Structure-based design, synthesis and in vitro antiproliferative effects studies of novel dual BRD4/HDAC inhibitors

Article abstract of DOI:10.1016/j.bmcl.2017.07.054

Histone acetylation marks play important roles in controlling gene expressions and are removed by histone deacetylases (HDACs). These marks are read by bromodomain and extra-terminal (BET) proteins, whose targeted inhibitors are under clinical investigation. BET and HDAC inhibitors have been demonstrated to be synergistically killing in Mycinduced murine lymphoma. Herein, we combine the inhibitory activities of BET and HDAC into one molecule through structure-based design method and evaluate its function. The majority of these synthesized compounds showed inhibitory activity against second bromdomains(BRD) of BRD4 and HDAC1. Among them, 16ae presented anti-proliferative effects against human acute myelogenous leukemia (AML) cell lines in vitro, and 16ae is confirmed to reduce the expression of Myc by Western blot analysis. Those results indicated that 16ae is a potent dual BRD4/HDAC inhibitor and deserves further investigation.

Full text of DOI:10.1016/j.bmcl.2017.07.054

Accepted Manuscript
Structure-based design, synthesis and in vitro antiproliferative effects studies of
novel dual BRD4/HDAC inhibitors
Mingfeng Shao, Linhong He, Li Zheng, Lingxiao Huang, Yuanyuan Zhou,
Taijing Wang, Yong Chen, Mingsheng Shen, Fang Wang, Zhuang Yang, Lijuan
Chen
PII:
S0960-894X(17)30761-8
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.07.054
BMCL 25167
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
17 March 2017
9 July 2017
20 July 2017
Please cite this article as: Shao, M., He, L., Zheng, L., Huang, L., Zhou, Y., Wang, T., Chen, Y., Shen, M., Wang,
F., Yang, Z., Chen, L., Structure-based design, synthesis and in vitro antiproliferative effects studies of novel dual
BRD4/HDAC inhibitors, Bioorganic & Medicinal Chemistry Letters (2017), doi: http://dx.doi.org/10.1016/j.bmcl.
2017.07.054
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Structure-based design, synthesis and in vitro antiproliferative
effects studies of novel dual BRD4/HDAC inhibitors
Mingfeng Shaoa,*, Linhong Hea,*, Li Zhenga,*, Lingxiao Huangb, Yuanyuan Zhoua,
Taijing Wanga, Yong Chena, Mingsheng Shena, Fang Wanga, Zhuang Yanga and
Lijuan Chena,#.
a Cancer Center, West China Hospital, Sichuan University and Collaborative
Innovation Center. Chengdu, Sichuan 610041, PR China
b Sichuan Provincial Key Laboratory for Organ Transplantation Translation Medicine,
Hospital of the University of Electronic Science and Technology of China and
Sichuan Provincial People's Hospital. Chengdu, Sichuan 610072, China;
* There authors contributed equally to this work.
#Corresponding Author
Lijiuan Chen
Tel: +86 28 85164063.
Fax: +86 28 85164060.
E-mail address: chenlijuan125@163.com (L-J. Chen).
Bromodomain and extraterminal (BET) and histone deacetylase (HDAC), which
are important epigenetic modulators, are influential targets in drug discovery and
development. The BET family proteins, including Brd2, Brd3, Brd4, and BrdT, utilize
tandem bromodomains (BD-1 and BD-2) to recognize specific acetylated lysine
residues in the N-terminal tails of histone proteins1, and play vital roles in the
regulation of cellular proliferation and apoptosis by regulating the genes transcription2,

