The chromodomain protein CBX2 binds directly to histone H3 trimethylation at lysine 27 (H3K27me3) and is a component of polycomb repressive complex 1 (PRC1). CBX2 plays a pivotal role in transcriptional repression by acting as a reader protein that recognizes H3K27me3. In this study, we performed in silico screening based on the crystal structure of CBX2 to identify small molecule compounds that target the chromodomain of CBX2. The ability of the selected compounds to inhibit the interaction between CBX2 and histone H3 in cells was validated. After three rounds of in silico screening, CG3-46 was ultimately identified as the most potent CBX2 inhibitor in this study. CG3-46 inhibited the growth of the triple-negative breast cancer cell line MDA-MB-231 in a concentration range comparable to that used to inhibit the interaction between CBX2 and H3K27me3. Our results indicate that CG3-46 represents the first nonpeptide small molecule inhibitor of CBX2, which not only serves as a valuable chemical tool for elucidating the role of CBX2 in cellular epigenetic regulation but also as a starting compound for the development of CBX2-targeted therapeutics for triple-negative breast cancer.
Article
Identification of substituted acetanilide compounds as small molecule CBX2 inhibitors via in silico screening
https://doi.org/10.21203/rs.3.rs-4904827/v1
This work is licensed under a CC BY 4.0 License
You are reading this latest preprint version
Breast cancer
CBX2
Chromodomain
Epigenetics
Histone methylation
Small molecule
Histone modifications, including methylation, are fundamental mechanisms involved in the epigenetic regulation of gene transcription within chromatin. The repression of gene transcription by histone H3 lysine 27 trimethylation (H3K27me3) is facilitated by two conserved polycomb repressor protein complexes (PRCs). First, the lysine methyltransferase EZH2, a component of the PRC2 complex, catalyzes H3K27me3 at the target gene locus1,2. Subsequently, the chromobox protein (CBX) in the PRC1 complex recognizes H3K27me3 via its chromodomain, inducing monoubiquitination of histone H2A at lysine 119 (H2AK119ub1) by BRG1 ubiquitin E3 ligase, a component of the PRC2 complex. Additionally, Jarid2, a PRC2 cofactor, facilitates the localization of the PRC2 complex to the H2AK119ub1 site3. These interactions with two histone modifications mediated by the PRC1 and PRC2 complexes lead to compaction of the chromatin conformation and consequently cause transcriptional repression at the target locus.
CBX proteins, consisting of five paralogs (CBX2, CBX4, CBX6, CBX7, and CBX8), form a part of the PRC1 complex and contain an N-terminal chromodomain responsible for binding to H3K27me3, as well as a C-terminal domain Pc-box facilitating integration into the PRC1 complex2,4. Recent studies suggest that the dysregulation of PRC complexes is associated with cancer5,6. In particular, the aberrant expression of CBX family members plays an important role in a variety of cancers. Among them, CBX2 has been identified as a positive regulator of cell proliferation and metastasis in several types of cancers. CBX2 is highly expressed in multiple cancers, including ovarian, breast, lung, and prostate cancer, and has been shown to be correlated with poor prognosis in prostate cancer7,8. In breast cancer tissues, CBX2 is associated with cell proliferation and metastasis, suggesting its important role in breast cancer development and progression9. Therefore, targeting CBX2 is anticipated to be a promising therapeutic strategy for breast cancer. Small molecule inhibitors targeting CBX7 have been developed recently10,11. To our knowledge, however, no small molecule compounds targeting CBX2 have been reported. In this study, we conducted in silico screening to identify CG3-46, which inhibits the interaction between CBX2 and H3K27me3. Because CG3-46 suppressed the growth of triple-negative breast cancer cells, this study demonstrates for the first time that small molecule compounds targeting CBX2 have the potential to serve as treatments for breast cancer.
