Comprehensive genomic analysis identifies SOX2 as a frequently amplified gene in small-cell lung cancer

Authors

Rudin CM

Durinck S

Stawiski EW

Poirier JT

Modrusan Z

Shames DS

Bergbower EA

Guan Y

Shin J

Guillory J

Rivers CS

Foo CK

Bhatt D

Stinson J

Gnad F

Haverty PM

Gentleman R

Chaudhuri S

Janakiraman V

Jaiswal BS

Parikh C

Yuan W

Zhang Z

Koeppen H

Wu TD

Stern HM

Yauch RL

Huffman KE

Paskulin DD

Illei PB

Varella-Garcia M

Gazdar AF

de Sauvage FJ

Bourgon R

Minna JD

Brock MV

Seshagiri S

Doi

10.1038/ng.2405

PMID: 22941189 · DOI: 10.1038/ng.2405 · Journal: Nature Genetics (2012)

TL;DR

Rudin et al. performed exome sequencing, whole-genome sequencing, RNA-seq, and copy-number analysis of ~80 small-cell lung cancer (SCLC) samples (36 primary tumor-normal pairs, 17 matched cell lines, plus additional unpaired samples). They identified 22 significantly mutated genes, discovered SOX2 amplification in ~27% of samples with functional dependence on SOX2 for proliferation, and identified a recurrent RLF-MYCL fusion transcript that drives cell growth.

Cohort & data

  • 80 human SCLC samples total: 36 primary tumor-normal pairs, 17 paired SCLC cell lines with matched lymphoblastoid lines, 4 unpaired primary tumors, and 23 unpaired cell lines.
  • 56 samples assayed by Illumina HumanOmni2.5 SNP arrays for copy-number analysis.
  • 110 independent primary SCLC tumor samples used for SOX2 IHC/FISH validation.
  • Cancer type: SCLC.
  • Dataset: sclc_jhu.
  • Methods: whole-exome-seq, whole-genome-seq, RNA-seq, GISTIC.
  • Reference genome: GRCh37/hg19.

Key findings

  • Exome sequencing of 42 tumor-normal pairs identified 26,406 somatic mutations, of which 7,977 were protein-altering; validation rate was 91% by Sequenom.
  • Mean nonsynonymous mutation rate was 5.5 mutations per megabase (excluding one hypermutated sample with 2,953 mutations).
  • G-to-T transversions were predominant, consistent with tobacco smoke carcinogenesis.
  • 22 significantly mutated genes were identified (q score >= 1, FDR <= 10%), including TP53, RB1, and several novel SCLC genes.
  • SOX2 amplification (copy number >= 4) detected in ~27% (15/56) of SCLC samples; SOX2 expression correlated with gene copy number and clinical stage.
  • shRNA knockdown of SOX2 in amplified cell lines (H446, H720) significantly reduced proliferation (P < 0.01).
  • Recurrent RLF-MYCL fusion found in 1 primary tumor and 4 cell lines; siRNA targeting MYCL reduced proliferation in fusion-positive lines.
  • Mutations clustered in PI3K pathway (PIK3CA, AKT1-3, MTOR), SOX family members (SOX3/4/5/6/9/11/14/17, mutually exclusive), Notch/Hedgehog (NOTCH1, NOTCH2, NOTCH3, SMO), and chromatin-modifying genes (EP300, MLL2, TRRAP).
  • Four kinase gene fusions identified: NPEPPS-EPHA6, SKP1-CDKL3, NEK4-SFMBT1, ZAK-RAPGEF4.

Genes & alterations

  • TP53 — inactivating mutations (75-90% of SCLC); significantly mutated in this cohort.
  • RB1 — inactivating mutations and copy-number loss; significantly mutated.
  • SOX2 — amplification in ~27% of samples; lineage-survival oncogene driving proliferation.
  • MYCL — recurrent RLF-MYCL fusion; functional oncogene when fused.
  • PIK3CA — rare activating mutations; hotspot mutations identified.
  • PTEN — inactivating mutations (2-4%); hotspot mutations and copy-number loss.
  • MYC — copy-number amplification (family members).
  • KIT — mutations including codon 761 (likely activating); copy-number gain.
  • NOTCH1 — mutations in Notch pathway members.
  • EP300 — hotspot mutations in chromatin-modifying gene.
  • CDKN2A — hotspot mutations identified.

Clinical implications

  • SOX2 amplification and overexpression identify a substantial SCLC subgroup (~27%) that may be targetable; SOX2 expression correlates with disease stage.
  • RLF-MYCL fusion represents a potential therapeutic target in fusion-positive SCLC.
  • Kinase gene fusions (NPEPPS-EPHA6 and others) may offer targeted therapy opportunities analogous to ALK/ROS1 fusions in NSCLC.
  • PI3K pathway mutations suggest potential susceptibility to PI3K/AKT/mTOR inhibitors.

Limitations & open questions

  • No matched clinical outcome data reported; prognostic significance of SOX2 amplification beyond stage correlation is unknown.
  • Functional validation limited to cell line models (shRNA/siRNA); in vivo validation not performed.
  • No direct therapeutic targeting of SOX2 demonstrated (SOX2 is a transcription factor, difficult to drug directly).
  • Unpaired samples (27/80) may have residual germline variants despite filtering.
  • The hypermutated sample was excluded from analysis, and its biology (possible mismatch repair deficiency) was not explored.

Citations from this paper used in the wiki

  • “we identified high levels of amplification (copy number of >=4) of SOX2 in ~27% (15/56) of the SCLC samples”
  • “Induction of SOX2 shRNA in both H446 and H720 resulted in lower amounts of SOX2 protein and reduced cell proliferation”
  • “A fusion involving RLF and MYCL1 was found in one primary SCLC tumor and four SCLC cell lines (H889, HCC33, H1092 and COR-L47)”
  • “We identified 22 significantly mutated genes in SCLC (q score >= 1; false discovery rate <= 10%)”
  • “the SCLC tumors had an average of 175 protein-altering single-nucleotide variants (range 31-388) with a mean nonsynonymous mutation rate of 5.5 mutations per megabase”

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