Vanderbilt Institute of Chemical Biology



Discovery at the VICB






A New Approach to a Deadly Cancer 


By: Carol A. Rouzer, VICB Communications
Published: JUne 12, 2019


Studies reveal that inhibition of the MYC transcription factor could be a powerful new therapy for malignant rhabdoid tumor and other cancers.


SWI/SNF is a multicomponent protein complex that plays an important role in chromatin remodeling (Figure 1). The various proteins that form this complex are frequently mutated in cancer. In fact, approximately 20% of human cancers bear a mutation in one or more SWI/SNF proteins. In most cases, the mutation leads to a loss of function in the protein, suggesting that SWI/SNF serves primarily as a tumor suppressor. Whether this tumor suppressor activity is always related to chromatin remodeling activity, however, is not always clear. Case in point: SNF5, an SWI/SNF component that is inactivated in a number of cancers, including malignant rhabdoid tumor (MRT). MRT is a particularly aggressive and almost uniformly fatal childhood cancer, and the fact that SNF5 is frequently the only mutation found in MRT cells isolated from patients indicates its importance in disease pathogenesis. However, the fact that SNF5 binds to the well-known oncoprotein c-MYC in addition to playing a role in the SWI/SNF complex leads one to question whether its role in MRT is due to an alteration in the function of MYC, SWI/SNF, or both. To answer this question, Vanderbilt Institute of Chemical Biology member William Tansey and his laboratory investigated. Their research reveals that inhibition of MYC function is an important mechanism by which SNF5 acts as a tumor suppressor, at least in the case of MRT [A. M. Weissmiller, et al. (2019) Nat. Commun., 10, 2014].



FIGURE 1. (a) Diagrammatic representation of the composition of two different SWI/SNF complexes. (b) Mechanism of chromatin remodeling by SWI/SNF complexes. Figure reproduced by permission from Springer Nature from B.G. Wilson and C.W.M. Roberts (2011) Nat. Rev. Cancer, 11, 481. Copyright 2011.




Through its C-terminal basic helix-loop-helix leucine zipper (bHLHZip) domain, MYC binds with a second protein, MAX, to form a heterodimeric transcription factor. MYC:MAX then binds to E-box sequences (CACGTG) in the proximal promoter regions of target genes, thereby stimulating transcription. SNF5 can also interact with the bHLHZip domain of MYC, but it does so without interfering with MYC:MAX heterodimer formation. Instead, it blocks the ability of MYC:MAX to bind to DNA. The researchers illustrated this in vitro using a DNA probe containing an E-box sequence in an electrophoretic mobility shift assay (EMSA). They demonstrated that increasing amounts of SNF-5 displaced MYC:MAX dimers from binding to the probe in a concentration-dependent manner. This did not occur if they used a mutant SNF5 that lacked the MYC-binding domain.

To see if SNF5 blocked MYC:MAX binding to DNA in intact cells, the researchers used CRISPR-Cas technology to knock out the gene encoding SNF5 (SMARCB1) in HEK293 cells. They then transduced the cells with an engineered gene encoding SNF5 fused with FKBP12F36V. The fusion protein retained the functionality of native SNF5, but was subject to degradation upon addition of a small molecule, dTAG-47 (Figure 2). This approach enabled them to control the levels of SNF5 in the cells by addition of dTAG-47. Using chromatin immunoprecipitation (ChIP) with an antibody directed against MYC, they discovered that chromatin from cells expressing high levels of the SNF5 fusion protein contained lower amounts of bound MYC than chromatin from dTAG-47-treated cells that expressed low levels of the fusion protein. Thus, they confirmed that SNF5 suppresses MYC binding to DNA in intact cells.



FIGURE 2. (Left) Mechanism of dTAG-mediated protein degradation. The target protein is expressed as a fusion protein with FKBP12F36V, a mutated form of the enzyme peptidyl-prolyl cis-trans isomerase that binds dTAG ligands tightly. Binding of dTAG to this protein results in its association with the E3 ubiquitin ligase cereblon (CBRN). As a result, the target protein is ubiquitinated, leading to its degradation by the proteasome. (Right) Structures of some dTAG ligands. Figure reproduced by permission from Springer Nature from B. Nabet, et al. (2018) Nat. Chem. Biol, 14, 431. Copyright 2018.



These findings led the investigators to hypothesize that loss of SNF5-mediated suppression of MYC's binding to DNA contributes to the malignant phenotype of MRT cells. To test this hypothesis, they first used shRNA to knock down expression of MYC in MRT cells in culture. Consistently, they found that a reduction in MYC levels resulted in a loss of viability and anchorage-independent growth of the cells. To further test the hypothesis, they expressed either enhanced green fluorescent protein (EGFP, a control protein), SNF5, or OmoMYC, a dominant-negative protein that blocks association of MYC with target genes in MRT cells. In cells expressing SNF5, they confirmed incorporation of the protein into SWI/SNF complexes. SNF5 expression had no effect on steady state levels of MYC expression. They also confirmed that expression of OmoMYC reduced the interaction of MYC with chromatin. Having established that both SNF5 and OmoMYC functioned as expected in the cells, the researchers used chromatin immunoprecipitation followed by next-generation sequencing (ChIP-seq) to investigate the effects of their expression on MYC binding to DNA. They identified ~900 sites of MYC binding in the control cells. This number was lower than seen in other cancer cells, but they were able to confirm that the binding occurred at E-box motifs in the proximal promoters of MYC-controlled genes. Thus, the binding sites were confirmed to be valid MYC targets. Expression of either SNF5 or OmoMYC reduced MYC binding at these sites across the genome.

