Next-generation sequencing analysis of receptor-type tyrosine kinase genes in surgically resected colon cancer: identification of gain-of-function mutations in the RET proto-oncogene

Next-generation sequencing analysis of receptor-type tyrosine kinase genes in surgically resected... Background: Improvement in genetic characterization of Colon Cancer (CC) patients is required to propose new potential targets, since surgical resection coupled to chemotherapy, presents several limits such as cancer recurrence and drug resistance. Targeted therapies have more efficacy and less toxicity than standard treatments. One of the most relevant cancer-specific actionable targets are receptor tyrosine kinases (RTKs) whose role in CC need to be better investigated. Methods: We have analysed 37 CC patients using the Ion AmpliSeq™ Comprehensive Cancer Panel (CCP). We have confirmed the somatic nature of RET variants through Sanger sequencing and assessed RET activation status and protein expression by immunofluorescence and western-blot analyses. We have used RET mutant expression vectors to evaluate the effect of selected mutations in HEK293 cells by performing proliferation, migration and clonogenic assays. Results: Among the 409 cancer-related genes included in the CCP we have focused on the RTKs. Overall, we have observed 101 different potentially damaging variants distributed across 31 RTK genes in 28 patients. The most frequently mutated RTKs were FLT4, ROS1, EPH7, ERBB2, EGFR, RET, FGFR3 and FGFR4. In particular, we have identified 4 different somatic variants in 10% of CC patients in RET proto-oncogene. Among them, we have demonstrated that the G533C variant was able to activate RET by promoting dimer formation and enhancing Y1062 phosphorylation. Moreover, we have demonstrated that RET G533C variant was able to stimulate anchorage-dependent proliferation, migration and clonogenic cell survival. Notably, the effects induced by the RET G533C variant were abolished by vandetanib. Conclusions: The discovery of pathogenic variants across RTK genes in 75% of the CC patients under analysis, suggests a previously underestimated role for RTKs in CC development. The identification of a gain-of-function RET mutation in CC highlights the potential use of RET in targeted therapy. Keywords: Next-generation sequencing, Receptor-type tyrosine kinases, Colon cancer, RET proto-oncogene * Correspondence: viglietto@unicz.it; malanga@unicz.it Department of Experimental and Clinical Medicine, University Magna Graecia of Catanzaro, Campus Salvatore Venuta -Viale Europa, Catanzaro 88100, Italy Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 2 of 12 Background effectors downstream EGFR such as KRAS, BRAF, Targeted therapy of cancer has more efficacy and less PIK3CA and PTEN [22]. Also the patients that initially toxicity than conventional chemotherapy because it is respond to the anti-VEGF therapy [23, 24] will eventu- directed against cancer-specific molecules and/or signal- ally show evidence of resistance to anti-angiogenic ther- ling pathways. So far the most relevant cancer-specific apy [25]. actionable targets are protein tyrosine kinases (PTKs) Clearly, we need to improve the molecular cha- [1]. PTKs regulate several cellular processes that include racterization of tumors from CRC patients in order to cell proliferation, cell death, motility and invasion [2]. identify the molecular alterations inside cancer cell ge- When subverted, these processes may give origin to nomes that drive cancer growth and determine the inef- cancer [2]. ficacy of anti-EGFR/VEGF targeted therapies. In a The catalogue of genes encoding PTKs within the hu- previous work we have reported the analysis of 37 colon man genome has been updated to 90 genes, of which 58 cancer patients using the Ion AmpliSeq™ Comprehen- are receptor type PTKs and 32 are non-receptor type sive Cancer Panel on the Ion Proton platform (Oliveira PTKs [3].Genes encoding PTKs - both receptor and et al., under review). In this manuscript, we have fo- non-receptor - can be oncogenically activated through a cused the analysis on the role of genes encoding RTKs variety of mechanisms that include gain-of-function mu- as potentially targeting genes in colon cancer. The dis- tations as in the case of EGFR in lung cancer and KIT in covery of pathogenic variants in this gene family in the GIST [4–7], translocations as in the case of BCR-ABL in majority of colon cancer patients suggests a previously CML, ALK in lung cancer and RET in thyroid, lung and underestimated role for RTKs in the development of colorectal cancer [8–10] and amplifications as in the colon cancer and thus the possibility to treat those pa- case of ERBB2 in breast cancer and EGFR in lung, tients with specific RTK inhibitors. In particular, the HSNCC and colon cancer [11–15]. identification of gain-of-function RET variants in a sub- In any case, the constitutive signalling from PTKs set of colon cancer, as described here, has led to the leads to aberrant activation of downstream pathways identification of a small subset of patients potentially that promote continued cell growth and survival. Cancer treatable with RET inhibitors. cells are particularly sensitive to intracellular pathways that are improperly hyperactive, a phenomenon known Methods as oncogene/pathway addiction. This phenomenon pro- Tumor samples vides a unique therapeutic opportunity to block cancer Patient accrual was conducted according to Institutional by inhibiting PTK activity [16]. Review Board of the Mater Domini/University Magna Colorectal cancer (CRC) is one of the most commonly Graecia (Catanzaro, Italy). The study was approved by diagnosed and lethal cancers worldwide, with 1.4 million the Institutional Review Board of the Mater Domini/ new cases and 690,000 deaths in 2012 [17]. Prognosis University Magna Graecia in the meeting of May 21st of CRC patients depends on cancer stage at diagnosis. 2014. Tumor, normal mucosa and peripheral blood sam- Detection of CRC at early stage may confer a 90% 5- ples were obtained from patients referring to General year survival rate, compared to 12% if distant metasta- Surgey Unit of University Hospital Magna Graecia of sis has occurred. Catanzaro (Catanzaro, Italy), who underwent surgical Current therapy for colon cancer patients include resection for colon cancer since January 2013. Biopsies surgical resection coupled to chemotherapy [18]. In ad- were immediately snap frozen and stored at − 80 °C. vanced disease, either FOLFOX (oxaliplatin/5-fluoro- In all cases diagnosis was confirmed by reviewing uracil [5-FU]/leucovorin [LV]) or FOLFIRI (irinotecan/ hematoxylin/eosin-stained slides. LV/5-FU) are typical regimens for first-line chemother- apy, but cancer recurs in many patients [19]. Targeted Patients’ demographics therapy is suggested for patients with metastatic disease General demographic information, clinical findings, sur- [18]. In these patients the use of monoclonal antibodies gical treatment, histo-patological examination and to epidermal growth factor receptor (EGFR) (cetuxi- follow-up data were collected for each patient. Detailed mab, panitumumab) and/or to vascular endothelial documentation of the histo-pathological findings allowed growth factor (VEGF) (bevacizumab) has increased the classification according to the current edition of UICC median survival beyond 20 months [20]. However, the [26]. None of the patients received chemotherapy or ra- responses to EGFR-targeted antibodies are relatively diation therapy prior to surgery. See Additional file 1 for low, with improvement in survival rates lasting only complete characteristics of patients. Briefly, among the few months [21]. 37 patients, 13 were women and 24 were males. The In addition, the efficacy of anti-EGFR antibodies is mean age was 68 years old (range 47–84). Stage was limited to patients with tumors lacking mutations in known for 36 of the 37 patients: 7 patients had stage I Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 3 of 12 disease, 13 patients had stage II disease, 12 patients had phospho-RET Y905 (rabbit polyclonal antibody anti- stage III disease and 4 patients had stage IV disease. pRET, 1:100 diluition, #3221 Cell Signaling Technology) Grade was known for 35 out of 37 patients: 26 patients or phospho-RET Y1062 (rabbit polyclonal antibody anti had tumors that were graded G2 and 9 patients had tu- p-RET, 1:100 dilution) [27] for 60 min, followed by incu- mors that were graded G3. Four patients presented bation with secondary Alexa Fluor 488F fragment of goat distant metastasis. anti-rabbit IgG (H + L) antibody; 1:500 dilution (Thermo- Fisher Scientific, MA, USA #A11070) for 45 min. Cells Next generation sequencing were counterstained with DAPI,1:500 dilution, (Thermo- Sequencing was performed with the Ion AmpliSeq™ Fisher Scientific, MA, USA), mounted using anti-fade Comprehensive Cancer Panel on the Ion Proton system mounting medium #53023 (DAKO, Glostrup, Denmark) (ThermoFisher Scientific, MA, USA) starting from 40 ng and observed at Fluorescence Microscopy (Leica Micro- of DNA and following standard protocols. Primary bio- systems, Wetzlar Germany). informatic analysis of data is described in a previous manuscript (Oliveira et al., under review). We set an Protein extraction and immunoblot average mean coverage of > 300, a coverage of > 40 for Protein extracts were prepared with lysis buffer contain- each variant and variant frequency > 5%. To further filter ing 50 mM HEPES pH 7.5, 5 mM EDTA, 250 mM NaCl, the number of non-pathogenic variants identified in the 1 mM dithiothreitol, 0.5% Nonidet P40, 1 mM Na VO , 3 4 study the following criteria were used: i) variants with 1 mM NaF supplemented with 10 μg of aprotinin/ml, quality score ≤ 30 were excluded, ii) variants that were 10 μg of leupeptin/ml, 1 mM PMSF and a mix of prote- not annotated as pathogenic or likely pathogenic by ase inhibitors (SIGMAFAST protease inhibitor Tablets SIFT and Polyphen2 algorithms were excluded, iii) for general Use; Sigma-Aldrich St. Louis, MO, USA). Ly- variants detected in both normal and tumor samples sates were centrifuged at 13,000 rpm for 30 min at 4 °C were excluded. Finally, manual observation of the and the supernatants were collected. Protein concentration variants using the Integrated Genomic Viewer (IGV – was estimated with a modified Bradford assay (Bio-Rad Broad Institute) was performed to exclude inconclusive Laboratories, Berkeley, CA, USA). Western blot analysis variants. was carried out by standard methods. To detect RET dimers, electrophoresis was carried out Sanger sequencing on a 6% SDS–PAGE in non-reducing conditions. In this DNA obtained from matched normal and tumor sam- case proteins extracts were prepared in the same lysis ples was amplified using specific primers. The purified buffer as above devoid of dithiothreitol, and with sample products were sequenced using BigDye terminator v3.1 loading buffer devoid of 2-mercaptoethanol. Samples (Applied Biosystems, Foster City, CA) with ABI 3100 were not heated upon loading. Genetic Analyser (Applied Biosystems, Foster City, CA). Proteins were revealed by enhanced chemilumines- cence detection using Clarity™ Western ECL Substrate RNA extraction and reverse-transcription (Bio-Rad Laboratories, Berkeley, CA, USA). The anti- Tumors obtained at surgery were snap frozen in liquid bodiesusedinthisstudy were:anti-RET(E1N8X, nitrogen and conserved at − 80 °C. RNA was prepared #14556), anti-phospho-RET, (#3221), anti-phospho- using Trizol (ThermoFisher Scientific, MA, USA) and ERK1/2 (#9101), anti-ERK1/2 (#9107) purchased from purified using RNeasy mini-eluate clean-up kit (QIAGEN Cell Signaling Technology, Danver, MA, USA. The Inc. CA, USA). cDNA was retro-transcribed with random anti-phospho-RET Y1062 used in this study has been hexamers using SuperScript®III First-Strand Synthesis previously described [27]. System (ThermoFisher Scientific, MA, USA). Construction of RET mutants expression vectors Immunostaining RET mutants were obtained by site-specific mutagenesis Deparaffinized 4-μm sections were hydrated in decreas- of RET51-WT construct (pBabe vector encoding for the ing ethanol gradient solutions and rinsed in wash solu- proto-RET gene long isoform). The RET51-C634R plas- tion (TBST, 0.05 mol/L Tris Buffered Saline with mid, used as positive control, is described elsewhere [28]. Tween20). Antigen retrieval was performed with citrate RET51-G533C and RET51-P1047S were obtained by site- buffer pH 6 for antibodies to RET and phospho-RET directed mutagenesis of RET51-WT using an in vitro Y905, and pH 9 for antibody to phospho-RET Y1062, re- oligonucleotide mutagenesis system (Quik-Change XL spectively, for 20 min at 98 °C followed by washing in site-directed mutagenesis; Agilent Technologies, CA, TBST. Sections were incubated with antibodies against USA). Plasmid DNA was extracted using the QIAGEN RET (rabbit monoclonal antibody anti-RET, 1:250 dilu- Plasmid Maxi Kit (QIAGEN CA, USA) as suggested