MAF amplification licenses ERα through epigenetic remodelling to drive breast cancer metastasis

Abstract

MAF amplification increases the risk of breast cancer (BCa) metastasis through mechanisms that are still poorly understood yet have important clinical implications. Oestrogen-receptor-positive (ER⁺) BCa requires oestrogen for both growth and metastasis, albeit by ill-known mechanisms. Here we integrate proteomics, transcriptomics, epigenomics, chromatin accessibility and functional assays from human and syngeneic mouse BCa models to show that MAF directly interacts with oestrogen receptor alpha (ERα), thereby promoting a unique chromatin landscape that favours metastatic spread. We identify metastasis-promoting genes that are de novo licensed following oestrogen exposure in a MAF-dependent manner. The histone demethylase KDM1A is key to the epigenomic remodelling that facilitates the expression of the pro-metastatic MAF/oestrogen-driven gene expression program, and loss of KDM1A activity prevents this metastasis. We have thus determined that the molecular basis underlying MAF/oestrogen-mediated metastasis requires genetic, epigenetic and hormone signals from the systemic environment, which influence the ability of BCa cells to metastasize.

Fig. 1. MAF amplification drives E2/ER signalling-dependent BCa metastasis.

a, Schema of the relationship between MAF amplification in primary BCa, bone metastasis and bisphosphonate (BSP) treatment response in patients. b,c, Normalized photon flux quantification (mean ± s.e.m.) of bone metastasis in vivo, in BALB/c nude mice injected intracardially (IC) with control (mock, n = 14 limbs) or MAF-overexpressing (MAF, n = 14 limbs) MCF7 cells (b), and ex vivo, of leg bones from mice in b (c; representative images). In c, the median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown. Statistical significance determined by two-tailed Mann–Whitney test. The colour scale indicates the intensity of radiance. d, Experimental design for obtaining control and Maf-overexpressing mTB BCa cells. GEMM, genetically engineered mouse model. e, Representative paraffin sections of MPA-DMBA-induced mBCa tumours stained with H&E or against ER, CK18, CK17 or p63. Scale bar, 50 µm. f, Immunoblot analysis of ER protein expression in mTB-derived cell lines. MCF7 cells, positive control; MDA-MB-231 cells, negative control; GAPDH, loading control. g, Representative paraffin sections of mTB-derived cell pellets stained for CK18, CK17 or p63. Scale bar, 50 µm. h,i, Control and Maf-overexpressing mTB-derived cell lines (see d) at five days after infection, analysed by immunoblot for MAF and green fluorescent protein (GFP) expression (h), and as MAF-stained paraffin sections (i). GAPDH, loading control (h). Scale bar, 50 µm (i). j, Left: schema of the bone colonization experiment. Right: bone lesion in vivo photon flux quantification (mean ± s.e.m.), control (n = 6 limbs) or Maf-overexpressing (n = 10 limbs) mTB cells. k, Normalized photon flux of ex vivo legs of mice injected intracardially with control (n = 6 limbs) or Maf-overexpressing (n = 6 limbs) mTB cells. Representative images are shown. The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown. Statistical significance determined by a two-tailed Mann–Whitney test. The colour scale indicates the intensity of radiance. l, Zoledronic acid (ZOL) experiment overview (left) and quantification of bone homing by MAF cells in mice treated with vehicle (PBS) (n = 9 mice) or ZOL (n = 9 mice) at day 65 (right). Statistical significance determined by a two-tailed Fisher’s exact test. m, Extraskeletal metastasis quantification in mice after intracardiac injection with mTB Maf-overexpressing cells with vehicle (n = 9 mice) or ZOL (n = 9 mice) at day 65. Source data

Fig. 2. MAF interacts with the ER transcriptional complex.

a, Network diagram of high-confidence MAF interactors (BFDR < 0.02; spectral counts show a threefold enrichment in BioID2-MAF samples as compared to the myc-BioID2 control) identified in MCF7 cells using BioID. Four BioID2-MAF fusion proteins (N- or C-terminal fusion, MAF-S or -L isoforms) were used as baits. The Venn diagrams show the MAF interactors discovered with each bait. The STRING database was used for visual representation using publicly available protein interaction data. Network edges indicate high-confidence protein–protein associations. Disconnected nodes in the network are hidden. The Markov cluster (MCL) algorithm was used. n = 2 biological replicates. For protein–protein interaction enrichment, P < 1 × 10⁻¹⁶ (two-sided). b, Biological process, molecular function and cellular component GO analyses of MAF interactors. Hypergeometric test, one-tailed. Significance was defined using the adjusted P value with the Benjamini and Hochberg multiple testing correction. c, Bait–prey dot plot for MAF interactors. Selected high-confidence MAF interactors discovered by BioID in MCF7 cells are organized by protein complexes or according to their known functions. Abundances are visualized in MCF7 and MDA-MB-231 cells. Dot colours indicate the average spectral counts for each indicated interactor. Dot size indicates the relative abundance of the interactor across the four different baits. Edge colour shows the BFDR value associated with each bait–prey interaction. d, Representative immunoblot (WB, western blot) showing anti-HA co-IP of endogenous ER with HA-tagged MAF (S or L isoforms). Co-immunoprecipitated ER band densities are normalized to the input. The means of three different experiments are shown below. Statistical significance determined by a two-tailed Mann–Whitney test. e, PLA of the HA or ER antibody, alone or together, in MAF-overexpressing MCF7 cells treated with DMSO or ER-PROTAC (1 μM) at 24 h before fixation. Representative confocal microscopy images for the PLA red signal and DAPI nuclear staining (with zoomed insets) are shown. Scale bar, 50 μm. Inset scale bar, 10 μm. f, PLA signal quantification. Each dot represents the average PLA signal from 123 to 201 nuclei per condition. n = 3 biological replicates. Bars represent mean ± s.e.m. Statistical significance determined by two-tailed Wilcoxon rank-sum test. g, Representative immunoblot showing ER degradation by different concentrations of ER-PROTAC in MCF7 cells. Tubulin, loading control. Source data