3. HDACs are a family of enzymes that catalyze the deacetylation of lysine residues
locating at the N-terminus of different protein substrates4. There are eighteen human
HDACs enzymes divided into four different categories: class I (HDACs 1,2,3,8), class
II (HDACs 4,5,6,7,9,10), class III (Sirtuins 1,2,3,4,5,6,7) and class IV (HDAC 11).
The class III HDACs enzymes are NAD+ dependent while the others belong to zinc
dependent metal enzymes5, 6. Actually, only zinc dependent HDACs, especially class I
and class II isozymes, are closely related to the proliferation, angiogenesis and
differentiation of tumor cells.
BET bromodomain inhibitors (BET inhibitors) can block cancer cell
proliferation and induce apoptosis in a wide range of tumor types7. To date, several
classes of potent and specific BET inhibitors have been reported, such as JQ-18,
I-BET-1519, I-BET72610 and RVX-20811 (Fig. 1A). These prospective BET inhibitors
have provided a set of powerful pharmacological tools for further investigation about
the mechanism of BET proteins in essential cellular processes and the therapeutic
potential of BET inhibition in human diseases12.
As have been approved by FDA for the treatment of haematological and solid
tumors (Fig. 1B), HDACs inhibitors have been becoming a promising therapy for
aberrant epigenetic changes associated with cancer, inflammation and
neurodegenerative diseases13. The hydroxamic acid vorinostat, (also known as SAHA
or Zolinza) was the first-in-class FDA approved HDAC-inhibitor to treat cutaneous
T-cell lymphoma (CTCL) in 200614. Oral panobinostat (also known as LBH589 or

Farydak) combined with bortezomib and dexamethasone was approved by FDA for
the treatment of recurrent multiple myeloma in 201515. Accordingly, the
pharmacophore of HDACs inhibitors (Fig. 1A, Vorinpstat) is usually composed of a
capping group, a zinc-binding group (ZBG) and a proper linker. And hydroxamic acid
is one of the most potent zinc2+ chelating groups among ZBGs which play important
roles in the binding efficiency between HDACs inhibitors and enzyme16.
Figure 1. (A) Structures of representative inhibitors of BET bromodomains. (B)
Structures of FDA approved HDAC inhibitors.
Recent studies have indicated that BET and HDAC inhibitors have similar target
genes and biological effects, thus synergizing to kill in Myc-induced murine
lymphoma17. Combination of BRD4 antagonist and histone deacetylase inhibitor have
been proven effective to against human acute myelogenous leukemia cells18.
Therefore, incorporation of the active group of BET and HDAC inhibitors into one
molecule will be a new way for cancer drug development. Atkinson et al.19 have
reported a dual active BRD/HDAC small molecule probe by fusing a BRD active
tetrahydroquinoline (THQ) core with a hydroxamic acid HDACi motif, while Zhang
et al.16 have also been synthesizing a series of novel 3, 5-dimethylisoxazole

derivatives as BRD4/HDAC dual inhibitors. These dual BET/HDACs inhibitors were
demonstrated to be efficacious and promising, suggesting that the drugs can be
rationally designed for disease applications. In addition, further investigation is
necessary for better effectiveness and specificity of BET/HDACs dual inhibitors.
For the treatment of atherosclerosis-associated cardiovascular diseases ，
Resverlogix Corporation has designed the quinazolone RVX-20820, 21 which has been
entering clinical trials on Alzheimer’s disease recently22. RVX-208 is a derivative of
resveratrol (3, 4’, 5-trihydroxy-transstilbene), which has good biological effects on
the BET proteins. Besides, RVX-208 prefer to binds with the second bromodomain of
BET proteins, presenting better selectivity on BD1 up to 23-fold23. It indicates that
quinazolone derivatives are worth of further investigations.
RVX-OH is another compound closely related to RVX208, but it lacks of
selectivity against BD1 and BD2. In order to maintain the efficacy of RVX-208 on the
BET proteins during the development of BRD4/HDAC dual inhibitors, we have
observed the previously reported X-ray crystal structure of RVX-208 and RVX-OH
within the first and second bromdomains of human BET proteins (Fig. 2). According
to the analysis of the crystal structure of the second bromodomain of BRD2, we find
that RVX-208 bound to the acetyl-lysine binding pocket in a peptide-competitive
manner and the hydroxy-ethylether moiety of RVX-208 is outside of the acetyl-lysine
binding pocket and rarely contacts with the bromodomain surface23. The binding
mode of RVX-OH is similar to RVX-208 in the crystal structure of the second

