Abstract
Identifying phase-separated structures remains challenging, and effective intervention methods are currently lacking1. Here we screened for phase-separated proteins in breast tumour cells and identified forkhead (FKH) box protein M1 (FOXM1) as the most prominent candidate. Oncogenic FOXM1 underwent liquid–liquid phase separation (LLPS) with FKH consensus DNA element, and compartmentalized the transcription apparatus in the nucleus, thereby sustaining chromatin accessibility and super-enhancer landscapes crucial for tumour metastatic outgrowth. Screening an epigenetics compound library identified AMPK agonists as suppressors of FOXM1 condensation. AMPK phosphorylated FOXM1 in the intrinsically disordered region (IDR), perturbing condensates, reducing oncogenic transcription, accumulating double-stranded DNA to stimulate innate immune responses, and endowing discrete FOXM1 with the ability to activate immunogenicity-related gene expressions. By developing a genetic code-expansion orthogonal system, we demonstrated that a phosphoryl moiety at a specific IDR1 site causes electrostatic repulsion, thereby abolishing FOXM1 LLPS and aggregation. A peptide targeting IDR1 and carrying the AMPK-phosphorylated residue was designed to disrupt FOXM1 LLPS and was shown to inhibit tumour malignancy, rescue tumour immunogenicity and improve tumour immunotherapy. Together, these findings provide novel and in-depth insights on function and mechanism of FOXM1 and develop methodologies that hold promising implications in clinics.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$32.99 / 30 days
cancel any time
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
The ATAC, CUT&Tag and RNA-seq data are available in the Genome Sequence Archive in the National Genomics Data Center, the China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences, BioProject: PRJC017309 (GSA-Human: HRA004723), which are publicly accessible (https://bigd.big.ac.cn/gsa-human/browse/HRA004723). Mass spectrometry data have been deposited into the OMIX, China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (accession numbers OMIX007561 and OMIX007562). All relevant data are available in a publicly accessible repository. Source data are provided with this paper.
Code availability
No customized code was generated in this study.
References
Lafontaine, D. L. J., Riback, J. A., Bascetin, R. & Brangwynne, C. P. The nucleolus as a multiphase liquid condensate. Nat. Rev. Mol. Cell Biol. 22, 165–182 (2021).
Shin, Y. & Brangwynne, C. P. Liquid phase condensation in cell physiology and disease. Science 357, eaaf4382 (2017).
Mittag, T. & Pappu, R. V. A conceptual framework for understanding phase separation and addressing open questions and challenges. Mol. Cell 82, 2201–2214 (2022).
Banani, S. F., Lee, H. O., Hyman, A. A. & Rosen, M. K. Biomolecular condensates: organizers of cellular biochemistry. Nat. Rev. Mol. Cell Biol. 18, 285–298 (2017).
Hnisz, D., Shrinivas, K., Young, R. A., Chakraborty, A. K. & Sharp, P. A. A phase separation model for transcriptional control. Cell 169, 13–23 (2017).
Klein, I. A. et al. Partitioning of cancer therapeutics in nuclear condensates. Science 368, 1386–1392 (2020).
Zhao, M. et al. GCG inhibits SARS-CoV-2 replication by disrupting the liquid phase condensation of its nucleocapsid protein. Nat. Commun. 12, 2114 (2021).
Xie, J. et al. Targeting androgen receptor phase separation to overcome antiandrogen resistance. Nat. Chem. Biol. 18, 1341–1350 (2022).
Chakravarty, A. K. et al. Biomolecular condensation: a new phase in cancer research. Cancer Discov. 12, 2031–2043 (2022).
York, A. Targeting viral liquid–liquid phase separation. Nat. Rev. Microbiol. 19, 550 (2021).
Risso-Ballester, J. et al. A condensate-hardening drug blocks RSV replication in vivo. Nature 595, 596–599 (2021).
Wang, S. et al. Targeting liquid–liquid phase separation of SARS-CoV-2 nucleocapsid protein promotes innate antiviral immunity by elevating MAVS activity. Nat. Cell Biol. 23, 718–732 (2021).
Shin, Y. et al. Spatiotemporal control of intracellular phase transitions using light-activated optoDroplets. Cell 168, 159–171.e14 (2017).
Patel, A. et al. A liquid-to-solid phase transition of the ALS protein FUS accelerated by disease mutation. Cell 162, 1066–1077 (2015).
Agarwal, A., Rai, S. K., Avni, A. & Mukhopadhyay, S. An intrinsically disordered pathological prion variant Y145Stop converts into self-seeding amyloids via liquid–liquid phase separation. Proc. Natl Acad. Sci. USA 118, e2100968118 (2021).
Hou, F. et al. MAVS forms functional prion-like aggregates to activate and propagate antiviral innate immune response. Cell 146, 448–461 (2011).
Littler, D. R. et al. Structure of the FoxM1 DNA-recognition domain bound to a promoter sequence. Nucleic Acids Res. 38, 4527–4538 (2010).
Mo, J. S. et al. Cellular energy stress induces AMPK-mediated regulation of YAP and the Hippo pathway. Nat. Cell Biol. 17, 500–510 (2015).
Hardie, D. G., Ross, F. A. & Hawley, S. A. AMPK: a nutrient and energy sensor that maintains energy homeostasis. Nat. Rev. Mol. Cell Biol. 13, 251–262 (2012).
Rai, A. K., Chen, J. X., Selbach, M. & Pelkmans, L. Kinase-controlled phase transition of membraneless organelles in mitosis. Nature 559, 211–216 (2018).
Han, T. W. et al. Cell-free formation of RNA granules: bound RNAs identify features and components of cellular assemblies. Cell 149, 768–779 (2012).
Whyte, W. A. et al. Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell 153, 307–319 (2013).
Creyghton, M. P. et al. Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proc. Natl Acad. Sci. USA 107, 21931–21936 (2010).
Zhang, N. et al. FoxM1 inhibition sensitizes resistant glioblastoma cells to temozolomide by downregulating the expression of DNA-repair gene Rad51. Clin. Cancer Res. 18, 5961–5971 (2012).
Guichard, G. et al. Antigenic mimicry of natural l-peptides with retro-inverso-peptidomimetics. Proc. Natl Acad. Sci. USA 91, 9765–9769 (1994).
Herce, H. D. & Garcia, A. E. Molecular dynamics simulations suggest a mechanism for translocation of the HIV-1 TAT peptide across lipid membranes. Proc. Natl Acad. Sci. USA 104, 20805–20810 (2007).
Alberti, S. & Hyman, A. A. Biomolecular condensates at the nexus of cellular stress, protein aggregation disease and ageing. Nat. Rev. Mol. Cell Biol. 22, 196–213 (2021).
Xie, F. et al. USP8 promotes cancer progression and extracellular vesicle-mediated CD8+ T cell exhaustion by deubiquitinating the TGF-β receptor TbetaRII. EMBO J. 41, e108791 (2022).
Xie, F. et al. Breast cancer cell-derived extracellular vesicles promote CD8+ T cell exhaustion via TGF-β type II receptor signaling. Nat. Commun. 13, 4461 (2022).