In silico screening to identify compounds that inhibit the interaction between CBX2 and H3K27me3
We performed in silico screening following the guidelines of Simhadri et al. to identify small molecules that inhibit the interaction between CBX2 and H3K27me312. Both H3K27me3 and H3K9me3 were truncated into five-residue fragments (Fig. S1). Specifically, the CBX2 H3K27me3 peptide was truncated to AAR(M3L)S, and the CBX5 H3K9me3 peptide was truncated to TAR(M3L)S, both of which were used as queries for shape and electrostatic potential similarity searches. The similarity search, followed by molecular docking, led to the identification of 77 compounds (Fig. 1A and Table S1). We established a NanoBRET assay to evaluate the inhibitory effects of the identified hit compounds on the interaction between CBX2 and H3K27me3 in cells. The NanoBRET assay system was validated using an active mutant of EZH2 (A677G)13 and CBX2 mutants (V11E/L50D, V11E/L50A) with a loss of binding activity to H3K27me3 (Fig. 1B)2. Among the 77 hit compounds, 4, CG1-24, CG1-30, CG1-44 and CG1-62, inhibited the CBX2-H3K27me3 interaction by more than 35% at 100 µM (Fig. 1C and Table S1).
Identification of CG3-46 by evaluation of structural analogs of active hits
We performed two additional rounds of in silico screening to identify more potent CBX2 inhibitors. Initially, compounds CG1-24 and CG1-44 were used as queries to identify structural analogs because of their promising initial activity profiles. Docking revealed that CG1-24 and CG1-44 bind to the CBX2 protein in a manner that mimics the H3K27me3 peptide (Fig. S2). The 91 compounds were selected and assessed for their ability to inhibit the CBX2-H3K27me3 interaction via the NanoBRET assay, resulting in the identification of four compounds (CG2-26, CG2-32, CG2-89, and CG2-90) with high inhibitory activity (Table S2). These four compounds and CG1-62 were then used as queries for a third round of screening. The docking of CG1-62 suggested that it possessed chemical moieties interacting with amino acid residues at the entrance of the H3K27me3 binding channel as well as with aromatic cage residues Phe12, Trp33, and Trp36 (Fig. S2). The third round of in silico screening led to the identification of another set of 49 compounds (Table S3).
To increase the reliability of our assessments, we employed the NanoBiT assay14 in addition to the NanoBRET assay to evaluate the inhibitory effects of selected compounds on the H3K27me3-CBX2 interaction. Similar to the NanoBRET assay (Fig. 1B), the specificity of the NanoBiT assay was validated via the use of an active mutant of EZH2 (A677G)13 and a binding-deficient CBX2 mutant (V11E/L50D)2 (Fig. 2A). Among the 49 third-round compounds, 6 compounds (CG3-25, -28, -40, -44, -46, and − 49) inhibited the interaction of CBX2-H3K27me3 by more than 50% at 100 µM, as confirmed by both NanoBRET and NanoBiT assays (Fig. 2B and Table S3). Counterassays using Nluc and HaloTag fusion proteins demonstrated that these hit compounds reduced the NanoBRET signal by inhibiting the binding of the CBX2 chromodomain to histone H3 rather than inhibiting energy transfer from the donor to the acceptor (Fig. S3). We then evaluated the concentration dependency of these 6 compounds and calculated their IC50 values via both NanoBRET and NanoBiT assays (Fig. 2C and D, Table 1). The IC50 values of CG3-46, as determined by NanoBRET and NanoBiT assays, were comparable, measuring 15.4 µM and 14.6 µM, respectively. These values were the most potent among the six compounds tested (Table 1). On the other hand, CG3-46 did not affect the cellular H3K27me3 level (Fig. 2E). These results suggest that CG3-46, identified as the most potent CBX2 inhibitor through three rounds of in silico screening, inhibited the interaction of CBX2 and H3K27me3 without altering the cellular H3K27me3 level.
Table 1. Chemical structures of hit compounds selected by the 3rd round of in silico screening and their IC50 values calculated via NanoBRET and NanoBiT assays.