bromodomain of BRD2 (Fig. 2A). Interestingly, RVX-OH reverses its binding mode
in the crystal structure of the first bromodomain of BRD4, the free hydroxyl group
that is from the phenyl ring system acts as an acetyl-lysine mimetic moiety, thus
forming a hydrogen bond with N140, while RVX-208 has a similar interaction where
the quinaxolinone functions as the acetyl-lysine mimetic moiety (Fig. 2B). Hence, in
order to form a series of dual BRD/HDAC inhibitors, we used various types of linkers
to link the ZBG group with the free hydroxyl group of the phenol ring (Fig. 3). The
fused molecules were expected to retain the essential interactions with both proteins
to exert desired biological functions.
Figure 2. Design of dual inhibitors of BRD4/HDAC (A) Overview of RVX-208 and
RVX-OH binding onto BD2 of BRD2 (PDB entry: 4mr6 and 4mr5). (B) Overview of
RVX-208 and RVX-OH binding onto BD1 of BRD4 (PDB entry: 4mr4 and 4mr3).
The proteins are shown as white surface and blue cartoon. Orange sticks represent
RVX-208 and green sticks represent RVX-OH.

Figure 3. Schematic diagram of construct novel dual BRD4/HDAC inhibitors.
As shown in Schemes 1, all compounds were synthesized via the general route.
Intermediate 10 was synthesized from the commercial available material
3,5-dimethoxyaniline (9) and 5-dimethoxyaniline, and together with the freshly made
HCl (g) to produce the corresponding hydrochloride salt, which was followed
condensed with oxalyl chloride to obtain the cyclization adduct. The 1H-pyrrole-2,
3-dione of compound 10 was hydrolyzed under the basic condition to give the
o-aminobenzoic acid derivative 11. The key intermediate 12 was synthesized by the
amidation of 11 with NH3 (g) in the presence of common peptide condensing agents
EDC and HOBt. The other important intermediates 14aa-14ah were simply made by
refluxing of 4-hydroxy-3, 5-dimethylbenzaldehyde (13a) with various alkyl bromides
in acetonitrile. Then 12 and 14aa-14ah were treated with sodium hydrogen sulfite and
PTSA under the high boiling point solvent DMAc to give compounds 15aa-15ah,
whose ester groups reacted with freshly prepared hydroxylamine in methanol to
produce target compounds 16aa-16ah. The rest compounds 16ba-16bc were prepared
in the same way from the distinct reagent 4-hydroxybenzaldehyde.

Scheme 1. Synthesis of Compounds 16aa-16ah and 16ba-16bc. Reaction conditions:
(a) HCl (g), ether, 0 °C, 2 h, oxalyl chloride, 70 °C, 1.5 h, MeOH, reflux, 1 h, 81%; (b)
NaOH (33% in water), H2O2 (30% in water), 1 h, 70 °C, 33% (c) EDC, HOBT, NMM,
THF, 4 h, NH3 (g), 1 h, rt, 75%; (d) K2CO3, alkyl bromide, acetonitrile, reflux, 24 h,
80–85%; (e) PTSA, NaHSO3, DMAc, 120 °C, 16 h, 75-89%; (f) 50% NH2OH aq,
NaOH, CH2Cl2/MeOH (1:2), rt, 80–85%.
To explore the biological activity, compounds 16aa-16ah were evaluated the
inhibitory effects for human BRD4/BD2 at 5 M and 0.5 M (Table 1). All of them
could inactive BRD4/BD2 over 50% at 5 M, and the majority still inhibited
BRD4/BD2 over 50% ration at 0.5 M. This proposes that our compounds have
maintained the biological activity to BRD4/BD2. We also evaluated compounds
16aa-16ah for their inhibition of HDAC1 at 1 M and 0.1 M (Table 1). In accord
with the results, we found that RVX-OH had no biological activity on HDAC1 as
SAHA had no biological activity on BRD4/BD2. Almost all synthesized compounds,

