The Importance of Being Me: Magic Methyls, Methyltransferase Inhibitors, and the Discovery of Tazemetostat

Article abstract of DOI:10.1021/acs.jmedchem.5b01501

Posttranslational methylation of histones plays a critical role in gene regulation. Misregulation of histone methylation can lead to oncogenic transformation. Enhancer of Zeste homologue 2 (EZH2) methylates histone 3 at lysine 27 (H3K27) and abnormal methylation of this site is found in many cancers. Tazemetostat, an EHZ2 inhibitor in clinical development, has shown activity in both preclinical models of cancer as well as in patients with lymphoma or INI1-deficient solid tumors. Herein we report the structure-activity relationships from identification of an initial hit in a high-throughput screen through selection of tazemetostat for clinical development. The importance of several methyl groups to the potency of the inhibitors is highlighted as well as the importance of balancing pharmacokinetic properties with potency.

Full text of DOI:10.1021/acs.jmedchem.5b01501

Subscriber access provided by GAZI UNIV
Article
The importance of being Me: Magic methyls, methyltransferase
inhibitors and the discovery of tazemetostat.
Kevin W Kuntz, John Emmerson Campbell, Heike Keilhack , Roy M Pollock, Sarah K
Knutson, Margaret Porter Scott, Victoria M Richon, Chris J Sneeringer, Tim J Wigle, Christina
J Allain, Christina R Majer, Mikel P Moyer, Robert A. Copeland, and Richard Chesworth
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01501 • Publication Date (Web): 15 Jan 2016
Downloaded from http://pubs.acs.org on January 18, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The importance of being Me: Magic methyls,
methyltransferase inhibitors and the discovery of
tazemetostat.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kevin W. Kuntz,†,* John E. Campbell, Heike Keilhack,† Roy M. Pollock,‡ Sarah K. Knutson,§
Margaret Porter-Scott,# Victoria M. Richon,† Chris J. Sneeringer,# Tim J. Wigle,‡ Christina J.
Allain, Christina R. Majer,† Mikel P. Moyer, Robert A. Copeland, Richard Chesworth
Epizyme, 400 Technology Square, 4th Floor, Cambridge, MA 02139
EZH2, methyltransferase, lymphoma, EPZꢀ6438, tazemetostat
ABSTRACT: Posttranslational methylation of histones plays a critical role in gene regulation.
Misregulation of histone methylation can lead to oncogenic transformation. Enhancer of Zeste
homolog 2 (EZH2) methylates histone 3 at lysine 27 (H3K27), and abnormal methylation of this
site is found in many cancers. Tazemetostat, an EHZ2 inhibitor in clinical development, has
shown activity in both preclinical models of cancer as well as in patients with lymphoma or
INI1ꢀdeficient solid tumors. Herein we report the structureꢀactivity relationships from
identification of an initial hit in a highꢀthroughput screen through selection of tazemetostat for
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 35
1
2
3
4
5
6
clinical development. The importance of several methyl groups to the potency of the inhibitors
is highlighted as well as the importance of balancing pharmacokinetic properties with potency.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(Me)THYLATION IN BIOLOGY
A major focus of the field of epigenetics is the study of the myriad postꢀtranslational
modifications (PTMs) installed “on top” of the chromatin complex to regulate the expression of
genes and thereby govern physiological processes from tissue differentiation to mood. Lysine
(K) acetylation of histones was the first PTM to be explored in depth, but in addition to
acetylation, numerous PTMs have now been identified, including methylation, ubiquitination,
SUMOylation, crotonylation, butyrylation and propionylation of lysine residues; methylation,
citrullination and ADPꢀribosylation of arginine residues; and phosphorylation and glycosylation
of serine and threonine residues.1 Methylation, however, remains one of the most critical of all
the histone PTMs.
The orchestrated collection of chromatinꢀmodifying activities must be tightly controlled to
enable normal, healthy cell growth and differentiation. In no small part, that control is regulated
by the number of methyl groups (0, 1, 2, or 3) appended to the lysine and arginine (R) residues
composing the polybasic amine tails of the histones – the protein components of chromation
upon which chromosomal DNA is wrapped. The four lysine methylation states when combined
with the number of residues on histone tails can result in billions of permutations for this single
PTM; for example, human H3 alone contains 12 lysine residues, resulting in over 16 million
(412) possible states of methylation.
2
ACS Paragon Plus Environment

