Swapping function of two chromatin remodeling complexes

SWI/SNF- and ISWI-based complexes have distinct yet overlapping chromatin-remodeling activities in vitro and perform different roles in vivo. This leads to the hypothesis that the distinct remodeling functions of these complexes are specifically required for distinct biological tasks. By creating and characterizing chimeric proteins of BRG1 and SNF2h, the motor proteins of human SWI/SNF- and ISWI-based complexes, respectively, we found that a region that includes the ATPase domain specifies the outcome of the remodeling reaction in vitro. A chimeric protein based on BRG1 but containing the SNF2h ATPase domain formed an intact SWI/SNF complex that remodeled like SNF2h. This altered-function complex was active for remodeling and could stimulate expression from some, but not all, SWI/SNF responsive promoters in vivo. Thus, we were able to separate domains of BRG1 responsible for function from those responsible for SWI/SNF complex formation and demonstrate that remodeling functions are not interchangeable in vivo.     

Distinct strategies to make nucleosomal DNA accessible

One hallmark of ATP-dependent remodeling complexes is the ability to make nucleosomal DNA accessible to regulatory factors. We have compared two prominent human ATP-dependent remodelers, BRG1 from the SWI/SNF family and SNF2h from the ISWI family, for their abilities to make a spectrum of nucleosomal sites accessible. By measuring rates of remodeling at seven different sites on a mononucleosome and at six different sites on the central nucleosome of a trinucleosome, we have found that BRG1 opens centrally located sites more than an order of magnitude better than SNF2h. We provide evidence that this capability of BRG1 is caused by its ability to create DNA loops on the surface of a nucleosome, even when that nucleosome is constrained by adjacent nucleosomes. This specialized ability to make central sites accessible should allow SWI/SNF family complexes to facilitate binding of nuclear factors in chromatin environments where adjacent nucleosomes might otherwise constrain mobility.     

The SWI/SNF Subunit/Tumor Suppressor BAF47/INI1 Is Essential in Cell Cycle Arrest upon Skeletal Muscle Terminal Differentiation

Myogenic terminal differentiation is a well-orchestrated process starting with permanent cell cycle exit followed by muscle-specific genetic program activation. Individual SWI/SNF components have been involved in muscle differentiation. Here, we show that the master myogenic differentiation factor MyoD interacts with more than one SWI/SNF subunit, including the catalytic subunit BRG1, BAF53a and the tumor suppressor BAF47/INI1. Downregulation of each of these SWI/SNF subunits inhibits skeletal muscle terminal differentiation but, interestingly, at different differentiation steps and extents. BAF53a downregulation inhibits myotube formation but not the expression of early muscle-specific genes. BRG1 or BAF47 downregulation disrupt both proliferation and differentiation genetic programs expression. Interestingly, BRG1 and BAF47 are part of the SWI/SNF remodeling complex as well as the N-CoR-1 repressor complex in proliferating myoblasts. However, our data show that, upon myogenic differentiation, BAF47 shifts in favor of N-CoR-1 complex. Finally, BRG1 and BAF47 are well-known tumor suppressors but, strikingly, only BAF47 seems essential in the myoblasts irreversible cell cycle exit. Together, our data unravel differential roles for SWI/SNF subunits in muscle differentiation, with BAF47 playing a dual role both in the permanent cell cycle exit and in the regulation of muscle-specific genes.     

MyoD can induce cell cycle arrest but not muscle differentiation in the presence of dominant negative SWI/SNF chromatin remodeling enzymes

Cell cycle arrest is critical for muscle differentiation, and the two processes are closely coordinated but temporally separable. SWI/SNF complexes are ATP-dependent chromatin-remodeling enzymes that have been shown to be required for muscle differentiation in cell culture and have also been reported to be required for Rb-mediated cell cycle arrest. We therefore looked more closely at how SWI/SNF enzymes affect the events that occur during MyoD-induced myogenesis, namely, cell cycle regulation and muscle-specific gene expression, in cells that inducibly express dominant negative versions of Brahma (BRM) and Brahma-related gene 1 (BRG1), the ATPase subunits of two distinct SWI/SNF complexes. Although dominant negative BRM and BRG1 inhibited expression of every muscle-specific regulator and structural gene assayed, there was no effect on MyoD-induced activation of cell cycle regulatory proteins, and thus, cells arrested normally. In particular, in the presence or absence of dominant negative BRM or BRG1, MyoD was able to activate expression of p21, cyclin D3, and Rb, all of which are critical for cell cycle withdrawal in the G1/G0 phase of the cell cycle. These findings suggest that at least one basis for the distinct mechanisms that regulate cessation of cell proliferation and muscle-specific gene expression during muscle differentiation is that SWI/SNF-mediated chromatin-remodeling enzymes are required only for the latter.     

A cooperative activation loop among SWI/SNF, γ‐H2AX and H3 acetylation for DNA double‐strand break repair

Although recent studies highlight the importance of histone modifications and ATP‐dependent chromatin remodelling in DNA double‐strand break (DSB) repair, how these mechanisms cooperate has remained largely unexplored. Here, we show that the SWI/SNF chromatin remodelling complex, earlier known to facilitate the phosphorylation of histone H2AX at Ser‐139 (S139ph) after DNA damage, binds to γ‐H2AX (the phosphorylated form of H2AX)‐containing nucleosomes in S139ph‐dependent manner. Unexpectedly, BRG1, the catalytic subunit of SWI/SNF, binds to γ‐H2AX nucleosomes by interacting with acetylated H3, not with S139ph itself, through its bromodomain. Blocking the BRG1 interaction with γ‐H2AX nucleosomes either by deletion or overexpression of the BRG1 bromodomain leads to defect of S139ph and DSB repair. H3 acetylation is required for the binding of BRG1 to γ‐H2AX nucleosomes. S139ph stimulates the H3 acetylation on γ‐H2AX nucleosomes, and the histone acetyltransferase Gcn5 is responsible for this novel crosstalk. The H3 acetylation on γ‐H2AX nucleosomes is induced by DNA damage. These results collectively suggest that SWI/SNF, γ‐H2AX and H3 acetylation cooperatively act in a feedback activation loop to facilitate DSB repair.