Enhanced migratory potential and invasiveness of cancer cells contribute crucially to cancer progression. These phenotypes are achieved by precise alteration of invasion-associated genes through local epigenetic modifications which are recognized by a class of proteins termed a chromatin reader. ZMYND8 [zinc finger MYND (myeloid, Nervy and DEAF-1)-type containing 8], a key component of the transcription regulatory network, has recently been shown to be a novel reader of H3.1K36Me2/H4K16Ac marks. Through differential gene expression analysis upon silencing this chromatin reader, we identified a subset of genes involved in cell proliferation and invasion/migration regulated by ZMYND8. Detailed analysis uncovered its antiproliferative activity through BrdU incorporation, alteration in the expression of proliferation markers, and cell cycle regulating genes and cell viability assays. In addition, performing wound healing and invasion/migration assays, its anti-invasive nature is evident. Interestingly, epithelial–mesenchymal transition (EMT), a key mechanism of cellular invasion, is regulated by ZMYND8 where we identified its selective enrichment on promoters of CLDN1/CDH1 genes, rich in H3K36Me2/H4K16Ac marks, leading to their up-regulation. Thus, the presence of ZMYND8 could be implicated in maintaining the epithelial phenotype of cells. Furthermore, syngeneic mice, injected with ZMYND8-overexpressed invasive breast cancer cells, showed reduction in tumor volume and weight. In concert with this, we observed a significant down-regulation of ZMYND8 in invasive ductal and lobular breast cancer tissues compared with normal tissue. Taken together, our study elucidates a novel function of ZMYND8 in regulating EMT and invasion of cancer cells, possibly through its chromatin reader function.

Introduction

Sustained proliferation and activated migration/invasion are two major hallmarks of cancer cells [1]. Normal tissues precisely control the cell growth and division cycle to ensure a homeostasis of cell number and thus maintain normal tissue architecture and function. There are distinct transcription programs for pluripotent stem cells and lineage committed cells of a particular tissue [2,3]. A fine balance in the levels of tumor suppressors and oncogenes, where the former is turned on and the latter is turned off, is instrumental in maintaining proper tissue identity. However, deregulated cell proliferation along with suppression of apoptosis primarily supports neoplastic transformation [1]. The epithelial–mesenchymal transition (EMT), originally described in the context of normal cell differentiation during early tissue development [4], contributes significantly in promoting cancer [5]. The epithelial cells that grow as a single layer/multilayer have a specific apical-basal polarity, adherent properties and cell/cell communication skills through specialized intracellular junctions [6]. During developmental stages and pathological conditions, the epithelial lining, which normally acts as a barrier constituting of tissues and organs, gets perturbed as they acquire mesenchymal properties. This transdifferentiation is mediated by key transcription factors like SNAIL, ZEB and bHLH proteins whose functions are regulated at several levels [7]. Distinct signal transduction pathways which can be modulated by extracellular cues are instrumental in reprogramming of gene expression during EMT [8].

Dynamicity in gene expression is largely controlled by posttranslational modifications of histones [9]. Several epigenetic regulators read chromatin by recognizing particular histone modifications through their specific domains and hence termed a ‘reader’ [10]. Many epigenetic readers have recently been found to be involved in a variety of diseases, including cancer [11]. ZMYND8 [zinc finger MYND (myeloid, Nervy and DEAF-1)-type containing 8], a recently discovered dual histone reader, has been shown to prevent metastasis in prostate cancer, as shown in mouse models and cell-line experiments [12]. Furthermore, loss of ZMYND8 (alias RACK7) causes overactivation of super-enhancers leading to tumorigenesis [13]. Along with this, ZMYND8 is a well-known component of the transcription coregulator complex and it exerts its function by being associated with several demethylase machinery components, including lysine demethylase 5A (KDM5A), KDM5C or LSD1 [14,15]. In addition, ZMYND8 plays a significant role in embryonic neural differentiation, through its ability to interact with Xenopus RCoR2 [16]. We have recently reported that ZMYND8 through its selective association with dual histone signatures stimulates the retinoic acid-responsive genes by assisting the recruitment of RNA Pol II phospho Ser5 [17]. Thus, ZMYND8, being a component of the transcription-initiation competent RNA polymerase II complex, influences the pathophysiological condition of the cell [17].

In the present work, we have demonstrated an antiproliferative and anti-invasive property of ZMYND8 in cervical and breast cancer cells. The genes critically involved in EMT are under the regulation of ZMYND8. We highlight the ability of ZMYND8 to positively regulate the expression of epithelial genes, through its chromatin reader function. Furthermore, we show that invasive ductal and lobular breast carcinoma has significantly reduced expression of ZMYND8. Taken together, we show for the first time that ZMYND8 is involved in maintaining the epithelial phenotype of cells through its histone binding ability and hence can be used as a selective candidate for reprogramming towards noninvasive mesenchymal–epithelial transition (MET) fate.

Materials and methods

Constructs

Full-length ZMYND8 and deletion of three domains of ZMYND8 [PHD (plant homeodomain), Bromo and PWWP; designated as ΔPBP-ZMYND8] were cloned in pCMV-FLAG using the Gateway Cloning System (Invitrogen) as described recently [17]. To overexpress ZMYND8, 1 µg of FLAG-ZMYND8 was transfected per 105 cells/well in a 6-well plate using Lipofectamine 2000 (Invitrogen). After 24 h of transfection, the cells were harvested for subsequent analysis.

Cell culture and transfection

HeLa and MCF7 cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM; Gibco, Invitrogen, Pittsburgh, PA, U.S.A.). Mouse metastatic breast cancer cells, 4T1, were maintained in DMEM. All media were supplemented with 10% fetal bovine serum (Gibco) and Pen-Strep (10 µl/ml of medium, Gibco) at 37°C in 5% (v/v) CO2.

To overexpress ZMYND8, 1 µg of FLAG-ZMYND8 was transfected per 105 cells/well in a 6-well plate using Lipofectamine 2000 (Invitrogen). After 24 h of transfection, the cells were harvested for subsequent analysis. For stable expression, FLAG-ZMYND8 was transfected into 4T1 cells using Lipofectamine 2000 and after 4 h, medium was replaced by fresh complete medium. After 48 h of transfection, fresh complete medium supplemented with G418 (Sigma) was added and cells were maintained in that medium. The stably transfected cells were thus selected in G418 (final concentration 400 µg/ml) and used for the in vivo experiment. The stable expression was verified in selected cells by quantitative real-time RT-PCR (Q-PCR) with the primer for ZMYND8.

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) assays were performed as per the standard protocol [18]. Briefly, cross-linking was done with 1% formaldehyde and then the reaction was stopped by 0.125 M glycine, followed by cell lysis in cell lysis buffer [5 mM PIPES (pH 8.0), 85 mM KCl and 0.5% NP40 (with a fresh protease inhibitor)]. This was followed by lysis with nuclei lysis buffer [50 mM Tris–HCl (pH 8.0), 10 mM EDTA and 1% SDS (with a fresh protease inhibitor)]. After sonication of the chromatin, pre-clearing was done. Immunoprecipitation was set with antibodies as mentioned in Table 1. Pre-blocked DYNA beads were put for binding to the pulled chromatin complex. Beads were washed with RIPA buffer, high-salt buffer, LiCl buffer and TE consecutively. Following RNaseA and Proteinase K treatment, the beads were kept for de-cross-linking at 65°C. Phenol–chloroform extraction followed by ethanol precipitation was performed. The DNA pellet was dissolved in H2O and used for Q-PCR analysis using gene specific primers (Table 2).

