Selective Inhibition of Lysine-Specific Demethylase 5A (KDM5A) Using a Rhodium(III) Complex for Triple-Negative Breast Cancer Therapy.

Angewandte A Journal of the Gesellschaft Deutscher Chemiker Chemie International Edition www.angewandte.org Accepted Article Title:Selective inhibition of lysine-specific demethylase 5A (KDM5A) using a rhodium(III) complex for triple-negative breast cancer therapy Authors:Chung-Hang Leung, Guan-Jun Yang, Wanhe Wang, Simon Wing Fai Mok, Chun Wu, Betty Yuen Kwan Law, Xiang-Min Miao, Ke-Jia Wu, Hai-Jing Zhong, Chun-Yuen Wong, Vincent Kam Wai Wong, and Dik-Lung Ma This manuscript has been accepted after peer review and appears as an Accepted Article online prior to editing, proofing, and formal publication of the final Version of Record (VoR). This work is currently citable by using the Digital Object Identifier (DOI) given below. The VoR will be published online in Early View as soon as possible and may be different to this Accepted Article as a result of editing. Readers should obtain the VoR from the journal website shown below when it is published to ensure accuracy of information. The authors are responsible for the content of this Accepted Article. To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201807305 Angew. Chem. 10.1002/ange.201807305 Link to VoR: http://dx.doi.org/10.1002/anie.201807305 http://dx.doi.org/10.1002/ange.201807305 10.1002/anie.201807305 COMMUNICATION Selective inhibition of lysine-specific demethylase 5A (KDM5A) using a rhodium(III) complex for triple-negative breast cancer therapy Guan-Jun Yang,#[a] Wanhe Wang,#[b] Simon Wing Fai Mok,[c] Chun Wu,[b] Betty Yuen Kwan Law,[c] Xiang-Min Miao[d], Ke-Jia Wu,[a]Hai-Jing Zhong,[a] Chun-Yuen Wong,[e] Vincent Kam Wai Wong,*[c] Dik- Lung Ma,*[b] and Chung-Hang Leung*[a] Abstract: Lysine-specific demethylase 5A (KDM5A) has recently become a promising target for epigenetic therapy. In this study, we designed and synthesized metal complexes bearing ligands with reported demethylase and p27 modulating activities. The Rh(III) complex 1 was identified as a direct, selective and potent inhibitor of KDM5A that directly abrogate KDM5A demethylase activity via antagonizing the KDM5A-tri-/di-methylated histone 3 protein–protein interaction (PPI) in vitro and in cellulo. Complex 1 induced accumulation of H3K4me3 and H3K4me2 levels in cells, causing growth arrest at G1 phase in the triple negative breast cancer (TNBC) cell lines, MDA-MB-231 and 4T1. Finally, 1 exhibited potent anti-tumour activity against TNBC xenografts in an in vivo mouse model, presumably via targeting of KDM5A and hence upregulating p27. Moreover, complex 1 was less toxic compared with two clinical drugs, cisplatin and doxorubicin. To our knowledge, complex 1 is the first metal-based KDM5A inhibitor reported in the literature. We anticipate that complex 1 may be utilized as a novel scaffold for the further development of more potent epigenetic agents against cancers, including TNBC. Figure 1. Chemical structures of the cyclometallated Rh(III) Triple negative breast cancers (TNBC) are characterized by and Ir(III) complexes 1–5 (racemates), ligands 6 and 7, positive control 8 (CPI-455), compounds 9 and 10 aberrations in the TP53 gene, enhancement in ability of (analogues of 6 and 7) are evaluated in this study. migration and metastasis, and high risk of recurrence, rendering it as a challenging target for molecular therapies.