The discovery of SKLB-0335 as a paralog-selective EZH2 covalent inhibitor

Article abstract of DOI:10.1039/d0cc04670a

By targeting the unique Cys663 of EZH2, SKLB-0335 displays high paralog-selectivity on EZH2. Biochemical studies show that SKLB-0335 can covalently bind to EZH2 at its S-adenosylmethionine (SAM) pocket and inhibit H3K27Me3. SKLB-0335 could be an effective chemical probe with which to further investigate the specific biological functions of EZH2.

Full text of DOI:10.1039/d0cc04670a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
The discovery of SKLB-0335 as a paralog-selective
EZH2 covalent inhibitor†
Cite this: Chem. Commun., 2021,
57, 3006
Qiangsheng Zhang,‡a Xi Hu,‡a Lu Li,‡b Lidan Zhang,a Guoquan Wan,a Qiang Feng,c
Yongxia Zhu,d Ningyu Wang,*e Zhihao Liu*a and Luoting Yu
*
a
Received 6th July 2020,
Accepted 1st February 2021
DOI: 10.1039/d0cc04670a
rsc.li/chemcomm
By targeting the unique Cys663 of EZH2, SKLB-0335 displays high significant anti-proliferative activity, which means that the
paralog-selectivity on EZH2. Biochemical studies show that SKLB-0335 inhibition of both EZH2 and EZH1 is unnecessary, such as in
can covalently bind to EZH2 at its S-adenosylmethionine (SAM) pocket diffuse large B-cell lymphomas.4 The functions of EZH2 and
and inhibit H3K27Me3. SKLB-0335 could be an effective chemical probe EZH1 mainly differ in two ways: (1) EZH2 and EZH1 are
with which to further investigate the specific biological functions expressed in opposite ways during development and cell differen-
of EZH2.
tiation. EZH2 is dominant in proliferating cells, while EZH1 is
dominant in postmitotic tissues;5 (2) PRC2 that contains EZH1
Polycomb repressive complex 2 (PRC2) is composed of one directly and robustly represses transcription from chromatinized
catalytic subunit, enhancer of zeste homologue (EZH) 1 or templates and compacts chromatin, while PRC2 that contains
EZH2, two core members, suppressor of zeste 12 (SUZ12) and EZH2 does not.6 EZH performs some other biological functions
embryonic ectoderm development (EED) and additional in a PRC2-independent manner. EZH2 trimethylates not only
subunits.1 PRC2 catalyzes the trimethylation of the Lys 27 of H3K27, but also nonhistone proteins such as STAT3 and RORa.7,8
histone H3 (H3K27Me3), which is an epigenetic mark of EZH1 also plays an important role in differentiation, regeneration
transcriptionally repressed and silenced target genes. Dysregu- and immunity.9–11
lation of the PRC2 function correlates with certain malignancies
The development of EZH2 inhibitors has been intensely studied.
and poor prognosis. The mutation, amplification, and/or over- Several EZH2 inhibitors have demonstrated promising antitumor
expression of EZH2 has been implicated in numerous blood and activity in preclinical and clinical trials, such as EPZ-6438 (Tazeme-
solid tumors.2
tostat), CPI-1205, GSK126, ZLD1039, (R)-OR-S1 and GNA002.12–14
EZH2 and EZH1 are highly homologous, sharing 76% overall GNA002, a derivative of gambogenic acid, is the first reported
sequence identity and 96% sequence identity in their catalytic covalent inhibitor of EZH2. However, most of these inhibitors are
su(var)3-9, enhancer-of-zeste and trithorax (SET) domains.3 less selective for EZH1/EZH2 (Fig. 1). As such, they are not effective
In specific tumors, the inhibition of EZH2 alone exerts chemical probes with which to study the specific biological func-
tions of EZH2. The distinct functions of EZH1 and EZH2 suggest
that selectively targeting EZH2 may result in a different efficacy and
tolerability profile compared with dual-EZH1/EZH2 inhibitors. To
enable the specific functions of EZH2 to be investigated, we have
a State Key Laboratory of Biotherapy/Collaborative Innovation Center for
Biotherapy, West China Hospital, West China Medical School,
Sichuan University, Chengdu, Sichuan 610041, P. R. China.
developed and characterized potent, highly paralog-selective EZH2
inhibitors.
