In Silico/In Vitro Hit-to-Lead Methodology Yields SMYD3 Inhibitor That Eliminates Unrestrained Proliferation of Breast Carcinoma Cells

SMYD3 is a lysine methyltransferase that regulates the expression of over 80 genes and is required for the uncontrolled proliferation of most breast, colorectal, and hepatocellular carcinomas. The elimination of SMYD3 restores normal expression patterns of these genes and halts aberrant cell proliferation, making it a promising target for small molecule inhibition. In this study, we sought to establish a proof of concept for our in silico/in vitro hit-to-lead enzyme inhibitor development platform and to identify a lead small molecule candidate for SMYD3 inhibition. We used Schrodinger® software to screen libraries of small molecules in silico and the five compounds with the greatest predicted binding affinity within the SMYD3 binding pocket were purchased and assessed in vitro in direct binding assays and in breast cancer cell lines. We have confirmed the ability of one of these inhibitors, Inhibitor-4, to restore normal rates of cell proliferation, arrest the cell cycle, and induce apoptosis in breast cancer cells without affecting wildtype cell behavior. Our results provide a proof of concept for this fast and affordable small molecule hit-to-lead methodology as well as a promising candidate small molecule SMYD3 inhibitor for the treatment of human cancer.


Introduction
SMYD (SET and MYND domain-containing) protein family members constitute a class of methyltransferases that regulate a wide range of normal cellular processes [1][2][3] and are also involved

Inhibitor-4 Decreases SMYD3-Mediated H3 Methylation
After our iterative in silico screening using Schrodinger software (Glide ® , Maestro ® , LigPrep ® , and Epik ® ), we purchased the top five hit compounds for testing. Hits were defined as the drug-like small molecules with the lowest free binding energy when docked in the protein-target binding pocket of SMYD3. The predicted free binding energies of the five lead compounds ranged from −7.2 kJ/mol to −9.1 kJ/mol, compared to the natural protein ligand's predicted free binding energy of only around −1 kJ/mol (fragment of VEGFR1). We used an in vitro methylation assay using purified Histone 3 (H3) to assess the ability of the five lead in silico-designed SMYD3 inhibitor candidates to decrease SMYD3 enzymatic activity. We demonstrated that Compound 4 (Inhibitor-4) significantly reduces SMYD3-mediated Histone 3 methylation (70% reduction), while the other novel compounds did not show significant differences. H3 was chosen because of previous studies that demonstrated SMYD3 methylates H3 preferentially ( Figure 1) [3]. Human epithelial breast cancer cells High Western blot, RT-qPCR H4K5, H3K4 [8,14,[18][19][20] MCF10A Human epithelial breast cells Very low Western blot, RT-qPCR H4K5, H3K4 [18,19]

Inhibitor-4 Decreases SMYD3-Mediated H3 Methylation
After our iterative in silico screening using Schrodinger software (Glide ® , Maestro ® , LigPrep ® , and Epik ® ), we purchased the top five hit compounds for testing. Hits were defined as the drug-like small molecules with the lowest free binding energy when docked in the protein-target binding pocket of SMYD3. The predicted free binding energies of the five lead compounds ranged from −7.2 kJ/mol to −9.1 kJ/mol, compared to the natural protein ligand's predicted free binding energy of only around −1 kJ/mol (fragment of VEGFR1). We used an in vitro methylation assay using purified Histone 3 (H3) to assess the ability of the five lead in silico-designed SMYD3 inhibitor candidates to decrease SMYD3 enzymatic activity. We demonstrated that Compound 4 (Inhibitor-4) significantly reduces SMYD3-mediated Histone 3 methylation (70% reduction), while the other novel compounds did not show significant differences. H3 was chosen because of previous studies that demonstrated SMYD3 methylates H3 preferentially ( Figure 1) [3].  Compound 4 is from here on referred to as Inhibitor-4. Error bars display standard error of means. Statistically significant differences from control are indicated by ** p < 0.01, *** p < 0.001 or ns p > 0.05.

Inhibitor-4 and BCl-121 are Stable in d 6 -DMSO Solution
Because of the limited solubility of selected molecules in aqueous solution and in media, we dissolved BCI-121 and Inhibitor-4 in d 6 -DMSO solution to record and analyze the 1D 1 H NMR spectra of both compounds. The major species attributed to Inhibitor-4 and BCI-121 were observed at time 0 and 24 h, as shown in Figure S1a (BCI-121) and Figure S1b (Inhibitor-4). The 1 H NMR peaks of the fresh and aged samples for Inhibitor-4 showed no observable difference in the presence of the major component (67%) and minor component (33%) peaks as a function of time, suggesting that no hydrolysis is taking place during the experiment for Inhibitor-4. For BCI-121, 70% of the major species was present at time 0, however after 24 h this decreased slightly to 68%, suggesting that the positive control may be slightly less stable than Inhibitor-4.