[1] Current chemotherapeutic strategies for TNBC treatment are limited by [a] G.-J. Yang, K.-J. Wu, H.-J. Zhong, Prof. Dr. C.-H. Leung drug resistance and serious side effects such as organ damage, Institute of Chinese Medical Sciences and State Key Laboratory of which increases the suffering of TNBC patients.[2] Therefore, Quality Research in Chinese Medicine University of Macau there is an urgent need to develop new targeted approaches to Macao SAR (China) treat this type of cancer. Recent studies have identified a strong E-mail: duncanleung@umac.mo association between lysine-specific demethylase 5A (KDM5A) [b] W.-H. Wang, C. Wu, Prof. Dr. D.-L. Ma and TNBC.[3] KDM5A belongs to the KDM5 subfamily of Jumonji Department of Chemistry Hong Kong Baptist University histone demethylases that removes di- and tri-methylation marks Kowloon Tong, Hong Kong (China) from lysine 4 of histone H3.[4] KDM5A is overexpressed in breast E-mail: edmondma@hkbu.edu.hk cancer[3a] and multiple other human cancer types, including [c] S.W.F. Mok, B.Y.K. Law, Pro. Dr. V.K.W. Wong glioblastoma,[5] lung,[6] and gastric[7] cancers, where it promotes State Key Laboratory of Quality Research in Chinese Medicine, Macau University of Science and Technology, tumorigenesis, metastasis and drug tolerance of these Avenida Wai Long, Taipa, Macao, China cancers.[8] Several KDM5A inhibitors have been reported in E-mail: bowaiwong@gmail.com recent years, but most lack selectivity for the enzyme over other [d] Dr. X.-M. Miao demethylases, which complicates the interpretation of biological School of Life Science, Jiangsu Normal University, Xuzhou 221116, PR China and clinical studies that make use of these compounds.[9] [e] Dr. C.-Y. Wong Department of Chemistry, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong SAR, China Supporting information for this article is given via a link at the end of the document. # Authors contribute equally to the work. * Corresponding authors tpircsunaM detpeccA Angewandte Chemie International Edition This article is protected by copyright. All rights reserved. 10.1002/anie.201807305 COMMUNICATION Figure 3. Thermal stabilization of KDM5A by complex 1 in Figure 2. Inhibition of KDM5A demethylase activity in cell lysates. (A) MDA-MB-231 cell lysates were treated with vitro was determined using a chemiluminescence assay. 1 or 8 at 1.0 or 3.0 μM. KDM5A content in the soluble (A) Effect of complexes 1–5 (3.0 µM) on KDM5A activity. fraction was determined by Western blotting. (B) (B) 1 inhibits KDM5A activity in a dose-dependent Densitometry analysis of KDM5A content. manner. *P < 0.05 compared to DMSO-treated group. ligand is important for KDM5A inhibitory activity. Consistent with Organometallic compounds based on transition metals such this, the Rh(III) complex 3 also bearing two 2-phenylquinoline as rhodium and iridium have emerged as promising scaffolds for C^N ligands as well as a 2,2’-bipyridine N^N ligand also showed antitumor leads.[10] Organometallic complexes can display great powerful KDM5A antagonist activity and was only slightly less structural diversity of geometrical shapes depending on the potent than complex 1. Additionally, the isolated ligands of the oxidation state of their metal center, and variation of their co- Rh(III) complex 1, 4,4'-diphenyl-2,2'-bipyridine (6) and 2- ligands.