Table 1
List of antibodies used in ChIP, Western blot and immunofluorescence staining
Name of antibody used Application Dilution used in Western blot Company Catalog no. Lot no. 
H3K36me2 ChIP  Abcam ab9049 GR75522-3 
H4K16AC ChIP  Abcam Ab109463 GR53193-14 
H3 ChIP, Western blot 1 : 10 000 Abcam ab1791 GR265016-1 
H4 ChIP, Western blot 1 : 5000 Abcam ab10158 GR264160-1 
ZMYND8 ChIP, IF, Western blot 1 : 200 Sigma HPA020949 A96431 
FLAG-M2 ChIP, Western blot 1 : 1000 Sigma F1804 SLBK1346V 
β-Actin Western blot 1 : 2000 Sigma A5316 052M4793 
GAPDH Western blot 1 : 2000 Sigma G8795 080M4806 
VIM Western blot, IF 1 : 2000 Abcam Ab92547 GR219216-10 
N-cadherin Western blot 1 : 500 Novus NBP1-51612 131012 
Ki67 IF  Abcam ab15580 GR233420-2 
BrdU IF  BD 555627 01974 
Claudin1 IF  Cell Signaling Technology 13255 
Anti-mouse IgG-HRP Western blot 1 : 5000 Promega W402B 0000089661 
Anti-rabbit Alexa Fluor 488 IF – Invitrogen A11034 1073084 
Anti-rabbit Alexa Fluor 594 IF – Invitrogen A11037 1003217 
Anti-mouse Alexa Fluor 488 IF – Invitrogen A11029 1073083 
Name of antibody used Application Dilution used in Western blot Company Catalog no. Lot no. 
H3K36me2 ChIP  Abcam ab9049 GR75522-3 
H4K16AC ChIP  Abcam Ab109463 GR53193-14 
H3 ChIP, Western blot 1 : 10 000 Abcam ab1791 GR265016-1 
H4 ChIP, Western blot 1 : 5000 Abcam ab10158 GR264160-1 
ZMYND8 ChIP, IF, Western blot 1 : 200 Sigma HPA020949 A96431 
FLAG-M2 ChIP, Western blot 1 : 1000 Sigma F1804 SLBK1346V 
β-Actin Western blot 1 : 2000 Sigma A5316 052M4793 
GAPDH Western blot 1 : 2000 Sigma G8795 080M4806 
VIM Western blot, IF 1 : 2000 Abcam Ab92547 GR219216-10 
N-cadherin Western blot 1 : 500 Novus NBP1-51612 131012 
Ki67 IF  Abcam ab15580 GR233420-2 
BrdU IF  BD 555627 01974 
Claudin1 IF  Cell Signaling Technology 13255 
Anti-mouse IgG-HRP Western blot 1 : 5000 Promega W402B 0000089661 
Anti-rabbit Alexa Fluor 488 IF – Invitrogen A11034 1073084 
Anti-rabbit Alexa Fluor 594 IF – Invitrogen A11037 1003217 
Anti-mouse Alexa Fluor 488 IF – Invitrogen A11029 1073083 
Table 2
The sequence, respective amplicon size and Tm of the oligonucleotide primers used in ChIP-Q-PCR
Gene name Forward primer (5′–3′) Reverse primer (5′–3′) Tm (°C) 
hZEB1 CCTCACTTACATAGCTGGCAG TTCTTTGCCCTTGGAACATTTC 55 
hSNAI2 TCACCCAGAAATTCACCTCG TGCTTGTCTGTAAATCTGTGGG 55 
hSNAIL1 AACAAGGAATACCTCAGCCTG GCAAGCCCAGATTCACAAAAG 55 
hCDH1 ACCTCTTGACCGTTGCTAAG CACTGTCCTGTTCCTTTAGGTC 55 
hCLDN1 AATGGAACAGAGCACAAACATG CACTGAACAAAACCTACGCAC 55 
Gene name Forward primer (5′–3′) Reverse primer (5′–3′) Tm (°C) 
hZEB1 CCTCACTTACATAGCTGGCAG TTCTTTGCCCTTGGAACATTTC 55 
hSNAI2 TCACCCAGAAATTCACCTCG TGCTTGTCTGTAAATCTGTGGG 55 
hSNAIL1 AACAAGGAATACCTCAGCCTG GCAAGCCCAGATTCACAAAAG 55 
hCDH1 ACCTCTTGACCGTTGCTAAG CACTGTCCTGTTCCTTTAGGTC 55 
hCLDN1 AATGGAACAGAGCACAAACATG CACTGAACAAAACCTACGCAC 55 

Quantitative real-time RT-PCR

Total RNA was isolated from cells using TRI Reagent (Sigma) following the standard protocol. First-strand cDNA synthesis was done with the RevertAid fast-strand cDNA synthesis kit (Thermo Scientific) followed by the Q-PCR assay performed using Power SYBR GREEN mix (Applied Biosystems). The comparative CT method (ΔΔCT) was used to measure relative gene expression where the fold enrichment was calculated as: , where ΔCT is the CT of target gene subtracted from the CT of the housekeeping gene. The primer sequences are mentioned in Table 3.

Table 3
The sequence, respective amplicon size and Tm of the oligonucleotide primers used in Q-PCR
Gene name Forward primer (5′–3′) Reverse primer (5′–3′) Tm (°C) 
hZMYND8 CAGAAAATGAAACAGCCAGGG ACTTTGCATCAGCCAGGAAG 55 
hKi67 AAGTTCACACGGACGTCAG GATGCTCTTGCCATCTCC 60 
hPCNA GAGGCCTGCTGGGATATTAGC GGGTGAGCTGCACCAAAGAG 59 
hSNAIL1 TCGGAAGCCTAACTACAGCGA AGATGAGCATTGGCAGCGAG 60 
hCDH1 GTCACTGACACCAACGATAATCCT TTTCAGTGTGGTGATTACGACGTTA 60 
hCLDN1 TTGACTCCTTGCTGAATCTGAG TTCTGCACCTCATCGTCTTC 54 
hCLDN7 GGTGGAGGCATAATTTTCATCG ATGTTGGTAGGGATCAAAGGG 54 
hZEB1 AAGAAAGTGTTACAGATGCAGCTG CCCTGGTAACACTGTCTGGTC 60 
hVIM ACACCCTGCAATCTTTCAGACA GATTCCACTTTGCGTTCAAGGT 60 
hCCNE1 AAGGCCGAAGCAGCAAGTAT TGGCGTTTAAGTCCCCTGAC 54 
hCCNA2 TAAACAGCCTGCGTTCACCA ACACTCACTGGCTTTTCATCTTC 54 
hCCNB1 ACTTTCGCCTGAGCCTATTT TCCCGACCCAGTAGGTATTT 54 
mCLDN4 CAGTGCAAGATGTACGACTC TACCACTGAGAGAAGCATCC 55 
mSNAIL1 CTCATCTGGGACTCTCTCCT ACAGCGAGGTCAGCTCTAC 55 
mVIM CAGACAGGATGTTGACAATG ATCTCTTCATCGTGCAGTTT 55 
mZEB1 AGTCATGATGAAAACGGAAC TTCTCATGGCGGTACTTAAT 55 
mCLDN7 GAGATGACAAAGCGAAGAAG GACAATCTGATGACCAATCC 55 
mPCNA GGACTTAGATGTGGAGCAAC TATCACAACAGCATCTCCAA 55 
mKi67 GTTAGCCAAGACTCATCAGG CCTGTAACTGCTCCCTGTAG 55 
Gene name Forward primer (5′–3′) Reverse primer (5′–3′) Tm (°C) 
hZMYND8 CAGAAAATGAAACAGCCAGGG ACTTTGCATCAGCCAGGAAG 55 
hKi67 AAGTTCACACGGACGTCAG GATGCTCTTGCCATCTCC 60 
hPCNA GAGGCCTGCTGGGATATTAGC GGGTGAGCTGCACCAAAGAG 59 
hSNAIL1 TCGGAAGCCTAACTACAGCGA AGATGAGCATTGGCAGCGAG 60 
hCDH1 GTCACTGACACCAACGATAATCCT TTTCAGTGTGGTGATTACGACGTTA 60 
hCLDN1 TTGACTCCTTGCTGAATCTGAG TTCTGCACCTCATCGTCTTC 54 
hCLDN7 GGTGGAGGCATAATTTTCATCG ATGTTGGTAGGGATCAAAGGG 54 
hZEB1 AAGAAAGTGTTACAGATGCAGCTG CCCTGGTAACACTGTCTGGTC 60 
hVIM ACACCCTGCAATCTTTCAGACA GATTCCACTTTGCGTTCAAGGT 60 
hCCNE1 AAGGCCGAAGCAGCAAGTAT TGGCGTTTAAGTCCCCTGAC 54 
hCCNA2 TAAACAGCCTGCGTTCACCA ACACTCACTGGCTTTTCATCTTC 54 
hCCNB1 ACTTTCGCCTGAGCCTATTT TCCCGACCCAGTAGGTATTT 54 
mCLDN4 CAGTGCAAGATGTACGACTC TACCACTGAGAGAAGCATCC 55 
mSNAIL1 CTCATCTGGGACTCTCTCCT ACAGCGAGGTCAGCTCTAC 55 
mVIM CAGACAGGATGTTGACAATG ATCTCTTCATCGTGCAGTTT 55 
mZEB1 AGTCATGATGAAAACGGAAC TTCTCATGGCGGTACTTAAT 55 
mCLDN7 GAGATGACAAAGCGAAGAAG GACAATCTGATGACCAATCC 55 
mPCNA GGACTTAGATGTGGAGCAAC TATCACAACAGCATCTCCAA 55 
mKi67 GTTAGCCAAGACTCATCAGG CCTGTAACTGCTCCCTGTAG 55 

siRNA and transfection

RNA interference was carried out by siRNA against ZMYND8 (Santa Cruz Biotechnology, Dallas, TX, U.S.A.) or negative control siRNA (Invitrogen) using the INTERFERin transfection reagent (Polyplus) according to the manufacturer's protocol and incubated for 24 h.