SMYD3 Is Overexpressed in Breast Cancer Cells
Western blot and immunocytochemistry were carried out to test the expression levels of SMYD3 using anti-SMYD3 antibody in normal and breast cancer cell lines. Western blot data have indicated that SMYD3 was highly expressed in breast cancer cell lines (1.8-fold in MCF7 and 2.6-fold in MDA-MB-231) compared to normal cell line ( Figure 2).

Inhibitor-4 and BCl-121 are Stable in d6-DMSO Solution
Because of the limited solubility of selected molecules in aqueous solution and in media, we dissolved BCI-121 and Inhibitor-4 in d6-DMSO solution to record and analyze the 1D 1 H NMR spectra of both compounds. The major species attributed to Inhibitor-4 and BCI-121 were observed at time 0 and 24 h, as shown in Figure S1a (BCI-121) and Figure S1b (Inhibitor-4). The 1 H NMR peaks of the fresh and aged samples for Inhibitor-4 showed no observable difference in the presence of the major component (67%) and minor component (33%) peaks as a function of time, suggesting that no hydrolysis is taking place during the experiment for Inhibitor-4. For BCI-121, 70% of the major species was present at time 0, however after 24 h this decreased slightly to 68%, suggesting that the positive control may be slightly less stable than Inhibitor-4.

SMYD3 Is Overexpressed in Breast Cancer Cells
Western blot and immunocytochemistry were carried out to test the expression levels of SMYD3 using anti-SMYD3 antibody in normal and breast cancer cell lines. Western blot data have indicated that SMYD3 was highly expressed in breast cancer cell lines (1.8-fold in MCF7 and 2.6-fold in MDA-MB-231) compared to normal cell line ( Figure 2).   Additionally, immunocytochemistry data have shown elevated levels of SMYD3 expression in breast cancer cell lines comparing to normal cell line (Figure 2a,b). Therefore, increased SMYD3 expression could be correlated with breast carcinogenesis.

Inhibitor-4 Inhibits Growth of Breast Cancer Cells
The impact of SMYD3 inhibitors on growth of breast cancer cells was tested by adding 50, 100 and (c) Western blot analysis shows fold change in SMYD3 expression in the cell lines. (d) Expression of SMYD3 protein using immunocytochemistry. (e) immunocytochemistry analysis shows SMYD3 intensity in the cell lines. Values are mean ± standard error of the means. Statistically significant differences from control are indicated by * p < 0.05, ** p < 0.01. Additionally, immunocytochemistry data have shown elevated levels of SMYD3 expression in breast cancer cell lines comparing to normal cell line (Figure 2a,b). Therefore, increased SMYD3 expression could be correlated with breast carcinogenesis.

Inhibitor-4 Inhibits Growth of Breast Cancer Cells
The impact of SMYD3 inhibitors on growth of breast cancer cells was tested by adding 50, 100   For MCF10A (normal) cells, the effect of the SMYD3 inhibitors was limited. The basal doubling time for MCF10A was 28 h. Interestingly, no delay was noticed with 50, or 100 µM concentrations of either inhibitor. Treatment of the normal cells with 200 µM of Inhibitor-4 resulted in a minor, not significant, growth delay (approximately 5%), while treatment with 200 µM BCI-121 resulted in a major growth delay (Figure 3e,f). These results suggest that Inhibitor-4 shows more growth inhibition than BCI-121 and causes significant inhibition in cancer cell growth while only modestly impacting healthy cells.

Inhibitor-4 Suppresses Breast Cancer Cell Colony Formation
To determine the effects of Inhibitor-4 on the colony formation of breast cancer cells and normal cell lines, the cells were treated with various concentrations of Inhibitor-4 and BCI-121 ( Figure S2c). Surprisingly, a significant decrease in colony formation of MCF10A (normal) cells was also observed at 200 µM concentration of BCI-121 ( Figure S2e) compared to Inhibitor 4, which did not affect MCF10A survival ( Figure S2f). This result again suggests the improved inhibition effect of Inhibitor-4 compared to BCI-121.

