Abstract
For an identification of novel candidate genes in thyroid tumourigenesis, we have investigated gene copy number changes in a Trk-T1 transgenic mouse model of thyroid neoplasia. For this aim, 30 thyroid tumours from Trk-T1 transgenics were investigated by comparative genomic hybridisation. Recurrent gene copy number alterations were identified and genes located in the altered chromosomal regions were analysed by Gene Ontology term enrichment analysis in order to reveal gene functions potentially associated with thyroid tumourigenesis. In thyroid neoplasms from Trk-T1 mice, a recurrent gain on chromosomal bands 1C4–E2.3 (10.0% of cases), and losses on 3H1–H3 (13.3%), 4D2.3–E2 (43.3%) and 14E4–E5 (6.7%) were identified. The genes Twist2, Ptma, Pde6d, Bmpr1b, Pdlim5, Unc5c, Srm, Trp73, Ythdf2, Taf12 and Slitrk5 are located in these chromosomal bands. Copy number changes of these genes were studied by fluorescence in situ hybridisation on 30 human papillary thyroid carcinoma (PTC) samples and altered gene expression was studied by qRT-PCR analyses in 67 human PTC. Copy number gains were detected in 83% of cases for TWIST2 and in 100% of cases for PTMA and PDE6D. DNA losses of SLITRK1 and SLITRK5 were observed in 21% of cases and of SLITRK6 in 16% of cases. Gene expression was significantly up-regulated for UNC5C and TP73 and significantly down-regulated for SLITRK5 in tumours compared with normal tissue. In conclusion, a global genomic copy number analysis of thyroid tumours from Trk-T1 transgenic mice revealed a number of novel gene alterations in thyroid tumourigenesis that are also prevalent in human PTCs.
Introduction
Papillary thyroid carcinoma (PTC) is the most common malignancy of the thyroid gland (Knauf & Fagin 2009). At the molecular level PTC is characterised by genetic alterations of components of the MAP kinase pathway (Nikiforov 2008). These include structural chromosome rearrangements affecting genes encoding RET (RET/PTC) and NTRK1 (TRK-T) tyrosine kinase receptors that undergo in-frame recombination with various partner genes (Bongarzone et al. 1989, Greco et al. 2010). Depending on the fusion partner, various RET/PTC and TRK-T chimeric oncogenes have been identified.
Specifically, the Trk-T1 oncogene results from a paracentric inversion of chromosome 1q25 that fuses the 5′ end of the translocated promoter region to the 3′ end of NTRK1 genes generating the constitutively active and oncogenic kinase NTRK1 (Greco et al. 2010). Transgenic mice featuring the thyroid-specific expression of Trk-T1 under the transcriptional control of the thyroid-specific bovine thyroglobulin (Tg) promoter were generated previously (Russell et al. 2000). Twenty-three percent of Trk-T1 mice of age ≤7 months and 78% of mice >7 months developed thyroid nodules characterised by malignant features, such as the proliferation of follicular epithelial cells containing scant cytoplasm, mitotic figures and papillae with fibrovascular stalks (Russell et al. 2000, Kim & Zhu 2009). Neoplasms observed in Trk-T1 thyroids were focal, suggesting the occurrence of clonal genetic events cooperating with the initiating Trk-T1 transgene. Such genetic lesions that cooperate with Trk-T1 in this process are still unknown.
Comparative genomic hybridisation (CGH) allows the identification of copy number gains and losses of chromosomal regions in tumours and subsequent discovery of genes that may play a role in the tumourigenic process. CGH can be applied at different resolutions depending on the targets used for the hybridisation such as metaphase chromosomes (metaphase CGH; resolution 5–10 Mb) or DNA microarrays (array CGH; 1 Mb to kb resolution, Chung et al. 2004).
In this study, we applied metaphase CGH and array CGH to study gene copy number alterations (CNA) in thyroid tumours developed in Trk-T1 transgenic mice. We identified four recurrent CNA. For these CNA, we used the synteny approach on www.ensemble.org to define the syntenic human regions. These human regions were reconciled with array CGH data generated in our lab from human PTCs (Unger et al. 2008, Hess et al. 2011). We further determined candidate genes located in these altered regions from the published literature.
In human PTC samples copy number gains of the genes TWIST2, PTMA and PDE6D were identified by fluorescence in situ hybridisation (FISH) in 10/12 samples for TWIST2 and in 12/12 samples for PTMA and PDE6D. Moreover, human thyroid tumours consistently featured loss of the genes SLITRK1, SLITRK5 and SLITRK6 (4/19, 3/19 and 4/19 samples).
Materials and methods
Transgenic mice
Mice were maintained under specific pathogen-free conditions in the Laboratory Animal Facility of the Università ‘Federico II’, Naples, Italy and all studies were conducted in accordance with Italian regulations for experiments on animals. Tg-Trk-T1 mice were euthanised and thyroids were immediately fixed in buffered formalin for 4 h at room temperature. Tissues underwent automated processing and paraffin embedding; 5 μm sections were cut and haematoxylin and eosin stained (H&E) for microscopic analysis of the tissue morphology. Thirty Tg-Trk-T1 thyroid samples were studied and classified according to the WHO criteria for the evaluation of mouse thyroid tumours (Jokinen & Botts 1994). The thyroid tissues were considered normal when composed by variable-sized follicles covered by a monolayer of flattened epithelial cells or as hyperplastic when showing secondary occurrence of small follicles with scant colloid and tall epithelial cells that were merged with normal areas. Finally, neoplastic thyroid lesions were defined as well-demarcated nodules with a distinct papillary and/or follicular architecture; in a few cases, invasion of the surrounding glandular parenchyma or stroma was detected, thereby allowing the classification of carcinomas of the papillary subtype.
Human tissues
For validation experiments human samples from already existing post-Chernobyl PTC cohorts were available at Helmholtz Zentrum München and Imperial College London. Thyroid tumour samples (FFPE tissue sections) from 30 patients presenting with PTC developed after the Chernobyl reactor accident were obtained from the Chernobyl Tissue Bank (CTB; 16 cases), from the Centre of Thyroid Tumours in Minsk, Belarus (seven cases) and from the Institute of Endocrinology and Metabolism, Kiev, Belarus (seven cases). In addition, RNA from frozen tissues of 67 PTC cases was obtained from the CTB for mRNA gene expression analyses. All PTC samples were characterised for the presence of RET/PTC rearrangements by RT-PCR and for the V600E BRAF mutation, the most frequent BRAF mutation in PTC, by direct DNA sequencing (Powell et al. 2005).
All tumours were diagnosed as PTC according to the TNM Classification of Malignant tumours and further classified according to their dominant histological architecture (papillary, follicular and solid) (Hedinger et al. 1989). Appropriate informed consent was obtained from the patients or their guardians and the study was approved by the ethics committee in accordance with the declaration of Helsinki.