Fig. 3. MAF regulates the E2/ER-induced metastasis transcriptional gene program.

a, RNA-seq heatmap of differentially expressed genes in MAF-overexpressing compared to control (mock-infected) MCF7 cells, HD or E2-treated (10 nM, 6 h). Expression profiles are grouped in six clusters based on comparisons between the four conditions. Cluster 1, MAF upregulated genes; cluster 2, MAF downregulated genes; cluster 3, E2 upregulated genes; cluster 4, E2 downregulated genes; cluster 5, E2 and MAF upregulated genes; cluster 6, E2 and MAF downregulated genes. The colour scale indicates expression levels. n = 3 biological replicates. b, Associated GO terms for genes in the clusters in a. Hypergeometric test, one-tailed. Significance is defined by the adjusted P value using the Benjamini and Hochberg multiple testing correction. c, Quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) expression analysis of selected genes of clusters 1–4 and 6 in control (mock-infected) and MAF-overexpressing MCF7 cells, either HD or E2-treated. Expression is normalized to the housekeeping gene GAPDH. MUCL, n = 3; SPANXA1, n = 3; PGR, n = 5; CCDN1, n = 5; GREB1, n = 6; BMF, n = 3; NFATC4, n = 3. Three to six biological replicates per gene. Data are presented as mean ± s.e.m. P values were calculated using a two-sided t-test. d, Differentially expressed genes in cluster 5 as determined by RNA-seq (left) (see also a) and significantly enriched hallmark GO terms for the indicated cluster (right). Hypergeometric test, one-tailed. Significance was defined by the adjusted P value using the Benjamini and Hochberg multiple testing correction. e, qRT–PCR expression analysis of selected genes of cluster 5 in control (mock-transfected) and MAF-overexpressing MCF7 cells, either HD or E2-treated. Expression was normalized to the housekeeping gene GAPDH. FGF18, n = 9; PTHLH, n = 8; JAG1, n = 8; TMEM2, n = 8; TGFA, n = 7; JAK1, n = 6; SHH, n = 7. Six to nine biological replicates per gene. Data are presented as mean ± s.e.m. P values were calculated using a two-sided t-test. f, Correlation of MAF expression with FGF18, PTHLH, JAG1, TMEM2, TGFA, JAK1 and SHH in patients with ER⁺ HER2⁻ BCa. Gene expression data were retrieved from TCGA and METABRIC. Data are presented as mean ± s.d. NA, not available. Correlation was determined using the two-tailed Spearman’s correlation test. Unadjusted P values. Source data

Fig. 4. MAF chromatin binding overlaps with and expands binding of ER.

a, ER and MAF ChIP-seq peaks in control and MAF-overexpressing MCF7 cells that were HD or E2-treated (for 1 h). Groups of binding sites are defined based on the positional overlap between ER and MAF binding. Colour scale bars indicate the scale for Reads Per Genome Content (RPGC) normalized coverage (deepTools). The number of peaks is indicated for each group. b, Shared MAF/ER binding sites from a are divided into common and increased ER binding sites in the presence of E2 (ER E2-gained) or in the presence of E2 and MAF overexpression (ER MAF/E2-gained). c, UCSC genome browser (http://genome.ucsc.edu, February 2009 (GRCh37/hg19)) screenshots of ChIP-seq profiles at representative target genes, showing ER ChIP-seq tracks from control and MAF-overexpressing MCF7 cells, either HD or E2-treated. p300 ChIP-seq peaks (from ref. ) depict active enhancer regions. MAF ChIP-seq tracks from MAF-overexpressing MCF7 cells under HD conditions are shown. Predicted MAF binding sites (using the MAF or MARE matrices) within ER peaks are represented in black. bs, binding sites. d, Venn diagrams showing the overlap between p300 and ER binding sites in control (left) or MAF-overexpressing MCF7 cells (right) after E2 treatment. e, qRT–PCR expression analysis of PTHLH (left) and JAG1 (right) in control (mock-infected < yellow/orange) and MAF-overexpressing (blue/purple) MCF7 cells transduced with a lenti-dCas9-KRAB and a lentiGuide-puro expressing sgScramble (dark) or specific sgRNAs against PTHLH and JAG1 herein uncovered enhancer sequences (light). Cells were cultured under HD conditions and stimulated with 10 nM E2. Expression was normalized to the housekeeping gene B2M. n = 5 (PTHLH) and n = 6 (JAG1) biological replicates. Data are presented as mean ± s.e.m. P values were calculated using a one-sided t-test. f, Circos plot summarizing the chromosomal distribution of ER ChIP-seq reads in control or MAF-overexpressing MCF7 cells under HD conditions and after E2 treatment (10 nM, 1 h). The outermost circle represents the ideograms of each chromosome with labels in Mb of physical distance. The inner two circles show the density of ER ChIP-seq reads. Source data