Inhibitor-4 Reduces Cell Viability in MCF7 Cells
The effect of Inhibitor-4 on the viability of wild type and cancer cell lines was evaluated using an MTT assay at different time points (24,48,72 Figure S3f). Collectively, these data suggest that Inhibitor-4 is a promising, cancer-specific inhibitor that reduces cancer cell line viability and growth without affecting normal cells.

Inhibitor-4 Induces Cell Cycle Arrest in Breast Cancer Cells
To investigate whether the growth inhibitory effect of Inhibitor-4 on breast cancer cells was due to cell cycle arrest, we conducted cell cycle analysis using Propidium Iodide (PI) staining. Cells were treated with 200 µM of BCI-121 or Inhibitor-4 for 24 h. As shown in Figure 4 and Figure S4   Values are mean ± standard error of the means. Statistically significant differences from control are indicated by * p < 0.05, ** p < 0.01 or ns p > 0.05.

Inhibitor-4 Promotes Apoptosis in Breast Cancer Cells
To reveal whether Inhibitor-4 induce apoptosis on breast cancer cell line or not, we performed apoptosis assay using APC Annexin V/PI followed by flow cytometry analysis. After 48 h of Inhibitor-4 treatment, the percentage of live cells decreased to 71% in both breast cancer cell lines (from 91% in MCF7 and 95% in MDA-MB-231) as demonstrated by flow cytometry (Figures S5d,f,j,l and 5b,d). Also, treatment with Inhibitor-4 showed increase in late apoptosis and necrosis percentages in MCF7 ( Figures S5d,f and 5b), while MDA-MB-231 showed early apoptosis with treatment of Inhibitor-4 ( Figures S5j,l and 5d). BCI-121 caused late apoptosis in MCF7 and both early and late apoptosis in

Inhibitor-4 Promotes Apoptosis in Breast Cancer Cells
To reveal whether Inhibitor-4 induce apoptosis on breast cancer cell line or not, we performed apoptosis assay using APC Annexin V/PI followed by flow cytometry analysis. After 48 h of Inhibitor-4 treatment, the percentage of live cells decreased to 71% in both breast cancer cell lines (from 91% in MCF7 and 95% in MDA-MB-231) as demonstrated by flow cytometry (Figure S55d ,k and 5a,c). Neither treatment caused significant differences in apoptosis nor necrosis in MCF10A cells ( Figures S5q,r and 5e,f). Apoptosis induction through SMYD3 inhibitors was also tested using Caspase-3/7 activity assay. The data have shown increases in Caspase-3/7 activity in MDA-MB-231, however, no significant differences in MCF7, which is Caspase-3/7 independent apoptosis pathway, and MCF10-A.

Discussion
Aberrant expression of SMYD3 has been shown to be oncogenic and is essential for the proliferation of most colorectal, hepatocellular, and breast carcinomas, as well as prostate cancer [3,8]. Over 80 genes (including highly regulated homeobox genes, cell cycle regulators, and oncogenes) display altered expression because of aberrant upregulation of SMYD proteins [1][2][3]. Specifically, SMYD3 over-expression is highly associated with cancer development by regulating tumor proliferation, metastasis, invasion, and apoptosis [22]. Several studies have shown that SMYD3 regulates the oncogenic RAS signaling pathway by integrating a cytoplasmic-kinase signaling cascade, resulting in accelerated cell proliferation and differentiation [9]. Another study demonstrated that SMYD3 is essential for estrogen receptor-mediated transcription in breast cancer cells by down-regulating SMYD3 via RNA interference [14]. SMYD3 mediated-H2A.Z methylation has also been shown to trigger cyclin A1 gene expression, leading to cell cycle activation in breast cancer cells [23]. Knock down of SMYD3 in ovarian cancer tissues leads to upregulation of CDKN2B (p15INK4B), CDKN2A (p16INK4), CDC25A and CDKN3 as members of cyclin-dependent kinase inhibitors (CDK) [24]. Inducing apoptosis via silencing of SMYD3 has also been observed in ovarian cancer in vivo and has been accredited to the upregulation of CD40LG and downregulation of BIRC3 [24]. BIRC3 is a member of the inhibitors of apoptosis proteins (IAP) family and relates to many cancers in cases of aberrant overexpression because it can prevent apoptotic signals [25,26]. Therefore, it is likely that SMYD3 inhibitors can trigger apoptosis by down-regulating BIRC3. In addition, another study demonstrated that the MCF7 cell line lacks Caspase 3, which is essential for apoptosis, however in the absence of Caspase 3, Caspase 6 can be activated as an alternative mechanism to trigger apoptosis. As a result, under cellular stress, MCF7 cells undergo apoptosis in response to Caspase 6 and necrosis in response to TNF-α stimulation [27][28][29]. Despite the connection between SMYD3-overexprssion and several types of carcinogenesis, few studies have targeted SMYD3 inhibition in the context of breast cancer through the design of the inhibitors.
In this study, we sought to design small molecule inhibitors for the inhibition of SMYD3-mediated methylation (Figure 1), proliferation (Figure 3), colony formation ( Figure S2), and viability ( Figure S3) in breast cancer cells. Specifically, we demonstrated that in silico enzyme models can predict effective competitive enzyme inhibitors by screening vast molecular libraries and predicting binding energies. This approach to small molecule design significantly reduces the time, expense, and equipment that have been required for traditional benchtop small molecule screening until now.
Using Schrodinger ® software and several in vitro assays, we demonstrated that one of the hit compounds identified in silico (Inhibitor-4) was able to reduce breast cancer cellular growth and viability without affecting normal breast epithelial cells. In vitro, Inhibitor-4 was shown to inhibit SMYD3-mediated histone methylation. In breast cancer cells, Inhibitor-4 extended cell doubling time (Figure 3). We also demonstrated that Inhibitor-4 arrests the cell cycle in breast cancer cells without affecting normal cells (Figure 4 and Figure S4), which demonstrates an improvement over BCl-121, a previously-developed SMYD3 inhibitor. Finally, the novel SMYD3 inhibitor presented here caused apoptosis in breast cancer cell lines without affecting the normal breast cell line ( Figure 5 and Figure S5). However, testing the in vivo SMYD3 specificity of Inhibitor-4 needs to investigate the impacts of Inhibitor-4 on SMYD3-knockdown cells. This could be performed for future characterization and validation.

