Discovery of AG-270, a First-in-Class Oral MAT2A Inhibitor for the Treatment of Tumors with Homozygous MTAP Deletion

Article abstract of DOI:10.1021/acs.jmedchem.0c01895

The metabolic enzyme methionine adenosyltransferase 2A (MAT2A) was recently implicated as a synthetic lethal target in cancers with deletion of the methylthioadenosine phosphorylase (MTAP) gene, which is adjacent to the CDKN2A tumor suppressor and codelet

Full text of DOI:10.1021/acs.jmedchem.0c01895

pubs.acs.org/jmc
Drug Annotation
Discovery of AG-270, a First-in-Class Oral MAT2A Inhibitor for the
Treatment of Tumors with Homozygous MTAP Deletion
Zenon Konteatis,*,# Jeremy Travins,# Stefan Gross, Katya Marjon, Amelia Barnett, Everton Mandley,
Brandon Nicolay, Raj Nagaraja, Yue Chen, Yabo Sun, Zhixiao Liu, Jie Yu, Zhixiong Ye, Fan Jiang,
Wentao Wei, Cheng Fang, Yi Gao, Peter Kalev, Marc L. Hyer, Byron DeLaBarre, Lei Jin, Anil K. Padyana,
Lenny Dang, Joshua Murtie, Scott A. Biller, Zhihua Sui, and Kevin M. Marks
Cite This: J. Med. Chem. 2021, 64, 4430−4449
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The metabolic enzyme methionine adenosyltrans-
ferase 2A (MAT2A) was recently implicated as a synthetic lethal
target in cancers with deletion of the methylthioadenosine
phosphorylase (MTAP) gene, which is adjacent to the CDKN2A
tumor suppressor and codeleted with CDKN2A in approximately
15% of all cancers. Previous attempts to target MAT2A with small-
molecule inhibitors identified cellular adaptations that blunted
their efficacy. Here, we report the discovery of highly potent,
selective, orally bioavailable MAT2A inhibitors that overcome
these challenges. Fragment screening followed by iterative
structure-guided design enabled >10 000-fold improvement in
potency of a family of allosteric MAT2A inhibitors that are
substrate noncompetitive and inhibit release of the product, S-
adenosyl methionine (SAM), from the enzyme’s active site. We demonstrate that potent MAT2A inhibitors substantially reduce
SAM levels in cancer cells and selectively block proliferation of MTAP-null cells both in tissue culture and xenograft tumors. These
data supported progressing AG-270 into current clinical studies (ClinicalTrials.gov NCT03435250).
devise effective MAT2A inhibitors have been challenging.
Methionine analogues such as cycloleucine7,8 as well as
stilbene derivatives9 have been reported in the literature to
be inhibitors of MAT2A; however, their weak biochemical
potency (>10 μM) and very weak cellular activity did not
enable their development into useful therapeutics. The recent
discovery of a moderately potent allosteric MAT2A inhibitor,
PF-9366,10 demonstrates the potential to drug MAT2A via an
allosteric mechanism. Unfortunately, PF-9366 treatment in
cells induced cellular adaptation, particularly upregulation of
MAT2A itself, which blunted cellular potency and led to
inadequate antiproliferative effects.
INTRODUCTION
■
The methylthioadenosine phosphorylase (MTAP) gene is
located adjacent to the CDKN2A tumor suppressor and is
codeleted with CDKN2A in approximately 15% of all cancers,
leading to aggressive tumors with poor prognoses for which no
effective molecularly targeted therapies exist.1−4 The metabolic
enzyme methionine adenosyltransferase 2A (MAT2A) has an
important role in metabolism and epigenetics because it is the
primary producer of the universal methyl donor S-adenosyl
methionine (SAM). Recent work has demonstrated that
depletion of MAT2A using RNA interference leads to a
selective antiproliferative effect in cancers with deletion of
MTAP.2,5,6 A simple explanation for this selective vulnerability
has been described, in which the activity of the SAM-utilizing
type II protein arginine N-methyltransferase 5 (PRMT5) is
inhibited by the MTAP substrate, 5′-methylthioadenosine
(MTA), which accumulates when MTAP is deleted. Within
this tumor environment, the catalytic activity of the PRMT5
enzyme is reduced, and it becomes vulnerable to further
inhibition by reduction of SAM levels, whereas its activity in
normal tissues and MTAP-proficient tumors remains largely
unaffected.2
Herein, we describe the drug discovery efforts that led to the
identification of AG-270, a first-in-class, orally bioavailable
MAT2A inhibitor currently in clinical development
(ClinicalTrials.gov NCT03435250). This class of MAT2A
Received: November 2, 2020
Published: April 8, 2021
Although these results suggest that targeting MAT2A may
prove beneficial in MTAP-deleted cancers, past efforts to
https://doi.org/10.1021/acs.jmedchem.0c01895
© 2021 American Chemical Society
J. Med. Chem. 2021, 64, 4430−4449
4430

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 1. Hit identification. (A) MAT2A enzyme inhibition IC50 (μM). Values are the mean of three experiments. (B) Surface plasmon resonance
sensorgrams (the result of only one experiment each shown). (C) Mechanism of action study demonstrates compound 2 is noncompetitive with
regard to ATP and L-methionine substrates; test concentrations of 2 were 0.5, 1.5, and 4.5 μM.
inhibitors is allosteric, substrate noncompetitive, and inhibits
MAT2A activity by preventing product release. We also report
findings from preclinical studies describing the in vitro and in
vivo characterization of this class of inhibitors with a focus on
our clinical molecule AG-270.
We cocrystallized compound 2 with the MAT2A protein in
complex with SAM (MAT2A·SAM·2) and determined the
structure of this ternary complex at atomic resolution of 1.14 Å
(Supporting Information Table S1). We observed two copies
of 2 bound at the interface of the obligate MAT2A dimer, one
molecule of 2 per monomer of MAT2A, in the same allosteric
pocket as previously reported for PF-936610 (Figure 2).
Compound 2 forms key interactions with MAT2A, including
bidentate hydrogen bonds from the amide carbonyl and the
nitrogen of the pyrazole of 2 to the guanidino group of Arg335,
and two hydrophobic van der Waals contacts between the two
phenyl groups of 2 and the protein. Another observation in this
crystal structure is the desolvation of the allosteric pocket upon
binding of the inhibitor. Whereas the unliganded pocket
contains multiple water molecules extending out to the solvent
front (PDB code: 2P02),13 this liganded structure contains
only a few remaining water molecules. Importantly, this
cocrystal structure determined in the presence of SAM
molecules bound in the MAT2A active sites demonstrates
that the inhibitor binding traps the reaction product by closure
of the α helix gates (Figure 3A). This confirms the mode of
action to be enhanced inhibition of enzyme product (SAM)
release, consistent with our kinetic studies demonstrating a
noncompetitive mechanism-of-action. These interactions are
reflected in the increase in enzymatic potency and established
this pyrazolo[1,5-a]pyrimidin-7(4H)-one class as worthwhile
for further exploration. The five substituents pursued in this
SAR campaign are shown in Figure 2C.
RESULTS AND DISCUSSION
■
Identification of MAT2A Inhibitors via Fragment
Screening. We screened a library of >2000 fragments for
the ability to bind to the MAT2A protein dimer using a mass
spectrometry-based ultrafiltration assay.11 Thirty-one fragment
hits were identified, a modest 1.55% hit rate, which were then
validated in orthogonal enzymatic and surface plasmon
resonance (SPR) assays. Of these fragments, compound 1
showed weak enzymatic inhibition of MAT2A at 620 μM
(Figure 1A), which was consistent with a weak SPR
sensorgram (Figure 1B). A hit expansion similarity search
(Tanimoto >80%; 350 < MW > 200) was carried out, and 54
commercially available compounds were purchased and tested
for enzymatic inhibition of MAT2A, followed by SPR binding.
Among the virtual hits, 2 was identified as the most potent
MAT2A inhibitor (Figure 1A), having an IC50 of 1.5 μM and
showing a robust, dose dependent SPR signal with a
dissociation constant (KD) of 70 μM (Figure 1B). Detailed
kinetic analysis12 with compound 2 demonstrated a non-
competitive inhibition modality with regard to both ATP and
L-methionine, suggesting a binding location outside the
catalytic active site (Figure 1C).
4431
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 2. Cocrystal structure of compound 2 with MAT2A and reaction product SAM (PDB code 7KCE). (A) Homodimeric form of MAT2A
with two protein chains shown in green and cyan, respectively. SAM molecules, in yellow, are in the active site and compound 2, in magenta, is in
an allosteric site. (B) Allosteric binding site of MAT2A. Compound 2, in magenta, and water interacting with residue Gly215 in red sphere. (C)
Structural class pyrazolo[1,5-a]pyrimidin-7(4H)-one and all the substituents to be explored in the SAR campaign.
Structure-Based Design Leading to Potent, MTAP-
Null Selective MAT2A Inhibitor AGI-24512. The cocrystal
structure of 2 indicated that mostly hydrophobic residues
occupied regions of the binding site surrounding R1 and R2
substituents of 2. This led to targeted substituents at the R1
and R2 positions to establish initial SARs for this series.