[11] With readily tunable steric and electronic phenylquinoline (7), as well as their non-heterocyclic analogues characteristics, organic complexes can be optimized to (9 and 10) were tested for their ability to inhibit KDM5A effectively target the active sites of proteins or enzymes through demethylase activity and p27 expression. The results showed shape-specific interactions.[12] Anchoring bioactive ligands to a that neither the isolated ligands, 6 and 7, nor their non- metal center is a versatile approach to generate new therapeutic heterocyclic analogues 9 and 10 displayed significant bioactivity metal complex leads. 2,2’-Bipyridine and 2-phenylquinoline have both in vitro and in cellulo (Figures S8). These results illustrate been reported as structural moieties for synthesizing compounds the importance of the Rh(III) center in arranging the ligands into displaying DNA demethylase and p27KIP1 modulating activities, a structurally active conformation within the architecture of the respectively.[13] p27, a cyclin-dependent kinase (Cdk) inhibitor organometallic complex. regulated by KDM5A,[6] represses cell proliferation and motility, Interestingly, the Rh(III) complex 1 was much more potent and induces apoptosis in TNBC.[14] In this study, we designed than the Ir(III) congener 4 despite having the same ligands and and synthesized Rh(III) and Ir(III) complexes 1–5 bearing same charge. To understand this difference, the stability, redox ligands related to 2,2’-bipyridine and 2-phenylquinoline (Figure activity, size and hydrophobicity of the two complexes were 1)[15] and explored their effect on KDM5A demethylase and p27 further studied. Complexes 1 and 4 showed similar stability in 80% activity. The rhodium(III) complex 1 was identified as the first acetonitrile/20% aqueous buffer as monitored by UV-Vis metal-based inhibitor of KDM5A activity that suppressed breast absorption spectroscopy (Figure S2) and 1H NMR spectroscopy cancer cell growth in vivo. (Figure S3). Cyclic voltammetry experiments showed almost Detailed methods for the synthesis of complexes 1, 2, 3, and 4 indistinguishable redox potentials (Figure S4), while theoretical are described in the ESI, while complex 5 has been previously calculations showed that the two complexes had similar sizes reported.[16] As methylated H3 peptides are known substrates for (Figure S5). However, 1 had a longer retention time than 4 KDM5A, we initially screened the complexes for their ability to (26.45 min vs. 24.55 min) in reversed phase HPLC, indicating modulate the protein-protein interaction (PPI) between tri- that 1 is more hydrophobic. Additionally, ICP-MS results showed methylated lysine histone 4 peptide (H3K4me3) and KDM5A in that complex 1 has significant higher cellular uptake in MDA-MB- vitro using a chemiluminescence assay. CPI-455 (8), a reported 231 cells than complex 4 (Figure S7). As hydrophobicity is an KDM5A inhibitor,[9b] was used as a positive control. From the important determinant of both binding activity and cellular preliminary screening results, the Rh(III) complex 1 emerged as permeability,[18] the enhanced activity and higher intracellular the most active candidate, with higher activity compared to the concentration of complex 1 compared to complex 4 could be due, positive control compound 8 at 3.0 µM (Figure 2A). In a dose- at least in part, to its higher hydrophobicity. response experiment, complex 1 inhibited KDM5A activity with To explore whether complex 