Molecular Cell

B-catenin functions as a molecular adapter for disordered cBAF interactions

Graphical abstract

ß-catenin serves as a modular hub for cBAF interactions

SF-1-dependent steroidogenesis

Transcription factor

MC2R (Receptor activation)

cBAF

p300

SCARB1 (Cholesterol import)

STAR (Cholesterol transporter)

ARID1A IDR

ß-catenin

B-catenin coordinates binding of cBAF, CBP/p300, and transcription factors

CYP17A1

(enzyme) CYP11A1 (enzyme)

Aldosterone Testosterone Cortisol

YAP1 SOX2 FOXO3 SF-1

Authors

Yuen San Chan, Qinyu Gao, Sarah A. Robinson, … , Vaclav Veverka, Katerina Cermakova, H. Courtney Hodges

Correspondence

katerina.cermakova@bcm.edu (K.C.), chodges@bcm.edu (H.C.H.)

In brief

By dissecting steroidogenic expression dependencies, Chan et al. reveal that B-catenin mediates physical association of cBAF with key binding partners. The authors show that a disordered region of cBAF subunit ARID1A interacts with Armadillo repeats on ß-catenin, highlighting roles for modular adapters in IDR-mediated interactions.

Highlights

. Dissection of functional interactions involving B-catenin and cBAF (SWI/SNF)

. Disordered sequences of ARID1A directly bind folded Armadillo repeats on B-catenin

. B-catenin mediates cBAF interaction with SF-1 (NR5A1), YAP1, SOX2, FOXO3, and CBP/p300

. Contact with structured protein mediates key IDR-driven cBAF interactions

Check for updates

CellPress

Molecular Cell

CelPress OPEN ACCESS

Article -catenin functions as a molecular adapter for disordered cBAF interactions

Yuen San Chan, 1,2 Qinyu Gao, 1,2 Sarah A. Robinson, 1,2 Wenzhi Wang, 1,2 Ruzena Filandrova,3 Lisa-Maria Weinhold,3,4 Mario Loeza Cabrera, 1,2 Miao Zhang,5 Chandra Shekar R. Ambati,6 Antonio M. Lerario,7 Nagireddy Putluri, 1,6,8 Katja Kiseljak-Vassiliades, 9,10 Margaret E. Wierman, 9,10 Mouhammed Amir Habra, 11 Gary D. Hammer,7,12 Vaclav Veverka,3,4 Katerina Cermakova,2,13,* and H. Courtney Hodges1,2,8,14,15,*

1Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, TX 77030, USA

2Center for Precision Environmental Health, Baylor College of Medicine, Houston, TX 77030, USA

3Institute of Organic Chemistry and Biochemistry, Czech Academy of Sciences, Prague, Czech Republic

4Department of Cell Biology, Faculty of Science, Charles University, Prague, Czech Republic

5Department of Pathology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA 6Advanced Technology Cores, Baylor College of Medicine, Houston, TX 77030, USA

7Department of Internal Medicine, Division of Metabolism, Endocrinology and Diabetes, University of Michigan, Ann Arbor, MI 48105, USA

8Dan L Duncan Comprehensive Cancer Center, Baylor College of Medicine, Houston, TX 77030, USA

9Division of Endocrinology, Metabolism and Diabetes, Department of Medicine, University of Colorado School of Medicine at Colorado Anschutz Medical Campus, Aurora, CO 80045, USA

10Research Service Rocky Mountain Regional Veterans Affairs Medical Center, Aurora, CO 80045, USA

11Department of Endocrine Neoplasia and Hormonal Disorders, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA

12Endocrine Oncology Program, Rogel Cancer Center, University of Michigan Health System, Ann Arbor, MI 48109, USA

13Department of Biochemistry and Molecular Pharmacology, Baylor College of Medicine, Houston, TX 77030, USA

14Department of Bioengineering, Rice University, Houston, TX 77030, USA

15Lead contact

*Correspondence: katerina.cermakova@bcm.edu (K.C.), chodges@bcm.edu (H.C.H.) https://doi.org/10.1016/j.molcel.2025.06.026

SUMMARY

BAF (SWI/SNF) chromatin remodelers engage binding partners to generate site-specific DNA accessibility. However, the basis for interaction between BAF and divergent binding partners has remained unclear. Here, we tested the hypothesis that scaffold proteins augment BAF’s binding repertoire by examining ß-cat- enin (CTNNB1) and steroidogenic factor 1 (SF-1, NR5A1), a transcription factor central to steroid production in human cells. BAF inhibition rapidly opposed SF-1/B-catenin enhancer occupancy, impairing SF-1 target activation and SF-1/B-catenin autoregulation. These effects arise due to ß-catenin’s role as a molecular adapter between SF-1 and an intrinsically disordered region (IDR) of the canonical BAF (cBAF) subunit ARID1A. In contrast to exclusively IDR-driven mechanisms, adapter function is mediated by direct associa- tion of ARID1A with B-catenin’s folded Armadillo repeats. B-catenin similarly linked cBAF to YAP1, SOX2, FOXO3, and CBP/p300, reflecting a general IDR-mediated mechanism for modular coordination between factors. Molecular visualization highlights ß-catenin’s adapter role for interaction of cBAF with binding partners.

INTRODUCTION

BAF (SWI/SNF) family ATP-dependent chromatin remodelers act as critical co-activators in many settings. The coupling of cata- lytic ATPase activity with non-catalytic interactions involving BAF complexes is essential to enhancer function, 1-3 opposition to constitutive and facultative heterochromatin, 4,5 replication, response to DNA damage,7,8 transcriptional pausing,9 and many other activities.10,11 The high degree of cell plasticity enabled by these processes arises in part from the diverse inter-

actomes of BAF complexes, 12,13 which include selective interac- tions with transcription factors (TFs),14 histone-modifying en- zymes,15 proteins that mediate chromatin architecture, 16 transcription elongation,17 RNA splicing,9 DNA damage,7 and many other cofactors. 12,13,18,19 Collectively, these interactions give rise to highly tissue- and state-specific functions.

The core architectures of BAF family complexes have revealed the structural basis for their capacity to mobilize nucleo- somes.20-23 BAF chromatin remodelers assemble the primary ATPase subunit SMARCA4 (BRG1) into at least three major

( Check for updates

Molecular Cell 85, 3041-3056, August 21, 2025 @ 2025 The Author(s). Published by Elsevier Inc. 3041 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

CelPress OPEN ACCESS

Molecular Cell Article

Figure 1. BAF (SWI/SNF) ATPase activity is required for SF-1-mediated steroidogenesis

SF-1 (NR5A1) expression in endocrine tissues

SMARCA4 high vs. low (TCGA adrenocortical carcinoma, ACC)

Adrenal gland-

Ovary-

Adrenal gland

Pituitary gland

Pituitary gland-

ß-catenin targets

Log2 fold change (steroid high vs. low)

MC2R

SF-1 targets p = 2.9e-6

Steroid biosynthesis p = 0.03

HSD3B2

Testis

CYP17A1

Hypothalamus

CYP21A2

Thymus

p = 0.0007

SCARB1

Stomach

0.4

CYP11A1

Pancreas Colon Lung

0.6

Ovary

Testis

R = 0.76, p < 2.2e-16

Prostate

0.0

-4

Liver

10000

50 100 150

Rank expression change

-2

Expression (TPM)

Log2 fold change (SMARCA4 high vs. low)

SES component

Protein levels (tissue microarray)

Steroid phenotype

Low

SF-1

ß-catenin

SMARCA4

High

Steroid

Age

<50

Low Steroid phenotype High

Age

n.s.

>50

Sex

n.s.

Sex

ß-catenin

SMARCA4

SF-1

Z-score

-4

SES

Low

Intermediate

High

27%

41%

100%

Steroid phenotype

Low

-20 um

High

N = 11

N = 17

N = 7

Log2 FC over DMSO

RNA-seq

300

p > 0.05 or not detected

STAR

· SF-1 targets

CYP11A1

· ß-catenin targets

Cholesterol

STAR

-5

-Log10 FDR p

VCAN

CYP11A1

CYP17A1

17alpha- hydroxypregnenolone

CYP17A1

200

Pregnenolone

DHEA

SCARB1

HSD3B2

Levels

4.3e-4

CYP17A1

17alpha- hydroxyprogesterone

100

TCF7

Progesterone

CCND2

HSD3B2

CYP21A2

BRM014

MYC

CYP21A2

MC2R

SP5

Deoxycorticosterone

HSD3B2

1.9e-4

-4

-2

Log2 fold change (BRM014 vs. DMSO)

Levels

11-deoxycortisol

Androstenedione

BRM014

Levels

2.1e-4

CYP11B1

Gene set enrichment analysis

Corticosterone

BRM014

HSD17B3 AKR1C3

Levels

1.4e-4

(BRM014 vs. DMSO)

BRM014

CYP11A1

-1

NES

CYP11B2

Steroid biosynthesis

-2

Aldosterone

Cortisol

Testosterone

West adrenocortical

tumor up

Levels

6.1e-4

Levels

4.4e-3

9.4e-3

Levels

-3

SF-1 targets

1000

2000

BRM014

Gene set rank

Mineralocorticoids

Glucocorticoids

Androgens

(A) Steroidogenic factor 1 (SF-1, NR5A1) expression across human tissues. Data from the human protein atlas.

(B) Gene set enrichment analysis (GSEA) in SMARCA4 high vs. low tumors in the adrenocortical carcinoma (ACC) TCGA cohort. ES, enrichment score. (C) Correlation of RNA expression changes between high (HSP) and low steroid phenotype (LSP) compared with RNA expression changes between SMARCA4- high and SMARCA4-low ACC tumors in TCGA. SF-1 targets are highlighted in red (SMARCA4-high, N = 28; SMARCA4-low, N = 51; HSP, N = 47; LSP, N = 31; see STAR Methods for details). For analysis of steroid output, only patients with annotated steroid phenotypes were considered.

Molecular Cell Article

CelPress OPEN ACCESS

sub-complexes that are defined by their unique subunit compo- sitions.24 For example, ARID1A is found within canonical (cBAF) complexes,25 whereas PBRM1 and BRD7 are found within poly- bromo-associated (PBAF) complexes,26 and BRD9 is found in non-canonical (ncBAF or GBAF) complexes.24,27 Instead of the well-ordered cores, which power nucleosome mobilization, intrinsically disordered regions (IDRs) have been attributed as the primary interaction surfaces that mediate these complexes’ capacity to engage diverse interaction partners.13,28,29 Unfortu- nately, the IDRs on these complexes are highly dynamic and remain structurally unresolved. As a result, many questions have remained regarding the biochemical basis for the extensive interactions of BAF complexes with TFs and other chromatin machinery.

Disordered protein has emerged as an important mediator of protein interactions, particularly for nuclear proteins involved in chromatin and transcription.28 IDRs have a high capacity to engage other IDRs through heterotypic interactions 13,30-33; how- ever, the biochemical mechanisms that enable selectivity in IDR- IDR interactions remain incompletely understood. By contrast, a growing number of selective interactions have been identified whereby IDRs engage well-ordered folded protein domains with high specificity.28,34-36 IDR-domain affinities are highly var- iable but substantially overlap with those observed between folded proteins37 and are sufficient to sustain protein-protein interactions. 34,38-40

We hypothesized that folded domains with more extensive interaction surfaces could mediate selective interactions with IDRs in BAF complexes and thereby expand the BAF interac- tion repertoire in a modular fashion. We therefore sought to identify a candidate protein that interacts both with TFs and BAF complexes and whose proximity is essential for tran- scription activation. These requirements were readily fit by B-catenin (CTNNB1), a protein primarily composed of folded Armadillo repeats and IDRs that plays scaffolding roles in other settings.41 ß-catenin interacts with TCF/LEF and BAF com- plexes during canonical Wnt signaling42-44 but also has selec- tive interactions with multiple TFs and co-activators, including those from YAP,45,46 SOx,45,46 and FOXO47 families, as well as the nuclear receptor TF known as steroidogenic factor 1 (SF-1, NR5A1).48,49 B-catenin also directly interacts with histone acetyltransferases CBP/p300,39 which cooperate with BAF complexes to regulate TF activity and gene expres- sion.50,51 The overlap with BAF’s binding partners as well as the lack of a known DNA-binding domain and the extensive interaction surface of ß-catenin prompted us to hypothesize that it may serve as a molecular adapter for interactions be- tween BAF complexes and shared interaction partners.

Compared with many other TFs, SF-1 activates a relatively consistent set of target genes.52 Ectopic expression of SF-1 in many settings is sufficient to activate the entire steroid biosyn- thesis pathway, including the cholesterol importer STAR and cy- tochrome P450 enzymes like CYP17A1 that are essential for ste- roid biosynthesis. 53-55 As a result, SF-1 expression is generally restricted to endocrine tissues (Figure 1A). The highly defined transcriptional effects of SF-1 across many cell types make SF-1 an ideal TF to test and dissect the potential adapter role of ß-catenin for BAF complexes and to evaluate currently unknown roles of BAF’s involvement in steroid hormone production.

Here, we evaluated a role for ß-catenin as a molecular adapter for chromatin remodeling by dissecting BAF-related functions in SF-1-mediated steroidogenesis. Our findings reveal that cBAF remodeling activity enables SF-1/ß-catenin binding at enhancers to activate steroidogenic gene expression, as well as to sustain autoregulation of the steroidogenic circuitry. We find that ß-cat- enin acts as an essential molecular adapter that mediates bind- ing of cBAF not only to SF-1 but also to an array of other ß-cat- enin binding partners, including diverse TFs and CBP/p300. By dissecting the specific protein elements needed for biochemical association between the involved factors, we determined that an IDR in the cBAF-specific subunit ARID1A mediates direct phys- ical association with folded Armadillo repeats on ß-catenin, high- lighting key roles for stably folded molecular adapters in orches- trating interactions of disordered regions on cBAF with TFs and chromatin regulators.

RESULTS

Excess SF-1-dependent hormone production is linked with high SMARCA4 and -catenin

To interrogate BAF- and ß-catenin-related mechanisms for SF-1 activity, we identified adrenocortical carcinoma (ACC), a tumor with highly activated SF-1/B-catenin transcriptional programs and a high frequency of excess steroid hormone production, 48,56 as an ideal experimental setting. Analysis of ACC data from The Cancer Genome Atlas (TCGA; Figures S1A-S1F)56 revealed that ACC tumors with high SMARCA4 expression exhibit higher expression of SF-1 targets (gene set enrichment analysis [GSEA] normalized enrichment score [NES] = 2.0, p = 2.9e-6), steroidogenic genes (NES = 1.6, p = 0.03), and ß-catenin targets (NES = 1.8, p = 7.2e-4) compared with tumors with low SMARCA4 expression (Figure 1B; Table S1). Moreover, global gene expression patterns between these two groups strongly correlated with expression changes between high (HSP) and low steroid phenotype (LSP) groups (R = 0.76, p < 2.2e-16,

(D) Representative IHC images of steroidogenic expression signature (SES) components SMARCA4, ß-catenin, and SF-1 in ACC by steroid phenotype. (E) Heatmap of unbiased hierarchical clustering of validation cohort of ACC tumors. Tumors are assigned to HSP or LSP based on clinical detection of excess steroid hormone production. Pie charts indicate the proportions of tumors with HSP in each category.

(F) Expression changes induced by BRM014 in H295R cells, as measured by RNA-seq. SF-1 targets are presented in red, and ß-catenin targets are presented in blue.

(G) Gene sets from differentially reduced transcripts upon BRM014 treatment in H295R cells ranked by normalized enrichment score (NES).

(H) BRM014-induced expression changes (filled circles) from RNA-seq presented in (F). Inset bar charts indicate DMSO-normalized production of individual steroid hormones or precursors in H295R cells, as assessed by ELISA or LC-MS. Error bars are mean ± SEM (n = 3 independent replicates).

*p < 0.05, ** p < 0.01, **** p < 2.2e-16. See also Figures S1-S3 and Table S1.

CelPress OPEN ACCESS

Molecular Cell Article

Figure 2. cBAF mediates DNA accessibility and SF-1/B-catenin binding at enhancers to sustain autoregulatory steroidogenic expression (A) SF-1 motif DNA accessibility (black) and SF-1 protein levels (red) measured after 1 h treatment with BRM014 or DMSO control. Error bars are mean ± SEM (ATAC-seq n = 2 independent replicates, protein levels n = 3 independent replicates). Protein levels were measured using western blot densitometry. p value for ATAC-seq from hypergeometric test and p value for protein levels from t test. (B) Ranking of TF motif enrichment based on p value from sites that lose accessibility. Volcano plot (inset) shows differential DNA accessibility in H295R cells treated with 1 h of BRM014 vs. DMSO control.

-Log10 FDR p

12358

102

SF-1

SF-1 ATAC motif score (Z scaled)

SMARCA4

ß-catenin

ATAC

p = 0.64

800

decr.

incr.

H3K4me1

BRM014

1.0

SF-1 protein level (norm.)

SF-1

SF-1 sites

DMSO

-Log10 p

12|

15.

