Novel Oxindole Sulfonamides and Sulfamides: EPZ031686, the First Orally Bioavailable Small Molecule SMYD3 Inhibitor

Article abstract of DOI:10.1021/acsmedchemlett.5b00272

SMYD3 has been implicated in a range of cancers; however, until now no potent selective small molecule inhibitors have been available for target validation studies. A novel oxindole series of SMYD3 inhibitors was identified through screening of the Epizyme proprietary histone methyltransferase-biased library. Potency optimization afforded two tool compounds, sulfonamide EPZ031686 and sulfamide EPZ030456, with cellular potency at a level sufficient to probe the in vitro biology of SMYD3 inhibition. EPZ031686 shows good bioavailability following oral dosing in mice making it a suitable tool for potential in vivo target validation studies.

Full text of DOI:10.1021/acsmedchemlett.5b00272

Letter
pubs.acs.org/acsmedchemlett
Novel Oxindole Sulfonamides and Sulfamides: EPZ031686, the First
Orally Bioavailable Small Molecule SMYD3 Inhibitor
Lorna H. Mitchell,* P. Ann Boriack-Sjodin, Sherri Smith, Michael Thomenius, Nathalie Rioux,
Michael Munchhof, James E. Mills, Christine Klaus, Jennifer Totman, Thomas V. Riera,
Alejandra Raimondi, Suzanne L. Jacques, Kip West, Megan Foley, Nigel J. Waters, Kevin W. Kuntz,
Tim J. Wigle,† Margaret Porter Scott,‡ Robert A. Copeland, Jesse J. Smith, and Richard Chesworth
Epizyme Inc., Fourth Floor, 400 Technology Square, Cambridge, Massachusetts 02139, United States
S
* Supporting Information
ABSTRACT: SMYD3 has been implicated in a range of
cancers; however, until now no potent selective small molecule
inhibitors have been available for target validation studies. A
novel oxindole series of SMYD3 inhibitors was identified
through screening of the Epizyme proprietary histone
methyltransferase-biased library. Potency optimization af-
forded two tool compounds, sulfonamide EPZ031686 and
sulfamide EPZ030456, with cellular potency at a level sufficient to probe the in vitro biology of SMYD3 inhibition. EPZ031686
shows good bioavailability following oral dosing in mice making it a suitable tool for potential in vivo target validation studies.
KEYWORDS: SMYD3, oxindole, methyltransferase, KMT, oncology, tool compound
leads to enhanced downstream MAPK activation and appears
to be critical for mutant KRAS driven oncogenesis.11
SMYD3’s role in cancer cell line proliferation, its effect on
known oncogenic signal transduction pathways, and the
association of SMYD3 mRNA expression with aggressive
transformed phenotypes make SMYD3 an attractive target for
therapeutic intervention. We report here the first potent and
selective small molecule inhibitors suitable for target validation
studies.
Compound 1 was identified as a micromolar inhibitor of
SMYD3 through screening of Epizyme’s proprietary histone
methyltransferase-biased library. The 1.5 Å resolution crystal
structure of 1 was solved in a ternary complex with SMYD3 and
SAM and revealed the oxindole headgroup made several
hydrophobic and electrostatic interactions with the lysine
pocket of SMYD3 (Figure 1). The oxindole carbonyl forms a
hydrogen bond to the side chain of Ser202 and is within 2.9 Å
of the SAM methyl group. The aromatic ring system forms a π-
stacking interaction with Phe183, while the oxindole nitrogen
makes water-mediated contacts with the protein. The amide of
the linker engages the protein via two hydrogen bonds that
could mimic a protein substrate. The piperidine tail of 1,
positioned in the large, solvent filled peptide binding site,
engaged a water mediated interaction with the side chain
carboxyl of Glu192.
et and Mynd Domain containing 3 (SMYD3) is a lysine
Smethyltransferase (KMT) expressed at high levels in a
number of different cancer histologies and is associated with a
poor clinical prognosis.1−10 While no single mechanism has
emerged to explain this correlation, a number of studies have
implicated SMYD3 in the regulation of gene transcription and
signal transduction pathways critical for cell survival in breast,
liver, prostate, pancreatic, and lung cancer models.4,7−9
In addition, considerable evidence has been reported in the
literature showing that genetic knockdown of SMYD3 leads to
a decrease in proliferation of a variety of cancer cell lines.4,7−9,11
Two studies, employing RNAi-based technologies, have shown
that ablation of SMYD3 in hepatocellular carcinoma cell lines
greatly reduces cell viability and that its pro-oncogenic role is
dependent on its catalytic activity.7,9 Moreover, SMYD3 has
also been shown to be a critical mediator of transformation
induced by a KRAS gain-of-function mutation in both
pancreatic and lung adenocarcinoma mouse models; these
models were likewise dependent on the catalytic activity of
SMYD3.11
The biological function of SMYD3 is still poorly understood.
