Cancer Letters 357 (2015) 498–501
Contents lists available at ScienceDirect
Cancer Letters j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c a n l e t
Original Articles
Activation of human telomerase reverse transcriptase through gene fusion in clear cell sarcoma of the kidney Jenny Karlsson a,*, Henrik Lilljebjörn a, Linda Holmquist Mengelbier a, Anders Valind a, Marianne Rissler a, Ingrid Øra b, Thoas Fioretos a, David Gisselsson a,c a
Department of Clinical Genetics, BMC C13, Lund University, Lund SE 221 84, Sweden Department of Pediatric Oncology and hematology, Clinical Sciences, Lund University, Lund 221 85, Sweden c Department of Pathology, Skåne Regional and University Laboratories, Lund SE 221 85, Sweden b
A R T I C L E
I N F O
Article history: Received 9 October 2014 Received in revised form 25 November 2014 Accepted 26 November 2014 Keywords: TERT IRX2 RNA sequencing Pediatric kidney tumors Fusion transcript
A B S T R A C T
Clear cell sarcoma of the kidney (CCSK) is a rare tumor type affecting infants and young children. Most CCSKs display few genomic aberrations, and no general underlying mechanism for tumor initiation has yet been identified, although a YWHAE-NUTM2B/NUTM2E fusion gene has been observed in a minority of cases. We performed RNA-sequencing of 22 CCSKs to investigate the presence of additional fusion transcripts. The presence of the YWHAE-NUTM2B/NUTM2E fusion was confirmed in two cases. In addition, a novel IRX2-TERT fusion transcript was identified in one case. SNP-array analyses revealed the underlying event to be an interstitial deletion in the short arm of chromosome 5 (5p15.33). TERT was dramatically upregulated under the influence of the IRX2 promoter. In line with TERT expression being driven by active IRX2 regulatory elements, we found a high expression of IRX2 in CCSKs irrespective of fusion gene status. IRX2 was also expressed in human fetal kidney – the presumed tissue of origin for CCSK. We conclude that in addition to promoter mutations and epigenetic events, TERT can also be activated in tumors via formation of fusion transcripts. © 2014 Elsevier Ireland Ltd. All rights reserved.
Introduction CCSK is a rare neoplasm, accounting for 2–5% of all pediatric renal tumors. The overall survival for patients with CCSK has historically been low, but has now improved by intensified chemotherapy treatment [1]. CCSK is often described as a puzzling tumor, since its histological appearance does not resemble that of the kidney and there is no specific positive biological diagnostic marker available [2]. Furthermore, most CCSKs lack large-scale chromosomal alterations, while a subset shows a translocation between chromosomes 10 and 17 [3–8], resulting in a fusion transcript between YWHAE and NUTM2B/NUTM2E (also known as FAM22B/FAM22E) [9]. Maintenance of telomere length is essential for the immortalization of tumor cells, and reactivation of telomerase is found in approximately 90% of all cancers [10]. Telomerase reverse transcriptase (TERT), the gene coding for the catalytic subunit of telomerase, can be activated by promoter mutations in different
Abbreviations: AK, adult kidney; CCSK, clear cell sarcoma of the kidney; FK, fetal kidney; IRX2, Iroquois homeobox 2; NUTM2B, NUT family member 2B; NUTM2E, NUT family member 2E; SNP, single nucleotide polymorphism; TERT, telomerase reverse transcriptase; WT, Wilms tumor; YWHAE, tyrosine 3-monooxygenase/ tryptophan 5-monooxygenase activation protein, epsilon. * Corresponding author. Tel.: +46 46 222 94 00; fax +46 46 13 10 61. E-mail address:
[email protected] (J. Karlsson). http://dx.doi.org/10.1016/j.canlet.2014.11.057 0304-3835/© 2014 Elsevier Ireland Ltd. All rights reserved.