Page 3 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Employing differential methylation as a transcriptional code presents a unique challenge for
biological systems because, unlike other PTMs, methylation does not change the overall charge
of the lysine/arginine side chain but instead results in subtle alterations in substrate size and
hydrophobicity. As a result, each of the extensive battery of enzymes that evolved to read, write,
and erase this code, requires a specific pocket architecture to optimize interactions with a
particular methylation state. For example, certain histone methyltransferase (HMT) complexes
position perpendicular aromatic rings to surround fully the charged polymethylated side chain.
Cationꢀπ interactions and hydrophobic and van der Waals contacts drive the formation of the
resulting complex between the methylammonium group and the aromatic rings in the binding
pocket. Methyl lysine reader domains select for monoꢀ or dimethylated lysine over trimethylated
lysine when one of the aromatic walls is replaced by a negatively charged aspartate or glutamate
residue, the carboxylate group of which makes additional favorable hydrogen bonding and
electrostatic contacts with the methylammonium moiety.1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Mutations in the active sites of protein lysine methyltransferases (PKMTs) near the lysine
binding region have been shown to occur in cancers.2 An example is the Y646X (X = F, N, S, C
or H) mutation in Enhancer of Zeste Homolog 2 (EZH2) found in nonꢀHodgkin Lymphoma
(NHL). The Y646X mutations lead to increased H3K27me3 through a change in function of the
mutant enzyme.3 The wildꢀtype enzyme prefers the zeroꢀmethyl lysine as its substrate whereas
the mutant enzyme prefers the dimethyl lysine. The combined effect of the pair of EZH2
enzymes in the heterozygous, mutantꢀbearing cells leads to decreased H3K27me2 and increased
K3K27me3. Similar mutations have been made in other PKMTs,4 where a tyrosine is converted
to a phenylalanine and these mutations also increase the ability of the PKMT to utilize dimethyl
lysine as a substrate. It is proposed that the loss of the tyrosine allows for one of the lysine
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 35
1
2
3
4
5
6
7
8
9
methyls to fit in the space previously occupied by the phenol and the lysine to be deprotonated
by a water channel, thereby completing the methylation reaction.5 The subtle change of Y to F
completely alters the substrate preference, resulting in aberrant histone methylation patterns in
EZH2 mutantꢀbearing lymphomas.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A similar selection preference exists among the protein arginine methyltransferases (PRMTs),
with different enzymes preferentially performing arginine monomethylation and/or selective
symmetric or asymmetric arginine dimethylation. This substrate specificity for arginine
methylation has recently been successfully exploited in the design of selective Protein Arginine
Methyltransferase 5 (PRMT5) peptideꢀcompetitive inhibitors that prevent the symmetric
dimethylation of H3R8 and H4R3.6 In addition to demonstrating the small but significant
differences inherent in the substrate binding pockets of PMTs of high homology, the
effectiveness of the PRMT5 inhibitor in tumor suppression highlights the importance of the
methylation pathways in cell function and their potential targeting for treatment of disease.
(Me)THYLATION IN CHEMISTRY
As any seasoned, synthetic chemist can attest, installing a single methyl group can be a
difficult task. Specific methylation of a carbon center can rely on transitionꢀmetal catalysis with
highly active and unstable reagents. Methylation of basic nitrogen might be considerably more
facile; however, specific placement of a single methyl group requires a multistep protocol ꢀ
including protecting group installation and removal ꢀ and this remains one of the few effective
means for this transformation. Not surprisingly then, a chemist engaged in the study of enzymes
that specifically add methyl groups to the basic residues on histone tails may feel justified
4
ACS Paragon Plus Environment

Page 5 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reverence toward the seeming effortless way with which these enzymes accomplish this
challenging transformation. Moreover, the ability of histone enzyme complexes to recognize the
difference between variably methylated amino acids belies the significance of a single methyl
group for substrateꢀtarget interaction. This significance is not lost on the medicinal chemistry
community responsible for designing inhibitors for these and other small molecule receptor
binding sites.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The literature is full of examples describing what has been termed the “magic methyl” effect
where the addition of a single methyl group in the “right” location on a chemical scaffold can
result in a significant activity shift.7 Given that on its own, a methyl group can only affect
binding via weak London dispersion interactions and a modest impact on lipophilicity, this is a
remarkable feat.8 Often, (but not always) the significant jump in activity that attends specific
addition of a methyl group results from the ability of this small hydrocarbon group to drive, via
steric hindrance, preference for a stereospecific conformation; indeed we highlight an example of
this phenomenon herein.
With respect to drug discovery, it is interesting that four companies identified hits for EZH2
inhibition containing a common 4,6ꢀdimethylpyridone.9ꢀ12 From our extensive experience with
EZH2 inhibitors, we predict that the methyls decorating the pyridone do not impact ligand
conformation, yet are critical to the activity of the chemical inhibitors. Hence, if the available
screening decks contained only pyridones lacking the methyl substitution, it is likely that none of
the chemical starting points identified would have been uncovered in the initial screens and far
fewer EZH2 inhibitors would be known today. That too is the magic of the humble methyl.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 35
1
2
3
4
(Me)THYLTRANSFERASE INHIBITORS
5
6
7
8
9
Differential expression of EZH2, the catalytic component of the polycomb repressive complex
2 (PRC2), has been linked to cancer in numerous studies. EZH2 is responsible for methylating
lysine 27 of histone H3 (H3K27) and overꢀexpression, activating mutations and paradoxically
lossꢀofꢀfunction mutations of EZH2, are all considered oncogenic. H3K27me3 is a
transcriptionally repressive mark, associated with genes that are not being actively transcribed.
Inappropriate levels of H3K27 methylation, however, can lead to the repression of key tumor
suppressor genes, thus allowing cancer cells to grow unchecked. Some cancer cells appear to
become addicted to the suppression of genes that ensues from the overabundance of H3K27me3,
leading to tumors resistant to the natural growth checkpoints and able to avoid apoptosis signals.
Thus inhibition of EZH2 should cause proliferation defects in cells which have become
dependent on excess H3K27me3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
HN
N
N
H
1
Figure 1. Validated screening hit, 1
To find inhibitors of EZH2, highꢀthroughput diversity screens were performed. The four
component PRC2 complex containing EZH2, EED, SUZ12 and RbAp48 was used with chicken
oligonucleosome as the substrate.9 The reaction was run under balanced conditions (at the Km
6
ACS Paragon Plus Environment

Page 7 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
for both SAM and nucleosome) to allow for discovery of inhibitors of all modes of inhibition
such as competitive with SꢀAdenosylmethionine (SAM) or with nucleosome.13 The diversity
screens provided many hits, but very little tractable chemical matter. A rigorous hit validation
procedure was employed to determine which compounds inhibited the enzyme in a useful
manner (i.e. the inhibition was driven by direct binding to the enzyme). Many of the hits proved
to be PAINS compounds,14 redox active,15 or had issues with solubility or micelle formation as
determined by differential lightꢀscattering.16 From one of the diversity screens, compound 1
(Fig. 1), containing a 4,6ꢀdimethyl pyridone, was identified as an IC50 = 3.4 + 0.9 µM inhibitor
of EZH2 in a biochemical assay. Since only a portion of the diversity library had been screened,
close analogs of this hit, which were not part of the initial screen, were selected from the library
and tested for activity against EZH2. From this hit expansion, 5 was identified as a subꢀµM
inhibitor of EZH2 (Table 1). This compound successfully passed a battery of bad actor tests,
and had retained activity from an independent synthesis. Subsequent mechanism of inhibition
studies showed that the inhibitor was SAMꢀcompetitive and nucleosome nonꢀcompetitive. This
gave us confidence that the inhibition was through a direct interaction with the enzyme and
efforts to optimize the activity of this template were initiated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 35
1
2
Figure 2. Structures of EZH2 inhibitors, 2 (EPZ005687),9 3 (EPZ006088),17 and 4 (EPZꢀ6438,
3
4
5
6
tazemetostat)17
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Identification of vector, R, to improve physicochemical properties
EZH2 IC50
ID#
5
R
cLogD
(µM)
0.5 ± 0.2
2.1
6
0.2 ± 0.2
0.1
7
0.4 ± 0.2
0.3 ± 0.2
2.1
8
-0.1
N
9
1.6 ± 0.4
2.3 ± 0.3
-0.7
1.9
NH
10
11
12
13
2.4 ± 0.1
0.9 ± 0.3
3.1 ± 1.0
0.9
1.5
Br
H
0.5
The first problem that needed to be overcome was the limited solubility (<10 µM at pH 7) of
the hit compound and its poor oral bioavailability (0.5 %F), likely caused by the low solubility.
Polarity was added to all of the available vectors of the molecule to improve aqueous solubility.
We quickly discovered that the 6ꢀposition of the 7ꢀazaindazole tolerated a variety of polar
functionality. As shown in Table 1, basic amines with (6-7) and without aryl spacers (8-9)
8
ACS Paragon Plus Environment