Western blot

Whole cell extracts were prepared with Laemmeli buffer [4% SDS, 20% glycerol and 120 mM Tris–HCl (pH 6.8)] and sonicated followed by boiling at 100°C for 5 min. The samples were electrophoresed on 7.5 or 11% SDS–PAGE and transferred on the nitrocellulose membrane followed by blocking with nonfat skimmed milk for 1 h and probed for specific antibodies (with dilutions indicated in the product data sheets of the respective antibodies). Primary antibodies used are listed in Table 1.

In vitro invasion assay

Transwell membranes coated with Matrigel (BD Biosciences, U.S.A.) were used to assess in vitro invasion by ZMYND8 as mentioned previously [19]. Cells were transiently transfected with respective construct or siRNAs and allowed to recover overnight. The next day, cells were trypsinized and 2.5 × 105 transfected cells were added in the upper chamber and allowed to invade in the presence of 10% FBS in the lower chamber. After incubating for 22 h at 37°C in 5% CO2, the invaded cells on the lower surface were fixed, stained with toluidine blue and counted under the microscope. Three independent experiments were performed followed by statistical analysis.

Wound healing assay

Cells at 70% confluency were transfected with either ZMYND8 construct or ZMYND8 siRNA as described above. Scratching was carried out with a 200 µl pipette tip prior to the transfection and referred to as t = 0 h in the figure. Cells were washed several times with PBS to remove the detached ones and supplied with new growth medium. Photographs of the scratches were taken at 0 and 24 h using an inverted microscope (Nikon) equipped with a digital camera and the in-built software (Nikon application suite v3.0). 5-Flurouracil (5-FU) was used at final concentration of 1 mM [20].

Confocal microscopy

Immunofluorescence staining was performed as per the standard protocol [21]. Briefly, the cells were fixed with 4% paraformaldehyde, permeabilized with 1% Triton X-100 and blocked with 3% BSA. Cells were then incubated with anti-VIM, anti-Claudin1, anti-ZMYND8, anti-FLAG and anti-Ki67 antibody (dilutions as per product data sheet specifications) for 1 h. Following washes with PBST, the cells were incubated with Alexa Fluor-488- and Alexa Fluor-594-conjugated secondary antibodies for 1 h at RT. The coverslips were washed with PBS-T and mounted after DAPI staining. A Nikon Ti-E confocal microscope with an A1RMP scanner head was used for imaging. For BrdU staining, the cells were treated/transfected for 24 h followed by incubation with 18 µg/ml BrdU for 30 min at 37°C. Then, the cells were washed, harvested in PBS, fixed with ice-cold ethanol for 30 min and then with 2 M HCl prior to blocking with 3% BSA followed by staining with anti-BrdU antibody.

MTT assay

HeLa and MCF-7 (both 5 × 104 cells) were seeded on a 12-well plate, transfected or treated as necessary and incubated for 24 h. After that 100 µg/ml MTT was added to each well and incubated for another 4 h. Then, media were carefully removed and 150 µl of acidic solvent (4 mM HCl and 0.1% NP40 in isopropanol) was added to each well to dissolve the blue formazan product. The absorbance of this product was measured at 570 nm, using the ELISA plate reader (Stat Fax™® 2100 Microplate Reader, U.S.A.) and the background reading at 650 nM was subtracted. As a blank, the cells received 200 µl complete medium.

Immunohistochemistry of tissue

Immunohistochemistry (IHC) staining was performed with breast cancer tissue microarray slide (Abcam; ab178111, Cambridge, MA, U.S.A.; Table 4). The slides were de-waxed by heating at 70°C for 30 min followed by a 5 min wash in xylene and rehydration with 100%, 95% and 80% ethanol upto pure distilled water. The heat-induced antigen retrieval was done in 10 mM sodium citrate (pH 6.0) followed by blocking in universal blocking solution (Thermo). The slides were then incubated with anti-ZMYND8 (1 : 50) antibody (confirmed to be IHC grade as per the product data sheet information) for 1 h, followed by washing in TBS-T and incubating in HRP-tagged secondary antibody (1 : 200). The stain was developed with 3,3′-diaminobenzidine (Sigma). After counterstaining with hematoxylin, the sections were mounted with cover slips with DPX (Thermo) for microscopic imaging. The intensity of the staining was quantified using the ImageJ software.

Table 4
Histotypes of human breast tissue samples used to check the expression pattern of ZMYND8
Histotype No. of samples ZMYND8 staining 
Normal tissue 12 
Invasive ductal carcinoma 41 − 
Invasive lobular carcinoma − 
Lymph node metastasis 36 − 
Histotype No. of samples ZMYND8 staining 
Normal tissue 12 
Invasive ductal carcinoma 41 − 
Invasive lobular carcinoma − 
Lymph node metastasis 36 − 

Animals and syngenic mouse model

Female Balb/c mice (6–8 weeks old) were housed in individually ventilated cages under alternate dark and light cycles, and maintained on food and water. Mice were anesthetized and the fur was shaved over the flanks and 1 × 105 4T1 cells (stably transfected with empty-FLAG or FLAG-ZMYND8 construct) suspended in 0.1 ml PBS were subcutaneously injected into either flank. Primary tumors were visible within 2 weeks in control (FLAG-transfected) and after 3 weeks in FLAG-ZMYND8 transfected sets. After 4 weeks of injection, mice were killed; tumors were excised and the length (L) and width (W) were measured with a slide caliper. The volume of primary tumor is measured using the ellipsoid formula: (L × W2)/2 [22]. All animals were treated in accordance with the guidelines of the Institutional Animal Ethics Committee (approved by the CPCSEA, Ministry of Environment & Forest, Government of India). For each group, five mice were used for statistical significance.

Statistical analysis

All data were expressed as mean ± SEM with error bars. The statistical significance was calculated by two-tailed Student's t-test. A value of P < 0.05 was considered as significant. The experiments were done at least three times in duplicate unless otherwise stated.

Microarray analysis

MCF7 and HeLa cells were transfected with non-targeting (NT)- and ZMYND8-siRNA and after 24 h, total cellular RNA was extracted. Quality analysis and microarray with the isolated RNA were done as described in [17]. Raw data were quantile normalized and baseline transformation was carried out on the median of all samples using the GeneSpring GX 12.5 software (Agilent Technologies, Inc., Santa Clara, U.S.A.).

Differentially expressed genes

Differentially expressed probe sets (genes) upon ZMYND8 siRNA in comparison with NT siRNA were identified by applying Volcano Plot using a fold-change threshold of absolute fold-change ≥1.5 and a statistically significant t-test P-value threshold adjusted for false discovery rate of < 0.001. Statistically significantly enriched transcripts with a P-value adjusted for false discovery rate of <0.05 derived using the hypergeometric distribution test corresponding to differentially expressed genes were determined using Student's t-test with the Benjamini–Hochberg FDR test. Unsupervised hierarchical clustering of differentially expressed genes upon ZMYND8 siRNA in comparison with NT siRNA upon treatment in comparison with untreated was done using Euclidian algorithm with the Centroid linkage rule to identify gene clusters whose expression levels are significantly reproduced across the replicates.

Biological pathways and gene ontology enrichment analysis

A differentially expressed gene list was subjected to biological significance analysis by the GOElite tool (Bionivid Technology Pvt Ltd, Bangalore, India). A total of 21 887 protein-coding genes were used as the background and the differentially expressed gene list was used as a query. The database of Gene Ontology categories, Wiki pathways, KEGG Pathways, Pathway Commons, Pheno Ontology, Diseases, Protein Domains, Transcription factor targets and Tissue expression was configured for significance analysis. Each query list was subjected to over-representation analysis against each of the above databases. A Z-score and permutation or Fisher's exact test P-value are calculated to assess over-representation of enriched biological categories.

