 1
INTRODUCTION
Non-small cell lung cancer (NSCLC) and ovarian cancer are two of the most deadly cancers plaguing our society: the former accounting for more than 80% of primary lung-cancer cases and the latter boasting the highest mortality of the gynecological malignancies worldwide (1). Although platinum-based chemotherapy still plays an important role in the treatment of these two tumor types, the disease progresses to a platinum-resistant state in a high percentage of the diagnosed cases (2,3). Cisplatin (CDDP) is a platinum compound that induces apoptosis in cancer cells (4,5) but leads to cytogenetic alterations, such as deletions or amplifications, which contribute to the development of CDDP-resistance (6,7). Cisplatin treatment also causes changes in DNA copy numbers, altering the expression of genes involved in the drug efflux (ABC transporters family) or the antiapoptotic pathway (BCL2) (8). Moreover cisplatin treatment induce complex aberrations, including deletions of genes involved in tumor progression, metastasis and drug resistance (9) as a consequence of its binding to the N7 position of the guanines and crosslinking DNA in cancer cells (4,5). In this study, by performing a high resolution million feature array-Comparative Genomic Hybridization (aCGH) with four NSCLC and ovarian cancer sensitive/resistant paired cell lines, previously reported by our group (10) we explored the chromosomic deletions that differ in the resistant subtypes. We were able to find a common deletion that includes the Transcription Factor 4, TCF4 (hereafter called ITF2). ITF2 belongs to the Class 1 of bHLH transcription factors involved in the neurogenesis (11). Chromosomic alterations and mutations in this gene are associated with intellectual disability (12–14).
Little is known about how ITF2 regulates different cellular pathways as epithelial-mesenchymal transition (11), besides its role in different tumors, such as being regulated by CXCR4 in breast cancer (15) or β-catenin in colon and ovarian cancers (16,17). Most importantly, ITF2 is a downstream target gene of the Wnt/β-catenin/TCF pathway acting as a tumor suppressor, negatively regulating its activity (16,17).
Therefore, further study of the role of ITF2 and the Wnt signaling pathway in tumor response to chemotherapy is the key to keep finding new ways to fight resistance to this widespread treatment. Here we report the downregulation of ITF2 in NSCLC patients to be frequently involved in cancer development by affecting RIOX1 through the Wnt/B-catenin pathway as well as to present HOXD9 as a potential therapeutic target.
MATERIALS AND METHODS
Cell culture and cell-viability assays
The NSCLC and ovarian cancer cell lines H23, H460, OVCAR3 and A2780 were purchased from the ATCC (Manassas, USA) and ECACC (Sigma-Aldrich, Spain) and cultured as recommended. Their CDDP-resistant variants H23R, H460R, OVCAR3R and A2780R were previously established in our laboratory (10,18). Cisplatin (Farma Ferrer, Spain) was used for CDDP-viability assays. Cells were seeded in 24-well dishes at 40,000 cells/well, treated with increasing doses of CDDP (0, 0.5, 1, 1.5, 2 and 3μg/ml) for an additional 72 or 48 hours and stained as described (19). Cell viability comparing sensitive vs. resistant cell lines was estimated relative to the density recorded over the same experimental group without drug exposure at same period of time. Cell authentication is included in Supplementary Table 1.
Clinical sample and data collection
We selected a representative number of fresh frozen surgical specimens from University Hospital La Paz-Hospital (HULP) Biobank, totaling 25 NSCLC and 10 ovarian cancer samples, belonging to previously reported cohorts of patients (10). Three normal ovarian samples obtained from sex reassignment surgery patients and two lung tissues from non-neoplastic origin were used as negative controls. Follow-up was conducted according to the criteria of the medical oncology divisions from each institution. The samples were processed following the standard operating procedures with the appropriate approval of the Human Research Ethics Committees at each contributing center, including informed consent within the context of research. Clinical, pathological and therapeutic data were recorded by an independent observer, and a blind statistical analysis was performed on these data.