E-mail: liuzhihao@scu.edu.cn, yuluot@scu.edu.cn
b Institute of Clinical Trials, West China Hospital, Sichuan University,
Chengdu, Sichuan, 610041, P. R. China
We identified a unique cysteine 663 (Cys663) residue in the
SET region of EZH2 as an effective target for selective inhibition
after aligning the sequences of EZH2 and EZH1 (Fig. 2A and B).15
A search of the International Cancer Genome Consortium
(ICGC) database also revealed that no Cys663 mutation in
EZH2 has yet been reported (Fig. S1, ESI†). Therefore, we hypo-
thesized that targeting Cys663 as a covalent coupling site might
achieve a higher selectivity for EZH2 than EZH1.
c College of Chemistry and Life Science, Chengdu Normal University,
Chengdu 611130, P. R. China
d Department of Clinical Pharmacy, Sichuan Cancer Hospital & Institute,
Sichuan Cancer Center, School of Medicine, University of Electronic Science and
Technology of China, Chengdu, 610041, P. R. China
e School of Life Science and Engineering, Southwest JiaoTong University,
Chengdu, Sichuan, 611756, P. R. China. E-mail: wangny-swjtu@swjtu.edu.cn
† Electronic supplementary information (ESI) available: Experimental proce-
dures, characterization data, and additional spectra. See DOI: 10.1039/
d0cc04670a
By analyzing the co-crystal structures of multiple compounds
complexed with PRC2, we found that indole derivatives were the
‡ These authors contributed equally to this work.
This journal is © The Royal Society of Chemistry 2021
3006 | Chem. Commun., 2021, 57, 3006–3009

View Article Online
Communication
ChemComm
Of the first five compounds generated, compounds b1 and
c2 inhibited the activity of EZH2 at the micromolar level
(Fig. 2C). To preliminarily verify whether these two compounds
exert an inhibitory effect by covalently binding to EZH2, their
reversible analogs, compounds b10 and c20, were synthesized
(Fig. S4, ESI†). Compared to compound b1, the inhibitory
activity of compound b10 on EZH2WT is significantly reduced
(Fig. 2D). The activities of compounds c20 and c2 are compar-
able. However, we could not determine whether this type of
compound binds covalently to EZH2 (Fig. S5, ESI†). We there-
fore focused on pyrazole derivatives in the following work.
To obtain more potent inhibitors, we studied the structure–
activity relationships of pyrazole derivatives (Table S1, ESI†). It
was found that the activity was significantly reduced if the
pyridone ring C4–C5 positions together formed a cyclohexane
or two ethyl groups were introduced at the C4 and C6 positions.
When a larger group, such as a biphenyl group, was introduced
at the R2 position, compounds b12–b15 were slightly less active
than when a benzene substituent was present. Compounds b4,
b6 and b7 with a methyl substitution at the R1 position were
found to be more potent than the unsubstituted compounds
b8, b10 and b11. If a methyl group was introduced on the
methylene at the benzyl substituted at the R2 position, com-
pound b4 was 12-fold more active than compound b1. Notably,
compound (R)-b4 exhibited the most potent activity among the
synthesized compounds. Compound b4 (SKLB-0335) was there-
fore selected for further research because it has comparable
activity to that of (R)-b4 and is easier to synthesize.
Fig. 1 Representative EZH2 inhibitors (n.d. = not determined).
The reversible analog of SKLB-0335 was synthesized as a
control (Fig. 3A). Compared to SKLB-0335, the inhibitory activ-
ity of SKLB-03350 on EZH2WT was significantly reduced. For
incubation times of 15 to 120 min, the inhibitory activity of
SKLB-0335 was improved, while the activities of GSK126 and
SKLB-03350 were not (Fig. 3B). Molecular modeling studies
show that SKLB-0335 binds to the SAM pocket of EZH2
(Fig. 3C). SKLB-0335 displays an extremely slow off-rate (koff
)
of 0.00347 minꢀ1, which results in a residence time half-life
(t1/2) of 199.8 min (Fig. 3D). To further demonstrate that SKLB-0335
binds to the SAM pocket of PRC2 and exerts its inhibitory activity by
forming a covalent bond, we performed competitive experiments
using different concentrations of SAM. When the incubation time
was 15 min, the inhibitory activities of SKLB-0335 and GSK126
decreased by around 10-fold as the SAM concentration increased
from 3 to 30 mM. When the incubation time was extended to
120 min, the inhibitory activity of GSK126 still decreased with an
Fig. 2 (A) The sequence alignment of the SET domains of EZH2 and EZH1.