Results

ZMYND8 reduces proliferation of cancer cells

ZMYND8, an important component of the transcription co-regulator network, has recently been identified as a dual histone reader [17]. In an attempt to understand the cellular processes which can be affected by ZMYND8, we performed microarray analysis in breast cancer cells, MCF7, transfected with NT- and ZMYND8-siRNA. Statistical analysis of gene expression data showed 191 transcripts up-regulated and 196 transcripts down-regulated with a fold change of ≥1.5 and P-value of ≤0.05. Unsupervised hierarchical clustering of differentially expressed genes showed distinct patterns of up- and down-regulated transcripts upon treatment (Figure 1). Through detailed biological network analysis, the differentially expressed genes could be categorized into cell proliferation, apoptosis, tumor suppression, cancer promotion and EMT pathways (Figure 1). Interestingly, all these pathways were highlighted in a similar fashion while analyzed the microarray data obtained from cervical cancer cells, HeLa transfected with NT-/ZMYND8-siRNA (Supplementary Figure S1). Considering the importance of proliferation and invasion in promoting cancer, we attempted to decipher the involvement of ZMYND8 in these aspects. Knockdown of ZMYND8 increased the expression of known proliferation markers, Ki67 and PCNA, by ∼14-fold and ∼4-fold in HeLa (Figure 2A), ∼8- and ∼3-folds in MCF7 cells (Figure 2B) and ∼9- and ∼2.5-folds in T47D (Figure 2C), cells. In contrast, the overexpression of ZMYND8 in both HeLa (Figure 2D) and MCF7 (Figure 2E) cells reduced the expression of Ki67 (∼0.1-fold in both the cell lines) and PCNA (∼0.4- or 0.3-fold in MCF7 or HeLa, respectively) significantly. Reduction in Ki67 and PCNA was also observed in overexpressed T47D cells (Figure 2F). In 4T1 cells, similar alteration in expression of Ki67 could be detected upon overexpressing ZMYND8 (Figure 2G). These observations suggest the possible involvement of ZMYND8 in regulating cell proliferation. The reduction in Ki67 protein level was confirmed by IF imaging in ZMYND8-overexpressed HeLa cells (Figure 2H). Statistical quantification of Ki67 expression has been performed in ZMYND8 overexpressed cells compared with adjacent control cells in the same coverslips (Figure 2I). The up- and down-regulation of ZMYND8 expression by respective transfection of FLAG-ZMYND8 and ZMYND8-siRNA was verified at mRNA (Supplementary Figure S2A,B) and protein levels (Supplementary Figure S2Ci,ii).

ZMYND8 differentially regulates global gene expression in MCF7 cells.

Figure 1.
ZMYND8 differentially regulates global gene expression in MCF7 cells.

Microarray was performed with RNA isolated from MCF7 cells transfected with NT- and ZMYND8-siRNA. (A) Clustering and heat maps of expression values for differentially expressed genes are shown where down-regulated genes are marked in green, and up-regulated genes are marked in red. From left to right, first two samples are control siRNA-treated and latter two samples are ZMYND8 siRNA-treated MCF7 cells. (B) The biological network was built indicating its involvement in several cancer promoting/suppressive pathways. The color gradation of each gene is based on their fold change. P < 0.05.

Figure 1.
ZMYND8 differentially regulates global gene expression in MCF7 cells.

Microarray was performed with RNA isolated from MCF7 cells transfected with NT- and ZMYND8-siRNA. (A) Clustering and heat maps of expression values for differentially expressed genes are shown where down-regulated genes are marked in green, and up-regulated genes are marked in red. From left to right, first two samples are control siRNA-treated and latter two samples are ZMYND8 siRNA-treated MCF7 cells. (B) The biological network was built indicating its involvement in several cancer promoting/suppressive pathways. The color gradation of each gene is based on their fold change. P < 0.05.

ZMYND8 regulates proliferation of cancer cells.

Figure 2.
ZMYND8 regulates proliferation of cancer cells.

(A) The change in the expression of Ki67 and PCNA genes was measured by Q-PCR in HeLa (A), MCF7 (B) and T47D (C) cells transfected with 20 nM ZMYND8-siRNA (labeled as ZM8-si). The control cells were transfected with NT-siRNA. (D–G) HeLa, MCF7, T47D and 4T1 cells were transiently transfected with FLAG-ZMYND8 and Q-PCR assay was performed to check the change in Ki67 and PCNA. Relative gene expression is indicated as ‘fold’ change in the y-axis (mean + SEM). The statistical analysis is done as described previously. *P < 0.05. (H) HeLa cells were transiently transfected with FLAG-ZMYND8, co-stained with anti-FLAG and anti-Ki67 antibodies. (I) The intensity of Alexa-488-label in ZMYND8-overexpressed cells was calculated in the same coverslips with statistical significance based on cell counts from five randomly chosen fields.

Figure 2.
ZMYND8 regulates proliferation of cancer cells.

(A) The change in the expression of Ki67 and PCNA genes was measured by Q-PCR in HeLa (A), MCF7 (B) and T47D (C) cells transfected with 20 nM ZMYND8-siRNA (labeled as ZM8-si). The control cells were transfected with NT-siRNA. (D–G) HeLa, MCF7, T47D and 4T1 cells were transiently transfected with FLAG-ZMYND8 and Q-PCR assay was performed to check the change in Ki67 and PCNA. Relative gene expression is indicated as ‘fold’ change in the y-axis (mean + SEM). The statistical analysis is done as described previously. *P < 0.05. (H) HeLa cells were transiently transfected with FLAG-ZMYND8, co-stained with anti-FLAG and anti-Ki67 antibodies. (I) The intensity of Alexa-488-label in ZMYND8-overexpressed cells was calculated in the same coverslips with statistical significance based on cell counts from five randomly chosen fields.

ZMYND8-dependent cell cycle regulation was subsequently assessed. The ectopic overexpression of ZMYND8 caused significant reduction in three major cell cycle regulatory Cyclin genes, CCNE1 (involved in G1–S-phase progression), CCNA2 (involved in S–G2-phase and G2–M-phase transition) and CCNB1 (involved in G2–M-phase progression; Figure 3A), while ZMYND8-siRNA transfection up-regulated their expression (Figure 3B). Subsequently, we monitored BrdU incorporation following ZMYND8 gene silencing in newly proliferating cells by anti-BrdU antibody (Figure 3Ci). The incorporation of BrdU increased significantly in ZMYND8-knocked down cells when compared with adjacent control cells which had not taken up the siRNA (imaged in the same coverslip; Supplementary Figure S3A). Statistical analysis of similar cell populations clearly showed an enhanced uptake of BrdU in the absence of ZMYND8 (Figure 3D). Reciprocal experiments upon overexpressing ZMYND8 indicated a reduced BrdU uptake (Figure 3Cii), and the statistical analysis is represented in Figure 3E. Thereafter, ZMYND8-mediated reduced cell viability was confirmed by performing MTT assay, where overexpression of the same reduced proliferation of HeLa cells by ∼3-fold (P < 0.05; Figure 3F). In contrast, the cell proliferation was found to be enhanced significantly by ∼2-fold (P < 0.01; Figure 3F) upon knockdown of ZMYND8. Similar alteration in cell viability was observed in MCF7 cells by the MTT assay, where ectopic expression of ZMYND8 reduced cell proliferation to ∼2.5-fold (Figure 3G) and silencing of ZMYND8 enhanced cell proliferation by ∼2-fold (Figure 3G). All these results collectively confirm that ZMYND8 affects cell cycle and proliferation of cancer cells.

ZMYND8 regulates cell cycle of cancer cells.

Figure 3.
ZMYND8 regulates cell cycle of cancer cells.

(A and B) Q-PCR assay was performed with cDNAs of FLAG-ZMYND8 (A) and ZMYND8-siRNA (B)-transfected HeLa cells to assess the change in CCNE1, CCNA2 and CCNB1 genes with specific primers. (C) The BrdU incorporation was measured in HeLa cells transfected with ZMYND8-siRNA (i) and FLAG-ZMYND8 (ii) followed by staining with anti-BrdU antibody and immunofluorescence imaging. (D and E) The intensity of Alexa-488-label in ZMYND8-silenced (D) and -overexpressed (E) cells in the same coverslips with statistical significance based on cell counts from five randomly chosen fields. (F and G) MTT assay was performed with HeLa and MCF7 cells treated/transfected as indicated followed by measuring the absorbance at 570 nM.