Binding mode of CG3-46 predicted by molecular docking
The binding mode of CG3-46 to CBX2 was explored via molecular docking, revealing a binding pose analogous to that of the H3K27me3 peptide and the broad-spectrum chromodomain inhibitor UNC3866 (Fig. 3A and B), which also demonstrated high inhibitory activity against CBX215. Docking analysis revealed several key interactions that stabilize CG3-46 binding. Notably, hydrogen bonds with Ala14 and Glu9, observed in the binding mode of UNC3866, were also conserved in the predicted binding mode of CG3-46 (Fig. 3B and C). The hydrogen bond formed between the carbonyl group of CG3-46’s acetamide moiety and the backbone nitrogen of Ala14 plays a crucial role in anchoring the compound within the binding pocket. Additionally, the butylamino group of CG3-46 establishes a hydrogen bond with the sidechain carboxylate of Glu9. The bromo- and methoxy-substituted phenyl rings of CG3-46 interact with several hydrophobic residues, including Val11, Trp33, Leu50, and Leu54, further stabilizing CG3-46 binding. Interestingly, the CF3 group of CG3-46 is exposed to the solvent region and does not directly interact with CBX2 residues. However, its presence can influence a compound's overall physicochemical properties, such as solubility and membrane permeability. A notable difference between CG3-46 and UNC3866 is the absence of a trimethylated lysine-mimicking group in CG3-46, which interacts with the aromatic cage residues in UNC3866. Most inhibitors of chromodomains possess a trimethylated lysine-mimicking moiety, yet CG3-46 was found to be reasonably potent despite lacking this moiety. These findings suggest that CG3-46 may utilize alternative mechanisms or interactions to achieve its inhibitory effect, highlighting its potential as a unique and effective CBX2 inhibitor.
Inhibition of the growth of breast cancer cells by CG3-46
Next, we evaluated the effect of CG3-46 on the expression of CBX2-targeted genes. Given that RhoB has been identified as a target gene of EZH2 in breast cancer cells16, it is plausible that RhoB is also a target of CBX2. Indeed, knockdown of CBX2 resulted in elevated RhoB expression in the human breast cancer cell line MDA-MB-231 (Fig. 4A and B). Similarly, treatment with CG3-46 also led to a concentration-dependent increase in RhoB expression in MDA-MB-231 cells, providing further evidence that CG3-46 inhibits CBX2 activity in cells (Fig. 4C). Finally, we examined the anti-breast cancer activity of CG3-46, as CBX2 is known to play a significant role in breast cancer cell growth9. Consistent with previous findings, CBX2 knockdown led to reduced proliferation of MDA-MB-231 cells (Fig. 4D). Importantly, CG3-46 inhibited the growth of MDA-MB-231 cells (Fig. 4E) in a dose range that reduced the interaction between CBX2 and H3K27me3 (Fig. 2C and D), as well as the expression of RhoB (Fig. 4C). These results suggest that CG3-46 inhibited the proliferation of breast cancer cells by suppressing cellular CBX2 activity.
Aberrant levels of H3K27me3 are strongly involved in carcinogenesis. Accordingly, modulators of H3K27me3 have attracted attention as targets for cancer therapy. Indeed, considerable progress has been made in the development of small molecule inhibitors targeting the histone H3K27 methyltransferase EZH2, with an approved therapeutic now available17,18. In contrast, drug development targeting CBX family members, which are directly involved in transcriptional repression by the H3K27me3 mark, has lagged far behind that of EZH2 inhibitors. Recently, peptide-based antagonists with selectivity and high affinity for CBX family members, including CBX2, have been reported12,15,19. However, these molecules are typically characterized by poor cellular permeability. In contrast, CG3-46, identified in this study, emerged as the first small molecule capable of effectively inhibiting CBX2 at the cellular level. CG3-46 inhibited breast cancer cell growth at similar concentrations and inhibited the activity of intracellular CBX2, providing novel support of the concept that small molecule inhibitors targeting CBX2 could be effective for breast cancer treatment. Thus, our findings provide not only a chemical tool to elucidate the cellular roles of CBX2 but also valuable insights for drug discovery research targeting CBX2.