neoplasms [11–13]. However, such mutations are not present in all investigated tumor types [13], indicating that TERT can be activated via other, unknown mechanisms. We here describe TERT overexpression resulting from gene fusion between TERT and the transcription factor IRX2 in a CCSK. The novel IRX2-TERT fusion was found during exploratory RNA sequencing of a CCSK cohort. Materials and methods Tumor and kidney tissue The study was reviewed and approved by the regional ethics committee (L11903) and by the review boards of the participating institutes. Frozen CCSKs were obtained from the Children’s Oncology Group (COG) of North America and from Skåne University Hospital, Lund, Sweden. CCSK diagnoses were corroborated by reference pathologists of COG for the American cases and the International Society of Pediatric Oncology for the two Swedish cases. Wilms tumors for comparative studies were obtained from Skåne University Hospital and from the Academic Medical Center, Amsterdam, the Netherlands. Frozen non-neoplastic kidney tissue from 6 adult nephrectomy patients were obtained from the biobank at the Department of Pathology at Lund University (Ethics Approval L2012-405) and the total RNA from 4 normal fetal kidneys was acquired from BioChain, Hayward, CA, Stratagene/Agilent Technologies, Santa Clara, CA and Clontech, Mountain View, CA. DNA extraction and Single nucleotide polymorphism (SNP)-array analyses DNA from CCSKs was extracted with the DNeasy Blood & Tissue Kit, (Qiagen, Valencia, CA), according to the manufacturer’s recommendations. The fusionpositive case was analyzed, as part of a bigger CCSK cohort, with the 1M Genotyping
J. Karlsson et al./Cancer Letters 357 (2015) 498–501
BeadChip SNP-array (Illumina Inc., San Diego, CA) according to standard methods, as previously described [14]. RNA extraction and gene expression array RNA was extracted from CCSKs, Wilms tumors and non-neoplastic kidney with the RNeasy Lipid Tissue Mini Kit (Qiagen, Valencia, CA) and hybridized to HumanHT12 v4 Expression BeadChips (Illumina) as previously described [14]. The CCSKs were first analyzed together with the non-neoplastic kidney tissue (the Gene Expression Omnibus database, accession number GSE49972). The Wilms tumors were analyzed in a second experiment that also included overlapping samples from the previous run (2 CCSKs, 2 adult kidneys and 4 fetal kidneys). Batch effects caused by running the samples in two different experiments were removed with the ComBat algorithm [15] and the adequacy of this method was confirmed by the similar localization of the duplicates in a principal component analysis plot (data not shown). The expressions of TERT and IRX2 were visualized with scatter plots generated with Qlucore Omics Explorer Software v2.3, (Qlucore AB, Lund, Sweden). RNA sequencing and data analysis mRNA libraries were created with the Truseq RNA sample preparation kit v2 (Illumina) according to the manufacturer’s instructions. 101 bp paired-end reads were generated using a HiScanSQ (Illumina). Fusion transcript candidates were identified with TopHat v 2.0.7 as previously described [16] and with ChimeraScan 0.4.5 [17]. Putative fusions reported by ChimeraScan were included in the study if they were covered by more than 5 spanning reads or more than 10 reads in total. Fusion transcript candidates generated by both TopHat and ChimeraScan were not included for further analyses if they involved non-coding genes, overlapping genes or were indicated as readthrough, together with candidates encompassing closely located genes with no copy number changes in the segment between them. Confirmation of potential fusion transcript candidates with RT-PCR and genomic PCR RNA from CCSKs was converted to cDNA with random hexamers and Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA). The IRX2TERT fusion product was amplified with primers binding to the third exon of IRX2 (5′-GACTCGCTCACGGATCACTC-3′) and the second exon of TERT (5′-ACCGTGTTGGG CAGGTAG-3′). The amplified fragment was sequenced with a forward primer (5′CAAGTATGACGACCTGGAGGA-3′) using the Bigdye Cycle Sequencing kit (Life Technologies, Carlsbad, CA, USA) on an ABI3130 genetic analyzer (Life Technologies). The chromatogram was visualized with Chromas Lite 2.01 (Technelysium Pty Ltd, South Brisbane, Australia). The genomic breakpoints were established by performing PCR on DNA which was extracted as described above. The same primers that were used to amplify cDNA, see above, were also used to amplify the genomic DNA and to sequence the PCR product.