Conclusions
In conclusion, we have successfully used in silico compound screening to identify a small molecule that inhibit SMYD3 activity in vitro and reduced cancer cell growth and proliferation. Future work will involve the application of this approach to other therapeutic targets and the continued development and optimization of therapeutics for SMYD3-related cancers.

In Silico Screening
We implemented the Small Molecule Drug Discovery Suite (Schrodinger, Inc., New York, NY, USA) to predict the binding affinity of a library of 137,990 molecules [15][16][17]. This library of molecules was downloaded from the free ZINC15 database, and included all "purchasable" molecules with reported or predicted activity in vitro [30]. The 3D structure of SMYD3 used for in silico docking was uploaded from the Protein Data Bank (PDB) under PDB identification code 5EX3 [31]. After an initial simulation which docked each molecule into SMYD3 s protein-target binding pocket (not its s-adenosylmethionine binding pocket), the top ten hits (most-negative binding energy) were entered into the ZINC15 molecular similarity search engine, and the 50 most-similar compounds to each of the ten leading candidates were again scored using the Schrodinger software (500 total compounds). From this iteration, the top five compounds were purchased and assessed in vitro using SMYD3 methyltransferase assays. After initial experiments, Inhibitor-4 was found to be the most promising and, consequently, it advanced to the cell line experiments described below.

Chemical Compounds
All screened compounds were dissolved in dimethyl sulfoxide (DMSO) as 5, 10 or 100 mM stock solutions. The positive control, BCI-121, is a previously-reported SMYD3 inhibitor shown to reduce the cellular proliferation of colorectal and ovarian cancer [21,24]. It was purchased from Millipore Sigma (1817, Burlington, MA, USA) and dissolved in DMSO at 10 and 100 mM. BCI-121 was used in all experiments to investigate its impacts against breast cancer cell lines and as a positive control inhibitor. All compounds were stored at −20 • C until used for the experiments. Some d 6 -dimethyl sulfoxide (D, 99.9%) containing 0.05% v/v TMS were used for the 1 H NMR stability study and the d 6 -dimethyl sulfoxide was purchased from Cambridge Isotope Laboratories, Inc. and used as is. The stock solutions of 10.0 mM of BCI-121 and Inhibitor-4 were prepared immediately before use in d 6 -DMSO.