Early studies of SMYD3 suggested that its primary function is
to methylate histones. Indeed, several reports have indicated
that SMYD3 modifies histone H3 on lysine 4,9,12 but have also
identified a novel modification of histone H4 on lysine 5.7 The
results of these studies have not yet yielded a clear picture of
how SMYD3 might be regulating chromatin, but a recent study
has strongly implicated SMYD3 as a direct regulator of MAPK
pathways in the cytoplasm and not as a regulator of
transcription. MAP3K2 (MEKK2) was shown to be trimethy-
lated at lysine 260 by SMYD3. Modification of this residue
Special Issue: Epigenetics
Received: July 7, 2015
Accepted: August 27, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acsmedchemlett.5b00272
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Decreased biochemical potency was seen when the oxindole
NH was methylated in compound 8, which is not unexpected
as this would interrupt the water mediated H-bond interaction
to protein observed from this position by crystallography.
There is some tolerance in placement of the basic center as the
3-methylpiperidines, 10 and 11, and the cyclohexylamine 12
retain good biochemical potency (Table 2).
Analogues with sulfamide linkers also showed potent
biochemical activity (Table 3). As seen for the sulfonamide
series, cellular potency could be improved by blocking H-bond
donors, compounds 14 and 15 compared to 13, and increasing
lipophilicity, EPZ030456. The 6-chloro- and 6-fluoro-sub-
stituted oxindoles (13 and 16) showed similar biochemical
potency to the compound with no substitution in this position
(18).
The two most cell potent compounds, EPZ030456 and
EPZ031686, were further characterized. Both compounds
showed <30% inhibition against 16 histone methyltransferase
targets at a 10 μM screening concentration.13 Determination of
IC50 against the highly homologous SMYD2 enzyme showed
both compounds to have IC50 > 50 μM.
A crystal structure of EPZ030456 was solved and revealed
unambiguous density for the oxindole headgroup, amide
sulfamide linker, and piperidine ring, while no density was
seen for the terminal benzyl tail (Figure 2). The oxindole
groups of 1 and EPZ030456 are superimposable and make all
the expected interactions previously described. The main chain
of Leu240 and side chains of Met190 and Val368 engage the
compound through van der Waals interactions as it spans the
width of the peptide binding site. Intriguingly, EPZ030456
binds with low nanomolar affinity although the compound
makes few specific interactions with the protein beyond the
headgroup and amide linker. Electrostatic and water mediated
interactions, unseen in the crystal structure, as well as the
rigidity of the compound, are likely important to the potency of
EP030456.
Figure 1. Structure of 1 (cyan) in complex with SMYD3 (green) and
SAM (yellow) (PDB ID 5CCL). Electron density (2Fo−Fc, 1σ) for
the compound is shown. Hydrogen bonds are indicated as dashed
lines.
Since crystallography revealed that both Glu192 and Asp241
were within 3.5 Å of the piperidine nitrogen of 1 initial design
and synthesis of analogues focused on introduction of a basic
center in an attempt to interact with one of these acidic
residues and boost potency. Indeed, the propylamino-
sulfonamide analogue, 2, was over 10-fold more potent in the
biochemical assay than the starting hit, 1. Introduction of a
chlorine group onto the oxindole ring, compound 3, was
tolerated. Constraint of the basic side-chain in the piperidine
derivative 4 resulted in a nominal 2-fold potency improvement
over the propylamine analogue 2 (Table 1).
Further restriction of conformational freedom with the
bridged piperidine core, compound 5, gave a 17-fold potency
improvement over 4. However, despite the single-digit
nanomolar biochemical potency of compound 5 and its
chloro-oxindole derivative 6 the cellular activity was still in
the micromolar range. This weak cell potency could be
attributed to the physicochemical property space of the
compounds, in particular the polarity and high number of
hydrogen bond donors. This hypothesis is supported by the
improvement in cell-to-biochemical shift seen with
EPZ031686, which has one of the H-bond donors blocked
and increased lipophilicity. EPZ031686 is the first SMYD3
inhibitor identified to show double-digit nanomolar cellular
activity.
