ATP-dependent chromatin remodeling complexes as novel targets for cancer therapy

Abstract

The progression to advanced stage cancer requires changes in many characteristics of a cell. These changes are usually initiated through spontaneous mutation. As a result of these mutations, gene expression is almost invariably altered allowing the cell to acquire tumor-promoting characteristics. These abnormal gene expression patterns are in part enabled by the posttranslational modification and remodeling of nucleosomes in chromatin. These chromatin modifications are established by a functionally diverse family of enzymes including histone and DNA-modifying complexes, histone deposition pathways, and chromatin remodeling complexes. Because the modifications these enzymes deposit are essential for maintaining tumor-promoting gene expression, they have recently attracted much interest as novel therapeutic targets. One class of enzyme that has not generated much interest is the chromatin remodeling complexes. In this review, we will present evidence from the literature that these enzymes have both causal and enabling roles in the transition to advanced stage cancers; as such, they should be seriously considered as high-value therapeutic targets. Previously published strategies for discovering small molecule regulators to these complexes are described. We close with thoughts on future research, the field should perform to further develop this potentially novel class of therapeutic target.

Keywords: Cancer; Chromatin; Chromatin remodeling; Epigenetics; Nucleosome.

Figure 5.1. Chromatin remodeling complexes are segregated into distinct families

The ATPase containing subunits found in chromatin remodeling complexes have common functional domains allowing their segregation into distinct families. All ATPase subunits have distinctive DEXDc ATP-binding domains and a HELICc helicase domain. The combination of these domains convert the chemical energy stored in ATP into mechanical energy used to remodel nucleosomes. The space between these two domains varies among family members and is distinctively long for the INO80 family of ATPases. Other domains which are diagnostic to specific families include the HSA domain (HSA) found in SWI/SNF and INO80 families, bromodomains (bromo) found in SWI/SNF family members, chromodomains (chromo) found in the CHD family, and the HAND–SANT–SLIDE (HSS) cluster found in ISWI family members (Becker & Workman, 2013).

Figure 5.2. Chromatin remodeling complexes vary subunit composition to increase diversity

The subunit composition of major characterized chromatin remodeling complexes from the SWI/SNF, INO80, CHD, and ISWI families are shown in cartoon. Complexes are organized into rows based on family. ATPase subunits are shown in red; all other subunits are shown in blue. Subunits which can be differentially incorporated into complexes are separated by a backslash symbol (/).

Figure 5.3. Heat map representation of somatic mutation rate for subunits of chromatin remodeling complexes in human cancers

The COSMIC database was queried using each of the subunits found in the chromatin remodeling complexes depicted in Fig. 5.2 (Forbes et al., 2011). The frequency of somatic mutation is expressed as a percentage of sequenced tumors from each tissue type. Percentages are not corrected for background mutation rates which vary with individual tumors and tumors from tissue types. The heat map range is shown on the lower right and varies from 0% to 10% of sequenced tumors in the COSMIC database.

Figure 5.4. Heat map representation of gene expression changes for subunits of chromatin remodeling complexes in human cancers

The ONCOMINE database was queried using each of the subunits found in the chromatin remodeling complexes depicted in Fig. 5.2 (Rhodes et al., 2004). Data is expressed as a percentage of unique analyses for which transcript levels of the subunit differ by greater than twofold between tumor and control normal tissue (P value≥1.0×10⁻⁴). Transcripts for which overexpression was observed in tumor tissue are colored in red, and those where an underexpression is observed are shown in green. The heat map range is shown on the lower right and varies from 0% to 40% of unique analyses found in the ONCOMINE database.