CGH of mouse tissues
CGH was applied to detect genomic copy number changes in murine thyroid nodules. Nodule (thyroid) and reference (liver) DNA were isolated using the QIAmp Mini Kit (Qiagen). DNA concentration and purity were determined photospectrometrically (NanoDrop, Thermo Scientific, Wilmington, DE, USA). Indirect labelling of the reference DNA with digoxigenin and of the tumour DNA with biotin was performed by nick translation using a standard protocol. Hybridisation of the labelled DNA on metaphase spreads with a normal karyotype from the 102/ELxC3H/EL mouse strain was performed according to Richter et al. (2003). The chromosomes were counterstained with DAPI and actinomycin D prior to image acquisition and analysis as described by Richter et al. (2003).
Array-based CGH
Genomic DNA was isolated from fresh frozen tissues or from cultured tumour cells and fluorescence labelled with Cy3 (Invitrogen). Genomic reference DNA was labelled with Cy5 (Invitrogen) and co-hybridised to 1 Mb BAC mouse arrays (Chung et al. 2004). DNA labelling, hybridisation and data processing were performed as described by Unger et al. (2008). Further data analysis was performed using the web-based array CGH evaluation platform CAPweb (Liva et al. 2006, http://bioinfo.curie.fr/CAPweb). After import of raw data into CAPweb, the normalisation procedure using the R-package MANOR (Neuvial et al. 2006) and segmentation of the data set using the R-package GLAD (Hupe et al. 2004) were performed using default parameters. The GLAD algorithm considers probes called as gain and a log2 ratio greater than one as high level amplifications. Data were exported to the array data visualisation tool VAMP (La Rosa et al. 2006). Common regions of alterations of the analysed cases were determined using the minimal-alteration algorithm (Rouveirol et al. 2006), which is implemented in the VAMP software (http://bioinfo-out.curie.fr/projects/vamp/).
Identification of candidate genes
Based on a synteny approach, the murine chromosome alterations were translated into the human syntenic chromosome bands using the (V58) Ensembl database (www.ensembl.org). A comparison of the human chromosome bands with CNA from PTC samples published by Unger et al. (2008) and Hess et al. (2011) revealed CNA derived from the mouse model that are prevalent either in PTC from young patients (Unger et al. 2008, Hess et al. 2011) or in PTC from adults (Unger et al. 2008).
Further, candidate genes located in these syntenic altered chromosomal regions were analysed by Gene Ontology (GO) term enrichment analysis (Zheng & Wang 2008) in order to reveal gene functions potentially associated with thyroid tumourigenesis.
Fluorescence in situ hybridisation
BAC clones were selected from the RPCI23 and -24 mouse libraries (Osoegawa et al. 2000) and from the human 32k re-array library (Osoegawa et al. 2001). After growth for 15 h at 37 °C plasmid DNA was isolated from the bacterial cultures according to standard protocols (Birnboim & Doly 1979). The following gene spanning (digoxigenin) and reference (biotin) BAC clones were selected and labelled (as described above): mouse: RP23-396K20 (Ptma, Pde6d, reference clone: RP23-34E15), RP23-441M15 (Twist2, reference clone: RP23-34E15), RP23-90A10 (Slitrk5, reference clone: RP23-34E15), RP11-127O5 (Taf12, Ythdf2, reference clone: RP23-59B17), RP23-331P21 (Srm, reference clone: RP23-59B17), RP23-254N4 (Trp73, reference clone: RP23-59B17); and human: RP11-690I21 (PTMA, PDE6D, reference clone: RP11-912D12), pooled RP23-717F15 and RP23-275G07 (TWIST2, reference clone: RP11-912D12), RP11-613F15 (SLITRK1, reference clone: RP11-139J24), RP11-243B06 (SLITRK5, reference clone: RP11-139J24), RP11-398A22 (SLITRK6, reference clone: RP11-139J24). Murine and human BACs were checked for binding specificity by hybridisation to normal metaphase spreads. Hybridisation of murine BAC clones to interphase cells was performed as described above. The human BAC clones were hybridised onto 8 μm paraffin-embedded tissue sections according to Unger et al. (2004). Images from the mouse interphase preparations were taken using the Metafer 4 System (MF-054) (V3.6.7; Metasystems, Altlussheim, Germany), while the human FFPE tissues were analysed by taking 3D images with the AxioImager system equipped with ApoTome (Zeiss, Oberkochen, Germany). The FISH signals of at least 100 cells were analysed and classified as positive for a deletion when cell nuclei showed two green and less then two red signals or as positive for an amplification when cell nuclei harboured more than two red signals in addition to two green signals.
qRT-PCR
Total RNA was isolated from fresh frozen tissue using the RNeasy kit (Qiagen) including on-column DNA digestion. RNA concentration and purity were measured using a spectrophotometer (NanoDrop) and integrity of RNA was determined on a microfluidic chip using an Agilent Bioanalyser. mRNA was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA, USA). Quantitative RT-PCR was carried out in an Applied Biosystems 7500 Fast System machine using exon-spanning gene-specific TaqMan assays with 40 cycles according to the manufacturer's protocols (Applied Biosystems). A TaqMan assay for Actin B was used as endogenous control for each measurement. Relative expression levels were calculated using the ΔΔCt method (Livak & Schmittgen 2001).
Results
Recurrent CNA in Trk-T1-transgenic mice
We investigated 30 thyroid lesions from Trk-T1 transgenics including four hyperplasias, three unspecified neoplasias and 23 thyroid carcinomas for genomic CNA by array CGH. We detected a recurrent gain of chromosome bands 1C4–E2.3 (10.0% of cases) and recurrent losses of chromosome bands 3H1–H3 (13.3%), 4D2.3–E2 (43.3%) and 14E4–E5 (6.7%) (Table 1).