DNA extraction and array of Comparative Genome Hybridyzation
DNA from cell lines was isolated as previously described (20) and was used to analyze copy number variations by the Array-CGH SurePrint G3 Human CGH Microarray 1×1M (Agilent, USA). Array experiments were performed as recommended by the manufacturer (Agilent, USA). Briefly, DNA (500 ng) from the resistant and the sensitive (used as reference) subtypes of each cell line were double-digested with RsaI and AluI for 2h at 37°C. After heat inactivation of the enzymes at 65°C for 20 min, each digested sample was labeled by random priming (Genomic DNA Enzymatic Labeling Kit Agilent) for 2 h using Cy5-dUTP for patient DNAs and Cy3-dUTP for reference DNAs. Labeled products were column purified (Microcon Ym-30 filters, Millipore Corporation). After probe denaturation and pre-annealing with Cot-1 DNA, hybridization was performed at 65°C with rotation for 24 h. After two washing steps, the microarray was then scanned by the Agilent Microarray Scanner and analyzed with Feature Extraction software (v9.1 Agilent Technologies) and Agilent Cytogenomic 3.0.1.1. The analysis and visualization of the microarray data was performed using Agilent Cytogenomic 3.0.1.1. Comprehensive description of the statistical algorithms is available in the user’s manual provided by Agilent Technologies. The Aberration Detection Method 2 (ADM-2) quality weighted interval score algorithm identifies aberrant intervals in samples that have consistently high or low log ratios based on their statistical score. The score represents the deviation of the weighted average of the normalized log ratios from its expected value of zero calculated with Derivative Log2 Ratio Standard Deviation algorithm. A Fuzzy Zero algorithm is applied to incorporate quality information about each probe measurement. Our threshold settings for the CGH analytics software to make a positive call were 6.0 for sensitivity, 0.45 for minimum absolute average log ratio per region, and 5 consecutive probes with the same polarity were required for the minimum number of probes per region.
RNA extraction, RT-PCR, qRT-PCR
Total RNA from human cancer cell lines and surgical samples was isolated, reverse transcribed and quantitative RT-PCR analysis was performed as previously described (10). Briefly, 500 ng of total RNA was used for RT reaction using PrimeScript™ RT Master (Clontech-Takara, USA). For RT-PCR, 2 μl of the RT product (diluted 1:5) was used for semi-quantitative PCR or qPCR reactions with Promega PCR Mix (Promega, USA) and SYBR Green PCR Mix (Applied Biosystems, USA), respectively. RT-PCR was performed under the following conditions: (a) 1 cycle of 95°C for 2 min; (b) 30–40 cycles of 95°C for 1 min, 56°C–60°C for 1 min, 72°C for 1 min; (c) an extension of 5 min at 72°C. RT-PCR products were run on a 1.5% agarose gel, using the 100 bp DNA ladder as a molecular weight marker (New England Biolabs) for appropriate identification of band size. qRT-PCR absolute quantification was calculated according to the 2−ΔCt method using GAPDH as endogenous control, whereas relative quantification was calculated with the 2−ΔΔCt using GAPDH as endogenous control and the sensitive-parental cell line as a calibrator. Samples were analyzed in triplicate using the HT7900 Real-Time PCR system (Applied Biosystems, USA). Primers and probes for qRT-PCR expression analysis were purchased from Applied Biosystems (TCF4: Hs00162613_m1; LRP1B: Hs01069120_m1; DKK1: Hs00183740_m1; GADPH: Hs03929097_g1). Primers for RT-PCR of Cyclin D1, DKK1, LEF, TCF3 genes were kindly donated by Pilar Santisteban Laboratory. Primers for HOXD9, CLDN6, XIST and RIOX1 were designed, when possible, to analyze the specific transcript that significantly showed changes in the RNA-seq; all primers and specific amplification conditions are listed in Supplementary Table 2.
