EHMT1 knockdown induces apoptosis and cell cycle arrest in lung cancer cells by increasing CDKN1A expression

Dozens of histone methyltransferases have been identified and biochemically characterized, but the pathological roles of their dysfunction in human diseases such as cancer remain largely unclear. Here, we demonstrate the involvement of EHMT1, a histone lysine methyltransferase, in lung cancer. Immunohistochemical analysis indicated that the expression levels of EHMT1 are significantly elevated in human lung carcinomas compared with non‐neoplastic lung tissues. Through gene ontology analysis of RNA‐seq results, we showed that EHMT1 is clearly associated with apoptosis and the cell cycle process. Moreover, FACS analysis and cell growth assays showed that knockdown of EHMT1 induced apoptosis and G1 cell cycle arrest via upregulation of CDKN1A in A549 and H1299 cell lines. Finally, in 3D spheroid culture, compared to control cells, EHMT1 knockdown cells exhibited reduced aggregation of 3D spheroids and clear upregulation of CDKN1A and downregulation of E‐cadherin. Therefore, the results of the present study suggest that EHMT1 plays a critical role in the regulation of cancer cell apoptosis and the cell cycle by modulating CDKN1A expression. Further functional analyses of EHMT1 in the context of human tumorigenesis may aid in the development of novel therapeutic strategies for cancer.


Introduction
Histone methylation plays important dynamic roles in regulating chromatin structure. Precise coordination and organization of open and closed chromatin are crucial for normal cellular processes such as DNA replication, repair, recombination, and transcription. Histone lysine methylation is considered to exert positive or negative regulatory effects on transcription depending on the methylation sites and status [1].
Despite a large body of information on the prominent role of histone lysine methyltransferases in transcriptional regulation, their involvement in human disease remains unclear.
EHMT1, also known as GLP, is mainly responsible for the monomethylation and dimethylation of histone H3 lysine 9 (H3K9), forming a heteromeric complex with G9a in euchromatin. Although G9a and EHMT1 can methylate H3K9 independently in vitro, the heteromeric complex appears to be essential for the methylation of euchromatin [2,3]. Several studies have indicated that methylation mediated by the G9a-EHMT1 complex is involved in transcriptional silencing, chromatin remodeling, and DNA methylation events, such as those affecting Mage-A2, in G9a and GLP knockout embryonic stem (ES) cells [4,5,2]. EHMT1 knockout mice show very early embryonic lethality and a substantial reduction in H3K9 monoand dimethylation as well as HP1 relocalization [6,3]. Moreover, loss-of-function mutations in EHMT1 are related to 9p34 subtelomeric deletion syndrome [7][8][9]. Evidence of an association with breast cancer has been found for variants of EHMT1 [10], although the physiological functions of this gene in cancer cells have not been elucidated.
Here, we demonstrate the possible involvement of EHMT1 in human tumorigenesis via direct regulation of CDKN1A by EHMT1. In a 3D spheroid culture system, EHMT1 knockdown interrupted 3D spheroid formation by upregulating CDKN1A expression. Therefore, EHMT1 could become a diagnostic and therapeutic marker for lung cancers, and our findings may contribute to the development of novel approaches for the treatment of cancers that overexpress EHMT1.

Cell culture and reagents
The human lung cancer cell lines H1299 and A549 were purchased from the Korean Cell Line Bank (Seoul, South Korea) and cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in a humidified atmosphere with 5% CO 2 at 37°C.

Cell viability assay
Cell Counting Kit-8 (CCK-8: Dojindo Laboratories, Rockville, MD, USA) assay was used to assess cell viability. Cells were seeded in 6-well plates starting at 2 9 10 5 cells per well and incubated for 24 h. After siRNA transfection for 48 h, a mixture of CCK-8 solution and RPMI-1640 medium supplemented with 10% FBS was added to each well and incubated in 5% CO 2 at 37°C for 10 min. The absorbance at 450 nm was assessed by a microplate reader. For crystal violet staining, the cells were fixed with cold 100% methanol for 5 min at À20°C and stained with 0.1% crystal violet solution (C0775, Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) for 5 min at room temperature after siRNA transfection for 48 h. Then, the cells were observed under a microscope (CELENA Ò S Digital Imaging System, Logos Biosystems, Anyang, Korea).

