Recurrent R-spondin fusions in colon cancer

Abstract

Identifying and understanding changes in cancer genomes is essential for the development of targeted therapeutics. Here we analyse systematically more than 70 pairs of primary human colon tumours by applying next-generation sequencing to characterize their exomes, transcriptomes and copy-number alterations. We have identified 36,303 protein-altering somatic changes that include several new recurrent mutations in the Wnt pathway gene TCF7L2, chromatin-remodelling genes such as TET2 and TET3 and receptor tyrosine kinases including ERBB3. Our analysis for significantly mutated cancer genes identified 23 candidates, including the cell cycle checkpoint kinase ATM. Copy-number and RNA-seq data analysis identified amplifications and corresponding overexpression of IGF2 in a subset of colon tumours. Furthermore, using RNA-seq data we identified multiple fusion transcripts including recurrent gene fusions involving R-spondin family members RSPO2 and RSPO3 that together occur in 10% of colon tumours. The RSPO fusions were mutually exclusive with APC mutations, indicating that they probably have a role in the activation of Wnt signalling and tumorigenesis. Consistent with this we show that the RSPO fusion proteins were capable of potentiating Wnt signalling. The R-spondin gene fusions and several other gene mutations identified in this study provide new potential opportunities for therapeutic intervention in colon cancer.

Figure 1. Somatic mutations in colon cancer

a, Number of mutations in each of the MSS, MSI and hypermutated subtypes. b, Transitions and transversions at the individual sample level for the three groups in the same order as shown in a. c, Transitions and transversions averaged across each of the three sample groups. d, Circos plot representation of the whole genome of an MSS and an MSI colon tumour. The outer blue band shows copy number as measured using WGS reads. Red dots indicate somatic non-synonymous, splice-site and stop-gained mutations (MSS genome, 17,651 mutations, average of 6.2 mutations per 10⁶ bases, 3,257 mutated genes; MSI genome, 97,968 mutations, average of 34.5 mutations per 10⁶ bases, 9,851 mutated genes). Grey dots indicate all other somatic mutation types. Orange and grey lines show predicted intra- and interchromosomal rearrangements, respectively.

Figure 2. Significantly mutated colon cancer genes

a, Genes evaluated for significance using Q score criteria for MSS samples are shown. Each circle represents a gene and the size of the circle is proportional to the mutation count for that gene. The genes are represented in order of increasing number of expected mutations from left to right on the x axis. Genes with a statistically significant Q score are labelled. b, TCF12 helix–loop–helix (HLH) domain structure bound to DNA showing the Arg603 hotspot and a schematic diagram depicting the various TCF12 mutations identified in this study. c, Somatic mutations in the TET gene family including TET1, TET2 and TET3 shown on the domain architecture of the TET genes. Recurrent mutations found in this study for TET3 and for TET2 by comparison to COSMIC data are shown in red. CD, Cysteine-rich domain; CXXC, Cys-X-X-Cys-type domain; DSBH, double-stranded β-helix 2OG-Fe(II)-dependent dioxygenase domain. Asterisks in b and c denote nonsense mutations that result in premature stop codons.

Figure 3. Recurrent R-spondin gene fusions

a, List of type and frequency of R-spondin gene fusions in colon cancer. b, Cartoon depicting the location, orientation and exon–intron architecture of EIF3E–RSPO2 fusion on the genome. The read evidence for the EIF3E(e1)–RSPO2(e2) fusion identified using RNA-seq data is shown. c, Independent reverse transcriptase (RT)–PCR-derived products confirming the EIF3E(e1)–RSPO2(e2) somatic fusion resolved on an agarose gel. The RT–PCR product was Sanger-sequenced confirming the fusion junction and a representative chromatogram is shown in b. d, Schematic of the resulting EIF3E(e1)–RSPO2(e2) fusion protein. e, Tumours harbouring R-spondin fusions show elevated expression of the corresponding RSPO gene as depicted by the heat-map. f, Cartoon depicting the location, orientation and exon–intron architecture of the PTPRK–RSPO3 gene fusion on the genome. The read evidence for PTPRK(e1)–RSPO3(e2) fusion identified using RNA-seq data is shown. g, Independent RT–PCR-derived products confirming the PTPRK(e1)–RSPO3(e2) somatic fusion resolved on an agarose gel. The RT– PCR product was Sanger-sequenced confirming the fusion junction, and a representative chromatogram is show in panel f. h, Schematic of PTPRK, RSPO3 and the resulting PTPRK(e1)–RSPO3(e2) fusion proteins. FN3, fibronectin type 3 domain; FU, furin-like repeats; IG, immunoglobulin; MAM, meprin A5 receptor protein tyrosine phosphatase mu domain; N, normal; PTPc, protein tyrosine phosphatase catalytic domain; T, tumour; TM, transmembrane domain; TSP1, thrombospondin type 1 repeats.

Figure 4. RSPO fusion products activate Wnt signalling

a, Secreted RSPO fusion proteins detected by western blot in media from 293T cells transfected with expression constructs encoding the fusion proteins. The arrowhead indicates the expected product for RSPO 1–387. b, RSPO fusion proteins activate and potentiate Wnt signalling as measured using a luciferase reporter assay (see Supplementary Methods). Data shown are from conditioned media derived from cells transfected with the fusion constructs. Representative data from at least three experiments are shown. Error bars represent mean ± s.d. from three replicate experiments. EGFP, enhanced green fluorescent protein. c, Cartoon representing R-spondin-mediated Wnt signalling pathway activation. d, Quilt plot depicting RSPO fusions, somatic mutations, copy-number alteration and expression changes across a select set of Wnt-signalling pathway genes. Each column represents a tumour sample. Het, heterozygous; Hom, homozygous; WT, wild type.