except for 16ad, inhibited 50% HDAC1 at 1 M, whereas only 16ac, 16af and 16ah
inhibited HDAC1 over 50% rate at 0.1 M. This indicates that the compounds which
use various alkanes as linkers have a better biological activity for HDAC1. Above all,
16ac and 16af showed powerful inhibitory function against HDAC1 and BRD4.
Table 1. Inhibition ratio of compounds 16aa-16ah on BRD4/BD2 and HDAC1.
Avg. Inh%
Avg. Inh%
(HADC1)
Compd
(BRD4/BD2)
5M
0.5M
42±3
41±5
73±3
51±2
62±2
90±1
29±5
59±2
87±1
2±1
1M
0.1M
5±1
90±1
93±2
91±2
93±1
83±3
97±1
88±3
99±1
98±1
10±3
61±3
66±4
96±1
14±4
74±3
92±2
81±3
89±2
5±1
16aa
16ab
20±3
79±3
-2±2
24±4
56±2
36±5
56±3
0±0
16ac
16ad
16ae
16af
16ag
16ah
RVX-OH
SAHA
99±1
96±1
Based on the results of inhibition rate, compounds 16ac and 16ae were evaluated
for their BRD4/BD1, BRD4/BD2 and HDAC1 inhibitory activity in vitro (Table 2).
The results have shown that the activity of compound 16ac and 16ae were decreased
comparing to the positive control, but still remained at a good level against
BRD4/BD2 and HDAC1. Similar to RVX-208, our compounds showed good
selectivity when they effect on BRD4/BD1 and BRD4/BD2, and the selectivity of

16ac and 16ae are more pronounced than RVX-208 due to the larger groups
introduced on the free hydroxyl group of the phenol ring. This indicates that our
designed compounds not only retains the original activity, but also enhances their
selectivity for BRD4/BD2.
Table 2. IC50 values of compounds 16ac and16ae on BRD4/BD1, BRD4/BD2 and
HDAC1.
IC50 (nM)
Compd
BRD4/BD1
>5000
BRD4/BD2
225±32
401±21
67±9
HDAC1
32±10
204±21
-
16ac
16ae
>5000
1985±233
-
RVX-208
SAHA
-
10±3
BET and HDAC inhibitors have been reported synergize to kill in Myc-induced
murine lymphoma, hence the antiproliferative activities against 3 kinds of human
acute myelogenous leukemia (AML) cell lines MV4-11, OCI-AML2 and OCI-AML3
of the representative compounds were evaluated by MTT. In order to investigate
whether the two meta-methyl groups on benzene have influence in the biological
activity of those compounds, the optimal compounds based on the previous result of
Table 1 and 16ba-16bc were selected to further tests. As shown in Table 3, RVX-208
and SAHA were used as positive controls. However, compounds 16ba-16bc did not
significantly inhibit the proliferation of human AML cell lines. It indicated that the

two meta-methyl on benzene really affected biological effects of the compounds. As
expected, compounds 16ac and 16af still exhibited remarkable anti-proliferative
activities. But the compound 16ae which contains the structure of styrene showed
better biological activity than compounds 16ac and 16af. The positive control
RVX-208 showed an inhibitory effect against AML cell lines, but the activity is not
very good. Compound 16ae showed better biological activity than RVX-208 and
SAHA, and it suggested that the dual inhibitors are more effective than the single
target inhibitors against AML cell lines.
Table 3. In vitro cell growth inhibitory effects of compounds on human AML cell
lines
IC50 (M)
Compd
MV4-11
5.00±0.32
1.67±0.21
1.39±0.21
0.56±0.09
0.92±0.12
2.22±0.36
9.88±0.45
4.28±0.31
1,97±0.35
4.48±0.21
0.98±0.27
OCI-AML2
4.53±0.48
1.52±0.21
6.36±0.37
0.38±0.08
0.98±0.17
2.21±0.39
12.25±1.26
13.60±1.33
1.78±0.36
8.31±0.32
0.78±0.32
OCI-AML3
5.53±0.34
1.09±0.18
3.04±0.23
0.43±0.18
0.67±0.10
1.83±0.21
9.31±0.38
11.13±0.23
1.80±0.17
7.17±0.34
0.85±0.29
16aa
16ac
16ad
16ae
16af
16ah
16ba
16bb
16bc
RVX-208
SAHA