FACS analysis
To measure caspase-3/7 activity, we used the Muse Caspase-3/7 Kit (MCH100108, Luminex, Austin, TX, USA). According to the user guide, cells were treated with siRNA for 48 h, and Muse Caspase-3/7 working solution was then added. The cells were incubated for 30 min in a 37°C incubator with 5% CO 2 . After incubation, approximately 5 9 10 4 cells were analyzed with the Muse Cell Analyzer (Millipore, Burlington, MA, USA). To measure the apoptotic cell portion, the cells were collected after transfection with siRNA for 48 h and incubated with Muse Annexin V & Dead Cell Kit reagent (MCH100105, Millipore) for 20 min at room temperature. After incubation, approximately 5 9 10 4 cells were analyzed by the Muse Cell Analyzer (Millipore). To identify shifts in the proportion of cell cycles, approximately 1 9 10 4 cells were treated with siRNA for 48 h, fixed overnight, and incubated with Muse Cell Cycle Kit reagent (MCH100106, Luminex) for 20 min in accordance with the instruction manual. After incubation, the cells were analyzed by the Muse Cell Analyzer (Millipore). The FACS data obtained were analyzed by using MUSE 1.5 Analysis software (Millipore).

Immunohistochemical staining
An EnVision + kit/HRP kit (Dako, Carpinteria, CA, USA) was used. Paraffin-embedded sections of lung tumor specimens were processed in a microwave (90°C) with antigen retrieval solution (pH 9) (S2367; Dako), treated with a peroxidase-blocking reagent, and then treated with a protein-blocking reagent (K130, X0909; Dako). Tissue sections were incubated first with a rabbit anti-EHMT1 antibody (BIO MATRIX RESEARCH, Chiba, Japan) and then with an HRP-conjugated secondary antibody (Dako). Immunoreactivity was visualized with a chromogenic substrate (Liquid DAB Chromogen; Dako). Finally, tissue specimens were stained with Mayer's hematoxylin solution (Hematoxylin QS; Vector Laboratories, Burlingame, CA, USA) for 20 s to distinguish nuclei from the cytoplasm. Human lung cancer tissues were purchased from BioChain [11].

Immunocytochemistry
Cultured cells were fixed in 4% paraformaldehyde at room temperature for 10 min, permeabilized with 0.5% Triton X-100 (Sigma-Aldrich) in PBS for 10 min, and blocked with 5% bovine serum albumin in PBS for 30 min. Fixed cells were incubated with anti-p21 antibody (sc-6246, Santa Cruz Biotechnology, Inc.) overnight at 4°C and stained with an Alexa Fluorconjugated secondary antibody (Life Technologies). Fluorescent images were obtained by using the CELENA Ò S Digital Imaging System (Logos Biosystems).

3D cell culture
To perform spheroid culture of lung cancer cell lines, ultra-low-attachment microplates were used (Corning, Cat. 7007; Corning, NY, USA). Gene knockdown was performed by siRNA transfection; after transfection, cells were seeded into plates at concentrations of 1 9 10 5 and 1.2 9 10 5 cells per well for the H1299 and A549 cell lines, respectively. The spheroids were cultured for 3 days and observed under a microscope (CKX53, Olympus Corporation, Tokyo, Japan).

EHMT1 is overexpressed in lung cancer tissues
To assess EHMT1 expression levels, we performed tissue-wide expression profile analysis of EHMT1 in the Gene Expression Database of Normal and Tumor Tissues (GENT) (http://gent2.appex.kr/gent2/) and showed that the expression of EHMT1 was higher in several types of cancers than in normal tissues (Fig. S1). Additionally, EHMT1 was significantly overexpressed in lung cancer samples (n = 533) compared to normal lung samples (n = 59) according to RNA sequencing (RNA-seq) data derived from The Cancer Genome Atlas (TCGA) portal (Fig. 1A). In addition, to verify the overexpression of EHMT1 in lung cancer, we performed immunohistochemical analysis with a tissue microarray of lung cancer and normal lung tissues. As shown in Fig. 1B,C, we clearly observed the nuclear staining pattern in lung cancer tissue, and EHMT1 was significantly overexpressed in lung cancer tissues compared with normal lung and placental tissues. Taken together, these results indicate that EHMT1 is significantly overexpressed in lung cancer at both transcript and protein levels. Thus, we suggest that EHMT1 overexpression is closely associated with lung cancer proliferation.