NGS (RNA-seq) and Wnt signaling pathways analysis
Total RNA from nine tumor tissues, three Non-tumor tissues from NSCLC patients and two normal lung samples from healthy donors were sent to Sistemas Genómicos Company (Valencia, Spain) for RNA-sequencing. 100-200 ng of RNA were analyzed by Bioanalyzer 2100 and Qubit 2.0 to determine the quality and quantity of the RNA. mRNA poli (A)+ fraction was isolated from total RNA to obtain the cDNA libraries, following Illumina recommendations. Briefly, mRNA poli (A)+ was isolated using magnetic beads bound to oligo-poly T following subsequent chemical fragmentation prior to retrotranscription. cDNA fragments were subjected to a final repair process through the addition of a unique Adenine base to the 3’ end, followed by adaptors ligation. Finally, products were purified and enriched by PCR to generate the double strand cDNA indexed library. Libraries quality was analyzed in bioanalyzer 2100, High Sensitivity Assay. Quantity of libraries was determined by Real Time PCR in LightCycler 480 (Roche). Libraries were sequenced by end-paired method (100 x 2) in the Illumina HiSeq2500 sequencer.
The bioinformatic analysis performed in the University Hospital la Paz, was based on the FastQC flow-work, to ensure the quality of the reads, and TRIMMOMATIC to eliminate the remnants from the adaptors and discard those sequences shorter than 750 bp. Reads were analyzed to quantify genes and isoforms through the RSEM v1.2.3 methodology (RNA-seq by Expectation Maximization) (21) and using the hg19 versions as reference for annotation. The differential expression was carried out with edgeR, which can estimate the common and individual dispersion (CMN and TGW, respectively) to obtain the variability of the data (22). p-values and FDR statistical analysis were performed by Cmn and Twg models and the statistical cut-off point was set as FDR < 0.05. Normalization was performed by the TMM method (Trimmed mean of M-values) (23). The bioinformatics analysis also included an efficiency analysis for every sample, considering the total efficiency as the percentage of reads annotated belonging to a transcript regarding the total fragments initially read. The Principal Component Analysis (PCA) was used to determine if the NT samples were too different from the lung control samples and no differences were observed, so these 5 samples were considered as controls. We analyzed three different contrasts including 1) comparison between Tumor samples with upregulated DKK1 versus Tumor samples with downregulated DKK1 expression, 2) Tumor samples with upregulated DKK1 versus control samples; 3) All tumor samples versus all control samples. The bioinformatics analysis included a study of all the genes annotated in the literature that have shown an unconfirmed relation with the Wnt-signaling pathway.
Transfection assays: top-fop, B-cat and TCF4 Overexpression
A Myc-DDK-tagged ORF clone of TCF4 and the negative control pCMV6 were used for in transient transfection (RC224345; OriGene, USA) using the previously described methodology (10). Briefly, H23 and A2780 cells were plated onto 60-mm dishes at 600,000 cells/dish and transfected with a negative control or TCF4 vectors, using jet-PEI DNA Transfection Reagent (PolyPlus Transfection, USA). For the Wnt reporter assay A2780S/R cells were plated at a concentration of 600,000 cells/well in 6-MW plates. Cells were serum starved overnight and co-transfected with 0.2 μg of either Super8xTopFlash (2 sets of 3 copies, the second set in the reverse orientation) of the TCF binding site) or Super8xFopFlash (2 full and one incomplete copy of the TCF binding site mutated followed by 3 copies in the reverse orientation) expression plasmids, kindly donated by Pilar Santisteban Laboratory, and 0.1μg pRL-TK (Renilla-TK-luciferase vector, Promega) as a control, using Lipofectamine 2000. Cells were subsequently treated with LiCl 10mM or transfected with S33Y B catenin for 48 hours prior to luciferase activities being measured using a Glomax 96 Microplate Luminometer (Turner Biosystems Instrument, Sunnyvale, USA). After 48 hours of transfection cells were cultured and collected for quantification of Luciferase and Renilla activity. Firefly luciferase activity was calculated as light units normalized with the Renilla activity generated by the pRL-CMV vector. B-catenin/TCF specific activity was calculated by obtaining the ratio of the activities of the Top/Fop promoters and is expressed in relative terms as the number of times of activation of the levels of untreated cells (= 1).