Fig. 5. MAF amplification causes a change in the chromatin landscape.

a, Heatmap of ATAC-seq normalized data showing most differential accessible peaks in MAF-overexpressing compared to control (mock-infected) MCF7 cells under HD or E2-treated (10 nM) conditions. Cluster A, MAF-dependent open chromatin; cluster B, MAF-dependent closed chromatin; cluster C, E2-dependent open chromatin; cluster D, E2-dependent closed chromatin; cluster E, MAF/E2-dependent open chromatin; cluster F, MAF/E2-dependent closed chromatin regions. n = 4 biological replicates. The colour scale indicates opening levels (regularized log (rlog) + z-score normalization of counts per peak matrix). b, ATAC-seq signals that were centred on TSSs in control (mock-infected) and MAF-overexpressing MCF7 cells, either HD or E2-treated (10 nM); signals were normalized to mock HD. c, Density distribution of distances to TSSs for peaks according to the clusters from a. d, Percentage of annotated peaks (promoter-TSS/gene body/intergenic) in the different cluster groups from a. Statistical significance determined by a permutation test (one-sided). e, Peak breadth plot according to peak annotation (promoter-TSS/gene body/intergenic) and cluster (from a). Statistical significance determined by two-sided Wilcoxon rank sum test. f, Plot depicting the percentage of promoter-TSS annotated ATAC-seq peaks from ATAC-seq data corresponding to clusters (from a) that overlap with H3K27ac (active enhancers and promoters), H3K4me3 (active promoters) or both (active promoters—broad peaks). Statistical significance determined by one-sided permutation test. g, Integration of ATAC-seq and RNA-seq data. For comparison, only peaks annotated as promoter-TSS were considered. ATAC-peak candidates were assigned a gene using Homer annotations; gene signatures corresponding to each cluster were used to perform Gene Set Analysis (GSA) on the RNA-seq data. The colour scale indicates DESeq2 test statistics in peaks near TSSs (±10 kb). ES, Enrichment Score; NES, Normalized Enrichment Score.

Fig. 6. Shared ER-MAF binding sites control the E2-induced metastasis gene program.

a, Enrichment of chromatin binding sites from MCF7 cells in the vicinity of E2-induced genes compared with an equal number of constitutively expressed genes in MCF7 cells. Lines illustrate the cumulative percentage of sites within a given distance from the TSSs. b, MAF/E2-shared ER-ChIP peaks that overlap with ATAC-seq peaks are enriched in the MAF/E2 condition (MAF/E2 versus mock comparison in ATAC-seq data; P < 0.10). Only one peak per gene is considered. Permutation test, two-tailed. c, Percentage of BCa ATAC-seq peaks (from TCGA) occupied in MCF7 cells by both MAF and ER, or only MAF or ER, that are connected to E2 target gene promoters through promoter–enhancer connections (red line). The distribution of connections between the same number of random ATAC-seq peaks and E2 target gene promoters is shown for comparison (grey); 10,000 iterations were performed to estimate the distribution to test the null hypothesis. Permutation test, one-tailed. d, Identified functional promoter–enhancer links between E2-induced genes and MAF/ER-, MAF- and ER-occupied BCa ATAC-seq peaks. e, Kaplan–Meier curves representing the probability of bone metastasis-free survival in ER⁺ BCa patients (MSKCC-EMC dataset), stratified according to the expression of a MAF-dependent gene signature generated from the integration of RNA-seq transcriptomics and MAF- and MAF/E2-dependent ER ChIP-seq data. Log-rank test, two-tailed.

