Recurrent SMARCA4 mutations in small cell carcinoma of the ovary

Authors

Petar Jelinic

Jennifer J Mueller

Narciso Olvera

Fanny Dao

Sasinya N Scott

Ronak Shah

Jian Jiong Gao

Nikolaus Schultz

Mithat Gonen

Robert A Soslow

Michael F Berger

Douglas A Levine

Doi

10.1038/ng.2922

PMID: 24658004 · DOI: 10.1038/ng.2922 · Journal: Nature Genetics (2014)

TL;DR

Jelinic et al. performed targeted sequencing of 279 cancer-related genes (MSK-IMPACT) on 12 paired tumor/normal samples from small cell carcinoma of the ovary, hypercalcemic type (SCCOHT) — a rare, aggressive ovarian malignancy of young women. They identified inactivating biallelic SMARCA4 mutations (nonsense, frameshift, splice-site, and intragenic deletions) in 100% of the 12 tumors, with only 4 additional non-recurrent somatic mutations across all other 278 genes combined. Immunohistochemistry and immunoblotting confirmed loss of SMARCA4 (BRG1) protein expression in 7 of 9 evaluable cases, establishing SMARCA4 as a defining tumor suppressor in SCCOHT and implicating the SWI/SNF chromatin-remodeling complex as the central driver of this disease.

Cohort & data

  • 12 paired tumor–normal SCCOHT samples (formalin-fixed, paraffin-embedded; from MSKCC 1998–2012 plus collaborating centers), cancer type SCCO (small cell carcinoma of the ovary, hypercalcemic type).
  • Dataset: scco_mskcc (Memorial Sloan-Kettering Cancer Center SCCOHT cohort).
  • Assay: targeted hybrid-capture sequencing of 279 cancer-associated genes using the msk-impact-panel (Integrated Mutation Profiling of Actionable Cancer Targets); Illumina HiSeq 2000, 2 × 75-bp reads; mean depth 442×; ≥100× depth at 97% of targeted exons in tumors.
  • Validation: sanger-sequencing of genomic DNA and cDNA (RT-PCR) for all 12 cases; immunohistochemistry (anti-SMARCA4) on 9 cases with available tissue; immunoblotting on frozen tumor samples (cases 101, 102).
  • Comparator: 4,784 non-hypermutated TCGA tumors, with mutation frequencies cross-referenced via the cBioPortal for Cancer Genomics ([PMID:22588877] / [PMID:23550210]).

Key findings

  • Inactivating biallelic SMARCA4 mutations were detected in 100% (12/12) of SCCOHT tumors. Probability of finding a single gene mutated in all 12 samples by chance across 279 genes: P < 2.22 × 10⁻¹⁶ (Bernoulli model with p = 0.015 from TCGA background).
  • Mutation classes spanned nonsense, frameshift, splice-site, and homozygous intragenic deletions of exons; no missense SMARCA4 mutations were identified in SCCOHT (contrast: 84% of 128 TCGA SMARCA4 mutations were missense).
  • 4 cases harbored two inactivating mutations each; the remaining 8 cases harbored a single inactivating mutation with loss of heterozygosity at the SMARCA4 locus (mutant allele frequency ≥0.75).
  • Only 4 additional non-recurrent somatic mutations were found in any of the other 278 sequenced genes across all 12 samples — strikingly low compared to a mean of 4.3 mutations per tumor (s.d. 4.4) in TCGA non-hypermutated cases sequenced over the same gene set.
  • Immunohistochemistry showed loss of SMARCA4 nuclear staining in all 4 cases with available tissue and nonsense mutations; splice-site mutation cases 101 and 102 also showed clear loss of protein on immunoblot and IHC; case 112 showed equivocal loss; case 103 (in-frame deletion of exons 25–26, removing 102 amino acids of the helicase domain) retained protein expression but is predicted to encode a catalytically dead helicase.
  • Across all TCGA tumors, SMARCA4 mutations were detected in 3% of 4,787 non-hypermutated samples, with mutation frequencies of 5–8% in bladder carcinoma, stomach adenocarcinoma, lung adenocarcinoma, and lower grade glioma. Inactivating SMARCA4 mutations were most common in lung adenocarcinoma and were associated with poor outcome.
  • Co-occurrence of inactivating SMARCA2 and SMARCA4 mutations was essentially absent in TCGA (only 1 case with SMARCA4 inactivation plus SMARCA2 missense), consistent with mutual exclusivity of the BAF complex catalytic subunits.
  • Functional studies: ectopic re-expression of SMARCA4 in SMARCA4-null H1299 lung adenocarcinoma cells caused dose-dependent growth suppression and induced p21; shRNA knockdown of SMARCA4 in 293T cells increased proliferation (XTT assay).
  • One of the 12 cases (case 111) carried a heterozygous germline SMARCA4 nonsense mutation with somatic loss of the wild-type allele, supporting a hereditary component to SCCOHT.