by tion, #14556 Cell Signaling Technology, Danver, MA), the supplier. The presence of the specific mutations was Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 4 of 12 verified by DNA direct sequencing. Mutant plasmids were Results sequenced entirely to exclude the presence of additional Mutations in genes encoding RTKs in colon cancer mutations. In a previous work, we have reported the analysis of 37 colon cancer patients using the Ion AmpliSeq™ Com- prehensive Cancer Panel (CCP) (Oliveira et al., under Transfections and colony formation assay review). In that work, patients who had undergone sur- Human HEK293T cells were maintained in Gibco™ gery for colon cancer at the General Surgery Unit of DMEM with 10% FBS and 1% penicillin/streptomycin University Magna Graecia of Catanzaro in the years (ThermoFisher Scientific, MA, USA). Empty pBABE vec- 2013–2015 were studied. Complete demographic and tor and/or plasmids encoding RET wild type or mutant clinical information of patients are reported in Additional alleles were transiently transfected using Lipofectamine file 1. DNA was extracted from matched normal and 3000 (ThermoFisher Scientific, MA, USA) according to pathological tissues and subjected to NGS analysis on the the manufacturer’s instructions. Forty-eight hours after Ion Torrent platform using the CCP (ThermoFisher transfection, cells were trypsinized and plated into 100-mm Scientific, MA, USA), that provides complete exon cover- plates. Transfected cells were selected with 0.75 μg/ml age of the 409 most important cancer-associated genes. puromycin for 1 week until all cells in the control plates As indicated in Oliveira et al., (under review), the se- were dead. Then cells were trypsinized, counted and 5000 6 quencing performance achieved was: 13 × 10 mean cells were plated into 60-mm plates. Cells were kept in number reads/sample, 791 (range 102.5–2656) mean DMEM with 10% bovine serum. The culture medium was sequence coverage depth of targeted exonic regions of changed every 3–4 days. Approximately 10 days after trans- the 409 genes analysed, with median uniformity of fection, cells were stained by crystal violet. The colony for- sequenced genes 97% (range 0.7–0.99). Common germ- mation experiments were repeated three times. line variants were removed by filtering sequentially through a pool of peripheral blood samples available from some patients (n = 12), the dbSNP141 and the MTT assay 1000 Genomes Project data sets. Variants were function- Stably transfected cells expressing RET wild type and ally annotated using the algorithms Sift and Polyphen2 mutant alleles were seeded in 96 wells round bottom to predict their pathogenicity. plates (1000 cells/well). After 24 h, 10 μl of MTT [3- In this manuscript, we have re-analysed the data gen- (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium brom- erated in the manuscript from Oliveira et al., focusing ide]; (Sigma-Aldrich St. Louis, MO, USA) was dissolved on genes encoding receptor-type tyrosine kinase (RTK). at the concentration of 5 mg/ml in warm assay medium The number of RTK genes contained in the Comprehen- was added to cells at the indicated time points (24 h, sive Cancer Panel that were object of the present study 48 h, 72 h and 96 h) to yield a final assay volume of is 31. Overall, upon analysis of the data reported in Oli- 110 μl/well. Plates were incubated for 4 h at 37 °C and veira et al. we have observed 101 different potentially 5% CO . After incubation, supernatants were removed, damaging variants distributed across 31 RTK genes. See and 100 μl of isopropanol was added. Plates were placed Additional file 2 for a general summary of the analysis. on an orbital shaker for 5 min, and the absorbance was We have found that 28 patients out of 37 presented vari- recorded at 570 nm. Triplicates were made for each time ants in at least one gene encoding RTK, with the point and two independent experiments were conducted. remaining 9 patients showing no variant in any of the 31 Statistical analysis was performed by two-way ANOVA RTK genes. The mean number of mutated RTK genes/ and Dunnett’s multiple comparison test with GraphPad patient was 3.4 (range 1–13). Of the 28 samples that Software (La Jolla, CA, USA). presented mutations in RTK genes, 8 patients had only one mutated gene, 6 had two different mutated genes Wound healing assay and 14 presented 3 or more mutated genes. See Stably HEK293 cells expressing RET wild type or mutant Additional file 3 for patient-by-patient analysis of the re- alleles were seeded in a 60 mm plates and cultured until sults. Among RTK genes that presented variants, we ob- reaching 90% confluence. A micropipette tip (200 μl) was served variants in FLT4 (G1131S, R104Q, E1052K, used to make a scratch and simulate a wound. Cells were D1003N, D728N, P707L, R362L, R282*, Q213*, G180R, monitored after 0 h, 24 h and 48 h from the scratch and P138S, E36* and 3 frameshifts) in 10 patients, ROS1 images of wound healing were captured (magnification of (S277Y, P437L, S653F, T804 N, Q1127*, G1709C and 10X) using the Leica DFC420 C and Leica Application QG1708HC) in 7 patients, EPHA7 (D839G, P805L and Suite Software (Leica Microsystems, Wetzlar Germany). P278S) in 6 patients, RET (R77C, P270L, G533C, P1047S) Subsequently, cell migration was quantified by measuring in 4 patients and MET (A320V, R988C) in 2 patients. In the wound opening area with ImageJ64 software. addition, ERBB2 variants (S310F, W482C, Q533*, T631A, Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 5 of 12 S819F, A1216D) were detected in 6 patients, EGFR variant 30.5%. Raw data showing the detection of the (R669*, W731*, D807E, L933P) and ERBB4 (R983K, different RET mutations by NGS are contained in M977 V, R847H, T244I) variants were detected in 4 pa- Additional file 4. tients, and ERBB3 variants (T389I, V850 M) were detected Interestingly, mutation G533C is located in the RET in 2 patients. FGFR1 (D784E, A131V) and FGFR2 extracellular domain and consists in the replacement of (G302 K, L104P) variants were detected in 2 patients, a glycine amino acid with a cysteine as often observed in whereas FGFR3 (D320N, A352E, A571V, T653I) and families with familial medullary thyroid carcinoma FGFR4 variants (R130C, KE144RG, F859 L, T912I) were (FMTC) or Multiple Endocrine Neoplasia (MEN) 2A. detected in 4 patients. Eight tumors presented variants in Notably, the G533C substitution was identified in pa- only one RTK gene (FLT4, EGFR, ERBB4, ROS1, and tients with FMTC [29], suggesting that it represents a EPHA7), which suggests that these variants can represent gain-of-function mutation. For the complete list of mu- driving mutations. Accordingly, the variants detected in tated genes in the patient harboring the G533C substitu- the EGFR, ERBB4 and FLT4 were present in the COSMIC tion see Additional file 5. First we confirmed the results database. Conversely, co-occurrence of variants in differ- obtained through NGS by traditional Sanger sequencing. ent genes encoding RTKs was frequent, occurring in 20 We performed DNA sequencing of RET exons on tumors outof37analysed(54%).Co-occurrenceof vari- matched normal and cancer samples to confirm the ants included 2 different genes (FGFR4/NTRK1, ERRB3/ presence of the specific mutations identified by NGS FLT3, ALK/ERBB4, EPHA3/EPHA7, EPHB4/ROS1, and (See Fig. 1). Among the 4 different RET mutants identi- FLT4/LTK) in 6 cases and three or more genes in the fied in colon cancer patients we excluded the P270L remaining 14 cases. Notably, some samples showed vari- variant because it represented less than 10% of the cell ants in numerous genes encoding RTK (CC27, 11 genes; population and the R77C since it has been observed in a CC34, 13 genes; CC12, 8 genes). case of Hirschsprung’s disease [30]. Conversely, variants G533C and P1047S were included in the subsequent Identification of RET mutations in colon cancer by NGS analysis. We found that in the two patients, mutations and validation of NGS results were cancer-specific and were not found in the corre- Among the RTKs showing variants identified in this sponding normal tissue. These findings led us to exclude study, we decided to biologically validate the variants the presence of germ-line mutations associated with identified in the gene encoding the RET receptor. inherited diseases such as FMTC, MEN2 type 2A, The reasoning for this choice was that, at difference MEN2 type 2B and/or Hirschsprung’s disease. Con- with what happens in thyroid and lung cancer, the in- versely mutations were somatically acquired and were volvement of RET proto-oncogene to the develop- apparently selected for within the bulk of tumor cells ment of colon cancer is unclear. Variants in the RET since they represented 20–49% of RET alleles, respect- gene were identified in four colon cancer patients ively. We analysed RET expression in the tumors and in (patients CC12, CC20, CC25 and CC33), as indicated the corresponding normal mucosa from patients CC12 in Additional file 2. and CC20 where RET mutations were identified. In the present study, we have identified a C/T transi- The expression of RET in the samples harbouring tion that led to the replacement of P1047 with S RET mutations was evaluated by semi-quantitative (P1047S) in the cytoplasmic tail of RET in patient CC12, reverse transcriptase-polymerase chain reaction (RT- a C/T transition that led to the replacement of P270 PCR) (Fig. 2) and by in situ immunofluorescence ana- with L (P270L) in the RET extracellular domain in pa- lysis (Fig. 3a). RT-PCR analysis revealed the presence of tient CC33, an A/T transversion that led to the replace- a RET transcript in normal mucosa and the corre- ment of G533 with C (G533C) in the cysteine-rich sponding cancer in all patients analysed. The detection domain of RET protein in patient CC20 and a G/A tran- of RET transcript by RT-PCR mRNA was corroborated sition that led to the replacement of R77 with C (R77C) by analysis of the RET protein in normal and cancerous in the RET extracellular domain in patient CC25. tissues by immunofluorescence. Expression of RET pro- As to the technical aspects of NGS, patient CC12 tein and the levels of its phosphorylation at residues showed total mean coverage of 1990, coverage of the Y905 and Y1062 were assessed in colon cancer and the variant of 45 and frequency of the variant 49%, patient corresponding normal mucosa of patients CC12 and CC33 showed total mean coverage of 789, coverage of CC20 using immunofluorescence. the variant of 349 and frequency of the variant 5.9%, pa- Tissues were scored as positive when showing uni- tient CC20 showed total coverage of 728, coverage of form, intense cytoplasmic labelling with marked accen- the variant of 442 and frequency of the variant 20.4% tuation of membranes, and as negative or weakly and patient CC25 showed total mean coverage of 761, positive when labelling was a faint cytoplasmic staining coverage of the variant of 89 and frequency of the with few positive cells interspersed among the majority Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 6 of 12 CC12 Normal CC12 Tumor (C/T), P1047S CC20 Normal CC20 Tumor (G/T), G533C Fig. 1 RET is somatically mutated in colon cancer samples. a. DNA Sanger sequencing of normal mucosa and the corresponding tumor tissue of patient CC12 [exon 12 (C/T)] (b). DNA Sanger sequencing of normal mucosa and the corresponding tumor tissue of patient CC20 [exon 7 (G/T)] of negative cells. We found that normal colon mucosa of residues while the normal mucosa samples were either the two samples analysed were moderately positive for negative or presented weak positivity (see Fig. 3b). These RET expression (see Fig. 3a). results suggest that, at least in the case of some RET Similarly, we found that the corresponding cancerous mutants, the expression of the receptor is not impaired by tissues of patients CC12 and CC20 retained expression selective pressure for loss of RET expression. of RET at levels that were comparable with normal mucosa. Moreover, tumor samples harbouring RET mutations RET activating mutations identified in colon cancer showed positivity for phospho-Y905 and phospho-Y1062 patients increase anchorage-dependent cell proliferation and clonogenic cell survival We investigated whether the RET variants identified in colon cancer patients are able to stimulate anchorage- dependent proliferation in HEK293 cells. This cellular system has already been successfully used to investigate the oncogenic characteristics of multiple oncoproteins [31]. To this aim, the G533C and P1047S RET variants were generated by site-specific mutagenesis of pBABE-puro plas- mid carrying the cDNA encoding human wild type RET (RET-G533C, RET-P1047S and RET-WT, respectively). HEK293 cells were transfected with empty vector (HEK293-C) or transfected with plasmids carrying wild type or mutant RET (HEK293-RET-WT, HEK293-RET- G533C and HEK293-RET-P1047S, respectively). Multiple Fig. 2 Semi-quantitative RT-PCR analysis of RET expression in colon clones of HEK293-C, HEK293-RET-WT, HEK293-RET- cancer samples. Semi-quantitative RT-PCR was performed to investigate RET expression in matched normal (N) and tumor (T) samples in patients G533C and HEK293-RET-P1047S cells from transfection CC12 and CC20. Actin mRNA was used as control for RNA integrity and experiments were expanded and/or frozen for further quantity. TT cells were used as positive control. Sample without input studies. HEK293-RET-C634R cells were used as positive RNA was used as negative control (C) control. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 7 of 12 Fig. 4 Analysis of the biological effects exerted by RET-G533C and RET-P1047S mutants. Plasmids encoding wild type or mutant RET variants (G533C and P1047S) were transfected into HEK293 cells and selected in puromycin. a. MTT assay was performed at different time points (24 h, 48 h, 72 h and 96 h) using empty pBabe plasmid as negative control and plasmid encoding RET C634R mutant as positive control. Values are shown as bar graphs and all results are the average of two independent experiments performed in triplicate, Error bars s.d.; n =6; **p < 0.01; ***p < 0.001 compared with control. b. Colony formation assay was performed with HEK293 cells transfected with plasmids encoding wild type or mutant RET variants (G533C and P1047S). We used empty pBabe plasmid as negative control and plasmid encoding RET C634R mutant as positive control. Similar results were observed in three independent experiments Similar results were obtained with colony assay experi- Fig. 3 Immunofluorescence analysis of RET expression and ments. Also in this type of assay, HEK293-RET-G533C phosphorylation in colon cancer samples. a. Immunofluorescence cells showed greater proliferative potential in clonogenic as- analysis of RET expression in normal and tumor tissue samples says than HEK293-C cells, generating more colonies than from patients CC12 and CC20 shown as merged images of green control cells. Similarly, HEK293-RET-WT and HEK293- (RET staining) and blue (DAPI staining). Magnification 63X. Scale RET-P1047S cells showed an intermediate behaviour bar, 200px. b. Immunofluorescence analysis of RET phosphorylation at Y905 and Y1062 in normal and tumor tissue samples from (Fig. 4b). These results indicate that RET mutant patients CC12 and CC20 shown as merged images of green G533C, but not wild type or P1047S RET mutants (phospho-RET staining) and blue (DAPI staining). Magnification expressed at similar level, is able to stimulate anchorage- 63X. Scale bar 200px dependent proliferation of human epithelial cells in culture. Subsequently, immunoblot analysis was performed to Analysis of cell proliferation by MTT demonstrated check for RET expression and for the activation status of that, similar to HEK293-RET-C634R cells, HEK293- MAPK pathway in transfected HEK293 cells (Fig. 5). Ly- RET-G533C duplicated with an increased rate in mono- sates from transfected HEK293 cells and derivatives were layer compared with HEK293-C (p < 0.01). Conversely, analysed by immunoblot with anti-phosphoY1062 or HEK293-RET-WT and HEK293-RET-P1047S cells showed anti-RET antibody. In agreement with the data from a more limited increase, being intermediate between nega- MTT and colony assay experiments, the G533C RET tive (HEK293-C) and positive controls (HEK293-RET- mutant showed increased RET phosphorylation than C634R), see Fig. 4a. wild type RET or P1047S mutant (Fig. 5a). Subsequently, Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 8 of 12 Fig. 5 Analysis of the biochemical effects exerted by RET-G533C and RET-P1047S mutants. Immunoblot analysis of RET expression and activity in transfected HEK293 cells (a). Lysates from transfected HEK293 cells were analysed by immunoblot with anti-phosphoY1062 or anti-RET antibody. The activation status of the MAPK pathway was assessed with anti phospho-Y202T204 ERK1/2 antibody. Anti-ERK1/2 antibody was used for normalization. b. Immunoblot analysis of RET dimers in transfected HEK293 cells. Samples were prepared under non-reducing (NR) or reducing (R) conditions to detect RET dimer formation and loaded onto SDS-PAGE in non-reducing conditions. Half the total amount of protein lysate was loaded in the case of RET634 we assessed the activation status of critical signalling Cells expressing the RET G533C mutant are sensitive to pathways that function downstream RET [32], the treatment with the RET specific inhibitor vandetanib. MAPK pathway. The activity of the MAPK pathway was Finally, we determined whether the RET inhibitor assessed with anti phospho-Y202/T204 ERK1/2 antibody vandetanib was able to abolish the effects exerted by (Fig. 5a). Anti-total ERK1/2 antibody was used for RET-533C mutant in transfected HEK293 cells. Treat- normalization. We found that the adoptive expression of ment of HEK393 cells expressing RET mutants with RET mutant G533C, at difference with wild type or vandetanib induced a dose-dependent reduction (al- P1047S RET, induced a marked increase in the activation most 50% using 500 nM Vandetanib) in viable cell of MAPK pathway. Finally, since the mutation carried by number of HEK293-RET-G533C (and HEK293-RET- the active mutant involved a cysteine residue in the C634R) cells, demonstrating growth inhibitory activity extracellular domain, we investigated whether this amino in vitro (Fig. 7a). Similarly, subsequent immunoblot acidic change facilitated the formation of RET dimers. analysis demonstrated that vandetanib markedly re- To detect RET dimer formation we separated cell lysates duced phosphorylation of RET Y1062 and of ERK1/2 from transfected HEK293 cells by SDS-PAGE under ei- Y202/T204 in HEK293-RET-G533C and HEK293- ther non-reducing or reducing condition. As shown in RET-C634R cells (Fig. 7b). Fig. 5b, we found that RET mutants G533C was particu- larly efficient in generating RET dimers in transfected Discussion HEK293 cells comparable to the positive control (PC), In this manuscript, we have re-analysed NGS data gener- suggesting that the oncogenic mechanism whereby RET ated on a cohort of Italian patients affected by colon mutant G533C contribute to colon cancer development cancer to identify variants in genes encoding RTKs. The is through an increase in the amount of covalently main findings of this analysis were: i) somatic variants linked dimers induced by the presence of the unpaired were detected in 31 RTK encoding genes, the most fre- cysteine induced by the mutation. quent of which were FLT4, ROS1, EPH7, ERBB2, EGFR, We also found that RET-G533C mutant stimulated RET, FGFR3 and FGFR4, ii) the majority of the observed cell migration in vitro (Fig. 6). We investigated the ef- variants were variants of unknown significance (VUS), fects of the RET-G533C mutant on the migration cap- being only 10% present in the COSMIC database, iii) ability of HEK293 cells by using wound healing assay somatic mutations in the RET proto-oncogene were (Fig. 6a). RET-WT and RET-C634R were used as con- identified in 4 colon cancer patients (10%), iv) only one trols. The results of a triplicate set of independent exper- RET variants were apparently active and able to stimu- iments demonstrated that RET-G533C mutant enhanced late anchorage-dependent proliferation of HEK293 cells, migration of HEK-293 cells of about 70% at 48 h at dif- resembling MEN2A-like mutations, since they efficiently ference with RET-WT, which was much less efficient in generate RET dimers in transfected HEK293 cells. These promoting migration (Fig. 6b). issues will be commented more in detail below. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 9 of 12 Fig. 7 The RET-G533C mutant is inhibited by vandetanib. a. MTT assay was performed in the presence of two different concentrations of vandetanib. HEK293 cells expressing pBabe were used as negative Fig. 6 Effect of RET-G533C on cell migration. a. Wound healing assay control and HEK293 cells expressing RET-C634R mutant were used as of HEK293 cells expressing different RET mutants or empty pBABE positive control. Values are shown as bar graphs and all results are the vector. Cells were scratch-wounded with a micropipette tip (200 μl). average of experiments performed in triplicate. Statistical significance Images of cellular migration were taken at times 0 h, 24 h and 48 h compared with vehicle was evaluated by Two-Way ANOVA (with using the Leica DFC420 C and Leica Application Suite Software. multiple comparison Dunnet’stest) (n =3; **p <0.01; ***p <0.001). Magnification, 10X. b. Wound healing was quantified by the ImageJ b. Immunoblot analysis of RET/MAPK pathway (with anti-phosphoY1062 64 software using the area of the wound of cells expressing empty RET, anti-RET, anti phospho-Y202T204 ERK1/2, anti-ERK1/2 antibodies) of pBabe vector at T0 as reference value. Final results represent mean ± SD HEK293 cells transfected with RET-G533C mutant treated with vehicle or of three independent experiments and are indicated as fold change over vandetanib (500 nM) for 2 h. HEK293 cells transfected with RET-634 the control in terms of reduction of the wound. Statistical significance mutant were used as control was evaluated by Two-Way ANOVA (with multiple comparison Tukey’s test) confronting at each time point the different conditions and indicating the difference over the control (***p <0.001; ****p <0.0001) are VUS. Only few variants could be easily categorized as known pathogenic mutations, such as those present Upon re-analysis of the data reported in Oliveira et al. in the COSMIC database [FLT4 (G1131S, R1041Q, (under review) we have focused on genes encoding RTKs, E1026K), EGFR (R669*, W731*), ERBB2 (S310F), ERBB4 and have observed that RTKs are potentially involved in (R847H), IGF1R (I1151M), EPHB1 (A922T) genes]. This the initiation and/or progression of at least 75% of colon observation raises the issue of the clinical significance of cancer. In fact, 28 patients out of 37 presented variants in the variants identified, and consequently of the use that at least one gene encoding RTKs. Co-occurrence of vari- oncologists should do to patients with colon cancer that ants in different RTK genes was observed in 20 tumors carry VUS in RTK genes. Although 30/101 of the vari- out of 37 (54%), with several samples showing variants in ants identified in this study promoted amino acid multiple RTK genes simultaneously (CC27, 11 genes; changes within the catalytic domains of the correspond- CC34, 13 genes; CC12, 8 genes), which raised the issue of ing RTKs, the ability to provide a clinical interpretation whether these variants were indeed functionally relevant. of VUS represents a daunting task because clinical or An important issues raised by this work is the finding experimental data that aid interpretation are rarely avail- that the vast majority of variants identified in RTK genes able. The only valid alternative for clinicians is to Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 10 of 12 evaluate significance by the use of computational tools Y202/T204 in parallel with a marked reduction of the that predict the biologic effects of novel variants [33]on proliferative effects induced by the RET G533C variant. the basis of available literature. In this manuscript we These results indicate that RET mutant G533C, but have used two different prediction algorithms, SIFT and not wild type or P1047S RET expressed at similar level, Polyphen2, to define a variant as pathogenic. However is able to stimulate proliferation and migration of human the oncogenic role of VUS remains unclear. epithelial cells in culture. In addition, the finding that The choice of RET for subsequent validation of RTK the sample with RET G533C variant presented variants variants was due to the fact that, at difference with what in the downstream effectors like KRAS or MAPK happens in thyroid and lung cancer, the involvement of whereas the samples that carries the RET P270L or the RET proto-oncogene to the development of colon can- RET P1047S variants presented also variants in other cer is still debated. On one hand, it has been proposed genes (KRAS, AKT1 and BRAF, PTEN, AKT1 and that RET is a potential tumor suppressor gene in colo- FGFR3, respectively) indicated that the RET G533C rep- rectal cancer [34]. RET is expressed in normal mucosa, resents a real gain-of-function mutation. and its expression is lost in adenomas and adenocarcin- omas because of an aberrantly methylation of CpG Conclusions islands within its promoter [34]. Notably, RET promoter In conclusion, we have re-analysed NGS data generated methylation is associated with poor prognosis in stage II on 37 colon cancer patients by use of Comprehensive and III colorectal cancer patients [35]. Moreover, al- Cancer Panel, focusing on genes encoding RTKs. Over- though RET mutations (V145G, R360W or G593E) in all, we have observed 101 different potentially damaging colorectal cancer have been identified [36] and are re- variants distributed across 31 RTK genes in 28 patients, ported in the COSMIC database, these mutations appar- suggesting that the remaining tumors represent a subset ently inactivate RET [34]. On the other hand, Le Rolle of colon cancer that may develop through pathways that and co-workers have recently reported on the identifica- are independent of RTK signalling. However, only few tion and characterization of RET fusions that juxtapose variants could be unambiguously categorized as patho- the C-terminal RET kinase domain to the N-terminal do- genic mutations, being the vast majority of RTK variants main of CCD6 or NCOA4 (CCDC6-RET and NCOA4- identified in this study VUS, whose clinical significance RET fusions) in advanced colorectal cancer. Importantly, a remains unclear. This is best exemplified by the validation patient carrying the CCDC6-RET fusion showed a thera- analysis performed on 2 different RET variants. The results peutic response when treated with the RET kinase inhibi- presented here indicate that RET variant G533C is clearly tor regorafenib [37]. oncogenic whereas RET variant P1047S is not. Yet, the In this manuscript, we have investigated whether the identification of gain-of-function RET variants in colon RET variants identified in colon cancer patients are able cancer patients provides novel data for the involvement of to confer pro-tumorigenic properties to epithelial cells. this RTK in the development of colon cancer and identifies We observed that HEK293 cells expressing RET-G533C, RET as a novel putative target for targeted therapy. but not the same cells expressing RET-WT or RET- P1047S, duplicated with an increased rate in monolayer Additional files and in clonogenic assays and promoted migration com- pared with control cells. Accordingly, the active RET Additional file 1: Table S1. Clinical-pathological characteristics of the 37 colon cancer patients included in the study. (XLSX 12 kb) mutant detected in colon cancer patients (G533C) may Additional file 2: Table S2. Summary of the variants identified in RTKs. resemble those described in MEN2A patients, of which (XLSX 16 kb) the C634R is the prototype [38]. In fact, RET mutant Additional file 3: Table S3. List of mutated RTKs in each patient G533C was particularly efficient in generating RET di- included in the study. (XLSX 11 kb) mers in transfected HEK293 cells, suggesting that the Additional file 4: Table S4. Sequencing metrics of the identified RET oncogenic mechanism whereby this RET mutant con- mutations. (XLSX 10 kb) tributes to colon cancer development occurs through an Additional file 5: Table S5. List of mutations identified in CC20 after the next generation sequencing analysis pipeline. (XLSX 17 kb) increase in the amount of covalently linked dimers in- duced by the presence of the unpaired cysteine caused by the mutations. Such increased amount of dimers Abbreviations CCP: Comprehensive Cancer Panel; CRC: Colorectal cancer; HEK293: Human would turn into increased activity of RET receptors (de- embryonic kidney cells 293; HSNCC: Head and neck squamous-cell carcin- tected as increased RET phosphorylation), which in turn oma; IGV: Integrated Genomic Viewer; MAPK,: Mitogen-activated protein would cause activation of the MAPK signalling pathway kinase; NGS: Next Generation Sequencing; PTKs: Protein tyrosine kinases; RTKs: Receptor-type tyrosine kinases; SDS-PAGE: Sodium Dodecyl Sulphate - (detected as increased ERK1/2 phosphorylation). PolyAcrylamide Gel Electrophoresis;; SIFT: Scale-invariant feature transform; It is of note that the RET inhibitor vandetanib abol- UICC: Union for International Cancer Control; VUS: Variants of unknown ished phosphorylation of RET Y1062 and of ERK1/2 significance Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 11 of 12 Acknowledgments transforming sequence in papillary thyroid carcinoma. Genes Chromosomes Not applicable. Cancer. 1994;9:244–50. 9. Soda M, Choi YL, Enomoto M, Takada S, Yamashita Y, Ishikawa S, Fujiwara S, Watanabe H, Kurashina K, Hatanaka H, et al. Identification of the Funding transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature. This work was supported by Centro di Genomica Funzionale e Patologia 2007;448:561–6. Molecolare of University Magna Graecia of Catanzaro and by PON01_02782 10. Rowley JD. Letter: a new consistent chromosomal abnormality in chronic to GV. myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature. 1973;243:290–3. Availability of data and materials 11. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL. Human All data generated or analysed during this study are included either in this breast cancer: correlation of relapse and survival with amplification of the article or in the Additional files. HER-2/neu oncogene. Science. 1987;235:177–82. 12. Herman SE, Gordon AL, Hertlein E, Ramanunni A, Zhang X, Jaglowski S, Authors’ contributions Flynn J, Jones J, Blum KA, Buggy JJ, et al. Bruton tyrosine kinase represents GV, DM and AR designed the research plan; RS and AR performed the a promising therapeutic target for treatment of chronic lymphocytic operations; DMO, KG, RL, FB and CM performed experiments; CM and RL leukemia and is effectively targeted by PCI-32765. Blood. 2011;117:6287–96. contributed to the collection of biological samples; CM, AR and MF 13. Cai H, Babic I, Wei X, Huang J, Witte ON. Invasive prostate carcinoma driven contributed to the collection and analysis of clinical data; CL, performed by c-Src and androgen receptor synergy. Cancer Res. 2011;71:862–72. bioinformatic data analysis; DMO, GV, DM, FC prepared manuscript draft and 14. Memarzadeh S, Xin L, Mulholland DJ, Mansukhani A, Wu H, Teitell MA, Witte participated in manuscript editing and review. All authors read and ON. Enhanced paracrine FGF10 expression promotes formation of multifocal approved the final manuscript prostate adenocarcinoma and an increase in epithelial androgen receptor. Cancer Cell. 2007;12:572–85. Ethics approval and consent to participate 15. Cartwright CA, Kamps MP, Meisler AI, Pipas JM, Eckhart W. pp60c-src All protocols performed in studies involving human participants were activation in human colon carcinoma. J Clin Invest. 1989;83:2025–33. reviewed and approved by the Institutional Review Board of the Mater 16. Grant S. FAM83A and FAM83B: candidate oncogenes and TKI resistance Domini/University Magna Graecia in the meeting of May 21st 2014. mediators. J Clin Invest. 2012;122:3048–51. 17. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer Consent for publication statistics, 2012. CA Cancer J Clin. 2015;65:87–108. Not applicable. 18. Boland PM, Fakih M. The emerging role of neoadjuvant chemotherapy for rectal cancer. J Gastrointest Oncol. 2014;5:362–73. Competing interests 19. Lubner SJ, Loconte NK, Holen KD, Schelman W, Thomas JP, Jumonville A, The authors declare that they have no competing interests. Eickhoff JC, Seo S, Mulkerin DL. A phase II study of oxaliplatin, 5-fluorouracil, leucovorin, and high-dose capecitabine in patients with metastatic colorectal cancer. Clin Colorectal Cancer. 2010;9:157–61. Publisher’sNote 20. Ramakrishnan MS, Eswaraiah A, Crombet T, Piedra P, Saurez G, Iyer H, Arvind Springer Nature remains neutral with regard to jurisdictional claims in AS. Nimotuzumab, a promising therapeutic monoclonal for treatment of published maps and institutional affiliations. tumors of epithelial origin. MAbs. 2009;1:41–8. 21. Siena S, Sartore-Bianchi A, Di Nicolantonio F, Balfour J, Bardelli A. Biomarkers Author details predicting clinical outcome of epidermal growth factor receptor- Department of Experimental and Clinical Medicine, University Magna targeted therapy in metastatic colorectal cancer. J Natl Cancer Inst. Graecia of Catanzaro, Campus Salvatore Venuta -Viale Europa, Catanzaro 2009;101:1308–24. 88100, Italy. Department of Health Sciences, University Magna Graecia of 22. Bardelli A, Siena S. Molecular mechanisms of resistance to cetuximab and Catanzaro, Campus Salvatore Venuta - Viale Europa, Catanzaro 88100, Italy. panitumumab in colorectal cancer. J Clin Oncol. 2010;28:1254–61. Department of Medical and Surgical Sciences, University Magna Graecia of 23. Hubbard J, Grothey A. Antiangiogenesis agents in colorectal cancer. Curr Catanzaro, Campus Salvatore Venuta - Viale Europa, Catanzaro 88100, Italy. Opin Oncol. 2010;22:374–80. Department of Molecular Medicine and Medical Biotechnologies, University 24. Marques I, Araujo A, de Mello RA. Anti-angiogenic therapies for metastatic Federico II, Naples, Italy. UOC Chirurgia Generale, Azienda Ospedaliera dei colorectal cancer: current and future perspectives. World J Gastroenterol. Colli, Naples, Italy. 2013;19:7955–71. 25. Mulder K, Scarfe A, Chua N, Spratlin J. The role of bevacizumab in colorectal Received: 20 December 2017 Accepted: 2 April 2018 cancer: understanding its benefits and limitations. Expert Opin Biol Ther. 2011;11:405–13. 26. UICC. TNM Classification of Malignant Tumours, 7th Edition. 7th edn. References Chichester: Wiley; 2009. 1. Arora A, Scholar EM. Role of tyrosine kinase inhibitors in cancer therapy. J 27. Carlomagno F, Vitagliano D, Guida T, Ciardiello F, Tortora G, Vecchio G, Ryan Pharmacol Exp Ther. 2005;315:971–9. AJ, Fontanini G, Fusco A, Santoro M. ZD6474, an orally available inhibitor of 2. Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature. 2001;411: KDR tyrosine kinase activity, efficiently blocks oncogenic RET kinases. Cancer 355–65. Res. 2002;62:7284–90. 3. Robinson DR, Wu YM, Lin SF. The protein tyrosine kinase family of the 28. Carlomagno F, Salvatore G, Cirafici AM, De Vita G, Melillo RM, de Franciscis human genome. Oncogene. 2000;19:5548–57. V, Billaud M, Fusco A, Santoro M. The different RET-activating capability of 4. Corless CL, Barnett CM, Heinrich MC. Gastrointestinal stromal tumours: mutations of cysteine 620 or cysteine 634 correlates with the multiple origin and molecular oncology. Nat Rev Cancer. 2011;11:865–78. endocrine neoplasia type 2 disease phenotype. Cancer Res. 1997;57:391–5. 5. da Cunha SG, Shepherd FA, Tsao MS. EGFR mutations and lung cancer. 29. Da Silva AM, Maciel RM, Da Silva MR, Toledo SR, De Carvalho MB, Cerutti Annu Rev Pathol. 2011;6:49–69. JM. A novel germ-line point mutation in RET exon 8 (Gly(533)Cys) in a large 6. Yamamoto Y, Kiyoi H, Nakano Y, Suzuki R, Kodera Y, Miyawaki S, Asou N, kindred with familial medullary thyroid carcinoma. J Clin Endocrinol Metab. Kuriyama K, Yagasaki F, Shimazaki C, et al. Activating mutation of D835 2003;88:5438–43. within the activation loop of FLT3 in human hematologic malignancies. Blood. 2001;97:2434–9. 30. Kjaer S, Kurokawa K, Perrinjaquet M, Abrescia C, Ibanez CF. Self-association 7. Hirota S, Isozaki K, Moriyama Y, Hashimoto K, Nishida T, Ishiguro S, Kawano of the transmembrane domain of RET underlies oncogenic activation by K, Hanada M, Kurata A, Takeda M, et al. Gain-of-function mutations of c-kit MEN2A mutations. Oncogene. 2006;25:7086–95. in human gastrointestinal stromal tumors. Science. 1998;279:577–80. 31. Bonfil RD, Reddel RR, Ura H, Reich R, Fridman R, Harris CC, Klein-Szanto JP. 8. Sozzi G, Bongarzone I, Miozzo M, Borrello MG, Blutti MG, Pilotti S, Della Invasive and metastatic potential of a v-ha-ras-transformed human Porta G, Pierotti MA. A t(10;17) translocation creates the RET/PTC2 chimeric bronchial epithelial cell line. J Natl Cancer Inst. 1989;81:587–94. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 12 of 12 32. Mulligan LM. RET revisited: expanding the oncogenic portfolio. Nat Rev Cancer. 2014;14:173–86. 33. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, Grody WW, Hegde M, Lyon E, Spector E, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17:405–24. 34. Luo Y, Tsuchiya KD, Il Park D, Fausel R, Kanngurn S, Welcsh P, Dzieciatkowski S, Wang J, Grady WM. RET is a potential tumor suppressor gene in colorectal cancer. Oncogene. 2013;32:2037–47. 35. Draht MX, Smits KM, Tournier B, Jooste V, Chapusot C, Carvalho B, Cleven AH, Derks S, Wouters KA, Belt EJ, et al. Promoter CpG island methylation of RET predicts poor prognosis in stage II colorectal cancer patients. Mol Oncol. 2014;8:679–88. 36. Sjoblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD, Mandelker D, Leary RJ, Ptak J, Silliman N, et al. The consensus coding sequences of human breast and colorectal cancers. Science. 2006;314:268–74. 37. Le Rolle AF, Klempner SJ, Garrett CR, Seery T, Sanford EM, Balasubramanian S, Ross JS, Stephens PJ, Miller VA, Ali SM, Chiu VK. Identification and characterization of RET fusions in advanced colorectal cancer. Oncotarget. 2015;6:28929–37. 38. Zou HJ, Li YS, Shan ZY. Genetic analysis and clinical investigation of a pedigree with multiple endocrine neoplasia type 2A: a case report. Oncol Lett. 2016;12:2657–9. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Experimental & Clinical Cancer Research Springer Journals

Next-generation sequencing analysis of receptor-type tyrosine kinase genes in surgically resected colon cancer: identification of gain-of-function mutations in the RET proto-oncogene

Free
12 pages

Loading next page...
 
/lp/springer_journal/next-generation-sequencing-analysis-of-receptor-type-tyrosine-kinase-BvjE4t5AVc
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Oncology; Cancer Research; Immunology; Apoptosis
eISSN
1756-9966
D.O.I.
10.1186/s13046-018-0746-y
Publisher site
See Article on Publisher Site

Abstract

Background: Improvement in genetic characterization of Colon Cancer (CC) patients is required to propose new potential targets, since surgical resection coupled to chemotherapy, presents several limits such as cancer recurrence and drug resistance. Targeted therapies have more efficacy and less toxicity than standard treatments. One of the most relevant cancer-specific actionable targets are receptor tyrosine kinases (RTKs) whose role in CC need to be better investigated. Methods: We have analysed 37 CC patients using the Ion AmpliSeq™ Comprehensive Cancer Panel (CCP). We have confirmed the somatic nature of RET variants through Sanger sequencing and assessed RET activation status and protein expression by immunofluorescence and western-blot analyses. We have used RET mutant expression vectors to evaluate the effect of selected mutations in HEK293 cells by performing proliferation, migration and clonogenic assays. Results: Among the 409 cancer-related genes included in the CCP we have focused on the RTKs. Overall, we have observed 101 different potentially damaging variants distributed across 31 RTK genes in 28 patients. The most frequently mutated RTKs were FLT4, ROS1, EPH7, ERBB2, EGFR, RET, FGFR3 and FGFR4. In particular, we have identified 4 different somatic variants in 10% of CC patients in RET proto-oncogene. Among them, we have demonstrated that the G533C variant was able to activate RET by promoting dimer formation and enhancing Y1062 phosphorylation. Moreover, we have demonstrated that RET G533C variant was able to stimulate anchorage-dependent proliferation, migration and clonogenic cell survival. Notably, the effects induced by the RET G533C variant were abolished by vandetanib. Conclusions: The discovery of pathogenic variants across RTK genes in 75% of the CC patients under analysis, suggests a previously underestimated role for RTKs in CC development. The identification of a gain-of-function RET mutation in CC highlights the potential use of RET in targeted therapy. Keywords: Next-generation sequencing, Receptor-type tyrosine kinases, Colon cancer, RET proto-oncogene * Correspondence: viglietto@unicz.it; malanga@unicz.it Department of Experimental and Clinical Medicine, University Magna Graecia of Catanzaro, Campus Salvatore Venuta -Viale Europa, Catanzaro 88100, Italy Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 2 of 12 Background effectors downstream EGFR such as KRAS, BRAF, Targeted therapy of cancer has more efficacy and less PIK3CA and PTEN [22]. Also the patients that initially toxicity than conventional chemotherapy because it is respond to the anti-VEGF therapy [23, 24] will eventu- directed against cancer-specific molecules and/or signal- ally show evidence of resistance to anti-angiogenic ther- ling pathways. So far the most relevant cancer-specific apy [25]. actionable targets are protein tyrosine kinases (PTKs) Clearly, we need to improve the molecular cha- [1]. PTKs regulate several cellular processes that include racterization of tumors from CRC patients in order to cell proliferation, cell death, motility and invasion [2]. identify the molecular alterations inside cancer cell ge- When subverted, these processes may give origin to nomes that drive cancer growth and determine the inef- cancer [2]. ficacy of anti-EGFR/VEGF targeted therapies. In a The catalogue of genes encoding PTKs within the hu- previous work we have reported the analysis of 37 colon man genome has been updated to 90 genes, of which 58 cancer patients using the Ion AmpliSeq™ Comprehen- are receptor type PTKs and 32 are non-receptor type sive Cancer Panel on the Ion Proton platform (Oliveira PTKs [3].Genes encoding PTKs - both receptor and et al., under review). In this manuscript, we have fo- non-receptor - can be oncogenically activated through a cused the analysis on the role of genes encoding RTKs variety of mechanisms that include gain-of-function mu- as potentially targeting genes in colon cancer. The dis- tations as in the case of EGFR in lung cancer and KIT in covery of pathogenic variants in this gene family in the GIST [4–7], translocations as in the case of BCR-ABL in majority of colon cancer patients suggests a previously CML, ALK in lung cancer and RET in thyroid, lung and underestimated role for RTKs in the development of colorectal cancer [8–10] and amplifications as in the colon cancer and thus the possibility to treat those pa- case of ERBB2 in breast cancer and EGFR in lung, tients with specific RTK inhibitors. In particular, the HSNCC and colon cancer [11–15]. identification of gain-of-function RET variants in a sub- In any case, the constitutive signalling from PTKs set of colon cancer, as described here, has led to the leads to aberrant activation of downstream pathways identification of a small subset of patients potentially that promote continued cell growth and survival. Cancer treatable with RET inhibitors. cells are particularly sensitive to intracellular pathways that are improperly hyperactive, a phenomenon known Methods as oncogene/pathway addiction. This phenomenon pro- Tumor samples vides a unique therapeutic opportunity to block cancer Patient accrual was conducted according to Institutional by inhibiting PTK activity [16]. Review Board of the Mater Domini/University Magna Colorectal cancer (CRC) is one of the most commonly Graecia (Catanzaro, Italy). The study was approved by diagnosed and lethal cancers worldwide, with 1.4 million the Institutional Review Board of the Mater Domini/ new cases and 690,000 deaths in 2012 [17]. Prognosis University Magna Graecia in the meeting of May 21st of CRC patients depends on cancer stage at diagnosis. 2014. Tumor, normal mucosa and peripheral blood sam- Detection of CRC at early stage may confer a 90% 5- ples were obtained from patients referring to General year survival rate, compared to 12% if distant metasta- Surgey Unit of University Hospital Magna Graecia of sis has occurred. Catanzaro (Catanzaro, Italy), who underwent surgical Current therapy for colon cancer patients include resection for colon cancer since January 2013. Biopsies surgical resection coupled to chemotherapy [18]. In ad- were immediately snap frozen and stored at − 80 °C. vanced disease, either FOLFOX (oxaliplatin/5-fluoro- In all cases diagnosis was confirmed by reviewing uracil [5-FU]/leucovorin [LV]) or FOLFIRI (irinotecan/ hematoxylin/eosin-stained slides. LV/5-FU) are typical regimens for first-line chemother- apy, but cancer recurs in many patients [19]. Targeted Patients’ demographics therapy is suggested for patients with metastatic disease General demographic information, clinical findings, sur- [18]. In these patients the use of monoclonal antibodies gical treatment, histo-patological examination and to epidermal growth factor receptor (EGFR) (cetuxi- follow-up data were collected for each patient. Detailed mab, panitumumab) and/or to vascular endothelial documentation of the histo-pathological findings allowed growth factor (VEGF) (bevacizumab) has increased the classification according to the current edition of UICC median survival beyond 20 months [20]. However, the [26]. None of the patients received chemotherapy or ra- responses to EGFR-targeted antibodies are relatively diation therapy prior to surgery. See Additional file 1 for low, with improvement in survival rates lasting only complete characteristics of patients. Briefly, among the few months [21]. 37 patients, 13 were women and 24 were males. The In addition, the efficacy of anti-EGFR antibodies is mean age was 68 years old (range 47–84). Stage was limited to patients with tumors lacking mutations in known for 36 of the 37 patients: 7 patients had stage I Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 3 of 12 disease, 13 patients had stage II disease, 12 patients had phospho-RET Y905 (rabbit polyclonal antibody anti- stage III disease and 4 patients had stage IV disease. pRET, 1:100 diluition, #3221 Cell Signaling Technology) Grade was known for 35 out of 37 patients: 26 patients or phospho-RET Y1062 (rabbit polyclonal antibody anti had tumors that were graded G2 and 9 patients had tu- p-RET, 1:100 dilution) [27] for 60 min, followed by incu- mors that were graded G3. Four patients presented bation with secondary Alexa Fluor 488F fragment of goat distant metastasis. anti-rabbit IgG (H + L) antibody; 1:500 dilution (Thermo- Fisher Scientific, MA, USA #A11070) for 45 min. Cells Next generation sequencing were counterstained with DAPI,1:500 dilution, (Thermo- Sequencing was performed with the Ion AmpliSeq™ Fisher Scientific, MA, USA), mounted using anti-fade Comprehensive Cancer Panel on the Ion Proton system mounting medium #53023 (DAKO, Glostrup, Denmark) (ThermoFisher Scientific, MA, USA) starting from 40 ng and observed at Fluorescence Microscopy (Leica Micro- of DNA and following standard protocols. Primary bio- systems, Wetzlar Germany). informatic analysis of data is described in a previous manuscript (Oliveira et al., under review). We set an Protein extraction and immunoblot average mean coverage of > 300, a coverage of > 40 for Protein extracts were prepared with lysis buffer contain- each variant and variant frequency > 5%. To further filter ing 50 mM HEPES pH 7.5, 5 mM EDTA, 250 mM NaCl, the number of non-pathogenic variants identified in the 1 mM dithiothreitol, 0.5% Nonidet P40, 1 mM Na VO , 3 4 study the following criteria were used: i) variants with 1 mM NaF supplemented with 10 μg of aprotinin/ml, quality score ≤ 30 were excluded, ii) variants that were 10 μg of leupeptin/ml, 1 mM PMSF and a mix of prote- not annotated as pathogenic or likely pathogenic by ase inhibitors (SIGMAFAST protease inhibitor Tablets SIFT and Polyphen2 algorithms were excluded, iii) for general Use; Sigma-Aldrich St. Louis, MO, USA). Ly- variants detected in both normal and tumor samples sates were centrifuged at 13,000 rpm for 30 min at 4 °C were excluded. Finally, manual observation of the and the supernatants were collected. Protein concentration variants using the Integrated Genomic Viewer (IGV – was estimated with a modified Bradford assay (Bio-Rad Broad Institute) was performed to exclude inconclusive Laboratories, Berkeley, CA, USA). Western blot analysis variants. was carried out by standard methods. To detect RET dimers, electrophoresis was carried out Sanger sequencing on a 6% SDS–PAGE in non-reducing conditions. In this DNA obtained from matched normal and tumor sam- case proteins extracts were prepared in the same lysis ples was amplified using specific primers. The purified buffer as above devoid of dithiothreitol, and with sample products were sequenced using BigDye terminator v3.1 loading buffer devoid of 2-mercaptoethanol. Samples (Applied Biosystems, Foster City, CA) with ABI 3100 were not heated upon loading. Genetic Analyser (Applied Biosystems, Foster City, CA). Proteins were revealed by enhanced chemilumines- cence detection using Clarity™ Western ECL Substrate RNA extraction and reverse-transcription (Bio-Rad Laboratories, Berkeley, CA, USA). The anti- Tumors obtained at surgery were snap frozen in liquid bodiesusedinthisstudy were:anti-RET(E1N8X, nitrogen and conserved at − 80 °C. RNA was prepared #14556), anti-phospho-RET, (#3221), anti-phospho- using Trizol (ThermoFisher Scientific, MA, USA) and ERK1/2 (#9101), anti-ERK1/2 (#9107) purchased from purified using RNeasy mini-eluate clean-up kit (QIAGEN Cell Signaling Technology, Danver, MA, USA. The Inc. CA, USA). cDNA was retro-transcribed with random anti-phospho-RET Y1062 used in this study has been hexamers using SuperScript®III First-Strand Synthesis previously described [27]. System (ThermoFisher Scientific, MA, USA). Construction of RET mutants expression vectors Immunostaining RET mutants were obtained by site-specific mutagenesis Deparaffinized 4-μm sections were hydrated in decreas- of RET51-WT construct (pBabe vector encoding for the ing ethanol gradient solutions and rinsed in wash solu- proto-RET gene long isoform). The RET51-C634R plas- tion (TBST, 0.05 mol/L Tris Buffered Saline with mid, used as positive control, is described elsewhere [28]. Tween20). Antigen retrieval was performed with citrate RET51-G533C and RET51-P1047S were obtained by site- buffer pH 6 for antibodies to RET and phospho-RET directed mutagenesis of RET51-WT using an in vitro Y905, and pH 9 for antibody to phospho-RET Y1062, re- oligonucleotide mutagenesis system (Quik-Change XL spectively, for 20 min at 98 °C followed by washing in site-directed mutagenesis; Agilent Technologies, CA, TBST. Sections were incubated with antibodies against USA). Plasmid DNA was extracted using the QIAGEN RET (rabbit monoclonal antibody anti-RET, 1:250 dilu- Plasmid Maxi Kit (QIAGEN CA, USA) as suggested by tion, #14556 Cell Signaling Technology, Danver, MA), the supplier. The presence of the specific mutations was Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 4 of 12 verified by DNA direct sequencing. Mutant plasmids were Results sequenced entirely to exclude the presence of additional Mutations in genes encoding RTKs in colon cancer mutations. In a previous work, we have reported the analysis of 37 colon cancer patients using the Ion AmpliSeq™ Com- prehensive Cancer Panel (CCP) (Oliveira et al., under Transfections and colony formation assay review). In that work, patients who had undergone sur- Human HEK293T cells were maintained in Gibco™ gery for colon cancer at the General Surgery Unit of DMEM with 10% FBS and 1% penicillin/streptomycin University Magna Graecia of Catanzaro in the years (ThermoFisher Scientific, MA, USA). Empty pBABE vec- 2013–2015 were studied. Complete demographic and tor and/or plasmids encoding RET wild type or mutant clinical information of patients are reported in Additional alleles were transiently transfected using Lipofectamine file 1. DNA was extracted from matched normal and 3000 (ThermoFisher Scientific, MA, USA) according to pathological tissues and subjected to NGS analysis on the the manufacturer’s instructions. Forty-eight hours after Ion Torrent platform using the CCP (ThermoFisher transfection, cells were trypsinized and plated into 100-mm Scientific, MA, USA), that provides complete exon cover- plates. Transfected cells were selected with 0.75 μg/ml age of the 409 most important cancer-associated genes. puromycin for 1 week until all cells in the control plates As indicated in Oliveira et al., (under review), the se- were dead. Then cells were trypsinized, counted and 5000 6 quencing performance achieved was: 13 × 10 mean cells were plated into 60-mm plates. Cells were kept in number reads/sample, 791 (range 102.5–2656) mean DMEM with 10% bovine serum. The culture medium was sequence coverage depth of targeted exonic regions of changed every 3–4 days. Approximately 10 days after trans- the 409 genes analysed, with median uniformity of fection, cells were stained by crystal violet. The colony for- sequenced genes 97% (range 0.7–0.99). Common germ- mation experiments were repeated three times. line variants were removed by filtering sequentially through a pool of peripheral blood samples available from some patients (n = 12), the dbSNP141 and the MTT assay 1000 Genomes Project data sets. Variants were function- Stably transfected cells expressing RET wild type and ally annotated using the algorithms Sift and Polyphen2 mutant alleles were seeded in 96 wells round bottom to predict their pathogenicity. plates (1000 cells/well). After 24 h, 10 μl of MTT [3- In this manuscript, we have re-analysed the data gen- (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium brom- erated in the manuscript from Oliveira et al., focusing ide]; (Sigma-Aldrich St. Louis, MO, USA) was dissolved on genes encoding receptor-type tyrosine kinase (RTK). at the concentration of 5 mg/ml in warm assay medium The number of RTK genes contained in the Comprehen- was added to cells at the indicated time points (24 h, sive Cancer Panel that were object of the present study 48 h, 72 h and 96 h) to yield a final assay volume of is 31. Overall, upon analysis of the data reported in Oli- 110 μl/well. Plates were incubated for 4 h at 37 °C and veira et al. we have observed 101 different potentially 5% CO . After incubation, supernatants were removed, damaging variants distributed across 31 RTK genes. See and 100 μl of isopropanol was added. Plates were placed Additional file 2 for a general summary of the analysis. on an orbital shaker for 5 min, and the absorbance was We have found that 28 patients out of 37 presented vari- recorded at 570 nm. Triplicates were made for each time ants in at least one gene encoding RTK, with the point and two independent experiments were conducted. remaining 9 patients showing no variant in any of the 31 Statistical analysis was performed by two-way ANOVA RTK genes. The mean number of mutated RTK genes/ and Dunnett’s multiple comparison test with GraphPad patient was 3.4 (range 1–13). Of the 28 samples that Software (La Jolla, CA, USA). presented mutations in RTK genes, 8 patients had only one mutated gene, 6 had two different mutated genes Wound healing assay and 14 presented 3 or more mutated genes. See Stably HEK293 cells expressing RET wild type or mutant Additional file 3 for patient-by-patient analysis of the re- alleles were seeded in a 60 mm plates and cultured until sults. Among RTK genes that presented variants, we ob- reaching 90% confluence. A micropipette tip (200 μl) was served variants in FLT4 (G1131S, R104Q, E1052K, used to make a scratch and simulate a wound. Cells were D1003N, D728N, P707L, R362L, R282*, Q213*, G180R, monitored after 0 h, 24 h and 48 h from the scratch and P138S, E36* and 3 frameshifts) in 10 patients, ROS1 images of wound healing were captured (magnification of (S277Y, P437L, S653F, T804 N, Q1127*, G1709C and 10X) using the Leica DFC420 C and Leica Application QG1708HC) in 7 patients, EPHA7 (D839G, P805L and Suite Software (Leica Microsystems, Wetzlar Germany). P278S) in 6 patients, RET (R77C, P270L, G533C, P1047S) Subsequently, cell migration was quantified by measuring in 4 patients and MET (A320V, R988C) in 2 patients. In the wound opening area with ImageJ64 software. addition, ERBB2 variants (S310F, W482C, Q533*, T631A, Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 5 of 12 S819F, A1216D) were detected in 6 patients, EGFR variant 30.5%. Raw data showing the detection of the (R669*, W731*, D807E, L933P) and ERBB4 (R983K, different RET mutations by NGS are contained in M977 V, R847H, T244I) variants were detected in 4 pa- Additional file 4. tients, and ERBB3 variants (T389I, V850 M) were detected Interestingly, mutation G533C is located in the RET in 2 patients. FGFR1 (D784E, A131V) and FGFR2 extracellular domain and consists in the replacement of (G302 K, L104P) variants were detected in 2 patients, a glycine amino acid with a cysteine as often observed in whereas FGFR3 (D320N, A352E, A571V, T653I) and families with familial medullary thyroid carcinoma FGFR4 variants (R130C, KE144RG, F859 L, T912I) were (FMTC) or Multiple Endocrine Neoplasia (MEN) 2A. detected in 4 patients. Eight tumors presented variants in Notably, the G533C substitution was identified in pa- only one RTK gene (FLT4, EGFR, ERBB4, ROS1, and tients with FMTC [29], suggesting that it represents a EPHA7), which suggests that these variants can represent gain-of-function mutation. For the complete list of mu- driving mutations. Accordingly, the variants detected in tated genes in the patient harboring the G533C substitu- the EGFR, ERBB4 and FLT4 were present in the COSMIC tion see Additional file 5. First we confirmed the results database. Conversely, co-occurrence of variants in differ- obtained through NGS by traditional Sanger sequencing. ent genes encoding RTKs was frequent, occurring in 20 We performed DNA sequencing of RET exons on tumors outof37analysed(54%).Co-occurrenceof vari- matched normal and cancer samples to confirm the ants included 2 different genes (FGFR4/NTRK1, ERRB3/ presence of the specific mutations identified by NGS FLT3, ALK/ERBB4, EPHA3/EPHA7, EPHB4/ROS1, and (See Fig. 1). Among the 4 different RET mutants identi- FLT4/LTK) in 6 cases and three or more genes in the fied in colon cancer patients we excluded the P270L remaining 14 cases. Notably, some samples showed vari- variant because it represented less than 10% of the cell ants in numerous genes encoding RTK (CC27, 11 genes; population and the R77C since it has been observed in a CC34, 13 genes; CC12, 8 genes). case of Hirschsprung’s disease [30]. Conversely, variants G533C and P1047S were included in the subsequent Identification of RET mutations in colon cancer by NGS analysis. We found that in the two patients, mutations and validation of NGS results were cancer-specific and were not found in the corre- Among the RTKs showing variants identified in this sponding normal tissue. These findings led us to exclude study, we decided to biologically validate the variants the presence of germ-line mutations associated with identified in the gene encoding the RET receptor. inherited diseases such as FMTC, MEN2 type 2A, The reasoning for this choice was that, at difference MEN2 type 2B and/or Hirschsprung’s disease. Con- with what happens in thyroid and lung cancer, the in- versely mutations were somatically acquired and were volvement of RET proto-oncogene to the develop- apparently selected for within the bulk of tumor cells ment of colon cancer is unclear. Variants in the RET since they represented 20–49% of RET alleles, respect- gene were identified in four colon cancer patients ively. We analysed RET expression in the tumors and in (patients CC12, CC20, CC25 and CC33), as indicated the corresponding normal mucosa from patients CC12 in Additional file 2. and CC20 where RET mutations were identified. In the present study, we have identified a C/T transi- The expression of RET in the samples harbouring tion that led to the replacement of P1047 with S RET mutations was evaluated by semi-quantitative (P1047S) in the cytoplasmic tail of RET in patient CC12, reverse transcriptase-polymerase chain reaction (RT- a C/T transition that led to the replacement of P270 PCR) (Fig. 2) and by in situ immunofluorescence ana- with L (P270L) in the RET extracellular domain in pa- lysis (Fig. 3a). RT-PCR analysis revealed the presence of tient CC33, an A/T transversion that led to the replace- a RET transcript in normal mucosa and the corre- ment of G533 with C (G533C) in the cysteine-rich sponding cancer in all patients analysed. The detection domain of RET protein in patient CC20 and a G/A tran- of RET transcript by RT-PCR mRNA was corroborated sition that led to the replacement of R77 with C (R77C) by analysis of the RET protein in normal and cancerous in the RET extracellular domain in patient CC25. tissues by immunofluorescence. Expression of RET pro- As to the technical aspects of NGS, patient CC12 tein and the levels of its phosphorylation at residues showed total mean coverage of 1990, coverage of the Y905 and Y1062 were assessed in colon cancer and the variant of 45 and frequency of the variant 49%, patient corresponding normal mucosa of patients CC12 and CC33 showed total mean coverage of 789, coverage of CC20 using immunofluorescence. the variant of 349 and frequency of the variant 5.9%, pa- Tissues were scored as positive when showing uni- tient CC20 showed total coverage of 728, coverage of form, intense cytoplasmic labelling with marked accen- the variant of 442 and frequency of the variant 20.4% tuation of membranes, and as negative or weakly and patient CC25 showed total mean coverage of 761, positive when labelling was a faint cytoplasmic staining coverage of the variant of 89 and frequency of the with few positive cells interspersed among the majority Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 6 of 12 CC12 Normal CC12 Tumor (C/T), P1047S CC20 Normal CC20 Tumor (G/T), G533C Fig. 1 RET is somatically mutated in colon cancer samples. a. DNA Sanger sequencing of normal mucosa and the corresponding tumor tissue of patient CC12 [exon 12 (C/T)] (b). DNA Sanger sequencing of normal mucosa and the corresponding tumor tissue of patient CC20 [exon 7 (G/T)] of negative cells. We found that normal colon mucosa of residues while the normal mucosa samples were either the two samples analysed were moderately positive for negative or presented weak positivity (see Fig. 3b). These RET expression (see Fig. 3a). results suggest that, at least in the case of some RET Similarly, we found that the corresponding cancerous mutants, the expression of the receptor is not impaired by tissues of patients CC12 and CC20 retained expression selective pressure for loss of RET expression. of RET at levels that were comparable with normal mucosa. Moreover, tumor samples harbouring RET mutations RET activating mutations identified in colon cancer showed positivity for phospho-Y905 and phospho-Y1062 patients increase anchorage-dependent cell proliferation and clonogenic cell survival We investigated whether the RET variants identified in colon cancer patients are able to stimulate anchorage- dependent proliferation in HEK293 cells. This cellular system has already been successfully used to investigate the oncogenic characteristics of multiple oncoproteins [31]. To this aim, the G533C and P1047S RET variants were generated by site-specific mutagenesis of pBABE-puro plas- mid carrying the cDNA encoding human wild type RET (RET-G533C, RET-P1047S and RET-WT, respectively). HEK293 cells were transfected with empty vector (HEK293-C) or transfected with plasmids carrying wild type or mutant RET (HEK293-RET-WT, HEK293-RET- G533C and HEK293-RET-P1047S, respectively). Multiple Fig. 2 Semi-quantitative RT-PCR analysis of RET expression in colon clones of HEK293-C, HEK293-RET-WT, HEK293-RET- cancer samples. Semi-quantitative RT-PCR was performed to investigate RET expression in matched normal (N) and tumor (T) samples in patients G533C and HEK293-RET-P1047S cells from transfection CC12 and CC20. Actin mRNA was used as control for RNA integrity and experiments were expanded and/or frozen for further quantity. TT cells were used as positive control. Sample without input studies. HEK293-RET-C634R cells were used as positive RNA was used as negative control (C) control. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 7 of 12 Fig. 4 Analysis of the biological effects exerted by RET-G533C and RET-P1047S mutants. Plasmids encoding wild type or mutant RET variants (G533C and P1047S) were transfected into HEK293 cells and selected in puromycin. a. MTT assay was performed at different time points (24 h, 48 h, 72 h and 96 h) using empty pBabe plasmid as negative control and plasmid encoding RET C634R mutant as positive control. Values are shown as bar graphs and all results are the average of two independent experiments performed in triplicate, Error bars s.d.; n =6; **p < 0.01; ***p < 0.001 compared with control. b. Colony formation assay was performed with HEK293 cells transfected with plasmids encoding wild type or mutant RET variants (G533C and P1047S). We used empty pBabe plasmid as negative control and plasmid encoding RET C634R mutant as positive control. Similar results were observed in three independent experiments Similar results were obtained with colony assay experi- Fig. 3 Immunofluorescence analysis of RET expression and ments. Also in this type of assay, HEK293-RET-G533C phosphorylation in colon cancer samples. a. Immunofluorescence cells showed greater proliferative potential in clonogenic as- analysis of RET expression in normal and tumor tissue samples says than HEK293-C cells, generating more colonies than from patients CC12 and CC20 shown as merged images of green control cells. Similarly, HEK293-RET-WT and HEK293- (RET staining) and blue (DAPI staining). Magnification 63X. Scale RET-P1047S cells showed an intermediate behaviour bar, 200px. b. Immunofluorescence analysis of RET phosphorylation at Y905 and Y1062 in normal and tumor tissue samples from (Fig. 4b). These results indicate that RET mutant patients CC12 and CC20 shown as merged images of green G533C, but not wild type or P1047S RET mutants (phospho-RET staining) and blue (DAPI staining). Magnification expressed at similar level, is able to stimulate anchorage- 63X. Scale bar 200px dependent proliferation of human epithelial cells in culture. Subsequently, immunoblot analysis was performed to Analysis of cell proliferation by MTT demonstrated check for RET expression and for the activation status of that, similar to HEK293-RET-C634R cells, HEK293- MAPK pathway in transfected HEK293 cells (Fig. 5). Ly- RET-G533C duplicated with an increased rate in mono- sates from transfected HEK293 cells and derivatives were layer compared with HEK293-C (p < 0.01). Conversely, analysed by immunoblot with anti-phosphoY1062 or HEK293-RET-WT and HEK293-RET-P1047S cells showed anti-RET antibody. In agreement with the data from a more limited increase, being intermediate between nega- MTT and colony assay experiments, the G533C RET tive (HEK293-C) and positive controls (HEK293-RET- mutant showed increased RET phosphorylation than C634R), see Fig. 4a. wild type RET or P1047S mutant (Fig. 5a). Subsequently, Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 8 of 12 Fig. 5 Analysis of the biochemical effects exerted by RET-G533C and RET-P1047S mutants. Immunoblot analysis of RET expression and activity in transfected HEK293 cells (a). Lysates from transfected HEK293 cells were analysed by immunoblot with anti-phosphoY1062 or anti-RET antibody. The activation status of the MAPK pathway was assessed with anti phospho-Y202T204 ERK1/2 antibody. Anti-ERK1/2 antibody was used for normalization. b. Immunoblot analysis of RET dimers in transfected HEK293 cells. Samples were prepared under non-reducing (NR) or reducing (R) conditions to detect RET dimer formation and loaded onto SDS-PAGE in non-reducing conditions. Half the total amount of protein lysate was loaded in the case of RET634 we assessed the activation status of critical signalling Cells expressing the RET G533C mutant are sensitive to pathways that function downstream RET [32], the treatment with the RET specific inhibitor vandetanib. MAPK pathway. The activity of the MAPK pathway was Finally, we determined whether the RET inhibitor assessed with anti phospho-Y202/T204 ERK1/2 antibody vandetanib was able to abolish the effects exerted by (Fig. 5a). Anti-total ERK1/2 antibody was used for RET-533C mutant in transfected HEK293 cells. Treat- normalization. We found that the adoptive expression of ment of HEK393 cells expressing RET mutants with RET mutant G533C, at difference with wild type or vandetanib induced a dose-dependent reduction (al- P1047S RET, induced a marked increase in the activation most 50% using 500 nM Vandetanib) in viable cell of MAPK pathway. Finally, since the mutation carried by number of HEK293-RET-G533C (and HEK293-RET- the active mutant involved a cysteine residue in the C634R) cells, demonstrating growth inhibitory activity extracellular domain, we investigated whether this amino in vitro (Fig. 7a). Similarly, subsequent immunoblot acidic change facilitated the formation of RET dimers. analysis demonstrated that vandetanib markedly re- To detect RET dimer formation we separated cell lysates duced phosphorylation of RET Y1062 and of ERK1/2 from transfected HEK293 cells by SDS-PAGE under ei- Y202/T204 in HEK293-RET-G533C and HEK293- ther non-reducing or reducing condition. As shown in RET-C634R cells (Fig. 7b). Fig. 5b, we found that RET mutants G533C was particu- larly efficient in generating RET dimers in transfected Discussion HEK293 cells comparable to the positive control (PC), In this manuscript, we have re-analysed NGS data gener- suggesting that the oncogenic mechanism whereby RET ated on a cohort of Italian patients affected by colon mutant G533C contribute to colon cancer development cancer to identify variants in genes encoding RTKs. The is through an increase in the amount of covalently main findings of this analysis were: i) somatic variants linked dimers induced by the presence of the unpaired were detected in 31 RTK encoding genes, the most fre- cysteine induced by the mutation. quent of which were FLT4, ROS1, EPH7, ERBB2, EGFR, We also found that RET-G533C mutant stimulated RET, FGFR3 and FGFR4, ii) the majority of the observed cell migration in vitro (Fig. 6). We investigated the ef- variants were variants of unknown significance (VUS), fects of the RET-G533C mutant on the migration cap- being only 10% present in the COSMIC database, iii) ability of HEK293 cells by using wound healing assay somatic mutations in the RET proto-oncogene were (Fig. 6a). RET-WT and RET-C634R were used as con- identified in 4 colon cancer patients (10%), iv) only one trols. The results of a triplicate set of independent exper- RET variants were apparently active and able to stimu- iments demonstrated that RET-G533C mutant enhanced late anchorage-dependent proliferation of HEK293 cells, migration of HEK-293 cells of about 70% at 48 h at dif- resembling MEN2A-like mutations, since they efficiently ference with RET-WT, which was much less efficient in generate RET dimers in transfected HEK293 cells. These promoting migration (Fig. 6b). issues will be commented more in detail below. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 9 of 12 Fig. 7 The RET-G533C mutant is inhibited by vandetanib. a. MTT assay was performed in the presence of two different concentrations of vandetanib. HEK293 cells expressing pBabe were used as negative Fig. 6 Effect of RET-G533C on cell migration. a. Wound healing assay control and HEK293 cells expressing RET-C634R mutant were used as of HEK293 cells expressing different RET mutants or empty pBABE positive control. Values are shown as bar graphs and all results are the vector. Cells were scratch-wounded with a micropipette tip (200 μl). average of experiments performed in triplicate. Statistical significance Images of cellular migration were taken at times 0 h, 24 h and 48 h compared with vehicle was evaluated by Two-Way ANOVA (with using the Leica DFC420 C and Leica Application Suite Software. multiple comparison Dunnet’stest) (n =3; **p <0.01; ***p <0.001). Magnification, 10X. b. Wound healing was quantified by the ImageJ b. Immunoblot analysis of RET/MAPK pathway (with anti-phosphoY1062 64 software using the area of the wound of cells expressing empty RET, anti-RET, anti phospho-Y202T204 ERK1/2, anti-ERK1/2 antibodies) of pBabe vector at T0 as reference value. Final results represent mean ± SD HEK293 cells transfected with RET-G533C mutant treated with vehicle or of three independent experiments and are indicated as fold change over vandetanib (500 nM) for 2 h. HEK293 cells transfected with RET-634 the control in terms of reduction of the wound. Statistical significance mutant were used as control was evaluated by Two-Way ANOVA (with multiple comparison Tukey’s test) confronting at each time point the different conditions and indicating the difference over the control (***p <0.001; ****p <0.0001) are VUS. Only few variants could be easily categorized as known pathogenic mutations, such as those present Upon re-analysis of the data reported in Oliveira et al. in the COSMIC database [FLT4 (G1131S, R1041Q, (under review) we have focused on genes encoding RTKs, E1026K), EGFR (R669*, W731*), ERBB2 (S310F), ERBB4 and have observed that RTKs are potentially involved in (R847H), IGF1R (I1151M), EPHB1 (A922T) genes]. This the initiation and/or progression of at least 75% of colon observation raises the issue of the clinical significance of cancer. In fact, 28 patients out of 37 presented variants in the variants identified, and consequently of the use that at least one gene encoding RTKs. Co-occurrence of vari- oncologists should do to patients with colon cancer that ants in different RTK genes was observed in 20 tumors carry VUS in RTK genes. Although 30/101 of the vari- out of 37 (54%), with several samples showing variants in ants identified in this study promoted amino acid multiple RTK genes simultaneously (CC27, 11 genes; changes within the catalytic domains of the correspond- CC34, 13 genes; CC12, 8 genes), which raised the issue of ing RTKs, the ability to provide a clinical interpretation whether these variants were indeed functionally relevant. of VUS represents a daunting task because clinical or An important issues raised by this work is the finding experimental data that aid interpretation are rarely avail- that the vast majority of variants identified in RTK genes able. The only valid alternative for clinicians is to Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 10 of 12 evaluate significance by the use of computational tools Y202/T204 in parallel with a marked reduction of the that predict the biologic effects of novel variants [33]on proliferative effects induced by the RET G533C variant. the basis of available literature. In this manuscript we These results indicate that RET mutant G533C, but have used two different prediction algorithms, SIFT and not wild type or P1047S RET expressed at similar level, Polyphen2, to define a variant as pathogenic. However is able to stimulate proliferation and migration of human the oncogenic role of VUS remains unclear. epithelial cells in culture. In addition, the finding that The choice of RET for subsequent validation of RTK the sample with RET G533C variant presented variants variants was due to the fact that, at difference with what in the downstream effectors like KRAS or MAPK happens in thyroid and lung cancer, the involvement of whereas the samples that carries the RET P270L or the RET proto-oncogene to the development of colon can- RET P1047S variants presented also variants in other cer is still debated. On one hand, it has been proposed genes (KRAS, AKT1 and BRAF, PTEN, AKT1 and that RET is a potential tumor suppressor gene in colo- FGFR3, respectively) indicated that the RET G533C rep- rectal cancer [34]. RET is expressed in normal mucosa, resents a real gain-of-function mutation. and its expression is lost in adenomas and adenocarcin- omas because of an aberrantly methylation of CpG Conclusions islands within its promoter [34]. Notably, RET promoter In conclusion, we have re-analysed NGS data generated methylation is associated with poor prognosis in stage II on 37 colon cancer patients by use of Comprehensive and III colorectal cancer patients [35]. Moreover, al- Cancer Panel, focusing on genes encoding RTKs. Over- though RET mutations (V145G, R360W or G593E) in all, we have observed 101 different potentially damaging colorectal cancer have been identified [36] and are re- variants distributed across 31 RTK genes in 28 patients, ported in the COSMIC database, these mutations appar- suggesting that the remaining tumors represent a subset ently inactivate RET [34]. On the other hand, Le Rolle of colon cancer that may develop through pathways that and co-workers have recently reported on the identifica- are independent of RTK signalling. However, only few tion and characterization of RET fusions that juxtapose variants could be unambiguously categorized as patho- the C-terminal RET kinase domain to the N-terminal do- genic mutations, being the vast majority of RTK variants main of CCD6 or NCOA4 (CCDC6-RET and NCOA4- identified in this study VUS, whose clinical significance RET fusions) in advanced colorectal cancer. Importantly, a remains unclear. This is best exemplified by the validation patient carrying the CCDC6-RET fusion showed a thera- analysis performed on 2 different RET variants. The results peutic response when treated with the RET kinase inhibi- presented here indicate that RET variant G533C is clearly tor regorafenib [37]. oncogenic whereas RET variant P1047S is not. Yet, the In this manuscript, we have investigated whether the identification of gain-of-function RET variants in colon RET variants identified in colon cancer patients are able cancer patients provides novel data for the involvement of to confer pro-tumorigenic properties to epithelial cells. this RTK in the development of colon cancer and identifies We observed that HEK293 cells expressing RET-G533C, RET as a novel putative target for targeted therapy. but not the same cells expressing RET-WT or RET- P1047S, duplicated with an increased rate in monolayer Additional files and in clonogenic assays and promoted migration com- pared with control cells. Accordingly, the active RET Additional file 1: Table S1. Clinical-pathological characteristics of the 37 colon cancer patients included in the study. (XLSX 12 kb) mutant detected in colon cancer patients (G533C) may Additional file 2: Table S2. Summary of the variants identified in RTKs. resemble those described in MEN2A patients, of which (XLSX 16 kb) the C634R is the prototype [38]. In fact, RET mutant Additional file 3: Table S3. List of mutated RTKs in each patient G533C was particularly efficient in generating RET di- included in the study. (XLSX 11 kb) mers in transfected HEK293 cells, suggesting that the Additional file 4: Table S4. Sequencing metrics of the identified RET oncogenic mechanism whereby this RET mutant con- mutations. (XLSX 10 kb) tributes to colon cancer development occurs through an Additional file 5: Table S5. List of mutations identified in CC20 after the next generation sequencing analysis pipeline. (XLSX 17 kb) increase in the amount of covalently linked dimers in- duced by the presence of the unpaired cysteine caused by the mutations. Such increased amount of dimers Abbreviations CCP: Comprehensive Cancer Panel; CRC: Colorectal cancer; HEK293: Human would turn into increased activity of RET receptors (de- embryonic kidney cells 293; HSNCC: Head and neck squamous-cell carcin- tected as increased RET phosphorylation), which in turn oma; IGV: Integrated Genomic Viewer; MAPK,: Mitogen-activated protein would cause activation of the MAPK signalling pathway kinase; NGS: Next Generation Sequencing; PTKs: Protein tyrosine kinases; RTKs: Receptor-type tyrosine kinases; SDS-PAGE: Sodium Dodecyl Sulphate - (detected as increased ERK1/2 phosphorylation). PolyAcrylamide Gel Electrophoresis;; SIFT: Scale-invariant feature transform; It is of note that the RET inhibitor vandetanib abol- UICC: Union for International Cancer Control; VUS: Variants of unknown ished phosphorylation of RET Y1062 and of ERK1/2 significance Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 11 of 12 Acknowledgments transforming sequence in papillary thyroid carcinoma. Genes Chromosomes Not applicable. Cancer. 1994;9:244–50. 9. Soda M, Choi YL, Enomoto M, Takada S, Yamashita Y, Ishikawa S, Fujiwara S, Watanabe H, Kurashina K, Hatanaka H, et al. Identification of the Funding transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature. This work was supported by Centro di Genomica Funzionale e Patologia 2007;448:561–6. Molecolare of University Magna Graecia of Catanzaro and by PON01_02782 10. Rowley JD. Letter: a new consistent chromosomal abnormality in chronic to GV. myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature. 1973;243:290–3. Availability of data and materials 11. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL. Human All data generated or analysed during this study are included either in this breast cancer: correlation of relapse and survival with amplification of the article or in the Additional files. HER-2/neu oncogene. Science. 1987;235:177–82. 12. Herman SE, Gordon AL, Hertlein E, Ramanunni A, Zhang X, Jaglowski S, Authors’ contributions Flynn J, Jones J, Blum KA, Buggy JJ, et al. Bruton tyrosine kinase represents GV, DM and AR designed the research plan; RS and AR performed the a promising therapeutic target for treatment of chronic lymphocytic operations; DMO, KG, RL, FB and CM performed experiments; CM and RL leukemia and is effectively targeted by PCI-32765. Blood. 2011;117:6287–96. contributed to the collection of biological samples; CM, AR and MF 13. Cai H, Babic I, Wei X, Huang J, Witte ON. Invasive prostate carcinoma driven contributed to the collection and analysis of clinical data; CL, performed by c-Src and androgen receptor synergy. Cancer Res. 2011;71:862–72. bioinformatic data analysis; DMO, GV, DM, FC prepared manuscript draft and 14. Memarzadeh S, Xin L, Mulholland DJ, Mansukhani A, Wu H, Teitell MA, Witte participated in manuscript editing and review. All authors read and ON. Enhanced paracrine FGF10 expression promotes formation of multifocal approved the final manuscript prostate adenocarcinoma and an increase in epithelial androgen receptor. Cancer Cell. 2007;12:572–85. Ethics approval and consent to participate 15. Cartwright CA, Kamps MP, Meisler AI, Pipas JM, Eckhart W. pp60c-src All protocols performed in studies involving human participants were activation in human colon carcinoma. J Clin Invest. 1989;83:2025–33. reviewed and approved by the Institutional Review Board of the Mater 16. Grant S. FAM83A and FAM83B: candidate oncogenes and TKI resistance Domini/University Magna Graecia in the meeting of May 21st 2014. mediators. J Clin Invest. 2012;122:3048–51. 17. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer Consent for publication statistics, 2012. CA Cancer J Clin. 2015;65:87–108. Not applicable. 18. Boland PM, Fakih M. The emerging role of neoadjuvant chemotherapy for rectal cancer. J Gastrointest Oncol. 2014;5:362–73. Competing interests 19. Lubner SJ, Loconte NK, Holen KD, Schelman W, Thomas JP, Jumonville A, The authors declare that they have no competing interests. Eickhoff JC, Seo S, Mulkerin DL. A phase II study of oxaliplatin, 5-fluorouracil, leucovorin, and high-dose capecitabine in patients with metastatic colorectal cancer. Clin Colorectal Cancer. 2010;9:157–61. Publisher’sNote 20. Ramakrishnan MS, Eswaraiah A, Crombet T, Piedra P, Saurez G, Iyer H, Arvind Springer Nature remains neutral with regard to jurisdictional claims in AS. Nimotuzumab, a promising therapeutic monoclonal for treatment of published maps and institutional affiliations. tumors of epithelial origin. MAbs. 2009;1:41–8. 21. Siena S, Sartore-Bianchi A, Di Nicolantonio F, Balfour J, Bardelli A. Biomarkers Author details predicting clinical outcome of epidermal growth factor receptor- Department of Experimental and Clinical Medicine, University Magna targeted therapy in metastatic colorectal cancer. J Natl Cancer Inst. Graecia of Catanzaro, Campus Salvatore Venuta -Viale Europa, Catanzaro 2009;101:1308–24. 88100, Italy. Department of Health Sciences, University Magna Graecia of 22. Bardelli A, Siena S. Molecular mechanisms of resistance to cetuximab and Catanzaro, Campus Salvatore Venuta - Viale Europa, Catanzaro 88100, Italy. panitumumab in colorectal cancer. J Clin Oncol. 2010;28:1254–61. Department of Medical and Surgical Sciences, University Magna Graecia of 23. Hubbard J, Grothey A. Antiangiogenesis agents in colorectal cancer. Curr Catanzaro, Campus Salvatore Venuta - Viale Europa, Catanzaro 88100, Italy. Opin Oncol. 2010;22:374–80. Department of Molecular Medicine and Medical Biotechnologies, University 24. Marques I, Araujo A, de Mello RA. Anti-angiogenic therapies for metastatic Federico II, Naples, Italy. UOC Chirurgia Generale, Azienda Ospedaliera dei colorectal cancer: current and future perspectives. World J Gastroenterol. Colli, Naples, Italy. 2013;19:7955–71. 25. Mulder K, Scarfe A, Chua N, Spratlin J. The role of bevacizumab in colorectal Received: 20 December 2017 Accepted: 2 April 2018 cancer: understanding its benefits and limitations. Expert Opin Biol Ther. 2011;11:405–13. 26. UICC. TNM Classification of Malignant Tumours, 7th Edition. 7th edn. References Chichester: Wiley; 2009. 1. Arora A, Scholar EM. Role of tyrosine kinase inhibitors in cancer therapy. J 27. Carlomagno F, Vitagliano D, Guida T, Ciardiello F, Tortora G, Vecchio G, Ryan Pharmacol Exp Ther. 2005;315:971–9. AJ, Fontanini G, Fusco A, Santoro M. ZD6474, an orally available inhibitor of 2. Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature. 2001;411: KDR tyrosine kinase activity, efficiently blocks oncogenic RET kinases. Cancer 355–65. Res. 2002;62:7284–90. 3. Robinson DR, Wu YM, Lin SF. The protein tyrosine kinase family of the 28. Carlomagno F, Salvatore G, Cirafici AM, De Vita G, Melillo RM, de Franciscis human genome. Oncogene. 2000;19:5548–57. V, Billaud M, Fusco A, Santoro M. The different RET-activating capability of 4. Corless CL, Barnett CM, Heinrich MC. Gastrointestinal stromal tumours: mutations of cysteine 620 or cysteine 634 correlates with the multiple origin and molecular oncology. Nat Rev Cancer. 2011;11:865–78. endocrine neoplasia type 2 disease phenotype. Cancer Res. 1997;57:391–5. 5. da Cunha SG, Shepherd FA, Tsao MS. EGFR mutations and lung cancer. 29. Da Silva AM, Maciel RM, Da Silva MR, Toledo SR, De Carvalho MB, Cerutti Annu Rev Pathol. 2011;6:49–69. JM. A novel germ-line point mutation in RET exon 8 (Gly(533)Cys) in a large 6. Yamamoto Y, Kiyoi H, Nakano Y, Suzuki R, Kodera Y, Miyawaki S, Asou N, kindred with familial medullary thyroid carcinoma. J Clin Endocrinol Metab. Kuriyama K, Yagasaki F, Shimazaki C, et al. Activating mutation of D835 2003;88:5438–43. within the activation loop of FLT3 in human hematologic malignancies. Blood. 2001;97:2434–9. 30. Kjaer S, Kurokawa K, Perrinjaquet M, Abrescia C, Ibanez CF. Self-association 7. Hirota S, Isozaki K, Moriyama Y, Hashimoto K, Nishida T, Ishiguro S, Kawano of the transmembrane domain of RET underlies oncogenic activation by K, Hanada M, Kurata A, Takeda M, et al. Gain-of-function mutations of c-kit MEN2A mutations. Oncogene. 2006;25:7086–95. in human gastrointestinal stromal tumors. Science. 1998;279:577–80. 31. Bonfil RD, Reddel RR, Ura H, Reich R, Fridman R, Harris CC, Klein-Szanto JP. 8. Sozzi G, Bongarzone I, Miozzo M, Borrello MG, Blutti MG, Pilotti S, Della Invasive and metastatic potential of a v-ha-ras-transformed human Porta G, Pierotti MA. A t(10;17) translocation creates the RET/PTC2 chimeric bronchial epithelial cell line. J Natl Cancer Inst. 1989;81:587–94. Mendes Oliveira et al. Journal of Experimental & Clinical Cancer Research (2018) 37:84 Page 12 of 12 32. Mulligan LM. RET revisited: expanding the oncogenic portfolio. Nat Rev Cancer. 2014;14:173–86. 33. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, Grody WW, Hegde M, Lyon E, Spector E, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17:405–24. 34. Luo Y, Tsuchiya KD, Il Park D, Fausel R, Kanngurn S, Welcsh P, Dzieciatkowski S, Wang J, Grady WM. RET is a potential tumor suppressor gene in colorectal cancer. Oncogene. 2013;32:2037–47. 35. Draht MX, Smits KM, Tournier B, Jooste V, Chapusot C, Carvalho B, Cleven AH, Derks S, Wouters KA, Belt EJ, et al. Promoter CpG island methylation of RET predicts poor prognosis in stage II colorectal cancer patients. Mol Oncol. 2014;8:679–88. 36. Sjoblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD, Mandelker D, Leary RJ, Ptak J, Silliman N, et al. The consensus coding sequences of human breast and colorectal cancers. Science. 2006;314:268–74. 37. Le Rolle AF, Klempner SJ, Garrett CR, Seery T, Sanford EM, Balasubramanian S, Ross JS, Stephens PJ, Miller VA, Ali SM, Chiu VK. Identification and characterization of RET fusions in advanced colorectal cancer. Oncotarget. 2015;6:28929–37. 38. Zou HJ, Li YS, Shan ZY. Genetic analysis and clinical investigation of a pedigree with multiple endocrine neoplasia type 2A: a case report. Oncol Lett. 2016;12:2657–9.

Journal

Journal of Experimental & Clinical Cancer ResearchSpringer Journals

Published: Apr 17, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off