Fig. 7. KDM1A inhibition disrupts E2/ER and MAF-dependent signalling and prevents metastasis.

a, PLA of HA, KDM1A or HA plus KDM1A antibodies together in HA-MAF-L-overexpressing MCF7 cells transduced with scrambled (shSc) or KDM1A-targeted shorthairpin RNA (shRNAs). Representative confocal microscopy images staining, with zoomed insets, are shown. Scale bars, 50 μm. Inset scale bars, 10 μm. b, Representative immunoblot showing MAF and KDM1A expression in control and MAF-overexpressing MCF7 cells with or without KDM1A knockdown. Tubulin and GAPDH, loading controls. c, PLA signal quantification. Each dot represents the average PLA signal of 140 to 252 nuclei/condition/biological replicates (n = 3 biological replicates). Bars indicate mean ± s.e.m. Statistical significance, two-tailed Wilcoxon rank-sum test. d, RNA-seq heatmap showing expression profiles of genes from clusters 5 and 6 in MAF-overexpressing cells treated with E2 and with or without KDM1A knockdown. e, qRT–PCR expression analysis of selected genes in cluster 5 in MAF-overexpressing MCF7 cells treated with E2 and with or without KDM1A knockdown. Expression was normalized to GAPDH. FGF18, n = 5; JAG1, n = 6; PTHLH, n = 5; SOX9, n = 4; TMEM2, n = 3; TGFA, n = 3; JAK1, n = 3 and SHH, n = 4 biological replicates per gene. Data are presented as mean ± s.d. Statistical significance, two-sided t-test. f,g, Immunoblot showing methylation of H3K9 after ORY-1001 treatment in MAF-overexpressing or control MCF7 cells (f) or mTB cells (g). Total H3, loading control. h, BICA after ORY-1001 treatment of MAF-overexpressing MCF7 bone lesions. Each dot represents an independent bone fragment; n = 24 to 30 bone fragments (from three different mice) in each group. Representative bioluminescence images show MAF-overexpressing MCF7 cells treated with DMSO or ORY-1001 at day 1 or day 43. Data are presented as mean ± s.e. Linear mixed model fit by REML with z-tests for individual comparisons, two-tailed. Significance was defined by the adjusted P value using the Benjamini and Hochberg multiple testing correction. i, Schematic diagram of bone colonization (left) and normalized ex vivo bone metastasis photon flux quantification (right), in which mice with intratibial (IT) injections of control or MAF-overexpressing mTB cells were treated with DMSO (Ad-ctrl, n = 16 limbs; Ad-cre, n = 14 limbs) or ORY-1001 (Ad-ctrl, n = 12 limbs; Ad-cre, n = 12 limbs) for 40 days. The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown. Statistical significance, two-tailed Mann–Whitney test. Source data

Extended Data Fig. 1. MAF overexpression metastasis models.

a, Schema of the tissue-colonization experiment (left). Quantification of metastasis in adrenal glands (Mock n = 14, MAF n = 16), ovaries (Mock n = 14, MAF n = 16), brain (Mock n = 7, MAF n = 8) and liver (Mock n = 7, MAF n = 8), from mice injected with control or MAF-overexpressing MCF7 cells (right). Statistical significance, two-tailed Fisher’s exact test. b, Schema of circulating tumor cell experiment. Normalized ex vivo photon flux quantification of mammary fat pad (MFP) growth in BALB/c mice injected with control (n = 10 mice) or MAF-overexpressing (n = 12 mice) MCF7 cells. The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown (left). MCF7 Mock and MAF-overexpressing circulating cells were measured by qPCR using a human GAPDH and mouse B2m mRNA probe (mean ± sd) in blood samples obtained from mice bearing mammary fat pad size-matched tumors (n = 9 tumors per group) (right). Statistical significances were calculated using two-sided Wilcoxon rank-sum tests. c, Schematic representation of the experimental design. Control (mock-infected) and MAF-overexpressing MCF7 cells are cultured in hormone-deprived (HD) media containing charcoal-stripped serum for 72 h to deprive cells of E2. Cells are then stimulated with 10 nM E2 or vehicle. d, Cell proliferation assay of control and MAF-overexpressing MCF7 cells cultured with HD medium and treated with vehicle or 10 nM E2, n = 3. Data is shown as mean ± sd from three biologically independent samples. Statistical significance was calculated using an unpaired two-tailed t-test. e, Representative images of culture plates (left) and BrdU stainings (right) of MCF7 control (mock-infected) and MAF-overexpressing MCF7 cells cultured under HD conditions and after 10 nM E2 administration. Scale bar, 50 μm. f, Quantitative analysis of BrdU positive cells per field. Data represents mean ± sd of 3 biological replicates (n = 3 fields per experiment). Statistical significance was calculated using a two-sided Wilcoxon rank-sum test. g, Gene targeting strategy. The floxed PGK-neo-STOP cassette followed by the Maf cDNA and an IRES-eGFP-Luciferase were targeted into the Rosa26 locus. The position of the external probes (5’P and 3’P) used for Southern blot analysis and relevant restriction enzyme cleavage sites (H, HindII and E, EcoRI) are shown. The diagnostic HindIII DNA fragments for the wild-type (4.4 kbp) and the targeted allele (7.3 kbp) as well as the diagnostic EcoRI DNA fragments for the wild-type (15.6 kbp) and the targeted allele (13.5 kbp) are represented by dotted arrows. h, Southern blot analysis of DNA isolated from recombinant ES cell clones carrying wild-type (WT) and/or recombinant alleles (LSLMaf, HR1-HR5). The migration and sizes of the diagnostic HindIII DNA fragments probed with probe 5’P for the WT (4.4 kbp) and the LSLMaf (7.3 kbp) alleles are indicated by arrowheads. The migration and sizes of the diagnostic EcoRI DNA fragments probed with probe 5’P for the WT (15.6 kbp) and the LSLMaf (13.5 kbp) alleles are indicated by arrowheads. i, Long PCR analysis of DNA isolated from recombinant ES cell clones carrying WT or LSLMaf (HR1-5) alleles. R26F primer is located in the mouse genome, upstream the knock-in cassette, and CMVR primer binds to the CAG promoter. R26R primer is located in the mouse genome, downstream the knock-in cassette, and LucF primer binds to the luciferase sequence. Lane B, blank control. Lane M, DNA size marker. DNA fragment size is indicated. Source data