Page 9 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
resulted in improved solubility without substantial loss in potency. Even smaller nonꢀbasic
groups (10-13) were well tolerated and more soluble than 1, but the improved solubility of 7 was
significant, as it allowed the demonstration of cellular activity for the first time (data not shown)
and afforded improved oral bioavailability (41 %F). From hitꢀtoꢀlead through lead optimization
hundreds of variants along this vector were synthesized, but ultimately the morpholine separated
from benzene by a methyl spacer in 7, discovered early in the program and present in the final
drug candidate, afforded the best balance of properties. Compound 7 represented an excellent
lead compound, possessing cellular activity, oral bioavailability and moderate clearance.
However, because removal of the solubilizing substituent resulted only in a modest reduction in
biochemical potency, we elected to optimize further structural activity relationships (SAR) with
a halogen handle or a hydrogen in the 6ꢀposition for ease of synthesis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. 5,6ꢀbicyclic core optimization
EZH2 IC50
ID#
R
cLogD
(µM)
3.1 ± 1.0
0.5
13
14
15
1.1 ± 0.3
3.1 ± 2.6
1.4
2.2
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 35
1
2
3
4
5
6
7
8
9
>50
>50
1.4
1.4
16
17
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22 ± 6
1.4
1.4
18
19
>50
Next, derivatives were synthesized to identify the preferred core composition. A variety of
heteroaryl systems were investigated (Table 2) and the activity of the indazole and indole was
improved relative to the pyrazolopyridine, but activity was ablated when the amide attachment
point was moved or when nitrogen was incorporated adjacent to the amide. Because the
benzimidazole (16) and triazole (17) can likely accept an internal hydrogen bond from the amide
we postulated that a planar amide may disfavor the active conformation. The lower activity of
18 and 19 may also be due to the preference of the amide and aryl ring to be planar in the
absence of an ortho substituent (see compounds 28 & 29 below) and not due to the position of
the 5ꢀmembered ring. However, at the time the 4ꢀsubstitutied indazole orientation appeared
favored and was selected for further exploration of the SAR of other regions of the scaffold.
10
ACS Paragon Plus Environment

Page 11 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Branched substitution is key to 5,6ꢀbicyclic core potency. R1 = 2,2,6,6ꢀtetramethylꢀ
1,2,3,6ꢀtetrahydroꢀpyridine attached at the 4ꢀposition, R2 = 3,4,6ꢀtrimethylpyridinꢀ2(1H)ꢀone
attached to the amide at the 3ꢀmethyl position
In an earlier series of pyrazolopyridines (Scheme 1), we demonstrated that branched
substitution was preferred for the N1ꢀsubstitution. Extending the methyl in 20 to the ethyl affords
a 10ꢀfold improvement in potency. Replacing the methyl with the more bulky isopropyl resulted
in 8, 0.3 ± 0.2 µM, which represents a 2.7 kcal advantage over the methyl alone. Therefore,
focus was given to branched substituents for the optimization of the indazole N1ꢀposition; these
studies demonstrated that hydrophobic substituents and larger nonꢀionizable polar groups were
preferred (Table 3). Charged groups such as piperidine or nonꢀbulky polar groups such as
tetrahydropyran appeared less favored at N1.
At this point in the SAR exploration, numerous diverse indazoles (including 2, Fig. 2, Table 4)
were synthesized without significant activity improvement and the marginal increases that were
observed, were accompanied by decreased bioavailability or increased clearance of the resultant
inhibitors. We decided to take a step back from substitution and reꢀevaluate the core. The 5ꢀ
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 35
1
2
3
4
5
6
7
8
9
membered ring of the 5,6 bicycle was “disconnected” to give a substituted aniline which would
allow access to different vectors to improve activity. As shown in Table 5, R1 substitution was
critical to achieve onꢀscale activity in the biochemical assay. For compound 28, lacking an R1
methyl, activity is significantly decreased from corresponding bicyclic compound 22. This
activity is completely restored, however, upon the addition of a methyl group (i.e. 29).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3. Branched substitution optimization of indazole core
ID#
22
R
EZH2 IC50 (µM)
0.08 ± 0.04
cLogD
2.7
0.2 ± 0.03
2.3
2.4
23
24
0.05 ± 0.03
0.05 ± 0.02
0.7
25
2.2 ± 1.1
0.7 ± 0.2
ꢀ1.6
1.3
26
27
12
ACS Paragon Plus Environment