Ran, F. A. et al. Genome engineering using the CRISPR–Cas9 system. Nat. Protoc. 8, 2281–2308 (2013).
Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y. & Greenleaf, W. J. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat. Methods 10, 1213–1218 (2013).
Kaya-Okur, H. S. et al. CUT&Tag for efficient epigenomic profiling of small samples and single cells. Nat. Commun. 10, 1930 (2019).
Baar, M. P. et al. Targeted apoptosis of senescent cells restores tissue homeostasis in response to chemotoxicity and aging. Cell 169, 132–147.e16 (2017).
Yuan, Y. et al. Targeting UBE4A revives viperin protein in epithelium to enhance host antiviral defense. Mol. Cell 77, 734–747.e7 (2020).
Borsello, T. et al. A peptide inhibitor of c-Jun N-terminal kinase protects against excitotoxicity and cerebral ischemia. Nat. Med. 9, 1180–1186 (2003).
Beydoun, T. et al. Subconjunctival injection of XG-102, a JNK inhibitor peptide, in patients with intraocular inflammation: a safety and tolerability study. J. Ocul. Pharmacol. Ther. 31, 93–99 (2015).
Suckfuell, M. et al. Efficacy and safety of AM-111 in the treatment of acute sensorineural hearing loss: a double-blind, randomized, placebo-controlled phase II study. Otol. Neurotol. 35, 1317–1326 (2014).
Deloche, C. et al. XG-102 administered to healthy male volunteers as a single intravenous infusion: a randomized, double-blind, placebo-controlled, dose-escalating study. Pharmacol. Res. Perspect. 2, e00020 (2014).
Brooks, P. C. et al. Antiintegrin αvβ3 blocks human breast cancer growth and angiogenesis in human skin. J. Clin. Invest. 96, 1815–1822 (1995).
Chen, K. & Chen, X. Integrin targeted delivery of chemotherapeutics. Theranostics 1, 189–200 (2011).
Murphy, E. A. et al. Nanoparticle-mediated drug delivery to tumor vasculature suppresses metastasis. Proc. Natl Acad. Sci. USA 105, 9343–9348 (2008).
Yang, J. et al. Conjugation of iron oxide nanoparticles with RGD-modified dendrimers for targeted tumor MR imaging. ACS Appl. Mater. Interfaces 7, 5420–5428 (2015).
Ruan, H. et al. Stapled RGD peptide enables glioma-targeted drug delivery by overcoming multiple barriers. ACS Appl. Mater. Interfaces 9, 17745–17756 (2017).
Li, Y. et al. c-Myb enhances breast cancer invasion and metastasis through the Wnt/β-catenin/Axin2 pathway. Cancer Res. 76, 3364–3375 (2016).
Zhang, L. et al. TRAF4 promotes TGF-β receptor signaling and drives breast cancer metastasis. Mol. Cell 51, 559–572 (2013).
Zong, Z. et al. Alanyl-tRNA synthetase, AARS1, is a lactate sensor and lactyltransferase that lactylates p53 and contributes to tumorigenesis. Cell 187, 2375–2392.e33 (2024).
Acknowledgements
The current work was supported by the Chinese National Natural Science Funds (31925013, 32125016, U20A20393, U24A20371, W2411011, T2321005, 92169122 and 82473119), programs from the Ministry of Science and Technology of China (2024YFC2707400, 2021YFA1101000, 2022YFA1105200 and 2023YFA1800200), a Key R&D Program of Zhejiang Province (2023C03044 and 2024C03142), the Excellent Youth Fund of Jiangsu Province (BK20240148), Starry Night Science Fund of Zhejiang University Shanghai Institute for Advanced Study (SN-ZJU-SIAS-006), the Suzhou Innovation and Entrepreneurship Leading Talent Program (ZXL2022442 and ZXL2022505), the Suzhou Medical College Basic Frontier Innovation Cross Project (YXY2303027 and YXY2302017), the Joint Project of Pinnacle Disciplinary Group, the Second Affiliated Hospital of Chongqing Medical, and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions. We thank G. Xiao and Z. Lin from the Core Facilities, Zhejiang University School of Medicine for their technical support.
Author information
Authors and Affiliations
Contributions
F.X. and L.Z. conceived the project and designed the studies. F.X. and X.Z. performed the molecular, biochemical and mouse experiments. Y.R. and B.Y. assisted with the mass spectrometry data analysis. F.X., R.L. and J.Z. analysed the ATAC-seq, CUT&Tag and RNA-seq data. S.W. and P.S. conducted the protein purification experiments with assistance from H.Y. and H.L. H.R. provided valuable suggestions for the design of the peptide mutants. H.H. and X.M. provided patient samples. Z.M. prepared the nanoparticles and performed the cellular uptake, release and biodistribution studies. F.X., F.Z. and L.Z. wrote and revised the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors have filed a patent application on this work (202411045018.3).
Peer review
Peer review information
Nature thanks Denes Hnisz, Xavier Salvatella and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Extended Data Fig. 1 FOXM1 undergoes phase separation.