Cell culture
HEK293T, PLAT-E and HeLa cells were maintained in Dulbecco’s modified Eagle’s medium (Thermo Fisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum (FBS) and penicillin/streptomycin. MDA-MB-231 cells were maintained in RPMI 1640 (Wako) supplemented with 10% FBS and penicillin/streptomycin. All the cells were cultured at 37°C under 5% CO2.
Plasmid construction
Histone H3 and CBX2 (chromodomain (CD), 1–73 aa) genes were amplified with PrimeSTAR Max DNA polymerase (Takara), subcloned, and inserted into the EcoRI/XhoI sites of the pBiT1.1-C TK-LgBiT vectors and pBiT2.1-C TK-SmBiT vectors (Promega), respectively. H3-LgBiT and CBX2-CD-SmBiT were subcloned and inserted into the EcoRI/NotI sites of the retroviral expression vectors pCX4-pur and pCX4-bls, respectively. The histone H3-HaloTag vector and Nluc-CBX2-CD vector for the NanoBRET assay were generated by Promega.
Establishment of a stable cell line
Retroviral infection was used as previously described to establish cell lines stably expressing H3-LgBiT and CBX2-CD-SmBiT20. Briefly, PLAT-E cells were cotransfected with pcDNA3.1-VSVG and pCX4-pur-H3-LgBiT and pCX4-bls-CBX2-CD-SmBiT in combination with 5 µg/mL polyethylenimine MAX (Polysciences) and incubated for 2 days. The viral supernatant containing 8 µg/mL polybrene (Sigma) was added to the HeLa cells. After incubation for 2 days, the infected cells were selected via treatment with 1 µg/mL puromycin and 1 µg/mL blasticidin S. Clonal puromycin/blasticidin S-resistant cells were established by limiting dilution. The desired clones were confirmed by a luciferase activity assay.
In silico screening
In silico screening was conducted following the protocol outlined in Fig. 1A. Initially, a library of approximately six million commercially available compounds was filtered based on shape and electrostatic potential similarities with queries derived from the H3K27me3 peptide-bound crystal structure of CBX2 (PDB code 3H91) and the H3K9me3 peptide-bound crystal structure of CBX5 (PDB code 3FDT). Shape similarity calculations were performed via the ROCS program21. Up to 200 ligand conformations were generated via the OMEGA program22. The screening library was ranked using TanimotoCombo. The top 2000 compounds were then evaluated for electrostatic potential similarities via the EON program (OpenEye, Cadence Molecular Sciences). The compounds were ranked via the ET_Combo score, and the top 1000 were selected for molecular docking. Molecular docking was carried out via the Glide program23, with ligands prepared through Schrodinger's LigPrep (Schrödinger, LLC.), considering ionization and tautomeric states at a target pH of 7.0 ± 2.0. Protein structures were prepared via Maestro (Schrödinger, LLC.), ensuring proper hydrogen addition, bond-order assignment, and protonation states at neutral pH. The compounds were ranked via Glide's docking score, incorporating Epik penalties. Following the identification of the initial active compounds, two additional rounds of in silico screening were performed to identify structural analogs. The most active compounds from each round served as queries for subsequent shape and electrostatic similarity searches. All graphics were generated via PyMOL (Schrödinger, LLC.).
NanoBRET assay
HEK293T cells were cotransfected with the NanoBRET H3-HaloTag vector, the NanoBRET assay vector Nluc-CBX2-CD and the pMSCV-hygro-FLAG-Ezh2-A677G vector via FuGENE HD (Promega) for 24 h. Cells with 100 nM NanoBRET 618 Ligand were seeded into 96-well white plates, and after incubation for 6 h, various concentrations of compounds were added. After incubation for an additional 24 h, 10 µM Nano-Glo Luciferase Assay Substrate was added, and detection was performed using an EnVision microplate reader (PerkinElmer). A true hit assay was performed using the NanoBRET positive control vector in HEK293T cells. The BRET ratio was calculated as described previously24.
NanoBiT assay
HeLa cells expressing H3-LgBiT, CBX2-CD-SmBiT were seeded into 96-well white plates. Various concentrations of compounds were added, and the plates were incubated for 6 h. The luciferase assay was performed via a Nano-Glo Live Cell Assay System (Promega), and the luminescence was measured via an EnSpire microplate reader (PerkinElmer).
siRNA transfection
MDA-MB-231 cells were transfected with the 10 nM siRNA listed in Table S4 via the Lipofectamine RNAiMAX reagent (Thermo Fisher Scientific). AllStars Negative Control siRNA (QIAGEN) was used as a negative control.
Western blotting
Cell lysis and western blotting were performed according to a previous protocol20. The primary antibodies used are listed in Table S4.
Quantitative PCR
Total cellular RNA was prepared via the Tissue Total RNA Mini Kit (FAVORGEN), and cDNA was synthesized via ReverTra Ace Master Mix (TOYOBO). The expression of target genes was analyzed via THUNDERBIRD SYBR qPCR Mix (TOYOBO) on a CFX Connect Real-Time System (Bio-Rad). The primers used are listed in Table S4. The primers for GAPDH were used as a reference.
Cell viability assay
Cell viability was analyzed via either the ATPlite (PerkinElmer) or the WST-8 assay, as described previously20.
Statistical analysis
Significance was determined via one-way analysis of variance (ANOVA) with Tukey’s multiple comparison test or two-tailed t test. A P value < 0.05 was considered statistically significant.
Acknowledgments
We thank Dr. T. Kitamura (The University of Tokyo) for providing the PLAT-E cells and Mr. T. Imaizumi and Ms. M. Kawakatsu (Tokyo University of Pharmacy and Life Sciences) for his technical help. We acknowledge RIKEN ACCC for the resources at the Hokusai Big Waterfall supercomputer used in this study. This work was supported by a Grant-in-Aid for Scientific Research (B) (24K01684) to A.I. and a Grant-in-Aid for Scientific Research (S) (23H05473) and to M.Y. from the Japan Society for the Promotion of Science (JSPS) and a Grant-in-Aid for Transformative Research Areas (A) (JP23H04882) to M.Y. from the Ministry of Education, Culture, Sports, Science and Technology, Japan.
Author contributions
S.T. and A.K. performed the experiments and wrote the initial manuscript. S.Y., M.A., H.M. and S.D. performed the experiments. Y.M., Y.S.Y., H.I. and M.H. guided and supervised this study. M.Y. designed this study and edited the manuscript. K.Y.J.Z. supervised this study and wrote and edited the manuscript. A.I. designed this study and wrote and edited the manuscript. All authors reviewed the manuscript.
Data availability statement
The authors confirm that all relevant data are included in the article.
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
- Schuettengruber, B., Chourrout, D., Vervoort, M., Leblanc, B. & Cavalli, G. Genome regulation by polycomb and trithorax proteins. Cell. 128, 735–745 (2007).
- Kaustov, L. et al. Recognition and specificity determinants of the human cbx chromodomains. J. Biol. Chem. 286, 521–529 (2011).
- Cooper, S. et al. Jarid2 binds mono-ubiquitylated H2A lysine 119 to mediate crosstalk between Polycomb complexes PRC1 and PRC2. Nat. Commun. 7, 13661 (2016).
- Whitcomb, S. J., Basu, A., Allis, C. D. & Bernstein, E. Polycomb Group proteins: an evolutionary perspective. Trends Genet. 23, 494–502 (2007).
- Mills, A. A. Throwing the cancer switch: reciprocal roles of polycomb and trithorax proteins. Nat. Rev. Cancer 10, 669–682 (2010).