To understand further the inhibition of SMYD3 by
EPZ030456 and EPZ031686, we determined their mechanism
of action (MOA). Both compounds were essentially non-
competitive with respect to both substrates. EPZ030456 is best
described by mixed-type inhibition with respect to SAM with a
Ki = 4.7
1.8 nM and αKi = 1.1
0.1 nM and is
noncompetitive versus MEKK2 with a Ki = 1.3
0.1 nM
Table 1. Piperidine Oxindole Compounds
a
b
IC50s were determined from at least two independent experiments; see Supporting Information for details. Trimethyl MEKK2 in Cell Western
Assay; details in Supporting Information.
B
DOI: 10.1021/acsmedchemlett.5b00272
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 2. Bridged Piperidine Sulfonamide Analogues
a
b
IC50s were determined from at least two independent experiments; see Supporting Information for details. Trimethyl MEKK2 in Cell Western
Assay; details in Supporting Information.
Table 3. Bridged Piperidine Sulfamide Analogues
SMYD3
SMYD3
b
biochemical
ICW IC
a 50
a
#
R2
R4
R5 cLogP IC50 (nM)
(nM)
13
14
15
Cl
Cl
Cl
Cl
F
H
H
−0.77
1
1100
420
300
48
H
Me −0.34
Me 0.04
2
Me
H
3
EPZ030456
Bn
H
1.38
4
16
17
18
H
−1.23
−0.86
−1.38
5
530
410
850
Me
H
H
H
16
1
H
H
a
IC50s were determined from at least two independent experiments;
Figure 2. (A) Structure of EPZ030456 (cyan) with SMYD3 (green)
and SAM (yellow) (PDB ID 5CCM). Electron density (2Fo−Fc, 1σ)
for the compound is shown. Hydrogen bonds are indicated as dashed
lines. (B) The headgroup of EPZ030456 (cyan) superimposes well
with the oxindole of compound 1 (gray), while the saturated ring
linkers take divergent paths in the active site.
b
see Supporting Information for details. Trimethyl MEKK2 in Cell
Western Assay; details in Supporting Information.
(Supplementary Figure 2, Supplementary Table 1).
EPZ031686 displays noncompetitive inhibition with respect
to SAM and MEKK2 with a Ki = 1.2 0.1 and 1.1 0.1 nM,
respectively (Supplementary Figure 3, Supplementary Table 1).
Observation of compound binding in the lysine pocket in the
crystal structures suggests that these compounds would be
competitive with respect to MEKK2; however, the observed
mechanism is noncompetitive. We hypothesize that MEKK2
derives significant binding affinity from interactions with
C
DOI: 10.1021/acsmedchemlett.5b00272
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
SMYD3 outside of the lysine pocket (i.e., exosite binding) such
that these small molecules cannot displace the protein
substrate; noncompetitive inhibition by active site-directed
small molecule inhibitors is a well established phenomenon for
enzymes acting on macromolecular substrates.14
ization of EPZ031686 showed it to have good bioavailability
following oral dosing in mice making it a suitable tool for
potential in vivo target validation studies.
ASSOCIATED CONTENT
■
The two lead compounds were evaluated for in vitro
metabolic stability and permeability to determine their
suitability for in vivo PK determination. With a mean scaled
clearance of 34 mL/min/kg in mouse liver microsomal
incubations, EPZ030456 was slightly less stable than
EPZ031686 (24 mL/min/kg). EPZ030456 also had a lower
apical-to-basolateral apparent permeability (Papp= 0.34 0.22 ×
10−6 cm/s) in Caco-2 cells than EPZ031686 (Papp= 0.64
0.20 × 10−6 cm/s). Both compounds were subject to active
efflux in the Caco-2 cells, with efflux ratio values of 104 and 41,
respectively. EPZ030456 and EPZ031686 had a free fraction
of 0.32 0.035 and 0.53 0.12 in mouse plasma, respectively.
Thus, overall, EPZ031686 had a more favorable in vitro ADME
profile than EPZ030456.
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acsmedchem-
lett.5b00272.