a, Images of GFP-FOXM1 droplets formation (left), quantification (middle) and phase separation (PS) diagram (right) at room temperature with the indicated NaCl concentrations (25 μM GFP-FOXM1 at pH 7.0); n = 5 fields from 3 independent experiments. b, Images of GFP-FOXM1 droplets formation (left) and quantification (right) at room temperature with the indicated pH (25 μM GFP-FOXM1 in 300 mM NaCl); n = 5 fields from 3 independent experiments. c, GFP-FOXM1 (25 μM) were treated with 5% 1,6-hexanediol (1,6-HD) and then subjected to droplet formation assay in vitro (300 mM NaCl, pH 7.0, room temperature); n = 5 fields from 3 independent experiments. d, Time-lapse micrographs and equivalent diameter (EqDiameter) of AF488-labelled FOXM1 (50 μM) liquid droplets formed at the indicated times with 300 mM NaCl at pH 7.0 at room temperature. e, Representative micrographs (left) and fluorescence recovery after photobleaching (FRAP). Quantification over a 50 s time course at room temperature (right) of GFP-FOXM1 (50 μM) droplets; n = 3 drops from 3 independent experiments. f, Quantified percentages (left) and average number (middle) of nuclear FOXM1 puncta as in Fig. 1g, n = 3 regions from 3 independent experiments; quantified average puncta size per cell (right) as in Fig. 1g, n = 20 cells from 3 independent experiments; cells were fixed. g, Quantified percentages (left) and average number (right) of nuclear FOXM1 puncta as in Fig. 1h, n = 5 regions from 3 independent experiments; cells were fixed. h, Endogenous FOXM1 staining with anti-FOXM1 antibody (green) in MDA-MB-231 cells treated with DMSO or 10% 1,6-HD for 5 min; n = 3 regions from 3 independent experiments; cells were fixed. i, Representative micrographs (left) and quantification (right) of purified GFP-FOXM1 wild-type (WT) and GFP-FOXM1 depleted of its N-terminalautorepressor domain (NRD), forkhead domain (FHD), intrinsically disordered region (IDR)1, IDR2 and transactivation domain (TAD) domain (25 μM); n = 5 fields from 3 independent experiments. j, Related to Fig. 1j: quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear FOXM1 puncta were shown; n = 5 regions from 3 independent experiments. k, Images of GFP-FOXM1-IDR1 droplets formation (left) and quantification (right) at room temperature with the indicated GFP-FOXM1-IDR1 concentrations (300 mM NaCl at pH 7.0); n = 5 fields from 3 independent experiments. l, m, Images of unlabeled (l) or AF488-labelled (m) FOXM1-IDR1 droplet formation (left) and quantification (right) at room temperature with the indicated protein concentrations with 300 mM NaCl at pH 7.0; n = 5 fields from 3 independent experiments. n, Iodixanol gradient ultracentrifugation analysis of TCL from HEK293T transfected with Myc-FOXM1 WT or d_IDR1 (top). Aliquots of the fractions were immunoblotted with anti-Myc antibody (bottom left). The band intensities were shown (bottom right). n = 3 independent experiments. o, SDD-AGE analysis of FOXM1 aggregation (top) and SDS-PAGE (bottom) of the TCL derived from HEK293T cells transfected with Myc-FOXM1 WT or d_IDR1. p, MST binding affinity between prokaryotic FOXM1-WT and FOXM1-WT (1 μM); FOXM1-WT and FOXM1-del_IDR1 (1 μM); FOXM1-del_IDR1 and FOXM1-del_IDR1 (1 μM); n = 3 independent experiments. q, MST binding affinity between FOXM1-IDR1 and FOXM1-WT (1 μM) or FOXM1-del_IDR1 (1 μM); n = 3 independent experiments. r, SDD-AGE analysis of endogenous FOXM1 aggregation (top) and SDS-PAGE (bottom) of the TCL derived from MDA-MB-231 cells. s, Purified FOXM1 WT or del_IDR1 proteins were imaged by electron microscopy using negative staining (left). Quantified average fiber length of proteins were shown as indicated (right); n = 5 fields from 3 independent experiments, Scale bar, 200 nm. t, Fluorescence of GFP-FOXM1 (20 μM) and Cy3-FKH DNA (500 nM) mixed at room temperature with 300 mM NaCl at pH 7.0 (top). Quantitative line profile of colocalization along a white arrow of the left image (bottom). u, GFP-FOXM1-d_NRD, FOXM1-d_IDR2, FOXM1-d_TAD (green) and Cy3-FKH DNA (FOXM1 forkhead DNA binding consensus sequence) (magenta) were mixed with the indicated module concentration and were imaged for fluorescence (left). PS diagram of different concentrations of indicated FOXM1 proteins with Cy3-FKH DNA (right). v, Quantitative (left) and representative images (right) of FRAP assay of GFP-FOXM1 WT (20 μM) with Cy3-FKH DNA (500 nM) (top) or GFP-FOXM1 WT (20 μM) alone (bottom); n = 3 drops from 3 independent experiments. w, Fold change of reporter gene activation in HEK293T cells transfected with TSC22D1 promoter- and 6DB promoter- luciferase (luc) constructs plus either empty vector, FOXM1 WT or FOXM1 d_IDR1 (left); quantitative PCR (qPCR) analysis of CDC25B and CyclinA2 mRNA level in HEK293T cells transfected with FOXM1 WT or FOXM1 d_IDR1 expression plasmids (right); n = 3 independent experiments. Data are representative of at least three independent experiments (a-c, e-n, p, q, s, v, w). Scale bar, 5 μm (a-e, h, i, k-m, t-v). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (a-c, f-m, s, w), two-way ANOVA (v). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 2 Hydrophobic residues in IDR1 are essential for FOXM1 LLPS.
a, Amino acid sequence analysis of IDR1 region. HM, hydrophobic motif. Pink denotes hydrophobic amino acids. IDR1-I (321-343 aa.), IDR1-II (344-364 aa.), IDR1-III (365-385 aa.). b, Percentage of hydrophobic amino acids in three subdomains of IDR1. c, Hydrophobicity profile of IDR1. The graph shows Kyte & Doolittle scale mean hydrophobicity values (scan window = 13). A scheme on the upper shows the three subdomains in IDR1. d, Domain structure of FOXM1-IDR1 truncations or fragment (left). The numbers above indicate the position of amino acid residues. Coomassie blue staining of bacterially purified GFP-FOXM1-IDR1-WT, D1, D2, D3 and III (right). e, MST analysis of IDR1-WT peptide (1 μM) binding to GFP-IDR1-WT, D1 (deletion aa. 321-344), D2 (deletion aa. 344-364), D3 (deletion aa. 365-385) and III (aa. 365-385); n = 3. f, Coomassie blue staining of GFP-FOXM1-IDR1-WT, HM3-4A (P369A/L370A/L371A/P372A), HM4-4A (L378A/V379A/P380A/I380A), HM5-3A (F383A/P384A/V385A) proteins purified from E. coli. g, MST analysis of prokaryotic purified GFP-IDR1-WT (1 μM) binding to IDR1-WT, HM3-4A, HM4-4A, HM5-3A; n = 3 independent experiments. h, Representative images (left) and quantification (right) of phase separation of GFP-FOXM1-IDR1-WT, HM3-4A, HM4-4A, HM5-3A (25 μM) at room temperature with 300 mM NaCl at pH 7.0; n = 5 fields from 3 independent experiments. i, Immunofluorescence and DAPI staining (left) and immunoblot analysis of the expression (top right) of HeLa cells transfected with GFP-tagged FOXM1 WT or HM3-4A, HM4-4A, HM5-3A (left). Quantified average number of the nuclear FOXM1 puncta and the percentage of cells with nuclear FOXM1 puncta were shown (bottom right); cells were fixed; n = 3 regions from 3 independent experiments. j, Coomassie blue staining of GFP-FOXM1-IDR1-WT or indicated single, double or triple mutations which included L370E, L378E, F383E, L370E/L378E, L370E/F383E, L378E/F383E, L370E/L378E/F383E, V379E and V385E proteins purified from E. coli. k, MST analysis of prokaryotic purified GFP-IDR1-WT (1 μM) binding to IDR1-WT or indicated mutants. Replacement of the hydrophobic residue (s) with negatively charged glutamic acid (E) in HM3-5, including L370E, L378E, F383E, L370E/L378E, L370E/F383E, L378E/F383E, L370E/L378E/F383E, V379E, V385E, could maximally disrupt the hydrophobic environment required for interaction. n = 3 independent experiments. l, Representative images (left) and quantification (right) of phase separation of GFP-FOXM1-IDR1-WT or indicated mutants (25 μM) at room temperature with 300 mM NaCl at pH 7.0; n = 5 fields from 3 independent experiments. m, Immunofluorescence and DAPI staining of HeLa cells transfected with GFP-tagged FOXM1 WT or indicated mutants (left). Quantified average number of the nuclear FOXM1 puncta and the percentage of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. Data are representative of at least three independent experiments (e, g, h, i, k-m). Scale bar, 5 μm (h, i, l, m). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (h, i, l, m). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 3 AMPK directly phosphorylates FOXM1 at S376 in IDR1.
