Fig 2: ELF5 is acetylated at K130, K134, K143, K197, K228 and K245.a Diagram outlining the steps involved in the preparation of sample for mass spectrometry (MS) to identify the acetylation lysine sites of ELF5. b Mass spectrometry identification of ELF5 acetylation sites in HEK293T cells transfected with expression plasmids for Flag-ELF5 and HA-p300. c MS analysis of the ELF5-derived peptides containing acetylated K130. d Diagram showing the acetylation sites of ELF5 that were mutated. e Effects of lysine to arginine at putative acetylated lysine residues on the acetylation of ELF5. Hela cells co-transfected with Flag-ELF5 WT or mutant Flag-ELF5 plus HA-p300 or not, cell lysis was immunoprecipitated with anti-Flag affinity gel and immunoblotted with the anti-AcK Ab. f Quantitative analyses of the acetylation of ELF5 WT and mutants. g The acetylation of ELF5 6KR was abolished. HEK293T cells co-transfected Flag-ELF5 WT or Flag-ELF5 6KR without and with HA-p300. Cell lysis was immunoprecipitated with anti-Flag affinity gel and immunoblotted with the anti-AcK Ab.

Fig 3: Acetylation promotes the proteasome-mediated degradation of ELF5.a Inhibition of deacetylation and protein synthesis on the protein level of ELF5. HEK293T cells were treated without and with 2 µM TSA and 5 mM NAM in the absence or presence of 20 µM MG132. Endogenous ELF5 was probed by anti-ELF5 Ab. b NAM- and TSA-mediated regulation of the ELF5 ubiquitination. Ubiquitination of Flag-ELF5 expressed in HEK293T cells the absence and presence of NAM + TSA with MG132. c Quantitation of the ubiquitination levels of Flag-ELF5 WT, Flag-ELF5 6KR, and Flag-ELF5 6KQ in HEK293T cells. d, e Half-lives of ELF5 and mutants. HEK293T cells expressing GFP-ELF5-WT, GFP-ELF5-6KR, or GFP-ELF5-6KQ were treated with 50 µM Cycloheximide, and the expression levels of these proteins were analyzed by western blot with anti-GFP Ab. The graph shown represents the ELF5 half-lives in Fig. 5e. Data are the means ± SDs from three determinations. *p < 0.05; **p < 0.01. f, g The expression of ELF5 and SIRT6 in human breast cancer tissues. Representative images tissue samples from breast cancer patients showing the expression levels of ELF5 and SIRT6 as assessed by immunohistochemistry. Both ELF5 and SIRT6 levels were classified as low, medium, or high based on the intensities of their IHC staining, and the percentages of patients classified in each category are depicted in the histogram in Fig. 5g.

Fig 4: Inhibition of breast cancer cell proliferation and tumorigenesis by ELF5 acetylation.a Growth curves of MCF7 cells that stably expressed GFP-ELF5 WT, GFP-ELF5 6KR, GFP-ELF5-6KQ as measured by CCK8 assay. Control cells were transfected with just the pEGFP-C1. Data are the means ± SDs from three determinations. *p < 0.05; **p < 0.01; ns, no significance. b Colony formation of MCF7 cells that stably expressed GFP-ELF5 WT, GFP-ELF5 6KR, GFP-ELF5-6KQ as measured by crystal violet staining assay. Control MCF7 cells were transfected with just the vector pEGFP-C1. Data are the means ± SDs from three determinations. *p < 0.05; **p < 0.01; ns, no significance. c Anchorage-independent growth activity of MCF7 cells as measured by soft-agar colony culture assay. Data are the means ± SDs from three determinations. *p < 0.05; **p < 0.01; ns, no significance. d Image showing the tumor formed in xenograft mice. e, f Weight and growth of tumor induced in mice by subcutaneously implanted MCF7 cells (5 × 106 cells injected per mice) that stably expressed GFP-ELF5 WT, GFP-ELF5 6KR, GFP-ELF5-6KQ. Control MCF7 cells were transfected with just the vector pEGFP-C1. Data are the means ± SDs from three determinations. *p < 0.05; **p < 0.01; ns, no significance.

Fig 5: Identification of ELF5 from p300 interaction network.a Silver-staining of p300 interacted protein. MCF7 cell lysis and T47D cell lysis were prepared and subjected to affinity purification using anti-p300 antibody or normal IgG. The purified proteins were resolved by SDS-PAGE and visualized by silver-staining. The protein bands were excised and analyzed by mass spectrometry. b ELF5 was immunoprecipitated with p300. Whole-cell lysates from T47D cells were subjected to co-immunoprecipitation performed with anti-p300 antibody or normal IgG followed by immunoblotting with performed with the indicated antibodies. Enrichment analysis (c) and cluster analysis (d) of biological processes according to the results of mass spectrometry. The graphs were downloaded from http://metascape.org/. e, f Comparison of ELF5 mRNA expression in healthy and various cancer tissues. Patients datasets were retrieved from the Oncomine database. OE: over-expressed, UE: under-expressed, Tot: total unique analyses. Numbers of significant analyses (p < 0.05; fold change >1.5) are shown inside the boxes and total numbers on the right side. Cell color indicates the best gene rank percentile for the analyses (red: over-expressed; blue: under-expressed). ELF5 mRNA expression as in Fig. 1e. Breast cancer datasets from Fig. 1e were displayed in details in Fig. 1f. g The expression of ELF5 in breast cancer based on breast cancer subclasses and data from http://ualcan.path.uab.edu/.