Copy number changes in 30 thyroid tumours from Trk-T1 transgenic mice
| Copy number changes (CGH/aCGHa) | ||||
|---|---|---|---|---|
| Case | Histology | Gain | Loss | FISH validationb (frequency of cells with aberrant FISH signals in tumour tissues) |
| TRK-T1-1 | PTC | 2(H4), 7(E2, F3→ter), 8(D1→ter), 14(E5) | 12(A1.3→A3, C1), 14(A1→A3), X(A2) | ND |
| TRK-T1-2 | PTC | 15(E3→F1), 17(E5) | 5(A3), 7(A2→B3), 12(A1.3→A3), 14(A1→A3), 16(C1.3→C2), X(A2→A4) | ND |
| TRK-T1-170 | PTC | 1(A1→A3), 6(B3), 10(D3), 16(C3.1→C3.2, C4) | 2(H1→H3), 4(D1→E2), 7(F2→F3), 15(E1→E2) | ND |
| TRK-T1-236 | Hyperplasia | 15(E3→F2) | 3(G2→G3, H3), 4(C6→D, E2), 19(D3) | ND |
| TRK-T1-238 | Hyperplasia | 11(E1→E2) | 8(A2→A3), 14(A1→B) | ND |
| TRK-T1-239 | PTC | 11(B1.3→B5), 15(E2→F1) | 4(E→ter), 13(D2.2), 14(C3, E4→E5), X(F4→F5) | del 4 (Taf12, Ythdf2): 24.0% of cells |
| 10(A1→D3)a, 15(A1→D1)a | 3(A1→H4)a, 4(A1→E2)a | del 4 (Trp73): 24.1% of cells | ||
| del 14 (Slitrk5): 27.3% of cells | ||||
| TRK-T1-250 | PTC | 15(A1→D1)a | 3(H1→H3), 4(C1→C2), 14(A2) | del 4 (Taf12, Ythdf2): 11.3% of cells |
| 2(A3→B, H1→H4)a, 4(C7→E2)a, 5(B2, F, G1→G3)a, 8(A1.2→E2)a, 10(B5.3→C1), 11(B1.2 →E2)a | del 4 (Srm): 25.5% of cells | |||
| del 4 (Trp73): 22.8% of cells | ||||
| TRK-T1-252 | PTC | 4(C7→E2)a, 5(A1→C3.2)a, 12(B3→C1)a | del 4 (Taf12, Ythdf2): 25.4% of cells | |
| TRK-T1-254 | PTC | 8(D1→E2), 16(C4) | 12(A2→B1), 13(B3), 14(A1→A3) | ND |
| TRK-T1-255 | Hyperplasia | 14(A1→A3, E2.2→ter) | ND | |
| 2(C1.2→D)a, 7(A1→A2, A3→C)a, 8(A4→B1.2)a | ||||
| TRK-T1-256 | PTC | 3(C→E2), 11(A3.3, B1.3, C→D) | 4(E1→E2)a | ND |
| 7(A1→F5)a | ||||
| TRK-T1-261 | PTC | 1(C5→E2), 10(B2, C1→D1) | 3(A1→A3, H3→ter), 4(D1→ter) | amp 1 (Ptma, Pde6d): 38.0% of cells |
| 1(A1→D, E2→F, G2→H6)a, | 3(A1→H4)a, 4(A1→E2)a, 5(A3→C2)a, | amp 1 (Twist2): 43.5% of cells | ||
| 10(A1→D3)a, 14(E4→E5)a, | 9(A1→C)a | del 4 (Taf12, Ythdf2): 16.5% of cells | ||
| 17(A3.3→B3, C)a, 19(A)a/c | del 4 (Srm): 23.3% of cells | |||
| TRK-T1-262 | PTC | 1(C3→E2), 9(B→D), 10(C1), | 3(cen→A3, F2→ter), 4(C5→ter), | amp 1 (Ptma, Pde6d): 56.4% of cells |
| 13(A5→C1) | 6(A1→A2) | amp 1 (Twist2): 59.7% of cells | ||
| 10(A1→D3)a, 13(A1→D2.3)a | 3(A1→H4)a, 4(A1→C1, C3→E2)a, 5(A3→E1)a, 7(A1→F5)a, 11(B1.3→B4, C→E2)a | del 4 (Taf12, Ythdf2): 10.5% of cells | ||
| TRK-T1-263 | PTC | 1(H3→ter), 4(C5, C7→ter), 11(E2→ter) | ND | |
| TRK-T1-265 | PTC | 1(C5→E1) | amp 1 (Ptma, Pde6d): 26.2% of cells | |
| 7(A1→F5)a | 4(A1→E2)a | amp 1 (Twist2): 19.2% of cells | ||
| TRK-T1-268 | PTC | 4(D3→ter) | del 4 (Taf12, Ythdf2): 76.3% of cells | |
| 15(A1→D1)a, 19(A→tD3)a | 2(H1→H4)a, 4(D2.3→E2)a, 6(A1→G3)a, | del 4 (Srm): 86.6% of cells | ||
| 7(A1→F5)a, 12(A1.1→F2)a | del 4 (Trp73): 76.3% of cells | |||
| TRK-T1-270 | PTC | 3(F1→H2)a | ND | |
| TRK-T1-024 | Neoplasia | 19(A→ter)a | ND | |
| TRK-T1-028 | Neoplasia | 19(A→ter)a | ND | |
| TRK-T1-033 | Neoplasia | 7(A1→F5)a | 5(F→G1.1)a, 19(A→ter)a | ND |
| TRK-T1-040 | Hyperplasia | 11(B5→C, D→E1)a | ND | |
| TRK-T1-054 | PTC | 7(C, D1→ter)a | ND | |
| TRK-T1-057 | PTC | 6(A1→G3)a, 7(A1→F5)a | 1(H5→H6)a, 4(A1→E2)a, 5(A1→G3)a, 9(A1→F4)a, 11(A1→ E2)a | ND |
| TRK-T1-061 | PTC | ND | ||
| TRK-T1-062 | PTC | 1(E4→F)a, 5(F, G1→G3), 7(A2→B1)a, 17(A3.3→B3)a | ND | |
| TRK-T1-063 | PTC | 1(E4→F)a | ND | |
| TRK-T1-066 | PTC | 16(A2→C4)a | ND | |
| TRK-T1-068 | PTC | ND | ||
| TRK-T1-070 | PTC | 5(D→G3)a/c | 4(D2.3→E2)a, 9(A1→F4)a | ND |
| TRK-T1-072 | PTC | 7(A1→F5)a, 8(A1.2→B1.2)a | 1(A1→H6)a, 4(A1→E2)a, 5(A1→F)a, 11(A1→ E2)a, 12(A1.1→F2)a | ND |
aCGH, array CGH. Bold: recurrent CNA revealing consensus regions 1C4–E2.3 (10.0% of cases), 3H1–H3 (13.3%), 4D2.3–E2 (43.3%) and 14E4–E5 (6.7%). ND, not determined.
Data from array CGH experiments.
Comparison of frequencies in normal and tumour tissues, FDR-adjusted Fisher's exact test, FDR<0.1.
High level amplification: the GLAD algorithm (Hupe et al. 2004) considers probes called as gain and a log2-ratio greater than one as high-level amplifications.
Genes located in the altered chromosome regions were identified using the Ensembl database (www.ensembl.org) and genes that were published in association with cancer were selected as candidate genes (Table 2 and Fig. 1). The list includes the genes Twist2, Ptma, Pde6d (chromosome 1), Slitrk5 (chromosome 14), Bmpr1b, Pdlim5, Unc5c (chromosome 3), and Srm, Trp73, Ythdf2, Taf12 (chromosome 4). Analysis of the GO terms of these genes showed gene functions linked to cell differentiation and development, apoptosis and intracellular signalling. Copy number changes for Twist2, Ptma, Pde6d, Slitrk5, Srm, Trp73, Ythdf2 and Taf12 were validated in FISH experiments (Fig. 2A and B and Table 1).