(B) Cys663 is in the vicinity of the ligand-binding site (PDB code: 5HYN);
carbons are shown in green for S-adenosylhomocysteine. (C) The
chemical structures of the first five compounds and their IC50 values on
EZH2WT
.
aData are expressed as mean ꢁ SEM from at least two indepen-
dent experiments. (D) Inhibition curves of compound b1 and its reversible
analog b10 on EZH2WT. Data are expressed as mean ꢁ SD (n = 2).
most suitable for introducing electrophilic warheads (Fig. S2, ESI†). increase in SAM. However, the inhibitory activity of SKLB-0335
An acrylamide group, which is easy to synthesize and is inert to showed no significant changes. Over time, SKLB-0335 bound
glutathione (GSH),16 was introduced at the 5 position of the indole covalently to EZH2 and its activity was no longer affected by
ring to form a covalent bond with Cys663. We also reduced the the concentration of its substrate, SAM (Fig. 3E). To confirm
indole ring to pyrazole and benzene rings, given that the increased that SKLB-0335 specifically reacts with cysteine, SKLB-0335 was
flexibility of the molecule makes it easier to form a covalent bond incubated with purified EZH2 containing the full SET domain and
through Michael addition (Fig. S3, ESI†). The target compounds then analyzed using mass spectrometry (MS). The MS analyses
were synthesized via a four–six step procedure from commercially reveal that a peak at +433.75 Da could be observed for the probe-
available 5-nitro-1H-indole-3-carboxylic acid (Scheme S1, ESI†), EZH2 adduct from the unmodified EZH2, with no detectable
5-amino-3-methyl-1H-pyrazole-4-carboxylate (Scheme S2, ESI†) and unmodified EZH2, indicating that SKLB-0335 efficiently forms a
methyl 5-bromo-2-methyl-3-nitrobenzoate (Scheme S3, ESI†).
single modified covalent adduct (Fig. 3F). We also tested the
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 3006–3009
| 3007

View Article Online
ChemComm
Communication
Fig. 3 (A) Chemical structures of SKLB-0335 and SKLB-03350. (B) The inhibition curves of EZH2WT for SKLB-0335, SKLB-03350, and GSK126 with 15 and
120 min of incubation. (C) The predicted binding mode of SKLB-0335 (PDB code: 5LS6). Green dashed lines: hydrogen bonds. (D) The koff progress curve
on SKLB-0335 against EZH2 (E = enzyme, I = inhibitor). (E) The SAM competitive inhibition of SKLB-0335, SKLB-03350, and GSK126 with 15 and 120 min
of incubation. (F) Mass spectra of EZH2 alone and EZH2 preincubated with SKLB-0335 (MW:433.51). (G) The inhibition curves for EZH2 and EZH1 of
SKLB-0335. (H) The SKLB-0335 inhibition of G9a, SUV39H1, SETD8, DOT1L, PRMT4, PRMT5, and PRMT7. Data are expressed as mean ꢁ SD (n = 2).
paralog-selectivity of SKLB-0335 for EZH1/EZH2. The results show SKLB-03350 and GSK126 for six days, then fully washed them
that the inhibitory activity of SKLB-0335 on EZH1 is 1500 times with compound-free medium, and assessed the changes in the
lower than that of EZH2 (Fig. 3G). They also show potent and H3K27Me3 levels over a period of 96 h. Recovery of H3K27Me3 in
concentration dependent inhibition of mutant EZH2, including cells treated with SKLB-0335 was significantly slower than that
EZH2Y641F, EZH2Y641N and EZH2A677G (Fig. S6, ESI†). SKLB-0335 those treated with GSK126 (Fig. 4C). We next analyzed the mRNA
does not exhibit significant inhibition of the seven other tested levels of CDX2, CCND2, BMP6 and EOMES, which are regulated
methyltransferases (Fig. 3H).
by EZH1. All of these examined genes were significantly upregu-
The ability of SKLB-0335 to reduce H3K27Me3 was evaluated in lated after GSK126 treatment. However, treatment with SKLB-0335
SU-DHL-6 cells by western blot analysis. Treatment with SKLB- did not affect the mRNA expression of these genes both in
0335 potently abolished the H3K27Me3 marker in a time- SU-DHL-6 cells (Fig. 4D) and WSU-DLCL-2 cells (Fig. S7, ESI†).
dependent (Fig. 4A) and concentration-dependent manner The results showed that SKLB-0335 could selectively inhibit EZH2
(Fig. 4B). Furthermore, we treated the cells with SKLB-0335, without affecting the biological function of EZH1.
This journal is © The Royal Society of Chemistry 2021
3008 | Chem. Commun., 2021, 57, 3006–3009

View Article Online
Communication
ChemComm
Conflicts of interest
There are no conflicts to declare.
Notes and references
1 A. Kuzmichev, K. Nishioka, H. Erdjument-Bromage, P. Tempst and
D. Reinberg, Genes Dev., 2002, 16, 2893–2905.
2 H. P. Mohammad, O. Barbash and C. L. Creasy, Nat. Med., 2019, 25,
403–418.