a, Fold change of reporter gene activity in HEK293T cells transfected with TSC22D1 promoter- and 6DB promoter-luc constructs non-treated or treated with metformin (2 mM) or AICAR (0.5 mM) for 16 h; n = 3 ndependent experiments. b, qPCR analysis of CDC25B and CyclinA2 mRNA level in HEK293T cells non-treated or treated with metformin (2 mM) or AICAR (0.5 mM) for 16 h; n = 3 independent experiments. c, Immunofluorescence and DAPI staining in FOXM1-GFP stable HeLa cells treated with or without 2-DG (10 mM) for 16 h (left). Quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. d, Immunofluorescence and DAPI staining in FOXM1-GFP stable HeLa cells released from 16 h of glucose starvation (left). Quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. e-g, Endogenous FOXM1 staining with anti-FOXM1 antibody (green) in MDA-MB-231 cells (e), MDA-MB-468 cells (f) and HCC1937 cells (g) treated with with metformin (2 mM) or AICAR (0.5 mM) for 16 h (left). Quantified average number of nuclear FOXM1 puncta and the percentage of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. h, Diagram showing intracellular crosslinking mass spectrometric strategies targeting FOXM1 (Methods) (left). MS analysis of interacting proteins captured by the resultant covalent linkage in situ (right). i, Immunoblot (IB) of total cell lysate (TCL) and IP derived from HEK293T cells transfected with Myc-FOXM1-WT and Flag-tagged AMPKα1 or AMPKα2 plasmids. j, Immunoblot (IB) of input and immunoprecipitates (IP) of control IgG or anti-FOXM1 antibodies derived from HEK293T cells. k, AMPK kinase complex (α2/β2/γ2) was purified with Flag-beads and eluted by Flag peptide. Flag-AMPKα2 or its kinase-dead mutant (K47R) was co-expressed with HA-AMPKβ2 and HA-AMPKγ2 plasmid in HEK293T cells for 48 h. Illustration in panel k adapted from ref. 28, EMBO Press. l, Images of GFP-FOXM1 droplets formation pre-incubated with AMPK complex (α2/β2/γ2) or its kinase dead form (α2K47R/β2/γ2) purified from HEK293T cells for 1 h at room temperature in kinase buffer with 5 mM ATP and 10 mM MgCl2, and then dialyzed into droplet buffer with 300 mM NaCl (25 μM GFP-FOXM1, at pH 7.0, at room temperature). n = 3 fields from 3 independent experiments. m, Fold change in TSC22D1-luc and 6DB-luc activity in HEK293T cells transfected with Flag-AMPKα2 constitutively active form (CA) or its kinase-dead (KD) mutant (T172A) along with FOXM1 expression plasmids; n = 3 independent experiments. n, qPCR analysis of CDC25B and CyclinA2 mRNA level in HEK293T cells transfected with AMPKα2-CA or AMPKα2-KD along with FOXM1 expression plasmids; n = 3 independent experiments. o, Immunoblot (IB) of TCL and anti-Flag IP derived from HEK293T cells transfected with Myc-FOXM1 and the indicated Flag-AMPKα constructs. p, Immunoblot (IB) analysis of the TCL and anti-Myc IP derived from HEK293T cells transfected with Myc-FOXM1 and Flag-AMPKα2 constitutively active (CA) form or the Flag-active AMPK truncation kinase-dead (KD) mutant (T172A), as indicated. q, Immunoblot (IB) of TCL and IP using control IgG or anti-FOXM1 antibodies derived from control Ampkα WT and Ampkα1−/−α2−/− double knock-out (DKO) MEFs treated with metformin (2 mM) for 16 h. r, Immunoblot (IB) analysis of the TCL derived from AMPKα WT and Ampkα1−/−α2−/− double knock-out (DKO) MEFs treated with metformin (2 mM) for 16 h. s, SDS-PAGE of anti-Flag (FOXM1) immunoprecipitants from HEK293T cells with Coomassie brilliant blue staining. FOXM1 band at the corresponding location was indicated. t, Immunoblot (IB) of TCL and anti-Myc IP derived from HEK293T cells transfected with Myc-FOXM1 WT/S376A and Flag-AMPKα2-CA (left), or Myc-FOXM1 and Flag-AMPKα2-CA/KD (right). u, Immunoblot (IB) analysis of the TCL and IP of control IgG or anti-FOXM1 antibodies derived from AMPKα WT and Ampkα1−/−α2−/− double knock-out (DKO) MEFs treated with AICAR (0.5 mM) for 16 h. v, Left: design of guide RNA for CRISPR-Cas9 mutation of both copies of FOXM1 resulting in FOXM1 (Ser376A) in HeLa cells: upward arrow indicates the position of the codon in exon 7 that encodes Ser376; red underline (WT allele) indicates the codon (TCA) encoding Ser376; PAM (blue underline, WT allele) indicates the protospacer-adjacent motif; red font (in single-stranded oligodeoxynucleotide [SS ODN] and protein sequence) indicates the replacement codon (GCT) and amino acid (alanine [A]). Right: sequencing verification of the codon replacement by CRISPR-Cas9 resulting in FOXM1 (Ser376A) was shown. w, Immunoblot (IB) the TCL and anti-FOXM1 IP derived from the homozygous S376A knock-in HeLa cells treated with or without metformin (2 mM) for 16 h. Data are representative of at least three independent experiments (a-g, l-n). Scale bar, 5 μm (c-g, l). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (a-g, l-n). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 4 Ser376 phosphorylation abrogates FOXM1 LLPS.
