IPMK physically binds to the SWI/SNF complex and modulates BRG1 occupancy

Inositol polyphosphate multikinase (IPMK), a key enzyme in the inositol polyphosphate (IP) metabolism, is a pleiotropic signaling factor involved in major biological events including transcriptional control. In yeasts, IPMK and its IP products were known to promote the activity of SWI/SNF chromatin remodeling complex, which plays a critical role in gene expression by regulating chromatin accessibility. However, the direct linkage between IPMK and chromatin remodelers remains unclear, raising a question on how IPMK contributes to the transcriptional regulation in mammals. By employing unbiased screenings and in vivo/in vitro immunoprecipitations, here we demonstrated that IPMK physically associates with native mammalian SWI/SNF complexes by directly binding to SMARCB1, BRG1, and SMARCC1. Furthermore, we identified the specific domains required for the IPMK-SMARCB1 binding. Notably, using CUT&RUN and ATAC-seq assays, we discovered that IPMK co-localizes with BRG1 and regulates BRG1 localization as well as BRG1-mediated chromatin accessibility in a genome-wide manner (including promoter-TSS) in mouse embryonic stem cells. Finally, our mRNA-seq analyses revealed that IPMK and SMARCB1 regulate common gene sets, validating a functional link between IPMK and SWI/SNF complex. Together, these findings establish an importance of IPMK in promoter targeting of the SWI/SNF complex, thereby contributing to SWI/SNF-meditated chromatin accessibility and transcription.


21
control. In yeasts, IPMK and its IP products were known to promote the activity of SWI/SNF chromatin 22 remodeling complex, which plays a critical role in gene expression by regulating chromatin accessibility.

23
However, the direct linkage between IPMK and chromatin remodelers remains unclear, raising a question 24 on how IPMK contributes to the transcriptional regulation in mammals. By employing unbiased screenings 25 and in vivo/in vitro immunoprecipitations, here we demonstrated that IPMK physically associates with 26 native mammalian SWI/SNF complexes by directly binding to SMARCB1, BRG1, and SMARCC1.

27
Furthermore, we identified the specific domains required for the IPMK-SMARCB1 binding. Notably, using 28 CUT&RUN and ATAC-seq assays, we discovered that IPMK co-localizes with BRG1 and regulates BRG1 29 localization as well as BRG1-mediated chromatin accessibility in a genome-wide manner (including 30 promoter-TSS) in mouse embryonic stem cells. Finally, our mRNA-seq analyses revealed that IPMK and 31 SMARCB1 regulate common gene sets, validating a functional link between IPMK and SWI/SNF complex.

32
Together, these findings establish an importance of IPMK in promoter targeting of the SWI/SNF complex,

157
In addition, we observed that Ipmk knockdown did not affect the protein levels of SWI/SNF complex

222
Our results demonstrated that IPMK directly binds to the core subunits of mammalian SWI/SNF 223 complexes (SMARCB1, BRG1, and BAF155) and physically associates with native SWI/SNF complexes.

224
Regarding this, one can speculate that IPMK may play an important role in chromatin regulation. However, 225 the region where this IPMK-SWI/SNF interaction occurs in vivo and the detailed localization and the role 226 of IPMK in the chromatin remains elusive. To decipher these issues, we first conducted a chromatin 227 fractionation assay using mESCs and MEFs. Notably, we found that IPMK is evenly distributed in all three

234
Together, these results indicate that IPMK, BRG1, and SMARCB1 reside together in the chromatin.

235
We then sought to determine where this event (IPMK, BRG1, and SMARCB1 localizing at the 236 chromatin) takes place within the chromatin. To investigate the localization of BRG1 and IPMK within the 237 chromatin, we performed CUT&RUN (cleavage under targets and release using nuclease) assays (Skene & 238 Henikoff, 2017) in mESCs. In accordance with our results of the IPMK-SWI/SNF complex's physical 239 association, we found that IPMK was co-localized with BRG1 in a genome-wide manner ( Figure 4B and 240 C). Next, we performed peak annotation to analyze the genomic regions (e.g., promoters or intergenic 241 regions) enriched with CUT&RUN peaks. Given that BRG1, a catalytic subunit of the SWI/SNF complex, 242 is known to localize at the promoter-transcription start site (TSS) (de Dieuleveult et al., 2016), we 243 confirmed that BRG1 was significantly enriched at promoters of the mouse genome ( Figure 4D and E).

244
Notably, we detected that IPMK was also significantly enriched at promoters ( Figure 4D and E).

245
Collectively, these results strongly indicate that IPMK and BRG1 are co-localized at the chromatin, 246 particularly at the promoter region, which further support our previous results of physical association 247 between IPMK and SWI/SNF complex.

249
IPMK regulates the BRG1 occupancy and impacts the BRG1-mediated chromatin accessibility

250
To further elucidate the role of IPMK in BRG1 localization, we performed BRG1 CUT&RUN assays upon IpmkKD and compared them to EgfpKD (control) in mESCs. Interestingly, we observed a decreased 252 BRG1 occupancy upon IpmkKD at specific promoter-TSS regions with enriched IPMK ( Figure 4B and F).

