e promoter, one of their targets, H3K18 is not acetylated when iAs is present. iAs Disrupts CARM1 but not GRIP1 Interaction with the MMTV promoter CARM1 3006665 is a protein methyltransferase that specifically targets H3R17 for methylation and acts synergistically with p160 coactivators to enhance transcription from steroid-regulated promoters including ER and GR. Because H3R17 was less methylated in response to iAs, CARM1 could either be displaced from the MMTV promoter or be at the promoter but not be enzymatically active. To distinguish between these possibilities, cells were treated with Dex6iAs and ChIP analysis was done with an antibody to CARM1. By 30 min of treatment with Dex alone CARM1 was at the promoter but was not when iAs was present. Western blot analysis of NEs confirmed that CARM1 was available for binding and not MMAE degraded after iAs treatment. Thus, iAs inhibits CARM1 interaction at the GR-activated MMTV promoter which can account for the lack of H3R17me. ChIP assays also showed that CARM1 interaction with the SGK promoter was inhibited by iAs, similarly to the MMTV promoter. The p160 coactivator GRIP1 interacts with GR and CARM1 interacts with the promoter by binding to the C-terminal domain of GRIP1. To determine if iAs affects GRIP1 interaction with GR, cells were treated with Dex6iAs and ChIP assays were done 22441874 with antibody to GRIP1. There was no difference in the amount of GRIP1 at NucB with iAS treatment. These data suggest that iAs affects the CARM1/GRIP1 interaction but not the GRIP1/GR interaction at the times tested. A sequential ChIP experiment was done to test whether the CARM1/GRIP1 interaction was disrupted by iAs. Cells were treated with 5 nM Dex68 mM iAs for 30 min and ChIP assays were done with antibody to GRIP1 followed by antibody to CARM1. In the first step, GRIP1 was found at NucB with Dex6iAs as seen in Fig. 5C. The GRIP1 immunoprecipitated material was then incubated with antibody to CARM1 to determine whether CARM1 was associated with GRIP1 at the promoter. The CARM1/GRIP1 interaction was intact in the presence of Dex alone but was not in the presence of Dex + iAs. In vitro pull-down assays further tested CARM1 interaction with the MMTV promoter in response to Dex + iAs. The assay utilizes a short DNA fragment of the MMTV promoter assembled into a regularly spaced nucleosomal array coupled to a magnetic bead. The in vitro MMTV template was incubated with NE isolated from cells treated with Dex6iAs at concentrations that inhibit transcription in these cells. CARM1 bound to the promoter with Dex alone, but even at the lowest concentration of iAs tested the amount of CARM1 bound did not exceed that in NEs from untreated cells . CARM1 seen in the untreated cells can be attributed to the low level of background activation by GR resulting from a small amount of GR in the nucleus before treatment. NEs used in the pull-down reactions show an approximately equal amount of CARM1 was present in all of the NEs. These in vitro data confirm results from the ChIP analysis and support the conclusion that CARM1 interaction with the MMTV promoter is inhibited by iAs. To determine whether the inhibition of CARM1 interaction 
with NucB was a direct or an indirect effect of iAs on CARM1, cells were treated with Dex alone and NE from the cells was incubated with the in vitro MMTV DNA. iAs was added directly to the in vitro reactions at concentrations equivalent to those found in the nucleus after treatment of cells with 5 nM De