Comparison of altered regions in PTC from transgenic mice and humans
| Murine chromosomal band | Human chromosomal band | Frequency of aberration (%) in PTC from young patientsa | Frequency of aberration (%) in PTC from adultsb | Candidate genes |
|---|---|---|---|---|
| 1C4→E2.3 | 2q14.1–14.3 | Gain (0) | Gain (0) | |
| 2q37.1–37.3 | Gain (61.3) | Gain (0)c | TWIST2, PTMA, PDE6D | |
| 5q21.1 | Gain (7.5) | Gain (40.0)c | ||
| 18q21.33–22.1 | Gain (0) | Gain (0) | ||
| 3H1→H3 | 1p31.1–22.2 | Loss (68.8) | Loss (50.0) | |
| 4q22.3 | Loss (55.9) | Loss (0)c | UNC5C, BMPR1B, PDLIM5 | |
| 4D2.3→E2d | 1p36.32–35.3 | Gain (66.7) | Gain (0)c | YTHDF2, SRM, TAF12, TP73 |
| 14E4→E5 | 13q31.2–33.1 | Loss (67.7) | Loss (45.0) | SLITRK5 |
Between males and females and clinico-pathological parameters (lymph node metastasis (N), tumour size (T), histological variant) no differences were found for the human chromosomal bands.
CNA in 93 PTC from young patients (data from Hess et al. (2011) and Unger et al. (2008)).
CNA in 20 PTC from adults (data from Unger et al. (2008)).
Significantly different from childhood cases with P<0.001.
Consensus region derived from CNA of 30 murine thyroid tumours.
Overview of syntenic chromosomal regions in mouse and human. Chromosomal bands and genes representing gains are shown in green, bands and genes showing a loss are red. (A) Amplified syntenic chromosomal bands on mouse chromosome 1 and human chromosome 2 with candidate genes revealed by database-driven analysis. (B, C and D) Deleted syntenic chromosomal bands and genes that were revealed by database-driven analysis.
Citation: Endocrine-Related Cancer 19, 3; 10.1530/ERC-11-0387
FISH analysis of CNA. (A and B) shows FISH analysis of murine cases (A) TRK-T1-261 and (B) TRK-T1-239. Reference FISH signals are labelled in green while gene-specific FISH signals are labelled in red. (A) A gain of Ptma and Pde6d and (B) a deletion of Slitrk5 are shown (upper row of cells) in comparison to cells with unchanged FISH signals (lower row of cells). (C, D and E) shows FISH analysis on human PTC (C) for TWIST2 showing gained red FISH signals in case 16, (D) for PTMA and PDE6D showing amplified FISH signals in case 20 and (E) for SLITRK5 showing a deleted red FISH signal in case 9. Examples of aberrant cells are marked by arrows.
Citation: Endocrine-Related Cancer 19, 3; 10.1530/ERC-11-0387
Gene CNA in human post-Chernobyl PTC
Recently, gene CNA were identified in human PTC (Unger et al. 2008, Hess et al. 2011). While Hess et al. (2011) reported global genomic copy number changes of PTCs from young patients from Ukraine, the publication of Unger et al. (2008) also included data from adult cases and, therefore, conclusions on age-related changes can be drawn. In order to check whether genomic copy number changes that were found in the mouse model are also prevalent in human PTCs, the appropriate mouse regions were translated into homologous human regions using the Ensembl synteny database. Table 2 shows a comparison of syntenic aberrant regions in murine and human tumours. Noteworthy, most of the gene copy aberrations identified in the mouse model were confirmed in human cancer samples. Gains were reported on 2q37.1–37.3 (syntenic to gain 1C4–E2.3) in 61.3% of human PTC from young patients, on 5q21.1 (syntenic to gain 1C4–E2.3) in 7.5% of PTC from young patients and in 40.0% of PTC from adults and on 1p36.32–35.3 (syntenic to 4D2.3–E2) in 66.7% of PTC from young patients. On the other hand, losses on 1p31.1–22.2 (syntenic to loss 3H1–H3) were reported in 68.8% of PTC from young patients and in 50.0% of PTC from adults, on 4q22.3 (syntenic to 3H1–H3) in 55.9% of PTC from young patients and on 13q31.2–33.1 (syntenic to loss 14E4–E5) in 67.7% of PTC from young patients and in 45.0% of PTC from adults.
We sought to confirm the most frequent gain on chromosome 1 and one of the most frequent losses on chromosome 14 identified in the mouse model by FISH analysis of the corresponding syntenic regions on paraffin-embedded sections from human PTC samples. Therefore, we selected cases with DNA copy number gains on 2q37.1–37.3 (investigation of candidate genes TWIST2, PTMA, PDE6D from syntenic murine region on 1C4–E2.3) and cases with DNA copy number losses on 13q31.2–33.1 (investigation of candidate gene SLITRK5 from syntenic murine band 14E4–E5) (Table 2). In the human genome SLITRK5 is part of a gene cluster comprising three homologs which are likely of being functionally similar (Aruga & Mikoshiba 2003). Hence, we also investigated the copy number status of the two genes SLITRK1 and SLITRK6 which are the other members of the SLITRK gene cluster on chromosome 13q31.1–31.2 (Aruga et al. 2003). In order to define the copy number status (gain, normal or loss) of a particular case, frequencies of cells with aberrant FISH signals were compared between normal thyroid tissue (controls) and tumour tissue. The differences of frequencies were tested using Fisher's exact test. A P value <0.1 was accepted to reject the null hypothesis (same frequencies of aberrant cells in normal and tumour tissues) and to consider a case as significantly altered. As shown in Table 3 all investigated human PTCs (n=12) showed a gain of PTMA and PDE6D and the majority of cases also showed a gain of TWIST2 (ten out of 12 cases; 83%). Moreover, in four out of 19 cases (21%) a deletion of SLITRK5 was observed (Fig. 2C, D and E). Finally, four out of 19 cases (21%) showed a deletion of SLITRK1 and three out of 19 cases (16%) a deletion of SLITRK6. Two cases (Table 3) showed a deletion of all SLITRK genes (case 9) or of two SLITRK genes (case 18) simultaneously.