0.5

400

TCF/LEF

-2 0

Mean signal

Log2 FC (BRM014 vs. DMSO)

Signal

0.0

3432

max

-20

p < 2.2e-16

220

440

sites

DMSO BRM014

Motif rank, decreased sites (BRM014 vs. DMSO)

min

4 kb

# + CYP11A1

Count (×1000)

chr15 74,330 74,350 kb

Input

ChromHMM state

ARID1A

kDa

lgG

aSF-1

PBRM1

1.0

Polycomb repressed

50-

SF-1

100

BRD9

Proportion

Heterochromatin

Insulator

ß-catenin

SMARCA4

250

0.5

Primed enhancer

ChIP

H3K4me1

Active enhancer

ATPase

SMARCA4

Weak enhancer

H3K4me3

Flank TSS

cBAF

250

ARID1A

260

SF-1 (DMSO)

0.0

Flank TSS downstream

180-

PBRM1

260

SF-1 (BRM014)

SF-1

Bivalent TSS

PBAF

ß-catenin (DMSO)

ß-catenin

Active TSS

BRD7

ß-catenin (BRM014)

Occupancy after 1 h BRM014

Retained

ncBAF

75-

BRD9

ATAC

125

Accessibility (DMSO)

O Lost

125

Accessibility (BRM014)

Enhancer-promoter contacts (4C)

Enh

HH NR5A1

BRM014 treatment time (h)

chr9

124,400

124,600

124,800 kb

0 1 2 4 8 24 48 72

kDa

chr3 |+ CTNNB1

SF-1

-50

41.2

41.3Mb

ß-catenin

-100

ARID1A

SMARCA4

250

PBRM1

-250

BRD9

SMARCA2

SMARCA4

ß-actin

ChIP

100

125

105

H3K4me1

100

125

850

H3K4me3 SF-1 (DMSO)

140

150

Enh

SF-1

140

150

SF-1 (BRM014)

210

200

ß-catenin (DMSO)

Enh

ß-catenin

cBAF

210

200

ß-catenin (BRM014)

ATAC

140

Accessibility (DMSO)

Enh

SF-1 targets

140

Accessibility (BRM014)

Enh

(C) Heatmap of genomic SMARCA4-regulated sites that show reduced SF-1 and -catenin occupancy along with reduced DNA accessibility following 1 h of BRM014 treatment.

(D) Classification of individual clusters with differential loss of SF-1/B-catenin at different genomic regions defined by ChromHMM using histone marks. Bar plot indicates the number of sites in each cluster.

(E) Co-immunoprecipitation (CoIP) of endogenous proteins using anti-SF-1 antibodies in H295R cells (n = 2 independent replicates).

(F) Browser tracks of SF-1 target CYP11A1 locus showing loss of SF-1/B-catenin and reduced DNA accessibility at cBAF-regulated enhancer (Enh) following 1 h of BRM014 treatment or DMSO control.

(legend continued on next page)

Molecular Cell Article

CelPress OPEN ACCESS

Figure 1C; Table S1). No association was observed for the minor BAF ATPase SMARCA2 (BRM, Figures S1C-S1F), under- scoring the unique relationship between SF-1, ß-catenin, and SMARCA4. Unsupervised clustering of tumors based on expres- sion levels of SF-1 (NR5A1), B-catenin (CTNNB1), and SMARCA4 enabled classification based on their joint expression, which we termed the steroidogenic expression signature (SES). Elevated RNA levels of this signature (SES RNA-high) were highly predic- tive of HSP (chi-square test, p = 1.7e-5; Figures S1G and S1H). Compared with SF-1 and ß-catenin RNA levels without SMARCA4, the addition of SMARCA4 levels significantly improved classification accuracy for HSP (ANOVA, p < 0.03). Hence, elevated expression of steroidogenic pathways in vivo is associated with high RNA levels of SF-1, ß-catenin, and SMARCA4.

We performed immunohistochemistry (IHC) of a tissue micro- array composed of an independent cohort of primary human ACC tumors (N = 35), adrenal adenomas (ACA, N = 11), and normal adrenal cortex tissue samples (N = 11; Figures S2A- S2D). In non-malignant ACA, neither ß-catenin nor SMARCA4 was overexpressed (p > 0.05). However, SMARCA4 (p = 2.7e-8) and ß-catenin (p = 2.7e-3) were both overexpressed at the protein level in malignant ACC (Figures S2C and S2D). Similarly, 100% of SES protein-high tumors overproduced ste- roid hormones (Figures 1D, 1E, and S2E-S2G; chi-square test, p < 0.01). Our analyses link high SMARCA4 and ß-catenin protein levels to SF-1-dependent hormone production in humans.

SMARCA4 regulates SF-1 transcriptional activity

To test whether BAF catalytic activity is required for SF-1- dependent gene expression, we treated two human steroido- genic ACC cell lines, H295R57 and CU-ACC1,58 with the SMARCA4/2 ATPase inhibitor BRM01417,59-61 for 72 h and per- formed RNA sequencing (RNA-seq). BAF ATPase inhibition strongly reduced the expression of SF-1 targets that play key roles in steroid metabolism, including the adrenocorticotropic hormone (ACTH) receptor MC2R, cholesterol receptor SCARB1, cholesterol transporter STAR, and cytochrome P450 enzymes CYP11A1 and CYP17A1, as well as ß-catenin targets (Figures 1F and S3A; Table S1). GSEA revealed that SF-1 targets (NES = - 2.9, p = 6.3e-27) and genes associated with steroidogenesis (NES = - 1.7, p = 0.02) were among the most significantly downregulated gene sets (Figures 1G, S3B, and S3C; Table S1). These effects were reproducible upon treatment with the orthogonal SMARCA4/2 ATPase inhibitor FHD-286 or degrader AU-1533062 (Figures S3D and S3E), con- firming on-target effects. Consistent with these changes, the levels of hormones aldosterone (t test p = 6.1e-4), cortisol (p=4.4e-3), and testosterone (p = 9.4e-3), as well as their pre- cursors, were significantly decreased upon BAF inhibition, as assessed by ELISA and liquid chromatography-mass spec-

trometry (LC-MS) (Figures 1H and S3F). Our results demon- strate that BAF ATPase activity is essential for steroid hormone production in steroid-producing cells.

Consistent with the human data, short hairpin RNA (shRNA) knockdown revealed that SF-1 targets primarily depend on SMARCA4 for their expression, while only a subset depends on the alternative ATPase SMARCA2 (Figures S3G and S3H). Our findings establish SMARCA4 as the primary BAF ATPase governing the SF-1 steroidogenic program.

cBAF enables enhancer binding of SF-1 and -catenin

To determine how BAF ATPase activity influences the chromatin occupancy of SF-1, we began by measuring changes in DNA accessibility at sites bearing SF-1 motifs following acute inhibi- tion using BRM014 in H295R cells. We observed pronounced genome-wide losses of DNA accessibility at sites matching SF-1 consensus motifs within 1 h of BRM014 treatment (p < 2.2e-16), a time when SF-1 protein levels were unchanged (p = 0.64; Figure 2A). While other motifs associated with ß-cate- nin binding partners, such as TCF/LEF, had reduced accessi- bility, reductions at sites matching the SF-1 motif were the most significantly associated with decreased accessibility (Figure 2B; Table S2).

Chromatin profiling revealed strongly overlapping genome- wide binding patterns of SF-1, B-catenin, and SMARCA4 (Figure S4A), with sites overlapping SMARCA4 having the stron- gest occupancy of SF-1 and ß-catenin (Figure S4B). We further- more identified 3,432 genomic sites jointly bound by SMARCA4, SF-1, and ß-catenin that showed reduced DNA accessibility along with reduced binding of SF-1 and ß-catenin following 1 h of BRM014 treatment (Figures 2C and S4C). These sites were also highly enriched in the enhancer mark H3K4me1 and depleted of the active promoter mark H3K4me3 (Figure S4D), consistent with reduced dependency on BAF at promoters.17,60,61 These sites were, moreover, associated with loss of CBP/p300 binding (Figure S4E), indicating coordination between BAF and histone acetyltransferase activity. Annotation of genome-wide positions using a 10-state ChromHMM63 classi- fier confirmed that sites bearing losses of SF-1 and ß-catenin were enriched at enhancers and depleted at transcription start sites (Figures 2D and S4F-S4H).

Immunoprecipitation (IP) of SF-1 confirmed its physical inter- action with -catenin48 and revealed its interaction with cBAF subunits ARID1A and SMARCA4 (Figure 2E). By contrast, no interaction was detected with PBAF subunits PBRM1 or BRD7, or with ncBAF subunit BRD9 (Figure 2E). In agreement, BAF in- hibition induced joint reduction of SF-1 and ß-catenin binding at enhancers of SF-1 target genes associated with high occu- pancy of cBAF subunit ARID1A and low levels of PBAF subunit PBRM1 or ncBAF subunit BRD9 (Figures 2F, S4D, and S4F). These positions were also significantly enriched near SF-1 target

(G and H) (G) Browser tracks of SF-1 (NR5A1) and (H) B-catenin (CTNNB1) loci. Enhancers (Enh) identified by 4C anchored to the SF-1 promoter or previously identified for ß-catenin are indicated in red. Identities of browser track rows for (G) are the same as in (H).

(I) Time course western blot of SF-1, B-catenin, SMARCA4, and SMARCA2 during 1 µM BRM014 treatment in H295R cells (n = 3 independent replicates for 1 and 72 h).

(J) Schematic of the autoregulatory steroidogenic loop sustained by cBAF activity. Red lines indicate SF-1 activity, and blue lines indicate ß-catenin activity. See also Figures S4 and S5 and Table S2.

CelPress OPEN ACCESS

genes whose expression levels were reduced upon BAF inhibi- tion (p = 2.7e-13; Figures S41-S4K). Despite earlier work sug- gesting a unique role of PBAF for nuclear receptor activity, 64 our results demonstrate that cBAF complexes act as direct co- activators of SF-1 at enhancers. We conclude that physical inter- action of cBAF complexes with SF-1 and ß-catenin generates chromatin accessibility for their binding at enhancers involved in steroidogenesis.

cBAF sustains feedforward expression of SF-1 and ß-catenin

cBAF-regulated SF-1/B-catenin sites included enhancers that make long-distance contact with the promoters of NR5A1 (en- coding SF-1) and CTNNB1 (ß-catenin; Figures 2G and 2H). We therefore hypothesized that cBAF complexes are required to maintain the autoregulatory expression loop of SF-1 and ß-cate- nin. We extended BRM014 treatment and confirmed that the protein levels of both SF-1 and ß-catenin progressively dimin- ished following sustained BAF inhibition without altering the expression of BAF ATPases SMARCA4 and SMARCA2 (Figures 2I and S5A). Additionally, shRNA knockdown confirmed that expression of SF-1 and ß-catenin was mutually dependent (Figures S5B and S5C) and that SMARCA4 activity is upstream of both SF-1 and ß-catenin (Figure S3G). Our results demon- strate that cBAF sustains feedforward autoregulatory expression of SF-1 and ß-catenin by maintaining their binding at their own enhancers (Figure 2J), thereby representing an essential compo- nent of the self-sustaining expression circuit observed in steroid- producing tissues. 48,65

B-catenin is a molecular adapter between cBAF and SF-1

To dissect functional interactions in a context where SF-1 and B-catenin are not genetically essential or subject to confounding autoregulation, we used HEK 293T cells with ectopic expression of SF-1 at physiological levels (Figures 3A and S6A). Compared with empty vector control, SF-1 expression induced strong upre- gulation of canonical SF-1 targets CYP17A1 and STAR (Figure 3B), which served as markers of SF-1 activity and provided a setting for biochemical dissection. Knockout (KO) of ß-catenin revealed that it was required for SF-1-dependent upregulation of SF-1 targets (Figures 3A and 3B) and for physical interaction between BAF complexes and SF-1 (Figure 3C). By contrast, physical interaction of cBAF and ß-catenin was inde- pendent of SF-1 (Figure S6B). This led us to hypothesize that B-catenin acts as a molecular adapter between cBAF and SF-1. To dissect the protein elements responsible for physical association with SF-1 and cBAF complexes, we generated a set of hemagglutinin (HA)-tagged ß-catenin truncation and dele- tion mutants (Figure 3D) and expressed them in HEK 293T cells. IP of these constructs revealed that B-catenin’s C-terminal IDR is necessary for binding SF-1 (Figures 3E and 3F). Using a similar approach, we demonstrated the SF-1 ligand-binding domain (LBD) was required for physical association between SF-1 and B-catenin (Figures 3G, 3H, and S6C). Our results in human cells agree with previous reports that a region in the LBD is essential for murine SF-1 to bind to ß-catenin in yeast two-hybrid studies.66 By contrast, all of ß-catenin’s disordered regions, including the C-terminal IDR, were dispensable for interaction

with BAF complexes via SMARCA4 IP, which was instead lost following deletion of ß-catenin’s well-ordered Armadillo repeats 4-6 (Figures 31 and 3J). IP of ARID1A confirmed complete loss of interaction between ß-catenin and cBAF complexes upon deletion of Armadillo repeats 4-6 (Figure 3K). Impaired contact between ß-catenin and either cBAF or SF-1 was associated with failure to fully activate SF-1 transcription activity (Figure S6D). Our studies confirm that ß-catenin acts as a molec- ular adapter between cBAF and SF-1 (Figure 3L).

IDR2 of ARID1A mediates association of cBAF with ß-catenin

BRM014 treatment (Figures 4A and 4B) or dual knockout of cBAF subunits ARID1A and ARID1B (Figures 4C, 4D, S7A, and S7B) abolished the activation of SF-1 targets in HEK 293T cells. In contrast to ARID1A/B, knockout of cBAF subunits DPF1, DPF2, or DPF3, PBAF subunit PBRM1, or ncBAF subunit BRD9 did not abolish SF-1 transcription activity (Figures S7C- S7F). Among non-catalytic subunits, these experiments confirm a unique role for cBAF’s ARID subunit in the activation of SF-1 targets.

To identify a minimal protein element of ARID1A that medi- ates SF-1 transcriptional activity, we designed full-length (FL) or sequential deletion mutants of FLAG-tagged ARID1A (Figure 4E). Each of these constructs contains the core binding region (CBR) to ensure complex integrity and inclusion of downstream subunits during cBAF assembly.24 We expressed these constructs in dual ARID1A/ARID1B knockout HEK 293T cells expressing SF-1 (Figure S8A). Using STAR as a reporter of SF-1 activity revealed that, compared with FL constructs, the N-terminal IDR (IDR1) and ARID domain were dispensable, but deletion of ARID1A’s second IDR (IDR2) strongly impaired SF-1 activity (Figure 4F). Loss of IDR2 furthermore uniquely eliminated cBAF binding to ß-catenin as observed using IP of either SMARCA4 (Figure 4G) or ARID1A (Figure S8B), while pre- serving cBAF complex integrity as evidenced by inclusion of downstream cBAF subunit DPF2 (Figures 4G and S8B). Tar- geted rather than sequential deletions (Figures S8C and S8D) revealed a degree of functional redundancy of the IDRs. How- ever, the sufficiency of IDR2 to mediate cBAF binding to ß-cat- enin (Figures 4G and S8B) defined it as a minimal region enabling this interaction.

The ARID1A IDR2 directly interacts with ß-catenin Armadillo repeats

The above results using cell-based assays raised two key ques- tions: (1) does loss of ß-catenin Armadillo repeats 4-6 induce only local structural changes, or does it impact the long-range organization of the Armadillo domain? (2) Which regions of ARID1A IDR2 and the Armadillo repeats, if any, directly interact? To answer these questions, we expressed and purified recombi- nant ARID1A IDR2, as well as the ß-catenin Armadillo repeat domain with or without repeats 4-6 (Figures 5A and S9A). We then performed cross-linking mass spectrometry (XL-MS; Figure 5B) using these constructs. Deletion of repeats 4-6 within the Armadillo domain led to significantly shorter homotypic cross-links detected between repeats (t test p = 0.0002; Figure 5C; Table S3). Pairwise contact plots revealed that

Molecular Cell Article

CelPress OPEN ACCESS

Figure 3. B-catenin functions as an adapter for cBAF interaction with SF-1

HEK 293T ß-catenin

STAR

CYP17A1

2.5e-6

1.5e-3

ß-catenin

kDa

1.9e-6

1.6e-4

0.15

0.20

kDa

ß-catenin

100

Fold expression

over TBP

-75

0.15

SF-1

0.10

SF-1

SF-1-FLAG

-50

0.10

ß-catenin

Vinculin

-120

0.05

SMARCA4

190

SF-1

0.00

Input

lgG

aFLAG

Input

lgG

aFLAG

ß-catenin

Armadillo repeats

C HA

kDa

SF-1 IP

SF-1-FLAG constructs

100

DBD

Hinge

LBD

FLAG

781

ß-catenin-HA constructs

ΔΝ

ß-catenin-

-75

Input

ALBD

DBD

Hinge

FLAG

134

781

-Log10 p

ΔIDR

41-3

SMARCA4

Input

kDa

133 267

781

-190

44-6

SF-1-FLAG

267 395

781

-50

ß-catenin

-90

-100

aFLAG IP

410-12

;A7-91

ß-actin

SF-1-FLAG

47-9

ß-catenin-

-75

44-6

41-3

ALBD

FL ALBD

kDa

394 521

781

FLAG

A10-12

521 665

781

-2

ALBD

-25

-50

100

Log2 (IP/FL), (input norm.)