Attempts to increase polarity in substituent R1 by modifying
the phenyl group to 2-pyridine in 3 or to 2-furan in 4 led to
modest losses in potency (Table 1). Alternatively, increasing
hydrophobicity and sp3 (aliphatic carbons) character led to
improvements in enzymatic potency, as shown in 5−8. We
rationalized that piperidine could be a good structural
homologue of cyclohexene (7), and indeed, 9 maintained
good potency at 0.26 μM.
Initial exploration of the R2 group centered around probing
the properties of the R2 binding pocket. Replacing phenyl with
heterocycles in 10 (2-furan), 11 (4-pyrazole), and 13 (2-
thiazole) led to significant decreases in potency, suggesting
that polarity in R2 substituents was not tolerated in this
nonpolar pocket. Nonaromatic moieties such as cyclohexene
(12) led to large losses in enzymatic potency (Table 2). Only
14 (3-hydroxyphenyl) and 15 (3-fluorophenyl) maintained
similar potency to 2, indicating a preference for aromatic
groups in this highly hydrophobic subpocket.
residue Gly215, and this water-436 was within 5 Å of the core
of the pyrazolo-pyrimidinone scaffold. Using structure-based
design,14 we added a phenol substituent attached at the C-6
position of the scaffold to replace this water interaction. We
made this R3 modification using 9 as the core because it had
better solubility than 2 (Table 1). The resulting analogue,
AGI-24512, showed improved potency with an enzymatic IC50
of 8 nM. Mode of action enzymatic studies were carried out,
confirming that AGI-24512, like compound 2, is non-
competitive with respect to both ATP and L-Met (Supporting
Information Figure S2). A few close analogues of AGI-24512
were synthesized to establish the value of this new interaction
(Table 3). The addition of a phenyl ring at R3 (16) showed
0.15 μM enzymatic inhibition, while the 4-pyridyl (compound
17) had an IC50 of 0.160 μM and a thiazole analogue (18) was
much less active. Only the 3-pyrazolyl substitution (19)
maintained acceptable potency, presumably by still interacting
with Gly193 via a water bridge (Table 3).
The structure of the ternary complex of MAT2A bound to
SAM and AGI-24512 (MAT2A·SAM·AGI-24512) was
determined at 1.10 Å resolution (Supporting Information
Table S1) and confirmed both the allosteric binding site and
the predicted replacement of the interaction with water-436 by
the phenol substituent at R3 (Figure 3B). Additional
interactions are made by the piperidine group at the C-3
position of the scaffold, which is encased in a hydrophobic
protein environment created by residues Phe18, Phe20 of one
protomer, and Phe139, Ser331, Phe333, and Leu315 of the
adjacent chain MAT2A dimer (Figure 3B and C). The atomic
resolution cocrystal structure also permitted assessment of the
effects of inhibitor binding on the active site of MAT2A. The
active site possesses a loop−α-helix−loop arrangement that
caps the entrance to the active site and closes upon SAM
binding (Figure 3A and D [SAM-free structure, PDB code
A cell assay using HCT116 MTAP-null cells was developed
to assess MAT2A inhibition in cancer cells by measuring the
abundance of the MAT2A product SAM following treatment
with increasing concentrations of the more potent compounds.
Compounds 14, 5, and 7 showed increasing cellular potency
(Supporting Information Figure S1, Tables 1 and 2) in line
with their biochemical potency, suggesting that the two assays
were correlated and thus validating the cellular assay.
The MAT2A-2 cocrystal structure (Figure 2B) showed a
water molecule interacting with the N−H of the protein
4432
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
adaptation. This pharmacological selective reduction of growth
of MTAP-null genotype cancer cells confirms our previous
observations using genetic tools targeting MAT2A,2 suggesting
that pharmacologic inhibition of MAT2A enzymatic function
represents a viable therapeutic approach with potential for a
superior therapeutic window by sparing the MTAP-WT
noncancerous tissue. A variety of further biological studies
with AGI-24512 are discussed in a recent publication.16
Discovery of in Vivo Lead Candidate AGI-25696.
Initial assessment of the pharmacokinetic properties of AGI-
24512 indicated poor oral absorption and a short half-life in
rats. At the same time, both mouse and human in vitro
metabolite identification studies of 19 indicated, as expected,
that the piperidine ring at R1 was mostly responsible for the
poor liver microsomal stability exhibited (human microsomal
extraction ratio, ER, 0.72, Supporting Information Figure S3).
The focus in these experiments was to identify major
metabolites to aid appropriate design to reduce metabolism
or abrogate any specific metabolic pathways. All compounds at
this stage showed rapid oxidative metabolism across preclinical
species and human microsomes, so we set out to improve
metabolic stability. We first sought to generate analogues in the
R3 position that contain other hydrogen-bond acceptors to
investigate their ability to maintain enzymatic potency while
improving their metabolic stability. Compounds 20−23
maintained reasonable enzymatic potency but did not improve
microsomal stability (ER, Table 4). We revisited 24, an earlier
analogue of AGI-24512, which had a phenyl group in place of
the piperidine at R1 and showed better metabolic stability with
a human microsomal ER of 0.54. The next set of analogues
kept the quinoline at R3 and modified R1 with all the previously
identified potency-enhancing substituents (25−27 and AGI-
25696). While most compounds showed good enzymatic
potency, only AGI-25696, which incorporated the phenyl
substituent at R1, gave a very stable analogue with human
microsomal ER of 0.16 (Table 4). This first metabolically
stable analogue showed good cellular activity at 150 nM
(Figure 4A, HCT116 cell assay, SAM IC50) and was further
profiled in vivo.
Oral treatment of tumor-bearing mice with AGI-25696 led
to sustained high exposure in pancreatic KP4 MTAP-null
xenograft tumors (Figure 4B; maximum serum concentration
[Cmax] of 179 000 21 500 ng/mL and 32 300 6100 ng/g;
AUC0−12h of 1 650 000 and 372 000 h.ng/g in plasma and
tumor, respectively, after 3 b.i.d. doses). To test whether
MTAP-null tumors are sensitive to the antiproliferative effects
of MAT2A inhibition in vivo, AGI-25696 was given orally once
daily (q.d.) to mice bearing subcutaneous KP4 MTAP-null
xenografts. A high dose of 300 mg/kg was given to ensure a
result would be obtained for proof of principle demonstration.
Consistent with in vitro data, compound treatment led to a
substantial reduction in growth of KP4 MTAP-null xenografts
(tumor growth inhibition [TGI] = 67.8%, p = 0.0001, Figure
4C) with no adverse effects on mouse body weight throughout
the 33 day study (Figure 4D). Having established in vivo proof
of principle in this mouse study, the emphasis shifted to
improving physical properties (Figure 4A) and cellular activity
to produce a compound with the potential for clinical
development.
Figure 3. Cocrystal structure of MAT2A·SAM·AGI-24512. (A)
Crystal structure of MAT2A SAM free form (PDB code: 5A19). (B)
Close-up of the allosteric binding site, highlighting AGI-24512 and its
interactions with MAT2A protein and a key water molecule. (C) The
current crystal structure obtained with AGI-24512 in the presence of
SAM (PDB code 7KCF). (D) Close-up of the apo-MAT2A region
indicating the “open” form of the α-helix gate that is unstructured in
the absence of SAM. (E) Close-up of the active site gate loop (protein
in green and yellow, SAM in atom color).
5A19]15 vs Figure 3C and E [inhibitor-bound structure]). In
all the inhibitor-bound MAT2A cocrystal structures, we
observed SAM in the active site and the gate loop preserved
in the closed position (Figure 3E).
This lead compound, AGI-24512, was fully characterized in
cellular assays and became an important in vitro tool that
enabled further cellular studies to evaluate MAT2A biology.16
AGI-24512 treatment in the HCT116 MTAP-null cell assay
led to a dose-dependent decrease in SAM levels with an IC50 of
100 nM (Table 3). AGI-24512 also demonstrated MTAP-null
selective antiproliferative activity in an HCT116 engineered
cell model.16 Consistent with prior results with PF-9366,10 we
observed upregulation of MAT2A protein upon treatment with
AGI-24512. Notably, this pathway feedback mechanism did
not prevent the antiproliferative activity of AGI-24512. We
hypothesized that the improved potency of AGI-24512 relative
to prior MAT2A inhibitors, i.e. PF-9366, enabled AGI-24512
to maintain antiproliferative effects despite this cellular
Discovery of Clinical Candidate AG-270. AGI-25696
had very high PPB (>99.9% in human plasma measured by
ultracentrifugation, Figure 4A) and a high efflux ratio in a
Caco-2 assay (6.0, Figure 4A), which projected to a very high
4433
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Table 1. R1 Modifications
a
Values are the mean of three experiments.
dose for adequate human exposure. The weakly acidic nature
of this chemotype, driven by the tautomerism described in
Figure 5, was hypothesized to contribute to the high plasma
protein binding. The pyrazolo[1,5-a]pyrimidin-7(4H)-one (A)
scaffold can easily tautomerize to both the pyrazolo[1,5-
a]pyrimidin-7-ol (B) and the pyrazolo[1,5-a]pyrimidin-7(1H)-
one (C), giving rise to various potential modes of binding to
plasma proteins because the acidic proton can locate at three
different positions.