Statistical analysis
The data were compared using the chi-squared test or Fisher’s exact test for qualitative variables, and Student’s t test or the Wilcoxon-Mann-Whitney test (nonnormal distribution) for quantitative variables. Correlation of quantitative variables was analyzed by Pearson’s test. For the translational analysis, overall survival was estimated according to the Kaplan-Meier method and compared between groups by means of the Log Rank test. All the p-values were two-sided, and the type I error was set at 5 percent. Statistical analyses were performed using SPSS 20 software.
RESULTS
ITF2 is frequently downregulated in NSCLC and ovarian cancer cell lines
To identify new cytogenetic alterations after cisplatin administration, we performed Comparative Genomic Hybridization arrays with the DNA of the four paired cell lines from NSCLC, H23S/R and H460S/R, and ovarian cancer, A2780S/R and OVCAR3S/R. The cytogenetic study showed different genomic alterations in the resistant subtypes, using the sensitive cell lines as reference genome. In the NSCLC cell lines, H23 and H460, there was a common deleted region on the q22.1 part of chromosome 2 that included the gene LRP1B. The resistant ovarian cancer cell lines showed common deletions in the q22.33 location of chromosome 9 and the region of chromosome 18 bands q21.2-q21.31. The deletion of chromosome 9 included part of the gene TMOD1 and the deletion in chromosome 18 affected the genes RAB27B, CCDC68, TCF4, TXNL1, WDR7 and BOD1P (Supplementary Table 3). We also found two deleted regions shared by both resistant tumor types, in at least three of the four cell lines (H23, A2780 and OVCAR3) located in chromosome 18: q21.2q-21.31 and q21.32, this one including ZNF532, SEC11C, GRP genes (Figure 1A and Supplementary Table 3).
We selected those genes that were completely deleted in the same tumor type or in at least three of the four cell lines, focusing on LRP1B for NSCLC and TCF4 (ITF2) in H23, A2780 and OVCAR3 and quantify the expression levels of LRPB1 and ITF2 to validate the results obtained in the arrays CGH by qRT-PCR. The deletion of ITF2 in H23R and A2780R cell lines resulted in a significant loss of ITF2 expression compared to the sensitive subtypes. No changes were observed in OVCAR3 cells (Figure 1B), while the deletion of LRPB1 in the resistant subtype of H23 and H460 did not change at the mRNA expression level (Supplementary Figure 1). These results confirm that ITF2 is frequently downregulated in lung and ovarian CDDP-resistant subtypes.
Wnt canonical signaling pathway is increased in CDDP resistant cells with ITF2 deletions
Due to the association of ITF2 with the Wnt/B-catenin/TCF pathway, we studied the involvement of Wnt pathway in cisplatin resistance in H23S/R and A2780S/R cells. We transfected H23S/R and A2780S/R cells with the B-catenin/Tcf4 reporter vectors Super8xTopFlash or Super8xFopFlash and induced the B-catenin activity either by LiCl treatment, which inhibits GSK3-b, or by co-transfection with the constitutively stable B-catenin-S33Y mutant. We observed higher luciferase activity in A2780R cells compared with the parental sensitive ones, indicating an increased transcriptional activity of B-catenin/TCF in response to both pharmacological and functional activation of the pathway (Figure 2A). We did not observed this result in H23 cells (Figure 2B). Expression changes of several downstream effectors of this pathway were analyzed in sensitive and resistant cells. We observed an increase in two of these effectors, DKK1 and Cyclin D1 expression in A2780R cell and a slightly increment of Cyclin D1 in H23R cells (Figure 2C).
Restoration of ITF2 increases the sensitivity to CDDP by decreasing β-catenin/TCF transcriptional activity
To test the role of ITF2 in cisplatin resistance, we transiently overexpressed ITF2 cDNA in the resistant subtypes of A2780 cell line, in which our previous results confirmed the ability to evaluate changes in the transcriptional activity of the Wnt pathway. Overexpression of ITF2 in A2780R resulted in a significant increase in sensitivity to cisplatin (Figure 3A), showing an intermediate phenotype between the parental resistant and sensitive subtypes. The overexpression of ITF2 in A2780R induced a dramatic decrease in cell viability (p<0.05) 24 hours after transfection compared with the parental resistant cells transfected with the empty vector (Figure 3B). ITF2 overexpression at 24 and 72 hours after transfection was confirmed by qRT-PCR (Figure 3C). ITF2 overexpression decreased B-catenin/TCF transcriptional activity in A2780R, recovering the levels of the parental sensitive subtype (Figure 3D). Next, we analyzed the expression of β-catenin/TCF4 downstream targets Cyclin D1 and DKK1 after ITF2 overexpression. RT-PCR analysis showed that ITF2 overexpression in A2780R (R-ITF2) restored in part the expression of only DKK1 at 24 and 72 hours after transfection (Figure 3E).