Extended Data Fig. 2. MAF overexpression provides a competitive advantage for bone colonization in ER+ BCa.

a, Schematic representation of the experimental design to obtain mouse BCa cell lines (mTB1-TB4). b, Cell proliferation assay of mTB BCa cells cultured with HD medium and treated with vehicle or 10 nM E2, n = 3 biological replicates. Data is shown as mean ± sd. Statistical significance, two-sided Wilcoxon rank-sum test. c, 4-OHT dose-response curve (mean ± sd) and IC50 measurement in mTB BCa cells. n = 3 biological replicates. d, Representative H&E and ER staining in a tumor generated by mTB BCa cells. Scale bar, 50 μm. e, Immunoblot showing MAF overexpression in different BCa cell lines. f, Schematic representation of the bone-colonization experiment (left). Normalized ex vivo leg photon flux of mice injected in the mammary fat pad (MFP) with control (n = 4 limbs) or Maf-overexpressing (n = 14 limbs) mTB cells (right). The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown. Statistical significance, two-tailed Mann–Whitney test. g, Schematic representation of the bone-colonization experiment (left). Quantification of bone colonization by control (n = 24 limbs) and Maf-overexpressing (n = 24 limbs) mTB cells at day 25 (right). Statistical significance, two-tailed Fisher’s exact test. h, Schematic representation of the bone-colonization experiment (left). Normalized ex vivo leg photon flux of mice injected intratibially with a mixture of Ad-ctrl-mCherry and Ad-cre MAF-GFP cells at different ratios (10:0 n = 16 limbs; 9:1 n = 16 limbs; 5:5 n = 15 limbs; 1:9 n 16 limbs; 0:10 n = 13 limbs)(right). The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown. i, Representative staining images of mCherry and GFP (left) and quantification of mock (mCherry) and GFP-positive metastatic cells (right) in sized-matched bone metastasis from mice injected intratibially with a mixture of different proportions of control and MAF-GFP MCF7 cells. Scale bar, 50 μm (left); n = 3 limbs in each group (right). Source data

Extended Data Fig. 3. MAF interacts with ER through its N-terminal transactivation domain in ER+ BCa cells.

a, Representative immunoblot showing MAF expression in control (mock-infected) or MAF-overexpressing (S and L isoforms) MCF7 cells (left), or MDA-MB-231 cells (right). Tubulin was used as a loading control. b, Representative immunoblot showing the expression of myc-BioID2 alone and BioID2-tagged MAF proteins (N- or C-terminal fusion, MAF S or L isoforms) in MCF7 cells (left) and MDA-MB-231 cells (right) analysed after 24 h biotin supplementation. Expression of the BioID2 biotin ligase leads to biotinylation of endogenous proteins, detected using streptavidin-HRP. c, Representative immunofluorescence images of MCF7 cells transfected with myc-BioID2 or myc-BioID2-MAF L plasmids. The myc-BioID2 tag alone is distributed throughout the nucleus and cytoplasm (top), whereas the myc-BioID2-MAF L fusion protein localizes primarily in the nucleus (bottom). Biotinylated proteins, detected with fluorescently-labelled streptavidin, colocalize with BioID2 when cells are cultured with excess biotin (50 μM). DNA is labelled with DAPI. Scale bar, 50 μm. d, Immunoblot showing anti-HA co-immunoprecipitation (co-IP) with HA-tagged MAF (S or L isoforms) of endogenous ARID1A, NCoR1, NCoA3, KDM1A and MTA1. Co-IP band densities were normalized to each respective input. e, Venn diagrams showing high-confidence MAF interactors (BFDR < 0.02 and spectral counts with a 3-fold enrichment in BioID2-MAF samples compared to the myc-BioID2 control) discovered by BioID in MDA-MB-231 cells. Four BioID2-MAF fusion proteins (N- or C-terminal fusion, MAF S or L isoform) were used as baits. n = 1 biological replicate. f, PLA of HA or ER antibody alone or both antibodies together in MAF L-overexpressing MDA-MB-231 cells. Representative confocal microscopy images for PLA red signal and DAPI nuclear staining with magnified inset are shown. Scale bar, 50 μm; inset scale bar, 10 μm. g, Immunoblot showing HA co-IP of HA-tagged MAF (S or L isoforms) with endogenous ER in MDA-MB-231 cells. h, Schematic representation of full-length MAF L and two N-terminal truncation mutants. i, PLA of HA or ER antibody alone or both antibodies together in MCF7 cells transfected with full-length HA-tagged MAF L and N-terminal truncation mutants 24 h prior to fixation. Representative confocal microscopy images for HA-MAF expression (green), PLA signal (red) and DAPI nuclear staining with magnified inset are shown. Scale bar, 50 μm; inset scale bar, 10 μm. j, PLA signal quantification. Each dot represents the average PLA signal from 65 to 135 nuclei per condition. n = 3 biological replicates. Bars represent mean ± sem. Statistical significance, two tailed Wilcoxon rank-sum test. Source data