EHMT1 knockdown suppressed proliferation in lung cancer cell lines
We applied siEHMT1 and siCont to cells and confirmed the knockdown of EHMT1 (Fig. 1D); next, we assessed cell growth by CCK-8 assay and crystal violet (CV) staining. The A549 and H1299 cell lines were treated with siEHMT1, and the growth of these lung cancer cell lines was suppressed by EHMT1 knockdown compared with siCont transfection (Fig. 1E,F). Next, to verify the function of EHMT1 in lung cancer, we performed RNA-seq analysis of siEHMT1-versus siCont-transfected H1299 cells. A total of 3093 dysregulated genes (2144 upregulated genes and 949 downregulated genes) (cutoff, 1.5-fold; P > 0.05) were identified (Table S1). We performed gene ontology (GO) term analysis of these differentially expressed genes (DEGs) using the ClueGO program in Cytoscape and found enrichment of apoptotic-related terms and metabolic process terms with EHMT1 knockdown ( Fig. 2A). Moreover, using the Database for Annotation, Visualization and Integrated Discovery (DAVID) version 6.8 (https://david.ncifcrf.gov/), we carried out GO term analysis of the genes that were up-and downregulated by EHMT1 knockdown. The 949 downregulated genes were enriched for cell cyclerelated terms (positive regulation of cell cycle, cell cycle arrest) and apoptosis-related terms (apoptotic process, regulation of apoptotic process). In addition, the 2144 upregulated genes also exhibited enrichment in cell cycle-and apoptosis-related terms (Fig. 2B). Thus, we suggest that the function of EHMT1 is

EHMT1 knockdown induced cell apoptosis and cell cycle arrest
To validate whether the suppression of cell growth by EHMT1 knockdown was related to apoptosis, we performed western blot analysis using an anti-PARP antibody. In Fig. 3A, the level of cleaved PARP was significantly increased by EHMT1 knockdown. Moreover, the caspase-3/7 activity assay revealed that the activity of caspase-3/7 was increased by EHMT1 knockdown compared with siCont transfection according to FACS analysis (Fig. 3B). Additionally, Annexin V staining showed that the rate of early and late apoptosis was increased in the EHMT1 knockdown groups compared to the siCont groups (Fig. 3C). Next, to assess the effect of EHMT1 on the cell cycle process, we performed FACS analysis in A549 and H1299 cell lines after transfection with siEHMT1. Figure 3D shows that G1 arrest was induced by EHMT1 knockdown in both cell lines compared to siCont. Taken together, these results indicate that EHMT1 is highly associated with the regulation of lung cancer proliferation. We hypothesized that EHMT1 plays an important role in lung cancer proliferation and could be a therapeutic target for lung cancer.

