Cytotoxic activity of bromodomain inhibitor NVS-CECR2-1 on human cancer cells

Bromodomain (BRD), a protein module that recognizes acetylated lysine residues on histones and other proteins, has recently emerged as a promising therapeutic target for human diseases such as cancer. While most of the studies have been focused on inhibitors against BRDs of the bromo- and extra-terminal domain (BET) family proteins, non-BET family BRD inhibitors remain largely unexplored. Here, we investigated a potential anticancer activity of the recently developed non-BET family BRD inhibitor NVS-CECR2-1 that targets the cat eye syndrome chromosome region, candidate 2 (CECR2). We show that NVS-CECR2-1 inhibits chromatin binding of CECR2 BRD and displaces CECR2 from chromatin within cells. NVS-CECR2-1 exhibits cytotoxic activity against various human cancer cells, killing SW48 colon cancer cells in particular with a submicromolar half maximum inhibition value mainly by inducing apoptosis. The sensitivity of the cancer cells to NVS-CECR2-1 is reduced by CECR2 depletion, suggesting that NVS-CECR2-1 exerts its activity by targeting CECR2. Interestingly, our data show that NVS-CECR2-1 also kills cancer cells by CECR2-independent mechanism. This study reports for the first time the cancer cell cytotoxic activity for NVS-CECR2-1 and provides a possibility of this BRD inhibitor to be developed as an anticancer therapeutic agent.


Scientific RepoRtS
| (2020) 10:16330 | https://doi.org/10.1038/s41598-020-73500-7 www.nature.com/scientificreports/ over other 48 BRD targets (https ://www.thesg c.org/chemi cal-probe s/NVS-1). Neither the activity of NVS-CECR2-1 against its target within cells nor its therapeutic potential in treatment of human diseases, such as cancer, was reported. Here, we demonstrated that NVS-CECR2-1 inhibits chromatin binding of CECR2 BRD and displaces CECR2 from chromatin within cells. We found that NVS-CECR2-1 has a strong cytotoxic activity on various human cancer cells, killing SW48 colon cancer cells in particular with a submicromolar value of half maximum inhibitory concentration (IC 50 ) by increasing apoptosis. Our data showed that the cytotoxic activity of NVS-CECR2-1 on cancer cells is exerted by targeting CECR2 as well as via CECR2-independent mechanism. We also performed analysis of CECR2 gene expression in colorectal cancer tissues using several online databases and discussed the significance of the results.

NVS-CECR2-1 inhibits chromatin binding of CECR2 BRD within cells. First, we determined
whether NVS-CECR2-1 (Fig. 1a) inhibits chromatin binding of CECR2 BRD. After transfection with the plasmid vector expressing a dimeric form of CECR2 BRD, which binds chromatin with higher affinity than a monomeric form 18 , cells were treated with 10 μM of NVS-CECR2-1 and cell lysates were fractionated into insoluble chromatin-bound and soluble chromatin-unbound fractions. Immunoblot analysis showed that CECR2 BRD roughly equally distributed into the both fractions and this distribution largely shifted to the chromatinunbound fractions with the CECR2-BRD band in the chromatin-bound fraction barely detected (Fig. 1b, lanes 1, 2, 5 and 6), indicating that NVS-CECR2-1 inhibited the chromatin binding of CECR2 BRD. In parallel experiments, pre-treatment of tricostatin A (TSA) histone deacetylase inhibitor rendered most of CECR2 BRD bound to chromatin, which was nearly completely abolished by NVS-CECR2-1 treatment (Fig. 1b, lanes 3, 4, 7 and 8), further confirming the inhibitory activity of NVS-CECR2-1 against CECR2 BRD. As control, we performed the similar chromatin fractionation experiments for PFI-3, the BRD inhibitor specific for the BRG1 ATPase of SWI/ SNF chromatin remodeling complex 19 . Whereas NVS-CECR2-1 effectively inhibited the chromatin binding of CECR2 BRD at the concentrations ranging from 5 to 15 μM, PFI-3 had no activity against CECR2 BRD even at 100 μM or a higher concentration (Fig. 1c). Conversely, PFI-3 inhibited the chromatin binding of a dimeric form of BRG1 BRD, which also has higher chromatin affinity than a monomeric form 20 , whereas NVS-CECR2-1 had little effect (Fig. 1d). These results showed that NVS-CECR2-1 has a potent activity against CECR2 BRD within cells.