N-Methylation of the Core Reduces Human PPB but
Decreases Percentage Maximal Inhibition. Indeed, the N-
methylated analogues 28 and 29 (Table 5, R5 = CH3), where
the acidic proton was eliminated and tautomerization was
prevented, showed reduced PPB relative to their correspond-
ing R5 = H counterparts (compare free fraction of 0.09% for
AG-24512 with 1.54% for 28 and 0.03% for AGI-25696 with
0.9% for 29, as shown in Table 5). However, analogues 28 and
29 showed dramatic decreases in the percent maximal
inhibition12 of MAT2A from >90% to <70%. This trend was
observed with other R5 analogues (not reported here), and
thus, this strategy for improving free fraction was not successful
in maintaining full efficacy of the inhibitors. Throughout this
discovery program, the enzymatic allosteric inhibition IC50
correlated very well with the 72 h cell assay EC50 but not with
the percent maximal inhibition, as expected for the mechanism
(Supporting Information Figure S4).
Masking of Core N−H by Intramolecular Hydrogen
Bond Improves Percentage Maximal Inhibition but
Affects Other ADME Parameters. During the early SAR
development phase of the program, R4 substituent variations
4434
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Table 2. R2 Modifications
a
Values are the mean of three experiments. No fit = no activity.
were explored with R1 and R2 = Ph and R3 = H. Among the
analogues explored (represented by 30−34, Table 6; addi-
tional analogues not shown), only 34 exhibited better potency
than 2 and increased percent maximal inhibition from partial
to 88%. Encouraged by this result, we determined the cocrystal
structure of this analogue with MAT2A (MAT2A·SAM·34) at
1.24 Å resolution (Supporting Information Table S1). New
interactions were noted between the amide carbonyl of residue
Gln190 and the N−H at R4, the backbone carbonyl of Ile332
with the hydrogen of the pyrazole of R4, along with an
intramolecular hydrogen bond between the sp2 (aromatic)
nitrogen of the R4 pyrazole and the N−H at R5 shown in
Figure 6A. This intramolecular hydrogen bond was then
hypothesized to be responsible for the improved maximal
inhibition (Table 6) and lowering human PPB (Table 6) by
masking the acidic hydrogen on the scaffold. This strategy then
became a more viable approach toward inhibitors with
maximal inhibition and lower human PPB than N-methylation.
Similar internal hydrogen-bonding strategies have been
employed in medicinal chemistry to mask polarity and have
been exploited to improve some physicochemical parameters
of multiple structural classes.17−19
Analogue 35, bearing a 4-hydroxyphenyl group at R3,
showed low nanomolar enzymatic potency and high percent
4435
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Table 3. R3 Exploration
a
Values are the mean of three experiments.
maximal inhibition (Table 7). The cocrystal structure of this
compound (MAT2A·SAM·35) was determined at 1.24 Å
resolution, confirming the expected binding mode (Figure 6B).
Gratifyingly, all of the interactions observed in the crystal
structures of 34 and AGI-24512 were maintained in 35,
validating the cumulative structural understanding of
MAT2A−inhibitor interactions. Analogues of 35 were
synthesized, varying both R1 and R3 while keeping the amino
pyrazole substituent at R4 (Table 7). Characterization of these
advanced compounds included cellular assays that measured
SAM production in MTAP-null cells at 72 h to account for the
MAT2A upregulation as well as cellular proliferation assays.
Microsomal stability and PPB were also measured. All
compounds in this subseries had excellent enzymatic potency
(4−14 nM) and maximal inhibition (92−96%) but either poor
inhibition of cell proliferation (35, 36, 39) or moderate mouse
and human microsomal stability (37 and 38). Another
deficiency of these analogues was the very high PPB observed
for all members of this subseries (>99.5%, Table 7).
Masking Core N−H and Decreasing the Number of
Hydrogen-Bond Donors Leads to AG-270. The successful
masking of polarity by internal hydrogen bonding encouraged
further modifications at R4. However, the presence of the
additional hydrogen-bond donor in this region presented a
new challenge for permeability, both masking polarity and
decreasing the number of hydrogen-bond donors needed to be
pursued concomitantly. The N−H in the pyrazole ring of 36
was protected via an ethyl carbamate to give 40, which
maintained enzymatic and cellular potency and showed mildly
improved activity in the cell proliferation assay (Table 8).
Oxadiazole ester 41 also showed good enzymatic potency and
percent maximal inhibition, establishing that the second
hydrogen bond to the backbone carbonyl of residue Ile332
was not contributing much to the potency of the earlier
inhibitors. This finding led to a subseries of analogues
containing over 70 heterocycles in the R4 position. Some of
these are reported in Table 8 as representatives of the results
obtained. The 3-amino-isoxazole substitution (42) maintained
good potency and occupancy in the enzymatic assay but
4436
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Table 4. Further R3 Exploration
a
Values are the mean of three experiments. ER = extraction ratio; HLM = human liver microsome.
showed weaker cell activity and poor inhibition of HCT116
MTAP-null cell proliferation. However, microsomal stability
was excellent, and PPB was much improved to the level
previously seen for the N-methyl analogues at R5 (98.92%,
Table 8). Similar trends were observed with a variety of
analogues, exemplified here by the 2-pyridine (43) and the 2-
pyridazine (44), demonstrating good enzymatic and cellular
potency, excellent microsomal stability, and good free fraction,
4437
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 4. Pharmacologic targeting of MAT2A with in vivo tool molecule selectively blocks growth of MTAP-null tumors in vivo. Fifteen mice per
arm (total 30 animals) were used in this study. (A) Key in vitro parameters of AGI-25696. (B) PK data from the study after 3 b.i.d. doses. (C)
Mean tumor volume in mice inoculated with KP4 MTAP-null pancreatic cells treated daily with 300 mg/kg AGI-25696 (black line) or vehicle (red
line). (D) Mean body weight of the mice. Error bars show standard error of the mean. HLM, human liver microsomes; MLM, mouse liver
microsomes.
Figure 5. Are tautomers responsible for the physical chemistry properties? Hypothesis: three possible tautomers of the pyrrolopyrimidinone
scaffold (A−C) share the acidic proton in different locations of the scaffold, which can potentially increase the ability of each compound to bind to
plasma proteins.
but weak inhibition of HCT116 MTAP-null cell proliferation
(Table 8). Pyridine at R4 showed a higher free fraction than
pyridazine at R4, presumably due to the better intramolecular
hydrogen bond with the hydrogen in the core structure. The
compound with the best overall combination of potency and
drug-like properties was AG-270, combining the 2-pyridine at
R4 with the cyclohexene at R1. The measured pKa of AG-270
was 8.56, a substantial increase from the measured pKa of 6.3
for AGI-25696. The impact of masking the acidic hydrogen
(pKa = 6.3) by the intramolecular interaction from the 2-
pyridine of the R4 substituent was to stabilize the tautomers to
mostly the desired one, where the hydrogen remained on the
core nitrogen at R5. This resulted in good potency in both
enzymatic and cellular assays while increasing the free fraction
to 1.54% (Table 8).
The cocrystal structure of MAT2A·SAM·AG-270 was
determined at 1.32 Å resolution (Supporting Information
Table S1) and confirmed the designed mode (Figure 7).
The expected interactions as seen in the cocrystal structure
of 35 were maintained and the internal hydrogen bond was
validated, suggesting that the bound conformation is a low
energy conformer of AG-270. Energy calculations on this
bioactive conformation confirmed that it is either at the global
minimum or near it, based on the method used (LowModeMD
or systematic, Supporting Figure S5).14 The only ligand−
protein interaction changes were the absence of the hydrogen
4438
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Table 5. N-Methylation Improves Human PPB Free Fraction
a
By removing the acidic proton via N-methylation, tautomerism is no longer possible, as seen in Figure 5, and the free fraction of compounds 28
and 29 increases.
bond from the compound to Ile332 and a new hydrophobic
interaction between the 2-pyridine group at R4 and the
MAT2A protein residues. These modifications led to the good
potency in in vitro assays. AG-270 demonstrated potent
reduction in levels of intracellular SAM, as well as MTAP-null−
selective antiproliferative activity in the HCT116 MTAP
isogenic cell model in vitro (Table 8).16 Further biochemical
studies confirmed that AG-270 maintained the same mode of
action as Compound 2 and AGI-24512, being noncompetitive
with respect to both ATP and L-Met (Supporting Information
Figure S2).
AG-270 was further characterized for ADME properties, in
vitro ancillary pharmacology, and oral bioavailability. The
properties of AG-270 (Table 9) were favorable overall,
including high microsomal stability and an improved plasma
free fraction (Table 8). However, AG-270 showed low
solubility and indeterminable permeability in the Caco-2
assay due to low recovery. Low solubility was addressed by
using spray dried dispersion,20 which resulted in good oral
bioavailability across species (monkey 31%, dog 48%, rat 92%;
Table 9 and Supporting Information Figure S6, Table S2).
Compounds synthesized in this program had variable solubility
that was not related to the calculated distribution coefficient
4439
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Table 6. R4 Exploration
a
Values are the mean of three experiments.