ß-catenin

-75

664

781

A1-3

44-6

47-9

A10-12

AIDR

SMARCA4

190

ΔIDR

lgG

aFLAG

lgG

aFLAG

134

664

781

kDa

100

SMARCA4 IP

ß-catenin-HA

44-6

ß-catenin-

Input

kDa

ARID1A

-Log10 p

44-6

250

SMARCA4

-190

SMARCA4

-190

aSMARCA4 IP

ß-catenin-HA

-100

100

SAN

lgG

-75

ß-catenin- HA

;410-12

Input

aARID1A

Input

aARID1A

47-9;

ΔDR

41-3

ß-catenin

-2

N IDR

C IDR

A1-3

44-6

A7-9

A10-12

AIDR

Log2 (IP/FL), (input norm.)

Armadillo repeats

cBAF

SF-1 LBD

(A) Western blot of wild-type (WT) or knockout (KO) of ß-catenin with ectopic expression of SF-1 (+) or empty vector (EV) control (-) in HEK 293T cells. (B) qPCR analysis of SF-1 targets STAR and CYP17A1 in HEK 293T cells with WT or KO -catenin and EV or SF-1. Error bars are mean ± SEM (n = 3 independent replicates).

(D) Design of HA-tagged ß-catenin deletion and truncation mutants.

(E) ColP of ß-catenin deletion constructs with FLAG-tagged SF-1 (n = 3 independent replicates).

(F) Volcano plot of co-immunoprecipitation densitometry in (E) (n = 3 independent replicates). Constructs without significant loss of interaction with SF-1 are shown in gray.

(G) Design of FLAG-tagged SF-1 constructs. Full length (FL) and deletion of the ligand-binding domain (ALBD) are indicated.

(H) ColP of SMARCA4, B-catenin, and SF-1 in HEK 293T cells expressing either FL or ALBD SF-1, using immunoglobulin G (lgG) control or anti-FLAG antibodies. Input samples are shown separately (inset).

(I) ColP of ß-catenin deletion constructs with SMARCA4 using SMARCA4 antibody (n = 3 independent replicates).

(J) Volcano plot of co-immunoprecipitation densitometry in (I) (n = 3 independent replicates). Constructs without significant loss of interaction with SMARCA4 are shown in gray.

(K) Validation of altered cBAF interaction by co-immunoprecipitation of FL or 44-6 B-catenin using IgG control or anti-ARID1A antibody.

(L) Schematic of interactions between cBAF, B-catenin, and SF-1.

Abbreviations are as follows: IDR, intrinsically disordered region; DBD, DNA-binding domain; LBD, ligand-binding domain. See also Figure S6.

deletion of repeats 4-6 eliminated long-range cross-links involving other repeats (Figure 5D). Additionally, the only homo- typic cross-links observed in both the intact domain and following deletion of repeats 4-6 were those found entirely within

repeats 1-3 (Figure 5D). Hence our XL-MS data are consistent with disrupted structural organization across the Armadillo domain. Importantly, we also detected multiple heterotypic cross-links between ARID1A IDR2 and ß-catenin Armadillo

CelPress OPEN ACCESS

Molecular Cell Article

Figure 4. IDR2 of ARID1A mediates interaction of cBAF and B-catenin

HEK 293T

STAR

CYP17A1

SF-1

9.6e-6

4.3e-3

kDa

1.1e-4

5.1e-3

SF-1

-50

-100

Fold expression over TBP

0.15

· DMSO

ß-catenin

· BRM014

-75

0.10

ß-actin

0.05

BRM014

0.00

SF-1

0.00

SF-1

ARID1A/B

STAR

CYP17A1

2.5e-6

5.9e-7

SF-1

EV SF-1

kDa

2.3e-6

5.6e-7

ARID1A

Fold expression over TBP

0.15

-250

0.20

SF-1

ARID1B

250

0.15

0.10

SF-1

0.10

-50

0.05

Vinculin

120

0.00

ARID1A/B

ARID1A-FLAG constructs

IDR1

ARID

IDR2

CBR

FLAG

ΔIDR1

ΔDR1

ΔARD

ΔARID

AIDR2

AIDR1

kDa

250

ΔIDR1ΔΑRID

SMARCA4

ΔIDR1ΔARIDAIDR2

971 1170 1611 2285 a.a.

-250

ΔIDR1ΔΑRID

-150

ΔIDR1ΔARIDAIDR2

-75

DPF2

-37

AIDR1

ß-catenin

100

AIDR14ARID

ΔIDR1ΔARIDAIDR2

Input

lgG

aSMARCA4

Input

lgG

aSMARCA4

Input

lgG

aSMARCA4

0.0

0.1

0.2

0.3

STAR expression over TBP

(A) Western blot of ectopic expression of EV SF-1 in HEK 293T cells treated with 1 µM BRM014 or DMSO for 72 h.

(B) qPCR analysis of SF-1 targets STAR and CYP17A1 in cells shown in (A). Error bars are mean ± SEM (n = 3 independent replicates).

(C) Western blot of HEK 293T cells expressing wild-type (WT) or dual knockout (KO) of ARID1A and ARID1B, with or without ectopic expression of SF-1.

(D) qPCR analysis of SF-1 targets STAR and CYP17A1 in cells shown in (C). Error bars are mean ± SEM (n = 3 independent replicates). (E) Design of truncation variants of ARID1A. Positions indicate amino acid (aa) number.

(F) qPCR analysis of SF-1 target STAR normalized to housekeeping gene TBP in HEK 293T cells stably expressing SF-1 with ectopic expression of constructs shown in (E). Constructs with significantly compromised SF-1 activity compared with FL ARID1A are shown in gray. Error bars are mean ± SEM (n = 3 independent replicates).

(G) ColP to investigate the interaction between BAF complexes and ß-catenin in cells expressing indicated ARID1A constructs (n = 4 independent replicates). Abbreviations are as follows: IDR, intrinsically disordered region; ARID, AT-rich interaction domain; CBR, core binding region.

**** p < 0.0001.

Molecular Cell Article

CellPress OPEN ACCESS

Figure 5. ARID1A IDR2 directly interacts with ß-catenin folded Armadillo repeats

GST-ß-catenin(Arm)

GST-ß-catenin(Arm)44-6

MBP-ARID1A(IDR2)

Mix at equimolar ratio

p = 0.0002

Homotypic cross-link sequence distance (a.a.)

Cross-linking, quenching

400

GST

MBP

Digestion

kDa

260

140

200

100

HPLC-MS/MS

Hetero- and homotypic cross-linked fragment ID

WT 44-6

ß-catenin(Arm)

B-catenin(Arm)44-6

ARID1A K1201 - B-catenin K281

Armadillo repeat

10-12

Armadillo repeat

ARID1A

r rrrrr Ffff.fr

SNSVGIQDAFNDGSDSTFQKR

7-9

10-12

(IDR2)

1182

1202

4-6

7-9

ß-catenin (Arm 4)

rrrrrrr

275

LAGGLOKMYALLNK

1-3

Intensity (103)

288

1-3 4-6 7-9 10-12 Armadillo repeat

1-3 7-9 10-12

Armadillo repeat

Observed with:

· GST-ß-catenin(Arm) only

· GST-ß-catenin(Arm)44-6 only

500

1000

1500

m/z

o Both

ARID1A

IDR1

ARID

IDR2

CBR

IDR1

ARID

IDR2

CBR

1170

1611

1170

1611

ß-catenin

Contacts

1-3 4-6 7-9 10-12

1-3

7-9 10-12

- Homotypic

267 396

Contacts

267 396 Armadillo repeats

— Heterotypic

Armadillo repeats

- Homotypic

— Heterotypic (n.d.)

(A) SDS-PAGE of purified recombinant proteins used for cross-linking mass spectrometry (XL-MS).

(B) Schematic summary of XL-MS workflow.

(D) Positions of homotypic cross-links between ß-catenin Armadillo repeats.

(E) Mass spectrum corresponding to a heterotypic cross-link between ARID1A IDR2 and ß-catenin (additional spectra in Figure S9). (F) Homotypic and heterotypic cross-links detected involving ARID1A IDR2 and intact ß-catenin Armadillo repeats.

(G) Homotypic and heterotypic cross-links involving ARID1A IDR2 and 44-6 B-catenin Armadillo repeats. Dashed black line indicates the deleted region within the Armadillo domain (repeats 4-6). In (F) and (G), solid black lines indicate homotypic cross-links, and dashed red lines indicate heterotypic cross-links.

Abbreviations are as follows: IDR, intrinsically disordered region; ARID, AT-rich interaction domain; CBR, core binding region; Arm, Armadillo repeat; n.d., not detected.

Molecular Cell Article

Figure 6. B-catenin acts as a general adapter for cBAF interactions with TFs and chromatin regulators

ß-catenin

kDa

IDR1

ARID

IDR2

CBR

SMARCA4

190

ARID1A

250

N IDR

Armadillo repeats

C IDR

ß-catenin

100

ß-catenin

YAP1

DBD

Hinge

LBD

SOX2

SF-1

FOXO3

CBP/p300

250

c-Jun

Input

lgG

aSMARCA4

Input

lgG

aSMARCA4

SF-1 DBD

p300

ß-catenin dependent

Control

Co-IP intensity (WT norm.)

n.s.

1.0

cBAF

ß-catenin

SF-1

ß-catenin

ARID1A

IDR2

LBD

C IDR

0.5

ß-catenin

SF-1

B-catenin

LBD

Arm

YAP1

SOX2

FOXO3

CBP/p300

c-Jun

300

Motif family

-Log10 FDR p (BRM014 vs. DMSO)

Mean CBP/p300 signal (over background)

YAP1 partner (TEAD)

SOX

FOXO

200

Motif family

DMSO

BRM014

YAP1 partner

100

(TEAD)

· SOX

· FOXO

-4

-Log10 FDR p (ß-catenin KO vs. WT)

Distance from decreased site (BRM014 vs. DMSO) bearing motif (kb)

(A) Summary of interactions between cBAF, B-catenin, and SF-1. Disordered regions are shown in light blue, and structured regions are shown in dark blue. (B) Integrative three-dimensional model based on observed protein contacts between SF-1, B-catenin, and cBAF bound to a dinucleosome. Contacts are supported by experimental data, but atomic-level configurations of molecular interfaces are hypothetical. The disordered N terminus of ß-catenin is not shown for clarity.

(C) Co-immunoprecipitation of YAP1, SOX2, FOXO3, CBP/p300, and c-Jun control with SMARCA4 in HEK 293T cells expressing wild-type (WT) or knockout (KO) B-catenin using IgG control or anti-SMARCA4 antibody.

(D) Quantification of co-immunoprecipitation densitometry in (C) (n = 3 independent replicates).

(E) Enrichment of motifs at sites that lose DNA accessibility upon BAF inhibition or B-catenin KO. Key TF families are highlighted.

(F) Metagene plots of CBP/p300 chromatin immunoprecipitation (ChIP) chromatin occupancy at sites bearing key TF motifs that lose DNA accessibility following 1 h treatment of BRM014 or DMSO.

Abbreviations are as follows: IDR, intrinsically disordered region; ARID, AT-rich interaction domain; CBR, core binding region; Arm, Armadillo repeat; DBD, DNA- binding domain; LBD, ligand-binding domain.

*p < 0.05, ** p < 0.01.

Molecular Cell Article

CelPress OPEN ACCESS

cBAF to multiple binding partners. Excitingly, we discovered that B-catenin-interacting TFs and co-activators YAP1 (t test p = 0.0095, n = 3), SOX2 (p = 0.039, n = 3), FOXO3 (p = 0.0076, n = 3), and histone acetyltransferase CBP/p300 (p = 0.0051, n = 3) were all dependent on ß-catenin for interaction with cBAF in HEK 293T cells (Figures 6C and 6D). By contrast, the AP-1 factor c-Jun, which binds directly to SMARCD1,74 interacts independently of ß-catenin (p = 0.52, n = 3; Figures 6C and 6D), serving as a control to illustrate specificity. Consistent with these findings, DNA accessibility at sites bearing motifs of TEAD (YAP1 binding partner), SOX, or FOXO families was reduced upon either ß-catenin knockout or 1 h of BRM014 treat- ment (Figure 6E; Table S4). In agreement with our integrative three-dimensional model, CBP/p300 also displayed reduced chromatin occupancy at these sites (Figure 6F). Our results reveal that ß-catenin serves as a general molecular adapter to cBAF used by many factors, physically linking the chromatin re- modeling activity of cBAF to TFs and other chromatin regulators.

DISCUSSION

BAF chromatin remodeling complexes are critical regulators of gene expression, yet biochemical characterization of their inter- actomes remains a significant challenge. This limitation is largely attributable to the prevalence of disordered and flexible regions within these complexes, which are important for mediating pro- tein-protein interactions. Despite substantial efforts, less than 40% of the mass of these complexes has been resolved to date.28 As a result, understanding how IDRs in molecular ma- chines such as BAF maintain their multifaceted protein interac- tomes represents a currently unresolved issue. Here, we discover a general IDR-mediated mechanism by which cBAF en- gages other transcriptional regulators by using ß-catenin as a molecular adapter.

ß-catenin harbors a folded domain composed of helical Arma- dillo repeats, a structure uniquely suited for supporting a wide range of IDR-mediated interactions.75-80 These properties make ß-catenin an effective molecular bridge, linking cBAF com- plexes to SF-1, YAP1, SOX2, and FOXO3, as well as histone ace- tyltransferases CBP/p300. In addition to the effects on steroido- genesis we highlight here, the dependency on ß-catenin for these interactions reveals a modular basis by which cBAF en- gages factors essential for pluripotency,81 aging,82 and tissue homeostasis.83 Because ß-catenin’s interaction partners include additional transcription regulators41 that are also BAF-depen- dent, 62,84,85 we anticipate that factors beyond those tested here may display similar dependency on its molecular adapter function. Our finding that structured elements mediate IDR- driven cBAF interactions further emphasizes the growing recog- nition that IDR selectivity is strongly influenced by interactions with folded domains. 28,34

Our work reveals an important chromatin regulatory axis un- derlying endocrine function and positions cBAF as a critical ste- roidogenic dependency. Our results suggest that effects on ste- roid production may be an important consideration for the use of BAF ATPase inhibitors currently in clinical trials as anti-cancer agents.86-88 In particular, production of the SF-1-dependent glucocorticoid cortisol suppresses inflammation,89 promotes

T cell exhaustion,90 and impairs T cell infiltration into tumors, thereby reducing the efficacy of immune checkpoint therapies. 91 Elevated glucocorticoid levels in adrenal cancer indeed repre- sent a key hurdle for the efficacy of immune therapies. 92-96 As a result, potential changes in steroid production following BAF inhibition have the potential to exert local or systemic effects on immune microenvironments, including in tumors. By focusing on dependencies of steroid production, our work uncovers a widely used avenue for disordered protein to influence endocrine activities via the regulation of chromatin and transcription.

Limitations of the study

Our data reveal a key role for ß-catenin in mediating IDR-depen- dent cBAF activities. Although we show this function is mediated by direct interaction, we cannot exclude the involvement of addi- tional factors that may influence this regulatory axis. Achieving atomic-level resolution of the involved interfaces will require further structural analyses, such as cryo-EM or X-ray crystallog- raphy. Moreover, our study does not address the temporal order of events by which BAF and B-catenin associate with SF-1, CBP/ p300, or other transcription regulators. Finally, although we did not detect a functional interaction between SF-1 and either PBAF or ncBAF complexes, we cannot rule out that these com- plexes may have direct or indirect roles in ß-catenin transcrip- tional regulation.

RESOURCE AVAILABILITY

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, H. Courtney Hodges (chodges@bcm.edu).

Materials availability

All reagents used in this study can be found in the key resources table. Plas- mids generated in this study are available upon request to the lead contact.

Data and code availability

· High-throughput sequencing data generated for this project have been deposited in the Gene Expression Omnibus (GEO) database with accession numbers GEO: GSE260866, GSE260867, and GSE261006. Proteomic data generated for this project have been deposited in the ProteomeXchange database with accession number PRIDE: PXD063603. Original western blot images, including independent repli- cates, have been deposited at Mendeley Data (https://doi.org/10. 17632/2xm4vsv6w8.1).

. This paper does not report original code. Scripts used for genome-wide analyses are available on Zenodo97 and at the following link: https:// github.com/hodgeslab/workflows.

. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

ACKNOWLEDGMENTS

This work was supported by National Institutes of Health (NIH) grants R35GM137996 (H.C.H.), R01CA272769 (H.C.H.), P30CA125123 (N.P.), and R01CA220297 (N.P.); the Helis Family Medical Research Foundation (H.C. H.); The Mark Foundation for Cancer Research ASPIRE award (H.C.H.); CPRIT Proteomics and Metabolomics Core Facility (RP210227); the Dan L Duncan Comprehensive Cancer Center (H.C.H.); the Charles University Grant Agency grant 710120 (L .- M.W.); and Czech Science Foundation grant 25-15442X (V. V.). Portions of the figures were created using BioRender. We thank S. Boey- naems (BCM) for helpful discussion, F. Filandr and P. Junková (IOCB, Czech

CelPress OPEN ACCESS

Molecular Cell Article

Academy of Sciences) for assistance, and the UTMB Next Gen Sequencing fa- cility for assistance.