Results A panel of 22 CCSK tumors was subjected to RNA sequencing to investigate the occurrence of fusion transcripts. A novel IRX2-TERT gene fusion was identified in a CCSK from a 4 year old boy (Fig. 1a). The finding was confirmed with RT-PCR followed by Sanger sequencing (Fig. 1c). The fusion occurred between amino acid 332 in exon 3 of IRX2 and amino acid 2 in exon 1 of TERT (IRX2 transcripts NM_001134222 and NM_033267; TERT transcripts NM_198253 and NM_001193376). As a result, this in-frame fusion transcript included the entire coding sequence of TERT with the exception of the start codon. The already published YWHAE-NUTM2B/ NUTM2E fusion transcript was detected in 2 CCSKs. The presence of these fusion transcripts was also confirmed by RT-PCR (data not shown). Four additional fusion transcript candidates from the ChimeraScan list met the inclusion criteria (see the Materials and Methods section). However, all of these were detected by unconventional sequence reads, such as breakpoints in UTR-regions. In addition, none of these potential fusions were supported by the SNP-array data from the same cases and as expected, we were unable to confirm the presence of any of these candidates with RT-PCR, despite running PCR reactions at different annealing temperatures. We also specifically searched our dataset for additional fusion transcript candidates involving IRX2, TERT, YWHAE or NUTM2B/NUTM2E in our 22 CCSKs, but no other potential fusions were discovered, despite investigating the RNA sequencing data without a prior removal of candidates
499
with a low number of reads. Despite having a unique fusion transcript, the IRX2-TERT positive case showed a global gene expression profile that matched other cases of CCSK in the cohort (Appendix: Supplementary Fig. S1a). SNP-array analysis of the tumor with IRX2-TERT fusion demonstrated a complex rearrangement in the short arm of chromosome 5 (5p), including a hemizygous deletion of the region between TERT and IRX2 (Fig. 1b). A detailed analysis of the genome profile revealed a hemizygous deletion of at least 1.4 Mbp between rs4635969 at genomic position 1308552 and rs204783 at position 2728882 (hg19), in clear agreement with the RNA sequencing data. PCR on genomic DNA from the IRX2-TERT positive case, followed by Sanger sequencing, revealed that the genomic breakpoints perfectly corresponded to the breakpoints found by RNA sequencing and RTPCR (Appendix: Supplementary Fig. S1b and c). No other case in the cohort of totally 37 CCSKs, including those subjected to RNA sequencing, showed 5p rearrangements by SNP array [14]. To explore potential alterations in gene expression due to the IRX2-TERT fusion, we compared IRX2 and TERT mRNA levels in the fusion-positive case with those in the fusion-negative CCSKs. In the same experiment we also included RNA from normal fetal and adult kidney tissue, together with mRNA from Wilms tumor – the most common pediatric renal tumor and a differential diagnosis to CCSK. The expression of IRX2 mRNA was higher in CCSK than in Wilms tumor and in normal kidney (Fig. 1d). The fusion-positive case displayed a lower IRX2 expression compared to the majority of the CCSKs, in line with the partial deletion of IRX2 in this sample. In accordance with influence from the IRX2 promoter region, the expression of TERT was dramatically increased in the IRX2-TERT fusion positive case, compared to other CCSKs, Wilms tumors, and normal kidney samples (Fig. 1e).