Figure 3.
ZMYND8 regulates cell cycle of cancer cells.

(A and B) Q-PCR assay was performed with cDNAs of FLAG-ZMYND8 (A) and ZMYND8-siRNA (B)-transfected HeLa cells to assess the change in CCNE1, CCNA2 and CCNB1 genes with specific primers. (C) The BrdU incorporation was measured in HeLa cells transfected with ZMYND8-siRNA (i) and FLAG-ZMYND8 (ii) followed by staining with anti-BrdU antibody and immunofluorescence imaging. (D and E) The intensity of Alexa-488-label in ZMYND8-silenced (D) and -overexpressed (E) cells in the same coverslips with statistical significance based on cell counts from five randomly chosen fields. (F and G) MTT assay was performed with HeLa and MCF7 cells treated/transfected as indicated followed by measuring the absorbance at 570 nM.

Absence of ZMYND8 promotes EMT in cancer cells

EMT is considered to be a key step in promoting cancer. From the identified genes of microarray data, several key EMT markers are found to be differentially changed upon ZMYND8-knockdown, predicting its possible involvement in this process. Epithelial phenotype of cells is governed by claudin-1 (CLDN1), claudin-7 (CLDN7) and E-cadherin (CDH1), while mesenchymal phenotype is regulated by ZEB1, Snail1 (SNAI1), Slug (SNAI2), Vimentin (VIM) genes [7]. Transfection of ZMYND8-siRNA reduced the expression of CLDN1/7, CDH1 genes in both MCF7 (Figure 4A) and HeLa (Figure 4B) cells. In T47D cells, reduced expression of CDH1 and CLDN7 by ZMYND8-siRNA was observed (Figure 4C). In contrast, the expression of VIM, SNAI1 and ZEB1 were significantly increased in MCF7 (Figure 4A), HeLa (Figure 4B) and T47D (Figure 4C) cells. Furthermore, overexpression of ZMYND8 reduced the expression of SNAI1, ZEB1, VIM and increased CLDN1, CDH1 expression in MCF7 (Figure 4D), HeLa (Figure 4E) and T47D (Figure 4F) cells. A similar change was observed in 4T1 (Figure 4G) upon ectopic overexpression of ZMYND8. The reduction in mesenchymal phenotype upon overexpressing ZMYND8 in HeLa cells was observed through the change in N-cadherin and Vimentin (Figure 4H) expression by Western blotting.

Absence of ZMYND8 induces EMT in cancer cells.

Figure 4.
Absence of ZMYND8 induces EMT in cancer cells.

(AC) MCF7, HeLa and T47D cells were transfected with NT- and ZMYND8-siRNAs and the changes in the mRNA of CLDN1, CLDN7, ZEB1, SNAI1, SNAI2 and VIM were measured by Q-PCR assay. RNA isolation was done from the HeLa (D), MCF7 (E), T47D (F) and 4T1 (G) cells, transiently transfected with FLAG-ZMYND8, to quantify SNAI1, ZEB1, VIM, CDH1 and CLDN1 by Q-PCR with specific primers. Relative gene expression is indicated as ‘fold’ change in the y-axis (mean ± SEM). The statistical analysis is done as described previously. *P < 0.05. (H) Western blot confirmed the change in the respective protein level in ZMYND8-overexpressed HeLa cells. The intensity of the bands in Western blot was calculated using the ImageJ software and is presented along with the image.

Figure 4.
Absence of ZMYND8 induces EMT in cancer cells.

(AC) MCF7, HeLa and T47D cells were transfected with NT- and ZMYND8-siRNAs and the changes in the mRNA of CLDN1, CLDN7, ZEB1, SNAI1, SNAI2 and VIM were measured by Q-PCR assay. RNA isolation was done from the HeLa (D), MCF7 (E), T47D (F) and 4T1 (G) cells, transiently transfected with FLAG-ZMYND8, to quantify SNAI1, ZEB1, VIM, CDH1 and CLDN1 by Q-PCR with specific primers. Relative gene expression is indicated as ‘fold’ change in the y-axis (mean ± SEM). The statistical analysis is done as described previously. *P < 0.05. (H) Western blot confirmed the change in the respective protein level in ZMYND8-overexpressed HeLa cells. The intensity of the bands in Western blot was calculated using the ImageJ software and is presented along with the image.

In addition, IF imaging clearly indicates a reduction in epithelial marker Claudin and a concomitant increase in mesenchymal marker Vimentin in ZMYND8-overexpressed HeLa cells (Supplementary Figure S3B). Statistical quantification of Claudin1 and Vimentin expression have been performed in ZMYND8 overexpressed cells compared with adjacent control cells in the same coverslips (Supplementary Figure S3C). Collectively, all these results confirm an involvement of ZMYND8 in maintaining the epithelial cell fate. Silencing of ZMYND8, consequently, enhances the metastatic potential of the cancer cells.

ZMYND8 regulates EMT genes through recognizing H3K36Me2/H4K16Ac epigenetic signatures on respective promoters

We previously identified ZMYND8 as a chromatin reader which can selectively recognize H3K36me2/H4K16ac marks onto chromatin through its PWWP/Bromo/PHD (PBP) module [17]. Here, we attempted to understand whether ZMYND8-mediated regulation of gene expression is through its reader function. ChIP assay demonstrated a significant enrichment of ZMYND8 at upstream regions of CDH1 and CLDN1 genes (Figure 5A), while the mesenchymal genes had lesser abundance of ZMYND8 on their respective promoter regions. Further analysis with ChIP–Q-PCR uncovered a higher abundance of H3K36me2 (Figure 5B) and H4K16ac (Figure 5C) marks on CDH1 and CLDN1 promoters compared with their levels on upstream regions of SNAI1, SNAI2 and ZEB1. Next, we investigated the requirement of the chromatin-binding modules of ZMYND8 (PBP) in the selective recognition of these epigenetic signatures. The above-mentioned genes harboring H3K36me2 and H4K16ac had high occupancy of wild-type but not the PBP-deleted construct as seen by ChIP assays (Figure 5D). Here, the expression of FLAG-ZMYND8 and FLAG-ΔPBP-ZMYND8 was comparable, as verified by Western blot with anti-FLAG antibody using the lysates of HeLa cells transfected with respective constructs (Supplementary Figure S3D). Taken together, our results indicate that ZMYND8 regulates epithelial gene expression through its chromatin reader function.

ZMYND8 regulates the EMT-associated genes through its reader function.

Figure 5.
ZMYND8 regulates the EMT-associated genes through its reader function.

(AC) ChIP assay was performed with antibodies against ZMYND8, H3K36me2 and H4K16Ac showing their occupancy on the upstream regulatory regions of CDH1, CLDN1, SNAI1, SNAI2 and ZEB1 genes. The non-specific (NS) primer was used as a negative control. (D) The HeLa cells transiently transfected with FLAG-ZMYND8 or FLAG-ΔPBP-ZMYND8 were used for ChIP analysis with antibodies against FLAG. The normalization was done with either IgG- or H3/H4-immunoprecipitated DNA.

Figure 5.
ZMYND8 regulates the EMT-associated genes through its reader function.

(AC) ChIP assay was performed with antibodies against ZMYND8, H3K36me2 and H4K16Ac showing their occupancy on the upstream regulatory regions of CDH1, CLDN1, SNAI1, SNAI2 and ZEB1 genes. The non-specific (NS) primer was used as a negative control. (D) The HeLa cells transiently transfected with FLAG-ZMYND8 or FLAG-ΔPBP-ZMYND8 were used for ChIP analysis with antibodies against FLAG. The normalization was done with either IgG- or H3/H4-immunoprecipitated DNA.

Invasive property of cancer cells is enhanced upon down-regulation of ZMYND8

Considering the potential role of ZMYND8 in maintaining epithelial cell fate, we tried to elucidate its possible role in regulating cellular invasion and migration. Matrigel invasion assay demonstrated that ZMYND8 silencing resulted in ∼1.6- and ∼2-fold increases in invasiveness of HeLa (Figure 6A–C) and MCF7 (Figure 6B–D) cells, respectively. In contrast, ectopic overexpression of ZMYND8 reduced the invasiveness by ∼1.3- and ∼1.5-fold (P < 0.01) in HeLa (Figure 6A–C) and MCF7 (Figure 6B–D) cells, respectively. Subsequently, wound healing assay was performed with HeLa (Figure 6E) cells. Knockdown of ZMYND8 exhibited faster wound closure compared with the control set (Figure 6E, compare panels ii and vi). Conversely, overexpression of ZMYND8 inhibited the wound healing (Figure 6E, compare panels iv and vi). To check whether change in cell migratory potential (Figure 6E) is due to ZMYND8-induced alteration in cell proliferation, 5-FU, a potent inhibitor of cell proliferation, was applied simultaneously in transfected cells. Subtle reduction in cell migration was observed by 5-FU (compare Figure 6, panels ii and iii; iv and v), suggesting that the change in cell migration is predominantly governed by the abundance of ZMYND8 in cells. The results have been quantitatively represented in Figure 6F.

