The spectrum of SWI/SNF mutations, ubiquitous in human cancers

Abstract

SWI/SNF is a multi-subunit chromatin remodeling complex that uses the energy of ATP hydrolysis to reposition nucleosomes, thereby modulating gene expression. Accumulating evidence suggests that SWI/SNF functions as a tumor suppressor in some cancers. However, the spectrum of SWI/SNF mutations across human cancers has not been systematically investigated. Here, we mined whole-exome sequencing data from 24 published studies representing 669 cases from 18 neoplastic diagnoses. SWI/SNF mutations were widespread across diverse human cancers, with an excess of deleterious mutations, and an overall frequency approaching TP53 mutation. Mutations occurred most commonly in the SMARCA4 enzymatic subunit, and in subunits thought to confer functional specificity (ARID1A, ARID1B, PBRM1, and ARID2). SWI/SNF mutations were not mutually-exclusive of other mutated cancer genes, including TP53 and EZH2 (both previously linked to SWI/SNF). Our findings implicate SWI/SNF as an important but under-recognized tumor suppressor in diverse human cancers, and provide a key resource to guide future investigations.

Figure 1. SWI/SNF mutations are deleterious and widespread across human cancers.

A. Bar graph depicts the frequency of nonsynonymous mutations in SWI/SNF (right; counting mutations in any of 20 subunit genes) and TP53 (left) for each of the 18 tumor diagnoses surveyed. The average frequency of the 18 tumor diagnoses is indicated in red. The small number of samples with mutations in two different SWI/SNF subunits was not double-counted. B. The frequency distribution by mutation class is indicated for SWI/SNF subunit genes (right) and for all exome-sequenced genes (left). Note, the class distribution of SWI/SNF mutations is significantly skewed towards deleterious mutations (P = 1.0×10⁻¹⁸, chi-square test). Refer to Methods for a detailed description of these data.

Figure 2. Some SWI/SNF subunits are preferentially mutated.

A. The average frequency of nonsynonymous SWI/SNF subunit mutations (for the 18 tumor diagnoses analyzed) is indicated superimposed on a schematic depiction of the SWI/SNF complex. Mutations preferentially hit the SMARCA4 enzymatic subunit and several targeting subunits (ARID1A, ARID1B, PBRM1, and ARID2). B. Heatmap (color scale indicated) depicting the number of nonsynonymous mutations found in each SWI/SNF subunit gene from the exome datasets analyzed. Note that some tumor types show selective mutation of single SWI/SNF subunits, e.g. ARID1A in ovarian clear cell carcinoma (CCC) and gastric cancer, while most other tumor types do not. For simplification, only those SWI/SNF subunits and tumor types having mutations are shown.

Figure 3. Co-occurrence of mutated SWI/SNF subunits.

Heatmaps depict the mutation status of each SWI/SNF subunit gene in each tumor sample, shown for the seven tumor types with the highest frequency of SWI/SNF mutations. Rows and columns represent tumor samples and SWI/SNF subunit genes, respectively. Blue indicates the presence of a nonsynonymous mutation. Samples with mutations in two different SWI/SNF subunits are identified by a red arrow. TP53 mutations are also indicated, as are EZH2 activating mutations for the DLBCL study (lower left panel).

Figure 4. SWI/SNF mutations are not mutually exclusive of mutations in other commonly mutated genes.

For each panel, rows correspond to tumor samples and columns correspond to genes. Within the matrices, blue corresponds to a nonsynonymous mutation while grey corresponds to no reported mutation. The rows are ordered first based on the SWI/SNF mutational status and second on the cancer subtype (annotated in alternating black and brown text, left). A. The mutational status of the 189 most-highly mutated genes across the exome studies, in relation to SWI/SNF mutational status. The 189 genes are rank-ordered from left to right, from those most mutationally-inclusive to those most mutationally-exclusive with SWI/SNF mutations. B. Zoomed-in view of the mutational status of the four most-exclusive gene mutations (FAT2, NEB, CSMD1, SF3B1); none reach statistical significance. Additional discussion is provided in Text S1. C. Zoomed-in view of the mutational status of select cancer genes. These genes are denoted by an asterisk in panel A. Additional discussion is provided in Text S1.