The expression of ITF2 is frequently downregulated in NSCLC and ovarian tumor samples
To validate our in vitro results and determine the possible clinical implication of ITF2 and DKK1 expression in NSCLC and ovarian cancer patients, we measured the relative expression of both genes in two cohorts of fresh frozen tumor samples (T) belonging to NSCLC (Table 1) and ovarian cancer patients (Supplementary Table 4) from HULP. We also obtained Adjacent Tumor Tissues (ATT) from some of the lung tumor samples.
We observed that ITF2 expression is frequently downregulated in NSCLC and ovarian tumor samples (Figure 4), validating our in vitro data. We did not observe differences between ATT and normal lung samples (LC) for ITF2 (p = 0.177) and DKK1 (p = 0.693) in the NSCLC cohort (Figure 4A). In addition, we found that 15 out of 25 tumor samples of NSCLC patients had lower expression of ITF2 compared to the normal lungs mean (NLM) (Figure 4A). Furthermore, 60% of NSCLC tumor samples showed the opposite expression profile between ITF2 and DKK1 as observed in our experimental data. However, this situation was found only in approximately 10% (1 out of 9) of the tumors on the ovarian cancer cohort of patients (Figure 4B).
The Kaplan Meier curves analyzing the patients overall survival according to the median of ITF2 and DKK1 expressions showed that NSCLC patients with high expression of ITF2 tend to have better overall survival, p=0.1 (Figure 4C). No differences in survival were observed for DKK1 expression p=0.6 (Supplementary Figure 2A). After such promising results, we also analyzed the expression of ITF2 and DKK1 in the Kaplan Meier plotter online tool (24) and we observed that those patients with high expression of ITF2 (Supplementary Figure 2B) and low expression of DKK1 (Supplementary Figure 2C) had a significantly better overall survival rate (p = 0.016 and p < 0.001, respectively).
Identification of candidate genes involved in the Wnt signaling pathway through the analysis of RNA-seq in NSCLC patients
To further explore the role of the Wnt-signaling pathway in the lung cancer tumorigenesis, we performed RNA-sequencing on 14 samples; nine were NSCLC samples, six of them with an opposite expression profile between ITF2 and DKK1 (Pat3, Pat6, Pat10, Pat22, Pat25 and Pat26) and three of them with the same expression profile (Pat8, Pat16 and Pat18). In addition, we included three ATT (Pat9, Pat21 and Pat25) and two normal lung samples (LC1 and LC2) that together were considered as controls for comparisons. ITF2 is mainly downregulated in NSCLC patients, while the expression levels of DKK1 showed a more heterogeneous pattern, therefore we considered DKK1 as the best parameter to decide on the bioinformatics analysis of the RNA-seq. This analysis focused on three main contrasts: Contrast A, the differential gene expression analysis between tumors and controls. Contrast B, comparison between tumors with high and low expression of DKK1. And finally, contrast C was used to compare the tumor samples with high expression of DKK1 with the controls (Supplementary Figure 3). The bioinformatics study also included an analysis of all the genes that have been related to the Wnt-pathway in the literature. We selected those genes that showed a significant expression differences (FDR<0.05) in at least two of the three contrasts, prioritizing contrast B. Thus, we identified three coding genes (HOXD9, RIOX1 and CLDN6) and one long-non coding RNA (XIST) for further validations ((Supplementary Figure 3 and Supplementary Table 5).