AUTHOR CONTRIBUTIONS

Y.S.C .: formal analysis, visualization, methodology, and writing - original draft. Q.G., S.A.R., W.W., and M.L.C .: investigation and writing - review and editing. R.F., L .- M.W., C.S.R.A., A.M.L., N.P., and G.D.H .: investigation. M.Z .: re- sources. K.K .- V., M.E.W., M.A.H., and V.V .: investigation and resources. K. C. and H.C.H .: conceptualization, resources, formal analysis, supervision, funding acquisition, visualization, methodology, writing - original draft, project administration, and writing - review and editing.

DECLARATION OF INTERESTS

G.D.H. reports unrelated personal fees from Radionetics and Orphagen Phar- maceuticals for consultation on projects outside the scope of this work.

STAR*METHODS

Detailed methods are provided in the online version of this paper and include the following:

· KEY RESOURCES TABLE

· EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS o Cell lines

· METHOD DETAILS

o Cell-line generation

o Inhibitors and degraders

o RT-qPCR

o Western blotting

o Co-immunoprecipitation

o Chromatin immunoprecipitation and sequencing

o RNA sequencing

o ATAC sequencing

o 4C library preparation, sequencing, and analysis

o Hormone profiling

o Protein purification

o Cross-linking mass spectrometry

o Analysis of patient samples

o Structural model of enhancer activation

. QUANTIFICATION AND STATISTICAL ANALYSIS

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j. molcel.2025.06.026.

Received: December 16, 2024

Revised: May 16, 2025

Accepted: June 27, 2025

Published: July 21, 2025

REFERENCES

1. Cermakova, K., Tao, L., Dejmek, M., Sala, M., Montierth, M.D., Chan, Y. S., Patel, I., Chambers, C., Loeza Cabrera, M., Hoffman, D., et al. (2024). Reactivation of the G1 enhancer landscape underlies core circuitry addiction to SWI/SNF. Nucleic Acids Res. 52, 4-21. https://doi.org/10. 1093/nar/gkad1081.

2. Chambers, C., Cermakova, K., Chan, Y.S., Kurtz, K., Wohlan, K., Lewis, A.H., Wang, C., Pham, A., Dejmek, M., Sala, M., et al. (2023). SWI/SNF Blockade Disrupts PU.1-Directed Enhancer Programs in Normal Hematopoietic Cells and Acute Myeloid Leukemia. Cancer Res. 83, 983-996. https://doi.org/10.1158/0008-5472.CAN-22-2129.

3. Hodges, H.C., Stanton, B.Z., Cermakova, K., Chang, C .- Y., Miller, E.L., Kirkland, J.G., Ku, W.L., Veverka, V., Zhao, K., and Crabtree, G.R. (2018). Dominant-negative SMARCA4 mutants alter the accessibility landscape of tissue-unrestricted enhancers. Nat. Struct. Mol. Biol. 25, 61-72. https://doi.org/10.1038/s41594-017-0007-3.

4. Stanton, B.Z., Hodges, C., Calarco, J.P., Braun, S.M.G., Ku, W.L., Kadoch, C., Zhao, K., and Crabtree, G.R. (2017). Smarca4 ATPase mu- tations disrupt direct eviction of PRC1 from chromatin. Nat. Genet. 49, 282-288. https://doi.org/10.1038/ng.3735.

5. Kadoch, C., Williams, R.T., Calarco, J.P., Miller, E.L., Weber, C.M., Braun, S.M.G., Pulice, J.L., Chory, E.J., and Crabtree, G.R. (2017). Dynamics of BAF-Polycomb complex opposition on heterochromatin in normal and oncogenic states. Nat. Genet. 49, 213-222. https://doi. org/10.1038/ng.3734.

6. Bayona-Feliu, A., Barroso, S., Muñoz, S., and Aguilera, A. (2021). The SWI/SNF chromatin remodeling complex helps resolve R-loop- mediated transcription-replication conflicts. Nat. Genet. 53, 1050- 1063. https://doi.org/10.1038/s41588-021-00867-2.

7. Watanabe, R., Ui, A., Kanno, S.I., Ogiwara, H., Nagase, T., Kohno, T., and Yasui, A. (2014). SWI/SNF Factors Required for Cellular Resistance to DNA Damage Include ARID1A and ARID1B and Show Interdependent Protein Stability. Cancer Res. 74, 2465-2475. https://doi.org/10.1158/ 0008-5472.CAN-13-3608.

8. Lee, H .- S., Park, J .- H., Kim, S .- J., Kwon, S .- J., and Kwon, J. (2010). A cooperative activation loop among SWI/SNF, y-H2AX and H3 acetylation for DNA double-strand break repair. EMBO J. 29, 1434-1445. https://doi. org/10.1038/emboj.2010.27.

9. Batsché, E., Yaniv, M., and Muchardt, C. (2006). The human SWI/SNF subunit Brm is a regulator of alternative splicing. Nat. Struct. Mol. Biol. 13, 22-29. https://doi.org/10.1038/nsmb1030.

10. Hodges, C., Kirkland, J.G., and Crabtree, G.R. (2016). The Many Roles of BAF (mSWI/SNF) and PBAF Complexes in Cancer. Cold Spring Harb. Perspect. Med. 6, a026930. https://doi.org/10.1101/cshperspect.a026930.

11. Karki, M., Jangid, R.K., Anish, R., Seervai, R.N.H., Bertocchio, J .- P., Hotta, T., Msaouel, P., Jung, S.Y., Grimm, S.L., Coarfa, C., et al. (2021). A cytoskeletal function for PBRM1 reading methylated microtu- bules. Sci. Adv. 7, eabf2866. https://doi.org/10.1126/sciadv.abf2866.

12. Ho, L., Ronan, J.L., Wu, J., Staahl, B.T., Chen, L., Kuo, A., Lessard, J., Nesvizhskii, A.I., Ranish, J., and Crabtree, G.R. (2009). An embryonic stem cell chromatin remodeling complex, esBAF, is essential for embry- onic stem cell self-renewal and pluripotency. Proc. Natl. Acad. Sci. USA 106, 5181-5186. https://doi.org/10.1073/pnas.0812889106.

13. Patil, A., Strom, A.R., Paulo, J.A., Collings, C.K., Ruff, K.M., Shinn, M.K., Sankar, A., Cervantes, K.S., Wauer, T., St Laurent, J.D., et al. (2023). A disordered region controls cBAF activity via condensation and partner recruitment. Cell 186, 4936-4955.e26. https://doi.org/10.1016/j.cell. 2023.08.032.

14. Barisic, D., Stadler, M.B., lurlaro, M., and Schübeler, D. (2019). Mammalian ISWI and SWI/SNF selectively mediate binding of distinct transcription factors. Nature 569, 136-140. https://doi.org/10.1038/ s41586-019-1115-5.

15. Alver, B.H., Kim, K.H., Lu, P., Wang, X., Manchester, H.E., Wang, W., Haswell, J.R., Park, P.J., and Roberts, C.W.M. (2017). The SWI/SNF chromatin remodelling complex is required for maintenance of lineage specific enhancers. Nat. Commun. 8, 14648. https://doi.org/10.1038/ ncomms14648.

16. Miller, E.L., Hargreaves, D.C., Kadoch, C., Chang, C.Y., Calarco, J.P., Hodges, C., Buenrostro, J.D., Cui, K., Greenleaf, W.J., Zhao, K., et al. (2017). TOP2 synergizes with BAF chromatin remodeling for both resolu- tion and formation of facultative heterochromatin. Nat. Struct. Mol. Biol. 24, 344-352. https://doi.org/10.1038/nsmb.3384.

17. Brahma, S., and Henikoff, S. (2024). The BAF chromatin remodeler syn- ergizes with RNA polymerase II and transcription factors to evict

Molecular Cell Article

CelPress OPEN ACCESS

nucleosomes. Nat. Genet. 56, 100-111. https://doi.org/10.1038/s41588- 023-01603-8.

18. Hainer, S.J., Gu, W., Carone, B.R., Landry, B.D., Rando, O.J., Mello, C. C., and Fazzio, T.G. (2015). Suppression of pervasive noncoding tran- scription in embryonic stem cells by esBAF. Genes Dev. 29, 362-378. https://doi.org/10.1101/gad.253534.114.

19. Saha, D., Animireddy, S., Lee, J., Thommen, A., Murvin, M.M., Lu, Y., Calabrese, J.M., and Bartholomew, B. (2024). Enhancer switching in cell lineage priming is linked to eRNA, Brg1’s AT-hook, and SWI/SNF recruitment. Mol. Cell 84, 1855-1869.e5. https://doi.org/10.1016/j.mol- cel.2024.03.013.

20. Mashtalir, N., Suzuki, H., Farrell, D.P., Sankar, A., Luo, J., Filipovski, M., D’Avino, A.R., St Pierre, R., Valencia, A.M., Onikubo, T., et al. (2020). A Structural Model of the Endogenous Human BAF Complex Informs Disease Mechanisms. Cell 183, 802-817.e24. https://doi.org/10.1016/j. cell.2020.09.051.

21. He, S., Wu, Z., Tian, Y., Yu, Z., Yu, J., Wang, X., Li, J., Liu, B., and Xu, Y. (2020). Structure of nucleosome-bound human BAF complex. Science 367, 875-881. https://doi.org/10.1126/science.aaz9761.

22. Han, Y., Reyes, A.A., Malik, S., and He, Y. (2020). Cryo-EM structure of SWI/SNF complex bound to a nucleosome. Nature 579, 452-455. https://doi.org/10.1038/s41586-020-2087-1.

23. Wagner, F.R., Dienemann, C., Wang, H., Stützer, A., Tegunov, D., Urlaub, H., and Cramer, P. (2020). Structure of SWI/SNF chromatin remodeller RSC bound to a nucleosome. Nature 579, 448-451. https://doi.org/10. 1038/s41586-020-2088-0.

24. Mashtalir, N., D’Avino, A.R., Michel, B.C., Luo, J., Pan, J., Otto, J.E., Zullow, H.J., Mckenzie, Z.M., Kubiak, R.L., St Pierre, R., et al. (2018). Modular Organization and Assembly of SWI/SNF Family Chromatin Remodeling Complexes. Cell 175, 1272-1288.e20. https://doi.org/10. 1016/j.cell.2018.09.032.

25. Maxwell, M.B., Hom-Tedla, M.S., Yi, J., Li, S., Rivera, S.A., Yu, J., Burns, M.J., McRae, H.M., Stevenson, B.T., Coakley, K.E., et al. (2024). ARID1A suppresses R-loop-mediated STING-type I interferon pathway activation of anti-tumor immunity. Cell 187, 3390-3408.e19. https://doi.org/10. 1016/j.cell.2024.04.025.

26. Grossi, E., Nguyen, C.B., Carcamo, S., Kirigin Callaú, V., Moran, S., Filipescu, D., Tagore, S., Firestone, T.M., Keogh, M.C., Sun, L., Izar, B., Hasson, D., and Bernstein, E. (2025). The SWI/SNF PBAF complex facil- itates REST occupancy at repressive chromatin. Mol. Cell. 85, 1714- 1729.e7. https://doi.org/10.1016/j.molcel.2025.03.026.

27. Alpsoy, A., and Dykhuizen, E.C. (2018). Glioma tumor suppressor candi- date region gene 1 (GLTSCR1) and its paralog GLTSCR1-like form SWI/ SNF chromatin remodeling subcomplexes. J. Biol. Chem. 293, 3892- 3903. https://doi.org/10.1074/jbc.RA117.001065.

28. Cermakova, K., and Hodges, H.C. (2023). Interaction modules that impart specificity to disordered protein. Trends Biochem. Sci. 48, 477-490. https://doi.org/10.1016/j.tibs.2023.01.004.

29. Kim, Y.R., Joo, J., Lee, H.J., Kim, C., Park, J .- C., Yu, Y.S., Kim, C.R., Lee, D.H., Cha, J., Kwon, H., et al. (2024). Prion-like domain mediated phase separation of ARID1A promotes oncogenic potential of Ewing’s sar- coma. Nat. Commun. 15, 6569. https://doi.org/10.1038/s41467-024- 51050-0.

30. Farag, M., Borcherds, W.M., Bremer, A., Mittag, T., and Pappu, R.V. (2023). Phase separation of protein mixtures is driven by the interplay of homotypic and heterotypic interactions. Nat. Commun. 14, 5527. https://doi.org/10.1038/s41467-023-41274-x.

31. Lyons, H., Veettil, R.T., Pradhan, P., Fornero, C., De La Cruz, N., Ito, K., Eppert, M., Roeder, R.G., and Sabari, B.R. (2023). Functional partitioning of transcriptional regulators by patterned charge blocks. Cell 186, 327- 345.e28. https://doi.org/10.1016/j.cell.2022.12.013.

32. Zamudio, A.V., Dall’Agnese, A., Henninger, J.E., Manteiga, J.C., Afeyan, L.K., Hannett, N.M., Coffey, E.L., Li, C.H., Oksuz, O., Sabari, B.R., et al.

(2019). Mediator Condensates Localize Signaling Factors to Key Cell Identity Genes. Mol. Cell 76, 753-766.e6. https://doi.org/10.1016/j.mol- cel.2019.08.016.

33. Pei, G., Lyons, H., Li, P., and Sabari, B.R. (2025). Transcription regulation by biomolecular condensates. Nat. Rev. Mol. Cell Biol. 26, 213-236. https://doi.org/10.1038/s41580-024-00789-x.

34. Cermakova, K., Demeulemeester, J., Lux, V., Nedomova, M., Goldman, S.R., Smith, E.A., Srb, P., Hexnerova, R., Fabry, M., Madlikova, M., et al. (2021). A ubiquitous disordered protein interaction module orchestrates transcription elongation. Science 374, 1113-1121. https://doi.org/10. 1126/science.abe2913.

35. Martin, E.W., Thomasen, F.E., Milkovic, N.M., Cuneo, M.J., Grace, C.R., Nourse, A., Lindorff-Larsen, K., and Mittag, T. (2021). Interplay of folded domains and the disordered low-complexity domain in mediating hnRNPA1 phase separation. Nucleic Acids Res. 49, 2931-2945. https://doi.org/10.1093/nar/gkab063.

36. Davey, N.E., Simonetti, L., and Ivarsson, Y. (2023). The next wave of in- teractomics: Mapping the SLIM-based interactions of the intrinsically disordered proteome. Curr. Opin. Struct. Biol. 80, 102593. https://doi. org/10.1016/j.sbi.2023.102593.

37. Lazar, T., Tantos, A., Tompa, P., and Schad, E. (2022). Intrinsic protein disorder uncouples affinity from binding specificity. Protein Sci. 31, e4455. https://doi.org/10.1002/pro.4455.

38. Ghoneim, M., Fuchs, H.A., and Musselman, C.A. (2021). Histone Tail Conformations: A Fuzzy Affair with DNA. Trends Biochem. Sci. 46, 564-578. https://doi.org/10.1016/j.tibs.2020.12.012.

39. Brown, A.D., Cranstone, C., Dupré, D.J., and Langelaan, D.N. (2023). ß — Catenin interacts with the TAZ1 and TAZ2 domains of CBP/p300 to acti- vate gene transcription. Int. J. Biol. Macromol. 238, 124155. https://doi. org/10.1016/j.ijbiomac.2023.124155.

40. Cermakova, K., Veverka, V., and Hodges, H.C. (2023). The TFIIS N-terminal domain (TND): a transcription assembly module at the inter- face of order and disorder. Biochem. Soc. Trans. 51, 125-135. https:// doi.org/10.1042/BST20220342.

41. Valenta, T., Hausmann, G., and Basler, K. (2012). The many faces and functions of ß-catenin. EMBO J. 31, 2714-2736. https://doi.org/10. 1038/emboj.2012.150.

42. Lickert, H., Takeuchi, J.K., Von Both, I., Walls, J.R., McAuliffe, F., Adamson, S.L., Henkelman, R.M., Wrana, J.L., Rossant, J., and Bruneau, B.G. (2004). Baf60c is essential for function of BAF chromatin remodelling complexes in heart development. Nature 432, 107-112. https://doi.org/10.1038/nature03071.

43. Barker, N., Hurlstone, A., Musisi, H., Miles, A., Bienz, M., and Clevers, H. (2001). The chromatin remodelling factor Brg-1 interacts with beta-cate- nin to promote target gene activation. EMBO J. 20, 4935-4943. https:// doi.org/10.1093/emboj/20.17.4935.

44. Sharma, T., Olea-Flores, M., and Imbalzano, A.N. (2023). Regulation of the Wnt signaling pathway during myogenesis by the mammalian SWI/ SNF ATPase BRG1. Front. Cell Dev. Biol. 11, 1160227. https://doi.org/ 10.3389/fcell.2023.1160227.