In order to verify whether our compounds inhibit Myc in human ALM cell lines,
the most promising compound 16ae was selected to measure its bioactivity in
OCI-AML2 and OCI-AML3 cells by Western blot. As shown in Figure 4,
corresponding with the relative potency in MTT assays, 16ae induced Myc in a
concentration-dependent manner. JQ1 and SAHA, the single-target inhibitors, have
weaker effect to reduce Myc level than 16ae at high concentration. The results
demonstrated that the designed dual inhibitors have achieved the desired goals.
Figure 4. Western blot analysis of Myc in OCI-AML2 and OCI-AML3 cell lines
after 24 h of treatment with compound 16ae at 250, 500, 1000 nM, JQ1 and SAHA at
1000 nM. GAPDH was used as a loading control.
The molecular docking with compound 16ae in BRD4/BD2 and HDAC1 were
performed and the results are shown in Figure 4. From the bonding mode of 16ae with
BRD4/BD2(Fig. 5A), we found that the compound bound to the acetyl-lysine binding
pocket of BRD4/BD2 in a peptide-competitive manner, and one of the nitrogen atoms
of the quinazolinone ring formed a hydrogen bond with the residue Asn-433, which
was similar to the crystal structure of RVX208. The docking mode of 16ae in
complex with HDAC1 is shown in Figure 5B. The hydroxamic acid group of 16ae
entered into the active site by chelating the essential catalytic zinc ion. The xylene
group of compound 16ae occupies the cap region and comfortably locks into the

surface groove, while the quinaxolinone moiety points out of the binding pocket and
makes only a few contacts with the HDAC1 surface. This outcome was consistent
with the results of biological activity experiments.
Figure 5. (A) Compound 16ae docked into BRD4/BD2 (PDB entry: 5u2c), (B)
Compound 16ae docked into HDAC1 (PDB entry: 4kbx). Protein is shown as surface.
Blue stick represents compound 16ae. Zinc ion is shown as black sphere, and the
hydrogen bonds were denoted by yellow dash lines.
In conclusion, utilizing structure-based design approach, we have successfully
generated a series of novel BRD4/HDAC dual inhibitors composed of 3, 4’,
5-trihydroxy-transstilbene with a hydroxamate group which is essential for chelation
with the zinc ion in the active site of HDAC. As expected, most of compounds
exhibited powerful inhibitory function against HDAC1 and BRD4. Subsequently, we
found 16ae exhibit remarkable effects on anti-proliferative activities in vitro. The
Western blot analysis further confirmed the inhibitory effect of compound 16ae on
Myc, and it induced Myc in a concentration-dependent manner in human AML cell

lines. All these observations manifest that 16ae is a potential and promising dual BET
and HDAC inhibitor to kill the Myc-induced murine lymphoma.
Acknowledgment
This work was supported by National Natural Science Foundation of China
(81673289).
Reference
1. Belkina, A.C. and G.V. Denis. Nat Rev Cancer. 2012; 12: 465.
2. Dey, A., et al. Mol Biol Cell. 2009; 20:4899.
3. Maruyama, T., et al. Mol Cell Biol. 2002; 22: 6509.
4. Zhang, Y., et al. J Med Chem. 2011; 54: 2823.
5. Davie, J.R., et al. Curr Opin Genet Dev. 1998; 8:173.
6. Grunstein, M., et al. Nature. 1997; 389: 349.
7. Michael C. Hewitt, Yves Leblanc, et al. Bioorg Med Chem Lett. 2015; 25: 1842.
8. Filippakopoulos P, et al. Nature. 2010; 468: 1067.
9. Dawson MA, et al. Nature. 2011; 478: 529.
10. Wyce A, et al. PLoS One. 2013; 8: 72967.
11. Picaud S, et al. Proc Natl Acad Sci U S A. 2013; 110: 19754.
12. Jonathan Seal, Yann Lamotte, et al. Bioorg Med Chem Lett. 2012; 22: 2968.
13. Thaler, F., et al. Eur J Med Chem. 2016; 108: 53.
14. Marks, P.A., et al. Oncogene. 2007; 26: 1351.
15. Garnock-Jones, K.P., et al. Drugs. 2015; 75: 695.
16. Zhang, Z., et al. Bioorg Med Chem Lett. 2016; 26: 2931.
17. Joydeep Bhadury, Lisa M. Nilsson, et al. PNAS. 2014; 111: 2721
18. Warren Fiskus, Sunil Sharma, et al. Mol Cancer Ther. 2014; 1142
19. Atkinson, S. J.; Soden, P. E., et al. MedChemComm. 2014; 5: 342.
20. Bailey D, et al. J Am Coll Cardiol. 2010; 55: 2580.
21. Nicholls SJ, et al. Cardiovasc Drugs Ther. 2012; 26: 181.
22. McNeill E, et al. Curr Opin Investig Drugs. 2010; 10: 357.
23. Sarah Picauda, Christopher Wellsa, et al. PNAS. 2013; 110: 19754.