NVS-CECR2-1 dissociates endogenous CECR2 from chromatin. Next, we determined whether NVS-CECR2-1 can dissociate endogenous CECR2 proteins from chromatin by chromatin fractionation as described before. The cells were treated with NVS-CECR2-1 at 5, 10 or 15 μM, or with PFI-3 as control at 100 or 200 μM. The results showed that NVS-CECR2-1 dissociated CECR2 from chromatin in a dose-dependent manner without affecting BRG1, and conversely, PFI-3 dissociated BRG1 from chromatin with no effect on CECR2 (Fig. 2a,b). To confirm these results, we conducted salt gradient chromatin fractionation. Most of the cellular CECR2 proteins were extracted by 0.15-and 0.3-M NaCl, and NVS-CECR2-1 treatment increased the CECR2 extraction at 0.15-M NaCl with its concomitant decrease at 0.3-M NaCl (Fig. 2c,d and Fig. S1a-c), indicating that NVS-CECR2-1 dissociated CECR2 from chromatin. As expected from the fact that CECR2 is a chromatin remodeler and thus a nuclear protein, CECR2 was barely detected in the cytoplasmic fractions in salt gradient chromatin fractionation (Fig. 2c,d and Fig. S1a-c) and stained exclusively within the nucleus in immunofluorescence microscopy ( Fig. S2a,b). Essentially no CECR2 was found in the soluble nuclear fractions (Fig. S1a,c), indicating that CECR2 exists bound to chromatin in the nucleus. The CECR2 staining pattern remained unchanged after NVS-CECR2-1 treatment in immunofluorescence microscopy (Fig. S2b). These results therefore demonstrated the activity of NVS-CECR2-1 to prevent its target protein CECR2 from binding chromatin within cells.

NVS-CECR2-1 exhibits cytotoxic activity on various human cancer cells.
Having verified the in vivo activity of NVS-CECR2-1, we investigated the possibility that it kills cancer cells. For this, we selected a set of human cancer cell lines that represent various tissue origins, such as colon (SW48, HT29 and HCT116), lung (H460), uroepithelium (SV-HUC-1), cervix (HeLa) and bone (U2OS). The cells were treated with increasing concentrations of NVS-CECR2-1 and the cell viability was determined after three days by MTS assay. Interestingly, the treatment of NVS-CECR2-1 decreased the viability of all cancer cells analyzed in a dose dependent manner (Fig. 3). NVS-CECR2-1 treatment also showed a dose-dependent cytotoxicity on human embryonic kidney (HEK) 293 T cells, an immortalized cell line originated from non-cancer tissue (Fig. 3). Notably, NVS-CECR2-1 exhibited cytotoxic activity on all the cell types tested at the pharmacologically significantly concentrations, seemingly with a few micromolar concentrations killing more than half of the cell population (Fig. 3). Immunoblot analysis showed that the levels of CECR2 expression varied among the tested cell lines, with some displaying very little (HT29, H460, HeLa and SV-HUC-1) and the others exhibiting different degree of CECR2 expression (SW48, HCT116, U2OS and HEK 293 T) (Fig. 3, last panel). Therefore, the cytotoxic activity of NVS-CECR2-1 on cancer cells is not necessarily dependent on CECR2 expression. At this moment, owing to the limited accuracy of the short-term viability assay, it was difficult to predict any correlation between the cytotoxic activity of NVS-CECR2-1 and CECR2 expression levels of cancer cells. www.nature.com/scientificreports/ www.nature.com/scientificreports/ up to 4 μM and their clonogenic ability was evaluated after ten days by colony formation assay. NVS-CECR2-1 inhibited the clonogenic ability of SW48 cells in a dose dependent manner and its IC 50 value was estimated to be 0.64 μM (Fig. 4a,b). Next, we determined the mechanisms by which NVS-CECR2-1 kills SW48 cells. The cells were treated with increasing concentrations of NVS-CECR2-1 up to 6 μM and analyzed for apoptosis and necrosis after 72 h by Annexin-V assay. NVS-CECR2-1 treatment increased apoptosis in a dose-dependent manner, with more than 80% cells undergoing apoptosis at 6 μM, and had virtually no effect on necrosis (Fig. 4c,d).