In Vitro Methylation Assay
In vitro methylation was investigated using a colorimetric assay (BioVision, K986-100, Milpitas, CA, USA). SMYD3 inhibitors (160 nM) were incubated with H3 recombinant protein (1.6 µM; Sigma-Aldrich, SRP0177, St. Louis, MO, USA) for 10 min at room temperature. Next, SMYD3 recombinant protein (100 nM; Sigma-Aldrich, SRP0153, St. Louis, MO, USA) and s-adenosylmethionine (SAM) cofactor (500 µM, methyl donor ligand) were added to the SMYD3 inhibitor and H3 solution in the methylation buffer that was provided with the kit. The absorbance was read using a microplate reader (BioTek, Cytation 5, Winooski, VT, USA) at 570 nm in kinetic mode every 30 s at 37 • C for 45 min. The optical density (OD) of the inhibitors was normalized to the optical density (OD) of the control [3].

NMR Spectroscopy Analysis
The 10.0 mM stock solutions of Inhibitor-4 and BCI-121 were prepared freshly in d 6 -DMSO containing 0.05% v/v tetramethylsilane (TMS) and diluted to the final concentration of 5.0 mM in d 6 -DMSO. The stabilities of Inhibitor-4 and BCI-121 were determined by 1D 1 H Nuclear Magnetic Resonance (NMR) spectroscopy on a Bruker 400 MHz NMR spectrometer at 25 • C using routine parameters [32]. 2D-NMR experiment will run to confirm the assignments (data not included) [32]. Chemical shifts were measured against TMS (0 ppm) as an internal reference. The spectra of Inhibitor-4 and BCI-121 were recorded at 0 and 24 h. The spectra were worked up and integrated using Mnova V.14 (MestreLab Research SL, Escondido, CA, USA). The signals in the aromatic region were used to measure the ratio of starting material and hydrolysis product.

Protein Extraction and Immunoblotting
Total protein was extracted from frozen cells using RIPA buffer (150 mM NaCl, 5 mM EDTA, 50 Tris-HCl pH 8.0, 1% NP-40, 0.5% Na-catecholate and 0.1% SDS) supplemented with protease inhibitor (PI 87785, Life tech, Carlsbad, CA, USA). Protein concentration was determined according to Bradford (1976) using bovine serum albumin as a standard [35]. Fifty micrograms of total protein were separated by SDS-PAGE, transferred to nitrocellulose membrane by electroblotting as described by [36] and probed with the antibodies specific for the indicated proteins [36]. Actin was used as an internal control for normalization. Antibodies for immunoblot detection of SMYD3 (Rabbit monoclonal antibody to SMYD3, ab183498, Abcam, Cambridge, MA, USA) and β-Actin (A5316-100 UL, Sigma-Aldrich, St. Louis, MO, USA) have been used as the primary antibodies. Bound antibodies on blots were detected by HRP-conjugated secondary antibodies (ab205718, Abcam, Cambridge, MA, USA). Detection was done using Clarity Western ECL Substrate (Bio-Rad, Hercules, MA, USA) and visualized using Image Lab Software (Bio-Rad). Densitometric evaluation was performed by ImageJ software (Version 2.0.0).

Immunocytochemistry
Exponentially growing cells were fixed with 4% paraformaldehyde for 15 min at room temperature. After PBS wash, cells were permeabilized with 0.1% SDS, 0.5% Triton X-100 in PBS for 10 min at room temperature. 10% goat serum in PBS was used for blocking for 30 min at room temperature. Primary antibody was diluted in 1:300 and treated for 1 h at 37 • C. Secondary antibody (Alexa488 conjugated anti-rabbit IgG) was diluted for 1:500 and treated for 30 min at 37 • C. DAPI in Vectashield antifade solution was used for mounting. The images were capture by Zeiss Axiophot microscope with Qimaging Exi Aqua camera with Qcapture pro software. Blue signal was obtained with 50 millisecond exposure. Green signal was obtained with 200 millisecond exposure. Signals were quantified by densitometry using ImageJ2 software (Version 2.0.0).

In Vitro Cell Growth Inhibition Assay
Normal and breast cell lines were plated at a density of 20,000 cells/well onto a 6-well plate with different concentrations of Inhibitor-4 and BCI-121. After trypsinization, cell numbers were counted and scored as the number of proliferating cells after treatments at different time points (24,48, 72 and 96 h) using a Coulter Counter Z2 (Beckman-Coulter Z2 Coulter Particle Count Counter and Size Analyzer, Brea, CA, USA). Data were analyzed and cell doubling time was calculated using GraphPad Prism 6 software (Graph Pad Software, La Jolla, CA, USA) through the exponential growing equation using the exponential growing stage [37].

Clonogenic Cell Survival
A colony formation assay was used to determine cell sensitivity to SMYD3 inhibitors. The self-renewal and proliferative capacities of cells were measured. To form colonies, cells were seeded onto 6 well plates and were treated with varying concentrations of BCI-121 and Inhibitor-4. The plated cells were incubated in a humidified incubator at 5% CO 2 and 37 • C for two weeks. Then, colonies were fixed with 100% ethanol and allowed to dry for 20 min at room temperature before staining. Colonies were stained using 0.1% crystal violet and allowed to dry before counting. Reproductively viable surviving cells were counted based on the microscopic colonies containing more than 50 cells. From the cell survival fraction, survival curves were drawn using Graph Pad Prism 6 software (Graph Pad Software, La Jolla, CA, USA). At least three independent experiments for each cell line were conducted [38].

MTT Assay
Cells were plated at a density of 5000 cells/well in 96 well plates. After seeding, cells were treated with the vehicle (DMSO 0.1%, 0.15% and 0.2% v/v) or various concentrations of the screening inhibitors. The plated cells were incubated in CO 2 and treated for 24, 48, 72 and 96 h. Then, 10 µL MTT solutions (5 mg/mL) were added to each well followed by a 4 h incubation in CO 2 in the dark. Formazan crystals formed were dissolved in 100 µL of SDS followed by a second 4 h incubation in CO 2 . The absorbance was read using a microplate reader (BioTek Instrument, Cytation 5, Winooski, VT, USA). The optical density (OD) of each sample was subtracted from the optical density (OD) of the background and the Formazan standard curve was determined. Cellular viability of all samples was calculated using the ratio of the inhibitor treated-groups versus vehicle-treated group. Graph bars were obtained using GraphPad Prism 6 software (Graph Pad Software, La Jolla, CA, USA) [39].

Cell Cycle Assay
Cell cycle distributions were analyzed using PI flow cytometry. Cells were plated at density of 5 × 10 5 cells per well onto 6-well plate. Cells were treated with 200 µM of Inhibitor-4 or BCI-121 and incubated in a humidified incubator at 5% CO 2 and 37 • C for 24 h. Following incubation, detached cells were collected, washed two times with phosphate buffered saline (PBS). Then, cells were fixed with 70% ethanol in PBS overnight at 4 • C. The fixed cells were washed with PBS twice to remove ethanol thoroughly. The cells were resuspended in propidium iodide staining solution consisting of 20 µg/mL propidium iodide and 200 µg/mL RNase in 0.1% Triton X-100. The stained cells were incubated for 15 min in an incubator at 37 • C. DNA contents were measured subsequently using CyAn ADP analyzer flow cytometry (Beckman Coulter, Fort Collins, CO, USA). Each cell line was gated at 10,000 events and the cell cycle distributions were determined using FLOWJO 10.6 software (FlowJo LLC, Ashland, OR, USA) [40].

Apoptosis Assays
Cell apoptosis was detected using Annexin V, which binds to translocated phosphatidylserine (PS) in the plasma membrane as previously described [41]. Necrosis and late apoptosis were detected using PI to test loss of cell membrane integrity. Briefly, cells were plated and treated with 200 µM of either BCI-121 or Inhibitor-4 for 48 h. Then, the cells were washed with PBS, trypsinized, pelleted, and resuspended in Annexin binding buffer. The cells were stained first with APC Annexin V for 15 min and then with 2.5 µL of PI. The cell mixture was analyzed using a Cytek 4-laser Aurora instrument (Cytek, Fremont, CA, USA). From each sample, a minimum of 3 × 10 4 events was collected. SpectroFlo software (Cytek, Fremont, CA, USA) was used to analyze the multivariate data. APC Annexin V+/PI+, APC Annexin V-/PI-, APC Annexin V+/PI-or APC Annexin V-/PI+ represented late apoptotic cells, viable cells (intact), early apoptotic cells or necrosis, respectively [42].
In addition to, apoptosis induction by SMYD3 treatments was also assessed using Caspase 3/7 activation. Exponentially growing cells were treated with 200 µL of BC1-121 and Inhibitor-4. After 48 h of incubation, the early apoptosis was measured with the activation of Caspase 3/7 by Caspase-Glo 3/7 kit (Promega, Madison, WI, USA). Glow luminesce of 15,000 cells was measured by Lumat LB9507 (Berthold technologies, Oak Ridge, TN, USA).

Statistical Analysis
The statistical significance of the results in this study was analyzed using GraphPad Prism 6 software (Graph Pad Software, La Jolla, CA, USA) for two-way ANOVA analysis. p value of less than 0.05 were considered statistically significant for all analyses.