(between octanol and water) of ionizable compounds
(Supporting Information Figure S7), or any other easily
understood parameter, so empirical data guided the efforts.
AG-270 showed excellent microsomal, hepatocyte, and in vivo
metabolic stability across species (human, mouse, rat, dog, and
monkey; Table 9). No overly concerning off-target activities in
the standard battery of in vitro ancillary pharmacology assays
(Eurofins Discovery 95 biochemical screen assays, hERG,
CYP450, PXR, Ames test) were observed for this compound.
The only liabilities of AG-270 identified were inhibition of
UGT1A1 (IC50 of 1.1 μM) which could lead to elevation of
bilirubin, and of the hepatocyte transporter OATP1B1 (IC50 of
2.1 μM). The overall superior properties of AG-270 led to its
selection for detailed in vivo profiling.
Oral administration of AG-270 at 200 mg/kg q.d. in a
pancreatic KP4 MTAP-null xenograft mouse model for 38 days
led to sustained high-level exposure in plasma (Figure 8A, solid
line, data from last 24 h of the study) with corresponding
reductions in levels of tumor SAM (Figure 8A, dashed line). In
this multiday study, AG-270 was administered at multiple dose
levels to test whether MTAP-deleted tumors are sensitive to
the antiproliferative effects of MAT2A inhibition in vivo.
4440
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 6. Intramolecular hydrogen bonding improves maximum percentage inhibition. (A) Cocrystal structure of MAT2A·SAM·34 (PDB code
7KDA) identifies intramolecular hydrogen bond and new interaction with the MAT2A protein. (B) Cocrystal structure of MAT2A·SAM·35 (PDB
code 7KDB) confirms both the intramolecular hydrogen bond and the same new interactions seen with 34.
Consistent with in vitro data, treatment with AG-270 led to
dose-dependent reductions in tumor SAM levels and tumor
growth of KP4 MTAP-null xenografts (Figure 8B and 8C, TGI
= 67% at 200 mg/kg, p < 0.01) and was well tolerated, with
mean body weight loss <5% (Figure 8D). This in vivo study
demonstrated that efficacy is related to SAM reduction (Figure
8C) and a range of 60−80% SAM reduction is adequate for
maximal efficacy.
A similar study (being reported concurrently elsewhere)
with mice bearing subcutaneous HCT116 MTAP-null
xenografts was also conducted, and treatment with AG-270
(200 mg/kg q.d.) led to a similar significant reduction in tumor
growth (TGI = 75%, p < 0.01).16 No antiproliferative effects
were observed in HCT116 MTAP-WT tumor-bearing mice
dosed equivalently with AG-270,16 in line with the earlier
results obtained with the in vivo tool compound AGI-25696.
Further in vivo studies in patient-derived mouse xenograft
models, along with mechanism-of-action studies, leading to
rational combination therapies with AG-270 and other
anticancer treatment modalities have been reported.16 Mouse
and cynomolgus monkey safety studies established that AG-
270 has the desired preclinical safety profile to proceed to
clinical studies (unpublished data). A phase I clinical trial
investigating AG-270 in advanced solid tumors or lymphomas
with MTAP loss is currently underway (Clinicaltrials.gov
identifier: NCT03435250).
malonate 46 is reacted at high temperatures in a suitable
solvent such as xylene with the 1H-pyrazol-5-amine 56 to
provide the 5-hydroxy substituted pyrazolopyrimidinone 47,
bringing in the R1 and R2 substituents. Bicyclic intermediate 47
is reacted with phosphoryl chloride to provide the dichloride
48. The 7-chloro group of 48 is selectively reacted with sodium
methoxide to give the 7-methoxy intermediate 49, which is
aminated at the 5-position of the pyrazolopyrimidinone ring in
step E by reacting with 50 in the presence of a palladium
catalyst to give 51. Finally, the 7-methoxy group is dealkylated
to a ketone in step F to give the final product AG-270.
The synthesis of the 4-(cyclohex-1-en-1-yl)-3-phenyl-1H-
pyrazol-5-amine 56 is outlined in Scheme 1B. A mixture of 3-
oxo-3-phenylpropanenitrile 52 and 4-(methoxybenzyl)-
hydrazine were reacted in acetic acid/ethanol to give the
protected amino pyrazole 53, followed by addition of
cyclohexanone 54 in acetic acid to give 4-(cyclohex-1-en-1-
yl)-1-(4-methoxybenzyl)-3-phenyl-1H-pyrazol-5-amine 55.
Final deprotection of the 4-methoxybenzyl group was carried
out by trifluoromethanesulfonic anhydride in trifluoroacetic
acid stirred at 30 °C to give the amino pyrazole 56.
Other analogues were synthesized following Scheme 2A. A
solution of the commercially available methyl 2-(quinolin-6-
yl)acetate 57 was deprotonated with LDA and reacted with
acetyl chloride 58 to give intermediate 59. The mixture of
intermediate 59 and amine 60 in acetic acid was reacted at 90
°C to generate the final product AGI-25696. Scheme 2 B
describes the general synthesis of amine 60. A solution of 2-
phenylacetonitrile 61 and methyl benzoate 62 in THF was
treated with sodium hydride to give intermediate 63. A mixture
of 3-oxo-2,3-diphenylpropanenitrile and hydrazine hydrate in
acetic acid/ethanol were reacted to produce the amino
pyrazole 60.
CHEMISTRY
■
The synthetic route to AG-270, which is of general use to
prepare related analogues, is described in Scheme 1A. The
commercially available methyl 2-(4-methoxyphenyl) acetate 45
incorporating R3 (step A of Scheme 1) is reacted with dimethyl
carbonate in the presence of a suitable strong base such as
potassium tert-butoxide to provide intermediate 46. Dimethyl
4441
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Table 7. Further R4 Exploration Improves Cell Potency
a
Values for the biochemical and cell results are the mean of three experiments. ER = extraction ratio; HLM = human liver microsome.
Scheme 3 shows the general synthetic route for several
additional compounds reported here. The intermediate 59,
methyl 3-oxo-2-(quinolin-6-yl)butanoate, prepared as in
Scheme 2 (step A), was reacted with 3-phenyl-1H-pyrazol-5-
amine 64 (step B) to generate intermediate 65. 2-
(Trimethylsilyl)ethoxymethyl (SEM) protection followed by
NBS bromination gave intermediate 66. Suzuki reaction of 66
with the R1 intermediate 67 produced intermediate 68 and,
after SEM deprotection, provided the final compound 69.
that is not synergistic with MTA21,22 and thus fail to selectively
inhibit the proliferation of MTAP-null versus MTAP-WT
cancer cells. In contrast, reduction of SAM levels via MAT2A
inhibition can act synergistically with MTA elevation to
selectively inhibit PRMT5 in MTAP-deleted cells, thereby
inhibiting cell growth. Thus, the inhibition of MAT2A allows
for the selective inhibition of PRMT5 activity in MTAP-null
cancer cells and tumors by limiting the availability of SAM and
is predicted to afford a greater therapeutic window than known
PRMT5 inhibitors by limiting the potential toxicity of PRMT5
inhibition in normal, MTAP-WT tissues.23−25
We have described here the discovery of potent and selective
inhibitors of MAT2A via structure-guided design, starting from
a fragment with very low potency. Further, we have
demonstrated that potent pharmacologic inhibition of
MAT2A inhibits the growth of MTAP-deleted cancers in vivo
with good tolerability, thus validating MAT2A as a therapeutic
CONCLUSIONS
■
Previous work investigating MTAP loss in cancer cells revealed
that MTA accumulation sensitizes cells to short hairpin RNA-
mediated depletion of MAT2A and the SAM-utilizing enzyme
PRMT5. However, existing clinical-stage inhibitors of PRMT5
fail to recapitulate this MTAP-dependent effect, likely because
the existing inhibitors have a SAM-uncompetitive mechanism
4442
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Table 8. Final R4 Exploration Generates Compounds Having Balanced Properties
a
Values for the biochemical and cell results are the mean of three experiments. ER = extraction ratio; HLM = human liver microsome.
mesh. Analytical and preparative TLC plates were HSGF 254 (0.15−
0.2 mm thickness, Yucheng Chemical Co., Ltd. China). NMR spectra
were obtained on a Bruker AMX-400 instrument. Chemical shifts
were reported in parts per million (ppm, δ) downfield from
tetramethylsilane. Mass spectra were obtained with ESI using a
Waters LCT TOF mass spectrometer. HPLC was conducted using an
Agilent 1200 liquid chromatography column (UltiMate 4.6 × 50 mm,
5 μm, mobile phase A: 0.1% formic acid in water; mobile phase B:
acetonitrile). All compounds were assessed for purity by this reverse
column HPLC method (photodiode-array detection at wavelengths of
254 and 280 nm) and shown to have purity >95%. Microwave
reactions were run on a Biotage Initiator 2.5 microwave synthesizer.
The synthetic protocols and characterization of key intermediates
in the synthesis of AG-270 are included below. Synthesis of all other
key analogues are reported in the Supporting Information.
target for a biomarker-selected patient population of significant
size. On the basis of these discoveries, AG-270, an oral, first-in-
class MAT2A inhibitor, has entered clinical development and
is under investigation in a phase 1 trial that is currently
enrolling patients with MTAP-deleted solid tumors and
lymphomas (Clinical Trial NCT03435250).