Although these results supported the hypothesis that SNF5 was acting by directly interfering with MYC:MAX binding to target DNA, the researchers could not rule out the possibility that the effects they were seeing were the result of chromatin remodeling via SWI/SNF complexes that had formed as a result of SNF5 expression. To address this question, they used an assay for transposase-accessible chromatin followed by next generation sequencing (ATAC-seq), which identifies open areas in the genome. They found ~25,000 accessible sites in control cells, and expression of OmoMYC had no effect on chromatin structure. In contrast, expression of SNF5 resulted in the addition of ~2,500 new accessible sites and the closure of 7 sites. These findings confirmed that expression of SNF5 led to chromatin remodeling. However, the new sites were distal to the transcription start site rather than proximal to the site as in the case of MYC:MAX binding sites. The genes affected were enriched for signal transduction, development, and differentiation, totally distinct from MYC-regulated genes. Thus, they concluded that the effects of SNF5 on MYC binding to DNA were not the result of SWI/SNF-mediated chromatin remodeling.

The initiation of DNA transcription is an extremely complex process that requires the coordinated action of multiple individual proteins and protein complexes. Among these is TFIIH, a transcription factor that is recruited to sites of initiation along with RNA polymerase II (Pol II). TFIIH works in conjunction with two additional proteins NELF (negative elongation factor) and DSIF (DRB-sensitivity factor) to stop Pol II progression at the junction between the promotor and the gene body. This promoter-proximal pause helps to coordinate 7-methyl-guanine capping of the 5´-end of the nascent mRNA with further elongation. One function of MYC:MAX binding is to release Pol II from the pause, enabling completion of gene transcription. This led the researchers to hypothesize that blockade of MYC function with SNF5 or OmoMYC should lead to an increase in the presence of paused Pol II in the genome. To test this hypothesis, the researchers turned to precision global run-on transcription coupled with deep sequencing (PRO-seq), which enabled them to measure the density of Pol II bound at the promoter-proximal region versus the body of the gene. The ratio between the two quantities serves as a "pausing index". The investigators discovered that OmoMYC increased the pausing index at ~4,500 genes, while decreasing the index at ~2,000 genes. The corresponding numbers for SNF5 were ~3,500 genes, and ~3,400 genes, respectively. There was strong similarity in the genes newly paused by OmoMYC and SNF5, with 70% shared by both. Approximately 35% of genes that lost a pause were shared between cells expressing OmoMYC and those expressing SNF5. Of greater interest was the finding that among MYC target genes, expression of OmoMYC or SNF5 resulted in a gain of pause only, and 80% of these newly paused genes were identical for the two proteins (Figure 3). Through comparison of these results with previously published data on the effects of SNF5 expression on mRNA levels in MRT cells, the researchers found that 40% of genes that gained a pause in their studies also exhibited lower mRNA levels after 7 days in culture. Thus, the SNF5-mediated increase in the pause of Pol II at the transcription initiation site correlated with reduced expression of a significant proportion of genes.




FIGURE 3. Venn diagrams illustrating the degree of overlap among genes exhibiting a gain of pause (left) or loss of pause (center) as a result of OmoMYC or SNF5 expression in MRT cells. Diagram illustrating the degree of overlap among MYC target genes exhibiting a gain of pause as a result of OmoMYC or SNF5 expression (right). Figure reproduced under the Creative Commons Attribution 4.0 International License 4.0 from A. M. Weissmiller, et al. Nat. Comm., (2019) 10, 2014.



In summary, the data demonstrate clearly that SNF5 carries out two functions in MRT cells. One function is to serve as a subunit in the SWI/SNF complex that promotes chromatin remodeling, leading to differentiation and development. The other function is to suppress the expression of MYC-regulated genes that promote cell proliferation (Figure 4). Although both of these functions play a role in tumor suppression, these studies show that expression of SNF5 in MRT cells results in a phenotype that is very similar to that of MYC inhibition. This finding suggests that one way to treat MRT or other cancers in which SNF5 mutation plays a role is through MYC blockade. We currently do not have a MYC inhibitor available for clinical use, but this is a focus of current active research.



FIGURE 4. Proposed mechanism by which SNF5 acts as a tumor suppressor. (a) In the absence of SNF5, MYC/MAX heterodimers promote transcription of genes that support cell proliferation, and the loss of SWI/SNF complexes leads to heterochromatin structure that prevents development and differentiation. (b) In the presence of SNF5, MYC/MAX heterodimers do not bind to target gene promoters, leading to a reduction in transcription of those genes. SWI/SNF complexes can form, opening heterochromatin to enable transcription of genes that promote development and differentiation. Figure reproduced under the Creative Commons Attribution 4.0 International License 4.0 from A. M. Weissmiller, et al. Nat. Comm., (2019) 10, 2014.




View Nature Communications article: Inhibition of MYC by the SMARCB1 tumor suppressor







The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2019 by Vanderbilt University