a, Representative micrographs (left) and quantification (right) of GFP-FOXM1 droplet formation. Prokaryotic GFP-FOXM1-WT or S376A (25 μM) was incubated with or without eukaryotic purified AMPK complex (α2/β2/γ2) (1 μg) for 1 h at room temperature in kinase buffer with 5 mM ATP and 10 mM MgCl2, and then dialyzed into droplet buffer with 300 mM NaCl; n = 5 fields from 3 independent experiments. b, Representative images (left) and quantification (right) of phase separation of GFP-FOXM1 WT, Ser376Asp (S376D) or Ser376Glu (S376E) (25 μM) at room temperature with 300 mM NaCl; n = 5 fields from 3 independent experiments. c, Bacterially purified GFP-FOXM1 WT, S376A, S376D and S376E proteins were analyzed by Coomassie blue staining (left) and immunoblot (right). d, MST binding affinity between prokaryotic expressed GFP-FOXM1 WT or S376E (1 μM) and FKH-DNA. n = 3 independent experiments. e, Representative images (left) and quantification (right) of droplet formation of various concentrations of GFP-FOXM1 WT or S376E with Cy3-FKH DNA (500 nM); n = 5 fields from 3 independent experiments. f, FRAP quantification of the droplet fused by GFP-FOXM1 WT or S376E (10 µM) with Cy3-FKH DNA (500 nM) before and after photobleaching over 80 s (n = 3 droplets). n = 3 droplets. g, Flow chart of the major steps of generating Ser376-phosphorylated recombinant FOXM1 (left): (1) Step 1: sub-cloning of FOXM1 into pGEX-6p/His vector; (2) Step 2: mutating the codon of Ser376 to TAG to genetically incorporate Phosphoserine into recombinant FOXM1 (see Methods for detail); (3) Step 3: expression and purification of the S376-phosphorylated FOXM1 (FOXM1 S376-Phos). Coomassie blue staining of the indicated proteins purified (right). h, Schematic diagram describing the generation of site-specifically phosphorylated recombinant FOXM1 proteins with a phosphoserine tRNA (see in Methods). Schematic in panel h adapted from ref. 46, Cell Press. i, Immunoblot (IB) of the purified proteins with antibody specific to phosphorylated-Ser376 of FOXM1. The anti-Strep blots indicate loading of lanes. j, MST Binding affinity between IDR1 and FITC-IDR1 (1 μM) (IDR1/IDR1), IDR1 and FITC-pSer376-IDR1 (1 μM) (p-IDR1/IDR1), pSer376-IDR1 and FITC- pSer376-IDR1 (1 μM) (p-IDR1/p-IDR1); n = 3 independent experiments. k, FRAP quantification of droplets fused by the non-phosphorylated or Ser376-phosphorylated GFP-FOXM1 (10 µM) mixted with FKH-DNA (500 nM) before and after photobleaching over 50 s (n = 3 droplets). l, Immunofluorescence and DAPI staining in FOXM1-GFP HeLa stable cells treated with or without metformin (2 mM) or AICAR (0.5 mM) for 16 h (left). Quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. m, Immunofluorescence and DAPI staining of MDA-MB-231 cells transfected with GFP-tagged FOXM1 WT or S376E (left). Quantified average number of the nuclear FOXM1 puncta and the percentage of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. n, Immunofluorescence and DAPI staining of HeLa cells transfected with GFP-tagged FOXM1 WT or S376E (left). Quantified average number of the nuclear FOXM1 puncta and the percentage of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. o, p, Immunofluorescence and DAPI staining of MDA-MB-468 cells (o) or HCC1937 cells (p) transfected with GFP-tagged FOXM1 WT or S376E (left). Quantified average number of the nuclear FOXM1 puncta and the percentage of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. Data are representative of at least three independent experiments (a, b, d-f, j-p). Scale bar, 5 μm (a, b, e, l-p). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (a, b, e, l-p), two-way ANOVA (f, k). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 5 Ser376 phosphorylation abrogates FOXM1 LLPS.
a, Scheme for the iodixanol gradient ultracentrifugation of HEK293T transfected with Myc-tagged FOXM1 WT or S376E (top). Immunoblot (IB) of the aliquoted fraction with anti-Myc antibodies (bottom left) and quantified band intensities were shown (bottom right). n = 3 independent experiments. b, Iodixanol-gradient ultracentrifugation analysis of TCL from HEK293T cells pre-treated with control, AMPKα2 CA or KD (top). The FOXM1 band intensities were shown (bottom); n = 3 independent experiments. c, SDD-AGE analysis of FOXM1 aggregation (top) and SDS-PAGE (bottom) of the TCL derived from HEK293T cells transfected with Myc-FOXM1 WT or S376E. d, SDD-AGE analysis of FOXM1 aggregation (top) and SDS-PAGE (bottom) of the TCL derived from HEK293T cells transfected with plasmids expressing Myc-FOXM1 WT or S376A and Flag-AMPKα2 CA. e, Immunofluorescence of endogenous FOXM1 and DAPI staining in parental and FOXM1-S376A knock-in HeLa cells treated with control DMSO, metformin (2 mM) or AICAR (0.5 mM) for 16 h (left). Quantified average number of the nuclear FOXM1 puncta and the percentage of cells with nuclear puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. f, Colocalization of active RNA Pol II with FOXM1-GFP in the nuclear puncta of HeLa cells was detected using immunofluorescence staining with antibodies targeting either the active RNA Pol II, which is phosphorylated at Ser 2 (S2P) in its CTD (red), or GFP (green). The images were captured using super-resolution structured illumination microscopy (left). Quantitative line profile of colocalization along a white arrow of the left image (right); cells were fixed. g, Design of guide RNA for CRISPR–Cas9 mutation of both copies of FOXM1 resulting in FOXM1-S376E knock-in (left); sequencing verification of the codon replacement by CRISPR-Cas9 resulting in FOXM1 (Ser376E) was shown (right). h, SDD-AGE analysis of FOXM1 aggregation (top) and SDS-PAGE analysis (below) of TCL derived from parental and FOXM1-S376E knock-in HeLa cells. i, ChIP-qPCR analysis of FOXM1 binding to CENPA and CyclinA2 promoter with control IgG or anti-FOXM1 antibodies (left) and the qPCR analysis of CENPA and CyclinA2 mRNA expression (right) in parental or FOXM1-S376E knock-in HeLa cells; n = 3 independent experiments. j, Fold change in TSC22D1-luc and 6DB-luc activity (left) in HEK293T cells transfected without or with AMPKα2 CA along with control, wild-type FOXM1 (WT) or the catalytically inactive FOXM1 mutant (S376A) expression plasmids; n = 3 independent experiments. k, Iodixanol-gradient ultracentrifugation analysis of TCL from parental or FOXM1-S376E knock-in HeLa cells (top). Quantified FOXM1 band intensities were shown (bottom); n = 3 independent experiments. l, qPCR analysis (right) of CyclinA2 and CDC25B mRNA level in HEK293T cells transfected with or without AMPKα2 CA along with control, FOXM1 (WT) or FOXM1 mutant (S376A) expression plasmids; n = 3 independent experiments. Data are representative of at least three independent experiments (a, b, e, i-l). Scale bar, 5 μm (e, f). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (e, j, k, l). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 6 Phospho-mimicking mutation of FOXM1 (S376E) attenuates tumor cell growth and metastasis.