ZMYND8 regulates the invasion/migration of cancer cells.

Figure 6.
ZMYND8 regulates the invasion/migration of cancer cells.

(AD) Transwell invasion assay was performed with HeLa (A) and MCF7 (B) cells after transient transfection with FLAG-ZMYND8 or siRNA as indicated. Cells at three independent fields for each well were counted and plotted with error bar for HeLa (C) and MCF7 (D) cells. (E and F) Wound healing assay was performed with HeLa cells after transient transfection of ZMYND8 (ii) or siRNA (iii) in the presence or absence of 1 mM 5-FU in the time course of 24 h. t = 0 h at control cells (i) signifies the time of scratching the cells with pipette tips. The arrows indicate the width of wound and the assay was repeated three times independently. Scale bar: 100 µm. (F) The % recovery of the wound for each treated cells compared with untranfected control cells at 24 h was calculated and represented with statistical significance.

Figure 6.
ZMYND8 regulates the invasion/migration of cancer cells.

(AD) Transwell invasion assay was performed with HeLa (A) and MCF7 (B) cells after transient transfection with FLAG-ZMYND8 or siRNA as indicated. Cells at three independent fields for each well were counted and plotted with error bar for HeLa (C) and MCF7 (D) cells. (E and F) Wound healing assay was performed with HeLa cells after transient transfection of ZMYND8 (ii) or siRNA (iii) in the presence or absence of 1 mM 5-FU in the time course of 24 h. t = 0 h at control cells (i) signifies the time of scratching the cells with pipette tips. The arrows indicate the width of wound and the assay was repeated three times independently. Scale bar: 100 µm. (F) The % recovery of the wound for each treated cells compared with untranfected control cells at 24 h was calculated and represented with statistical significance.

Alteration in ZMYND8 levels in cancer

Having evaluated the biological implications of ZMYND8 overexpression in cancer cells, we next attempted to check the impact of ZMYND8-stable expression on the growth of primary tumors in vivo. For that, 4T1 cells were stably transfected with either FLAG or FLAG-ZMYND8 and selected by G418. The stable expression in selected cells was verified by Q-PCR (Supplementary Figure S2D). We observed that whereas control 4T1 cells (with the stable empty vector) formed tumors that grew steadily, 4T1-ZMYND8 cells caused tumors with reduced volume (Figure 7A). The comparative volume and weight of tumors induced by 4T1-FLAG and 4T1-ZMYND8 has been presented in Figure 7Bi and ii, respectively. We further monitored the expression of ZMYND8 by IHC from a commercially available tissue microarray slide (Abcam, ab178111) containing human breast normal and tumor tissue sections (summarized in Table 4). Interestingly, the normal tissues showed abundant expression of ZMYND8, while the ductal and lobular breast carcinomas showed significantly reduced expression (Figure 7C). The intensity of ZMYND8 staining is represented in Figure 7D upon quantification by the ImageJ software. Breast cancer with lymph node metastasis, which is highly aggressive in nature, also showed remarkable down-regulation in ZMYND8 expression. Since, the invasive breast tumors of ductal or lobular origin show an exclusion of ZMYND8, and an enhanced expression of ZMYND8 has a direct consequence in patient's survivability, futuristic studies can be designed to modulate ZMYND8 expression as a therapeutic strategy. Thus, we conclude that ZMYND8 attributes anti-invasive property to the cell and, consequently, misregulated expression of this epigenetic reader could be implicated in tumorigenicity.

ZMYND8 expression in breast cancer.

Figure 7.
ZMYND8 expression in breast cancer.

(A) 4T1-FLAG (stably transfected with empty-FLAG) and 4T1-ZMYND8 (stably transfected with FLAG-ZMYND8) were injected subcutaneously into female Balb/c mouse. The primary tumor of both sets after 4 weeks of injection is shown. (B) The volume (i) and weight (ii) of each tumor is presented. (C) The level of ZMYND8 was checked by IHC using array containing human breast tissue sections of normal individuals and breast cancer patients (Abcam, catalog #ab178111). The slide was developed with 3,3′-diaminobenzidine followed by hematoxylin staining of nuclei. Scale bar: 20 µm. (D) The quantification of ZMYND8 staining in each tissue type was performed with the ImageJ software and represented as bar diagram.

Figure 7.
ZMYND8 expression in breast cancer.

(A) 4T1-FLAG (stably transfected with empty-FLAG) and 4T1-ZMYND8 (stably transfected with FLAG-ZMYND8) were injected subcutaneously into female Balb/c mouse. The primary tumor of both sets after 4 weeks of injection is shown. (B) The volume (i) and weight (ii) of each tumor is presented. (C) The level of ZMYND8 was checked by IHC using array containing human breast tissue sections of normal individuals and breast cancer patients (Abcam, catalog #ab178111). The slide was developed with 3,3′-diaminobenzidine followed by hematoxylin staining of nuclei. Scale bar: 20 µm. (D) The quantification of ZMYND8 staining in each tissue type was performed with the ImageJ software and represented as bar diagram.

Discussion

During neoplastic transformation, the cells acquire increased migratory and invasive abilities and gradually become therapeutically challenging [5]. Such physiological alterations can be steered by epigenetic cues [23,24]. Previous reports have suggested a possible involvement of ZMYND8 in cancer, although the mechanism is not unraveled in detail. It has been reported to form fusion proteins with oncogene RELA in acute erythroid leukemia [25] or CEP250 in breast cancer [26,27]. These fusion constructs are proposed to interfere with the ability of ZMYND8 to associate with the NuRD complex which manifests in the disease pathology. ZMYND8 has also been implicated as a cutaneous T-cell lymphoma-associated antigen [28] as well as a novel candidate in high-risk human papillomavirus-induced cervical carcinogenesis [29]. In the present report, we attempt to decipher a mechanism by which ZMYND8 can regulate epithelial phenotype of cells by selective recognition of the epigenetic marks in the chromatin landscape.

Drawing a cue from our microarray studies, we found a significant involvement of ZMYND8 in a network of different biological processes, including cell proliferation, apoptosis, cancer promotion/suppression and EMT (Figure 1). Considering the significant involvement of induced cell proliferation and invasion in promoting cancer, we focused on deciphering the role of ZMYND8 in these aspects. Interestingly, ZMYND8 was observed to regulate the expression of proliferation markers namely Ki67 and PCNA (Figure 2). Furthermore, a significant alteration in cell cycle-regulating Cyclin gene expression (Figure 3) and BrdU incorporation was done by ZMYND8. In addition, cell viability assay also confirmed its active involvement in cell proliferation. Although no gross alteration in cell cycle progression (in osteosarcoma cell line U2OS; [30]) and cell proliferation (in ZR-75-30 ductal breast carcinoma cell line; [12]) was observed upon silencing of ZMYND8, our model (in HeLa and MCF7 cell lines) clearly indicates that the involvement of ZMYND8 in cell cycle regulation is more intricate and cell type specific.