1 could target KDM5A in cellulo, an IC value of 23.2 ± 1.8 nM (Figure 2B). Moreover, complex 1 the ability of complex 1 to induce histone methylation in MDA- 50 was more potent against KDM5A activity (84.3% inhibition) MB-231 cells was investigated by Western blotting (Figure S9A). compared to other demethylases including KDM1A (61.1% Complex 1 increased the levels of H3K4me2 (Figure S9B) and inhibition), KDM4A (10.2% inhibition) and KDM6B (58.4% H3K4me3 (Figure S9C) without affecting the expression level of inhibition) (Figure S1). Therefore, 1 may function as a selective KDM5A (Figure S9D). Moreover, a KDM5A knockdown assay KDM5A inhibitor. showed that ablation of KDM5A phenocopied the The Rh(III) complex 1, bearing two 2-phenylquinoline C^N pharmacological effects of KDM5A inhibition by complex 1 ligands and a 4,4'-diphenyl-2,2'-bipyridine N^N ligand, showed (Figure S10A-B). Both KDM5A knockdown and complex 1 higher potency than the Rh(III) complex 2, which contains the treatment induced a significant increase in H3K4me3 (Figure same N^N ligand as 1 but bears two 2-phenylpyridine C^N S10C) and H3K4me2 (Figure S10D) levels. Moreover, KDM5A ligands instead. This result suggests that the 2-phenylquinoline knockdown sharply reduced the cytotoxicity of 1 in MDA-MB-231 tpircsunaM detpeccA Angewandte Chemie International Edition This article is protected by copyright. All rights reserved. 10.1002/anie.201807305 COMMUNICATION Figure 4. Complex 1 raises p27 level by disrupting the interaction between KDM5A and its targets. Interactions between KDM5A and its targets H3K4me3, H3K4me2, and H3 in MDA-MB-231 cells were examined by (A) Western blotting and (B) co-IP. MDA-MB-231 cells were treated with DMSO and indicated concentrations of 1 or 8 for 6 h. Cell lysates Figure 5. Complex 1 exhibits antitumour effects in an in were subjected to immunoprecipitation with the appropriate vivo mouse model. Female BALB/c mice were antibody. (C) Effect of 1 or 8 on the binding ability of subcutaneously inoculated with 1.0×106 4T1 mouse breast H3K4me3 to p27 promoter in MDA-MB-231 cells by ChIP- cancer cells, and injected intraperitoneally daily with qPCR. After incubation with DMSO, 1 (3.0 μM) or 8 (3.0 μM) complex 1, cisplatin, and doxorubicin hydrochloride when for 6 h, cells were harvested and ChIP-qPCR was performed. the tumour volume reached 60 mm3. (A) Tumour volume of (D-F) Effect of 1 or 8 on the p27 levels in MDA-MB-231 and 4T1 tumour-bearing mice after control (vehicle), 1 (2.0 or 4T1 cells by Western blotting after 6 h treatment using the 4.0 mg/kg), cisplatin (1.5 mg/kg), and doxorubicin (Dox, 1.5 indicated concentrations of 1 and 8. *P < 0.05 compared with mg/kg) treatments (n = 6-8). (B) Representative photograph control group. images of tumour tissues in vehicle control, complex 1- treated groups, cisplatin- and Dox-treated groups after 9 cells (Figure S8E) compared to control cells (Figure S10F). days. (C) Body weight of 4T1 tumour-bearing mice in Finally, a cellular thermal shift assay (CETSA)[20] showed that vehicle control and teatment groups. (D-E) Mean weight of complex 1 could significantly stabilize KDM5A in MDA-MB-231 spleen and kidney from drug-treated mice after 9 days. (F- cell lysates compared to DMSO-treated controls (ΔTm: 1: 6.5 °C G) Survival curves of healthy C57 and BALB/C mice after at 1.0 µM and 16.0 °C at 3.0 µM) (Figure 3), suggesting that 1 treatment