FISH on PTC with CNA on human chromosome 2q37.1–37.3 and 13q31.1–31.2
| Case number | Sex | Age at surgery (years) | Histological variant | TNM | RET/PTC rearrangement/BRAF mutation (V600E) | Copy number alterationsa | FISH aberrant cells (%) | ||
|---|---|---|---|---|---|---|---|---|---|
| 1 | Female | 51 | Mixed (FP) | T1N0Mx | Negative/negative | Gain 2q37.1–37.3 | TWIST2 (19.8)b | PTMA/PDE6D (63.0)b | |
| 2 | Female | 34 | Classic | T3N1aMx | NA/positive | Gain 2q37.1–37.3 | TWIST2 (6.9)b | PTMA/PDE6D (59.4)b | |
| 3 | Female | 30 | Classic | T1N1aMx | Negative/negative | Gain 2q37.1–37.3 | TWIST2 (20.0)b | PTMA/PDE6D (54.9)b | |
| 4 | Male | 11 | Follicular | T4N1M0 | NA/negative | Gain 2q37.1–37.3 | TWIST2 (5.7) | PTMA/PDE6D (73.3)b | |
| 5 | Male | 61 | Mixed (FS) | T3N0Mx | Negative/positive | Gain 2q37.1–37.3 | TWIST2 (20.6)b | PTMA/PDE6D (88.1)b | |
| 6 | Female | 32 | Classic | T2N1Mx | Negative/positive | Gain 2q37.1–37.3 | TWIST2 (11.9)b | PTMA/PDE6D (69.3)b | |
| 7 | Male | 41 | Mixed (PS) | T3N1Mx | Negative/positive | Gain 2q37.1–37.3 | TWIST2 (15.8)b | PTMA/PDE6D (79.4)b | |
| 8 | Female | 14 | Mixed (FS) | T3N1abM1 | Positivec/negative | Loss 13q31.1–31.2 | SLITRK1 (17.8) | SLITRK5 (7.9) | SLITRK6 (19.5) |
| 9 | Female | 14 | Solid | T3N1abM0 | Positived/c/negative | Loss 13q31.1–31.2 | SLITRK1 (26.7)b | SLITRK5 (27.0)b | SLITRK6 (28.7)b |
| 10 | Female | 14 | Follicular | T3N1aM0 | Positived/negative | Loss 13q31.1–31.2 | SLITRK1 (17.5) | SLITRK5 (17.8) | SLITRK6 (19.8) |
| 11 | Male | 15 | Mixed (SF) | T1N1aM0 | Negative/negative | Loss 13q31.1–31.2 | SLITRK1 (14.9) | SLITRK5 (13.9) | SLITRK6 (15.4) |
| 12 | Female | 16 | Classic | T2N0M0 | Negative/positive | Loss 13q31.1–31.2 | SLITRK1 (15.7) | SLITRK5 (25.0)b | SLITRK6 (14.0) |
| 13 | Female | 18 | Follicular | T3N1aM0 | Negative/negative | Loss 13q31.1–31.2 | SLITRK1 (21.0) | SLITRK5 (13.5) | SLITRK6 (17.5) |
| 14 | Female | 17 | Classic | T3N1aM0 | Negative/negative | Loss 13q31.1–31.2 | SLITRK1 (5.0) | SLITRK5 (19.0) | SLITRK6 (17.0) |
| 15 | Male | 16 | Follicular | T3N1aM0 | Negative/negative | Loss 13q31.1–31.2 | SLITRK1 (6.5) | SLITRK5 (20.4) | SLITRK6 (17.2) |
| 16 | Female | 13 | Follicular | T1N0M0 | Positived/negative | Gain 2q37.1–37.3 | TWIST2 (17.6)b | PTMA/PDE6D (93.1)b | |
| 17 | Female | 18 | Solid | T2N1aM0 | Positive/negative | Loss 13q31.1–31.2 | SLITRK1 (9.9) | SLITRK5 (16.0) | SLITRK6 (12.7) |
| 18 | Male | 15 | Classic | T1N0M0 | Negative/NA | Loss 13q31.1–31.2 | SLITRK1 (20.6)b | SLITRK5 (12.0) | SLITRK6 (28.4)b |
| 19 | Male | 15 | Mixed (FS) | T3N1abM0 | Negative/NA | Loss 13q31.1–31.2 | SLITRK1 (25.5)b | SLITRK5 (20.7) | SLITRK6 (15.1) |
| 20 | Male | 17 | Classic | T1N0M0 | Negative/negative | Loss 13q31.1–31.2 | SLITRK1 (20.8) | SLITRK5 (9.3) | SLITRK6 (15.1) |
| Gain 2q37.1–37.3 | TWIST2 (5.8) | PTMA/PDE6D (94.0)b | |||||||
| 21 | Female | 17 | Mixed (PS) | T1N0M0 | Negative/positive | Loss 13q31.1–31.2 | SLITRK1 (16.8) | SLITRK5 (13.0)b | SLITRK6 (20.7) |
| 22 | Male | 13 | Solid | T3N0M0 | Positivec/negative | Loss 13q31.1–31.2 | SLITRK1 (16.8) | SLITRK5 (21.7) | SLITRK6 (17.8) |
| 23 | Female | 12 | Solid | T2N0M0 | NA/negative | Gain 2q37.1–37.3 | TWIST2 (7.7)b | PTMA/PDE6D (93.1)b | |
| 24 | Female | 35 | Classic | T4bN0M0 | Negative/negative | Loss 13q31.1–31.2 | SLITRK1 (15.9) | SLITRK5 (9.9) | SLITRK6 (25.7)b |
| 25 | Female | 35 | Classic | T4aN1aM0 | Positived/positive | Loss 13q31.1–31.2 | SLITRK1 (22.5)b | SLITRK5 (19.0) | SLITRK6 (25.0) |
| 26 | Female | 32 | Follicular | T4bN1aM0 | Positived/negative | Loss 13q31.1–31.2 | SLITRK1 (12.6) | SLITRK5 (23.0)b | SLITRK6 (14.5) |
| 27 | Male | 32 | Classic | T4aN1aM0 | Positived/negative | Loss 13q31.1–31.2 | SLITRK1 (18.1) | SLITRK5 (6.8) | SLITRK6 (6.7) |
| 28 | Female | 36 | Solid | T4N1bMx | Positived/negative | Loss 13q31.1–31.2 | SLITRK1 (19.4) | SLITRK5 (13.9) | SLITRK6 (8.0) |
| 29 | Female | 32 | Follicular | T2aN0Mx | Positive/negative | Gain 2q37.1–37.3 | TWIST2 (9.2)b | PTMA/PDE6D (10.0)b | |
| 30 | Female | 31 | Solid | T4aN1bMx | Negative/positive | Gain 2q37.1–37.3 | TWIST2 (44.1)b | PTMA/PDE6D (28.6)b |
FP, follicular-papillary; FS, follicular–solid; PS, papillary–solid; SF, solid–follicular variant; NA, not available. Sex, age median, histological variants, TNM status are not significantly different for the two CNA gain 2q37.1–37.3 and loss 13q31.1–31.2 (Wilcoxon rank test for age, χ2 test for the other parameters, P>0.05).
Copy number data from Unger et al. (2008) and Hess et al. (2011).
Comparison of frequencies in normal and tumour tissues, FDR-adjusted Fisher's exact test, FDR<0.1.
RET/PTC3 rearrangement.
RET/PTC1 rearrangement.
Altered expression in human PTC samples
We characterised the mRNA expression of ten selected cancer-related genes. The genes PDE6D, PTMA, BMPR1B, PDLIM5, UNC5C, SRM, TAF12, TP73, YTHDF2 and SLITRK5 were selected from syntenic frequently altered chromosomal regions on 1p, 2q, 4q and 13q (Table 2) in human PTC. Due to the limited amount of RNA available a maximum of ten genes could be investigated by qRT-PCR in 67 human PTC cases and five normal thyroid tissues (Hess et al. 2011). We have selected these ten genes because they were equally distributed on the above-mentioned altered four chromosomal bands. Although TWIST2 appeared to be an interesting candidate gene as well, the above-mentioned material limitations and the fact that it was very closely located to PTMA and PDE6D within the same chromosomal band led to the exclusion of TWIST2 from the gene expression analyses. The gene expression data are together with the available copy number data (55/67 cases; Hess et al. 2011) summarised in Supplementary Table 1 and Figure 1, see section on supplementary data given at the end of this article. As demonstrated in Fig. 3, the genes UNC5C and TP73 showed a significantly elevated expression in tumours compared to normal tissues, and SLITRK5 a significantly reduced gene expression in tumours compared to normal tissues (P<0.05; Mann–Whitney U test, Fig. 3). Thus, for SLITRK5 and TP73 a robust alteration in gene expression was in accordance with the genomic CNA.
Expression levels of candidate genes in normal thyroid tissue compared to PTC. The box plot for SLITRK5 shows a significant down-regulation of this gene in tumour tissues (P<0.05), while a significant over-expression of the genes TP73 and UNC5C is visible in tumour tissues.
Citation: Endocrine-Related Cancer 19, 3; 10.1530/ERC-11-0387
Discussion
Rearrangements of the tyrosine kinase-encoding proto-oncogenes NTRK1 (TRK-T) and RET (RET/PTC) have been found in human PTC (Bongarzone et al. 1989, Nikiforov 2008, Knauf & Fagin 2009, Greco et al. 2010). Transgenic mouse strains proved that these oncogenes are able to initiate thyroid tumourigenesis (Jhiang et al. 1996, Santoro et al. 1996, Powell et al. 1998, Russell et al. 2000). However, genetic lesions that cooperate with these frequent gene alterations of the MAP kinase pathway in thyroid carcinogenesis have not been identified so far. To tackle this issue, we have selected the Trk-T1 mouse model in our study in order to investigate clonal gene changes that may occur somatically to cooperate with the Trk-T1 transgene. Thus, initially we have investigated CNA in 30 thyroid tumours from Trk-T1 transgenic mice and found recurrent genomic CNA on chromosomes 1C4–E2.3 (10%), 3H1–H3 (13.3%), 4D2.3–E2 (43.3%) and 14E4–E5 (6.7%). A validation of potentially affected candidate genes located in these chromosomal regions was exemplarily performed for Ptma, Pde6d, Twist2 on murine chromosome 1, Taf12, Ythdf2, Srm, Tp73 on murine chromosome 4 and Slitrk genes on murine chromosome 14 by FISH.
We further asked whether gene copy alterations that have been detected in thyroid tumours from Trk-T1 transgenic mice were also involved in human thyroid carcinomas. Therefore, gene copy number and mRNA gene expression of these candidate genes were investigated in human PTC samples. From our data, we found that alterations of TWIST2, PTMA, PDE6D (murine chromosome 1D, human chromosome 2q37), UNC5C (murine chromosome 3H1, human chromosome 4q22.3), TP73 (murine chromosome 4D2.3, human chromosome 1p36.32) and SLITRK5 (murine chromosome 14E4, human chromosome 13q31.2) were located within the most frequently altered chromosomal regions in human thyroid carcinomas (murine CNA on 1C4–E2.3, 3H1–H3, 4D2.3–E2, 14E4–E5 corresponding to human CNA on 2q37.1–37.3, 4q22.3, 1p36.32–35.3 and 13q31.2–33.1 respectively) (Table 2). It has been shown in a recent publication on a homogeneous, age-matched tumour set of exposed and non-exposed PTC (Hess et al. 2011) that these syntenic CNA are not associated with exposure to ionising radiation. A de-regulated mRNA expression could be confirmed for the candidate genes UNC5C, TP73 and SLITRK5 in 67 PTC compared with normal thyroid tissues, thus, suggesting an impact of the genes in thyroid tumourigenesis in particular for TP73 and SLITRK5 for which the de-regulated gene expression (up-regulation for TP73; down-regulation for SLITRK5) was in accordance with the genomic CNA (gain of TP73; loss of SLITRK5). For a deeper mechanistic insight we have compared genomic copy number data and relative gene expression of the ten candidate genes in individual cases (Supplementary Figure 1). The resulting bean plots show that for individual cases the copy number status is in accordance with a de-regulated gene expression. However, in a sub-group of cases the de-regulated expression cannot be explained by a genomic CNA. Thus, from this comparison it is apparent that an inter-individual variation in the regulation of gene expression may exist.
The observation that there is no absolute correlation of mRNA expression with the observed CNA is not surprising because it is known from a study by Jarvinen et al. (2008) that the extent of correlation between CNA and corresponding mRNA expression is highly variable due to other factors regulating gene expression such as miRNA expression. A surprising result from our study is that genomic CNA are very similar between thyroid neoplasms from Trk-T1 transgenic mice and human PTC with predominant involvement of RET/PTC rearrangements or BRAF mutations (Table 3). It has been shown in an earlier study (Unger et al. 2008) that major differences in CNA in PTC depending on the involvement of alterations in the MAP kinase pathway exist. The detected CNA in our study might either interact with the MAP kinase pathway or in PTC cases negative for BRAF mutations and RET/PTC rearrangements indicate alternative routes of tumour development involving different pathways.
Another interesting finding of the present study is the involvement of specific CNA in different age groups of human PTC. While the DNA gains on 2q37.1–37.3 and 1p36.32–35.3 and the DNA loss on 4q22.3 are exclusively observed in PTC from young patients (Unger et al. 2008, Hess et al. 2011) the DNA gain on 5q21.1 is more frequently (P<0.001) observed in PTC from adults (Unger et al. 2008) and thus pointing to different age-related genetic routes of tumour development in these patients.
The most prominent alterations in mice were copy number gain on murine chromosome 1C4–E2.3 and the copy number loss on murine chromosome 14E4–E5 because the syntenic regions on 2q37.1–37.3 and 13q31.2–33.1 were most frequently altered in human PTC from young patients. The genes TWIST2, PTMA and PDE6D map to chromosomal band 2q37. An association with cancer has been reported for prothymosin-α (PTMA) that is involved in chromatin de-condensation (Karetsou et al. 2004). An up-regulation of this gene has been observed in human gastric cancer (Leys et al. 2007) and has been described as a prediction marker for radiosensitivity in rectal cancer (Ojima et al. 2007). Phosphodiesterase 6D (PDE6D) interacts with HRAS (Hanzal-Bayer et al. 2002, Nancy et al. 2002). HRAS itself is frequently involved in thyroid carcinogenesis (Nikiforov 2008). TWIST2 over-expression overcomes pre-mature senescence of epithelial cells and cooperates with oncoproteins (such as RAS and ERBB2) thereby promoting epithelial–mesenchymal transition and invasiveness (Ansieau et al. 2008). GO-term enrichment analysis also revealed a significant association of TWIST2 with biological processes such as cell differentiation and cellular development that are assumed to be important in tumour development. Of note, the highly related TWIST1 was recently found to be up-regulated in anaplastic thyroid cancer (Salerno et al. 2011). As far as the copy number loss on 13q31.2–33.1 is concerned, the SLITRK5 gene and the linked SLITRK gene cluster appear to be the most promising candidates because no other known gene is located within this chromosomal band. Thus, several SLITRK family members may be involved in human PTC development while only SLITRK5 seems to be involved in mouse tumours, because it is the only member located within murine 14E4–E5, while SLITRK1 and SLITRK6 are located on the chromosomal band 14E3 which is not affected by the observed deletion in Trk-T1 transgenic thyroid tumours. Members of the SLITRK family are integral membrane proteins with homology to the SLIT family in the N-terminal extracellular domain and to the TRK neurotrophin receptors in the C-terminal intracellular domain (Aruga & Mikoshiba 2003, Aruga et al. 2003). While the biological function of SLITRK genes is not elucidated so far, it is interesting to note that these genes are expressed in haematopoietic stem cells and leukaemia and are novel markers of haematopoietic stem and progenitor cells (Milde et al. 2007).
In conclusion, molecular cytogenetic analyses of thyroid tumours from the Trk-T1 transgenic mouse model have revealed four recurrent CNA on murine chromosomes 1C4–E2.3, 3H1–H3, 4D2.3–E2 and 14E–E5 that point to Ptma, Pde6d, Twist2, Taf12, Ythdf2, Srm, Tp73 and Slitrk genes as novel candidate genes in thyroid tumourigenesis. Syntenic human chromosomal regions are frequently affected by CNA in human PTC indicating a role of these candidate genes in human carcinogenesis as well. Moreover, some specific CNA can be attributed to different age groups of PTC patients indicating age-dependent routes of tumour development. Thus, this mouse model provides a powerful tool to unravel the molecular mechanisms of thyroid tumourigenesis.
Supplementary data
This is linked to the online version of the paper at http://dx.doi.org/10.1530/ERC-11-0387.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This study was supported by the EU grant FI6R-CT-2006-36495 (GENRISK-T). This study has also been in part supported by a grant of the Italian Association for Cancer Research (AIRC) to M Santoro.
Acknowledgements
We thank the International Pathology Panel of the Chernobyl Tissue Bank for confirmation of diagnosis: Dr A Abrosimov, Prof. T I Bogdanova, Prof. V LiVolsi, Prof. M Ito, Prof. J Rosai and Prof. E D Williams. We further thank F Basolo for non-radiation associated thyroid tumour RNAs and J Rothstein for the Tg-Trk-T1 mice. We thank A Greco and M A Pierotti for the Trk-T1 molecular construct. The excellent technical assistance of E Konhäuser is gratefully acknowledged.
References
Ansieau S, Bastid J, Doreau A, Morel AP, Bouchet BP, Thomas C, Fauvet F, Puisieux I, Doglioni C & Piccinin S et al. 2008 Induction of EMT by twist proteins as a collateral effect of tumor-promoting inactivation of premature senescence. Cancer Cell 14 79–89. doi:10.1016/j.ccr.2008.06.005.
Aruga J & Mikoshiba K 2003 Identification and characterization of Slitrk, a novel neuronal transmembrane protein family controlling neurite outgrowth. Molecular and Cellular Neurosciences 24 117–129. doi:10.1016/S1044-7431(03)00129-5.
Aruga J, Yokota N & Mikoshiba K 2003 Human SLITRK family genes: genomic organization and expression profiling in normal brain and brain tumor tissue. Gene 315 87–94. doi:10.1016/S0378-1119(03)00715-7.
Birnboim HC & Doly J 1979 A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Research 7 1513–1523. doi:10.1093/nar/7.6.1513.
Bongarzone I, Pierotti MA, Monzini N, Mondellini P, Manenti G, Donghi R, Pilotti S, Grieco M, Santoro M & Fusco A et al. 1989 High frequency of activation of tyrosine kinase oncogenes in human papillary thyroid carcinoma. Oncogene 4 1457–1462.
Chung YJ, Jonkers J, Kitson H, Fiegler H, Humphray S, Scott C, Hunt S, Yu Y, Nishijima I & Velds A et al. 2004 A whole-genome mouse BAC microarray with 1-Mb resolution for analysis of DNA copy number changes by array comparative genomic hybridization. Genome Research 14 188–196. doi:10.1101/gr.1878804.
Greco A, Miranda C & Pierotti MA 2010 Rearrangements of NTRK1 gene in papillary thyroid carcinoma. Molecular and Cellular Endocrinology 321 44–49. doi:10.1016/j.mce.2009.10.009.
Hanzal-Bayer M, Renault L, Roversi P, Wittinghofer A & Hillig RC 2002 The complex of Arl2-GTP and PDE delta: from structure to function. EMBO Journal 21 2095–2106. doi:10.1093/emboj/21.9.2095.
Hedinger C, Williams ED & Sobin LH 1989 The WHO histological classification of thyroid tumors: a commentary on the second edition. Cancer 63 908–911. doi:10.1002/1097-0142(19890301)63:5<908::AID-CNCR2820630520>3.0.CO;2-I.
Hess J, Thomas G, Braselmann H, Bauer V, Bogdanova T, Wienberg J, Zitzelsberger H & Unger K 2011 Gain of chromosome band 7q11 in papillary thyroid carcinomas of young patients is associated with exposure to low-dose irradiation. PNAS 108 9595–9600. doi:10.1073/pnas.1017137108.
Hupe P, Stransky N, Thiery JP, Radvanyi F & Barillot E 2004 Analysis of array CGH data: from signal ratio to gain and loss of DNA regions. Bioinformatics 20 3413–3422. doi:10.1093/bioinformatics/bth418.
Jarvinen AK, Autio R, Kilpinen S, Saarela M, Leivo I, Grenman R, Makitie AA & Monni O 2008 High-resolution copy number and gene expression microarray analyses of head and neck squamous cell carcinoma cell lines of tongue and larynx. Genes, Chromosomes & Cancer 47 500–509. doi:10.1002/gcc.20551.
Jhiang SM, Sagartz JE, Tong Q, Parker-Thornburg J, Capen CC, Cho JY, Xing S & Ledent C 1996 Targeted expression of the ret/PTC1 oncogene induces papillary thyroid carcinomas. Endocrinology 137 375–378. doi:10.1210/en.137.1.375.
Jokinen MP & Botts S 1994 Tumours of the thyroid gland. IARC Scientific Publications 111 565–594.