- Koppens, M. & van Lohuizen, M. Context-dependent actions of Polycomb repressors in cancer. Oncogene 35, 1341–1352 (2016).
- Clermont, P. -L. et al. Genotranscriptomic meta-analysis of the Polycomb gene CBX2 in human cancers: initial evidence of an oncogenic role. Br. J. Cancer 111, 1663–1672 (2014).
- Clermont, P. -L. et al. Identification of the epigenetic reader CBX2 as a potential drug target in advanced prostate cancer. Clin. Epigenetics 8, 16 (2016).
- Zheng, S. et al. Overexpression of CBX2 in breast cancer promotes tumor progression through the PI3K/AKT signaling pathway. Am. J. Transl. Res. 11, 1668–1682 (2019).
- Ren, C. et al. Small-molecule modulators of methyl-lysine binding for the CBX7 chromodomain. Chem. Biol. 22, 161–168 (2015).
- Ren, C. et al. Structure-Guided Discovery of Selective Antagonists for the Chromodomain of Polycomb Repressive Protein CBX7. ACS Med. Chem. Lett. 7, 601–605 (2016).
- Simhadri, C. et al. Chromodomain antagonists that target the polycomb-group methyllysine reader protein chromobox homolog 7 (CBX7). J. Med. Chem. 57, 2874–2883 (2014).
- McCabe, M. T. et al. Mutation of A677 in histone methyltransferase EZH2 in human B-cell lymphoma promotes hypertrimethylation of histone H3 on lysine 27 (H3K27). Proc. Natl. Acad. Sci. U. S. A. 109, 2989–2994 (2012).
- Dixon, A. S. et al. NanoLuc Complementation Reporter Optimized for Accurate Measurement of Protein Interactions in Cells. ACS Chem. Biol. 11, 400–408 (2016).
- Stuckey, J. I. et al. A cellular chemical probe targeting the chromodomains of Polycomb repressive complex 1. Nat. Chem. Biol. 12, 180–187 (2016).
- Lei, C. et al. LncRNA DUXAP8 induces breast cancer radioresistance by modulating the PI3K/AKT/mTOR pathway and the EZH2-E-cadherin/RHOB pathway. Cancer Biol. Ther. 23, 1–13 (2022).
- Bradley, W. D. et al. EZH2 inhibitor efficacy in non-Hodgkin's lymphoma does not require suppression of H3K27 monomethylation. Chem. Biol. 21, 1463–1475 (2014).
- Campbell, J. E. et al. EPZ011989, A Potent, Orally-Available EZH2 Inhibitor with Robust in Vivo Activity. ACS Med. Chem. Lett. 6, 491–495 (2015).
- Wang, S. et al. A Potent, Selective CBX2 Chromodomain Ligand and Its Cellular Activity During Prostate Cancer Neuroendocrine Differentiation. Chembiochem 22, 2335–2344 (2021).
- Sekine, S. et al. Identification of a derivative of the alkaloid emetine as an inhibitor of the YAP-TEAD interaction and its potential as an anticancer agent. Biosci. Biotechnol. Biochem. 87, 501–510 (2023).
- Hawkins, P. C. D., Skillman, A. G. & Nicholls, A. Comparison of shape-matching and docking as virtual screening tools. J. Med. Chem. 50, 74–82 (2007).
- Hawkins, P. C. D., Skillman, A. G., Warren, G. L., Ellingson, B. A. & Stahl, M. T. Conformer generation with OMEGA: algorithm and validation using high quality structures from the Protein Databank and Cambridge Structural Database. J. Chem. Inf. Model 50, 572–584 (2010).
- Friesner, R. A. et al. Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy. J. Med. Chem. 47, 1739–1749 (2004).
- Machleidt, T. et al. NanoBRET–A Novel BRET Platform for the Analysis of Protein-Protein Interactions. ACS Chem. Biol. 10, 1797–1804 (2015).
No competing interests reported.