with 1, cisplatin or Dox at 4 mg/kg per day (n = 6). can engage KDM5A even in the intricate biochemical environment of the cell lysates. Taken together, our data (H) Rhodium content in acid digests of mouse tumuor demonstrate that 1 is a direct, potent, and selective inhibitor of tissues as determined by ICP-MS. *P < 0.05 compared to KDM5A both in vitro and in cellulo. the vehicle-treated group analysed by t-test. (I) 1 impaired Experiments were performed to further explore the KDM5A activity in tumour tissues. mechanism of action and downstream effects of complex 1 in a dose-dependent fashion in both MDA-MB-231 and 4T1 cells regulating KDM5A activity. A co-immunoprecipitation (co-IP) (Figure 4C-F). Additionally, complex 1 significantly increased the assay showed that 1 impaired the interaction between KDM5A ratio of G1 cell population and raised the expression level of p27 and H3K4me3 or H3K4me2 in cellulo, as indicated by a in both MDA-MB-231 and 4T1 cell lines (Figure S11A-B and reduction of KDM5A co-precipitated with H3K4me3/H3K4me2 in Figure 4D-F). Taken together, these results suggest that 1 may MDA-MB-231 cells (Figure 4A-B). This result could account for function as a transcriptional modulator of p27 expression, the ability of complex 1 to induce H3K4me3/H3K4me2 presumably via suppression of KDM5A demethylase activity. accumulation in cellulo as described above. KDM5A is overexpressed in many breast cancer cell lines.[3a, p27, a cell cycle-related gene, is regulated by KDM5A.[6-7] 3b, 21] The cytotoxicity of complex 1 was assessed in different KDM5A impairs the binding ability of H3K4me3 to the p27 breast cancer cell lines (4T1 cells, MCF-7, MDA-MB-231, and promoter, thus downregulating its expression in cancer cells.[6-7] MCF-10A) and normal cell lines (HEK293T and LO2) using the Given that both KDM5A knockdown and complex 1 treatment MTT assay. Complex 1 exhibited potent anti-proliferative activity induced a significant increase in H3K4me3 levels (Figure S9D), against the breast cancer cell lines, particularly the TNBC cell we hypothesized that 1 might upregulate p27 via enhancing the lines MDA-MB-231 (IC : 90.1 ± 8.4 nM) and 4T1 (68.2 ± 9.7 50 interaction between H3K4me3 and the p27 promoter. To verify nM), while it showed lower cytotoxicity to the normal human cell this hypothesis, the chromatin immunoprecipitation (ChIP) assay lines HEK239T and LO2, and non-TNBC breast cancer cell lines was performed. ChIP and Western blotting analysis indicated MCF-10A and MCF-7 (Figure S12A-F). Moreover, complex 1 that 3.0 μM of 1 could increase H3K4me3 binding to the p27 significantly inhibited colony formation in both MDA-MB-231 gene promoter in MDA-MB-231 cells and raise its protein level in (Figure S13A) and 4T1 (Figure S13B) cell lines. We consider tpircsunaM detpeccA Angewandte Chemie International Edition This article is protected by copyright. All rights reserved. 