253
Strikingly, we found that the genome-wide BRG1 localization was severely disrupted (decreased BRG1 254 occupancy) upon IpmkKD at BRG1 CUT&RUN peaks with low BRG1 enrichment in EgfpKD mESCs 255 (bottom half of the heatmaps, termed as Low) ( Figure 4C, G, and H). In addition, we detected that BRG1 256 CUT&RUN peaks' genomic distributions were unaffected by IpmkKD ( Figure 4D and E), suggesting that 257 IpmkKD do not affect the global distribution (changes in peak positions) of BRG1 but impacts the global 258 occupancy of BRG1.

259
It is previously known that BRG1 regulates chromatin accessibility at NFR (nucleosome free regions)

263
we observed that both ATAC-seq signals and BRG1 occupancy were reduced upon IpmkKD at promoter-264 TSS regions of Nanog ( Figure 4F). Furthermore, we observed some discrepancies in major peak positions 265 when comparing BRG1 CUT&RUN peaks (EgfpKD, termed as BRG1 peaks) and ATAC-seq peaks 266 (EgfpKD, termed as ATAC peaks) ( Figure 4F). To precisely assess the effect of IpmkKD-induced decreased 267 BRG1 occupancy on chromatin accessibility, we assigned ATAC peaks to the nearby (within 2kb) BRG1 268 peaks and selected these BRG1 peaks for further analysis (we excluded BRG1 peaks without nearby ATAC 269 peaks) ( Figure G). In addition, BRG1 peaks containing or assigned with multiple ATAC peaks were 270 included without deduplication to match the same ordering as ATAC peaks (the same ordering -alignment 271 of heatmaps' row -was applied for BRG1 and ATAC peaks in Figure 4G). As expected, we observed that

294
and Bivalent (high H3K4me3, high H3K27me3). Considering nucleosome intensity and ATAC-seq signals, 295 we detected that chromatin was highly accessible in H3K4me3-Only, moderately accessible in bivalent, 296 and inaccessible in H3K4me3-Low promoters ( Figure 5A, right). Next, we categorized the 9,042 TSS with 297 decreased BRG1 occupancy upon IpmkKD (1.5 fold changes in BRG1 occupancy compared to EgfpKD, 298 see also Figure 5D, left Total) into the three promoter types ( Figure 5B). Since our goal was to examine the 299 IpmkKD-induced 'decreased' BRG1 occupancy (changes from enriched to depleted BRG1 signal upon 300 IpmkKD as seen in Figure 4C and G-I), we excluded the H3K4me3-Low promoters, which exhibited the 301 extremely low BRG1 signals ( Figure 5A, top right). Furthermore, regarding the fact that TSS divided the 302 BRG1 intensity into two ( Figure 5A, top left), we defined two genomic regions (Upstream and Downstream) 303 that coincided with the two major BRG1 intensity (which also coincided with -1 and +1 nucleosomes), 304 respectively ( Figure 5A, top). We then categorized the 9,042 TSS with IpmkKD-induced decreased BRG1

312
Previously, it is reported that BRG1, localized at -1 nucleosome in wide NFR (median length 808 bp) 313 of H3K4me3-Only and bivalent promoters, positively regulates the chromatin accessibility of NFR, 314 whereas BRG1 that localized at +1 nucleosome in narrow NFR (median length 28 bp) of H3K4me3-Only 315 promoters tends to inhibit the chromatin accessibility of NFR in mESCs (de Dieuleveult et al., 2016). To 316 elucidate the effect of IpmkKD-induced decreased BRG1 occupancy on chromatin accessibility at two 317 promoters and five clusters, we analyzed two ATAC-seq data of ours (IpmkKD) and publically released 318 (Brg1KD, GSE64825). We calculated the differential ATAC-seq signals (KD vs. controls) at Center regions,

319
where ATAC-seq intensity is highly enriched ( Figure 5A). Consistent with our previous genome-wide 320 results ( Figure 4G-I), we found that ATAC-seq signals (chromatin accessibility) were significantly reduced 321 upon IpmkKD at two promoters ( Figure 5E) and most clusters ( Figure 5F), where these promoters/clusters 322 were defined by decreased BRG1 occupancy upon IpmkKD. Notably, we observed that ATAC-seq signals  H3K4me3-Only promoters, BRG1 levels are relatively similar at both Up/Downstream regions ( Figure 5D).

337
Interestingly, although we applied the same criteria when categorizing the five clusters, H3K4me3-Only 338 and bivalent promoters exhibited different BRG1 localizations in EgfpKD mESCs at Cluster2 and Cluster3,

339
indicating that these two promoters each possess distinct BRG1 localizations in mESCs. Together, these 340 results suggest that IPMK plays a pivotal role in maintaining the chromatin accessibility of bivalent 341 promoters, particularly by safeguarding the BRG1 occupancy at the -1 nucleosome.