Page 13 of 35
Journal of Medicinal Chemistry
1
2
3
4
Table 4. Cellular methyl mark EC50 (H3K27me3 ELISA) and Mouse PK data (IV dose 2
5
6
mg/kg; PO dose 10 mg/kg)
7
8
9
ELISA
Cl (mL/
min/kg)
AUC
Vss
ID#
%F
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EC50 (µM)
(hr*ng/mL)
(L/kg)
ND
24
35
11
700
950
1.3
1.5
0.5
41
47
5*
7
10.5 ± 2
2.9 ± 1
2900
0.49
2**
*1 mg/kg IV, ** cLogD = 3.1
In the indole and indazole scaffolds, modelling suggests that the fused heterocycle forces the
amide bearing the 3,4,6ꢀtrimethyl pyridone out of plane from the core, presenting an optimal
mode for binding. In 29, the fused heterocycle is efficiently replaced by a single methyl group.
In other recent examples, the benzamide has been fused into a lactam ring with the adjacent
gearing methyl or chloro group to achieve the same effective shape.18 It is worth noting,
however, that only 6ꢀ and 7ꢀmembered lactams have been reported.
We next revisited the alternative amide substitution on the bicyclic core from Table 2. As a
further demonstration of the power of a magic methyl, we showed that activity can be increased
by installation of a methyl group adjacent to the amide, 38 to 39 (Scheme 2), suggesting that
forcing the amide out of plane is a significant driver of potency. The observed 10ꢀfold shift here
does not, however, account fully for the >100ꢀfold benefit observed between 28 and 29 (Scheme
2).
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 35
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. A steric methyl restores activity even in subꢀoptimal bicyclic cores. R = 3,4,6ꢀ
trimethylpyridinꢀ2(1H)ꢀone attached to the amide at the 3ꢀmethyl position
Looking further at the data in Table 5, there is evidence for a distinct preference for
disubstituted anilines. For compound 30, when R1 = Me, but the aniline is monosubstituted (R2
= H), the result is an IC50 of 630 nM, a >40ꢀfold shift from 31, where R2 = Me. Restoring the
small aliphatic group at R2 and removing the tetrahydropyran (THP) moiety at R3 has an even
larger negative effect as demonstrated by analogs 32 and 33. A survey of the Protein Data Bank
and the Cambridge Structural Database shows a preference for secondary amides ortho to a
methyl group to have a torsion angle of 60ꢀ140o or 220ꢀ300o (See Figure S2). This is confirmed
by calculations which show a planar configuration is disfavored for an amide ortho to a methyl
group (See Figure S1). The stericꢀdirecting effect of the methyl is not limited to the amide
moiety, but also has a profound effect on the adjacent aniline substituent. Calculations show a
preference for the aniline to be planar with respect to the aryl ring when R1 = H, but the aniline
prefers to be twisted out of plane when R1 = Me (See Figure S3). Furthermore, when R2 is
increased in size to the ethyl group, the THPꢀsubstituted variant, 34, has an IC50 of 15 nM against
14
ACS Paragon Plus Environment

Page 15 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EZH2 when R4 = Cl. Now removal of the R1 methyl in 35, results in a loss of >600ꢀfold
potency, which accounts for the combined impact of the orthogonally directed amide and the
favored disubstituted aniline orientation. This optimized substitution pattern led to a significant
potency breakthrough allowing us, for the first time, to drive potency to the threshold of singleꢀ
digit nanomolar levels.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. Optimization of substitution on the benzene core
EZH2 IC50
ID#
R1
R2
R3
R4
cLogD
(µM)
5.7 ± 1.6
H
Me
Me
H
Br
Br
Br
Br
3.4
3.9
1.9
2.5
28
29
30
31
0.03 ± 0.02
0.5 ± 0.2
Me
Me
Me
0.03 ± 0.02
Me
2.7 ± 1.2
1.7 ± 0.9
Me
Me
Me
Me
H
Cl
Cl
1.7
2.4
32
33
Me
0.01 ± 0.01
14 ± 2
Me
H
Et
Et
Et
Cl
Cl
Br
2.7
2.2
2.9
34
35
36
0.01 ± 0.01
Me
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 35
1
2
3
4
5
0.03 ± 0.02
Me
Me
Br
ꢀ0.3
37
NH
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
At first glance, the aniline SAR appears to compliment that of the indazole where branched
substitution is favored, ie. 33 has >50ꢀfold less activity than that observed for cyclopentane
substituted 29. A variety of polar and even charged cyclic systems, 37, are tolerated, but the
potency of the THPꢀsubstituted aniline suggests that the vector probed by the aniline is distinct
from the limited planar orientation enforced by the bicyclic indazole core (Scheme 3). While the
difference between the cyclopentyl substituted analogs is minimal, the THP, which is poorly
tolerated in the bicyclic system, is equipotent to the cyclopentane on the benzene core. This
small change, enabled again by the steric methyl effect, proved to be critical, because the THP
allows for lower logD and lower clearance in the ultimate development candidate. However, as
can be seen in Table 7, while the clearance was good, the oral bioavailability of compound 34
remains poor.
To optimize the bioavailability of compound 34, we considered whether the highly polar
benzamidomethyl pyridone might be a liability adversely affecting absorption. Initially we
focused on removing or mitigating the Hꢀbond donors that were hypothesized to be limiting
absorption of the class (Fig. 3). First, substitution of the benzamide NH led to a 10ꢀfold decrease
in potency, presumably due to the loss of the Eꢀamide conformational preference. Furthermore,
addition of a methyl group to the benzylic carbon (41) reduced the potency of the compound by
100ꢀfold and further limited our options for lowering the polarity of the amide NH.
16
ACS Paragon Plus Environment