NVS-CECR2-1 kills SW48 cells with a submicromolar IC
Increase of apoptosis by NVS-CECR2-1 was confirmed by PARP1 cleavage (Fig. 4e). Therefore, NVS-CECR2-1 kills SW48 cells mostly by inducing apoptosis. For the purpose of comparison, we analyzed the toxicity of doxorubicin on SW48 cells and observed that this well-known potent anticancer drug exhibited the IC 50 value of 0.54 μM (Fig. 4f,g). These results suggest that, given its IC 50 value compatible to doxorubicin, NVS-CECR2-1 can be a potentially good candidate of anticancer drug for colon cancer treatment.
NVS-CECR2-1 kills SW48 cells by targeting CECR2. Next, we asked whether NVS-CECR2-1 kills SW48 cells by targeting CECR2. If NVS-CECR2-1 exerts its cytotoxic activity by inhibiting CECR2, depletion of CECR2 should affect the cells' sensitivity to NVS-CECR2-1. SW48 cells were depleted for CECR2 by specific siRNA (Fig. 5a) and treated with increasing NVS-CECR2-1 from 0.5 to 2 µM. The clonogenic ability of the cells www.nature.com/scientificreports/ were then determined by colony formation assay. NVS-CECR2-1 showed a dose-dependent cytotoxic activity on the cells transfected with nonspecific control siRNA with the IC 50 value of 0.62 μM. Depletion of CECR2 by specific siRNA largely decreased the cell viability, indicating that CECR2 was essential for the survival of SW48 cells. Notably, NVS-CECR2-1 exhibited very little effect on the CECR2-depleted cells compared to the control cells, likely due to the large decrease of their viability (Fig. 5b,c), indicating that the sensitivity of the cells to NVS-CECR2-1 was reduced by CECR2 depletion. Nonetheless, NVS-CECR2-1 treatment still decreased the viability of the CECR2-depleted cells in a dose-dependent manner and this decreasing effect was slightly but www.nature.com/scientificreports/ significantly less than that on the control cells, accompanied with a shift of the IC 50 value from 0.62 to 0.81 µM (Fig. 5b,c). We confirmed these findings by analyzing apoptosis by Annexin-V assay. NVS-CECR2-1 increased apoptosis in a dose-dependent manner as observed before. CECR2 depletion largely increased apoptosis and the rate NVS-CECR2-1 increased apoptosis of the CECR2-depleted cells was lower than that of the control cells ( Fig. 5d-f). Consistently, whereas NVS-CECR2-1 increased PARP1 cleavage in a dose-dependent manner, CECR2 depletion resulted in a large increase of PARP1 cleavage, which was not further enhanced by NVS-CECR2-1 (Fig. 5g). These results showed that NVS-CECR2-1 exerts the cytotoxic activity on SW48 cells by targeting CECR2 although it still has this activity to some extent even after CECR2 depletion.
CECR2-independent cytotoxic activity of NVS-CECR2-1. The reason that CECR2-depleted cells were still sensitive to NVS-CECR2-1 could be due to incomplete CECR2 knockdown or an off-target activity of NVS-CECR2-1. To address this issue, we determine the effect of NVS-CECR2-1 depletion on HEK 293 T cells, which were shown to be independent of CECR2 for their viability 18 . As expected, CECR2 depletion had no effect on the viability of HEK 293 T cells (Fig. 6a, b-d). Notably, NVS-CECR2-1 exhibited a dose-dependent cytotoxic activity on the control and CECR2-depleted cells to a similar degree in both short-term (Fig. 6b) and longterm viability assays (Fig. 6c,d), indicating that NVS-CECR2-1 exerted the cytotoxic activity independently of CECR2. To further substantiate these results, we analyzed the HAP1 cells, a near-haploid human cell line derived from the chronic myelogenous leukemia cell line KBM-7, in which the gene coding for CECR2 was knocked out by the CRISP/Cas9 system. Absence of CECR2 was confirmed for this cell (Fig. 6E). The establishment of this cell line itself indicates that CECR2 is dispensable for cell survival and proliferation. Indeed, we observed no difference in cell growth between the CECR2-knockout HAP1 and the parental CECR2-wild type HAP1 cells (data not shown). NVS-CECR2-1 showed a similar degree of cytotoxic activity on the both cell lines in shortterm (Fig. 6f) and long-term viability assays (Fig. 6g,h). The IC 50 values of NVS-CECR2-1 for 293 T and HAP1 cells were 2.45 and 2.73 µM, respectively, which were approximately four times higher than that for SW48 cells.
Taking thus far results together, we concluded that NVS-CECR2-1 exerts cytotoxic activity not only by targeting CECR2 but also via CECR2-independent mechanism, possibly the off-target effects.

Cytotoxic activity of NVS-CECR2-1 on various human colon cancer cells. The results thus far
suggested that cells dependent on CECR2 for their viability tend to be more sensitive to NVS-CECR2-1 than CECR2-independent cells. To determine whether this tendency can be generalized, we analyzed a set of human colon cancer cell lines, including DLD-1, LoVo, HCT116, HT29, SW480 and HCT15 as well as SW48 cells. Immunoblot analysis revealed that SW48 and HCT116 cells expressed CECR2 at the similar levels as HEK 293 T cells whereas the remaining cell lines exhibited very low levels of CECR2 expression (Fig. 7a). CECR2 depletion reduced the viability of SW48 and HCT116 cells by approximately 72% and 42%, respectively, and had no effect on the viability of the other five cell lines (Fig. 7b,c). Then, we determined the IC 50 values of NVS-CECR2-1 against these cancer cells as described before. Notably, HCT116 cells exhibited 1.30 µM of IC 50 for NVS-CECR2-1, which was higher than SW48 cells (0.64 µM), whereas the IC 50 values against the other five CECR2-independent cell lines ranged from 2.1 to 3.3 µM (Fig. 7d). These results showed a good correlation between the cytotoxic activity of NVS-CECR2-1 on the cancer cells and the extent to which their survival relies on CECR2.

Analysis of CECR2 gene expression in colorectal cancer tissues.
To evaluate the clinical significance of our findings, we analyzed the CECR2 gene expression in human colorectal cancer using several online databases. First, the data from TCGA Pan-Cancer Atlas showed that CECR2 mRNA expression in colorectal cancer ranked at a moderate to medium level among the 32 different cancer types, comprising 10,967 samples from 10,953 patients (Fig. S3). A differential analysis of normal and cancer tissues in the 11 colorectal datasets comprising 2033 samples showed that CECR2 mRNA expression was upregulated in colorectal cancer tissues compared to colorectal normal tissues (Fig. 8a). Immunohistochemical staining data from the Human Protein Atlas revealed that colorectal cancer belonged to a group in which 100% of examined patients displayed medium or high levels of CECR2 protein expression (Fig. S4). However, both colorectal normal and colorectal adenocarcinoma tissues showed medium or high levels of CECR2 staining (Fig. 8b). A Kaplan-Meier survival curve indicated that the group of colon adenocarcinoma patients with high CECR2 expression showed significantly worse survival than that with low CECR2 expression (P = 0.041), but CECR2 was not categorized as a prognostic gene in colon cancer (gene with P < 0.001 is defined as prognostic) (Fig. 8c). There was no significant correlation between patient survival probability and CECR2 expression in rectal adenocarcinoma (Fig. 8c). Thus, it is not conclusive at the moment whether CECR2 is a therapeutic target for colorectal cancer and NVS-CECR2-1 is clinically applicable in treatment of this cancer type.

Discussion
In the present study, we investigated the recently developed BRD inhibitor NVS-CECR2-1 for its potential cytotoxic activity on human cancer cells. First, we demonstrated that NVS-CECR2-1 inhibits chromatin binding of CECR2 BRD and dissociates CECR2 from chromatin within cells. Then, we found that NVS-CECR2-1 has cytotoxic activity against various human cancer cells with different tissue origins at pharmacologically significant concentrations. In particular, NVS-CECR2-1 kills SW48 colon cancer cells with a submicromolar IC 50 value and this activity is exerted mainly by increasing apoptosis. The cytotoxic activity of NVS-CECR2-1 against the cancer cells is reduced by CECR2 depletion and is positively correlated with the extent to which the cells rely on CECR2 for their survival, suggesting that NVS-CECR2-1 exerts its activity by targeting CECR2. This is the first study that reports an anticancer cytotoxic activity for NVS-CECR2-1. www.nature.com/scientificreports/ NVS-CECR2-1 was originally developed as an inhibitor specific for CECR2 BRD. It has no cross reactivity in a BRD panel of 48 targets, with only weak interactions with BRD4 and BRD7, as demonstrated by a thermal shift assay, (https ://www.thesg c.org/chemi cal-probe s/NVS-1). However, our results showed that siRNA knockdown of CECR2 in 293 T cells as previously reported 18 or CRISPR/cas9-derived inactivation of the CECR2 gene in haploid leukemia cells did not affect the cytotoxic activity of NVS-CECR2-1 on these cells. Given the fact that CECR2 is dispensable for survival and proliferation of these two cell types, these results suggest that this activity of NVS-CECR2-1 could be exerted by off-targeting some other proteins, which are essential for cell survival and proliferation. Thus, our results have left the question open whether NVS-CECR2-1 is specific for CECR2 within cells.
CECR2 is predominantly expressed in the neural system and involved in neurulation and inner ear development during mouse embryogenesis 15,21 . CECR2 forms a complex with the ISWI family ATP-dependent chromatin remodeler SNF2L and functions in transcriptional regulation of genes involved in embryonic and post-natal brain development 15,22 . CECR2 also forms a complex with SNF2H, another ISWI family chromatin remodeler, in the testis and contributes to spermatogenesis in adult tissue 23 . The CECR2-SNF2L complex, named CERF, has been shown to remodel chromatin in vitro and displace the ATP hydrolysis activity that is stimulated by nucleosomes 15 . A genome-wide screen for human BRD-containing proteins identified CECR2 to be involved in DNA damage response 18 . However, the exact cellular functions of CECR2 remain largely unknown. Our results show that CECR2 is important for cell viability. Interestingly, the role of CECR2 in cell viability depends on cell types such that it is important for SW48 and HCT116 cells but dispensable for 293 T and KBM-7 cells. In addition, CECR2 expression levels vary even among different cancer cells with same tissue origin-a panel of colon cancer cell lines, and the requirement of CECR2 for cell viability is correlated with its expression levels. Therefore, the role of CECR2 in cell viability appears to depend on both cell type and CECR2 level. The exact role of CECR2 in the cancer cell viability and its effects on the cell sensitivity to NVS-CECR2-1 remain to be clearly defined in future study.