Extended Data Fig. 4. Interaction of ER AF1 and AF2 domains with MAF.

a, Immunofluorescence showing the nuclear localization of AF1 and AF2 ER domains through the detection of the 6x His-tag. Scale bar, 20 μm. b, PLA of HA or 6x His-tag antibody alone or both antibodies together in MAF-overexpressing MDA-MB-231 cells transduced with AF1 or AF2 ER domains. Representative confocal microscopy images for PLA red signal and DAPI nuclear staining with magnified inset are shown. Scale bar, 50 μm; inset scale bar, 10 μm. c, PLA quantification. Each dot represents the average PLA signal from 64 to 208 nuclei per condition. n = 3 biological replicates. Bars represent mean ± sem. Statistical significance, two-tailed Wilcoxon rank-sum test. Source data

Extended Data Fig. 5. MAF-E2 induced transcriptional changes in BCa cell lines and ER dependency.

a, Principal component analysis (PCA) on the rlog normalized expression matrix (showing first two components). Batch effect was adjusted gene-wise using a linear model. b, RNA-seq expression of a selected gene from each cluster represented by fold-change of RPKM values. Data represents the average of 3 biological replicates. c, Analysis of MAF expression in control and MAF-overexpressing MCF7 and T47D cell lines under hormone deprivation (HD) conditions and after 10 nM E2 administration by qRT-PCR. Expression was normalized to the housekeeping gene GAPDH. n = 3 independent biological replicates. Data are mean ± s.d. d, qRT-PCR expression analysis of a selected gene in each cluster in control (mock-infected) and MAF-overexpressing T47D cells under HD conditions and after 10 nM E2 administration. Expression was normalized to the housekeeping gene GAPDH. MUCL, n = 3; SPANXA1, n = 3; GREB1, n = 5; BMF, n = 4; NFATC4, n = 4; FGF18, n = 7; PTHLH, n = 7; JAG1, n = 7; TMEM2, n = 3; TGFA, n = 3 and JAK1, n = 3. 3 to 7 biological replicates per gene. Data are mean ± s.d. The P value was calculated using a two-sided t-test. e, qRT-PCR expression analysis of selected genes in control (mock-infected) and MAF-overexpressing T47D cells transduced with Androgen Receptor (AR)-targeted shRNA and cultured under HD conditions and after 10 nM E2 administration. Expression was normalized to the housekeeping gene GAPDH. n = 3 independent biological replicates. Data are mean ± s.d. The P value was calculated using a two-sided t-test. f, qRT-PCR expression analysis of selected genes in control (Ad-ctrl) and MAF-overexpressing (Ad-cre) mouse BCa cells transduced with AR-targeted shRNA and cultured under HD conditions or E2 treated (10 nM). Expression was normalized to the housekeeping gene GAPDH. n = 3 independent biological replicates. Data are mean ± s.d. The P value was calculated using a two-sided t-test. g, Immunoblot showing ER degradation by using ER-PROTAC in control (mock) and MAF-overexpressing MCF7 cells 1 and 6 hours after E2 stimulation. GAPDH was used as a loading control. h, Principal component analysis (PCA) on the rlog normalized expression matrix (showing first two components). Batch effect was adjusted gene-wise using a linear model. i, RNA-seq heatmap of differentially expressed genes, already defined clusters 5 and 6, in MAF-overexpressing compared to control (mock-infected) MCF7 cells upon ER degradation followed by administration of E2 (10 nM for 6 h). Color scale indicates expression levels. Red indicates upregulation and blue shows downregulation. n = 3 biological replicates. j, qRT-PCR expression analysis of PTHLH and JAG1 in MAF-overexpressing MCF7 cells transduced with PTHLH and JAG1-targeted shRNAs. Expression was normalized to the housekeeping gene GAPDH. k, Schematic representation of the experimental design. Normalized in vivo photon flux quantification (mean ± sem) of bone metastasis in BALB/c nude mice injected intracardially with control (scramble, n = 14 limbs) and knockdown of PTHLH or JAG1 (n = 14 limbs) in MAF-overexpressing MCF7 cells (left). Normalized ex vivo photon flux quantification of bone metastasis. The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown. Statistical significance, two-tailed Mann–Whitney test (right). Source data