Page 17 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The importance of the pyridone amide was also explored by methylation of the nitrogen or the
oxygen to capture either the pyridone or hydroxy pyridine tautomer. Nitrogen substitution, 42,
resulted in a comparable decrease in potency to that observed for the benzamide methylation.
The methoxy pyridine 43, however, was much less potent with no inhibition of EZH2 measured
up to 10 µM. This implies that the tautomer which confers activity conserves the carbonyl
oxygen and makes a highly favorable interaction with the enzyme. Attempts to completely
replace the pyridone confirmed this binding speculation (data not shown) and will be the subject
of a future communication.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Pairwise analysis of switch from bicyclic to monocyclic core highlights divergent
SAR on aniline nitrogen substitution. R = 3,4,6ꢀtrimethylpyridinꢀ2(1H)ꢀone attached to the
amide at the 3ꢀmethyl position
Exploring substitutions on the pyridone itself proved that the 4,6ꢀdisubstitution was equally
critical to activity. As we alluded to in the introduction, removal of either methyl group (45, 46)
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 35
1
2
3
4
5
6
7
8
9
resulted in a 10ꢀfold decrease in activity (Table 6) and removing both methyls in 44, resulted in
a >200ꢀfold decrease. While growing the size of either the 4ꢀ or 6ꢀgroup to ethyl (49, 51)
maintained activity, changing to 4ꢀCF3 (48) led to a slight decrease in potency, but both
substitutions have a net negative effect on logD. Conversely, increasing the polarity at the 4ꢀ
position by adding an alcohol, 47, or an amine, 50, decreased the activity by 10 and 100ꢀfold,
respectively. Addition of a third methyl group in the 5ꢀposition was also shown to be
detrimental, resulting in a 10ꢀfold decrease in activity and larger groups in this position led to
further loss of activity (data not shown). While in combination with other cores, there has been a
reported benefit by replacing the 4ꢀmethyl with a 4ꢀmethoxy,10 here we saw a decrease in oral
bioavailability and increased clearance with this substitution (data not shown). As a result, we
elected to proceed with the 4,6ꢀdimethyl pyridone intact, adjusting the other vectors to optimize
in vivo properties.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Effects of methylation of the benzamide pyridone
Table 6. Pyridone substitution plays a significant role in the EZH2 activity of small molecule
inhibitors
18
ACS Paragon Plus Environment

ID# R1
R2
EZH2 IC50 (µ
0.01 ± 0.01
3.3 ± 1.2
M)
cLogD
2.7
Page 19 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Me
H
Me
H
34
44
45
46
47
48
49
50
51
2.2
0.1 ± 0.04
0.2 ± 0.01
0.2 ± 0.03
0.03 ± 0.01
0.01 ± 0.01
0.9 ± 0.4
H
Me
2.5
Me
Me
Me
Me
Me
Et
H
2.4
CH2OH
CF3
1.4
3.0
Et
3.1
CH2NMe2
Me
0.9
0.02 ± 0.01
3.2
Fortunately, the increase in potency found in the monocyclic series did not come at the
expense of the in vivo profile of the series. In 2010, bicyclic compound 7 and tool compound 29
were shown to have good oral bioavailability, making them good lead candidates, but with subꢀ
optimal functional activity (IC50 = 3 µM in cell methylation assay). Early monocyclic compound
29 showed a dramatic improvement in this activity; however compounds lacking large
substituents in the 5ꢀposition of the phenyl ring tended to have high in vitro and in vivo
clearance. Installing the benzyl morpholine on compound 29 led to 3 (Fig. 2), which was able to
achieve significantly higher exposure as a result of the improved physicochemical properties.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 35
1
2
3
4
5
6
7
8
9
Similarly, compound 31, incorporating a THP group to lower the logD of 29, produced improved
oral bioavailability and, despite being limited by high clearance, had the additional advantage of
improved cellular response. Ultimately, thoughtful design utilizing our SARꢀlearnings, led to the
combination of the benzyl morpholine tail with the THPꢀdecorated aniline resulting in the
development candidate 4, now known as tazemetostat, which is appreciably more potent, has low
clearance and has good bioavailability.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The desired combination of potency and mouse PK was obtained with tazemetostat, thus
enabling in vivo efficacy studies. We were able to show reduction of H3K27me3 in both tumor
and surrogate tissue in mouse xenograft experiments.19 We hypothesized that integrated time
over a threshold concentration would be an important parameter for sustained efficacy. We
therefore compared once daily (QD), twice daily (BID) and thrice daily (TID) dosing schedules
in a mouse xenograft model. BID and TID schedules of the same total daily dose gave the same
tumor growth delay, and both trended superior to QD.19 This result supports the hypothesis of
time above a threshold being an important parameter for efficacy. For a combination of
simplicity (vs TID) and efficacy (vs QD) the subsequent mouse experiments were performed
using BID schedules.
Table 7. Cellular methyl mark EC50 (H3K27me3 ELISA) and Mouse PK data (IV dose 2
mg/kg; PO dose 10 mg/kg)
Cl
AUC
ID#
29
ELISA EC50
(
µM)
Vss (L/kg) %F
cLogD
(mL/min/kg)
(hr*ng/mL)
0.5*
18
46
0.97
2
3.9
20
ACS Paragon Plus Environment

Page 21 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
0.3 ± 0.1
0.7 ± 0.2
49
16
680
0.45
0.66
1.6
24
2.7
4.2
2.5
34
3
2500
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.3 ± 0.1
0.2 ± 0.1
90
13
350
2.7
1.0
18
55
31
4
2500
3.2
*no replicate
CONCLUSIONS
In a 2012 J Med Chem paper, Leung et al. stated that it is rare to find a factor of ten
improvement in activity for addition of a methyl group. In fact, in their survey of BMCL & J
Med Chem from 2006ꢀ2011 they found 10ꢀfold improvements in only 8% of the cases and 100ꢀ
fold or better in <0.5% of the cases. In the current chemical series, we have found three
locations where addition of a methyl group leads to a 10ꢀ to 20ꢀfold improvement in activity and
one location where addition of a methyl group adds >100ꢀfold improvement in activity. When
combined, the four methyl groups add up to a stunning >100,000ꢀfold improvement. This also
illustrates the difficulty in finding hits for enzymes like EZH2 from diversity screening.
Compounds with subtle changes, such as the absence of one methyl in the initial hit may have
been missed in the highꢀthroughput screen. When considering diversity collections of
compounds, it is important to keep in mind that methyl groups, especially when substituted on
aryl rings, can have profound effects on activity.
Biological systems have evolved to utilize the methyl group as one of the key drivers of
transcription control, generating gene expression instructions that can be passed through mitotic
and meiotic replication. Proteins are able to recognize and differently bind to methylated amines
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 35
1
2
3
4
5
6
7
8
9
and guanidines to control cell fate. This process is a major mechanism to orchestrate the
exceedingly complex path of organism growth and development, homeostasis, and aging. Hence
it is not surprising that mechanisms involved in methylation and demethylation reactions, as well
as methylmark recognition, are being hijacked in disease states, such as cancer. This makes
proteins involved in these processes, like EZH2, attractive drug targets.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In summary, we have described the optimization of a series of EZH2 inhibitors resulting in 4
(tazemetostat), a potent and selective small molecule inhibitor of EZH2, from hit to clinical
candidate for the use in patients with genetically defined cancers. Through the development of
the SAR of the series we realized that the bicyclic core of our lead candidate was limited by the
diversity of tolerated substituents. This factor when combined with the few modifiable vectors
on the critical pyridone component, severely restrained our ability to balance the physiochemical
properties necessary to develop a viable oral drug with the potency required against EZH2.
Disconnecting the 5ꢀmembered ring led not only to an increase in potency, but also allowed for
an additional vector where polarity could be incorporated. Furthermore, the addition of the
benzylꢀmorpholine group along the solubility vector consistently decreased clearance and
improved bioavailability leading to a candidate that has demonstrated robust activity in
preclinical models of disease for nonꢀHodgkin lymphoma (NHL),19,20 and malignant rhabdoid
tumors (MRT),21 tumors. More importantly, tazemetostat has demonstrated clinical tumor
regressions in relapsed/refractory human patients with FL, NHL and MRT.22 At the time of
writing, tazemetostat has entered a Phase II expansion in patients with relapsed/refractory diffuse
large Bꢀcell lymphoma (DLBCL) and follicular lymphoma (FL) and an IND has been approved
for a Phase II trial in INI1ꢀdeficient and DLBCL patients in the US.
22
ACS Paragon Plus Environment