Validation of the four candidate genes was initially performed by semi-qPCR (Supplementary Figure 4). HOXD9 expression showed a heterogeneous pattern among the samples analyzed, with almost no representation in normal samples. XIST expression decreased in eight out of nine tumors, compared with the controls. Regarding CLDN6 and RIOX1, we observed higher band intensity in the tumor samples compared with the non-tumor ones. We further confirm these results by qRT-PCR, validating the results from the RNA-sequencing (Figure 5, left panel). HOXD9 was significantly upregulated in 44% (four out of nine) of the tumors compared to the controls and correlated with the RNA-sequencing data accurately (r = 0.83) (Figure 5A, left panel). This correlation was even higher for CLDN6 and XIST genes (r = 0.97 and r = 0.97, respectively) (Figure 5C and 5D, left panel). Conversely, RIOX1 was downregulated in the tumors compare to the controls although the correlation coefficient was less marked (r = 0.58) (Figure 5B, left panel). We next correlated the expression of these candidates with the expression of ITF2 in the 14 samples analyzed, and we observed a negative correlation of HOXD9 (Pearson = - 0.24) and a positive correlation of RIOX1 (Pearson = 0.68) with ITF2 (Figure 5A and 5B, right panel). No correlation was found for XIST and CLDN6 (Pearson = −0.14 and 0.0, respectively) (Figure 5C and 5D, right panel).
In order to gain insight into the role of ITF2 regulating the expression of the selected candidates in NSCLC, we overexpressed ITF2 in H23 cell line. Validation of the transfection efficiency was done by qRT-PCR at 24 and 72h after transfection (Figure 6A). Changes in the expression of the candidate genes were determined by semi-qPCR and qRT-PCR. We observed that overexpression of ITF2 induced a decrease in the expression of HOXD9 at 72h after transfection (p = 0.04) and, as expected, increased the expression of RIOX1 also at 72h (p = 0.08) (Supplementary Figure 5 and Figure 6B). No changes were observed for CLDN6 (p = 0.1) or XIST (p = 0.59) expression after ITF2 overexpression, but we found significant differences in the expression of these genes between the sensitive and the resistant subtypes (p < 0.01). Having identified HOXD9 and RIOX1 as possible candidate genes of the Wnt pathway regulated by ITF2, we studied their translational application in terms of treatment response in the cohort of NSCLC patients used for the RNA-seq analysis. We observed a negative correlation between HOXD9 and the overall survival of the patients (Figure 6C, top panel) and a positive correlation between RIOX1 and same clinical feature (Figure 6D, top panel). The survival analysis performed by the Kaplan Meier method and stratifying the patients according to the median of the expression of the genes, showed that patients with lower expression of HOXD9 present a significant better overall survival rate (p = 0.046) (Figure 6C, bottom panel) whereas patients with low RIOX1 expression tend to live less, although no statistical significance was found in this case (Figure 6D, bottom panel).
DISCUSSION
The search for new biomarkers for cancer progression and response to treatment in cancer has been a wide open field for research in the past years. In this work, we studied the involvement of the Wnt signaling pathway in tumorigenesis through a combined experimental approach by using both comparative genome hybridization arrays and RNA-sequencing, while previous genomic alteration studies on platinum resistance in different types of cancer cell lines, have used only Comparative Genomic Hybridization array methodology (7,8,25,26). We explored the deleted genes after cisplatin administration, as they could have a role in the response to the drug. We identified a common deletion in chromosome 18 involving the region q21.2-q21.31, in H23R, OVCAR3R and A2780R cells. Previous studies analyzing the cytogenetic profile of A2780S/R cells showed different profile probably due to the lower specificity of the CGH-array used to perform those experiments (7,8). From the two genes selected for further validation, LRP1B deletion was shown in H23 and H460 cells, although we could not find changes in expression by using an alternative technique. This is not the first time that the deletion of LRP1B is not validated by gene expression assays (27). In our case, it could be due to the mosaicism observed in this region that occurs in less than 20% of the resistant cells. The level of mosaicism that can be detected is dependent on the sensitivity and spatial resolution of the clones and rearrangements present (28). Nevertheless, LRP1B could still play a role in tumor progression as several studies link its downregulation through deletion and the development of esophageal, renal and NSCL cancers (29–31). The