10.1002/anie.201807305 COMMUNICATION that the cytotoxicity displayed by 1 could be attributed, at least in This work is supported by Hong Kong Baptist University part, to the inhibition of KDM5A demethylase activity in cellulo. (FRG2/15-16/002), the Health and Medical Research Fund The effect of 1 on tumour growth was further investigated in (HMRF/14130522), the Research Grants Council an in vivo model of TNBC. Cisplatin and doxorubicin (Dox), two (HKBU/12301115, HKBU/204612, HKBU/201913, clinical chemotherapeutic drugs for TNBC, were used as positive CityU/11228316 and CityU/11207117), the National Natural controls. As shown in Figure 5A-B, daily intraperitoneal (IP) Science Foundation of China (21575121), the Guangdong injection of complex 1, cisplatin or doxorubicin led to significant Province Natural Science Foundation (2015A030313816), the inhibition of 4T1 tumour growth in vivo, with up to 35% (P < 0.05) Hong Kong Baptist University Century Club Sponsorship and 48% (P < 0.05) inhibition of tumour growth for 2.0 mg/kg Scheme 2016, the Interdisciplinary Research Matching Scheme and 4.0 mg/kg of 1 respectively over the course of treatment (RC-IRMS/15-16/03), the Science and Technology Development (Figures 5A-B). Although cisplatin and Dox exhibited slightly Fund, Macao SAR (098/2014/A2), the University of Macau more potency than complex 1 (Figures 5A-E), they also (MYRG2015-00137-ICMS-QRCM, MYRG2016-00151-ICMS- produced significant toxicity to mouse organs (spleen and kidney) QRCM), National Natural Science Foundation of China and caused a significant drop of body weight. In contrast, 1 was (21628502). well tolerated in vivo as indicated by its mild effects on mouse body weight change (Figure 5C) and organ weight change Keywords: KDM5A • epigenetics • metal complex • triple- (Figure 5D-E) over the duration of treatment. In addition, a negative breast cancer toxicological study showed that both cisplatin and Dox, but not complex 1, significantly lowered survival of both C57 and Balb/c [1] a) E. Andreopoulou, S. J. Schweber, J. A. Sparano, H. M. mice over 9 days of treatment (Figure 5F-G). McDaid, Expert Opin. Pharmacother. 2015, 16, 983-998; b) To further validate the mechanism of anti-tumour activity of H. Liu, C. Murphy, F. Karreth, K. Emdal, F. White, O. Elemento, A. Toker, G. Wulf, L. Cantley, Cancer Discov. complex 1, tumour specimens were examined after sacrifice. 2018, 8, 354-369. Tumours in complex 1-treated mice showed the increase of [2] A. Lee, M. Djamgoz, Cancer Treat. Rev. 2018, 62, 110-122. rhodium content in a dose-dependent fashion (Figure 5H). [3] a) J. Cao, Z. Liu, W. K. Cheung, M. Zhao, S. Y. Chen, S. W. Moreover, H3K4me3 and H3K4me2 levels were elevated in Chan, C. J. Booth, D. X. Nguyen, Q. Yan, Cell Rep. 2014, 6, tumour tissues treated with complex 1, suggesting that KDM5A 868-877; b) J. Hou, J. Wu, A. Dombkowski, K. Zhang, A. demethylase activity was dramatically inhibited (Figure 5I). Holowatyj, J. L. Boerner, Z. Q. Yang, Am. J. Transl. Res. Consistent with the in vitro results, p27 levels in tumour tissues 2012, 4, 247-256; c) L. Wang, J. Chang, D. Varghese, M. Dellinger, S. Kumar, A. M. Best, J. Ruiz, R. Bruick, S. Pena- were also increased by complex 1 (Figure 5I). Overall, the data Llopis, J. Xu, D. J. Babinski, D. E. Frantz, R. A. Brekken, A. suggest that 1 exhibits potent in vivo anti-cancer activity via its M. Quinn, A. Simeonov, J. Easmon, E. D. Martinez, Nat. ability to inhibit KDM5A activity, and could therefore represent a Commun. 