Genes & alterations

  • SMARCA4 (BRG1) — biallelic inactivating mutations in 12/12 SCCOHT tumors (nonsense, frameshift, splice-site, intragenic exonic deletions). Most mutations affected the helicase catalytic domains. The paper establishes SMARCA4 as the defining tumor suppressor of SCCOHT; truncating mutations correlate with reduced SMARCA4 expression in TCGA, consistent with nonsense-mediated decay.
  • SMARCA2 (BRM) — discussed as the mutually exclusive ATP-dependent BAF subunit. Co-occurrence with SMARCA4 inactivation is essentially absent across TCGA. The paper highlights prior evidence (Oike et al., 2013) that SMARCA2 silencing suppresses growth of SMARCA4-deficient lung cancer cells, suggesting a synthetic-lethal vulnerability potentially exploitable in SCCOHT.
  • ARID1A (BAF250A) and ARID1B (BAF250B) — non-catalytic BAF subunits invoked to contextualize SWI/SNF biology; not mutated in this cohort but discussed as relevant to ovarian tumorigenesis (citing prior ARID1A work).
  • PBRM1 (BAF180) — PBAF-specific subunit referenced in the discussion of complex composition; not mutated in this cohort.
  • TP53 — referenced via citation 15 (Guan et al., 2011) suggesting SMARCA4 loss may be sufficient for transformation through p53-dependent BAF mechanisms.

Clinical implications

  • Diagnostic biomarker: Loss of SMARCA4 protein expression by immunohistochemistry is proposed as a useful diagnostic adjunct for SCCOHT, especially given the challenges of distinguishing it from histological mimics (small cell neuroendocrine carcinoma, juvenile granulosa cell tumor, Sertoli-Leydig cell tumor, DSRCT, metastatic melanoma, lymphoma, rhabdomyosarcoma).
  • Hereditary risk: Identification of a germline SMARCA4 nonsense mutation in case 111 supports germline testing in SCCOHT patients and families.
  • Therapeutic hypothesis: The authors propose inhibitors of the SMARCA2 ATPase as a synthetic-lethal strategy for SMARCA4-deficient tumors, including SCCOHT — leveraging mutual exclusivity of BAF catalytic subunits. They note SCCOHT is otherwise refractory to standard adjuvant therapy (long-term survival 33% even with disease confined to the ovary).
  • Cross-tumor relevance: Inactivating SMARCA4 mutations in LUAD are associated with poor outcome, suggesting the same synthetic-lethal logic may apply beyond SCCOHT.

Limitations & open questions

  • Small sample size (n = 12) due to the rarity of SCCOHT; the 100% mutation rate is statistically robust (P < 2.22 × 10⁻¹⁶) but may not capture rarer molecular subsets.
  • Targeted panel only: the 279-gene IMPACT panel cannot rule out recurrent driver mutations in genes outside the panel; whole-exome or whole-genome characterization would refine the mutational landscape.
  • Tissue of origin remains speculative for SCCOHT; SMARCA4 loss alone may not explain the unique clinical presentation (young age, hypercalcemia).
  • Functional validation of synthetic lethality between SMARCA2 and SMARCA4 specifically in SCCOHT models is not performed in this paper — the hypothesis is extrapolated from prior work in lung cancer.
  • No clinical correlation: HIPAA restrictions limited collected clinical data to age and year of diagnosis; outcome correlations with specific mutation classes are not possible.
  • Case 103 (in-frame exonic deletion with retained protein) illustrates that SMARCA4 IHC alone can miss functionally inactivating alterations — sequencing is required for complete ascertainment.

Citations from this paper used in the wiki

  • “Sequencing of all protein-coding exons in 279 cancer-related genes for 12 paired tumor and normal SCCOHT samples identified inactivating biallelic SMARCA4 mutations in each case.” (p. 2)
  • “The probability of identifying SMARCA4 mutations in all 12 samples is less than 2.22 × 10⁻¹⁶.” (p. 2)
  • “Only 4 additional non-recurrent somatic mutations were identified in any of the other 278 genes sequenced across all 12 samples.” (p. 2)
  • “Across all tumors characterized by TCGA thus far, SMARCA4 mutations have been detected in 3% of the 4,787 non-hypermutated samples. Mutation frequencies of 5–8% are present in bladder carcinoma, stomach adenocarcinoma, lung adenocarcinoma and lower grade glioma.” (p. 4)
  • “Inhibitors of the SMARCA2 ATPase may be an effective approach for the treatment of SMARCA4-deficient tumors.” (p. 5)
  • “Mutation frequencies across TCGA tumor types were collated from data contained within the cBioPortal for Cancer Genomics.” (Methods, p. 7)

This page was processed by crosslinker on 2026-05-11.