Fatality from any cancer type is largely due to cell invasiveness and metastasis to distant locations. EMT is considered to be crucial in promoting invasive behavior of cancer cells. Claudin/Occludin genes are responsible for maintaining epithelial cell fate, and low expression of these genes is implicated in metastatic potential of cells [3133]. Transcription factor Snail is instrumental in blocking the expression of E-cadherin by direct binding to its promoter thereby promoting the mesenchymal phenotype [34]. Snail also induces the expression of transcription repressor ZEB1 which, in turn, represses the E-cadherin promoter [35]. Overexpression of Snail also reduces the expression of Claudin/Occludin [32]. Interestingly, our findings demonstrated that ZMYND8 regulates the expression of Claudin1, E-cadherin and Zeb/Snail1 genes in HeLa, MCF7 and 4T1 cells (Figure 4), thereby inhibiting EMT fate of cancer cells (Figure 5). Furthermore, ZMYND8 occupancy in concert with H3K36Me2/H4K16Ac modified histone signatures is quite robust in Claudin1 and E-cadherin gene promoters (Figure 6). A recent work shows that ZMYND8 (alias RACK7) in association with H3K4Me3-specific demethylase KDM5C regulates super-enhancers and loss of either factors leads to derepression of S100A oncogene in ZR-75-30 ductal carcinoma cells [13]. Furthermore, ZMYND8, in complex with JARID1D, acts as a repressor of metastasis-related genes [12], thereby eliciting antitumorigenic function. In both of these studies, ZMYND8 has been shown to be a corepressor of histone demethylases KDM5C or JARID1D. In contrast, our finding highlights the positive regulation of epithelial genes by this dynamic transcription factor through an anti-invasive mechanism. We have indeed observed enhanced cellular invasion ability upon silencing of ZMYND8 and a reciprocal situation upon its overexpression (Figure 6A–D). The wound closure assay also supported the inhibitory role of ZMYND8 in this process (Figure 6E,F). We subsequently assessed the antitumorigenic ability of ZMYND8 in mouse models, where ZMYND8 overexpressing 4T1 cells led to a significant reduction in tumor weight and volume (Figure 7A,B). These results indicate that indeed in the complex tumor microenvironment, overexpression of ZMYND8 can cause alteration in signal transduction pathways inhibiting the incessant growth of cancer cells and hence has a great future therapeutic potential. Furthermore, IHC analysis of human TMA demonstrated remarkably low expression of ZMYND8 in invasive luminal and ductal breast cancer when compared with the normal tissues (Figure 7C and Table 4). Cumulatively, ZMYND8 is an important player in epithelial cell fate maintenance and reprogramming the cells towards noninvasive MET phenotype could be modulated through this dynamic transcription factor. The cancer epigenetic-based drug discovery efforts could thus be applied in targeting the chromatin readers and their physiological co-complexes in cells. In this regard, specific small anchoring peptides that stabilize the histone modification–chromatin reader association could be developed as a novel therapeutic strategy against malignancy. As an alternative approach, the use of artificial transcription factors that could mimic the chromatin reader function and maintain the transcription of epithelial genes could be employed as an antitumorigenic mechanism.

Abbreviations

     
  • 5-FU

    5-flurouracil

  •  
  • bHLH

    basic Helix-loop-helix

  •  
  • BSA

    bovine serum albumin

  •  
  • CDH1

    E-cadherin

  •  
  • ChIP

    chromatin immunoprecipitation

  •  
  • CLDN1

    claudin-1

  •  
  • CLDN7

    claudin-7

  •  
  • DEAF-1′

    deformed epidermal autoregulatory factor-1

  •  
  • DMEM

    Dulbecco's Modified Eagle's Medium

  •  
  • EMT

    epithelial–mesenchymal transition

  •  
  • FBS

    fetal bovine serum

  •  
  • HRP

    horse radish peroxidase

  •  
  • IF

    immunofluorescence

  •  
  • IHC

    immunohistochemistry

  •  
  • KDM

    lysine demethylase

  •  
  • LSD

    Lysine-Specific Histone demethylase

  •  
  • MET

    mesenchymal–epithelial transition

  •  
  • MTT

    methylthiazolyl-tetrazolium

  •  
  • NT

    non-targeting

  •  
  • PBP

    PWWP/Bromo/PHD

  •  
  • NuRD

    nucleosome remodelling and deacetylase

  •  
  • PHD

    plant homeodomain

  •  
  • PIPES

    piperazine-N,N′-bis(2-ethanesulfonic acid)

  •  
  • Q-PCR

    quantitative real-time RT-PCR

  •  
  • RELA

    V-Rel avian reticuloendotheliosis viral oncogene homolog A

  •  
  • RIPA buffer

    Radioimmunoprecipitation assay buffer

  •  
  • SNAI1

    Snail1

  •  
  • SNAI2

    Slug

  •  
  • TBS-T

    tris-buffered saline with Tween-20

  •  
  • TMA

    tissue microarray

  •  
  • VIM

    Vimentin

  •  
  • ZEB1

    Zinc Finger E-Box Binding Homeobox 1

  •  
  • ZMYND8

    zinc finger MYND (myeloid, Nervy and DEAF-1)-type containing 8

Author Contribution

M.B. designed, performed experiments, analyzed data and drafted the manuscript. I.S. did the confocal imaging and analysis. D.K.S. helped in performing experimental procedures. M.W.K. and P.C. performed experiments on mouse model and reviewed the manuscript. S.R. designed experiments and drafted the manuscript. C.D. designed and performed experiments, and drafted the manuscript.

Funding

The study was supported by research grants entitled ‘Biomolecular Assembly, Recognition and Dynamics (BARD)’ project [Grant 12-R&D-SIN-5.04-0103] by the Department of Atomic Energy (DAE), Govt. of India and Science & Engineering Research Board (SERB) File No. EMR/2014/000335 by the Department of Science and Technology (DST), Govt. of India, provided to C.D. S.R. acknowledges Network Project (UNSEEN) funded by the Council of Scientific and Industrial Research, Govt. of India. C.D. and S.R. also acknowledge Ramalingaswami Fellowship provided by the Department of Biotechnology (DBT) and Ramanujan Fellowship provided by the Department of Science and Technology (DST), respectively. M.B. and M.W.K. acknowledge financial support from the INSPIRE faculty scheme of the Department of Science and Technology (DST), Govt. of India.