Discussion The mechanism behind the reactivation of telomerase in tumor cells has for a long time been unknown, but the recent discovery of activating mutations in the TERT promoter has offered an explanation applicable to several tumor types [11–13]. In this study, we show that TERT also can be activated via gene fusion. By RNA sequencing of 22 cases of the uncommon pediatric tumor CCSK, we found a novel in-frame fusion transcript between IRX2 and TERT in one of the tumors, resulting from a hemizygous interstitial deletion. The RNA sequencing data revealed a breakpoint in exon 2 of the 5′-fusion partner IRX2 and in the first exon following the TERT start codon. Due to the scarce distribution of reporters in the SNP-array, it could not be determined from this analysis if the unusual breakpoint, truncating two exons, represents the actual genomic breakpoint of the deletion. We could by PCR on genomic DNA followed by Sanger sequencing confirm that the breakpoints were identical on the RNA- and DNA levels. The IRX2-TERT fusion positive tumor displayed a dramatic increase in TERT-expression compared to the other CCSKs, Wilms tumors and the normal kidney tissue. However the global gene expression profile of this tumor is in clear agreement with a CCSK diagnosis (Appendix: Supplementary Fig. S1a and [14]). Based on the low TERT-expression levels in the fusion negative CCSKs, TERT activation through high expression levels does not appear to be a general characteristic of CCSK. IRX2 is a transcription factor that together with IRX1 and IRX4 constitutes the IRXA cluster. IRX2 is involved in various embryologic processes such as cerebellum formation [18] and retinogenesis [19]. It is expressed in the embryonic kidney of Xenopus and mouse, but is not crucial for proper nephrogenesis [20]. We found that IRX2 also is expressed in the human fetal kidney, although the expression was not as high as in CCSKs.
500
J. Karlsson et al./Cancer Letters 357 (2015) 498–501
A
B
1.0
Log2 ratio
BAF
chromosome 5
deletion 5’
3’
IRX2
TERT
0.5 0.0 2.0 0.0 -2.0
TERT
IRX2
C exon 3
exon 1
IRX2
TERT
D
E IRX2
TERT
Gene expression (Log2)
1.5
7.0 6.0
1.0
5.0 0.5
4.0 3.0
0.0
2.0 -0.5
1.0 0.0
-1.0
-1.0
FK
AK
WT
CCSK
FK
AK
WT
CCSK
Fig. 1. TERT activation through interstitial deletion and gene fusion. (a) TERT and IRX2 are located on the minus strand on the distal part of the short arm of chromosome 5 (5p15.33). In one case of CCSK, an IRX2-TERT fusion transcript was formed through a hemizygous deletion of a segment between these genes. (b) SNP-array analysis of the IRX2-TERT fusion positive case with focus on the area surrounding TERT and IRX2. The top graph demonstrates the loss of heterozygous SNP markers between TERT and IRX2, leading to a shift in B-allele frequency (BAF) from around 0.5 to 0 or 1. This is accompanied by a drop in the log2-ratio from the diploid state (log2 positioned close to 0) in the same segment. Reporters for SNPs in the deleted area are shown in red. (c) Chromatogram showing the IRX2-TERT fusion junction obtained by Sanger sequencing of cDNA. (d and e) Scatter plots demonstrating the expression levels of (d) IRX2 and (e) TERT (mean of the signal from two reporters on the gene expression array) in clear cell sarcoma of the kidney (CCSK), Wilms tumor (WT), and normal kidney from adults (AK) and fetuses (FK). The IRX2-TERT fusion positive CCSK is indicated with an arrow in (d) and (e). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Based on its presence in the fetal kidney, IRX2 is an interesting fusion partner in the context of CCSK, as CCSK appears to originate from primitive nephrogenic cells [14]. In addition to the IRX2-TERT fusion transcript we detected a fusion between YWHAE-NUTM2B/NUTM2E in two other cases, both previously been reported by O’Meara et al. to contain that particular fusion [9]. The ability to reproduce these findings corroborated the solidity of our RNA-sequencing platform. O’Meara et al. detected
the YWHAE-NUTM2B/NUTM2E fusion transcript by RT-PCR in 6 out of 50 (12%) analyzed tumors, a slightly higher frequency than the 2/22 (9%) cases analyzed here. No other fusion transcript candidates were detected in the remaining 19/22 (86%) analyzed CCSKs in the present study. This is consistent with the mostly normal karyotype and the few segmental aberrations reported for this tumor type [3–8]. A common genetic aberration behind the development of CCSK is yet to be discovered.
J. Karlsson et al./Cancer Letters 357 (2015) 498–501
In summary we report a proof-of-principle case demonstrating that TERT, in addition to promoter mutations and epigenetic events [21], also can be activated via formation of fusion transcripts. Acknowledgements We are grateful to the COG Renal Tumor Study Group and the Tissue Bank for contributing with CCSK samples and Dr Martin Johansson at the Center for Molecular Pathology, Lund University, SUS Malmö, Sweden, for providing kidney tissue. We would also like to thank the Swegene Centre for Integrative Biology at Lund University (SCIBLU). This work was supported by the Swedish Childhood Cancer Foundation (PROJ 2013-0011), the Swedish Cancer Society (CAN 2012/ 318), the Swedish Research Council (2013-2543), the Crafoord Foundation, the Gunnar Nilsson Cancer Foundation, the Royal Physiographic Society, and BioCare Sweden. Conflict of interest None. Appendix: Supplementary material Supplementary data to this article can be found online at doi:10.1016/j.canlet.2014.11.057. References [1] N.L. Seibel, S. Li, N.E. Breslow, J.B. Beckwith, D.M. Green, G.M. Haase, M.L. Ritchey, P.R. Thomas, P.E. Grundy, J.Z. Finklestein, T. Kim, S.J. Shochat, P.P. Kelalis, G.J. D’Angio, Effect of duration of treatment on treatment outcome for patients with clear-cell sarcoma of the kidney: a report from the National Wilms’ Tumor Study Group. J. Clin. Oncol. 22 (2004) 468–473, doi:10.1200/JCO.2004.06. 058JCO.2004.06.058 [pii]. [2] P. Argani, E.J. Perlman, N.E. Breslow, N.G. Browning, D.M. Green, G.J. D’Angio, J.B. Beckwith, Clear cell sarcoma of the kidney: a review of 351 cases from the National Wilms Tumor Study Group Pathology Center. Am. J. Surg. Pathol. 24 (2000) 4–18. [3] E.C. Douglass, J.A. Wilimas, A.A. Green, A.T. Look, Abnormalities of chromosomes 1 and 11 in Wilms’ tumor. Cancer Genet. Cytogenet. 14 (1985) 331–338. [4] H.H. Punnett, G.E. Halligan, N. Zaeri, N. Karmazin, Translocation 10;17 in clear cell sarcoma of the kidney. A first report. Cancer Genet. Cytogenet. 41 (1989) 123–128. [5] W.W. Sheng, S. Soukup, K. Bove, B. Gotwals, B. Lampkin, Chromosome analysis of 31 Wilms’ tumors. Cancer Res. 50 (1990) 2786–2793.
501
[6] Y. Kaneko, C. Homma, N. Maseki, M. Sakurai, J. Hata, Correlation of chromosome abnormalities with histological and clinical features in Wilms’ and other childhood renal tumors. Cancer Res. 51 (1991) 5937–5942. [7] D. Rakheja, A.G. Weinberg, G.E. Tomlinson, K. Partridge, N.R. Schneider, Translocation (10;17)(q22;p13): a recurring translocation in clear cell sarcoma of kidney. Cancer Genet. Cytogenet. 154 (2004) 175–179. [8] N.A. Brownlee, L.A. Perkins, W. Stewart, B. Jackle, M.J. Pettenati, P.P. Koty, S.S. Iskandar, A.J. Garvin, Recurring translocation (10;17) and deletion (14q) in clear cell sarcoma of the kidney. Arch. Pathol. Lab. Med. 131 (2007) 446–451. [9] E. O’Meara, D. Stack, C.H. Lee, A.J. Garvin, T. Morris, P. Argani, J.S. Han, J. Karlsson, D. Gisselson, I. Leuschner, M. Gessler, N. Graf, J.A. Fletcher, M.J. O’Sullivan, Characterization of the chromosomal translocation t(10;17)(q22;p13) in clear cell sarcoma of kidney. J. Pathol. 227 (2012) 72–80, doi:10.1002/path.3985. [10] J.W. Shay, S. Bacchetti, A survey of telomerase activity in human cancer. Eur. J. Cancer 33 (1997) 787–791, doi:10.1016/S0959-8049(97)00062-2. [11] F.W. Huang, E. Hodis, M.J. Xu, G.V. Kryukov, L. Chin, L.A. Garraway, Highly recurrent TERT promoter mutations in human melanoma. Science 339 (2013) 957–959, doi:10.1126/science.1229259. [12] S. Horn, A. Figl, P.S. Rachakonda, C. Fischer, A. Sucker, A. Gast, S. Kadel, I. Moll, E. Nagore, K. Hemminki, D. Schadendorf, R. Kumar, TERT promoter mutations in familial and sporadic melanoma. Science 339 (2013) 959–961, doi:10.1126/ science.1230062. [13] J. Vinagre, A. Almeida, H. Populo, R. Batista, J. Lyra, V. Pinto, R. Coelho, R. Celestino, H. Prazeres, L. Lima, M. Melo, A.G. da Rocha, A. Preto, P. Castro, L. Castro, F. Pardal, J.M. Lopes, L.L. Santos, R.M. Reis, J. Cameselle-Teijeiro, M. Sobrinho-Simoes, J. Lima, V. Maximo, P. Soares, Frequency of TERT promoter mutations in human cancers. Nat. Commun. 4 (2013) 2185, doi:10.1038/ ncomms3185. [14] J. Karlsson, L. Holmquist Mengelbier, C.D. Ciornei, A. Naranjo, M.J. O’Sullivan, D. Gisselsson, Clear cell sarcoma of the kidney demonstrates an embryonic signature indicative of a primitive nephrogenic origin. Genes Chromosomes Cancer (2014) 53 281–391, doi:10.1002/gcc.22149. [15] J.T. Leek, W.E. Johnson, H.S. Parker, A.E. Jaffe, J.D. Storey, The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics 28 (2012) 882–883, doi:10.1093/bioinformatics/ bts034. [16] C. Walther, J. Tayebwa, H. Lilljebjorn, L. Magnusson, J. Nilsson, F.V. von Steyern, I. Ora, H.A. Domanski, T. Fioretos, K.H. Nord, C.D. Fletcher, F. Mertens, A novel SERPINE1-FOSB fusion gene results in transcriptional up-regulation of FOSB in pseudomyogenic haemangioendothelioma. J. Pathol. (2013) 232 534–540. doi:10.1002/path.4322. [17] M.K. Iyer, A.M. Chinnaiyan, C.A. Maher, ChimeraScan: a tool for identifying chimeric transcription in sequencing data. Bioinformatics 27 (2011) 2903–2904, doi:10.1093/bioinformatics/btr467. [18] K. Matsumoto, S. Nishihara, M. Kamimura, T. Shiraishi, T. Otoguro, M. Uehara, Y. Maeda, K. Ogura, A. Lumsden, T. Ogura, The prepattern transcription factor Irx2, a target of the FGF8/MAP kinase cascade, is involved in cerebellum formation. Nat. Neurosci. 7 (2004) 605–612, doi:10.1038/nn1249. [19] S.W. Choy, C.W. Cheng, S.T. Lee, V.W. Li, M.N. Hui, C.C. Hui, D. Liu, S.H. Cheng, A cascade of irx1a and irx2a controls shh expression during retinogenesis. Dev. Dyn. 239 (2010) 3204–3214, doi:10.1002/dvdy.22462. [20] C. Heliot, A. Desgrange, I. Buisson, R. Prunskaite-Hyyrylainen, J. Shan, S. Vainio, M. Umbhauer, S. Cereghini, HNF1B controls proximal-intermediate nephron segment identity in vertebrates by regulating Notch signalling components and Irx1/2. Development 140 (2013) 873–885, doi:10.1242/dev.086538. [21] M. Daniel, G.W. Peek, T.O. Tollefsbol, Regulation of the human catalytic subunit of telomerase (hTERT). Gene 498 (2012) 135–146, doi:10.1016/j.gene. 2012.01.095.