3 X. Yang, F. Li, K. D. Konze, J. Meslamani, A. Ma, P. J. Brown,
M. M. Zhou, C. H. Arrowsmith, H. U. Kaniskan, M. Vedadi and J. Jin,
J. Med. Chem., 2016, 59, 7617–7633.
4 S. K. Knutson, T. J. Wigle, N. M. Warholic, C. J. Sneeringer,
C. J. Allain, C. R. Klaus, J. D. Sacks, A. Raimondi, C. R. Majer,
J. Song, M. P. Scott, L. Jin, J. J. Smith, E. J. Olhava, R. Chesworth,
M. P. Moyer, V. M. Richon, R. A. Copeland, H. Keilhack,
R. M. Pollock and K. W. Kuntz, Nat. Chem. Biol., 2012, 8, 890–896.
5 E. Ezhkova, W. H. Lien, N. Stokes, H. A. Pasolli, J. M. Silva and
E. Fuchs, Genes Dev., 2011, 25, 485–498.
6 X. Shen, Y. Liu, Y. J. Hsu, Y. Fujiwara, J. Kim, X. Mao, G. C. Yuan and
S. H. Orkin, Mol. Cell, 2008, 32, 491–502.
7 G. Cavalli, Science, 2012, 338, 1430–1431.
Fig. 4 (A) Time course of H3K27Me3 inhibition by SKLB-0335. (B) Dose-
dependent inhibition of H3K27 methylation in SU-DHL-6 cells after 6 days of
SKLB-0335 treatment. (C) Time course of H3K27Me3 inhibition after SKLB-0335,
SKLB-03350, and GSK126 were washed out of SU-DHL-6 cells (N = not washed
out). Results are expressed as mean ꢁ SD from at least three independent
experiments. (D) RT-qPCR analysis of the mRNA levels of EZH1-related genes in
SU-DHL-6 cells following treatment with compounds (10 mM for 5 days). Results
are expressed as mean ꢁ SD from at least three independent experiments.
8 E. Kim, M. Kim, D.-H. Woo, Y. Shin, J. Shin, N. Chang,
Y. T. Oh, H. Kim, J. Rheey and I. Nakano, Cancer Cell, 2013, 23,
839–852.
9 L. T. Vo, M. A. Kinney, X. Liu, Y. Zhang, J. Barragan, P. M. Sousa,
D. K. Jha, A. Han, M. Cesana and Z. Shao, Nature, 2018, 553,
506–510.
10 S. Ai, X. Yu, Y. Li, Y. Peng, C. Li, Y. Yue, G. Tao, C. Li, W. T. Pu and
A. He, Circ. Res., 2017, 121, 106–112.
11 Y. Liu, Q. Zhang, Y. Ding, X. Li, D. Zhao, K. Zhao, Z. Guo and X. Cao,
J. Immunol., 2015, 194, 2838–2846.
12 R. Fioravanti, G. Stazi, C. Zwergel, S. Valente and A. Mai, Chem. Rec.,
2018, 18, 1818–1832.
13 X. J. Song, T. T. Gao, N. Y. Wang, Q. Feng, X. Y. You, T. H. Ye, Q. Lei,
Y. X. Zhu, M. H. Xiong, Y. Xia, F. F. Yang, Y. J. Shi, Y. Q. Wei,
L. D. Zhang and L. T. Yu, Sci. Rep., 2016, 6.
14 X. Wang, W. Cao, J. J. Zhang, M. Yan, Q. Xu, X. B. Wu, L. X. Wan,
Z. Y. Zhang, C. P. Zhang, X. Qin, M. Xiao, D. X. Ye, Y. Y. Liu,
Z. G. Han, S. M. Wang, L. Mao, W. Y. Wei and W. T. Chen, EMBO J.,
2017, 36, 1243–1260.
15 N. Justin, Y. Zhang, C. Tarricone, S. R. Martin, S. Chen,
E. Underwood, V. De Marco, L. F. Haire, P. A. Walker,
D. Reinberg, J. R. Wilson and S. J. Gamblin, Nat. Commun., 2016,
7, 11316.
16 D. G. Hoch, D. Abegg and A. Adibekian, Chem. Commun., 2018, 54,
4501–4512.
In this study, a series of novel EZH2 covalent inhibitors was
designed and synthesized. SKLB-0335 was found to be a potent
and paralog-selective inhibitor of EZH2. This selectivity effec-
tively circumvents concerns about toxicity due to the inhibition
of EZH1 and other methyltransferases. From our findings, we
proposed that SKLB-0335 covalently binds to the SAM pocket of
EZH2. The findings that support this conclusion include MS
results, kinetic studies, and some other biochemical assays. In
summary, SKLB-0335 represents a potent, paralog-selective,
covalent EZH2 inhibitor that could be an effective therapy for
diseases driven by EZH2. SKLB-0335 could also be used as a
chemical probe to further investigate the specific roles of EZH2,
not EZH1/EZH2, in cancer and other related diseases.
This work was supported by the National Natural Science
Foundation of China (81702946, 82073687) and the National
S&T Major Special Project on Major New Drug Innovations of
China (2018ZX09201018).
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 3006–3009 | 3009