Extended Data Fig. 6. MAF expands ER binding throughout the genome.

a, ChIP-seq heatmap showing the distribution of ER reads (in control and MAF-overexpressing MCF7 cells) and p300 on the 5264 ER binding sites found in MAF-overexpressing cells after E2 stimulation (peak summit ±1 kb). Enrichment levels were normalized for the total number of mapped reads of each sample. Peaks were ranked by the intensity of ER signal in MAF-overexpressing cells after E2 stimulation. b, Venn diagrams showing ER binding sites (ERbs) (ChIP-seq peaks) identified in control (mock-infected) MCF7 cells under hormone deprivation (HD) conditions and after 1 h 10 nM E2 administration (top). Boxplot showing ER ChIP-seq signal intensity for common peaks in both HD conditions and after E2 stimulation (543 peaks) and for peaks that appeared only after E2 administration (2837 peaks) (bottom). Boxes represent the first, second (median) and third quartiles. Whiskers indicate maximum and minimum values with outliers being excluded. n= number of peaks (594 in Mock HD, 3380 in Mock E2) from 1 biological replicate per condition. Two-tailed Mann-Whitney test, unadjusted P values. c, Venn diagrams showing ERbs identified in control and MAF-overexpressing MCF7 cells after 1 h E2 stimulation (top). Boxplot represents ER ChIP-seq signal intensity for common ER binding sites in both conditions (3101 peaks) and for specific ER binding sites in MAF-overexpressing cells (2163 peaks) (bottom). Boxes represent the first, second (median) and third quartiles. Whiskers indicate maximum and minimum values with outliers being excluded. n= number of peaks (3380 in Mock E2, 5264 in MAF E2) from 1 biological replicate per condition. Two-tailed Mann-Whitney test, unadjusted P values. d, Venn diagrams showing ERbs (ChIP-seq peaks) identified in control and MAF-overexpressing MCF7 cells cultured in HD conditions (left). Boxplot represents ER ChIP-seq signal intensity for mock-specific ERbs (184 peaks), common ERbs in both conditions (410 peaks), and MAF-specific ERbs (247 peaks) (right). Boxes represent the first, second (median) and third quartiles. Whiskers indicate maximum and minimum values with outliers being excluded. n= number of peaks (594 in Mock HD, 657 in MAF HD) from 1 biological replicate per condition. Two-tailed Mann-Whitney test, unadjusted P values. e to l, UCSC genome browser (http://genome.ucsc.edu, Feb. 2009 (GRCh37/hg19)) screenshot of ChIP-seq profiles of the indicated genes. ER ChIP-seq tracks from control and MAF-overexpressing MCF7 cells under HD conditions and after E2 stimulation are shown. MAF ChIP-seq tracks from MAF-overexpressing MCF7 cells under HD conditions are shown. p300 ChIP-seq peaks depict active enhancer regions. Predicted MAF binding sites (using the MAF or MARE matrices) within ER peaks are represented in black.

Extended Data Fig. 7. Characterization of ER and MAF chromatin binding regions.

a, Position-weight matrices of two described MAF binding motifs obtained from the MatBase database (Genomatix software). MARE, MAF response element (half sites) (left). Prediction of MAF binding sites in the vicinity of ER ChIP-seq peaks identified in MAF-overexpressing cells after E2 stimulation generated with the MatInspector program from Genomatix. The number of input sequences with at least one match of the MARE or MAF matrix, the total number of matches in all input sequences, the expected match numbers in an equally sized sample of the genome and its standard deviation, the overrepresentation and the Z-score are shown. A Z-score below –2 or above 2 can be considered statistically significant. b, Boxplots showing enrichment of p300, a marker of active enhancers, in ERbs identified in control (left) and MAF-overexpressing MCF7 cells (right) after E2 administration. Boxes represent the first, second (median) and third quartiles. Whiskers indicate maximum and minimum values with outliers being excluded. n= number of peaks (3380 in Mock E2, 5264 in MAF E2) from 1 biological replicate per condition. c, Percentage of ER ChIP-seq peaks and MAF ChIP-seq peaks annotated according to promoter-TSS / intergenic / gene body.