Page 23 of 35
Journal of Medicinal Chemistry
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
9
Chemical Synthesis: All compounds were prepared using standard organic chemistry
techniques. Analytical NMR data were collected on a 400 MHz,Varian NMR in DMSOꢀd6
unless otherwise specified. LCMS conditions: YMC ODS A, C18, 50 mm x 4.6 mm, 3 µM;
Column Temp., 30 oC; Mobile phase, A, 0.05% TFA in water, B, 0.05% TFA in acetonitrile;
Flow rate, 1.2 mL/min; Gradient, 20% B to 95% B in 3 min, Hold for 0.5 min,3.51ꢀ4.5 min 20%
B. All test compounds reported in this publication were synthesized to a ≥95% purity level
unless otherwise stated, purity level as assessed by the following methodology. HPLC
Conditions: YMC ODSꢀA column 150 mm x 4.6 mm x 5 ꢁ; Mobile phase, A, 0.05% TFA in
water/ B, 0.05% TFA in acetonitrile; Inj. Vol., 10 ꢁL, Col. Temp., 30 oC; Flow rate: 1.4
mL/min.; Gradient: 5% B to 95% B in 8 min, Hold for 1.5 min, 9.51ꢀ12 min 5% B.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Standard aliphatic coupling conditions: To a stirred solution of ethyl 6ꢀbromoꢀ1Hꢀ
pyrazolo[3,4ꢀb]pyridineꢀ4ꢀcarboxylate (20 g, 74.34 mmol) in acetonitrile (200 mL), K2CO3
(15.39 g, 111.52 mmol) and 2ꢀbromopropane (18.13 g, 148.64 mmol) was added. The reaction
mixture was refluxed for 8 h. On completion, acetonitrile was removed under reduced pressure
and water was added. Extraction was carried out using ethyl acetate; the combined organic layers
were washed with water, brine and dried over anhydrous Na2SO4. Solvent was removed under
reduced pressure and residue was purified by silica gel column chromatography to obtain the
desired compound (13.5 g, 58%).
Standard reductive amination conditions: To a stirred solution of methyl 3ꢀaminoꢀ5ꢀ
bromobenzoate (2 g, 8.73 mmol) and cyclopentanone (2.2 g, 26.1 mmol) in methanol (20 mL),
acetic acid (1.04 g, 17.4 mmol) was added and reaction stirred at room temperature for 3 h. Then
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 35
1
2
3
4
5
6
7
8
9
sodium cyanoborohydride (1.37 g, 21.83 mmol) was added and the reaction was stirred
overnight. On completion, the solvent was removed under reduced pressure and the crude
material was purified by column chromatography to afford the desired compound (1.6 g, 61%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Standard Suzuki-coupling conditions: To a stirred solution of 5ꢀbromoꢀNꢀ((4,6ꢀdimethylꢀ2ꢀ
oxoꢀ1,2ꢀdihydropyridinꢀ3ꢀyl)methyl)ꢀ3ꢀ(ethyl(tetrahydroꢀ2Hꢀpyranꢀ4ꢀyl)amino)ꢀ2ꢀ
methylbenzamide (14 g, 29.5 mmol) in dioxane: water mixture (70 mL: 14 mL) was added 4ꢀ(4ꢀ
(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)benzyl)morpholine (13.4 g, 44.2 mmol) followed
by addition of Na2CO3 (11.2 g, 106.1 mmol). The solution was purged with argon for 15
minutes and then Pd(PPh3)4 (3.40 g, 2.94 mmol) was added and the solution was again purged
with argon for a further 10 min. The reaction mixture was heated at 100 °C for 4 h. After
completion (monitored by TLC), the reaction mixture was diluted with water and extracted with
10% MeOH/DCM. The combined organic layers were dried over anhydrous Na2SO4, filtered and
concentrated inꢀvacuo. The crude compound was purified by column chromatography (100ꢀ200
mesh silica gel) eluting with methanol: DCM to afford the desired compound (12 g, 71 %).
Standard ester hydrolysis and coupling conditions: Aqueous NaOH (5 equiv., 10%) was
added to a solution of ethyl 6ꢀ(2,3ꢀdihydrobenzo[b][1,4]dioxinꢀ6ꢀyl)ꢀ1ꢀisopropylꢀ1Hꢀ
pyrazolo[3,4ꢀb]pyridineꢀ4ꢀcarboxylate (1 equiv.) in EtOH (10 times by volume) and stirred at 60
oC for 3ꢀ4 h. After completion of the reaction, ethanol was removed under reduced pressure and
the resulting mixture was acidified using 10% citric acid solution. Extraction was carried out
using EtOAc (15 mL x 3) the combined organic layers were washed with water (30 mL x 2)
brine (30 mL) dried over anhydrous Na2SO4; filtered and concentrated under reduced pressure.
The residue was then dissolved in DMSO (10 times by volume) and 3ꢀ(aminomethyl)ꢀ4,6ꢀ
dimethylpyridinꢀ2ꢀol (2 equiv.) was added. The reaction mixture was stirred at room temperature
24
ACS Paragon Plus Environment