Motivated by the success in developing potent and highly selective inhibitors for the BET family BRDs and their entering into clinical trials in cancer treatment, research activities have been extended to developing non-BET family BRD inhibitors, including NVS-CECR2-1 13,14 . CECR2 has been ranked top among the 24 human BRD-containing proteins for the score of druggability 24 and our work showed that NVS-CECR2-1 has < 10 μM of the IC 50 values in cytotoxicity against various human cancer cells. Therefore, it is possible that NVS-CECR2-1 can be developed as a candidate of potential cancer therapeutic agent. In particular, our studies using a panel of human colon cancer cell lines raised the possibility that NVS-CECR2-1 has a potential to be used as a therapeutic agent in colon cancer treatment. In support of this, a differential analysis showed that CECR2 mRNA expression is upregulated in colorectal cancer tissues compared to normal counterparts. However, immunohistochemical staining of CECR2 scored high or medium levels in both colorectal normal and adenocarcinoma tissues from several patients, and CECR2 was not defined as a prognostic gene in colorectal cancer based on the Kaplan-Meier survival curves. Thus, the answer to the question whether NVS-CECR2-1 has a potential of clinical application for treatment of colon cancer is indecisive for now and reserved for future study. Given our data demonstrating the potent activity of NVS-CECR2-1 to displace CECR2 from chromatin within cells, this BRD inhibitor can also be used as a useful tool for studying the cellular functions of CECR2.  Seoul). HCT15, SW48, SW480, DLD-1, LoVo, H460, and SV-HUC-1 cells were maintained in RPMI1640 medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 µg/ml streptomycin. HCT116, HT29, and U2OS cells were cultured in McCOY's 5A modified medium supplemented with 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin. HEK 293 T and HeLa cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin. HAP1 parental (C631) and CECR2-KO (HZGHC000946c001) cell lines were purchased from Horizon Discovery (Cambridge, UK) and cultured in Iscove's modified Dulbecco's medium (IMDM) supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin. All cells were maintained at 37 °C in a 5% CO 2 humidified incubator. The sources of antibody used in this work are as follows; anti-CECR2 (LifeSpan Bio- Figure 5. CECR2 depletion reduces the sensitivity of SW48 cells to NVS-CECR2-1. SW48 cells were transfected with control or CECR2-specific siRNAs for 48 h, and treated with increasing concentrations of NVS-CECR2-1 before being subjected to colony formation, annexin V/PI staining and PARP1 cleavage assays. (a) Immunoblot analysis of CECR2 knockdown of three independent experiments for colony formation assay. Arrow, CECR2 band. (b) Representative results of colony formation assay. (c) Relative cell survival in the colony formation assay was depicted as a graph by setting the value of DMSO-treated (0 μM) si-control cells as 100 (left) or by setting each of the DMSO-treated si-control and si-CECR2 cells as 100 (right). n = 3 (each performed in triplicate); error bars, mean ± s.d. *, p < 0.05; **, p < 0.01; ***, p < 0.001. (d) Immunoblot analysis of CECR2 knockdown of three independent experiments for apoptosis assay. Arrow, CECR2 band. BRD inhibitors and doxorubicin. NVS-CECR2-1 was obtained from the SGC and PFI-3 was purchased from Tocris (5072, UK). Doxorubicin was purchased from Sigma. The inhibitors and drugs were dissolved in DMSO, kept in aliquots at -20 °C, and thawed immediately before use for experiments to minimize unexpected chemical reactions.
MtS assay. 2-4 × 10 3 cells were seeded in 96-well plate in 100 µl of medium per well. After 72 h, MTS assays were performed according to the protocols of Cell Tilter 96 aqueous One Solution Cell Proliferation assay (Promega). The absorbance at 490 nm was measured using a spectraMax i3X plate reader (Molecular Devices, I3X-SC-ACAD). colony formation assay. Colony formation assay was performed as previously described 20 . After staining with 0.5% crystal violet, colonies were dissolved in 10% acetic acid before measuring absorbance at 590 nm with a spectraMax i3X plate reader (I3X-SC-ACAD, Molecular Devices).

Scientific RepoRtS
| (2020) 10:16330 | https://doi.org/10.1038/s41598-020-73500-7 www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.