2013, 4, 2035. potential drug candidate for TNBC treatment. [4] a) E. V. Benevolenskaya, H. L. Murray, P. Branton, R. A. In conclusion, we have developed a selective and potent Young, W. G. Kaelin, Jr., Mol. Cell 2005, 18, 623-635; b) W. KDM5A inhibitor 1 by coupling a Rh(III) core with bioactive Lin, J. Cao, J. Liu, M. L. Beshiri, Y. Fujiwara, J. Francis, A. motifs previously reported as demethylase inhibitors or p27KIP1 D. Cherniack, C. Geisen, L. P. Blair, M. R. Zou, X. Shen, D. Kawamori, Z. Liu, C. Grisanzio, H. Watanabe, Y. A. modulators. The hybrid complex 1 showed high selectivity and Minamishima, Q. Zhang, R. N. Kulkarni, S. Signoretti, S. J. potency for KDM5A over KDM1A, KDM4A, and KDM6A. SAR Rodig, R. T. Bronson, S. H. Orkin, D. P. Tuck, E. V. analysis indicated that both the nature of the metal ion and the Benevolenskaya, M. Meyerson, W. G. Kaelin, Jr., Q. Yan, character of the ligands were important for the biological activity Proc. Natl. Acad. Sci. USA 2011, 108, 13379-13386. of 1 with the identity of the metal center playing a particularly key [5] B. Banelli, E. Carra, F. Barbieri, R. Wurth, F. Parodi, A. role in governing the hydrophobicity and cell permeability of the Pattarozzi, R. Carosio, A. Forlani, G. Allemanni, D. Marubbi, T. Florio, A. Daga, M. Romani, Cell Cycle 2015, 14, 3418- complex. In terms of mechanism, 1 inhibited the PPI between 3429. KDM5A and H3K4me3/2 as revealed by multiple biochemical [6] Y. C. Teng, C. F. Lee, Y. S. Li, Y. R. Chen, P. W. Hsiao, M. assays, leading to the accumulation of H3K4me3/2 and the Y. Chan, F. M. Lin, H. D. Huang, Y. T. Chen, Y. M. Jeng, C. activation of the p27 gene. Additionally, complex 1 exerted H. Hsu, Q. Yan, M. D. Tsai, L. J. Juan, Cancer Res. 2013, antiproliferative activity and induced cell-cycle arrest in multiple 73, 4711-4721. breast cancer cell lines, presumably through inhibition of KDM5A [7] J. Zeng, Z. Ge, L. Wang, Q. Li, N. Wang, M. Bjorkholm, J. Jia, D. Xu, Gastroenterology 2010, 138, 981-992. activity. Finally, complex 1 significantly repressed tumor [8] P. B. Rasmussen, P. Staller, Epigenomics 2014, 6, 277-286. xenograft growth in a murine model of TNBC presumably [9] a) X. Wu, Z. Fang, B. Yang, L. Zhong, Q. Yang, C. Zhang, S. through inhibition of KDM5A activity and hence upregulating p27, Huang, R. Xiang, T. Suzuki, L. Li, S. Yang, Bioorg. Med. without causing observable toxicity to the mice. Moreover, Chem. Lett. 2016, 26, 2284-2288; b) M. Vinogradova, V. S. despite having slightly lower anti-tumour efficacy than cisplatin Gehling, A. Gustafson, S. Arora, C. A. Tindell, C. Wilson, K. or doxorubicin, complex 1 showed much lower toxicity in vivo E. Williamson, G. D. Guler, P. Gangurde, W. Manieri, J. than the two clinical drugs, making it a potential lead scaffold for Busby, E. M. Flynn, F. Lan, H. J. Kim, S. Odate, A. G. Cochran, Y. Liu, M. Wongchenko, Y. Yang, T. K. Cheung, T. TNBC drug development. M. Maile, T. Lau, M. Costa, G. V. Hegde, E. Jackson, R. Pitti, D. Arnott, C. Bailey, S. Bellon, R. T. Cummings, B. K. Albrecht, J. C. Harmange, J. R. Kiefer, P. Trojer, M. Classon, Acknowledgements Nat. Chem. Biol. 2016, 12, 531-538 tpircsunaM detpeccA Angewandte Chemie International Edition This article is protected by copyright. All rights reserved. 10.1002/anie.201807305 COMMUNICATION [10] a) C.-H. Leung, H.-J. Zhong, D. S.-H. Chan, D.-L. Ma, Wong, W. H. Chung, F. Y. Chan, P. K. So, T. S. Lai, Z. Y. Coord. Chem. Rev 2013, 257, 1764-1776; b) Y. Gothe, T. Zhou, Y. C. Leung, K. Y. Wong, Angew. Chem. Int. Ed. Engl. Marzo, L. Messori, N. Metzler-Nolte, Chem.–Eur. J. 2016, 2008, 120, 3795-3799. 22, 12487-12494. c) J. S. Nam, M.-G. Kang, J. Kang, S.-Y. [16] a) L. J. Liu, L. Lu, H. J. Zhong, B. He, D. W. Kwong, D. L. Park, S. J. C. Lee, H.-T. Kim, J. K. Seo, O.