downregulation of the second gene, TCF4 was confirmed by qRT-PCR assays in H23R and A2780R cells, whereas no changes were observed in OVCAR3R, again probably due in part to the low level of mosaicism (36% cells harboring the deletion). Array CGH drawbacks are its limited ability to detect mosaicism and its inability to detect aberrations that do not result in copy number changes. Our results indicate that low levels of mosaicisms would make the validations of expression changes by another quantitative technique difficult, probably because the alteration at expression levels are not significant enough to be detected. The fact that ITF2 is deleted and downregulated after platinum treatment, provides us with a new insight regarding its importance in resistance to chemotherapy. ITF2 is a transcription factor belonging to the basic Helix Loop Helix (bHLH) family, also referred to as an E-protein due to its ability to bind to the Ephrussi-box in DNA. It can act as a transcriptional activator or repressor (32,33) and its regulation still is unknown. It is important to note that ITF2, whose real name is TCF4, is not the same as the T-Cell Factor 4 (TCF7L2), also known as TCF4, which is a downstream transcription factor of the Wnt/B-catenin/TCF pathway (34). In fact, ITF2 is transcribed by the B-catenin/TCF axis and acts as a molecular repressor of the transcription mediated by this complex through interfering the binding of B-catenin with TCF4, leading to the repression of cell proliferation (17). Consistent with these studies, we have observed that the resistant A2780 cells have an increased activity of the B-catenin/TCF transcription, probably due to the absence of ITF2. In contrast, we did not observe differences in H23 cells, which could be explained by previous observations that H23 has a higher basal activity of the Wnt signaling pathway and no exogenous activation could imply a difference (35). Moreover, we have observed that the overexpression of ITF2 in A2780R cells leads to a decrease in cell viability, rescuing the sensitive phenotype, maybe through the inhibition of the excessive proliferation and the activity levels of the B-catenin/TCF transcription. In fact, our results indicate that resistant cells respond better to the activation of the Wnt pathway and that Wnt signaling pathway plays an important role in the resistance to cisplatin through ITF2 in cancer cell lines.
The translational analysis, based on the expression levels of ITF2 and DKK1, a downstream effector of Wnt pathway in two different cohorts of patients was aimed to elucidate the role of this pathway in tumor progression and chemotherapy response. ITF2 expression was frequently downregulated in NSCLC and ovarian tumor samples, validating our in vitro data. The expression levels of DKK1, however, showed a more heterogeneous pattern in the NSCLC tumor samples, while no differences were observed in the ovarian tumors. These results suggest an aberrant activation of the Wnt signaling pathway in lung cancer. In fact, our in silico analysis of 1926 NSCLC patients indicates a significant increased overall survival when patients present high expression levels of ITF2 and low expression of DKK1. We observed the same findings from our own data although without statistical significance, probably due to the sample size. However, one of the strengthens of our cohort is that is comprised by fresh frozen samples, allowing to perform high quality RNA-sequencing in a group of NSCLC patients, in order to determine the involvement of the Wnt-pathway components in lung cancer development. The differential expression of DKK1 within the tumor samples allowed us to performed three different bioinformatics contrasts in order to explore all the possibilities focused on tumor development and Wnt signaling pathway. Contrast A aimed to identify genes with a possible involvement in lung cancer development by comparing differential expression in tumors versus control samples. Contrast B was made to identify alterations in the Wnt pathway in NSCLC tumors and those which could be used as potential therapeutic targets. Finally, contrast C was able to identify genes regulated by the Wnt pathway and others involved in NSCLC development. Using this approach with RNA-sequencing allowed us to identify new coding, non-coding genes and transcripts that have not been functionally characterized previously (36,37). Indeed, in this study we have identified three coding genes, HOXD9, CLDN6 and RIOX1, and one non-coding gene, XIST, which can be potentially involved in NSCLC progression through the Wnt signaling pathway. Our data was validated by two alternative methodologies, finding a strong positive correlation among these technologies for three of the candidates HOXD9, CLDN6 and XIST and slightly weaker correlation for RIOX1.