EXPERIMENTAL SECTION
■
Chemistry. General Experimental Notes. In the following
examples, all reagents were purchased from commercial sources
(such as Alfa, Acros, Sigma-Aldrich, TCI, and Shanghai Chemical
Reagent Company) and used without further purification. Flash
chromatography was performed on an Ez Purifier III (Lisure Science
Co., Ltd., China) using a column with silica gel particles of 200−300
4443
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 7. Cocrystal structure of AG-270 (PDB code 7KCC). The crystal structure of AG-270 (in yellow) confirms the binding mode, the
intramolecular hydrogen bond, and all the key interactions with MAT2A.
from Vitas-M (catalog number STK591300). Compounds AGI-24512
Table 9. AG-270 Characterization: Enzymatic Activity, Cell
and AGI-25696 were synthesized as recently reported.26,27
Potency, and Cross-Species ADME Properties Show a
a
Synthesis of AG-270 (Scheme 1 A). Dimethyl 2-(4-
Methoxyphenyl)malonate (46). Step A: To dimethyl carbonate
Balance of Potency and Properties
(3.2 L) was slowly added potassium tert-butoxide (500 g) at 0 °C and
the mixture was stirred for 1 h at rt. Then, methyl 2-(4-
methoxyphenyl)acetate 45 (400 g) was added dropwise over 2 h
and stirred at rt overnight. The reaction was quenched with water (1.5
L), followed by extraction with ethyl acetate (1 L × 3). The combined
organic layers were washed with brine (1 L), dried over anhydrous
Na2SO4, and concentrated in vacuo. The residue was purified by flash
chromatography eluting with petroleum ether/ethyl acetate (20/1−5/
1) to obtain the desired product as a white solid (400 g, 75%). m/z =
MAT2A enzyme IC50 (μM)
HCT116 MTAP-null cell SAM
inhibition at 72 h IC50 (μM)
ER liver microsome, human/
monkey/dog/rat/mouse
ER hepatocyte, human/
monkey/dog/rat
Eh in vivo, monkey/dog/rat/
mouse (1 mg/kg IV)
PPB, human/monkey/dog/rat/ 1.5/1.1/2.8/1.6/0.28
0.014
0.02
0.05/0.20/0/0.21/0.38
0.16/0.13/0.12/0.18
0.023/0.022/0.142/0.015
1
239.1 [M + H]+, LC-MS: m/z [M + H]+ 239; H NMR (400 MHz,
mouse (% free)
solubility (μM) pH 2/7.4
chloroform-d): δ 7.33 (d, J = 8.6 Hz, 2H), 6.90 (d, J = 8.9 Hz, 2H),
4.61 (s, 1H), 3.80 (s, 3H), 3.75 (s, 6H).
b
<0.5/<1.0
0.13/0.76
Caco-2 permeability, A > B/
3-(Cyclohex-1-en-1-yl)-5-hydroxy-6-(4-methoxyphenyl)-2-
phenylpyrazolo[1,5-a]pyrimidin-7(4H)-one (47). Step B: To a
solution of dimethyl 2-(4-methoxyphenyl)malonate 46 (39.6 g, 166
mmol) in tri-n-butylamine (80 mL) at 198 °C was added 4-
cyclohexenyl-3-phenyl-1H-pyrazol-5-amine 56 (47.3 g, 199 mmol) in
portions, and the resultant mixture was stirred for 1 h at 198 °C. The
mixture was cooled to rt, and solvent was decanted. THF (150 mL)
and HCl (6N, 600 mL) were added with stirring vigorously for 0.5 h.
The precipitates were collected by filtration, washed with methanol,
and dried under reduced pressure to give 3-(cyclohex-1-en-1-yl)-5-
hydroxy-6-(4-methoxyphenyl)-2-phenylpyrazolo[1,5-a]pyrimidin-
7(4H)-one 47 (48 g, 70%) as a yellow solid. m/z = 414.17 [M + H]+,
b
efflux ratio (10−6cm/s)
Oral bioavailability at 5 mg/kg, 31/48/92
monkey/dog/rat (%F)
PXR (fold activation at 10 μM) 3.6 (22% of rifampin)
UGT1A1 inhibition IC50 (μM) 1.1
OATP1B1 inhibition IC50 (μM) 2.1
CYP inhibition, 3A4/2C9/
2C19/2D6/3A4 (μM)
CYP inactivation
all >10
>3.26× shift for 2B6 (15.6 and >50 μM
NADPH); >50 μM for other CYPs
a
b
Eh = hepatic extraction ratio; ER = extraction ratio. Solubility
measurements were unreliable; Caco-2 data are unreliable due to low
recovery.
1
LC-MS: m/z 414.2 [M + H]+; H NMR (400 MHz, DMSO-d6): δ
7.74 (d, J = 6.98 Hz, 2H), 7.31−7.49 (m, 5H), 6.94 (d, J = 8.60 Hz,
2H), 5.80 (br s, 1H), 3.78 (s, 3H), 2.15 (br s, 2H), 2.02 (br s, 2H),
1.65 (br s, 4H).
Synthesis of Key Pyrazolo-pyrimidine Analogues. The original
fragment compound 1 was identified via MS-based detection of small
fragments that bound to MAT2A in an affinity capture assay using an
ultracentrifugation assay.11 Further validation studies utilized
compound 1 material purchased from Vitas-M (catalog number
STK851087). Compound 2 was purchased in the hit expansion phase
5,7-Dichloro-3-(cyclohex-1-en-1-yl)-6-(4-methoxyphenyl)-2-
phenylpyrazolo[1,5-a]pyrimidine (48). Step C: A solution of 3-
(cyclohex-1-en-1-yl)-5-hydroxy-6-(4-methoxyphenyl)-2-
phenylpyrazolo[1,5-a]pyrimidin-7(4H)-one 47 (47.0 g, 104 mmol) in
phosphorus oxychloride (100 mL) was stirred at reflux for 16 h. The
solvent was removed in vacuo. The residue was added slowly to
4444
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 8. Dose-dependence study in KP4 xenograft mouse model reveals that maximal efficacy is related to SAM reduction. Twelve mice were used
per arm (60 total animals) in this study. AG-270 was given orally q.d. (on days 24−38) to mice inoculated subcutaneously with KP4 MTAP-null
cells, and both compound and SAM levels were monitored. (A) Plasma pharmacokinetics of AG-270 were monitored up to 24 h post last 200 mg/
kg dose (solid blue line) and displayed very good coverage; tumor SAM levels were also monitored in the same time period and showed stable, low
concentrations (blue dotted line). (B) Tumor volume over a 38-day study. (C) Tumor growth inhibition (TGI) and percentage tumor SAM
reduction at various doses indicate that SAM reduction between 60 and 80% leads to the same level of TGI at ∼66%. Reduction in SAM
concentration was dose dependent from 10 to 200 mg/kg and in tumor volume from 10 to 100 mg/kg. (D) Mean change in mouse body weight.
Error bars show standard error of the mean.
Scheme 1. (A) Synthesis of AG-270 and (B) Synthesis of the 4-(Cyclohex-1-en-1-yl)-3-phenyl-1H-pyrazol-5-amine 56
methanol (100 mL) cooled at 0 °C. The precipitates were collected
by filtration, washed with methanol, and dried under reduced pressure
to give 5,7-dichloro-3-(cyclohex-1-en-1-yl)-6-(4-methoxyphenyl)-2-
phenylpyrazolo[1,5-a]pyrimidine (50 g, 97%) as a yellow solid. m/z
450.2 [M + H]+, LC-MS: m/z 450.1 [M + H]+; 1H NMR (400 MHz,
DMSO-d6): δ 7.82 (d, J = 7.25 Hz, 2H), 7.36−7.56 (m, 5H), 7.10 (d,
J = 8.60 Hz, 2H), 5.87 (br s, 1H), 3.84 (s, 3H), 2.20 (br s, 4H), 1.70
(d, J = 4.57 Hz, 4H).
5-Chloro-3-(cyclohex-1-en-1-yl)-7-methoxy-6-(4-methoxyphen-
yl)-2-phenylpyrazolo[1,5-a]pyrimidine (49). Step D: To a solution of
5,7-dichloro-3-(cyclohex-1-en-1-yl)-6-(4-methoxyphenyl)-2-
4445
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Scheme 2. Synthesis of (A) AGI-25696 and (B) Amino Pyrazole 60
Scheme 3. General Synthetic Scheme for R1 Analogues
phenylpyrazolo[1,5-a]pyrimidine 48 (40 g, 88 mmol) in DCM (400
mL) at 0 °C was added sodium methoxide (30% in methanol, 80 g)
dropwise. The resultant mixture was stirred for 10 min at 0 °C. The
reaction was quenched by adding ice water (100 mL) and extracted
with DCM (200 mL) three times. The combined organic layers were
washed with brine (200 mL), dried over anhydrous Na2SO4, and
concentrated in vacuo. The residue was suspended in methanol (50
mL). The precipitates were collected by filtration, washed with
methanol, and dried under reduced pressure to give 5-chloro-3-
(cyclohex-1-en-1-yl)-7-methoxy-6-(4-methoxyphenyl)-2-
phenylpyrazolo[1,5-a]pyrimidine as a yellow solid (37.6 g, 96%). m/z
phenyl)-2-phenyl-N-(pyridin-2-yl)pyrazolo[1,5-a] pyrimidin-5-amine
(51) (120 mg, 0.22 mol) in 4 M HCl/1.4-dioxane (3 mL) was stirred
at rt for 16 h. The reaction mixture was basified with NaHCO3
solution to pH = 8 and filtered to afford the title compound (97 mg,
90%). m/z [M + H]+ 490.2, LC-MS: m/z 490.2 [M + H]+; 1H NMR
(400 MHz, DMSO-d6): δ 9.09 (br s, 1H), 8.20 (d, J = 4.3 Hz, 1H),
7.82 (t, J = 7.3 Hz, 1H), 7.73 (d, J = 7.3 Hz, 2H), 7.38−7.51 (m, 3H),
7.28−7.38 (m, 3H), 7.10−7.17 (m, 1H), 7.03 (d, J = 8.5 Hz, 2H),
6.06 (br s, 1H), 3.82 (s, 3H), 2.34 (br s, 2H), 2.06 (br s, 2H), 1.65−
1.79 (m, 4H).
Molecular Modeling. All computations were carried out using
the molecular operating environment (MOE).14 All graphics are
produced using PyMol.28
1
446.16 [M + H]+, LC-MS: m/z 446.1 [M + H]+; H NMR (400
MHz, DMSO-d6): δ 7.78−7.91 (m, 2H), 7.42−7.58 (m, 3H), 7.33−
7.42 (m, J = 8.9 Hz, 2H), 7.00−7.14 (m, J = 8.9 Hz, 2H), 5.83 (br s,
1H), 4.14 (s, 3H), 3.84 (s, 3H), 2.20 (d, J = 5.9 Hz, 4H), 1.61−1.77
(m, 4H).
3-(Cyclohex-1-en-1-yl)-6-(4-methoxyphenyl)-2-phenyl-5-(pyri-
din-2-ylamino)pyrazolo[1,5-a]pyrimidin-7(4H)-one (51). Step E
stoichiometry: 5-chloro-3-(cyclohex-1-en-1-yl)-7-methoxy-6-(4-me-
thoxyphenyl)-2-phenylpyrazolo[1,5-a]pyrimidine (200 mg, 0.449
mol), pyridin-2-amine 50 (63.4 mg, 0.674 mol, 1.5 equiv), Pd(OAc)2
(20.2 mg, 0.0898 mol, 0.2 equiv), Xantphos (52 mg, 0.0898 mol, 0.2
equiv) and K2CO3 (265 mg, 1.12 mol, 2.5 equiv) in dioxane (5 mL)
under heating at 120 °C for 1 h under N2 atmosphere. After workup,
160 mg (71%) was collected. m/z [M + H]+ 504.2, LC-MS: m/z
504.3 [M + H]+.
Biochemical Assays. Cloning, Protein Expression, and Purifi-
cation. MAT2A protein for biochemical assay was expressed by
recombinant baculovirus in SF9 cells using the Bac to Bac system
cloned into the pFASTBAC1 vector (Invitrogen). Recombinant
MAT2A was isolated from the cell lysate of 150 g of infected cells
using high performance Ni Sepharose column chromatography.
Recombinant MAT2A homodimer was eluted with 250 and 500
mM imidazole, and fractions containing MAT2A were identified by
SDS-PAGE and pooled.
For crystallization, full-length MAT2A was cloned into a pET21a-
based vector with N-terminal (His)6-tag and tobacco etch virus
(TEV) protease cleavage site. The protein was expressed in
Escherichia coli strain BL21 (DE3) with 0.5 mM isopropyl β-D-1-
thiogalactopyranoside at 18 °C for 16 h. The cells were harvested,
suspended in lysis buffer containing 20 mM Tris-HCl, pH 7.5, 500
mM NaCl, 5% glycerol, 1 mM tris(2-carboxyethyl) phosphine
3-(Cyclohex-1-en-1-yl)-6-(4-methoxyphenyl)-2-phenyl-5-(pyri-
din-2-ylamino)pyrazolo[1,5-a]pyrimidin-7(4H)-one (AG-270). Step
F: A solution of 3-(cyclohex-1-en-1-yl)-7-methoxy-6-(4-methoxy-
4446
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
(TCEP), and opened by sonication. The supernatant was obtained by
centrifugation and loaded on a nickel−nitrilotriacetic acid (Ni−NTA)
column (Qiagen, Cat. #30230). The column was washed with buffer
containing 20 and 50 mM imidazole in the lysis buffer prior to elution
with 250 mM imidazole. (His)6-tag was cleaved by TEV protease
(1:20 w/w ratio protease to protein) while dialyzing against the lysis
buffer at 4 °C overnight. The sample was loaded onto an Ni−NTA
column. The flow-through and eluate with 15 mM imidazole in the
lysis buffer were collected and concentrated. The protein was further
purified by HiLoad 16/600 Superdex 200 pg column (GE Healthcare
Life Sciences) in the lysis buffer, concentrated to 24 mg/mL, and flash
frozen with liquid nitrogen for storage.
Fragment Screening. Library screening was performed by
ultrafiltration as previously described11 using Centricon 10 mW
cutoff devices (EMD Millipore). MAT2A protein at a concentration
of 2 μM was incubated with mixtures of 20 compounds at 50 μM each
in 20 mM Tris-HCl, pH 7.5, and 500 mM NaCl. Relative retention
was determined after three successive rounds of concentration and
redilution by LC-MS.
SPR Analysis. MAT2A protein was immobilized to CM5 sensor
chips (GE Healthcare Life Sciences) via a covalently attached anti-His
IgG (Qiagen). The chip surface was activated by 1-ethyl-3-(3-
(dimethylamino)propyl)carbodiimide (EDC)/N-hydroxysuccinimide
(NHS) in HBS-EP+ buffer at 10 μL/min for 540 s. Penta-anti-His
IgG was diluted 1:10 in sodium acetate pH 5.0 and injected for 360 s.
Active sites were then blocked with 1 mM ethanolamine-HCl. Final
immobilization level was about ∼16 000 response units for each flow
cell. MAT2A protein was captured on the prepared surface after
dilution in running buffer (50 mM Tris pH 8.0, 150 mM NaCl, 10
mM MgCl2, 0.005% P20, 0.5 mM TCEP), to a final capture level of
about 7500 RU. Binding analysis was performed using a 100 mM
stock compound solution prepared in DMSO diluted in running
buffer.
IC50 Determinations. For determination of the inhibitory potency
of compounds against the MAT2A homodimer, protein was diluted to
1.25 μg/mL in assay buffer (50 mM Tris, pH 8.0, 50 mM KCl, 15
mM MgCl2, 0.3 mM EDTA, 0.005% [w/v] BSA). Test compound
was prepared in 100% DMSO at 50× the desired final concentration.
A 1 μL volume of compound dilution was added to 40 μL of enzyme
dilution and the mixture was allowed to equilibrate for 60 min at 25
°C. The enzymatic assay was initiated by the addition of 10 μL of
substrate mix (500 μM ATP, pH 7.0, 400 μM L-methionine in 1×
assay buffer), and the mixture was incubated for a further 60 min at 25
°C. The reaction was halted and the liberated phosphate released by
the enzyme in stoichiometric amounts by the production of SAM was
measured using the PiColorLock Gold kit (Innova Biosciences).
Absolute product amounts were determined by comparison with a
standard curve of potassium phosphate buffer, pH 8.0.
Mechanism-of-Action Studies. To determine the mechanism-of-
action12 with regard to the L-methionine substrate, reactions were
performed as described above with the modification that L-methionine
substrate concentrations were varied in the final reaction from 6.25 to
400 μM, and the ATP concentration was set at 100 μM. For the
determination of the mechanism-of-action with regard to ATP,
reactions were performed as described above, with the modification
that ATP was varied from 15 μM to 1 mM final concentration and L-
methionine was fixed at 50 μM.
Crystallization, X-ray Data Collection, Processing, Structure
Refinement, and Analysis. MAT2A·SAM complex was generated by
mixing 20 mg/mL of MAT2A with 2 mM SAM and incubating on ice
for 2 h. Crystals were obtained by hanging-drop vapor diffusion with 2
μL of MAT2A·SAM complex mixed with 1 μL of crystallization well
buffer containing 0.2 M LiCl, 0.1 M Tris-HCl, pH 7.8−8.4, 16−22%
PEG6000 and 10% ethylene glycol at 18 °C. Crystals of inhibitor
complexes were obtained by cocrystallization by mixing 20 mg/mL of
MAT2A with 2−25 mM of compound, followed by the addition of 2
mM SAM and incubation on ice for 2 h prior to crystallization as
described above. The crystals were cryoprotected in the mother liquor
with 10% DMSO and flash frozen in liquid nitrogen. The data for
MAT2A·SAM·2 crystal were collected at Advanced Photon Source
beamline 21-ID-F with a Rayonix MX300 detector. MAT2A·SAM·
AGI-24512, MAT2A·SAM·34, and MAT2A·SAM·35 crystals were
collected at the Shanghai Synchrotron Radiation Facility beamline
BL17U1 with an ADSC Quantum 315r detector; MAT2A·SAM·AG-
270 crystal was collected at the Shanghai Synchrotron Radiation
Facility beamline BL18U1 with a Pilatus3 6 M detector. All data were
processed either with HKL200029 or XDS.30 Initial phases were
obtained by performing molecular replacement with the coordinates
derived from PDB code 2P02 as a search template using Phaser31 in
CCP4 Suite. The restraints and coordinates of the compounds were
generated by eLBOW.32 Iterative model building was performed using
COOT33 and refined using REFMAC534 initially and using
PHENIX35 at the final stages. The data collection and structure
refinement statistics are summarized in Supplemental Data Table 1.
All inhibitors were modeled at an occupancy of 1.0 with the exception
of inhibitor 2, which was refined to an occupancy of 0.65. All
structures contain one molecule of the ternary complex containing
MAT2A, SAM and the inhibitor in the asymmetric unit. Dimeric
MAT2A structures used in the analysis were generated by applying
crystallographic symmetry operation. All figures representing
structures were prepared with PyMOL.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.jmedchem.0c01895.
■
sı
Cellular activity; mechanism-of-action studies; metabolic
hotspot identification; enzymatic correlations; confor-
mations; plasma concentration−time profiles; hydro-
phobicity and solubility correlations; crystallography
data collection and refinement statistics; key pharmaco-
kinetic parameters; cellular assay methods; in vivo study
methods; synthetic procedures for select compounds;
1H NMR spectra; HPLC and LC-MS chromatograms;
AG-270 structures (PDF)
Molecular formula strings (CSV)
AUTHOR INFORMATION
Corresponding Author
■
Zenon Konteatis − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States; orcid.org/0000-
0002-5421-4907; Email: zenon.konteatis@agios.com
Authors
Jeremy Travins − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Stefan Gross − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Katya Marjon − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Amelia Barnett − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Everton Mandley − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Brandon Nicolay − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Raj Nagaraja − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Yue Chen − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Yabo Sun − Viva Biotech, Shanghai 201203, China
Zhixiao Liu − Viva Biotech, Shanghai 201203, China
Jie Yu − Viva Biotech, Shanghai 201203, China
Zhixiong Ye − Viva Biotech, Shanghai 201203, China
4447
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
S. A.; Marks, K. M. MTAP deletions in cancer create vulnerability to
targeting of the MAT2A/PRMT5/RIOK1 axis. Cell Rep. 2016, 15,
574−587.
Fan Jiang − Viva Biotech, Shanghai 201203, China
Wentao Wei − Viva Biotech, Shanghai 201203, China
Cheng Fang − ChemPartner, Shanghai 201203, China
Yi Gao − ChemPartner, Shanghai 201203, China
Peter Kalev − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Marc L. Hyer − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Byron DeLaBarre − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Lei Jin − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Anil K. Padyana − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Lenny Dang − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Joshua Murtie − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
(3) Powell, E. L.; Leoni, L. M.; Canto, M. I.; Forastiere, A. A.;
Iocobuzio-Donahue, C. A.; Wang, J. S.; Maitra, A.; Montgomery, E.
Concordant loss of MTAP and p16/CDKN2A expression in
gastroesophageal carcinogenesis: evidence of homozygous deletion
in esophageal noninvasive precursor lesions and therapeutic
implications. Am. J. Surg. Pathol. 2005, 29, 1497−1504.
(4) Zhang, H.; Chen, Z. H.; Savarese, T. M. Codeletion of the genes
for p16INK4, methylthioadenosine phosphorylase, interferon-α1,
interferon-β1, and other 9p21 markers in human malignant cell
lines. Cancer Genet. Cytogenet. 1996, 86, 22−28.
(5) Kryukov, G. V.; Wilson, F. H.; Ruth, J. R.; Paulk, J.; Tsherniak,
A.; Marlow, S. E.; Vazquez, F.; Weir, B. A.; Fitzgerald, M. E.; Tanaka,
M.; Bielski, C. M.; Scott, J. M.; Dennis, C.; Cowley, G. S.; Boehm, J.
S.; Root, D. E.; Golub, T. R.; Clish, C. B.; Bradner, J. E.; Hahn, W. C.;
Garraway, L. A. MTAP deletion confers enhanced dependency on the
PRMT5 arginine methyltransferase in cancer cells. Science 2016, 351,
1214−1218.
(6) Mavrakis, K. J.; McDonald, E. R., 3rd; Schlabach, M. R.; Billy, E.;
Hoffman, G. R.; deWeck, A.; Ruddy, D. A.; Venkatesan, K.; Yu, J.;
McAllister, G.; Stump, M.; deBeaumont, R.; Ho, S.; Yue, Y.; Liu, Y.;
Yan-Neale, Y.; Yang, G.; Lin, F.; Yin, H.; Gao, H.; Kipp, D. R.; Zhao,
S.; McNamara, J. T.; Sprague, E. R.; Zheng, B.; Lin, Y.; Cho, Y. S.; Gu,
J.; Crawford, K.; Ciccone, D.; Vitari, A. C.; Lai, A.; Capka, V.; Hurov,
K.; Porter, J. A.; Tallarico, J.; Mickanin, C.; Lees, E.; Pagliarini, R.;
Keen, N.; Schmelzle, T.; Hofmann, F.; Stegmeier, F.; Sellers, W. R.
Disordered methionine metabolism in MTAP/CDKN2A-deleted
cancers leads to dependence on PRMT5. Science 2016, 351, 1208−
1213.
(7) Lombardini, J. B.; Coulter, A. W.; Talalay, P. Analogues of
methionine as substrates and inhibitors of the methionine
adenosyltransferase reaction. Deductions concerning the conforma-
tion of methionine. Mol. Pharmacol. 1970, 6, 481−499.
(8) Lombardini, J. B.; Sufrin, J. R. Chemotherapeutic potential of
methionine analogue inhibitors of tumor-derived methionine
adenosyltransferases. Biochem. Pharmacol. 1983, 32, 489−495.
(9) Sviripa, V. M.; Zhang, W.; Balia, A. G.; Tsodikov, O. V.; Nickell,
J. R.; Gizard, F.; Yu, T.; Lee, E. Y.; Dwoskin, L. P.; Liu, C.; Watt, D. S.
2′,6′-Dihalostyrylanilines, pyridines, and pyrimidines for the inhib-
ition of the catalytic subunit of methionine S-adenosyltransferase-2. J.
Med. Chem. 2014, 57, 6083−6091.
(10) Quinlan, C. L.; Kaiser, S. E.; Bolanos, B.; Nowlin, D.; Grantner,
R.; Karlicek-Bryant, S.; Feng, J. L.; Jenkinson, S.; Freeman-Cook, K.;
Dann, S. G.; Wang, X.; Wells, P. A.; Fantin, V. R.; Stewart, A. E.;
Grant, S. K. Targeting S-adenosylmethionine biosynthesis with a
novel allosteric inhibitor of Mat2A. Nat. Chem. Biol. 2017, 13, 785−
792.
(11) Shibata, S.; Zhang, Z.; Korotkov, K. V.; Delarosa, J.; Napuli, A.;
Kelley, A. M.; Mueller, N.; Ross, J.; Zucker, F. H.; Buckner, F. S.;
Merritt, E. A.; Verlinde, C. L. M. J.; Van Voorhis, W. C.; Hol, W. G. J.;
Fan, E. Screening a fragment cocktail library using ultrafiltration. Anal.
Bioanal. Chem. 2011, 401, 1585.
(12) Copeland, R. A. Evaluation of Enzyme Inhibitors in Drug
Discovery: A Guide for Medicinal Chemists and Pharmacologists., 2nd
ed.; Wiley: New York, 2013; pp 345−382.
(13) Shafqat, N.; Muniz, J. R. C.; Pilka, E. S.; Papagrigoriou, E.; von
Delft, F.; Oppermann, U.; Yue, W. W. Insight into S-adenosylme-
thionine biosynthesis from the crystal structures of the human
methionine adenosyltransferase catalytic and regulatory subunits.
Biochem. J. 2013, 452, 27−36.
Scott A. Biller − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Zhihua Sui − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Kevin M. Marks − Agios Pharmaceuticals, Inc., Cambridge,
Massachusetts 02139, United States
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.jmedchem.0c01895
Author Contributions
#Z.K. and J.T. contributed equally to this work.
Notes
The authors will release the crystallographic atomic coor-
dinates and experimental data upon article publication (PDB
codes 7KCE, 7KDA, 7KCF, 7KDB, and 7KCC).
The authors declare the following competing financial
interest(s): All Agios authors were employees and stockholders
at the time of the study.
ACKNOWLEDGMENTS
■
Agios Pharmaceuticals, Inc. funded these studies. Editorial
assistance was provided by Christine Ingleby, PhD, CMPP, of
Excel Medical Affairs, Horsham, UK, and supported by Agios
Pharmaceuticals, Inc.
ABBREVIATIONS USED
■
Eh, hepatic extraction ratio; ER, extraction ratio; HLM, human
liver microsome; KD, dissociation constant calculated for SPR
data; MAT2A, methionine adenosyltransferase 2A; MLM,
mouse liver microsome; MTA, 5′-methylthioadenosine;
MTAP, methylthioadenosine phosphorylase; Ni−NTA, nick-
el−nitrilotriacetic acid; PRMT5, protein arginine N-methyl-
transferase 5; SAM, S-adenosyl methionine; SEM, 2-
(trimethylsilyl)ethoxymethyl; SPR, surface plasmon resonance;
TEV, tobacco etch virus; TGI, tumor growth inhibition
REFERENCES
■
(1) Crespo, I.; Tao, H.; Nieto, A. B.; Rebelo, O.; Domingues, P.;
Vital, A. L.; Patino Mdel, C.; Barbosa, M.; Lopes, M. C.; Oliveira, C.
R.; Orfao, A.; Tabernero, M. D. Amplified and homozygously deleted
genes in glioblastoma: impact on gene expression levels. PLoS One
2012, 7, No. e46088.
(14) Molecular Operating Environment (MOE); Chemical Computing
Group Inc.: Montreal, QC, Canada, 2016.
(15) Murray, B.; Antonyuk, S. V.; Marina, A.; Lu, S. C.; Mato, J. M.;
Hasnain, S. S.; Rojas, A. L. Crystallography captures catalytic steps in
human methionine adenosyltransferase enzymes. Proc. Natl. Acad. Sci.
U. S. A. 2016, 113, 2104−9.
(2) Marjon, K.; Cameron, M. J.; Quang, P.; Clasquin, M. F.;
Mandley, E.; Kunii, K.; McVay, M.; Choe, S.; Kernytsky, A.; Gross, S.;
Konteatis, Z.; Murtie, J.; Blake, M. L.; Travins, J.; Dorsch, M.; Biller,
4448
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
(16) Kalev, P.; Hyer, L. M.; Gross, S.; Konteatis, Z.; Chen, C.-C.;
Fletcher, M.; Lein, M.; Aguado-Fraile, E.; Frank, V.; Barnett, A.;
Mandley, E.; Goldford, J.; Chen, Y.; Sellers, K.; Hayes, S.; Lizotte, K.;
Quang, P.; Tuncay, Y.; Clasquin, M.; Peters, R.; Weier, J.; Simone, E.;
Murtie, J.; Liu, W.; Nagaraja, R.; Dang, L.; Sui, Z.; Biller, S. A.;
Travins, J.; Marks, K. M.; Marjon, K. MAT2A inhibition blocks the
growth of MTAP-deleted cancer cells by reducing PRMT5-dependent
mRNA splicing and inducing DNA damage. Cancer Cell 2021, 39,
209−224.
(17) Kuhn, B.; Mohr, P.; Stahl, M. Intramolecular hydrogen bonding
in medicinal chemistry. J. Med. Chem. 2010, 53, 2601−2611.
(18) Nielsen, D. S.; Shepherd, N. E.; Xu, W.; Lucke, A. J.; Stoermer,
M. J.; Fairlie, D. P. Orally absorbed cyclic peptides. Chem. Rev. 2017,
117, 8094−8128.
(19) Giordanetto, F.; Tyrchan, C.; Ulander, J. Intramolecular
hydrogen bond expectations in medicinal chemistry. ACS Med. Chem.
Lett. 2017, 8, 139−142.
(20) Leuner, C.; Dressman, J. Improving drug solubility for oral
delivery using solid dispersions. Eur. J. Pharm. Biopharm. 2000, 50,
47−60.
(21) Chan-Penebre, E.; Kuplast, K. G.; Majer, C. R.; Boriack-Sjodin,
P. A.; Wigle, T. J.; Johnston, L. D.; Rioux, N.; Munchhof, M. J.; Jin,
L.; Jacques, S. L.; West, K. A.; Lingaraj, T.; Stickland, K.; Ribich, S. A.;
Raimondi, A.; Scott, M. P.; Waters, N. J.; Pollock, R. M.; Smith, J. J.;
Barbash, O.; Pappalardi, M.; Ho, T. F.; Nurse, K.; Oza, K. P.;
Gallagher, K. T.; Kruger, R.; Moyer, M. P.; Copeland, R. A.;
Chesworth, R.; Duncan, K. W. A selective inhibitor of PRMT5 with in
vivo and in vitro potency in MCL models. Nat. Chem. Biol. 2015, 11,
432−437.
(31) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M.
D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software. J. Appl.
Crystallogr. 2007, 40, 658−674.
(32) Moriarty, N. W.; Grosse-Kunstleve, R. W.; Adams, P. D.
electronic Ligand Builder and Optimization Workbench (eLBOW): a
tool for ligand coordinate and restraint generation. Acta Crystallogr.,
Sect. D: Biol. Crystallogr. 2009, 65, 1074−1080.
(33) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features
and development of Coot. Acta Crystallogr., Sect. D: Biol. Crystallogr.
2010, 66, 486−501.
(34) Murshudov, G. N.; Skubák, P.; Lebedev, A. A.; Pannu, N. S.;
Steiner, R. A.; Nicholls, R. A.; Winn, M. D.; Long, F.; Vagin, A. A.
REFMAC5 for the refinement of macromolecular crystal structures.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2011, 67, 355−367.
(35) Adams, P. D.; Afonine, P. V.; Bunkóczi, G.; Chen, V. B.; Davis,
I. W.; Echols, N.; Headd, J. J.; Hung, L. W.; Kapral, G. J.; Grosse-
Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.
J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P.
H. PHENIX: a comprehensive Python-based system for macro-
molecular structure solution. Acta Crystallogr., Sect. D: Biol. Crystallogr.
2010, 66, 213−221.
(22) Fedoriw, A.; Rajapurkar, S. R.; O’Brien, S.; Gerhart, S. V.;
Mitchell, L. H.; Adams, N. D.; Rioux, N.; Lingaraj, T.; Ribich, S. A.;
Pappalardi, M. B.; Shah, N.; Laraio, J.; Liu, Y.; Butticello, M.;
Carpenter, C. L.; Creasy, C.; Korenchuk, S.; McCabe, M. T.;
McHugh, C. F.; Nagarajan, R.; Wagner, C.; Zappacosta, F.; Annan, R.;
Concha, N. O.; Thomas, R. A.; Hart, T. K.; Smith, J. J.; Copeland, R.
A.; Moyer, M. P.; Campbell, J.; Stickland, K.; Mills, J.; Jacques-
O’Hagan, S.; Allain, C.; Johnston, D.; Raimondi, A.; Porter Scott, M.;
Waters, N.; Swinger, K.; Boriack-Sjodin, A.; Riera, T.; Shapiro, G.;
Chesworth, R.; Prinjha, R. K.; Kruger, R. G.; Barbash, O.;
Mohammad, H. P. Anti-tumor activity of the type I PRMT inhibitor,
GSK3368715, synergizes with PRMT5 inhibition through MTAP loss.
Cancer Cell 2019, 36, 100−114.
(23) Bezzi, M.; Teo, S. X.; Muller, J.; Mok, W. C.; Sahu, S. K.;
Vardy, L. A.; Bonday, Z. Q.; Guccione, E. Regulation of constitutive
and alternative splicing by PRMT5 reveals a role for Mdm4 pre-
mRNA in sensing defects in the spliceosomal machinery. Genes Dev.
2013, 27, 1903−1916.
(24) Liu, F.; Cheng, G.; Hamard, P.-J.; Greenblatt, S.; Wang, L.;
Man, N.; Perna, F.; Xu, H.; Tadi, M.; Luciani, L.; Nimer, S. D.
Arginine methyltransferase PRMT5 is essential for sustaining normal
adult hematopoiesis. J. Clin. Invest. 2015, 125, 3532−3544.
(25) Tee, W.-W.; Pardo, M.; Theunissen, T. W.; Yu, L.; Choudhary,
J. S.; Hajkova, P.; Surani, M. A. Prmt5 is essential for early mouse
development and acts in the cytoplasm to maintain ES cell
pluripotency. Genes Dev. 2010, 24, 2772−2777.
(26) Konteatis, Z. D.; S, Z.; Travins, J. M.; Ye, Z. Pyrazolo[1,5-
a]pyrimidin-7(4H)-ones as Inhibitors of Cellular Metabolic Processes
and Their Preparation. Patent US 20180079753 A1 US 20180322,
2018.
(27) Konteatis, Z. D.; S, Z.; Travins, J. M.; Ye, Z. Preparation of
Pyrazolo[1,5-a]pyrimidin-7(4H)-ones as Methionine Adenosyltrans-
ferase 2A Inhibitors for Treatment of Cancer. Patent WO
2018045071 A1 WO 20180308, 2018.
(28) The PyMOL Molecular Graphics System, Version 2.0;
̈
Schrodinger LLC: New York, NY.
(29) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data
collected in oscillation mode. Methods Enzymol. 1997, 276, 307−326.
(30) Kabsch, W. XDS. Acta Crystallogr., Sect. D: Biol. Crystallogr.
2010, 66, 125−132.
4449
https://doi.org/10.1021/acs.jmedchem.0c01895
J. Med. Chem. 2021, 64, 4430−4449