Acknowledgments

We acknowledge Prof. Jessica K. Tyler for critical comments on the manuscript. We thank Ms Payal Chakraborty (Bionivid Technology Pvt Ltd, Bangalore, India) for microarray data analysis.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Evan
,
G.I.
and
Vousden
,
K.H.
(
2001
)
Proliferation, cell cycle and apoptosis in cancer
.
Nature
411
,
342
348
doi:
2
Jiang
,
W.G.
,
Sanders
,
A.J.
,
Katoh
,
M.
,
Ungefroren
,
H.
,
Gieseler
,
F.
,
Prince
,
M.
et al. 
(
2015
)
Tissue invasion and metastasis: molecular, biological and clinical perspectives
.
Semin. Cancer Biol.
35
(
Suppl
),
S244
S275
doi:
3
Talmadge
,
J.E.
and
Fidler
,
I.J.
(
2010
)
AACR centennial series: the biology of cancer metastasis: historical perspective
.
Cancer Res.
70
,
5649
5669
doi:
4
Micalizzi
,
D.S.
,
Farabaugh
,
S.M.
and
Ford
,
H.L.
(
2010
)
Epithelial-mesenchymal transition in cancer: parallels between normal development and tumor progression
.
J. Mammary Gland Biol. Neoplasia
15
,
117
134
doi:
5
Heerboth
,
S.
,
Housman
,
G.
,
Leary
,
M.
,
Longacre
,
M.
,
Byler
,
S.
,
Lapinska
,
K.
et al. 
(
2015
)
EMT and tumor metastasis
.
Clin. Transl. Med.
4
,
6
doi:
6
Lamouille
,
S.
,
Xu
,
J.
and
Derynck
,
R.
(
2014
)
Molecular mechanisms of epithelial-mesenchymal transition
.
Nat. Rev. Mol. Cell Biol.
15
,
178
196
doi:
7
Kalluri
,
R.
and
Weinberg
,
R.A.
(
2009
)
The basics of epithelial-mesenchymal transition
.
J. Clin. Invest.
119
,
1420
1428
doi:
8
Wu
,
J.
,
Ivanov
,
A.I.
,
Fisher
,
P.B.
and
Fu
,
Z.
(
2016
)
Polo-like kinase 1 induces epithelial-to-mesenchymal transition and promotes epithelial cell motility by activating CRAF/ERK signaling
.
eLife
5
,
e10734
doi:
9
Dong
,
X.
and
Weng
,
Z.
(
2013
)
The correlation between histone modifications and gene expression
.
Epigenomics
5
,
113
116
doi:
10
Taverna
,
S.D.
,
Li
,
H.
,
Ruthenburg
,
A.J.
,
Allis
,
C.D.
and
Patel
,
D.J.
(
2007
)
How chromatin-binding modules interpret histone modifications: lessons from professional pocket pickers
.
Nat. Struct. Mol. Biol.
14
,
1025
1040
doi:
11
Chi
,
P.
,
Allis
,
C.D.
and
Wang
,
G.G.
(
2010
)
Covalent histone modifications—miswritten, misinterpreted and mis-erased in human cancers
.
Nat. Rev. Cancer
10
,
457
469
doi:
12
Li
,
N.
,
Li
,
Y.
,
Lv
,
J.
,
Zheng
,
X.
,
Wen
,
H.
,
Shen
,
H.
et al. 
(
2016
)
ZMYND8 reads the dual histone mark H3K4me1-H3K14ac to antagonize the expression of metastasis-linked genes
.
Mol. Cell
63
,
470
484
doi:
13
Shen
,
H.
,
Xu
,
W.
,
Guo
,
R.
,
Rong
,
B.
,
Gu
,
L.
,
Wang
,
Z.
et al. 
(
2016
)
Suppression of enhancer overactivation by a RACK7-Histone demethylase complex
.
Cell
165
,
331
342
doi:
14
Malovannaya
,
A.
,
Lanz
,
R.B.
,
Jung
,
S.Y.
,
Bulynko
,
Y.
,
Le
,
N.T.
,
Chan
,
D.W.
et al. 
(
2011
)
Analysis of the human endogenous coregulator complexome
.
Cell
145
,
787
799
doi:
15
Eberl
,
H.C.
,
Spruijt
,
C.G.
,
Kelstrup
,
C.D.
,
Vermeulen
,
M.
and
Mann
,
M.
(
2013
)
A map of general and specialized chromatin readers in mouse tissues generated by label-free interaction proteomics
.
Mol. Cell
49
,
368
378
doi:
16
Zeng
,
W.
,
Kong
,
Q.
,
Li
,
C.
and
Mao
,
B.
(
2010
)
Xenopus RCOR2 (REST corepressor 2) interacts with ZMYND8, which is involved in neural differentiation
.
Biochem. Biophys. Res. Commun.
394
,
1024
1029
doi:
17
Adhikary
,
S.
,
Sanyal
,
S.
,
Basu
,
M.
,
Sengupta
,
I.
,
Sen
,
S.
,
Srivastava
,
D.K.
et al. 
(
2016
)
Selective recognition of H3.1K36 dimethylation/H4K16 acetylation facilitates the regulation of All-trans-retinoic acid (ATRA)-responsive genes by putative chromatin reader ZMYND8
.
J. Biol. Chem.
291
,
2664
2681
doi:
18
Basu
,
M.
,
Khan
,
M.W.
,
Chakrabarti
,
P.
and
Das
,
C.
(
2017
)
Chromatin reader ZMYND8 is a key target of all trans retinoic acid-mediated inhibition of cancer cell proliferation
.
Biochim. Biophys. Acta, Gene Regul. Mech.
1860
,
450
459
doi:
19
Basu
,
M.
,
Mukhopadhyay
,
S.
,
Chatterjee
,
U.
and
Roy
,
S.S.
(
2014
)
FGF16 promotes invasive behavior of SKOV-3 ovarian cancer cells through activation of mitogen-activated protein kinase (MAPK) signaling pathway
.
J. Biol. Chem.
289
,
1415
1428
doi:
20
Pariente
,
R.
,
Pariente
,
J.A.
,
Rodríguez
,
A.B.
and
Espino
,
J.
(
2016
)
Melatonin sensitizes human cervical cancer HeLa cells to cisplatin-induced cytotoxicity and apoptosis: effects on oxidative stress and DNA fragmentation
.
J. Pineal Res.
60
,
55
64
doi:
21
Das
,
C.
,
Gadad
,
S.S.
and
Kundu
,
T.K.
(
2010
)
Human positive coactivator 4 controls heterochromatinization and silencing of neural gene expression by interacting with REST/NRSF and CoREST
.
J. Mol. Biol.
397
,
1
12
doi:
22
Lou
,
Y.
,
McDonald
,
P.C.
,
Oloumi
,
A.
,
Chia
,
S.
,
Ostlund
,
C.
,
Ahmadi
,
A.
et al. 
(
2011
)
Targeting tumor hypoxia: suppression of breast tumor growth and metastasis by novel carbonic anhydrase IX inhibitors
.
Cancer Res.
71
,
3364
3376
doi:
23
Wang
,
Y.
and
Shang
,
Y.
(
2013
)
Epigenetic control of epithelial-to-mesenchymal transition and cancer metastasis
.
Exp. Cell Res.
319
,
160
169
doi:
24
Lujambio
,
A.
and
Esteller
,
M.
(
2009
)
How epigenetics can explain human metastasis: a new role for microRNAs
.
Cell Cycle
8
,
377
382
doi:
25
Panagopoulos
,
I.
,
Micci
,
F.
,
Thorsen
,
J.
,
Haugom
,
L.
,
Buechner
,
J.
,
Kerndrup
,
G.
et al. 
(
2013
)
Fusion of ZMYND8 and RELA genes in acute erythroid leukemia
.
PLoS ONE
8
,
e63663
doi:
26
Edgren
,
H.
,
Murumagi
,
A.
,
Kangaspeska
,
S.
,
Nicorici
,
D.
,
Hongisto
,
V.
,
Kleivi
,
K.
et al. 
(
2011
)
Identification of fusion genes in breast cancer by paired-end RNA-sequencing
.
Genome Biol.
12
,
R6
doi:
27
Wada
,
Y.
,
Matsuura
,
M.
,
Sugawara
,
M.
,
Ushijima
,
M.
,
Miyata
,
S.
,
Nagasaki
,
K.
et al. 
(
2014
)
Development of detection method for novel fusion gene using GeneChip exon array
.
J. Clin. Bioinforma
4
,
3
doi:
28
Eichmuller
,
S.
,
Usener
,
D.
,
Dummer
,
R.
,
Stein
,
A.
,
Thiel
,
D.
and
Schadendorf
,
D.
(
2001
)
Serological detection of cutaneous T-cell lymphoma-associated antigens
.
Proc. Natl Acad. Sci. U.S.A.
98
,
629
634
doi:
29
Bierkens
,
M.
,
Krijgsman
,
O.
,
Wilting
,
S.M.
,
Bosch
,
L.
,
Jaspers
,
A.
,
Meijer
,
G.A.
et al. 
(
2013
)
Focal aberrations indicate EYA2 and hsa-miR-375 as oncogene and tumor suppressor in cervical carcinogenesis
.
Genes Chromosom. Cancer
52
,
56
68
doi:
30
Gong
,
F.
,
Chiu
,
L.-Y.
,
Cox
,
B.
,
Aymard
,
F.
,
Clouaire
,
T.
,
Leung
,
J.W.
et al. 
(
2015
)
Screen identifies bromodomain protein ZMYND8 in chromatin recognition of transcription-associated DNA damage that promotes homologous recombination
.
Genes Dev.
29
,
197
211
doi:
31
Kwon
,
M.J.
(
2013
)
Emerging roles of claudins in human cancer
.
Int. J. Mol. Sci.
14
,
18148
18180
doi:
32
Suh
,
Y.
,
Yoon
,
C.-H.
,
Kim
,
R-.K.
,
Lim
,
E.-J.
,
Oh
,
Y.S.
,
Hwang
,
S.-G.
et al. 
(
2013
)
Claudin-1 induces epithelial-mesenchymal transition through activation of the c-Abl-ERK signaling pathway in human liver cells
.
Oncogene
32
,
4873
4882
doi:
33
Philip
,
R.
,
Heiler
,
S.
,
Mu
,
W.
,
Buchler
,
M.W.
,
Zoller
,
M.
and
Thuma
,
F.
(
2015
)
Claudin-7 promotes the epithelial–mesenchymal transition in human colorectal cancer
.
Oncotarget
6
,
2046
2063
doi:
34
Ikenouchi
,
J.
,
Matsuda
,
M.
,
Furuse
,
M.
and
Tsukita
,
S.
(
2003
)
Regulation of tight junctions during the epithelium-mesenchyme transition: direct repression of the gene expression of claudins/occludin by Snail
.
J. Cell Sci.
116
,
1959
1967
doi:
35
Guaita
,
S.
,
Puig
,
I.
,
Franci
,
C.
,
Garrido
,
M.
,
Dominguez
,
D.
,
Batlle
,
E.
et al. 
(
2002
)
Snail induction of epithelial to mesenchymal transition in tumor cells is accompanied by MUC1 repression and ZEB1 expression
.
J. Biol. Chem.
277
,
39209
39216
doi:

Author notes

*

Present address: Biophysics Department, Bose Institute, P 1/12, CIT Road, Kankurgachi, Kolkata- 700054, India.

Supplementary data