342
By comparing with the BRG1 unchanged locus ( Figure 5G, left), we confirmed the close association 343 between reduced BRG1 level upon IpmkKD and decreased ATAC-seq signals upon IpmkKD and Brg1KD 344 at specific loci of H3K4me3-Only and bivalent promoters ( Figure 5G and H). However, at a specific locus,

372
Similarly, we observed that all three clusters exhibited more reduction at bivalent promoters compared to    Figure 6A). Next, we confirmed that 396 mRNA expression of genes, exhibiting decreased BRG1 occupancy upon IpmkKD and reduced ATAC-seq 397 signals upon IpmkKD/Brg1KD at H3K4me3-Only (Nmral1) or bivalent (Phactr3) promoters, were 398 significantly down-regulated upon IpmkKD ( Figure 6B). To examine this in a genome-wide manner, we 399 identified differentially expressed genes (DEGs) by comparing the gene expression of IpmkKD with that reduced near TSS of down-regulated DEGs upon IpmkKD compared to EgfpKD ( Figure 6D). We further 403 confirmed this by monitoring the specific down-regulated DEG loci, including Phactr3, Lrrc61, and 404 Arhgap44 ( Figure 6B, right, E, and F). Together, these results suggest that IPMK maintains the expression 405 of a subset of genes by safeguarding the appropriate BRG1 occupancy and BRG1-mediated chromatin 406 accessibility at promoter-TSS in mESCs.

408
IPMK and SMARCB1 regulate a common set of genes in mESCs and MEFs

409
To investigate the functional interactions between IPMK and SMARCB1 in a transcription aspect, we 410 performed mRNA-seq using mESCs upon IpmkKD or Smarcb1KD. Using the same criteria (fold change ≥

441
To further investigate whether the functional interaction between IPMK and SMARCB1 is a specific 442 feature of mESCs or a more general feature, we performed identical analyses in MEFs (NIH3T3). We first

485
The N-terminal β sheets of the Rpt2 domain seem to be quite buried at the frontal view of nucleosome-486 bound SWI/SNF complex (He et al., 2020). However, if we rotate the structure and see the back view, we 487 could observe that the beginning part (that is close to C-terminal α helices of Rpt1) of N-terminal β sheets 488 of Rpt2 is being exposed, which may allow some space for IPMK binding. More importantly, we also

516
Despite our findings, the mechanism on how IPMK modulates the localization of BRG1 (or SWI/SNF 517 complex) remains unclear. We believe that the physical association of the IPMK-SWI/SNF complex is 518 strongly connected with the BRG1 localization, but the detailed mechanism is still elusive. IPMK may 519 facilitate the recruitment of the SWI/SNF complex's subunits from cytoplasm to chromatin. However, our 520 observations from chromatin fractionation assays indicated that IPMK depletion did not affect the BRG1 occupancy upon IpmkKD ( Figure 4C and G-I). In support of this, IPMK depletion exhibited more impact 531 on the chromatin accessibility and transcription of the relatively-low-BRG1-harboring bivalent promoters 532 than H3K4me3-Only promoters, which harbor high BRG1 levels ( Figure 5 and 6). Together, these results 533 indicate that IPMK primarily regulates the BRG1 occupancy (which resembles the SWI/SNF-nucleosome 534 interactions) and its downstream effects (chromatin accessibility and transcription) at the region where the 535 BRG1 level is originally low in mESCs but does not affect the regions with high BRG1 level. This 536 contextual discrepancy in IPMK-dependent or IPMK-independent BRG1 occupancy may provide key clues 537 to the mechanism on how IPMK modulates the BRG1 localization, but further experiments/analyses are 538 required.

539
Previous studies showing that IPMK regulates several target proteins (including cytosolic signaling 540 factors and transcription factors) through protein-protein interaction also support our view. In earlier studies,

541
IPMK was found to bind to mTOR and raptor, maintaining the mTOR-raptor association and amino acid-

571
Our study is the first to elucidate the physical association between IPMK and core subunits of SWI/SNF 572 complex, and the first to define the molecular function of IPMK in coordinating the BRG1 localizations           NaCl, 20 mM EDTA, 4 mM EGTA, 0.05% Digitonin, 0.1 mg/ml RNase A, 50 μg/ml glycogen) and 788 incubated at 37°C for 30min on a nutator. Beads were placed on a magnet stand and the liquid was removed 789 to a fresh tube, followed by addition of 2 μl 10% SDS and 2.5 μl proteinase K (20mg/ml) and incubated at 790 50°C for 1 hour. DNA was extracted using phenol chloroform as described at

826
The list of 47,382 mouse genes was obtained from the UCSC genome browser (

828
Among these genes, we selected protein-coding genes (gene name starting with NM_) that are longer than 829 2 kb. To classify the promoter-TSS regions precisely, we removed redundancies by merging the genes,                     (blue) and Brg1KD (green) at TSS with two promoter types.