-H. Kwon, M. H. Ma, C. H. Leung, J. Med. Chem. 2015, 58, 6697-6703; b) C. Lim, H.-W. Rhee, T.-H. Kwon, J. Am. Chem. Soc. 2016, 138, Yang, W. Wang, J. X. Liang, G. Li, K. Vellaisamy, C. Y. 10968-10977. Wong, D. L. Ma, C. H. Leung, J. Med. Chem. 2017, 60, [11] a) D.-L. Ma, D. S.-H. Chan, C.-H. Leung, Acc. Chem. Res. 2597-2603. 2014, 47, 3614-3631; b) S. M. Meier, M. Novak, W. [17] a) J. Maksimoska, D. S. Williams, G. E. Atilla‐Gokcumen, Kandioller, M. A. Jakupec, V. B. Arion, N. Metzler‐Nolte, B. K. S. Smalley, P. J. Carroll, R. D. Webster, P. K. Keppler, C. G. Hartinger, Chem.–Eur. J. 2013, 19, 9297- Filippakopoulos, S. Knapp, M. Herlyn, E. Meggers, Chem.– 9307; c) H. Huang, P. Zhang, Y. Chen, L. Ji, H. Chao, Eur. J. 2008, 14, 4816-4822. b)Fernandes, F. Carta, A. Dalton Trans. 2015, 44, 15602-15610. Innocenti, I. Santos, C. T. Supuran, Angew. Chem. Int. Ed. [12] K. J. Kilpin, P. J. Dyson, Chem. Sci. 2013, 4, 1410-1419. Engl. 2012, 51, 3354-3357. [13] a) D. Rotili, S. Tomassi, M. Conte, R. Benedetti, M. Tortorici, [18] a) S. Miyamoto, P. A. Kollman, Proc. Natl. Acad. Sci. U.S.A. G. Ciossani, S. Valente, B. Marrocco, D. Labella, E. 1993, 90, 8402-8406; b) D. Eisenberg, A. D. McLachlan, Novellino, J. Med. Chem. 2013, 57, 42-55; b) J. Zuo, T. Nature 1986, 319, 199-203. Chen, B. Walters, B. Kuo, B. Walters. U.S. Patent. 2017, No. [19] X. Huang, X. Li, X. Xie, K. Harms, R. Riedel, E. Meggers, 9,572,815. Nat. Commun. 2017, 8, 2245. [14] a) D. Horiuchi, R. Camarda, A. Y. Zhou, C. Yau, O. [20] D. Hu, Y. Liu, Y. T. Lai, K. C. Tong, Y. M. Fung, C. N. Lok, Momcilovic, S. Balakrishnan, A. N. Corella, H. Eyob, K. C. M. Che, Angew. Chem. Int. Ed. Engl. 2016, 55, 1387- Kessenbrock, D. A. Lawson, L. A. Marsh, B. N. Anderton, J. 1391. Rohrberg, R. Kunder, A. V. Bazarov, P. Yaswen, M. T. [21] a) H. Choi, H. Joo, H. Won, K. Min, H. Kim, T. Son, Y. Oh, McManus, H. S. Rugo, Z. Werb, A. Goga, Nat. Med. 2016, J. Lee, G. Kong, J. Natl. Cancer Inst. 2018, 110, 400-410; 22, 1321-1329; b) I. Chu, L. Hengst, J. Slingerland, Nat. b) E. Paolicchi, F. Crea, W. L. Farrar, J. E. Green, R. Rev. Cancer 2008, 8, 253-267. Danesi, Crit. Rev. Oncol. Hematol. 2013, 86, 97-103. [15] a) M. S. Lowry, W. R. Hudson, R. A. Pascal, S. Bernhard, J. Am. Chem. Soc. 2004, 126, 14129-14135; b) D. L. Ma, W. L. tpircsunaM detpeccA Angewandte Chemie International Edition This article is protected by copyright. All rights reserved. 10.1002/anie.201807305 COMMUNICATION Entry for the Table of Contents (Please choose one layout) Layout 1: COMMUNICATION We report herein the discovery of a Guan-Jun Yang, #[a] Wanhe Wang,#[b] rhodium(III)-based complex 1 as an Simon Wing Fai Mok,[c] Chun Wu, [b] inhibitor of KDM5A, an epigenetic Betty Yuen Kwan Law, [c] Xiang-Min target for triple-negative breast Miao, [d] Ke-Jia Wu, [a]Hai-Jing Zhong, cancer. Complex 1 inhibited the [a] Chun-Yuen Wong,[e] Vincent Kam KDM5A-H3K4me3 interaction and Wai Wong,*[c] Dik-Lung Ma,*[b] and suppressed proliferation of TNBC ((Insert T Graphic here)) Chung-Hang Leung*[a] tumors in mice and may be utilized as a novel scaffold for the further Page No. – Page No. development of more potent Selective inhibition of lysine-specific epigenetic agents against cancers, demethylase 5A (KDM5A) using a including TNBC. rhodium(III)-based complex for triple-negative breast cancer therapy tpircsunaM detpeccA Angewandte Chemie International Edition This article is protected by copyright. All rights reserved.