HOXD9 and CLDN6 were significantly downregulated in tumors with high expression levels of DKK1 and upregulated in tumors compared with controls, indicating that these genes could be involved in tumor progression through an aberrant activation of the Wnt signaling pathway. For HOXD9, the tumor samples had higher expression levels than the controls, as previously reported (38). We also observed a negative correlation of HOXD9 and ITF2 expression levels. In addition, patients with lower levels of HOXD9 had better overall survival than those with upregulated expression of this gene. These results are consistent with previous studies linking a high expression of HOXD9 with glioblastoma and hepatocarcinoma (39,40). Our functional analysis showed that ITF2 overexpression in H23R cells decreased the expression of HOXD9. However, these results do not indicate a direct regulation of HOXD9 by ITF2, as we expected a recovery of its expression to the sensitive subtype levels. We believe, that an alternative regulatory mechanism affected by ITF2 is modulating the expression of HOXD9. It has been shown that the expression of HOXD9 is also regulated by epigenetic mechanisms, such as the long-non coding RNA HOTAIR (41,42), which has been linked with cisplatin resistance (43). Although CLDND6 has been previously related with carcinogenesis of different tumor types (44,45), we did not find any correlation with ITF2 expression levels. In addition, the overexpression of ITF2 in H23R cells did not change the expression of this gene, altogether suggesting that CLDN6 is not involved in NSCLC development or Wnt signaling pathway.
The coding gene RIOX1 and the non-coding RNA XIST, showed expression inhibition in tumor samples with high levels of DKK1 and controls. Two different bioinformatic contrasts imply both genes in tumor progression through their regulation by the Wnt signaling pathway. Although an upregulation of XIST, which is involved in X-chromosome inactivation (46), has been shown to be involved in tumor development through the Wnt pathway (47), our results do not show a clear correlation with ITF2 expression or a relationship with clinical features of NSCLC patients. Conversely, RIOX1 showed increased expression levels in the tumor samples and a positive correlation of its expression with ITF2. In addition, patients with higher levels of RIOX1 tended to have better overall survival than the patients with lower expression levels. Moreover, overexpression of ITF2 in the resistant subtype of H23 cell line increased the expression level of RIOX1 up to the sensitive subtype. Although little is known about RIOX1, it has recently been shown to be involved in renal and colorectal carcinogenesis (48,49).
In summary, we have identified that ITF2 is frequently downregulated in cisplatin-resistant cancer cells as well as in NSCLC and ovarian cancer patients. Moreover, our identification of the Wnt-signaling pathway as a molecular mechanism behind the development of cisplatin resistance in cancer cells through the action of ITF2 provides novel insights into the molecular biology of the cellular mechanisms involved in the acquired resistance to the most widely use chemotherapy agent, as well as to present two potential therapeutic targets for further studies, ITF2 and HOXD9.
AUTHOR CONTRIBUTIONS
IIC: Conception and design
OP, ASP, AGG, MPB, CR, RMA, VP and OVP: development of methodology
OP, ASP, CRA, AGG, RR, DSC, CR, JdC and IIC: acquisition of data
OP, ASP, AGG, RR, DSC, CR, PS, JdC, OVP and IIC: analysis and interpretation of data
CRA and MPB: bioinformatic analysis and interpretation of bioinformatic data
OVP: wrote the manuscript
All authors reviewed and/or revised the manuscript.
ACKNOWLEDGEMENTS
The authors thank Hayley Pickett for the English language correction. The authors also acknowledge Biobank from HULP for sample processing. This study was supported by the ‘Fondo de Investigación Sanitaria-Instituto de Salud Carlos III’ [PI15/00186 and PI18/00050 to I.I.C.]; and the European Regional Development Fund/European Social Fund FIS [FEDER/FSE, Una Manera de Hacer Europa]. MINECO funds support C.R.A. and O.P. contracts through RTC-2016-5314-1 project.
Footnotes
Declaration of interest: All the authors have read the journal’s authorship statement and declare no potential conflicts of interest.