Extended Data Fig. 8. Characterization of ER and MAF impact on chromatin accessibility landscape.

a, Total number of ATAC peaks in the different cluster groups from 5a according to promoter-TSS / gene body / intergenic annotations. b, Scatter plot reporting the correlation between PROTAC and DMSO conditions in MAF-E2 ATACseq peaks. c, Peak breadth plot of ATAC-seq peaks of clusters from 5a. d, Plot depicting the percentage of ATAC-seq peaks from non-promoter regions corresponding to clusters described in 4a that overlap with H3K27ac (active enhancers and promoters), H3K4me3 (active promoters) or both (active promoters – broad peaks). Statistical significance, one-sided permutation test. e, UCSC genome browser (http://genome.ucsc.edu, Feb. 2009 (GRCh37/hg19)) screenshots of ATAC-seq profiles of PTHLH (left) and FGF18 (right) target genes. ATAC-seq tracks from control and MAF-overexpressing MCF7 cells under HD conditions and after E2 stimulation are shown. Predicted MAF binding sites (using the MAF or MARE matrices) are represented in black. f, Enhancers were ranked based on increasing ER (left) and MAF (right) signal. Genes within super-enhancers in both conditions cells are highlighted. g, Venn diagram showing overlap of defined super-enhancers occupied by ER (in MAF/E2 conditions) and MAF. h, Kaplan Meier curves representing the probability of non-bone metastasis-free survival in ER+ BCa patients (MSKCC-EMC data set; Gawrzak et al., 2018) stratified according to the expression of a MAF-dependent gene signature generated from the integration of RNA-seq transcriptomic and MAF- and MAF/E2-dependent ER ChIP-seq data. Long-rank test, two tailed.

Extended Data Fig. 9. ER and MAF interaction dependency on KDM1A.

a, Proximity ligation assay (PLA) of HA or KDM1A antibody alone or both antibodies together in MAF-overexpressing MCF7 cells treated with DMSO or ER-PROTAC (50 nM) 24 h prior to fixation. Representative confocal microscopy images for PLA red signal and DAPI nuclear staining with magnified inset are shown. Scale bar, 50 μm; inset scale bar, 10 μm. b, PLA signal quantification. Each dot represents the average PLA signal from 50 to 142 nuclei per condition. n = 3 biological replicates. Bars represent mean ± sem. Statistical significance, two-tailed Wilcoxon n-test. c, Representative immunoblots showing HA co-IP of HA-tagged MAF (S or L isoforms) with endogenous KDM1A upon ER degradation with a PROTAC. d, PLA of HA or ER antibody alone or both antibodies together in MAF-overexpressing MCF7 cells transduced with scrambled (shSc) or KDM1A-targeted shRNA. Representative confocal microscopy images for PLA red signal and DAPI nuclear staining with magnified inset are shown. Scale bar, 50 μm; inset scale bar, 10 μm. e, PLA quantification. Each dot represents the average PLA signal from 53 to 126 nuclei per condition. n = 3 biological replicates. Bars represent mean ± sem. Statistical significance, two-tailed Wilcoxon rank-sum test. f, Representative immunoblots showing HA co-IP of HA-tagged MAF (S or L isoforms) with endogenous ER in cells transduced with scrambled or KDM1A-targeted shRNAs. Source data

Extended Data Fig. 10. KDM1A supports MAF-dependent metastatic functions.

a, UCSC genome browser (http://genome.ucsc.edu, Feb. 2009 (GRCh37/hg19)) screenshot of ChIP-seq profiles of the PTHLH, JAG1, FGF18 and SOX9 target genes. MAF ChIP-seq tracks from MAF-overexpressing MCF7 cells under HD conditions are shown. KDM1A ChIP-seq peaks (published data, Sheng et al., 2018) in MCF7 cells are shown. b, Schematic representation of BICA experimental design. Luciferase-tagged control (scramble shRNA, shSc) and KDM1A-knockdown MCF7 cells with MAF overexpression were inoculated into mice through intra-iliac artery injection. Femur and tibia bones were extracted, segmented into pieces and arranged into 96-well plates filled with cell culture media. Cancer cell growth was traced weekly using bioluminescence imaging (IVIS imaging system). c, BICA assay showing the reduction of KDM1A-knockdown cell growth compared to control MAF-overexpressing cells. Each dot represents an independent bone fragment. n = 17 to 21 bone pieces (from 3 different mice) for each condition. The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown (middle). d, Representative bioluminescence images of bone fragments containing control (shSc) and KDM1A-knockdown MCF7 cells with MAF overexpression at day 1 and at day 42. e, Immunoblot showing methylation of H3K9 (as a surrogate for KDM1A activity) in mouse BCa cells after treatment with different concentrations of ORY-1001. Total histone H3, loading control. f, Schematic representation of the experimental design. Normalized ex vivo photon flux quantification of bone metastasis in BALB/c nude mice injected intracardially with control (Ad-ctrl) and MAF-overexpressing (Ad-cre) mouse BCa cells and treated with DMSO (Ad-ctrl, n = 12 limbs; Ad-cre, n = 12 limbs) or ORY-1001 (Ad-ctrl, n = 12 limbs; Ad-cre, n = 14 limbs) for 23 days. The median (centre line), first and third quartiles (box limits) and the minimum to maximum values (whiskers) are shown (middle). Statistical significance, two-tailed Mann–Whitney test. Source data