Recurrent R-spondin fusions in colon cancer

Authors

Seshagiri S

Stawiski EW

Durinck S

Modrusan Z

Storm EE

Conboy CB

Chaudhuri S

Guan Y

Janakiraman V

Jaiswal BS

Guillory J

Ha C

Dijkgraaf GJP

Stinson J

Gnad F

Huntley MA

Degenhardt JD

Haverty PM

Bourgon R

Wang W

Koeppen H

Gentleman R

Starr TK

Zhang Z

Largaespada DA

Wu TD

de Sauvage FJ

Doi

10.1038/nature11282

PMID: 22895193 · DOI: 10.1038/nature11282 · Journal: Nature (2012)

TL;DR

This study performed systematic multi-platform genomic characterization (exome sequencing, RNA-seq, SNP arrays, whole-genome sequencing) of more than 70 primary human colon tumour-normal pairs. The key discovery is recurrent gene fusions involving R-spondin family members RSPO2 and RSPO3 in approximately 10% of colon cancers. These fusions are mutually exclusive with APC mutations and activate Wnt signalling, identifying a new subset of colorectal cancers driven by an alternative mechanism of Wnt pathway activation that may be therapeutically targetable.

Cohort & data

  • 74 colon tumour-normal pairs from commercial sources (fresh-frozen primary tumours)
  • 72 pairs analysed by exome sequencing (15 MSI, 57 MSS); 68 pairs by RNA-seq; 74 pairs by Illumina 2.5M SNP arrays; 2 pairs by whole-genome sequencing
  • Dataset: coadread_genentech
  • Cancer type: COAD / COADREAD
  • Methods: whole-exome-seq, rna-seq, whole-genome-seq
  • Platform: SeqCap EZ v2.0 (NimbleGen) for exome; HiSeq 2000 (Illumina); mean exome coverage 179x

Key findings

  • Identified 36,303 protein-altering somatic changes across the cohort.
  • MSS samples: mean non-synonymous mutation frequency of 2.8 mutations per 10^6 bases (31-149 coding mutations per sample in 55 MSS samples).
  • MSI samples: 47 mutations per 10^6 bases (764-3,113 coding mutations per sample in 15 MSI samples).
  • 23 significantly mutated cancer genes identified in MSS samples (Q score >= 1; FDR <= 5%), including known drivers (KRAS, APC, TP53, SMAD4, FBXW7, PIK3CA) and new candidates including ATM.
  • Recurrent RSPO2 fusions (EIF3E-RSPO2): 3% (2/68 tumours).
  • Recurrent RSPO3 fusions (PTPRK-RSPO3): 8% (5/68 tumours).
  • Combined RSPO fusions occur in ~10% of colon tumours and are mutually exclusive with APC mutations (P = 0.038, Fisher’s exact test).
  • RSPO fusion proteins activate and potentiate Wnt signalling in luciferase reporter assays.
  • IGF2 upregulation with focal amplification in 12% (8/68) of samples.
  • MYC amplification on 8q in 23% (17/74) of samples; KRAS amplification in 13% (10/74).
  • ERBB3 mutated in 8% (6/72) with multiple oncogenic hotspot mutations.
  • New recurrent mutations found in TCF7L2 (Wnt pathway), TET2, and TET3 (chromatin remodelling / DNA demethylation).

Genes & alterations

  • RSPO2 — EIF3E(e1)-RSPO2(e2) gene fusion; drives Wnt signalling activation; 3% of tumours.
  • RSPO3 — PTPRK(e1)-RSPO3(e2) and PTPRK(e7)-RSPO3(e2) fusions; 8% of tumours.
  • APC — mutations and copy loss; mutually exclusive with RSPO fusions; deletion in 8% (6/74).
  • KRAS — significantly mutated; amplified in 13% (10/74); co-occurs with RSPO fusions.
  • TP53 — significantly mutated in both MSS and MSI samples.
  • SMAD4 — significantly mutated; altered (with SMAD2) in 27% (20/74) on chromosome 18q.
  • FBXW7 — significantly mutated in MSS CRC.
  • PIK3CA — significantly mutated in MSS CRC.
  • ATM — newly identified as significantly mutated cancer gene in CRC.
  • TCF7L2 — new recurrent mutations (Wnt pathway gene).
  • TET2 — somatic mutations; first report of TET family mutations in CRC.
  • TET3 — somatic mutations with recurrent hotspots.
  • ERBB3 — mutated in 8% (6/72) with multiple hotspot mutations; oncogenic.
  • IGF2 — focal amplification and overexpression in 12% (8/68).
  • MYC — broad amplicon on chromosome 8q; 23% (17/74).
  • BRAF — mutations co-occur with RSPO fusions in some samples.
  • CTNNB1 — Wnt pathway gene mutated in CRC; mutually exclusive with RSPO fusions.
  • PTEN — deletion in 4% (3/74).

Clinical implications

  • RSPO fusions define a new molecular subset (~10%) of colon cancers with alternative Wnt pathway activation, distinct from APC-mutant tumours.
  • RSPO fusion-positive tumours represent attractive targets for antibody-based therapy directed against R-spondin proteins.
  • Therapeutic strategies targeting downstream components of the Wnt signalling cascade may be effective against RSPO fusion-positive tumours.
  • IGF2 amplification/overexpression subset may benefit from IGF pathway-targeted therapies.
  • ATM mutations identify a subset potentially sensitive to DNA damage response-targeted agents.

Limitations & open questions

  • Relatively small cohort (74 pairs); limited statistical power for MSI subgroup analysis (no gene achieved significance in MSI alone).
  • Functional validation of RSPO fusions limited to in vitro Wnt reporter assays; in vivo therapeutic efficacy not demonstrated.
  • The clinical response of RSPO fusion-positive patients to Wnt pathway inhibitors or anti-RSPO antibodies remains untested.
  • Whether RSPO fusions confer differential prognosis compared to APC-mutant tumours is unknown.
  • Two hypermutated MSS samples were excluded from background mutation-rate calculations; their biology is unexplored.

Citations from this paper used in the wiki

  • “The RSPO2 and RSPO3 fusions were mutually exclusive and occurred in tumours that did not contain APC mutations or copy loss (Fig. 4d; P = 0.038, Fisher’s exact test)”
  • “R-spondin gene fusions involving RSPO2 and RSPO3 that together occur in 10% of colon tumours”
  • “We applied a previously described Q score metric to rank significantly mutated cancer genes… In MSS samples, we identified 23 significantly mutated cancer genes”
  • “ERBB3, which was mutated in 8% (6 out of 72) of the samples, contained multiple hotspot mutations that were oncogenic”
  • “IGF2 upregulation in 12% (8 out of 68) of the samples… Most (7 out of 8) of the tumours with IGF2 overexpression also showed focal amplification of the IGF2 locus”

This page was processed by paper-compiler on 2026-05-06.