a, Immunoblot (IB) analysis of the parental and FOXM1 knock-out MDA-MB-231 cells stable expressing Myc-FOXM1 WT, S376A and S376E. b, Growth curves of FOXM1 knock-out MDA-MB-231 cells stable expressing Myc-FOXM1 WT, S376A and S376E; n = 3 independent experiments. c, The relative tumor volume (fold) (left) and tumor weights (right) at the endpoint (day 35) of a. n = 8 mice per group. d, Experimental procedure in vivo: nude mice were subcutaneously injected with FOXM1 WT or S376A MDA-MB-231 cells (2 × 106 cells per mouse) on day 0, followed by intraperitoneal injection of metformin (200 mg/kg) every other day as indicated (top). The visible tumors were measured at the indicated days. Bright view of tumors from each group at week 5 (bottom); n = 6 mice per group. e, Tumor volume from each group was measured at the indicated times in d. f, The relative tumor volume (fold) (left) and tumor weights (right) at the endpoint (day 35) of d. n = 6 mice per group. g, Immunoblot (IB) analysis of the parental and FOXM1 knock-out 4T1 cells stable expressing Myc-FOXM1 WT, S376A and S376E. h-j, Experimental procedure in vivo: nude mice were subcutaneously injected with FOXM1 knock-out 4T1 cells stably expressing FOXM1-WT or FOXM1-S376E (2 × 106 cells per mouse) on day 0 and tumors were grown for 4 weeks (h, top); n = 5 mice per group. The visible tumors were measured at the indicated days. Bright view of tumors from each group at week 4 (h, bottom). Tumor volume from each group was measured at the indicated times (i). The relative tumor volume (fold) (left) and tumor weights (right) at the endpoint (day 28) (j). k, l, Experimental analysis in vivo: 4T1.2 cells stably expressing FOXM1-WT or FOXM1-S376E (2 × 105 cells per mouse) were tail vein-injected into Balb/C mice (n = 5 mice per group). Lung metastasis was measured by bioluminescence imaging (BLI) (k). Photon flux (left) and representative images (right) in each experimental group followed in time (l). Illustration in panel k adapted from ref. 29, Springer Nature. m, Experimental procedure in vivo: nude mice were subcutaneously injected with FOXM1 knock-out 4T1 cells stably expressing FOXM1-WT or FOXM1-S376A (2 × 106 cells per mouse) on day 0, followed by intraperitoneal injection of metformin (200 mg/kg) every other day as indicated (top left). The visible tumors were measured at the indicated days. Bright view of tumors from each group at week 4 (bottom left); n = 6 mice per group. Tumor volume from each group was measured at the indicated times (right). Illustrations in panels d,h,m adapted from ref. 28, EMBO Press. n, The relative tumor volume (fold) (left) and tumor weights (right) at the endpoint (day 28) in m. n = 6 mice per group. o, Expression of p-S376-FOXM1 (left) and p-T172-AMPK (right) of normal tissues (n = 10) and breast cancer patients (n = 100). H score of p-S376-FOXM1 and p-T172-AMPK in every tissue sample was calculated. p, Expression of p-S376-FOXM1 (left) and p-T172-AMPK (right) of normal tissues (n = 10) and breast cancer patients (n = 100) at different malignant stage (I-III). H score of p-S376-FOXM1 and p-T172-AMPK in every tissue sample was calculated. q, Immunoblot (IB) of TCL and anti-FOXM1 IP derived from paired samples of non-tumorous breast (N) and breast tumor (T) from the same patient; n = 6 patients. Data are representative of at least three independent experiments (b). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (c, f, j, n-p) or two-way ANOVA (b, e, i, l, m). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 7 The chromatin landscapes and enhancer remodeling mediated by S376E knock-in reduce malignancy and improve immunogenicity of tumor cells.
a, Scatter plot (left) and enriched motifs (right) for FOXM1 CUT&Tag in parental HeLa cells. b, Normalized read distribution profiles of CUT&Tag signal across typical enhancers (TEs) (left) and super enhancers (SEs) (right) in parental and FOXM1-S376E KI HeLa cells. c, Venn diagram analysis shows the overlap of FOXM1-bound peaks and H3K27ac-bound peaks in parental and FOXM1-S376E-KI HeLa cells (left) as determined by CUT&Tag; The FOXM1-bound peak number (middle) and the percentage of H3K27ac-bound peaks co-occupied with FOXM1 (right) were shown. d, Venn diagram showing the overlap of increased binding genes occupied by FOXM1 and H3K27ac in parental cells (left) and FOXM1-S376E-KI cells (right) as determined by CUT&Tag. e, Gene ontology (GO) functional clustering of increased binding genes co-occupied by FOXM1 and H3K27ac in parental and FOXM1-S376E-KI cells in d. f, KEGG pathway analysis of different expressed genes between the parental and FOXM1-S376E-KI cells by RNA-seq. g, Pathway enrichment map for significantly enriched gene sets in FOXM1-S376E-KI cells; shown by pre-ranked gene-set enrichment analysis (GSEA). h, Immunofluorescence and quantification of endogenous MAGEB2 (left) and SPA17 (right) staining in parental and FOXM1-S376E-KI HeLa cells; cells were fixed; n = 3 regions from 3 independent experiments. i, Immunoblot (IB) analysis of parental and FOXM1-S376E-KI HeLa cells. j, Quantification of endogenous dsDNA in parental and FOXM1-S376E-KI HeLa cells in Fig. 3h (n = 3 regions from 3 independent experiments). k, Immunofluorescence (left) and quantification (right) of endogenous dsRNA and DAPI staining in parental and FOXM1-S376E-KI HeLa cells transfected without or with polyinosinic-polycytidylic acid (poly I:C) as indicated; cells were fixed; n = 3 regions from 3 independent experiments. l, cGAMP concentrations in cytoplasmic fractions of WT or cGAS−/− HT1080 transfected without or with cytoplasmic DNA isolated from FOXM1-S376E-KI HeLa cells, measured by cGAMP ELISA. n = 3 independent experiments. m, qPCR analysis of IFNB1, CCL5 and CXCL10 mRNA in parental and FOXM1-S376E-KI HeLa cells; n = 3 independent experiments. n, Quantification of endogenous γH2AX in parental and FOXM1-S376E-KI HeLa cells in Fig. 3k (n = 3 regions from 3 independent experiments). o, Immunofluorescence (left) and quantification (right) of endogenous dsDNA and DAPI staining of FOXM1 knock-out MDA-MB-231 cells stably expressing Myc-FOXM1 WT or S376E; cells were fixed; n = 3 regions from 3 independent experiments. p, Immunofluorescence (left) and quantification (right) of endogenous dsDNA and DAPI staining of control and FOXM1 knock-out MDA-MB-231 cells stably expressing Myc-FOXM1 WT or S376E and transfected without or with poly (I:C) as indicated; cells were fixed; n = 3 regions from 3 independent experiments. q, qPCR analysis of IFNB1, CCL5 and CXCL10 mRNA in FOXM1 knock-out MDA-MB-231 cells stably expressing Myc-FOXM1 WT or S376E; n = 3 independent experiments. r, Immunoblot (IB) analysis of the indicated proteins in FOXM1 knock-out MDA-MB-231 cells stable expressing Myc-FOXM1 WT or S376E. s, Immunofluorescence (left) and quantification (right) of endogenous γH2AX and DAPI staining in FOXM1 knock-out MDA-MB-231 cells stable expressing Myc-FOXM1 WT and S376E; cells were fixed; n = 3 regions from 3 independent experiments. t, Transwell migration assay (left) and transwell invasion assay (right) using FOXM1 knock-out MDA-MB-231 cells stable expressing Myc-FOXM1-WT or S376E. Representative figures from three independent experiments are shown. The quantification of migrated cells per field is shown in the bar graphs; Scale bar, 50 μm. u, T cell‐mediated cancer cell killing assay: FOXM1 knock-out MDA-MB-231 cells stable expressing Myc-FOXM1-WT or S376E co‐cultured with or without activated human peripheral blood mononuclear cells (PBMC) (see in Methods for detail) for 72 h were subjected to crystal violet staining (left). MDA‐MB‐231‐to‐T cell ratio = 1:3. Quantification of cytotoxicity of T cells (right). n = 3 wells from 3 independent experiments. Data are representative of at least three independent experiments (h, j-q, s, u). Scale bar, 10 μm (h, k, o, p, s). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (h, j-q, s-u) or Weighted Kolmogorov–Smirnov statistic test (g). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 8 Direct association of FIP4 to aa. 365-385 of FOXM1 in the IDR1 domain (FOXM1365-385) disrupts FOXM1 LLPS and inhibits the transcriptional activity of FOXM1.
a, Droplet formation (left) and quantification (right) of GFP-FOXM1 (20 μM) mixed with FIPs (5 μM); n = 5 fields from one of three independent experiments. b, MST analysis of prokaryotic purified GFP-FOXM1-WT and del-IDR1 (1 μM) binding to FIP4 or FIP3; n = 3 independent experiments. c, MST analysis of prokaryotic purified GFP-FOXM1-WT and IDR1 (1 μM) binding to FIP3 or FIP4; n = 3 independent experiments. d, e, MST binding affinity between FIP3 (d) or FIP4 (e) and the prokaryotic purified GFP-FOXM1-WT, D1, D2, D3 (1 μM); n = 3 independent experiments. f, g, MST analysis of prokaryotic purified GFP-IDR1-WT (1 μM) binding to FIP3 (f) or FIP4 (g) -WT or indicated mutants. Replacement of the hydrophobic residue (s) with negatively charged glutamic acid (E), including L378E, F383E, L378E/F383E, V379E or V385E, could maximally disrupt the hydrophobic environment required for interaction; n = 3 independent experiments. h, MST binding affinity between FIP3/FIP4 and FITC-labeled IDR1/p-IDR1 as indicated; n = 3 independent experiments. i, A schematic model describing FIP4 that bind to IDR1, causes electrostatic repulsion and thus dissolves the liquid phase separation of FOXM1. j, Comparison of size exclusion chromatography profiles of GFP-FOXM1 alone (Black) and GFP-FOXM1 mixed with FIP4 (0.5 μg/μl) (Red) run under identical buffer conditions on a SuperdexTM 200 increase 10/300 GL column. Approximately 1 mg GFP-FOXM1 was run on a SuperdexTM 200 analytical gel filtration column and UV traces measured at 280 nm were recorded and analyzed (top). Fractions corresponding to each peak were collected and analyzed on 10% denaturing SDS-PAGE followed by Coomassie staining (bottom). k, MST binding affinity between FKH-DNA and prokaryotic purified GFP-FOXM1 (1 μM) pre-incubated without (PBS) or with indicated FIPs (2.5 μM) for 2 h. l, Quantification of the droplet fuse by GFP-FOXM1 and Cy3-FKH DNA and incubated with indicatd FIPs in Fig. 4f; n = 5 fields from one of three independent experiments. m, Quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear FOXM1 puncta were shown in Fig. 4g; n = 3 fields from one of three independent experiments. n, FOXM1-GFP condensates in HeLa cells treated with PBS or FIPs (50 μM) for 16 h (left). Quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from one of three independent experiments. o, p, FOXM1-GFP condensates in MDA-MB-468 cells (o) or HCC1937 cells (p) treated with PBS or FIPs (50 μM) for 16 h (left). Quantified average number of nuclear FOXM1 puncta and the percentages of cells with nuclear FOXM1 puncta were shown (right); cells were fixed; n = 3 regions from 3 independent experiments. q, Iodixanol gradient ultracentrifugation analysis of TCL derived from MDA-MB-231 cells treaded with FIP4 (50 μM) for 16 h (top). Quantified FOXM1 band intensities were shown (bottom); n = 3 independent experiments. r, SDD-AGE analysis of FOXM1 aggregation (top) and SDS-PAGE of protein expression (bottom) in MDA-MB-231 cells treated with increasing amount of FIP4 (25 μM, 50 μM) for 16 h. s, Immunoblot (IB) analysis of streptavidin biotin-FKH-DNA pull-down derived from HEK293T cells transfected with Myc-FOXM1 and treated without (-) or with indicated FIPs (50 μM) for 16 h. t, Normalized CDC25B and Cyclin A2 mRNA expression (determined by qPCR analysis) in HEK293T cells transfected with FOXM1 expression plasmids and treated without (PBS) or with the indicated FIPs (50 μM) for 12 h (left).Fold change of TSC22D1-luc and 6DB-luc activity in HEK293T cells treated without or with FIPs (50 μM) for 24 h (right); n = 3 independent experiments. u, qPCR analysis of CENPB, CENPF, CDC25B, KRAS, MYC and TGFBR1 mRNA expressions in MDA-MB-231 cells non-treated (PBS) or treated with FIP4 (50 μM) for 48 h; n = 3 independent experiments. v, Representative images (left) and quantification (right) of tumorspheres at day 14. MCF10A-RAS cells treated without (PBS) or with FIP4 (50 μM) were analyzed in mammosphere assay. The number of mammospheres with a diameter > 60 μm was quantified. Scale bar 50 μm; n = 3 wells from one of three independent experiments. Data are representative of at least three independent experiments (a-h, k-q, t-v). Scale bar, 5 μm (a, n-p). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (a, l-p, t-v). Uncropped gel images are provided in Supplementary Fig. 1.
Extended Data Fig. 9 FIP4 inhibits malignancy and enhances immunogenicity in tumor cells.
a-b, Transwell migration assay (a) and transwell invasion assay (b) using MDA-MB-231 cells non-treated (control) or treated with FIP4 (50 μM) for 48 h. Representative figures from three independent experiments are shown. The quantification of migrated cells per field is shown in the bar graphs; Scale bar, 50 μm. c, Growth curves of MDA-MB-231 (left) and 4T1 (right) cells non-treated (PBS) or treated with FIP4 (50 μM); n = 3 independent experiments. d, Heatmap and volcano plot illustrating differentially regulated gene expression from RNA-seq analysis in MDA-MB-231 cells treated for 48 h with control PBS or FIP4 (50 μM) (left). Genes upregulated and downregulated are shown in red and blue, respectively (fold change > 2) (right). Values are presented as the log2 of tag counts. e, Gene ontology (GO) functional clustering of significantly changed genes in MDA-MB-231 cells stimulated with FIP4 (50 μM) for 48 h. f, Gene signatures indicating the breast cancer-related genome amplification, mutation of BRCA1 and squamous breast tumor are significantly enriched in control MDA-MB-231 cells and the decreased breast cancer relapse in lung are significantly enriched in MDA-MB-231 cells treated with FIP4 (50 μM) for 48 h; shown by pre-ranked gene set enrichment analysis (GSEA). g, Gene signatures representing increased tumor invasiveness, metastasis and oncogenic transcription were significantly enriched in control MDA-MB-231 cells versus the cells treated with FIP4 (50 μM) for 48 h; shown by pre-ranked gene set enrichment analysis (GSEA). h, The relative tumor volume (left) and tumor weight (right) at the endpoint (day 35) in Fig. 4j. n = 6 mice per group. i, Immunofluorescence microscopy of FOXM1 and DAPI staining of the tumor isolated from mice (left) as in Fig. 4j. Quantified average number of nuclear FOXM1 puncta, the percentages of cells with nuclear puncta and the average puncta size were shown (right); cells were fixed; n = 24 tumor cells pooled from 6 control mice and 6 FIP4-treated mice as in Fig. 4j. Scale bar, 5 μm. j-n, Experimental analysis in vivo: 4T07-met cells (5 × 105 cells per mouse) were tail vein-injected into nude mice (n = 5 mice per group). Mice were also given injections of PBS or FIP4 (10 mg kg −1; i.p.) per mouse three times one week (j). Lung metastasis was measured by bioluminescence imaging (BLI). Photon flux (k) and representative images (l) in each experimental group followed in time. Representative H&E stained lung sections from each group at week 4 (m). Scale bar, 1 mm. Percentage of lung metastasis area (left), average lung lesion surface area (middle) and number of lung metastasis nodules (right) per H&E stained lung section of all mice from each group at week 4 (n). o, Representative lung nodules from MMTV-PyMT mice injected with control and FIP4 at the endpoint (week 12) in Fig. 4n. Black arrowheads indicate lung metastasis nodules. Scale bar, 2 mm. Data are representative of at least three independent experiments (c, f, i, k, n). Mean ± s.d., statistical analysis was performed using two-way ANOVA (c, k), Wald test (d), Hypergeometric test (e),Weighted Kolmogorov-Smirnov statistic test (f, g) or two-tailed Student’s t-test (h, i, n).
Extended Data Fig. 10 Assessments of the FIP4-loaded nanoparticles in their biosafety, potential effectiveness in suppressing tumor malignancy and boosting tumor immunogenicity.
a, Transmission electron microscope (TEM) image of FIP4-NPs. Scale bar, 100 nm. b, Cy5.5-FIP4 released from Cy5.5-FIP4 containing NPs at different pH conditions using a dialysis strategy; n = 3 independent experiments. c, Intracellular FIP4 amount of 4T1 breast cancer cells after treatment with the Cy5.5-FIP4 (20 μM) or Cy5.5-FIP4 containing NPs for different time; n = 3 biologically independent experiments. d, Plasma concentration of FIP4 versus time after injection of FIP4 or FIP4-NPs (2.5 mg FIP4/kg; i.v., n = 3 mice per group). Data are expressed as percent injected dose per gram of tissue (% ID/g). e, Pathway enrichment map for significantly enriched gene sets in MDA-MB-231 cells treated with FIP4-NPs (20 μM) for 24 h versus control; shown by pre-ranked gene-set enrichment analysis (GSEA). f-h, Healthy Balb/c mice were injected with PBS or FIP4-NPs (10 mg kg−1; i.v.) every three days for 21 days. n = 3 mice per group. Body weight changes of the mice from each group (f). Blood markers of the liver (alanine transaminase, ALT; aspartate transaminase, AST) and kidney (creatinine, CREA; blood urea nitrogen, BUN) function in the serum harvested at the termination of the experiment from each group (g). Histological analysis of the major organs using H&E staining (h). Scale bar, i-l, Experimental analysis in vivo: 4T1.2 cells (2 × 105 cells per mouse) were tail vein-injected into Balb/C mice (n = 5 mice per group) on day 0 and tumors were grown for 3 weeks. Lung metastasis was measured by bioluminescence imaging (BLI). Mice were also given injections with FIP4-NPs (10 mg kg −1 FIP4; i.v.) and/or anti-PD1 (100 μg per mouse; i.v.) twice weekly for 3 weeks. Treatment protocol is summarized (i). photon flux in each experimental group followed in time (j). BLI imaging of representative mice (k) and representative lungs (l) in each group at day 21. Illustration in panel i adapted from ref. 29, Springer Nature. m, Representative H&E stained lung sections from each group at week 3 from mice as in Fig. 5d. Scale bar, 1 mm. n, Metastatic index (normalized metastatic tumor burden) was calculated by dividing the total metastasis area (mm2) from lungs of each animal by the primary tumor volume (cm3) as in Fig. 5d. Data are representative of at least three independent experiments (b, c). Mean ± s.d., statistical analysis was performed using two-tailed Student’s t-test (c, g, n), two-way ANOVA (b, j) or Weighted Kolmogorov-Smirnov statistic test (e).
Supplementary information
Supplementary information
This file contains Supplementary Notes 1-5, Supplementary Figures 1-5, legends for Supplementary Tables, legends to Supplementary Videos and Supplementary References
Supplementary Table 1
Summary of the mass spectrometry screen for phase-separated proteins in MDA-MB-231 breast tumor cells (related to Fig. 1b).
Supplementary Table 2
mRNA expression of antigens and ISGs in parental and S376E-knockin HeLa cells (related to Fig. 3g)
Supplementary Table 3
mRNA expression of antigens and ISGs in MDA-MB-231 breast tumor cells treated with or without FIP4 (50 μM) for 24 h (related to Fig. 4h); mRNA expression of antigens and ISGs in MDA-MB-231 breast tumor cells treated with 14A or FIP4 (50 μM) for 48 h (related to Supplementary Fig. 4g).
Supplementary Table 4
Primers list
Supplementary Video 1
Representative time-lapse fluorescence images of MDA-MB-231 cells transfected with FOXM1-GFP
Supplementary Video 2
Representative time-lapse fluorescence images of MDA-MB-468 cells transfected with FOXM1-GFP
Supplementary Video 3
Representative time-lapse fluorescence images of HCC1937 cells transfected with FOXM1-GFP
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Xie, F., Zhou, X., Ran, Y. et al. Targeting FOXM1 condensates reduces breast tumour growth and metastasis. Nature 638, 1112–1121 (2025). https://doi.org/10.1038/s41586-024-08421-w
Received:
Accepted:
Published:
Issue date:
DOI: https://doi.org/10.1038/s41586-024-08421-w
This article is cited by
-
Metformin modulates mitochondrial metabolism and epigenetic dysregulation in DNMT3A-mutated clones
Molecular Biomedicine (2025)
-
Biomolecular phase separation in tumorigenesis: from aberrant condensates to therapeutic vulnerabilities
Molecular Cancer (2025)
-
Regulation of RIG-I activity by phase separation reveals new therapeutic opportunities
Nature Cell Biology (2025)
-
Peptide blocks phase separation and oncogenic transcription
Nature Reviews Drug Discovery (2025)
-
Super-enhancers orchestrate transcriptional dysregulation and metabolic reprogramming in uveal melanoma
Communications Biology (2025)