Karetsou Z, Martic G, Tavoulari S, Christoforidis S, Wilm M, Gruss C & Papamarcaki T 2004 Prothymosin alpha associates with the oncoprotein SET and is involved in chromatin decondensation. FEBS Letters 577 496–500. doi:10.1016/j.febslet.2004.09.091.
Kim CS & Zhu X 2009 Lessons from mouse models of thyroid cancer. Thyroid 19 1317–1331. doi:10.1089/thy.2009.1609.
Knauf JA & Fagin JA 2009 Role of MAPK pathway oncoproteins in thyroid cancer pathogenesis and as drug targets. Current Opinion in Cell Biology 21 296–303. doi:10.1016/j.ceb.2009.01.013.
La Rosa P, Viara E, Hupe P, Pierron G, Liva S, Neuvial P, Brito I, Lair S, Servant N & Robine N et al. 2006 VAMP: visualization and analysis of array-CGH, transcriptome and other molecular profiles. Bioinformatics 22 2066–2073. doi:10.1093/bioinformatics/btl359.
Leys CM, Nomura S, LaFleur BJ, Ferrone S, Kaminishi M, Montgomery E & Goldenring JR 2007 Expression and prognostic significance of prothymosin-alpha and ERp57 in human gastric cancer. Surgery 141 41–50. doi:10.1016/j.surg.2006.05.009.
Liva S, Hupe P, Neuvial P, Brito I, Viara E, La Rosa P & Barillot E 2006 CAPweb: a bioinformatics CGH array analysis platform. Nucleic Acids Research 34 W477–W481. doi:10.1093/nar/gkl215.
Livak KJ & Schmittgen TD 2001 Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) method. Methods 25 402–408. doi:10.1006/meth.2001.1262.
Milde T, Shmelkov SV, Jensen KK, Zlotchenko G, Petit I & Rafii S 2007 A novel family of slitrk genes is expressed on hematopoietic stem cells and leukemias. Leukemia 21 824–827. doi:10.1038/sj.leu.2404525.
Nancy V, Callebaut I, El Marjou A & de Gunzburg J 2002 The delta subunit of retinal rod cGMP phosphodiesterase regulates the membrane association of Ras and Rap GTPases. Journal of Biological Chemistry 277 15076–15084. doi:10.1074/jbc.M109983200.
Neuvial P, Hupe P, Brito I, Liva S, Manie E, Brennetot C, Radvanyi F, Aurias A & Barillot E 2006 Spatial normalization of array-CGH data. BMC Bioinformatics 7 264 doi:10.1186/1471-2105-7-264.
Nikiforov YE 2008 Thyroid carcinoma: molecular pathways and therapeutic targets. Modern Pathology 21 (Suppl 2) S37–S43. doi:10.1038/modpathol.2008.10.
Ojima E, Inoue Y, Miki C, Mori M & Kusunoki M 2007 Effectiveness of gene expression profiling for response prediction of rectal cancer to preoperative radiotherapy. Journal of Gastroenterology 42 730–736. doi:10.1007/s00535-007-2089-x.
Osoegawa K, Tateno M, Woon PY, Frengen E, Mammoser AG, Catanese JJ, Hayashizaki Y & de Jong PJ 2000 Bacterial artificial chromosome libraries for mouse sequencing and functional analysis. Genome Research 10 116–128. doi:10.1101/gr.10.1.116.
Osoegawa K, Mammoser AG, Wu C, Frengen E, Zeng C, Catanese JJ & de Jong PJ 2001 A bacterial artificial chromosome library for sequencing the complete human genome. Genome Research 11 483–496. doi:10.1101/gr.169601.
Powell DJ Jr, Russell J, Nibu K, Li G, Rhee E, Liao M, Goldstein M, Keane WM, Santoro M & Fusco A et al. 1998 The RET/PTC3 oncogene: metastatic solid-type papillary carcinomas in murine thyroids. Cancer Research 58 5523–5528.
Powell N, Jeremiah S, Morishita M, Dudley E, Bethel J, Bogdanova T, Tronko M & Thomas G 2005 Frequency of BRAF T1796A mutation in papillary thyroid carcinoma relates to age of patient at diagnosis and not to radiation exposure. Journal of Pathology 205 558–564. doi:10.1002/path.1736.
Richter H, Slezak P, Walch A, Werner M, Braselmann H, Jaramillo E, Ost A, Hirata I, Takahama K & Zitzelsberger H 2003 Distinct chromosomal imbalances in nonpolypoid and polypoid colorectal adenomas indicate different genetic pathways in the development of colorectal neoplasms. American Journal of Pathology 163 287–294. doi:10.1016/S0002-9440(10)63652-8.
Rouveirol C, Stransky N, Hupe P, Rosa PL, Viara E, Barillot E & Radvanyi F 2006 Computation of recurrent minimal genomic alterations from array-CGH data. Bioinformatics 22 849–856. doi:10.1093/bioinformatics/btl004.
Russell JP, Powell DJ, Cunnane M, Greco A, Portella G, Santoro M, Fusco A & Rothstein JL 2000 The TRK-T1 fusion protein induces neoplastic transformation of thyroid epithelium. Oncogene 19 5729–5735. doi:10.1038/sj.onc.1203922.
Salerno P, Garcia-Rostan G, Piccinin S, Bencivenga TC, Di Maro G, Doglioni C, Basolo F, Maestro R, Fusco A & Santoro M et al. 2011 TWIST1 plays a pleiotropic role in determining the anaplastic thyroid cancer phenotype. Journal of Clinical Endocrinology and Metabolism 96 E772–E781. doi:10.1210/jc.2010-1182.
Santoro M, Chiappetta G, Cerrato A, Salvatore D, Zhang L, Manzo G, Picone A, Portella G, Santelli G & Vecchio G et al. 1996 Development of thyroid papillary carcinomas secondary to tissue-specific expression of the RET/PTC1 oncogene in transgenic mice. Oncogene 12 1821–1826.
Unger K, Zitzelsberger H, Salvatore G, Santoro M, Bogdanova T, Braselmann H, Kastner P, Zurnadzhy L, Tronko N & Hutzler P et al. 2004 Heterogeneity in the distribution of RET/PTC rearrangements within individual post-Chernobyl papillary thyroid carcinomas. Journal of Clinical Endocrinology and Metabolism 89 4272–4279. doi:10.1210/jc.2003-031870.
Unger K, Malisch E, Thomas G, Braselmann H, Walch A, Jackl G, Lewis P, Lengfelder E, Bogdanova T & Wienberg J et al. 2008 Array CGH demonstrates characteristic aberration signatures in human papillary thyroid carcinomas governed by RET/PTC. Oncogene 27 4592–4602. doi:10.1038/onc.2008.99.
Zheng Q & Wang XJ 2008 GOEAST: a web-based software toolkit for Gene Ontology enrichment analysis. Nucleic Acids Research 36 W358–W363. doi:10.1093/nar/gkn276.
(K-J Heiliger and J Hess contributed equally to this work)