ADME/PK methods, cell assay conditions, biochemical
assay conditions and MOA data, crystallography methods
and data, synthesis and data for EPZ030456 and
EPZ031686 (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: lmitchell@epizyme.com.
Pharmacokinetic (PK) studies in CD-1 mice were performed
by intravenous (i.v.) bolus and oral gavage (p.o.) admin-
istration. EPZ031686 blood PK parameters derived from
noncompartmental analysis are displayed in Table 4. No
Present Addresses
†Warp Drive Bio, 400 Technology Square, Cambridge,
Massachusetts 02139, United States.
‡Genentech, Inc., 1 DNA Way, South San Francisco, California
94080, United States.
Table 4. Pharmacokinetic Parameters for EPZ031686
Following i.v. and p.o. Bolus Administration to Male CD-1
Mice; Expressed as Mean SD, n = 3
Notes
The authors declare the following competing financial interest:
L.H.M., P.A.B., S.S., M.T., N.R., C.K., J.T., T.V.R., A.R., S.L.J.,
M.F., N.J.W., K.W.K., R.A.B., J.J.S., and R.C. are employees and
stockholders of Epizyme Inc.
parameter
1 mg/kg i.v. 5 mg/kg p.o.
50 mg/kg p.o.
CL (mL/min/kg)
CLr (mL/min/kg)
Vss (L/kg)
27 3.9
5.3 1.6
2.3 0.29
REFERENCES
■
t1/2 (h)
1.7 0.13
2.7 0.98
0.89 0.19
345 89
2.2 0.087
1.3 0.58
(1) Vieira, F. Q.; Costa-Pinheiro, P.; Almeida-Rios, D.; Graca, I.;
Monteiro-Reis, S.; Simoes-Sousa, S.; Carneiro, I.; Sousa, E. J.;
Godinho, M. I.; Baltazar, F.; Henrique, R.; Jeronimo, C. SMYD3
contributes to a more aggressive phenotype of prostate cancer and
targets Cyclin D2 through H4K20me3. Oncotarget 2015, 6, 13644.
(2) Liu, Y.; Liu, H.; Luo, X.; Deng, J.; Pan, Y.; Liang, H.
Overexpression of SMYD3 and matrix metalloproteinase-9 are
associated with poor prognosis of patients with gastric cancer.
Tumor Biol. 2015, 36, 4377.
tmax (h)
Cmax (ng/mL)
AUC0‑last (h·ng/mL)
AUC0‑inf (h·ng/mL)
F (%)
4693 826
21158 2517
21170 2522
69 8.2
603 88
616 90
1281 243
1479 167
48 5.4
significant blood to plasma partitioning was observed for this
compound. Male mice administered a single dose of
EPZ031686 at 1 mg/kg by i.v. bolus showed a moderate
(3) Liu, Y.; Deng, J.; Luo, X.; Pan, Y.; Zhang, L.; Zhang, R.; Liang, H.
Overexpression of SMYD3 was associated with increased STAT3
activation in gastric cancer. Med. Oncol. 2015, 32 (1), 404.
(4) Vieira, F. Q.; Costa-Pinheiro, P.; Ramalho-Carvalho, J.; Pereira,
A.; Menezes, F. D.; Antunes, L.; Carneiro, I.; Oliveira, J.; Henrique, R.;
Jeronimo, C. Deregulated expression of selected histone methylases
and demethylases in prostate carcinoma. Endocr.-Relat. Cancer 2014,
21 (1), 51−61.
(5) Sponziello, M.; Durante, C.; Boichard, A.; Dima, M.; Puppin, C.;
Verrienti, A.; Tamburrano, G.; Di Rocco, G.; Redler, A.; Lacroix, L.;
Bidart, J. M.; Schlumberger, M.; Damante, G.; Russo, D.; Filetti, S.
Epigenetic-related gene expression profile in medullary thyroid cancer
revealed the overexpression of the histone methyltransferases EZH2
and SMYD3 in aggressive tumours. Mol. Cell. Endocrinol. 2014, 392
(1−2), 8−13.
(6) Liu, C.; Wang, C.; Wang, K.; Liu, L.; Shen, Q.; Yan, K.; Sun, X.;
Chen, J.; Liu, J.; Ren, H.; Liu, H.; Xu, Z.; Hu, S.; Xu, D.; Fan, Y.
SMYD3 as an oncogenic driver in prostate cancer by stimulation of
androgen receptor transcription. J. Natl. Cancer Inst 2013, 105 (22),
1719−28.