45. Rosenbluh, J., Nijhawan, D., Cox, A.G., Li, X., Neal, J.T., Schafer, E.J., Zack, T.I., Wang, X., Tsherniak, A., Schinzel, A.C., et al. (2012). ß — Catenin-Driven Cancers Require a YAP1 Transcriptional Complex for Survival and Tumorigenesis. Cell 151, 1457-1473. https://doi.org/10. 1016/j.cell.2012.11.026.

46. Mukherjee, S., Luedeke, D.M., McCoy, L., Iwafuchi, M., and Zorn, A.M. (2022). SOX transcription factors direct TCF-independent WNT/B — catenin responsive transcription to govern cell fate in human pluripotent stem cells. Cell Rep. 40, 111247. https://doi.org/10.1016/j.celrep.2022. 111247.

47. Essers, M.A.G., de Vries-Smits, L.M.M., Barker, N., Polderman, P.E., Burgering, B.M.T., and Korswagen, H.C. (2005). Functional Interaction

CelPress OPEN ACCESS

Molecular Cell Article

Between B-Catenin and FOXO in Oxidative Stress Signaling. Science 308, 1181-1184. https://doi.org/10.1126/science.1109083.

48. Mohan, D.R., Borges, K.S., Finco, I., LaPensee, C.R., Rege, J., Solon, A. L., Little, D.W., Else, T., Almeida, M.Q., Dang, D., et al. (2023). ß-Catenin- Driven Differentiation Is a Tissue-Specific Epigenetic Vulnerability in Adrenal Cancer. Cancer Res. 83, 2123-2141. https://doi.org/10.1158/ 0008-5472.CAN-22-2712.

49. Berthon, A., Martinez, A., Bertherat, J., and Val, P. (2012). Wnt/ß-catenin signalling in adrenal physiology and tumour development. Mol. Cell. Endocrinol. 351, 87-95. https://doi.org/10.1016/j.mce.2011.09.009.

50. Doughty, B.R., Hinks, M.M., Schaepe, J.M., Marinov, G.K., Thurm, A.R., Rios-Martinez, C., Parks, B.E., Tan, Y., Marklund, E., Dubocanin, D., et al. (2024). Single-molecule states link transcription factor binding to gene expression. Nature 636, 745-754. https://doi.org/10.1038/s41586-024- 08219-w.

51. Yang, J.H., and Hansen, A.S. (2024). Enhancer selectivity in space and time: from enhancer-promoter interactions to promoter activation. Nat. Rev. Mol. Cell Biol. 25, 574-591. https://doi.org/10.1038/s41580-024- 00710-6.

52. Doghman, M., Figueiredo, B.C., Volante, M., Papotti, M., and Lalli, E. (2013). Integrative analysis of SF-1 transcription factor dosage impact on genome-wide binding and gene expression regulation. Nucleic Acids Res. 41, 8896-8907. https://doi.org/10.1093/nar/gkt658.

53. Ruiz-Babot, G., Eceiza, A., Abollo-Jiménez, F., Malyukov, M., Carlone, D. L., Borges, K., Da Costa, A.R., Qarin, S., Matsumoto, T., Morizane, R., et al. (2023). Generation of glucocorticoid-producing cells derived from human pluripotent stem cells. Cell Rep. Methods 3, 100627. https:// doi.org/10.1016/j.crmeth.2023.100627.

54. Val, P., Lefrançois-Martinez, A .- M., Veyssière, G., and Martinez, A. (2003). SF-1 a key player in the development and differentiation of ste- roidogenic tissues. Nucl. Recept. 1, 8. https://doi.org/10.1186/1478- 1336-1-8.

55. Schimmer, B.P., and White, P.C. (2010). Minireview: Steroidogenic Factor 1: Its Roles in Differentiation, Development, and Disease. Mol. Endocrinol. 24, 1322-1337. https://doi.org/10.1210/me.2009-0519.

56. Zheng, S., Cherniack, A.D., Dewal, N., Moffitt, R.A., Danilova, L., Murray, B.A., Lerario, A.M., Else, T., Knijnenburg, T.A., Ciriello, G., et al. (2016). Comprehensive pan-genomic characterization of adrenocortical carci- noma. Cancer Cell 29, 723-736. https://doi.org/10.1016/j.ccell.2016. 04.002.

57. Logié, A., Boudou, P., Boccon-Gibod, L., Baudin, E., Vassal, G., Schlumberger, M., Le Bouc, Y., and Gicquel, C. (2000). Establishment and characterization of a human adrenocortical carcinoma xenograft model. Endocrinology 141, 3165-3171. https://doi.org/10.1210/endo. 141.9.7668.

58. Kiseljak-Vassiliades, K., Zhang, Y., Bagby, S.M., Kar, A., Pozdeyev, N., Xu, M., Gowan, K., Sharma, V., Raeburn, C.D., Albuja-Cruz, M., et al. (2018). Development of new preclinical models to advance adrenocor- tical carcinoma research. Endocr. Relat. Cancer 25, 437-451. https:// doi.org/10.1530/ERC-17-0447.

59. Papillon, J.P.N., Nakajima, K., Adair, C.D., Hempel, J., Jouk, A.O., Karki, R.G., Mathieu, S., Möbitz, H., Ntaganda, R., Smith, T., et al. (2018). Discovery of Orally Active Inhibitors of Brahma Homolog (BRM)/ SMARCA2 ATPase Activity for the Treatment of Brahma Related Gene 1 (BRG1)/SMARCA4-Mutant Cancers. J. Med. Chem. 61, 10155- 10172. https://doi.org/10.1021/acs.jmedchem.8b01318.

60. Martin, B.J.E., Ablondi, E.F., Goglia, C., Mimoso, C.A., Espinel-Cabrera, P.R., and Adelman, K. (2023). Global identification of SWI/SNF targets re- veals compensation by EP400. Cell 186, 5290-5307.e26. https://doi.org/ 10.1016/j.cell.2023.10.006.

61. Duplaquet, L., So, K., Ying, A.W., Pal Choudhuri, S., Li, X., Xu, G.D., Li, Y., Qiu, X., Li, R., Singh, S., et al. (2024). Mammalian SWI/SNF complex ac- tivity regulates POU2F3 and constitutes a targetable dependency in

small cell lung cancer. Cancer Cell 42, 1352-1369.e13. https://doi.org/ 10.1016/j.ccell.2024.06.012.

62. Xiao, L., Parolia, A., Qiao, Y., Bawa, P., Eyunni, S., Mannan, R., Carson, S.E., Chang, Y., Wang, X., Zhang, Y., et al. (2022). Targeting SWI/SNF ATPases in enhancer-addicted prostate cancer. Nature 601, 434-439. https://doi.org/10.1038/s41586-021-04246-z.

63. Ernst, J., and Kellis, M. (2017). Chromatin-state discovery and genome annotation with ChromHMM. Nat. Protoc. 12, 2478-2492. https://doi. org/10.1038/nprot.2017.124.

64. Lemon, B., Inouye, C., King, D.S., and Tjian, R. (2001). Selectivity of chro- matin-remodelling cofactors for ligand-activated transcription. Nature 414, 924-928. https://doi.org/10.1038/414924a.

65. Zubair, M., Ishihara, S., Oka, S., Okumura, K., and Morohashi, K. (2006). Two-Step Regulation of Ad4BP/SF-1 Gene Transcription during Fetal Adrenal Development: Initiation by a Hox-Pbx1-Prep1 Complex and Maintenance via Autoregulation by Ad4BP/SF-1. Mol. Cell. Biol. 26, 4111-4121. https://doi.org/10.1128/MCB.00222-06.

66. Mizusaki, H., Kawabe, K., Mukai, T., Ariyoshi, E., Kasahara, M., Yoshioka, H., Swain, A., and Morohashi, K.I. (2003). Dax-1 (Dosage- Sensitive Sex Reversal-Adrenal Hypoplasia Congenita Critical Region on the X Chromosome, Gene 1) Gene Transcription Is Regulated by Wnt4 in the Female Developing Gonad. Mol. Endocrinol. 17, 507-519. https://doi.org/10.1210/me.2002-0362.

67. Poy, F., Lepourcelet, M., Shivdasani, R.A., and Eck, M.J. (2001). Structure of a human Tcf4-ß-catenin complex. Nat. Struct. Biol. 8, 1053-1057. https://doi.org/10.1038/nsb720.

68. Abramson, J., Adler, J., Dunger, J., Evans, R., Green, T., Pritzel, A., Ronneberger, O., Willmore, L., Ballard, A.J., Bambrick, J., et al. (2024). Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630, 493-500. https://doi.org/10.1038/s41586- 024-07487-w.

69. Omidi, A., Møller, M.H., Malhis, N., Bui, J.M., and Gsponer, J. (2024). AlphaFold-Multimer accurately captures interactions and dynamics of intrinsically disordered protein regions. Proc. Natl. Acad. Sci. USA 121, e2406407121. https://doi.org/10.1073/pnas.2406407121.

70. Hecht, A., Vleminckx, K., Stemmler, M.P., van Roy, F., and Kemler, R. (2000). The p300/CBP acetyltransferases function as transcriptional co- activators of beta-catenin in vertebrates. EMBO J. 19, 1839-1850. https://doi.org/10.1093/emboj/19.8.1839.

71. Field, A., and Adelman, K. (2020). Evaluating Enhancer Function and Transcription. Annu. Rev. Biochem. 89, 213-234. https://doi.org/10. 1146/annurev-biochem-011420-095916.

72. Blassberg, R., Patel, H., Watson, T., Gouti, M., Metzis, V., Delás, M.J., and Briscoe, J. (2022). Sox2 levels regulate the chromatin occupancy of WNT mediators in epiblast progenitors responsible for vertebrate body formation. Nat. Cell Biol. 24, 633-644. https://doi.org/10.1038/ s41556-022-00910-2.

73. Hoogeboom, D., Essers, M.A.G., Polderman, P.E., Voets, E., Smits, L.M. M., and Burgering, B.M.Th. (2008). Interaction of FOXO with ß-Catenin Inhibits ß-Catenin/T Cell Factor Activity. J. Biol. Chem. 283, 9224- 9230. https://doi.org/10.1074/jbc.M706638200.

74. Ito, T., Yamauchi, M., Nishina, M., Yamamichi, N., Mizutani, T., Ui, M., Murakami, M., and Iba, H. (2001). Identification of SWI.SNF Complex Subunit BAF60a as a Determinant of the Transactivation Potential of Fos/Jun Dimers. J. Biol. Chem. 276, 2852-2857. https://doi.org/10. 1074/jbc.M009633200.

75. Kong, C., Qu, X., Liu, M., Xu, W., Chen, D., Zhang, Y., Zhang, S., Zhu, F., Liu, Z., Li, J., et al. (2023). Dynamic interactions between E-cadherin and Ankyrin-G mediate epithelial cell polarity maintenance. Nat. Commun. 14, 6860. https://doi.org/10.1038/s41467-023-42628-1.

76. Rudeen, A.J., Douglas, J.T., Xing, M., McDonald, W.H., Lamb, A.L., and Neufeld, K.L. (2020). The 15-Amino Acid Repeat Region of Adenomatous Polyposis Coli Is Intrinsically Disordered and Retains Conformational

Molecular Cell Article

CelPress OPEN ACCESS

Flexibility upon Binding ß-Catenin. Biochemistry 59, 4039-4050. https:// doi.org/10.1021/acs.biochem.0c00479.

77. Nong, J., Kang, K., Shi, Q., Zhu, X., Tao, Q., and Chen, Y .- G. (2021). Phase separation of Axin organizes the ß-catenin destruction complex. J. Cell Biol. 220, e202012112. https://doi.org/10.1083/jcb.202012112.

78. Richter, G., Gui, T., Bourgeois, B., Koyani, C.N., Ulz, P., Heitzer, E., von Lewinski, D., Burgering, B.M.T., Malle, E., and Madl, T. (2021). ß-catenin regulates FOXP2 transcriptional activity via multiple binding sites. FEBS Journal 288, 3261-3284. https://doi.org/10.1111/febs.15656.

79. Bourgeois, B., Gui, T., Hoogeboom, D., Hocking, H.G., Richter, G., Spreitzer, E., Viertler, M., Richter, K., Madl, T., and Burgering, B.M.T. (2021). Multiple regulatory intrinsically disordered motifs control FOXO4 transcription factor binding and function. Cell Rep. 36, 109446. https:// doi.org/10.1016/j.celrep.2021.109446.

80. Gui, T., Fleming, C., Manzato, C., Bourgeois, B., Sirati, N., Heuer, J., Papadionysiou, I., Montfort, D.I. van, Gijzen, M. van, Smits, L.M.M., et al. (2023). Targeted perturbation of signaling-driven condensates. Mol. Cell 83, 4141-4157.e11. https://doi.org/10.1016/j.molcel.2023. 10.023.

81. Huangfu, D., Osafune, K., Maehr, R., Guo, W., Eijkelenboom, A., Chen, S., Muhlestein, W., and Melton, D.A. (2008). Induction of pluripotent stem cells from primary human fibroblasts with only Oct4 and Sox2. Nat. Biotechnol. 26, 1269-1275. https://doi.org/10.1038/nbt.1502.

82. Morris, B.J., Willcox, D.C., Donlon, T.A., and Willcox, B.J. (2015). FOXO3: A Major Gene for Human Longevity - A Mini-Review. Gerontology 61, 515-525. https://doi.org/10.1159/000375235.

83. Varelas, X. (2014). The Hippo pathway effectors TAZ and YAP in develop- ment, homeostasis and disease. Development 141, 1614-1626. https:// doi.org/10.1242/dev.102376.

84. Xu, G., Chhangawala, S., Cocco, E., Razavi, P., Cai, Y., Otto, J.E., Ferrando, L., Selenica, P., Ladewig, E., Chan, C., et al. (2020). ARID1A determines luminal identity and therapeutic response in estrogen-recep- tor-positive breast cancer. Nat. Genet. 52, 198-207. https://doi.org/10. 1038/s41588-019-0554-0.

85. Kenneth, N.S., Mudie, S., van Uden, P., and Rocha, S. (2009). SWI/SNF regulates the cellular response to hypoxia. J. Biol. Chem. 284, 4123- 4131. https://doi.org/10.1074/jbc.M808491200.

86. Dreier, M.R., Walia, J., and de la Serna, I.L. (2024). Targeting SWI/SNF Complexes in Cancer: Pharmacological Approaches and Implications. Epigenomes 8, 7. https://doi.org/10.3390/epigenomes8010007.

87. DiNardo, C.D., Savona, M.R., Kishtagari, A., Fathi, A.T., Bhalla, K.N., Agresta, S., Reilly, S., Almon, C., Hentemann, M., Hickman, D., et al. (2023). Preliminary Results from a Phase 1 Dose Escalation Study of FHD-286, a Novel BRG1/BRM (SMARCA4/SMARCA2) Inhibitor, Administered As an Oral Monotherapy in Patients with Advanced Hematologic Malignancies. Blood 142, 4284. https://doi.org/10.1182/ blood-2023-178090.

88. DiNardo, C.D., Fathi, A.T., Kishtagari, A., Bhalla, K.N., Quintás-Cardama, A., Reilly, S.A., Almon, C., Patriquin, C., Nabhan, S., Healy, K., et al. (2025). A Phase 1 Study of FHD-286, a Dual BRG1/BRM (SMARCA4/ SMARCA2) Inhibitor, in Patients With Advanced Myeloid Malignancies. Clin. Cancer Res. 31, 2327-2338. https://doi.org/10.1158/1078-0432. CCR-24-3790.

89. Martinez, G.J., Appleton, M., Kipp, Z.A., Loria, A.S., Min, B., and Hinds, T. D. (2024). Glucocorticoids, their uses, sexual dimorphisms, and dis- eases: new concepts, mechanisms, and discoveries. Physiol. Rev. 104, 473-532. https://doi.org/10.1152/physrev.00021.2023.

90. Acharya, N., Madi, A., Zhang, H., Klapholz, M., Escobar, G., Dulberg, S., Christian, E., Ferreira, M., Dixon, K.O., Fell, G., et al. (2020). Endogenous Glucocorticoid Signaling Regulates CD8+ T Cell Differentiation and Development of Dysfunction in the Tumor Microenvironment. Immunity 53, 658-671.e6. https://doi.org/10.1016/j.immuni.2020.08.005.

91. He, X .- Y., Ng, D., Van Aelst, L., and Egeblad, M. (2019). Stressing Out about Cancer Immunotherapy. Cancer Cell 36, 468-470. https://doi. org/10.1016/j.ccell.2019.10.013.

92. Borgers, J.S.W., Tobin, R.P., Torphy, R.J., Vorwald, V.M., Van Gulick, R. J., Amato, C.M., Cogswell, D.T., Chimed, T .- S., Couts, K.L., Van Bokhoven, A., et al. (2021). Melanoma Metastases to the Adrenal Gland Are Highly Resistant to Immune Checkpoint Inhibitors. J. Natl. Compr. Canc. Netw. 19, jnccn20283. https://doi.org/10.6004/jnccn. 2020.7800.

93. Pegna, G.J., Roper, N., Kaplan, R.N., Bergsland, E., Kiseljak-Vassiliades, K., Habra, M.A., Pommier, Y., and Del Rivero, J. (2021). The Immunotherapy Landscape in Adrenocortical Cancer. Cancers (Basel) 13, 2660. https://doi.org/10.3390/cancers13112660.

94. Araujo-Castro, M., Pascual-Corrales, E., Molina-Cerrillo, J., and Alonso- Gordoa, T. (2021). Immunotherapy in Adrenocortical Carcinoma: Predictors of Response, Efficacy, Safety, and Mechanisms of Resistance. Biomedicines 9, 304. https://doi.org/10.3390/biomed- icines9030304.

95. Landwehr, L .- S., Altieri, B., Schreiner, J., Sbiera, I., Weigand, I., Kroiss, M., Fassnacht, M., and Sbiera, S. (2020). Interplay between glucocorti- coids and tumor-infiltrating lymphocytes on the prognosis of adrenocor- tical carcinoma. J. Immunother. Cancer 8, e000469. https://doi.org/10. 1136/jitc-2019-000469.

96. Wilmouth, J.J., Olabe, J., Garcia-Garcia, D., Lucas, C., Guiton, R., Roucher-Boulez, F., Dufour, D., Damon-Soubeyrand, C., Sahut- Barnola, I., Pointud, J .- C., et al. (2022). Sexually dimorphic activation of innate antitumor immunity prevents adrenocortical carcinoma develop- ment. Sci. Adv. 8, eadd0422. https://doi.org/10.1126/sciadv.add0422.

97. Hodges, H.C. (2021). Code: ‘hodgeslab/workflows’, Version 20210915. Zenodo. https://doi.org/10.5281/zenodo.5511049.

98. Morin, P.J., Sparks, A.B., Korinek, V., Barker, N., Clevers, H., Vogelstein, B., and Kinzler, K.W. (1997). Activation of beta-catenin-Tcf signaling in colon cancer by mutations in beta-catenin or APC. Science 275, 1787- 1790. https://doi.org/10.1126/science.275.5307.1787.

99. Langmead, B., and Salzberg, S.L. (2012). Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357-359. https://doi.org/10.1038/ nmeth. 1923.

100. Zhang, Y., Liu, T., Meyer, C.A., Eeckhoute, J., Johnson, D.S., Bernstein, B.E., Nusbaum, C., Myers, R.M., Brown, M., Li, W., et al. (2008). Model- based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137. https://doi. org/10.1186/gb-2008-9-9-r137.

101. Stark, R., and Brown, G. (2011). DiffBind: differential binding analysis of ChIP-Seq peak data. http://bioconductor.org/packages/release/bioc/ vignettes/DiffBind/inst/doc/DiffBind.pdf.

102. Love, M.I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550. https://doi.org/10.1186/s13059-014-0550-8.

103. Ritchie, M.E., Phipson, B., Wu, D., Hu, Y., Law, C.W., Shi, W., and Smyth, G.K. (2015). limma powers differential expression analyses for RNA- sequencing and microarray studies. Nucleic Acids Res. 43, e47. https://doi.org/10.1093/nar/gkv007.

104. Quinlan, A.R., and Hall, I.M. (2010). BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841-842. https:// doi.org/10.1093/bioinformatics/btq033.

105. Pohl, A., and Beato, M. (2014). bwtool: a tool for bigWig files. Bioinformatics 30, 1618-1619. https://doi.org/10.1093/bioinformatics/ btu056.

106. Bankhead, P., Loughrey, M.B., Fernández, J.A., Dombrowski, Y., McArt, D.G., Dunne, P.D., McQuaid, S., Gray, R.T., Murray, L.J., Coleman, H.G., et al. (2017). QuPath: Open source software for digital pathology image analysis. Sci. Rep. 7, 16878. https://doi.org/10.1038/s41598-017- 17204-5.

CelPress OPEN ACCESS

Molecular Cell Article

107. Colaprico, A., Silva, T.C., Olsen, C., Garofano, L., Cava, C., Garolini, D., Sabedot, T.S., Malta, T.M., Pagnotta, S.M., Castiglioni, I., et al. (2016). TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 44, e71. https://doi.org/10.1093/nar/ gkv1507.

108. Davis, S., and Meltzer, P.S. (2007). GEOquery: A bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics 23, 1846-1847. https://doi.org/10.1093/bioinformatics/btm254.

109. Gu, Z., Eils, R., and Schlesner, M. (2016). Complex heatmaps reveal pat- terns and correlations in multidimensional genomic data. Bioinformatics 32, 2847-2849. https://doi.org/10.1093/bioinformatics/btw313.

110. Heinz, S., Benner, C., Spann, N., Bertolino, E., Lin, Y.C., Laslo, P., Cheng, J.X., Murre, C., Singh, H., and Glass, C.K. (2010). Simple Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for Macrophage and B Cell Identities. Mol. Cell 38, 576-589. https://doi.org/10.1016/j.molcel.2010.05.004.

111. Schep, A.N., Wu, B., Buenrostro, J.D., and Greenleaf, W.J. (2017). ChromVAR: Inferring transcription-factor-associated accessibility from single-cell epigenomic data. Nat. Methods 14, 975-978. https://doi. org/10.1038/nmeth.4401.

112. Venables, W.N., and Ripley, B.D. (2002). Modern Applied Statistics with S, Fourth Edition (Springer). https://doi.org/10.1007/978-0-387-21706-2.

113. Krieger, E., and Vriend, G. (2014). YASARA View-molecular graphics for all devices-from smartphones to workstations. Bioinformatics 30, 2981-2982. https://doi.org/10.1093/bioinformatics/btu426.

114. Sarpe, V., Rafiei, A., Hepburn, M., Ostan, N., Schryvers, A.B., and Schriemer, D.C. (2016). High Sensitivity Crosslink Detection Coupled With Integrative Structure Modeling in the Mass Spec Studio. Mol. Cell. Proteomics 15, 3071-3080. https://doi.org/10.1074/mcp.0116.058685.

115. Lopez-deLisle, L., Rabbani, L., Wolff, J., Bhardwaj, V., Backofen, R., Grüning, B., Ramírez, F., and Manke, T. (2021). pyGenomeTracks: repro- ducible plots for multivariate genomic datasets. Bioinformatics 37, 422-423. https://doi.org/10.1093/bioinformatics/btaa692.

116. Therneau, T.M., and Grambsch, P.M. (2000). Modeling Survival Data: Extending the Cox Model (Springer). https://doi.org/10.1007/978-1- 4757-3294-8.

117. Robin, X., Turck, N., Hainard, A., Tiberti, N., Lisacek, F., Sanchez, J .- C., and Müller, M. (2011). pROC: An open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 12, 77. https:// doi.org/10.1186/1471-2105-12-77.

118. Krijger, P.H.L., Geeven, G., Bianchi, V., Hilvering, C.R.E., and de Laat, W. (2020). 4C-seq from beginning to end: A detailed protocol for sample preparation and data analysis. Methods 170, 17-32. https://doi.org/10. 1016/j.ymeth.2019.07.014.

119. Sanjana, N.E., Shalem, O., and Zhang, F. (2014). Improved vectors and genome-wide libraries for CRISPR screening. Nat. Methods 11, 783-784. https://doi.org/10.1038/nmeth.3047.

120. Kim, D., Paggi, J.M., Park, C., Bennett, C., and Salzberg, S.L. (2019). Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 37, 907-915. https://doi.org/10. 1038/s41587-019-0201-4.

121. Anders, S., Pyl, P.T., and Huber, W. (2015). HTSeq-a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166-169. https://doi.org/10.1093/bioinformatics/btu638.

122. Korotkevich, G., Sukhov, V., Budin, N., Shpak, B., Artyomov, M.N., and Sergushichev, A. (2019). Fast gene set enrichment analysis. Preprint at bioRxiv. https://doi.org/10.1101/060012.

123. Schoenfelder, S., Javierre, B .- M., Furlan-Magaril, M., Wingett, S.W., and Fraser, P. (2018). Promoter Capture Hi-C: High-resolution, Genome-wide Profiling of Promoter Interactions. J. Vis. Exp. 136, 57320. https://doi. org/10.3791/57320.

124. Montefiori, L.E., Sobreira, D.R., Sakabe, N.J., Aneas, I., Joslin, A.C., Hansen, G.T., Bozek, G., Moskowitz, I.P., McNally, E.M., and Nóbrega, M.A. (2018). A promoter interaction map for cardiovascular disease ge- netics. eLife 7, e35788. https://doi.org/10.7554/eLife.35788.

125. Muzzi, J.C.D., Magno, J.M., Cardoso, M.A., de Moura, J., Castro, M.A.A., and Figueiredo, B.C. (2021). Adrenocortical Carcinoma Steroid Profiles: In Silico Pan-Cancer Analysis of TCGA Data Uncovers Immunotherapy Targets for Potential Improved Outcomes. Front. Endocrinol. (Lausanne) 12, 672319. https://doi.org/10.3389/fendo.2021.672319.

126. Hatazawa, S., Liu, J., Takizawa, Y., Zandian, M., Negishi, L., Kutateladze, T.G., and Kurumizaka, H. (2022). Structural basis for binding diversity of acetyltransferase p300 to the nucleosome. iScience 25, 104563. https:// doi.org/10.1016/j.isci.2022.104563.

127. Hornbeck, P.V., Zhang, B., Murray, B., Kornhauser, J.M., Latham, V., and Skrzypek, E. (2015). PhosphoSitePlus, 2014: mutations, PTMs and reca- librations. Nucleic Acids Res. 43, D512-D520. https://doi.org/10.1093/ nar/gku1267.

CellPress OPEN ACCESS

Molecular Cell Article

STAR*METHODS

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
SMARCA4 (IP, WB)	Santa Cruz Biotechnology	Cat# sc-374197; RRID:AB_10990135
SMARCA4 (ChIP, WB)	Proteintech	Cat# 21634-1-AP; RRID:AB_10858784
SMARCA4 (IHC)	Abcam	Cat# ab108318; RRID:AB_10889900
SMARCA2 (WB)	Abcam	Cat# ab15597; RRID:AB_443214
B-catenin (ChIP, WB)	Invitrogen	Cat# 71-2700; RRID:AB_88035
ß-catenin (IHC)	Cell Signaling Technology	Cat# 8480; RRID:AB_11127855
SF-1 (ChIP, IP)	R&D Systems	Cat# PPN16650C
SF-1 (WB)	Millipore	Cat# 07-618; RRID:AB_310756
SF-1 (IHC)	Thermo Fisher	Cat# 2516-MSM7-P1
IgG (IP)	Santa Cruz Biotechnology	Cat# sc-2025; RRID:AB_737182
FLAG (co-IP, WB)	Millipore	Cat# F1804; RRID:AB_262044
ARID1A (ChIP, WB)	Cell Signaling Technology	Cat# 12354S; RRID:AB_2637010
ARID1A (IP)	Santa Cruz Biotechnology	Cat# sc-32761; RRID:AB_673396
ARID1B (WB)	Santa Cruz Biotechnology	Cat# sc-32762; RRID:AB_2060367
DPF1 (WB)	Proteintech	Cat# 21769-1-AP; RRID:AB_3085672
DPF2 (WB)	Proteintech	Cat# 12111-1-AP; RRID:AB_2095574
DPF3 (WB)	Proteintech	Cat# 68646-1-Ig; RRID:AB_3670389
PBRM1 (ChIP, WB)	Bethyl Laboratories	Cat# A301-591A; RRID:AB_1078808
BRD9 (ChIP, WB)	Abcam	Cat# ab137245
H3K4me1 (ChIP)	Abcam	Cat# ab8895; RRID:AB_306847
H3K4me3 (ChIP)	Millipore	Cat# 05-745R; RRID:AB_1587134
H3K27ac (ChIP)	Abcam	Cat# ab4729; RRID:AB_2118291
H3K27me3 (ChIP)	Active Motif	Cat# 39155; RRID:AB_2561020
CTCF (ChIP)	Millipore	Cat# 07-729; RRID:AB_441965
FOXO3 (WB)	Cell Signaling Technology	Cat# 12829; RRID:AB_2636990
SOX2 (WB)	Millipore	Cat# AB5603; RRID:AB_2286686
YAP1 (WB)	Proteintech	Cat# 13584-1-AP; RRID:AB_2218915
CBP/p300 (WB)	Proteintech	Cat# 22277-1-AP; RRID:AB_3085692
CBP/p300 (ChIP)	Abcam	Cat# ab14984; RRID:AB_301550
c-Jun (WB)	Cell Signaling Technology	Cat# 9165; RRID:AB_2130165
Vinculin (WB)	Sigma-Aldrich	Cat# V9131; RRID:AB_477629
ß-actin (WB)	GeneTex	Cat# GTX629630; RRID:AB_2728646
HRP conjugated mouse anti-rabbit	Santa Cruz Biotechnology	Cat# sc-2357; RRID:AB_628497
HRP conjugated anti-mouse	Santa Cruz Biotechnology	Cat# sc-516102; RRID:AB_2687626
Chemicals, peptides, and recombinant proteins
BRM014	This paper	N/A
FHD-286	MedChemExpress	Cat# HY-144835
AU-15330	MedChemExpress	Cat# HY-145388
cOmplete ULTRA	Roche	Cat# 5892791001
NuPAGE 4-12% bis tris polyacrylamide gel	Thermo Fisher	Cat# NP0323BOX
NuPAGE 3-8% Tris-Acetate Protein Gels	Thermo Fisher	Cat# EA03785BOX
MOPS SDS running buffer	Thermo Fisher	Cat# NP0001
Novex NuPAGE Tris-Acetate SDS running buffer	Thermo Fisher	Cat# LA0041

(Continued on next page)

CelPress OPEN ACCESS

Molecular Cell Article

Continued
REAGENT or RESOURCE	SOURCE	IDENTIFIER
PVDF membrane	Millipore	Cat# IPFL00010
BLOTTO	Santa Cruz Biotechnology	Cat# SC2325
Clarity Western ECL	Bio-Rad	Cat# 1705061
Clarity Max ECL	Bio-Rad	Cat# 1705062
Protein G Dynabeads	Invitrogen	Cat# 10003D
DpnII	New England Biolabs	Cat# R0543S
T4 DNA ligase	New England Biolabs	Cat# M0202
Csp6I	Thermo Scientific	Cat# ER0211
NheI-HF	New England Biolabs	Cat# R3131S
AgeI-HF	New England Biolabs	Cat# R3552S
MluI-HF	New England Biolabs	Cat# R3198S
Lipofectamine 3000	Thermo Scientific	Cat# L3000001
Polyethylenimine hydrochloride	Polysciences	Cat# 247651
TRIzol	Thermo Scientific	Cat# 15596026
IPTG	Merck	Cat# I6758
Glutathione affinity resin	Merck	Cat# GE17-5132-02
Amylose resin	New England BioLabs	Cat# E8021L
Trypsin/LysC	Promega	Cat# V5073
Superdex 200 Increase 10/300 GL	Merck	Cat# GE28-9909-44
BS3 cross-linker	Thermo Scientific	Cat# 21580
C18 PepMap Neo Trap Cartridge	Thermo Scientific	Cat# 174500
Aurora Ultimate Nano reversed phase column	Ion Opticks	Cat# AUR3-25075C18
DMEM/F12	ATCC	Cat# 30-2006
Nu-serum	Corning	Cat# 355500
ITS	Corning	Cat# 354352
Penicillin/streptomycin	Gibco	Cat# 15140122
Hydrocortisone	Sigma-Aldrich	Cat# H0888
Insulin	Sigma-Aldrich	Cat# I0516
Cholera toxin	Sigma-Aldrich	Cat# C8052
Epidermal growth factor	Gibco	Cat# PHG0311L
Adenine	Sigma-Aldrich	Cat# A2786
Critical commercial assays
Testosterone ELISA	Enzo Life Sciences	Cat# ADI-900-065
DHEA ELISA	Enzo Life Sciences	Cat# ADI-900-093
Cortisol ELISA	Enzo Life Sciences	Cat# ADI-900-071
Aldosterone ELISA	Enzo Life Sciences	Cat# ADI-900-173
CYP11A1 TaqMan probe	Thermo Fisher	Hs00167984_m1
CYP17A1 TaqMan probe	Thermo Fisher	Hs01124136_m1
SCARB1 TaqMan probe	Thermo Fisher	Hs00969821_m1
STAR TaqMan probe	Thermo Fisher	Hs00264912_m1
MC2R TaqMan probe	Thermo Fisher	Hs01563483_s1
NR5A1 TaqMan probe	Thermo Fisher	Hs00610436_m1
TBP TaqMan probe	Thermo Fisher	Hs00427620_m1
TaqMan Fast Advanced Master Mix	Applied Biosystems	Cat# 4444557
Pierce Detergent Compatible Bradford Assay Kit	Thermo Fisher	Cat# 23246
Illumina Tagment DNA Enzyme and Buffer Large Kit	Illumina	Cat# 20034198
High-Fidelity 2X PCR Master Mix	New England Biolabs	Cat# M0541S
AMPure XP beads	Beckman Coulter	Cat# A63880

(Continued on next page)