Page 25 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
for 15 min before PyBOP (1.5 equiv.) was added and stirring was continued for 5 to 16 h. After
completion of the reaction, saturated NaHCO3 solution was added and extraction was carried out
using EtOAc. The combined organic layers were washed with water (3x), brine, dried over
anhydrous Na2SO4, filtered and concentrated under reduced pressure. The residue was purified
by silica gel column chromatography followed by ether wash to provide pure final target
molecule.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of compound 29: 5-bromo-3-(cyclopentyl(methyl)amino)-N-((4,6-dimethyl-2-
oxo-1,2-dihydropyridin-3-yl)methyl)-2-methylbenzamide
Synthesis of methyl 5-bromo-3-(cyclopentylamino)-2-methylbenzoate: Standard reductive
amination conditions. (90%).
Synthesis of methyl 5-bromo-3-(cyclopentyl(methyl)amino)-2-methylbenzoate: Standard
aliphatic coupling conditions (0.39 g).
Synthesis of 5-bromo-3-(cyclopentyl(methyl)amino)-N-((4,6-dimethyl-2-oxo-1,2-
dihydropyridin-3-yl)methyl)-2-methylbenzamide: Standard hydrolysis and coupling
conditions (60%).
Analytical Data for 29: LCMS: 446.15 [M+1]; HPLC: 95% (@ 254 nm) (Rt; 5.257); 1H
NMR (DMSOꢀd6, 400 MHz) δ 11.47 (s, 1H), 8.20 (t, 1H, J=4.8 Hz), 7.25 (s, 1H), 7.04 (s, 1H),
5.85 (s, 1H), 4.23 (d, 2H, J=4.8 Hz), 3.37ꢀ3.45 (m, 1H), 2.49 (3H merged in solvent peak), 2.17
(s, 3H), 2.13 (s, 3H), 2.10 (s, 3H), 1.53ꢀ1.67 (m, 4H), 1.38ꢀ1.50 (m, 4H).
Synthesis of compound 34: 5-chloro-N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-
yl)methyl)-3-(ethyl(tetrahydro-2H-pyran-4-yl)amino)-2-methylbenzamide
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 35
1
2
3
4
Synthesis of methyl 5-chloro-2-methyl-3-((tetrahydro-2H-pyran-4-yl)amino) benzoate:
5
6
Standard reductive amination conditions (0.5 g, 36%).
7
8
9
Synthesis of methyl 5-chloro-3-(ethyl (tetrahydro-2H-pyran-4-yl) amino)-2-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
methylbenzoate: Standard aliphatic coupling conditions. (0.18 g, 34%).
Synthesis of 5-chloro-N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-yl)methyl)-3-
(ethyl(tetrahydro-2H-pyran-4-yl)amino)-2-methylbenzamide: Standard hydrolysis and
coupling conditions. (0.10 g, 44%).
Analytical Data for 34: LCMS: 432.25 [M+1]; HPLC: 94% (@ 254 nm) (Rt; 4.833); 1H
NMR (DMSOꢀd6, 400 MHz) δ 11.46 (s, 1H), 8.22 (t, 1H), 7.19 (s, 1H), 6.96 (s, 1H), 5.85 (s,
1H), 4.23 (d, 2H, J=4 Hz), 3.81 (d, 2H, J=10 Hz), 3.20ꢀ3.21 (m, 2H), 2.94ꢀ3.02 (m, 3H), 2.18 (s,
3H), 2.15 (s, 3H), 2.10 (s, 3H), 1.48ꢀ1.61 (m, 4H), 0.78 (t, 3H, J=6.4 Hz).
Synthesis of compound 3: 5-(cyclopentyl(methyl)amino)-N-((4,6-dimethyl-2-oxo-1,2-
dihydropyridin-3-yl)methyl)-4-methyl-4'-(morpholinomethyl)-[1,1'-biphenyl]-3-
carboxamide
Synthesis of 5-(cyclopentyl(methyl)amino)-N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-
yl)methyl)-4-methyl-4'-(morpholinomethyl)-[1,1'-biphenyl]-3-carboxamide: Standard
Suzukiꢀcoupling conditions.
Analytical Data for 3: LCMS: 543.22 [M+1]; HPLC: 99% (@ 254 nm) (Rt;4.181); 1H NMR
(DMSOꢀd6, 400 MHz) δ 11.46 (s, 1H), 8.17 (t, 1H, J=4.4 Hz), 7.98 (s, 1H), 7.73 (d, 1H, J=7.6
Hz), 7.57 (d, 2H, J=7.6 Hz), 7.37 (s, 2H), 7.17 (s, 1H), 5.85 (s, 1H), 4.27 (d, 2H, J=4.8 Hz), 3.44ꢀ
26
ACS Paragon Plus Environment