CDKN1A expression is regulated by EHMT1 in lung cancer cell lines
Next, we searched the RNA-seq results for potential EHMT1 targets, and we selected CDKN1A as a candidate that was upregulated by EHMT1 knockdown (Fig. 3E). CDKN1A is associated with cell cycle regulation and cell apoptosis in several types of cancers. As a cyclin-dependent kinase (CDK) inhibitor, CDKN1A plays an important role in the cell cycle, promoting cell cycle arrest and apoptosis in response to DNA damage. In cancer, CDKN1A acts as a tumor suppressor gene [12,13]. Under conditions of DNA damage, upregulation of CDKN1A by p53 suppresses tumor cell proliferation [14]. Thus, we established the hypothesis that induction of CDKN1A expression by EHMT1 knockdown induces cell cycle arrest and apoptosis in lung cancer cells. To confirm the upregulation of CDKN1A after EHMT1 knockdown, we performed qRT-PCR and western blot analysis after knockdown of EHMT1 in A549 and H1299 cell lines. As shown in Fig. 3F,G, we validated the upregulation of CDKN1A expression after EHMT1 knockdown in both cell lines. Moreover, in the immunocytochemical analysis, induction of CDKN1A expression by EHMT1 knockdown was observed in both cell lines (Fig. 3H). Taken together, these results demonstrated that EHMT1 knockdown increased CDKN1A expression in lung cancer cell lines and that upregulation of CDKN1A expression induced cell cycle arrest and apoptosis.
In the FACS analysis, we found the same recovery patterns with CDKN1A knockdown. Caspase-3/7 activity and Annexin V staining assays showed that the elevated caspase-3/7 activity and apoptosis rates were decreased to the levels of the siCont group upon cotransfection (Fig. 4C,D). Moreover, cell cycle analysis revealed that G1 arrest by EHMT1 knockdown was attenuated by CDKN1A knockdown in lung cancer cell lines (Fig. 4E). However, the knockdown of CDKN1A alone did not affect cell growth in either cell line (Fig. 4F). Next, to assess whether CDKN1A is a direct target for EHMT1, we performed a ChIP assay using a histone H3 lysine 9 dimethylation antibody after EHMT1 knockdown in A549 and H1299 cells. Figure 4G shows the ChIP primers for the promoter region of CDKN1A; our results showed that the status of H3K9 dimethylation was decreased in the promoter region of CDKN1A in EHMT1 knockdown cells compared to siCont cells (Fig. 4G). Thus, we suggest that downregulation of CDKN1A expression by epigenetic regulation of EHMT1 plays an important role in lung cancer proliferation.
3.6. EHMT1 knockdown reduced spheroid formation in 3D culture The 3D culture system can reflect the structural complexity of cancers and the tumor environment. Thus, 3D culture systems, such as spheroid culture and patient-derived cancer organoids, can be used as successful models for drug screening and target gene studies for cancer research [15,16]. In this study, we constructed 3D spheroid models with lung cancer cell lines using an ultra-low-attachment (ULA) plate system and assessed spheroid formation and target interactions with an EHMT1 knockdown model. Figure 5A shows that the cells treated with siCont formed spheroids well in 3D culture, but the EHMT1 knockdown cells exhibited a loosening of this 3D formation. Moreover, in the qRT-PCR analysis, the expression of E-cadherin increased in siCont cells compared to knockdown cells, suggesting that the cells were well aggregated for spheroid formation. However, in EHMT1 knockdown samples, E-cadherin expression was decreased in accordance with the decrease in spheroid formation. Additionally, as shown by 2D culture assays, CDKN1A expression was induced by EHMT1 knockdown (Fig. 5B). Next, to validate rescue by siCDKN1A, we cotransfected cells with siEHMT1 and siCDKN1A in 3D culture and found that the aggregation of spheroids was restored (Fig. 5A). Moreover, the expression level of E-cadherin was restored by cotransfection (Fig. 5B). Thus, we suggest that dysregulation of EHMT1 is clearly associated with lung cancer proliferation via regulation of CDKN1A.
Here, we confirmed that EHMT1 expression is upregulated in lung cancers via immunohistochemistry and that EHMT1 expression is upregulated in several types of cancers via analysis of GENT and RNA-seq lung cancer data from the TCGA portal. In addition, we observed growth suppression in lung cancer cell lines upon knockdown of EHMT1 via FACS analysis of Annexin V staining and caspase-3/7 activity. Moreover, we confirmed the effects of EHMT1 on lung cancer proliferation via GO analyses of RNA-seq data and found that it is particularly involved in the cell cycle and cell apoptosis. Collectively, our findings indicate that dysregulation of EHMT1 may play an important role in the process of lung cancer cell growth, and suggest that EHMT1 may be a rational drug target in the treatment of lung cancer. CDKN1A is regulated by EHMT1 in 3D spheroid culture system. (A) 3D spheroid formation assay. Cells transfected with siEHMT1, siCont, and siEHMT1/siCDKN1A were loaded onto ULA plates and incubated for 48 h. The cells were photographed under a microscope each day, Scale bar, 500 lm (left), and the size of spheroids was measured by the ImageView software program (right). Mean AE SD of three independent experiments. P values were calculated using Student's t-test (***P < 0.001). (B) qRT-PCR analysis of CDKN1A, CDH1, and EHMT1 expression after transfection with siEHMT1, siCont, and siEHMT1/siCDKN1A. Mean AE SD of three independent experiments. P values were calculated using Student's t-tests (**P < 0.01 and ***P < 0.001). 'Day 0' means the samples treated with siEHMT1 for 24 h. The results of the current study indicate that CDKN1A expression is epigenetically controlled by EHMT1 expression in lung cancer. Mono-, di-, and trimethylation of H3K9 are associated with the formation of the heterochromatin structure and with the transcriptional repression of activated genes. Thus, the direct targets of EHMT1 were upregulated upon EHMT1 knockdown. Herein, we found that EHMT1 knockdown upregulated CDKN1A expression and confirmed the reduction in H3K9 dimethylation at the promoter region of CDKN1A.
3D culture models have been recognized as alternative methods for anticancer drug development that mimic the physiological status and tumor formation process of cancer [15,16,27]. We used a 3D spheroid system to confirm the results from the 2D culture system study, implying that we could expect a similar anti-EHMT1 effect in a xenograft model derived from lung cancer cell lines. However, patient-derived lung cancer organoids are more effective than 3D spheroid systems for estimating the clinical effect of EHMT1 expression in lung cancer patients because cancer organoids more accurately resemble cancer cells and the tumor environment [28,29].

Conclusion
In this study, we found that EHMT1 is overexpressed in lung cancer and that EHMT1 modulates CDKN1A gene expression through the regulation of chromatin functions, subsequently promoting the proliferation of cancer cells by allowing them to evade apoptosis and cell cycle arrest (Fig. 6). Therefore, the development of EHMT1-specific inhibitors is needed for lung cancer and other cancers.

Supporting information
Additional Supporting Information may be found online in the Supporting Information section at the end of the article.