Molecular Cell Article

CelPress OPEN ACCESS

Continued
REAGENT or RESOURCE	SOURCE	IDENTIFIER
PCR beads	Nucleomag	Cat# 744100.1
MinElute PCR purification kit	Qiagen	Cat# 28004
In-Fusion	Takara Bio	Cat# 639649
Clean-Blot IP Detection Reagent	Thermo Fisher	Cat# 21230
Deposited data
ATAC-seq data	This study	GEO: GSE260866
RNA-seq data	This study	GEO: GSE261006
ChIP-seq data	This study	GEO: GSE260867
Raw western blot data	This study	Mendeley Data: https://doi.org/10.17632/2xm4vsv6w8.1
XL-MS data	This study	PRIDE: PXD063603
Experimental models: Cell lines
H295R	ATCC	Cat# CRL-2182, RRID:CVCL_0458
CU-ACC1	Kiseljak-Vassiliades et al.58	RRID:CVCL_RQ00
HEK 293T	ATCC	Cat# CRL-3216, RRID:CVCL_0063
Oligonucleotides
SMARCA4 shRNA #1	Dharmacon	Cat# TRCN0000015549
SMARCA4 shRNA #2	Dharmacon	Cat# TRCN0000015550
SMARCA2 shRNA #1	Dharmacon	Cat# TRCN20330
SMARCA2 shRNA #2	Dharmacon	Cat# TRCN20332
SF-1 shRNA #1	Dharmacon	Cat# TRCN0000019599
SF-1 shRNA #2	Dharmacon	Cat# TRCN0000019600
B-catenin shRNA #1	Rosenbluh et al. 45	Addgene plasmid #42543
-catenin shRNA #2	Rosenbluh et al.45	Addgene plasmid #42544
Non-targeting control	Dharmacon	Cat# RHS6848
Recombinant DNA
piggyBac transposase	C. Bashor	N/A
SF-1 ORF	Genscript	Cat# OHu26165
ARID1A ORF	NCBI	GenBank: NM_006015
B-catenin ORF	Morin et al.98	Addgene plasmid #16518; RRID:Addgene_16518
SF-1 ALBD	This study	N/A
ARID1A deletion constructs	This study	N/A
B-catenin deletion constructs	This study	N/A
Software and algorithms
Bowtie 2.4.1	Langmead and Salzberg9	RRID:SCR_005476
MACS 2.1.1	Zhang et al. 100	RRID:SCR_013291
Diffbind	Stark and Brown 10	RRID:SCR_012918
DESeq2	Love et al. 102	RRID:SCR_015687
Limma	Ritchie et al. 103	RRID:SCR_010943
Bedtools 2.28.0	Quinlan and Hall104	RRID:SCR_006646
bwtool	Pohl and Beato 105	RRID:SCR_003035
QuPath 0.2.3	Bankhead et al. 106	RRID:SCR_018257
TCGAbiolinks	Colaprico et al. 107	RRID:SCR_017683
GEOquery	Davis and Meltzer 108	RRID:SCR_000146
ComplexHeatmap	Gu et al. 109	RRID:SCR_017270
HOMER 4.11	Heinz et al. 110	RRID:SCR_010881

(Continued on next page)

CelPress OPEN ACCESS

Molecular Cell Article

Continued
REAGENT or RESOURCE	SOURCE	IDENTIFIER
ChromVAR	Schep et al. 111	RRID:SCR_026570
nnet	Venables and Ripley	https://cran.r-project.org/web/ packages/nnet/index.html
YASARA 24.4.10	Krieger and Vriend113	RRID:SCR_017591
Mass Spec Studio	Sarpe et al. 114	PMID: 27412762
pyGenomeTracks	Lopez-Delisle et al. 115	RRID:SCR_025312
survival	Therneau et al. 116	RRID:SCR_021137
survminer	Kassambara, Kosinski, Biecek	RRID:SCR_021094
PROC	Robin et al. 117	RRID:SCR_024286
pipe4C	Krijger et al. 118	https://doi.org/10.1016/j.ymeth.2019.07.014

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell lines

H295R cells were acquired (ATCC CRL-2182) and maintained in DMEM/F12 medium (ATCC 30-2006) supplemented with 2.5% of Nu-serum (Corning #355500), 1% ITS (Corning #354352) and 1% penicillin/streptomycin (Gibco #15140122). CU-ACC1 cells58 were maintained in F medium [3:1 (v/v) F12:DMEM], 5% FBS, 0.4 µg/mL hydrocortisone (Sigma-Aldrich H0888), 5 µg/mL insulin (Sigma- Aldrich #10516), 8.4 ng/ml cholera toxin (Sigma-Aldrich C8052), 10 ng/mL epidermal growth factor (Gibco PHG0311L), 24 µg/mL adenine (Sigma-Aldrich A2786) and 1% Pen/Strep as previously reported.58 HEK 293T cells were acquired (ATCC CRL-3216) and maintained in DMEM supplemented with 5% FBS.

METHOD DETAILS

Cell-line generation

CRISPR-Cas9 gene knockout

HEK 293T cells were transfected with lentiCRISPR v2 plasmids encoding mCherry, Cas9 protein and specific single guide RNAs (sgRNAs) selected from human GeCKO libraries.119 The following gRNA sequences were employed for gene editing in this study: ARID1A: AAT ACT CAC AGG CAA GCT GG; ARID1B: GTC CGA CCC TGG ATG CCA AT; PBRM1: GAA ACC ACT TCA TAA TAG TC; BRD9: CTT GAC GGA CAG TAC CGC AG; DPF1: GTC CGA CCC TGG ATG CCA AT; DPF2: TGG ATG GAA AAG CGA CAC CG; DPF3: CAA GTA GGC ACT CAC GCC TG. For CTNNB1 knockout, the following two gRNAs were used: GAA AAG CGG CTG TTA GTC AC and TTC CCA CTC ATA CAG GAC TT. All knockout experiments included non-targeting control (NTC) gRNA: GGG AGG TGG CTT TAG GTT TT. Following transfection, mCherry-positive cells were sorted into individual wells of a 96-well plate to obtain monoclonal cell lines. Knockout of targets was confirmed by western blot.

shRNA knockdown

To achieve stable knockdown of SMARCA4, SMARCA2, SF-1, or ß-catenin, H295R cells were transduced with lentivirus expressing pLKO.1 shRNAs. Transduced cells were selected using 3 µg/mL puromycin. All shRNAs used in this study are provided in key resources table.

piggyBac transposition

An acceptor plasmid for piggyBac transposition was constructed for ectopic protein expression. ARID1A (NCBI NM_006015) or SF-1 (Genscript OHu26165) were cloned into this plasmid with C-terminal FLAG tag. SF-1 lacking the ligand-binding domain (ALBD) was created by In-Fusion (Takara Bio #639649). Full-length ß-catenin was obtained from pCI-neo, a gift from Bert Vogelstein98 (Addgene plasmid #16518; http://n2t.net/addgene:16518; RRID:Addgene_16518), and cloned into the piggyBac expression plasmid with an HA tag through ligation. All truncation and deletion constructs of ARID1A and B-catenin were synthesized with boundaries denoted in main figures. Vectors were validated using Sanger and/or whole-plasmid sequencing.

Initial piggyBac transposition was performed using Lipofectamine 3000 (Thermo Fisher L3000001). Cells were selected using 10 µg/mL blasticidin. Stable integration and protein expression were confirmed by western blot.

Lentiviral transduction

HEK 293T cells were co-transfected with packaging plasmids (psPAX2, pMD2.G) and transfer plasmid using polyethylenimine hy- drochloride (PEI, Polysciences #247651, MW 40,000). Culture medium was changed to DMEM with 2% FBS 16 h after transfection. Lentiviral particle-containing culture medium was harvested 72 h after transfection. The medium was filtered through a 5-um filter (Millipore SLSV025LS), concentrated by ultrafiltration on a 100-kDa Amicon filter (Millipore UFC903008), and stored at -80℃. Cells were transduced at 50% confluency using spinfection (1,000 g, 30 min, 32℃).

Molecular Cell Article

CelPress OPEN ACCESS

Inhibitors and degraders

Sourcing and synthesis

FHD-286 (MedChemExpress HY-144835) and AU-15330 (MedChemExpress HY-145388) were purchased commercially. BRM014 was synthesized as described earlier. 1,2,59 For BRM014, the expected molecular weight of 318 Da was confirmed, and pu- rity of >99% was confirmed based on LC/MS measurement.

Treatment conditions

Unless stated otherwise, 500,000 cells were seeded per well in 6-well plates. After attachment, cells were treated with 1 µM BRM014 or equal volume of DMSO as vehicle control. Cells were then incubated for up to 72 hours. 1 µM FHD-286 or 1 µM AU-15330 were added in the same manner to evaluate their effects on SF-1 activity.

RT-qPCR

RNA was harvested using TRIzol reagent (Invitrogen #15596026). 1 µg of RNA was then reverse-transcribed into cDNA using a high- capacity reverse transcription kit (Fisher 43-688-14). TaqMan qPCR reactions contained 1x TaqMan Fast Advanced Master Mix (Applied Biosystems #4444557), 1x target primers (FAM-MGB), 1x housekeeping primers (VIC-MGB), and 50 ng of cDNA in a volume of 20 ul. TaqMan qPCR assays used in this study are provided in key resources table. PCR was performed with a QuantStudio 3 Real- Time PCR system (Applied Biosystems A28567). Quantification was internally normalized to TBP expression levels using the AACt method and control comparisons were made simultaneously using equivalently passaged cells.

Western blotting Immunoblotting

Total protein was isolated from cells using RIPA lysis buffer (50 mM Tris HCl pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% sodium deox- ycholate, 0.1% SDS, complete ULTRA protease inhibitors (Roche #5892791001), sonicated, and cleared by centrifugation. Protein concentration was determined using Bradford assay (Thermo Fisher #23246). Subsequently, 20 µg of total protein per well was run on a Novex NuPAGE 4-12% bis tris polyacrylamide gel (Thermo Fisher NP0323BOX) in MOPS SDS running buffer (Thermo Fisher #NP0001); or Nupage 3-8% Tris-Acetate Protein Gels (Thermo Fisher EA03785BOX) in Novex NuPAGE Tris-Acetate SDS running buffer (Thermo Fisher LA0041). Gel content was transferred to 0.45-um PVDF membrane (Millipore IPFL00010) overnight. Detection

Western blots were blocked in 5% BLOTTO (Santa Cruz Biotechnology SC2325) dissolved in Tris-buffered saline with 0.1% Tween 20 Detergent (TBS-T), incubated in primary antibody overnight at 4℃, washed 3 times in TBS-T, and probed with secondary antibody for 1-2 hours at room temperature (RT). Antibodies are provided in key resources table. Clean-Blot IP Detection Reagent (Thermo Fisher #21230, 1:1,000 dilution) was used for detecting co-IP samples to avoid contamination from denatured antibodies. Detection of HRP-conjugated antibodies was performed using Clarity Western or Clarity Max ECL substrates (Bio-Rad). Visualization was per- formed on a Bio-Rad ChemiDoc MP imager with densitometry analysis performed using Image Lab software (6.1).

Co-immunoprecipitation

To assess protein interactions via co-immunoprecipitation (coIP), H295R or HEK 293T cells were first lysed with IP buffer (25 mM Tris HCI PH 8.0, 150 mM NaCI, 1% NP-40, 1 mM EDTA) for 10 minutes on ice. Lysed samples were sonicated for 6 cycles of 10 seconds on and 30 seconds off at 4℃ and cleared by centrifugation. 500 µg of whole-cell lysate was pre-incubated with protein G dynabeads (Invitrogen 10003D) for 30 minutes and removed using a magnetic stand. Lysates were then incubated with 5 µg antibodies over- night. The targeted proteins and antibodies were incubated with protein G beads for 1 hour followed by magnetic separation. Protein was eluted by adding 1x sample buffer (Thermo Fisher NP0007) and boiled at 95℃ for 5 minutes. Antibodies are provided in key resources table. Before running western blot, eluted samples were supplemented with 2.5% ß-mercaptoethanol and incubated at 95℃ for 5 minutes, except for SOX2, where it was omitted to avoid signal overlap with antibodies.

Chromatin immunoprecipitation and sequencing

Library preparation

Chromatin immunoprecipitation (ChIP) was performed as previously described.3 Briefly, 107 H295R or 8x106 HEK 293T cells were seeded into 15-cm dishes. After attachment, cells were treated with 1 µM BRM014 or equal volume of DMSO for 1 hour. ChIP libraries were prepared from a single-cell suspension fixed for 10 minutes in 1% formaldehyde (histone markers, CTCF, SF-1), or dual-fixed for 30 minutes in 2 mM disuccinimidyl glutarate followed by a 10-minute incubation in 1% formaldehyde (BAF subunits, ß-catenin, CBP/p300). Excess formaldehyde was quenched by the addition of glycine to a final concentration of 125 mM. Fixed cells were washed, pelleted, and snap-frozen using liquid nitrogen. Antibodies used are provided in key resources table.

Analysis

Sequencing reads were mapped to the hg38 human reference genome using Bowtie (2.4.1).99 Duplicate fragments and reads with mapping quality <10 were discarded. Peak calling was performed by MACS (2.1.1).100 Diffbind101 and DESeq2102 were used for dif- ferential peak calling and for estimating size factor. Differential peak calls were made by requiring fold changes of >2 fold in either direction and FDR-adjusted p-values <0.05. Overlap of peaks was assessed using bedtools (2.28.0).104 Calculation of mean densities

CelPress OPEN ACCESS

Molecular Cell Article

and preparation of genome-wide heatmaps was performed using bwtool.105 Genome annotation was done with ChromHMM63 using genome-wide profiles of CTCF, H3K4me1, H3K4me3, H3K27ac, H3K4me3, and H3K27me3.

RNA sequencing

Library preparation

500,000 H295R or 300,000 CU-ACC1 cells were seeded per well in 6-well plates. After attachment, cells were treated with 1 µM BRM014 or equal volume of DMSO for 72 hours. Cells were harvested using TRIzol reagent. mRNA was isolated by polyA- enrichment, fragmented, and reverse transcribed into cDNA. Paired-end sequencing was performed on an Illumina NovaSeq 6000.

RNA-seq data analysis

As previously described,3 reads were aligned to the hg38 human reference genome utilizing HISAT2.120 Reads with mapping quality below 10 were excluded from further analysis. Reads mapping to specific genes were quantified using HTSeq.121 Differential gene expression analysis was processed with DESeq2,102 using default parameters. Genes with expression fold changes exceeding 1.5- fold in either direction, coupled with FDR-adjusted p values less than 0.05, were deemed differentially expressed.

Gene set enrichment analysis

Gene set enrichment analysis was performed using the R package fgsea (1.27.0).122 The following gene sets were used to generate enrichment plots: KEGG_STEROID_BIOSYNTHESIS, WEST_ADRENOCORTICAL_TUMOR_UP, PID_BETA_CATENIN NUC_PATHWAY, and SF-1 targets (derived from published annotations.52)

ATAC sequencing

Library preparation

500,000 H295R or HEK 293T cells were seeded per well in 6-well plates. After attachment, cells were treated with 1 µM BRM014 or an equal volume of DMSO for 1 hour. As previously described,1,2 nuclei were isolated by resuspending cells in 50 ul lysis buffer (comprising 0.1% Tween-20, 0.1% NP-40, 0.01% digitonin, and RSB buffer with 10 mM Tris-HCl pH 7.5, 10 mM NaCl, 3 mM MgCl2) and incubated on ice for 3 min. Cells were then washed with RSB buffer supplemented with 0.1% Tween-20 to remove re- sidual lysis buffer. Nuclei were centrifuged at 500 g for 10 min to pellet.

For transposition, nuclear pellets were resuspended in 50 ul of Transposition mix (2.5 ul Tagment DNA enzyme (Illumina #20034198), 25 ul of Tagment DNA buffer (Illumina #20034198), 0.1% Tween-20 and 0.01% Digitonin) and incubated at 37℃ for 30 min. Post-incubation, transposed DNA was purified using the MinElute PCR purification kit (Qiagen #28004). Libraries were ampli- fied using barcoded Nextera primers (Illumina) and 2x NEBNext High-Fidelity PCR Master Mix (NEB M0541S). Finally, libraries were purified and size-selected with AMPure XP beads (Beckman Coulter A63880), selecting for fragment sizes ranging from ~100 to 1,000 bp.

Analysis

As described previously,1,2 ATAC sequencing (ATAC-seq) reads were mapped to the hg38 human reference genome using Bowtie (2.4.1).99 Duplicate fragments and reads with mapping quality <10 were discarded. Peak calling was performed by MACS (2.1.1). 100 DESeq2 was used for differential peak calling, with size factors determined using the top 10% of accessible sites. Differential peak calls were made by requiring fold changes of >1.5-fold in either direction and FDR-adjusted p values <0.10. TF motif enrichment and significance were measured using HOMER (4.11)110 and ChromVAR. 111

4C library preparation, sequencing, and analysis

Identification of enhancers for NR5A1 and CTNNB1

We used 4C to experimentally identify enhancers that interact with the NR5A1 promoter. Enhancers for CTNNB1 were identified by the overlap of ChIP-seq histone modification data with adrenal promoter capture Hi-C (pcHi-C), as previously described. 48,123

Cross-linking and nuclei isolation

As previously described, 118 ~10 million H295R cells were cross-linked using 1% formaldehyde at RT for 10 minutes. The reaction was quenched with 0.13M glycine. Nuclei were then isolated using a lysis buffer composed of 50 mM Tris-HCl (pH 7.5), 0.5% NP-40, 1% TX-100, 150 mM NaCl, 5 mM EDTA, and 1x protease inhibitors. Isolated nuclei were resuspended in 1.2x NEBuffer sup- plemented with 0.3% SDS and 2.5% Triton.

DNA digestion and ligation

Resuspended nuclei were digested with 100 U of Dpnll (NEB R0543S) at 37 ℃ for 3 hours, followed by an additional overnight in- cubation with another 100 U of Dpnll. Dpnll was inactivated by incubating at 65℃ for 20 minutes. The digested samples were then ligated in 7 mL of ligation buffer using 3333 U of T4 DNA ligase (NEB M0202) at 16℃ overnight. De-crosslinking was achieved by adding 30 ul of Proteinase K (10 mg/ml) and incubating overnight at 65℃. DNA was purified using PCR beads (Nucleomag #744100.1). A subsequent digestion step was performed using 50 U of Csp61 (Thermo Scientific ER0211) at 37℃ overnight. For the second ligation, 25 µg of template DNA was mixed with 500 ul of 10x ligation buffer (660 mM Tris-HCI, 50 mM MgCl2, 50 mM DTT, 10 mM ATP), 3333 U of T4 Ligase, and adjusted to a total volume of 5 ml with milli-Q water. The mixture was then incubated overnight at 16℃ and the 4C templates were purified using PCR beads.

CelPress OPEN ACCESS

Molecular Cell Article

PCR amplification

Inverse PCR was used to amplify fragments ligated to the viewpoint. Primers were specifically designed using 4C-primer designer. 124 One reading primer and two non-reading primers were designed as follows: reading primer: 5’-TAC ACG ACG CTC TTC CGA TCT TCC CAC GGT GAC TGG ATC-3’; Non-reading primer 1: 5’-ACT GGA GTT CAG ACG TGT GCT CTT CCG ATC TTG TGG TGG GGA GCT GTT CA-3’; Non-reading primer 2: 5’-ACT GGA GTT CAG ACG TGT GCT CTT CCG ATC TGG AGC TGT TCA GAG GCG G-3’. For each non-reading primer, four 50-ul PCR reactions were set up using the Expand Long Template PCR system (Roche #11681834001), with each containing 200 ng of 4C template. The PCR program settings were: 94℃ for 2 min, 16 cycles of 94℃ for 15 s, 55℃ for 1 min, 68℃ for 3 min, followed by a final elongation at 68℃ for 5 min. The 4 PCR reactions were pooled together, and the DNA was purified with AMPure XP beads (Beckman Coulter A63880). A second round of PCR was conducted using a uni- versal primer and index primers for Illumina sequencing, following the PCR conditions: 94℃ for 2 min, 20 cycles of 94℃ for 10 s, 60℃ for 1 min, 68℃ for 3 min, and a final step at 68℃ for 5 min.

Sequencing

Final libraries were sequenced on the Illumina MiSeq platform using the MiSeq Reagent Kit v3 (Illumina MS-102-3001) to obtain 75 bp single-end reads.

Analysis

Sequencing reads from the two non-reading primers were combined. The combined fastq files were then processed by pipe4C118 and mapped to hg38 for peak calling. Genome browser tracks were generated using pyGenomeTracks. 115

Hormone profiling

Conditioned medium

500,000 H295R or 300,000 CU-ACC1 cells were seeded into each well of a 6-well plate. After overnight seeding, cells were treated with 1 µM of BRM014 or equal volume of DMSO. Conditioned media was collected after 72 hours of BRM014 or DMSO treatment. ELISA

Testosterone, DHEA, cortisol, and aldosterone were measured from conditioned media by the following ELISA kits from Enzo Life Sciences: ADI-900-065, ADI-900-093, ADI-900-071, and ADI-900-173 according to the manufacturer’s instructions.

Liquid chromatography/mass spectrometry

Deoxycorticosterone, corticosterone, and 11-deoxycortisol were measured via LC/MS. Steroids from conditioned media were ex- tracted using organic and aqueous solvents [water: methanol: acetonitrile (1:1.5:2.25)] with isotopically labeled internal standard. Following extraction from conditioned media, samples were vortexed for 30 s and allowed to settle for 5 min at RT. Methyl tert- butyl ether (MTBE) was added, and the upper layer was transferred into a glass vial. The samples were dried at a low boiling point using a speed vacuum and reconstituted with 100 µL of methanol: water (50/50 v/v). Separation and analysis were performed using the Agilent 1290 Infinity Liquid Chromatography (LC) system integrated with a 6495 Triple Quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA). Separation was achieved with a LC column (RESTEK Pinnacle DB Biphenyl 1.9 pm [50 x 2.1 mm]) and mobile phase A and B consisting of 0.25 mM ammonium fluoride in water and 0.25 mM ammonium fluoride in methanol respectively. Data were acquired via multiple reaction monitoring (MRM) with positive electrospray ionization (ESI) mode using a 6495 Triple Quadrupole mass spectrometry through Agilent Mass Hunter Software (Agilent Technologies). Identified peaks and retention time were manually reviewed. The relative peak area was log2 transformed, followed by internal standard normal- ization for each method.

Protein purification

Recombinant expression plasmids were synthesized, then transformed into BL21(DE3) E. coli, and cultured in LB medium supple- mented with ampicillin at 37°C. Protein expression was induced with 1 mM IPTG (Merck 16758-10G) for GST-tagged constructs and MBP, and with 0.25 mM IPTG for MBP-ARID1A(IDR2). GST-tagged proteins and MBP were expressed for 4 h at 37℃, while MBP- ARID1A(IDR2) was expressed overnight at 18℃. Cells were harvested by centrifugation and lysed by sonication (30 s on, 2 min off per cycle, 16 cycles, on ice) in lysis buffer containing 20 mM HEPES pH 8.0, 500 mM NaCl, and 1 mM DTT. Clarified lysates were applied to glutathione affinity resin (Merck GE17-5132-02) or amylose resin (New England Biolabs E8021L) and purified using manufac- turer’s instructions. Eluted fractions were dialyzed into gel filtration buffer (20 mM HEPES pH 8.0, 150 mM NaCl, 1 mM TCEP) and further purified on a size-exclusion column (Merck GE28-9909-44).

Cross-linking mass spectrometry

Sample preparation

The following pairs of protein constructs were mixed in equimolar ratios: GST-B-catenin(Arm) with MBP-ARID1A(IDR2), GST-ß-cat- enin(Arm) 44-6 with MBP-ARID1A(IDR2), GST with MBP-ARID1A(IDR2), and MBP with GST. Following mixing in equimolar ratios, 5 µg per replicate (n=2) of each complex was treated with BS3 (bis(sulfosuccinimidyl)suberate) cross-linker (ThermoFisher #21580) in 2x molar excess over lysines present in each sample. BS3 cross-linker was dissolved in water and added in two incre- ments with each reaction taking 30 minutes at RT, followed by quenching by addition of 1 M Tris pH 8.8 to 50-fold molar excess over cross-linker. The resulting mixture was further incubated for 20 min at RT.

CelPress OPEN ACCESS

In-solution digestion

TEAB buffer (0.1 M triethyl ammonium bicarbonate, pH 8.5) in final concentration 85 mM was added to each sample to adjust pH followed by addition of sodium deoxycholate (2% final concentration). TCEP was dissolved in 100 mM TEAB and added in final con- centration 10 mM. After addition of chloracetamide (25 mM final concentration), samples were incubated 30 min at 37℃ in the dark. Samples were digested with Trypsin/LysC (Promega V5073, 1:10 w/w ratio to protein) overnight at 37°C. After incubation, samples were acidified with TFA to pH 2 and the precipitated sodium deoxycholate was removed by ethyl acetate extraction. Samples were dried on speed-vac and desalted by using C18 stage tips. All samples were dried again and dissolved in 25 ul of 0.1% TFA.

Liquid chromatography/mass spectrometry

Peptides were analyzed using an UltiMate 3000 RSLC nano HPLC system integrated with an Orbitrap Fusion Lumos Tribrid mass spectrometer (Thermo Fisher Scentific). Samples were loaded onto a trap column (Thermo Scientific #174500) for 2.000 min at 25.000 µl/min. Loading buffer was composed of water, 2% acetonitrile and 0.1% trifluoroacetic acid. Nano reversed phase column (Ion Opticks AUR3-25075C18) was used for LC/MS analysis. Peptides were eluted by mobile phase A (0.1 % formic acid in water, v/v) and mobile phase B (0.1 % formic acid in ACN, v/v) over 140-min (for IDR2- and Armadillo-containing proteins), or 60-min (for MBP- and GST-only control samples) gradient profiles.

Eluted peptides were immediately ionized (1600 V, 275℃) and scanned using the Orbitrap analyzer between 350-1400 m/z at a resolution (R) of 120,000, normalized AGC target 250% with automatic maximum injection time mode, normalized AGC target set to 500%, R=30,000 and dynamic maximum injection time mode. Data-dependent acquisition mode was used. Ions were isolated in a 1.6-Th window and fragmented by higher-energy collisional dissociation (HDC) with a normalized collision energy of 30%. Cycle time was set to either 1.5 s (60-min gradient) or 3 s (140-min gradient). Precursor ions with charge states ranging from +2 to +8 were selected for fragmentation. Dynamic exclusion was enabled after 2 scans (140-min gradient) or 3 scans (60-min gradient) with exclu- sion duration of 15 s and an exclusion window of 10 ppm.

Data analysis

Mass Spec Studio (2.4.0.35615, Trajan Scientific and Medical)114 was used to identify cross-linked peptides. MS mass tolerance was set to 5 ppm, MS2 mass tolerance 10 ppm, and results were filtered for cross-linked peptide level with FDR-corrected p < 0.05. Cysteine carbamidomethylation was used as a static modification and methionine oxidation was employed as a variable modifica- tion. Peptides with a minimum length of 5 amino acids and maximum length of 40 were included in the search. Trypsin K/R was selected as enzyme. The first linkable site for BS3 cross-linker was defined as the amino group of lysine and the N-terminal amine group whereas the second linkable site included lysine, serine, threonine, and tyrosine. All acquired cross-links were validated manu- ally. Cross-links from both replicates were merged and filtered for peptides occurring outside the GST or MBP tags. Protein se- quences of the recombinant constructs were used to identify cross-linked peptides. Residue positions reported in Table S3 were converted to the corresponding positions in the canonical UniProt sequences for human ARID1A (O14497) and CTNNB1 (P35222).

Analysis of patient samples

Tissue microarray immunohistochemistry

Human tumor samples and de-identified clinical data were obtained under protocols approved by the IRB of MD Anderson Cancer Center. Slides were sectioned at 4 microns, hydrated with water, and antigen retrieved at pH 6.0. TMA slides were stained with pri- mary antibodies provided in key resources table, followed by anti-rabbit or anti-mouse HRP. Slides were then visualized with diami- nobenzidine, counterstained with hematoxylin, dehydrated, and mounted with coverslips. H-score quantification of intensities was performed using QuPath (0.2.3). 106 Z-scaled H-scores were used for unsupervised hierarchical clustering using ComplexHeatmap10 to classify specimens into SES-high, -intermediate, and -low.

Analysis of TCGA data

Public mRNA datasets and metadata were downloaded using R packages TCGAbiolinks (2.16.1)107 and GEOquery (2.56.0).108 SMARCA4-high was defined based on maximum difference in log-rank test statistic. SMARCA2-high and SMARCA2-low were deter- mined by ranking tumors based on expression levels and matching the group sizes to those of SMARCA4-high and SMARCA4-low. Hazard ratio was calculated using survival.116 Kaplan-Meier curves were plotted with survminer (0.4.9). Differential analyses involving SMARCA4 expression or steroid phenotype were performed using limma. 103 HSP and LSP were previously annotated.56,125 Log2 counts per million (CPM) of SF-1, B-catenin, SMARCA4 of each tumor sample were Z-scaled, and were then used to classify spec- imens into SES-high, -intermediate, and -low based on unsupervised hierarchical clustering using ComplexHeatmap. To evaluate the efficiency of predicting HSP, specimens were divided randomly in equal portion into training and testing datasets. Multinomial log- linear models were developed and optimized using the training dataset with nnet (7.3-19).112 Model performance was subsequently evaluated on the testing dataset. ROC analysis was performed using pROC (1.18.0).117

RNA expression in human tissues

The normalized transcripts per million (TPM) of NR5A1 expression in human tissues were acquired from The Human Protein Atlas consensus data.

Structural model of enhancer activation

The three-dimensional structural model was generated using YASARA (24.4.10)3,113 informed by published structures as well as coIP and XL-MS results. The structures of BAF (PDB 6LTJ)21 and p300 (PDB 7W9V)126 bound to nucleosomes were used as

Molecular Cell Article

CellPress OPEN ACCESS

starting templates. The dinucleosome was modeled with linear 46-bp DNA containing the SF-1 motif obtained from HOMER (CAAGGTCACG). The structure of the zinc finger DNA-binding domain (DBD) of SF-1 (aa 1-118) in complex with DNA containing the SF-1 motif was predicted using AlphaFold3 (AF3). The ß-catenin C-terminal domain (CTD, aa 665-781) was modeled via AF3 in complex with a region on the SF-1 ligand-binding domain (LBD, aa 221-243) previously identified as critical for this interaction by mutagenesis, 66 incorporating modifications reported on PhosphoSitePlus. 127 The final 30 amino acids of the ß-catenin C-terminal IDR are sufficient for interaction with the p300 TAZ2 domain by protein NMR39 while also being dispensable for SF-1 binding.66 Based on these data, we generated a hypothetical interface between the p300 TAZ2 domain and the final 30 amino acids of ß-catenin CTD using AlphaFold3 (AF3). Subsequently, the disordered CTD was linked to the structure of -catenin Armadillo repeats (aa 134-664, PDB 1JPW). 67 We used AF3 to model the interaction of ARID1A IDR2 (aa 1210-1230 and 1597-1610) and @-catenin Armadillo repeats 1-5 (aa 161-350), which we detected to have intermolecular cross-links. The individual complexes were aligned and linked using YASARA’s build tool, then subjected to energy minimization.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical analyses were performed in R (3.6.1) as two-sided tests. Unless otherwise specified, statistical analyses between groups are done by t test. For multiple-comparison tests, p values were adjusted using the Benjamini-Hochberg FDR correction procedure. All n values indicate independent biological replicates, genome-wide studies used n = 2 per condition in agreement with recommen- dations of the ENCODE Consortium.

Quartz 4

Explorer

40695292

Molecular Cell

B-catenin functions as a molecular adapter for disordered cBAF interactions

Graphical abstract

Authors

Correspondence

In brief

Highlights

Molecular Cell

Article -catenin functions as a molecular adapter for disordered cBAF interactions

SUMMARY

INTRODUCTION

Molecular Cell Article

Molecular Cell Article

RESULTS

Excess SF-1-dependent hormone production is linked with high SMARCA4 and -catenin

Molecular Cell Article

Molecular Cell Article

SMARCA4 regulates SF-1 transcriptional activity

cBAF enables enhancer binding of SF-1 and -catenin

cBAF sustains feedforward expression of SF-1 and ß-catenin

B-catenin is a molecular adapter between cBAF and SF-1

IDR2 of ARID1A mediates association of cBAF with ß-catenin

The ARID1A IDR2 directly interacts with ß-catenin Armadillo repeats

Molecular Cell Article

Molecular Cell Article

Molecular Cell Article

Molecular Cell Article

Molecular Cell Article

DISCUSSION

Limitations of the study

RESOURCE AVAILABILITY

Lead contact

Materials availability

Data and code availability

ACKNOWLEDGMENTS

AUTHOR CONTRIBUTIONS

DECLARATION OF INTERESTS

STAR*METHODS

SUPPLEMENTAL INFORMATION

REFERENCES

Molecular Cell Article

Molecular Cell Article

Molecular Cell Article

Molecular Cell Article

STAR*METHODS

KEY RESOURCES TABLE

Molecular Cell Article

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell lines

METHOD DETAILS

Cell-line generation

CRISPR-Cas9 gene knockout

shRNA knockdown

piggyBac transposition

Lentiviral transduction

Molecular Cell Article

Inhibitors and degraders

Sourcing and synthesis

Treatment conditions

RT-qPCR

Western blotting Immunoblotting

Co-immunoprecipitation

Chromatin immunoprecipitation and sequencing

Library preparation

Analysis

RNA sequencing

Library preparation

RNA-seq data analysis

Gene set enrichment analysis

ATAC sequencing

Library preparation

Analysis

4C library preparation, sequencing, and analysis

Identification of enhancers for NR5A1 and CTNNB1

Cross-linking and nuclei isolation

DNA digestion and ligation

Molecular Cell Article