Page 27 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
3.57 (m, 7H), 2.54 (s, 3H), 2.32ꢀ2.36 (m, 4H), 2.23 (s, 3H), 2.19 (s, 3H), 2.09 (s, 3H), 1.69ꢀ1.72
(m, 2H), 1.61 (m, 2H), 1.43ꢀ1.50 (m, 4H).
7
8
9
Synthesis of compound 31: 5-bromo-N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-
yl)methyl)-2-methyl-3-(methyl(tetrahydro-2H-pyran-4-yl)amino)benzamide
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of methyl 5-bromo-2-methyl-3-(methyl(tetrahydro-2H-pyran-4-
yl)amino)benzoate: Standard aliphatic coupling conditions (0.33 g, 80%).
Synthesis of 5-bromo-N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-yl)methyl)-2-methyl-
3-(methyl(tetrahydro-2H-pyran-4-yl)amino)benzamide: Standard hydrolysis and coupling
conditions.
Analytical Data for 31: LCMS: 462.23 [M+1]; HPLC: 99% (@ 254 nm) (Rt; 5.276); 1H
NMR (DMSOꢀd6, 400 MHz) δ 11.46 (s, 1H), 8.20 (t, 1H, J=4.8 Hz), 7.24 (d, 1H, J=2 Hz), 7.04
(d, 1H, J=2 Hz), 5.85 (s, 1H), 4.24 (d, 2H, J=4.8 Hz), 3.84 (d, 2H, J=10.8 Hz), 3.22ꢀ3.28 (m,
2H), 2.96 (m, 1H), 2.56 (s, 3H), 2.18 (s, 3H), 2.13 (s, 3H), 2.10 (s, 3H), 1.52ꢀ1.61 (m, 4H).
Synthesis of compound 4: N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-yl)methyl)-5-
(ethyl(tetrahydro-2H-pyran-4-yl)amino)-4-methyl-4'-(morpholinomethyl)-[1,1'-biphenyl]-
3-carboxamide
Synthesis of methyl 5-bromo-3-(ethyl (tetrahydro-2H-pyran-4-yl)amino)-2-
methylbenzoate: To a stirred solution of methyl 5ꢀbromoꢀ2ꢀmethylꢀ3ꢀ((tetrahydroꢀ2Hꢀpyranꢀ4ꢀ
yl)amino)benzoate (14 g, 42.7 mmol) in dichloroethane (150 mL) was added acetaldehyde (3.75
g, 85.2 mmol) and acetic acid (15.3 g, 256 mmol). The resulting reaction mixture was stirred at
room temperature for 15 minutes. Then sodium triacetoxyborohydride (27 g, 128 mmol) was
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 35
1
2
3
4
5
6
7
8
9
added at 0 °C and reaction stirred at room temperature for 3 h. After completion (monitored by
TLC), aqueous sodium bicarbonate was added to the reaction mixture till pH 7ꢀ8, the organic
phase was separated and the aqueous phase was extracted with ethyl acetate. The combined
organic layers were dried over anhydrous sodium sulfate, filtered and concentrated under
reduced pressure. The crude compound was purified by column chromatography (100ꢀ200 mesh
silica gel) eluting with ethyl acetate: hexane to afford the title compound as viscous liquid (14 g,
93.3%). 1H NMR (DMSOꢀd6, 400 MHz) δ 7.62 (s, 1H), 7.52 (s, 1H), 3.80 (bs, 5H), 3.31 (t, 2H),
2.97ꢀ3.05 (m, 2H), 2.87ꢀ2.96 (m, 1H), 2.38 (s, 3H), 1.52ꢀ1.61 (m, 2H), 1.37ꢀ1.50 (m, 2H), 0.87
(t, 3H, J=6.8 Hz).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of 5-bromo-N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-yl)methyl)-3-
(ethyl(tetrahydro-2H-pyran-4-yl)amino)-2-methylbenzamide: Standard hydrolysis and
coupling conditions. (14 g, 74%). 1H NMR (DMSOꢀd6, 400 MHz) δ 11.47 (s, 1H), 8.23 (t, 1H),
7.30 (s, 1H), 7.08 (s, 1H), 5.85 (s, 1H), 4.23 (d, 2H, J=4.4 Hz), 3.81 (d, 2H, J=10.4 Hz), 3.20ꢀ
3.26 (m, 2H), 3.00ꢀ3.07 (m, 1H), 2.91ꢀ2.96 (m, 2H), 2.18 (s, 3H), 2.14 (s, 3H), 2.10 (s, 3H),
1.58ꢀ1.60 (m, 2H), 1.45ꢀ1.50 (m, 2H), 0.78 (t, 3H, J=6.8 Hz).
Synthesis of N-((4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-yl)methyl)-5-(ethyl(tetrahydro-
2H-pyran-4-yl)amino)-4-methyl-4'-(morpholinomethyl)-[1,1’-biphenyl]-3-carboxamide:
Standard Suzukiꢀcoupling conditions.
Analytical Data for 4 (EPZ-6438, tazemetostat): LCMS: 573.35 [M+1]; HPLC: 99% (@ 254
nm) (Rt; 3.999); 1H NMR (DMSOꢀd6, 400 MHz) δ 11.46 (s, 1H), 8.19 (t, 1H), 7.57 (d, 2H, J=7.2
Hz), 7.36ꢀ7.39 (m, 3H), 7.21 (s, 1H), 5.85 (s, 1H), 4.28 (d, 2H, J=2.8 Hz), 3.82 (d, 2H, J=9.6
Hz), 3.57 (bs, 4H), 3.48 (s, 2H), 3.24 (t, 2H, J=10.8Hz), 3.07ꢀ3.09 (m, 2H), 3.01 (m, 1H), 2.36
28
ACS Paragon Plus Environment

Page 29 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
(m, 4H), 2.24 (s, 3H), 2.20 (s, 3H), 2.10 (s, 3H), 1.64ꢀ1.67 (m, 2H), 1.51ꢀ1.53 (m, 2H), 0.83 (t,
3H, J=6.4 Hz).
7
8
9
ASSOCIATED CONTENT
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Supporting Information: Torsion angle calculations for orthoꢀsubstituted benzamide and
aniline cores. Selectivity data and all relevant cellular data for compound 4, synthesis and
analytical data for all synthesized compounds 3-51 and complete compound data summary for
compounds 1-51. CSV file: Contains SMILES strings for compounds 1-51.
This material is available free of charge via the Internet at http://pubs.acs.org.”
AUTHOR INFORMATION
Corresponding Author
*Kevin W. Kuntz
kkuntz@ribontx.com
Present Addresses
† Ribon Therapeutics, 99 Hayden Ave. Lexington, MA 02421
‡ Warp Drive Bio, 400 Technology Square, 2nd Floor, Cambridge, MA 02139
§ Syros Pharmaceuticals, 620 Memorial Dr. Cambridge, MA 02139
# Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080.
Author Contributions
29
ACS Paragon Plus Environment