(7) Van Aller, G. S.; Reynoird, N.; Barbash, O.; Huddleston, M.; Liu,
S.; Zmoos, A. F.; McDevitt, P.; Sinnamon, R.; Le, B.; Mas, G.; Annan,
R.; Sage, J.; Garcia, B. A.; Tummino, P. J.; Gozani, O.; Kruger, R. G.
Smyd3 regulates cancer cell phenotypes and catalyzes histone H4
lysine 5 methylation. Epigenetics 2012, 7 (4), 340−3.
clearance (CL) of 27
agreement with the mouse microsomal data, with a volume
of distribution at steady state (Vss) of 2.3 0.29 L/kg,
3.9 mL/min/kg, in very good
translating to a mean terminal half-life (t1/2) of 1.7 0.13 h.
Approximately 20% of the administered dose was excreted
unchanged in urine after 24 h, equivalent to a renal clearance
(CLr) of 5.3 1.6 mL/min/kg. Following 5 and 50 mg/kg p.o.
dosing, both Cmax and AUC0‑last increased in a slightly higher
than dose-proportional manner, while t1/2 remained unchanged,
suggesting a possible saturation of intestinal efflux. Bioavail-
ability (F) of 48 5.4% and 69 8.2% was observed at 5 and
50 mg/kg, respectively, leading to EPZ031686 unbound blood
concentration remaining above the SMYD3 ICW IC50 value for
more than 12 h after a 50 mg/kg p.o. administration. In
contrast, EPZ030456 solubility was not sufficient to be
formulated for oral dosing at >30 mg/mL using a vehicle
amenable to repeat dosing.
In conclusion, we report here the first potent, selective, small
molecule SMYD3 inhibitors, sulfonamide EPZ031686 and
sulfamide EPZ030456, with cellular potency suitable to probe
the in vitro biology of SMYD3 inhibition. Further character-
D
DOI: 10.1021/acsmedchemlett.5b00272
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
(8) Hamamoto, R.; Silva, F. P.; Tsuge, M.; Nishidate, T.; Katagiri, T.;
Nakamura, Y.; Furukawa, Y. Enhanced SMYD3 expression is essential
for the growth of breast cancer cells. Cancer Sci. 2006, 97 (2), 113−8.
(9) Hamamoto, R.; Furukawa, Y.; Morita, M.; Iimura, Y.; Silva, F. P.;
Li, M.; Yagyu, R.; Nakamura, Y. SMYD3 encodes a histone
methyltransferase involved in the proliferation of cancer cells. Nat.
Cell Biol. 2004, 6 (8), 731−40.
(10) Liu, Y.; Luo, X.; Deng, J.; Pan, Y.; Zhang, L.; Liang, H. SMYD3
overexpression was a risk factor in the biological behavior and
prognosis of gastric carcinoma. Tumor Biol. 2015, 36 (4), 2685−94.
(11) Mazur, P. K.; Reynoird, N.; Khatri, P.; Jansen, P. W.; Wilkinson,
A. W.; Liu, S.; Barbash, O.; Van Aller, G. S.; Huddleston, M.; Dhanak,
D.; Tummino, P. J.; Kruger, R. G.; Garcia, B. A.; Butte, A. J.;
Vermeulen, M.; Sage, J.; Gozani, O. SMYD3 links lysine methylation
of MAP3K2 to Ras-driven cancer. Nature 2014, 510 (7504), 283−7.
(12) Silva, F. P.; Hamamoto, R.; Kunizaki, M.; Tsuge, M.; Nakamura,
Y.; Furukawa, Y. Enhanced methyltransferase activity of SMYD3 by
the cleavage of its N-terminal region in human cancer cells. Oncogene
2008, 27 (19), 2686−92.
(13) The 16 targets screened are DOT1L, EHMT1, EHMT2, EZH1,
EZH2, NSD1, PRDM9, PRMT3, PRMT6, PRMT7, PRMT8, SETD7,
SETDB1, SUV39H1, WHSC1, and WHSC1L1.
(14) Copeland, R. A. Evaluation of Enzyme Inhibitors in Drug
Discovery: A Guide for Medicinal Chemists and Pharmacologists, 2nd ed.;
Wiley: Hoboken, NJ, 2013.
E
DOI: 10.1021/acsmedchemlett.5b00272
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX