Discovery of PF-06835919: A Potent Inhibitor of Ketohexokinase (KHK) for the Treatment of Metabolic Disorders Driven by the Overconsumption of Fructose

Article abstract of DOI:10.1021/acs.jmedchem.0c00944

Increased fructose consumption and its subsequent metabolism have been implicated in metabolic disorders such as nonalcoholic fatty liver disease and steatohepatitis (NAFLD/NASH) and insulin resistance. Ketohexokinase (KHK) converts fructose to fructose-1

Full text of DOI:10.1021/acs.jmedchem.0c00944

Subscriber access provided by University of Massachusetts Amherst Libraries
Drug Annotation
Discovery of PF-06835919: A Potent Inhibitor of Ketohexokinase (KHK) for the
Treatment of Metabolic Disorders Driven by the Overconsumption of Fructose
Kentaro Futatsugi, Aaron C. Smith, Meihua Tu, Brian Raymer, Kay Ahn, Steven B. Coffey,
Matthew S. Dowling, Dilinie P. Fernando, Jemy A Gutierrez, Kim Huard, Jayasankar Jasti, Amit
S. Kalgutkar, John D. Knafels, Jayvardhan Pandit, Kevin D Parris, Sylvie Perez, Jeffrey A.
Pfefferkorn, David A Price, Tim Ryder, Andre Shavnya, Ingrid A Stock, andy s. tsai, Gregory J Tesz,
Benjamin A. Thuma, Yan Weng, Hanna Wisniewska, Gang Xing, Jun Zhou, and Thomas V Magee
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00944 • Publication Date (Web): 10 Sep 2020
Downloaded from pubs.acs.org on September 12, 2020
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 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 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.
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 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of PF-06835919: A Potent Inhibitor of Ketohexokinase (KHK) for the Treatment
of Metabolic Disorders Driven by the Overconsumption of Fructose
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
◊
║
◊
◊
┴
Kentaro Futatsugi, * Aaron C. Smith, * Meihua Tu, Brian Raymer, Kay Ahn, Steven B.
║
║
║
┴
◊
Coffey, Matthew S. Dowling, Dilinie Fernando, Jemy A. Gutierrez, Kim Huard, Jayasankar
║
◊
║
║
║
┴
Jasti, Amit S. Kalgutkar, John Knafels, Jayvardhan Pandit, Kevin D. Parris, Sylvie Perez,
┴
◊
║
║
║
Jeffrey A. Pfefferkorn, David A. Price, Tim Ryder, Andre Shavnya, Ingrid A. Stock, Andy
║
┴
║
◊
║
┴
Tsai, Gregory J. Tesz, Benjamin Thuma, Yan Weng, Hanna Wisniewska, Gang Xing, Jun
‡
◊
Zhou, and Thomas V. Magee
◊Pfizer Inc. Medicine Design, 1 Portland Street, Cambridge, Massachusetts USA 02139
┴Pfizer Inc. Internal Medicine Research Unit, 1 Portland Street, Cambridge,
Massachusetts USA 02139
║
Pfizer Inc. Medicine Design, Eastern Point Road, Groton, Connecticut USA 06340
‡Pfizer Inc. Drug Safety R&D, Eastern Point Road, Groton, Connecticut USA 06340
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Key words: Ketohexokinase, Fructokinase, Structure-based drug design, Fructose
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
metabolism, Hepatic steatosis, Non-alcoholic steatohepatitis
ACS Paragon Plus Environment

Page 3 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT: Increased fructose consumption and its subsequent metabolism have been
implicated in metabolic disorders such as non-alcoholic fatty liver disease and steatohepatitis
(NAFLD/NASH) and insulin resistance. Ketohexokinase (KHK) converts fructose to fructose-1-
phosphate (F1P) in the first step of the metabolic cascade. Herein we report the discovery of a
first-in-class KHK inhibitor, PF-06835919 (8), currently in phase 2 clinical trials. The discovery
of 8 was built upon our originally reported, fragment-derived lead 1 and the recognition of an
alternative, rotated binding mode upon changing the ribose-pocket binding moiety from a
pyrrolidinyl to an azetidinyl ring system. This new binding mode enabled efficient exploration of
the vector directed at the Arg-108 residue, leading to the identification of highly potent 3-
azabicyclo[3.1.0]hexane acetic acid-based KHK inhibitors by combined use of parallel medicinal
chemistry and structure-based drug design.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
The utilization of the simple sugar fructose as a common food additive and sweetener
has increased in recent decades. As a result, fructose is now a common sweetening
ingredient in many foods, usually in the form of high-fructose corn syrup (HFCS).
Fructose metabolism – unlike its isomer glucose – is not subject to feedback regulation
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
and excess fructose consumption promotes metabolic disease in humans. Specifically,
fructose intake has been associated with nonalcoholic fatty liver disease (NAFLD),2
3
4
insulin resistance, and cardiovascular disease (CVD). In support of the concept that
fructose may promote deleterious metabolic effects, when dietary fructose was replaced
with starch (a glucose polymer), steatosis and insulin resistance were reversed in
human subjects.5
Ketohexokinase (KHK) initiates the metabolic cascade by phosphorylation of the 1-
hydroxyl position of fructose utilizing adenosine triphosphate (ATP) as a cofactor,
generating fructose-1-phosphate (F1P) and adenosine diphosphate (ADP).6 Two
6b
isoforms of KHK exist, KHK-A and KHK-C. KHK-C is the primary enzyme for fructose
ACS Paragon Plus Environment

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
metabolism as it has both a lower K and a higher V
than KHK-A. Upon oral
M
max
ingestion, fructose is primarily metabolized by the intestines, followed by the liver,
pancreas and kidney, all tissues expressing KHK-C.7-8 Inhibition of KHK has been
considered a promising therapeutic hypothesis for the treatment of metabolic diseases
supported by studies in KHK-null mice demonstrating protection from fructose induced
hyperlipidemia, insulin resistance, obesity and NAFLD.7, 9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
To date, three structurally-distinct small molecule KHK inhibitors have been reported in
literature.10 A previous report from Pfizer disclosed compound 1 discovered through a
combination of fragment-based screening, structure-based drug design (SBDD) and
parallel medicinal chemistry (PMC).10c Compound 1 demonstrated several positive
attributes of a lead (Figure 1A), including low metabolic turnover in human liver
microsomes (HLM), high aqueous solubility, and minimal off-target activity. The
favorable pharmacokinetic profile of 1 in rats enabled preclinical in vivo pharmacology
studies, which demonstrated robust target engagement as determined by a dose-
dependent decrease in F1P formation in liver and kidney with concomitant increases in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
plasma fructose concentrations. As previously described, compound 1 was shown to
bind at the ATP site of KHK with four notable interactions within this binding site (Figure
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1B): (1) the (S)-3-methylpyrrolidin-3-ol group occupancy of the ATP-ribose pocket with
the methyl group filling a small but crucial sub-pocket defined by Phe-260 (normally
filled by the methylene of the ATP ribose sidechain); (2) a hydrogen-bonding interaction
between the nitrile group and a conserved water that forms a hydrogen-bond with the
backbone NH of Phe-245 and the backbone CO of Cys-282; (3) the trifluoromethyl
group filling the hydrophobic pocket defined by the Pro-246-247 residues and (4) the
(3S,4S)-pyrrolidine-3,4-diol hydrogen bonding interactions between the hydroxyls and
Glu-227 and Asn-107 – thus polar interactions at the solvent exposed edge opening of
the ATP pocket. Our goal for a clinical candidate was to increase the potency
significantly by taking full advantage of these elements of the binding pocket and
possibly adding additional interactions, all while retaining the advantageous clearance
and permeability properties of 1. Herein we report the discovery of a first-in-class clinical
candidate PF-06835919 (8), a potent inhibitor of KHK demonstrating >40-fold
improvement of potency over the lead (1).
ACS Paragon Plus Environment

Page 7 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
B
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OH
N
N
N
F
F
N
OH
F
OH
1
hKHK IC50 = 0.45 M
HLM Clint,app < 13 L/min/mg
Papp = 3.5x10 cm/s
-6
LogD = 2.2
Figure 1. (A) In vitro profile of 1. (B) Co-crystal structure of 1 in human KHK (PDB code
5WBZ) highlighting four different interaction sub-regions in the ATP pocket.
RESULTS AND DISCUSSION
Our initial observation was that the (3S,4S)-pyrrolidine-3,4-diol (trans) substituent could be
converted to the achiral cis analog 2 without compromising potency or changing the binding
mode. This change simplified subsequent analog synthesis, thus 2 was used as a template for
optimizing other changes in the molecule – the ribose binding element in particular. Limited
examination to modify the trifluoromethyl group of 2 and its related system did not lead to
ACS Paragon Plus Environment

Page 9 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potency improvement among tested (data not shown). Hence, the trifluoromethyl group was kept
unchanged during the course of potency optimization reported here.
Structurally diverse primary and secondary amine monomers were chosen for replacement of the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(S)-3-methylpyrrolidin-3-ol – with emphasis on ring structures with varying methyl
substituents using standard S Ar chemistry. Docking-based calculations along with estimated
N
physicochemical properties of the target molecules were also factored into the choice of amines
for coupling. This effort led to the rapid identification of 3, a (2S,3R)-2-methylazetidin-3-ol, with
comparable potency relative to 2 (Figure 2, A).
Significantly, X-ray crystallographic analysis after soaking 3 into KHK crystals revealed a
rotated binding mode relative to 2 (Figure 2, B). The azetidinyl group of 3 was positioned
deeper in the ribose pocket while maintaining the hydrophobic interaction of the methyl
substituent in the Phe-260 sub-pocket. The hydroxyl group of the azetidine ring of 3 formed a
hydrogen-bonding interaction with the backbone CO of Ala-226. The trifluoromethyl group was
bound in the Pro-246-247 hydrophobic region, and thus unchanged relative to the previous
poses. Notably, the conserved water molecule was displaced by the rotated binding mode such
that the nitrile of 3 bound directly to the backbone NH of Phe-245. In addition, the rotation
placed the vector of the cis-dihydroxypyrrolidine substituent in line with the Arg-108 residue at
the polar back portion of the ATP binding site normally occupied by the triphosphate sidechain
in KHK.
We sought to exploit this vector in the direction of Arg-108 in an effort to improve potency,
11
12
possibly utilizing a Coulombic interaction. Additionally, a HydroSite calculation predicted a
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
high-energy virtual ‘water’ molecule in this pocket (G, H and ‒TS are 3.04, 0.64 and 2.4
kcal/mol, respectively) present within the hydrogen-bonding distance between Arg-108, a
backbone carbonyl oxygen atom of Thr-253 and a backbone NH of Gly-257 respectively,
suggesting the possibility of making energetically favorable interactions with an appropriately
chosen sidechain attachment in this region (Figure 2, C).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OH
N
OH
A
N
N
N
N
N
F
F
F
F
N
N
OH
OH
F
F
OH
OH
2
3
hKHK IC50 = 0.41 M
hKHK IC50 = 0.47 M
B
C
Figure 2. (A) Genesis of the azetidine-based KHK inhibitor 3. (B) Compound 3 (shown in
green) in the KHK crystal structure (PDB code 6W0W). For comparison, compound 2 (shown in
dark gray) was overlaid (PDB code 6W0N). (C) Detailed view of compound 3 (shown in green)
in the KHK crystal structure. A sphere represents a simulated water by HydroSite analysis.
ACS Paragon Plus Environment

Page 11 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
By employing previously described sequential S Ar reactions, a diverse array of primary- and
N
secondary amines possessing polar functional groups were rapidly examined through PMC after
1
3
both calculated property- and docking-based prioritization. The initial exploration of this vector
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
culminated in the identification of the pivotal compound 4 possessing 3-
a
azabicyclo[3.1.0]hexane ring system as one of the most potent neutral azetidine analogues
among those tested from this PMC effort (Figure 3, A).
A
B
OH
N
OH
N
N
N
N
N
F
F
F
F
H
N
N
OH
F
F
OH
OH
H
3
4
hKHK IC50 = 0.47 M
hKHK IC50 = 0.14 M
Figure 3. (A) Genesis of 4 bearing the 3-azabicyclo[3.1.0]hexane group. (B) The binding mode
of compound 4 (shown in green; PDB code 6W0X). Arg-108 and Thr-253 are highlighted.
The crystal structure of 4 in human KHK (Figure 3, B) showed that the terminal hydroxyl group
on the newly introduced 3-azabicyclo[3.1.0]hexane ring system effectively reached proximally to
Arg-108. This terminal hydroxyl group of 4 indirectly forms a hydrogen-bonding interaction
with Arg-108 through a bound water, while within the hydrogen-bonding distance with the
backbone carbonyl oxygen atom of Thr-253 respectively. Relative binding poses between the
progenitor analog 3 and 4 are otherwise largely comparable suggesting that the rest of the key
protein-ligand interactions identified in 3 were maintained for 4. Taken together, the
conformationally rigid 3-azabicyclo[3.1.0]hexane ring of 4 was considered as a more efficient
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
spacer to reach the targeted buried area nearby Arg-108 relative to the monocyclic pyrrolidine-
based ring system.
With compound 4 in hand, further optimization of the terminal polar groups on the 3-
azabicyclo[3.1.0]hexane ring system of 4 was targeted as the next design strategy to improve
potency. Among the polar groups known to interact with the arginine residue, the carboxylic acid
group was considered as one of the top candidates given its potential to form a strong ionic
interaction with the guanidinium group of Arg-108.14
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Carboxylic acid analogues of 4 with various spacer length.
Compound
4
5
6
OH
N
OH
N
OH
N
N
N
N
N
N
N
F
F
H
F
F
H
F
F
H
N
N
N
F
F
F
OH
CO H
2
CO H
2
H
H
H
a
0.14 (0.027-0.72)b
hKHK IC (M)
0.39 (0.27 – 0.58)
0.006 (0.005 – 0.006)
5
0
LogDc
3.3
0.6
1.0
aHuman KHK potency reported as the geometric mean of at least 3 replicates. Numbers in parenthesis are
b
c
9
5% confidence interval. The geometric mean of two replicates. The experimental logD using shake-
flask method at pH=7.4.
With this in mind, the spacer length between the 3-azabicyclo[3.1.0]hexane ring and the terminal
carboxylic acid group was varied to optimize the targeted interaction with Arg-108 (compounds
5 and 6, Table 1). Compound 6, with a one carbon spacer between the 3-
ACS Paragon Plus Environment

Page 13 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
azabicyclo[3.1.0]hexane ring and the terminal carboxylic acid group, demonstrated a significant
potency improvement over 4 (>20-fold potency boost). It was apparent that the optimal spacer
length between the terminal carboxylic acid group and the 3-azabicyclo[3.1.0]hexane unit was
critical in order to achieve significant potency improvement (compounds 5 vs 6, Table 1). X-ray
crystallographic analysis of the compound 6 bound in human KHK protein was instructive
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Figure 4).
Figure 4. Co-crystal structure of 6 in human KHK (PDB code 6W0Y) highlighting key
hydrogen-bonding interactions. Gly-257 and Gly-255 (backbone NH’s within the hydrogen-
bonding distances), a bound water, and Arg-108 are highlighted. The dotted lines indicate
potential hydrogen-bonding distances.
The overall binding mode of 6 was determined to be comparable to that of the neutral azetidine
analogue 4 except that additional polar interactions with the newly introduced carboxylic acid
group of compound 6 were evident from the X-ray structure. Within the proximity of the
carboxylic acid group of 6, one of the oxygen atoms of carboxylate group was well-positioned to
interact with cationic guanidinium group of Arg-108. In addition, the backbone NH of Gly-255
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and Gly-257, respectively, as well as a bound water nearby were all within the hydrogen-bonding
distance. Thus, a complex network of polar interactions through the carboxylic acid group of 6 is
likely responsible for the significant potency improvement achieved by 6, and not solely a
Coulombic interaction.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The significant potency improvement demonstrated by 6, coupled with its favorable
physicochemical property space (MW=396, logD=1, TPSA=101), prompted us to profile
compound 6 in depth as an advanced lead (Table 2). Since the potency of 6 approached the
enzyme concentration used in the initial biochemical assay ([KHK]=10 nM), a more
sensitive biochemical assay was developed using the same substrate concentrations
but with 10-fold less enzyme while monitoring the reaction over a longer period of time.
Compound 6 showed KHK IC50 of 1 nM under this new assay format. Compound 6
demonstrated moderate cell permeability and exhibited low metabolic turnover in HLM
supplemented with cytochrome P450 (CYP) cofactor NADPH. Incubation of 6 in human
hepatocytes (HHEPs), however, resulted in relatively high turnover, indicating that 6 was subject
to metabolism by a non-CYP enzyme(s). Follow-up metabolite identification studies in human
and rat hepatocytes revealed that 6 was principally subject to a uridine glucuronosyl transferase
(UGT) mediated glucuronidation on its secondary hydroxyl group (Figure S1). The carboxylic
acid moiety was ruled out as the site of glucuronidation since inclusion of hydroxylamine in the
incubation mixture did not result in the formation of the corresponding hydroxamic acid product,
which is expected from the reaction of the amine nucleophile with the acyl glucuronide
1
5
metabolite (Figure S1-2 in Supporting Information).
ACS Paragon Plus Environment

Page 15 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. Profile of the compound 6.
a
MW 396, LogD 1 , TPSA 101, pK 4.3
a
[
KHK]=10 nM
0.006 (0.005-0.006)
0.001 (0.001-0.003)c
b
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hKHK IC (M)
5
0
[KHK]=1 nM
HLM CLint,app (µL/min/mg)d
<11
HHEPs CLint,app (µL/min/million cells)e
26
-6
f
Papp (10 cm/s)
5.6
a
b
The experimental logD using shake-flask method at pH=7.4. Human KHK potency reported as the
c
geometric mean of at least 3 replicates. Numbers in parenthesis are 95% confidence interval. The
d
geometric mean of two replicates. Apparent intrinsic clearance (CLint,app) obtained from isolated human
e
f
liver microsomes. Apparent intrinsic clearance (CL
) obtained from human hepatocytes cells. A
int,app
permeability assay using low-efflux MDCKII cells.16.
Key design objective:
Reduce glucuronidation-based clearance
OH
OH
Modify the site
of metabolism
N
N
N
N
N
A
N
Reduce logD:
A = C-CN to N
F
F
H
F
H
N
N
F
F
F
CO H
CO H
2
2
H
H
6
Figure 5. Design strategy to reduce glucuronidation of 6 by UGTs.
The next phase of the optimization was centered on balancing potency, metabolic clearance
attributed to glucuronidation, and passive permeability. Metabolite identification studies
ACS Paragon Plus Environment

Page 17 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
described above were instrumental in formulating the design strategy towards reducing
glucuronidation (Figure 5). Specifically, modifications of the cyanopyridine core and the
hydroxyl azetidine group of 6 were targeted to reduce glucuronidation through reduction of logD
and/or removal of the hydroxyl group on the azetidine ring (the proposed site of
glucuronidation). The pyrimidine core was considered as a bioisostere of the cyanopyridine
group with lower logD.17 The optimization effort within the scope of this design strategy is
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
summarized in Table 3. The more sensitive biochemical assay format described above was
used for the optimization efforts presented in Table 3.
Table 3. Optimization of compound 6.
Compound
6
7
8
OH
N
OH
N
N
N
N
N
N
N
N
F
F
H
F
F
H
F
F
H
N
N
N
F
F
F
H
H
H
CO H
CO2H
2
CO H
2
a
0.001 (0.001-0.003)b 0.014 (0.010 – 0.018) 0.010 (0.008 – 0.011)
hKHK IC (M)
5
0
LogDc
1.0
0.1
<2
1.0
<3
25
HHEPs CLint,appd
26
(µL/min/million cells)
-6
e
Papp (10 cm/s)
5.6
1.4
aHuman KHK potency using 1 nM KHK protein, reported as the geometric mean of at least 3 replicates.
b
c
The geometric mean of two replicates. The experimental logD measured using the shake-flask method at
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
d
e f
pH=7.4. Apparent intrinsic clearance (CL
) measured from human hepatocytes cells. . A
int,app
permeability assay using low-efflux MDCKII cells.16.
The pyrimidine analogue 7 showed significant improvement in metabolic stability in human
hepatocytes presumably due to the reduction of lipophilicity (logD (6  7)= ‒0.9). Against
backdrops of decrease in KHK potency and permeability relative to 6, a significant improvement
in clearance prompted us to shift the focus to the pyrimidine analogue 7 with the goal to increase
permeability while maintaining low metabolic clearance in human hepatocytes. To this end, des-
hydroxyazetidine analogue 8 was targeted with the expectation to improve permeability through
reduction of hydrogen bond donor counts while concomitantly removing the site of
glucuronidation in order to mitigate an increase in clearance. Compound 8 showed comparable
KHK potency relative to 7. Gratifyingly, compound 8 maintained low metabolic turnover in
human hepatocytes while significantly improving permeability. Based upon improvements in
clearance and permeability over 6, compound 8 was selected for further in vitro characterizations
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
Table 4).
Table 4. Summary table of pharmacology and in vitro ADME data for compound 8.
a
Compound 8: MW 356, LogD 1 , TPSA 70, pK 4.4, 3.6
a
[ATP], [fructose] at Km
0.010 (0.008-0.011)
0.028 (0.024-0.031)
0.17 (0.15-0.19)
0.21(0.18-0.23)
1.2
b
Human KHK-C IC (M)
5
0
b
Human KHK-A IC (M)
[ATP]=2mM, [fructose]=8mM
5
0
b
Rat KHK IC (M)
Human hepatocytes CLint,app (µL/min/million cells)c
5
0
-6
d
Papp (10 cm/s)
25
Thermodynamic solubilitye
2
.6 mg/mL
f
fu,p (human/rat)
0.08/0.02
ACS Paragon Plus Environment

Page 19 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
b
The experimental logD measured using the shake-flask method at pH=7.4. Human or rat KHK
potency reported as the geometric mean of at least 3 replicates. Numbers in parenthesis are 95%
c
confidence interval. Apparent intrinsic clearance (CLint,app) measured from human hepatocytes
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
cells using the relay method for low clearance compounds.18 d. A permeability assay using low-
f
1
6
e
f
efflux MDCKII cells. . Solubility of crystalline 8 in FaSSIF (final pH=6.2). fraction unbound in
plasma.
(i)
(ii)
60
Saturating
Km
1
.0
.5
4
0
0
0
0
2
0.0
0.01
1
100
10000
0
20
40
60
[
PF-06835919], nM
[Human KHK-C], nM
(iii)
(iv)
100
[
PF-06835919], uM
8
0
2.5
5
1
.0
.6
2.0
1.5
1.0
60
0
0
0
.50
.16
4
0
0
0
.050
0.016
2
0
0
.0050
.0016
0.5
n
-10
-5
5
10
15
ion
1/[ATP] (mM-1
)
After dilut
Before dilutio
Figure 6. Biochemical characterizations of 8. (i) Effect of substrate concentrations on KHK-C
IC (ii) tight binding inhibition of KHK-C by 8. (iii) Modality of compound 8 inhibition. (iv)
5
0
Reversibility of compound 8 inhibition.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Selected in vitro characterization of 8 is summarized in Table 4. To evaluate potency of 8
under physiologically relevant substrate concentrations, the IC was measured at 2 mM
50
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ATP and 8 mM fructose. In this setting, the IC of 8 increased to 0.028 M from 0.01
50

M under substrate concentrations at K , suggesting a degree of competition with at
m
least one of the substrates (Table 4; Figure 6(i)). At physiological substrate
concentrations, compound 8 showed a 6-fold higher IC50 value against human KHK-A
relative to human KHK-C while inhibiting rat KHK with an IC50 of 0.21 M. The higher
IC50 value of compound 8 in KHK-A can be reasoned with a mutation of Leu-112 in
KHK-C to Tyr-112 in KHK-A. Tyr-112 in KHK-A would likely attenuate the critical Arg-
108 interaction with the carboxyl group of compound 8 through a hydrogen bond
network (Arg-108Glu-29Lys-174Tyr-112). In addition, compound 8 was highly
selective against other sugar kinases, showing >3000-fold selectivity for human KHK-C
compared to human ribokinase, glucokinase, hexokinase-1, and hexokinase-2 (Table
18
S1). By using a relay method for low clearance compounds in human hepatocytes, the
apparent intrinsic clearance of 8 was determined to be 1.1 µL/min/million cells, suggestive
of very low metabolic turnover. The thermodynamic solubility of the crystalline form of 8 was
ACS Paragon Plus Environment

Page 21 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
measured in simulated fasted intestinal fluid with bile salts (FaSSIF). Consistent with its
amphoteric nature (acidic pK =4.4, basic pK =3.4), it showed high solubility in this media.
a
a
Compound 8 showed low, but measurable fraction unbound in plasma across the species.
Detailed biochemical characterization of 8 was further conducted to reveal its inhibition
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
mechanism (Figure 6). The IC50 dependence of 8 on enzyme concentration indicates
tight binding inhibition under these assay conditions (Figure 6(ii)). Varying
concentrations of ATP, but not fructose, had a significant effect and the Lineweaver-
Burke plot (1/rate vs 1/[ATP]) reveals a mixed noncompetitive mechanism (Figure 6(iii)).
While a purely noncompetitive binding modality displays equivalent affinities for the free
enzyme and the enzyme-substrate complex, mixed-type noncompetitive inhibition
describes cases where these affinities are different.19 Despite the dual affinity of the
inhibitor for both free and substrate-bound forms of the enzyme, compound 8 prefers
binding to the free enzyme by 19-fold, and essentially behaves like a competitive
inhibitor of ATP (Figure S3). Applying the correct mathematical model for mixed
noncompetitive inhibition gave an inhibition constant (K) of 4.53 nM for 8 (Figure S4).
i
Despite tight binding inhibition of 8, there was no evidence of time-dependence, and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
rapid dilution conditions allowed measurement of a regain in activity which supports a
reversible mechanism (Figure 6(iv); Figure S5 and Table S2). Taken together,
compound 8 was determined to be a potent reversible inhibitor of recombinant human
KHK-C through a mixed noncompetitive mechanism.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. The binding mode of 8 in the crystal structure highlighting key hydrogen-bonding
interactions (PDB code 6W0Z). Backbone NHs of Phe-245, Gly-257 and Gly-255, two bound
waters, and Arg-108 are highlighted. The dotted lines indicate potential hydrogen-bonding
distances.
The co-crystal structure of 8 in human KHK is shown in Figure 7. The binding mode of 8 was
comparable to that of 6, retaining multiple polar interactions by the carboxylate group (Figure
7
). Due to the change from the cyano pyridine core to the pyrimidine core (6  8), the replaced
conserved water molecule seen in co-crystal structure of 1 in human KHK was observed in this
ACS Paragon Plus Environment

Page 23 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
binding mode with 8. Compound 8 makes a hydrogen-bonding interaction with a conserved
water through the pyrimidine nitrogen while a conserved water makes a hydrogen-bonding
interaction with the backbone NH of Phe-245. A pairwise comparison (compound 7 
compound 6, Table 3) suggests replacing a conserved water molecule likely contributed to 14-
fold improvement in potency (0.014 M  0.001 M). However, this permutation is also
associated with increase in lipophilicity (LogD 0.1  1) and metabolic clearance (HHEP Clint,app
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
<
2  26) in this matched molecular pair. From the optimization perspective, the value of
replacing this conserved water molecule is likely chemo-type dependent, and it needs to be
balanced with other design parameters beyond potency.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 5. in vivo Pharmacokinetic profile of 8 in rat and dog.
Dose
CLp
t1/2
Vdss
Cmax
tmax
(h)
AUC(0-24)
F
Species Route
(mg/kg) (mL/min/kg)
(h)
(L/kg) (ng/mL)
(ng.h/mL) (%)
41800
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
.71
(6.89, (0.18,
.52) 0.22)
.7
(5.46,
.93)
0.20
NA
NA
0
.358 (0.361,
Rat
IV
PO
IV
1
5
1
5
(42500,
41000)
210000
NA
95
0
.354)
8
5
24100
2.1
NA
NA
(34400, (0.25, (245000,
5
13700)
NA
4.0)
NA
175000)
14600 ±
6900
5.57
±
0
.165
0.04
Dog
1.31 ± 0.55
NA
±
1
.77
31900
0.5
157000
>
10
PO
NA
3.4
NA
(31300, (0.5, (137000,
2500) 0.5) 177000)
0
3
a
All experiments involving animals were conducted in our AAALAC-accredited facilities and were reviewed and
approved by Pfizer Institutional Animal Care and Use Committee. Pharmacokinetic parameters were calculated from
plasma concentration–time data and are reported as mean ± S.D. (for n=3) and mean (for n=2) with individual
values provided in parenthesis. Intravenous (IV) pharmacokinetics were conducted in male gender of each species
(Wistar Han rats and beagle dogs). Solution formulation for the IV dose included 12.5% SBECD in 50 mM TRIS
buffer (pH 8.0). The PO (oral) studies, crystalline 8 was administered as suspension in 0.5 % methylcellulose.
Based on attractive in vitro disposition attributes of 8, the in vivo pharmacokinetic properties of
8 was determined in rats and dogs, the preclinical species for safety assessments in the program
(Table 5). Following IV administration, 8 demonstrated low total plasma clearance (CL ) and
p
ACS Paragon Plus Environment

Page 25 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
low steady state distribution volume (V ) leading to moderate-to-long elimination half-lives
dss
(
t ) in rats and dogs respectively (Table 5). Moreover, 8 demonstrated negligible renal
1/2
excretion as unchanged parent in both animal species. The corresponding oral bioavailability (F)
of the crystalline form of 8 was high (> 90%) in both species, consistent with a combination of
high aqueous solubility, high passive permeability and low metabolic clearance as shown in
Table 4. With the robust in vivo pharmacokinetic profile demonstrated, compound 8 was
evaluated for KHK inhibition in vivo by utilizing Sprague Dawley rats as previously described
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
Figure 8).10c All procedures performed on these animals were in accordance with regulations
and established guidelines and were reviewed and approved by the Pfizer Institutional Animal
Care and Use Committee. Sprague Dawley rats were administered vehicle or increasing doses of
compound 8 followed by iv administration of fructose. Two minutes after the administration of
fructose, rats were euthanized and liver and kidney tissues were removed and snap-frozen in
liquid nitrogen. F1P was quantified in liver and kidney homogenates by liquid
chromatography/mass spectrometry (LC/MS) as previously described.10c F1P decreased dose
dependently with measured ED values of 1.93 mg/kg in liver and 4.74 mg/kg in kidney (Figure
5
0
8
). The observed dose-dependent decrease in hepatic and kidney F1P as depicted in Figure 8
indicates inhibition of liver and kidney KHK.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
000
Liver
6000
4000
Kidney
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
000
0
0.1
1
10
100
Dose Compound 8 (mg/kg)
Liver Kidney
ED50
1.93
0.29
4.74
0.68
pED50 ± SE
Figure 8. Hepatic and renal concentration of fructose 1-phosphate (F1P) following oral
administrations of 8 and i.v. fructose bolus in Sprague Dawley rats. A nonlinear dose–response
curve was fitted to a variable slope (four parameter) using GraphPad Prism, and the calculated
ED values for kidney and liver F1P reduction are reported. Data are presented as mean values
5
0
(
±standard error) with n=7–8 per data points.
Compound 8 was further profiled in a battery of in vitro assays to determine its potential as a
development candidate. No measurable reversible and/or time-dependent inhibition of
cytochrome P450 enzymes (IC50 >100 M against CYP1A2, 2B6, 2C8, 2C9, 2C19, 2D6 and
3A4) and no induction of CYP1A2, 2B6 and 3A4 at 30 M was observed. Compound 8
exhibited a clean off-target selectivity profile as assessed by the CEREP panel (<53% at 30 M).
No significant inhibitions were observed on hERG (human ether-a-go-go-related gene, patch
clamp IC >100 µM), various kinases in the kinase selectivity panel, and PDE selectivity panel
5
0
(
the lowest IC =11 M against PDE10A, Table S3). Compound 8 did not present genetic
5
0
toxicology risks as evidenced by negative findings in an Ames test and an in vitro micronucleus
ACS Paragon Plus Environment

Page 27 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
assay in TK6 cells. Furthermore, in vivo exploratory toxicology studies in rats and dogs revealed
no development-limiting toxicity findings. On the basis of its favorable overall profile,
compound 8 (PF-06835919) was selected as a clinical candidate.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
SYNTHESIS
Scheme 1. General Synthesis of Pyridines 2-6a
R
R
N
Cl
N
N
N
Cl
N
a
N
b or c
N
F
F
F
F
N
N
F
F
OH
OH
Cl
F
F
F
9
10
2-3
OH
OH
OH
Cl
N
1
4, R = CH OH
2
N
N
1
5
6
, R = CO H
d,e, or f
2
N
N
1
, R = CH CO H
2
2
F
F
F
F
H
Cl
N
F
F
9
R1
H
a.
Reagents and conditions: (a) 11, Et N, DMF; (b) 12, K CO , DMF, 60 °C; (c) i. rac-13, Et N, DMF, 60
3
2
3
3
°
C; ii. SFC separation; (d) i. 14, K CO , NaHCO , 85%; ii. 13/(R)-CSA, i-Pr NEt, CH CN, 85 °C, 53%;
2 3 3 2 3
(e) i. 15, Et N, DMF; ii. 13·HCl, Et N, CH CN, 60 °C; iii. 2 M NaOH, EtOH; (f) i. 16, NaHCO , EtOH;
3
3
3
3
ii. 13/(R)-CSA, NaHCO , EtOH, 80 °C; iii. 1 M NaOH, EtOH, 44% over three steps;
3
The chemistry to make the described pyridine analogs is shown in Scheme 1. Amines described
in Schemes 2–4 were employed in sequential S Ar chemistry beginning with commercially-
N
available dichloropyridine 9. The first S Ar reaction proceeds at room temperature to install the
N
6-position substituent selectively. The second S Ar reaction proceeds at elevated temperatures to
N
install the 2-position substituent leading to pyridines 2–4. For ester-containing intermediates, a
hydrolysis reaction was used in step 3 to form pyridine acids 5 and 6.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 2. Precedented amines used in S Ar chemistry
N
HO
H
H
HN
OH
HN
HN
CO Me
2
N
H
N
OH
OH
11
H
H
H
14
15
17ref.21
bis-L-dihydroorotic
acid salt
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(12)ref.10c
Scheme 3. Synthesis of Azetidine 13a
OH
OH
Br
Br
a
b
c
e
HO
HO
N
R
N
R
O
Br
rac-19
18
rac-20
R = CHPh , rac-21 R = CHPh , 21/(R)-CSA
2
2
d
f
R = H, rac-13 or
R = H, 13/(R)-CSA
13·HCl
a
Reagents and Conditions: (a) Br , CHCl , 0 °C to 15 °C; (b) aq. KOH, THF, 15 °C to rt; (c) Ph CHNH ,
2
3
2
2
NaHCO , EtOH, rt to 80 °C; 37% over three steps; (d) Pd(OH) /C, H (55 psi), EtOH, quant.; or i. SFC
3
2
2
separation ii.4 M HCl in dioxane, EtOH, Pd(OH) /C, H (50 psi), 89%; (e) i. (R)-CSA, EtOH; ii. MTBE
2
2
crystallization; iii. CH Cl , EtOAc recrystallization; 32% over two steps; (f) Pd(OH) /C, H (60 psi),
2
2
2
2
MeOH, 95%.
The synthesis of azetidine 13 is shown in Scheme 3. Following modifications of literature
conditions,20,21 (2E)-buten-1-ol 18 was treated with bromine to from dibromide 19 which was
then reacted with potassium hydroxide to form epibromohydrin 20. Treatment with benzhydryl
amine led to displacement of the bromide and concomitant epoxide opening to form protected
azetidine rac-21, which could be separated via SFC on a chiral column or taken forward to
hydrogenolysis to remove the benzhydryl protecting group and provide rac-13. Treatment of
rac-21 with (R)-camphor sulfonic acid (CSA) under classical chiral resolution conditions led to
the isolation of the single enantiomer of 21 as a salt with (R)-CSA. Hydrogenolysis of the
ACS Paragon Plus Environment

Page 29 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
protecting group then led to the isolation of the 13/(R)-CSA salt that could be used directly in the
SNAr chemistry.
Scheme 4. Synthesis of Amino esters 16 and 26a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
O
H
H
H
H
OH a or b
c or d
OR1
BnN
BnN
HN
H
H
1
22
23
R =Et, 16
1
R =Me, 26·HCl
a
Reagents and Conditions: (a) i. Et N, MsCl, DMF, THF, 0–10 °C; ii. NaCN, DMF, 73%, over two steps;
3
(
b) i. HCl, MeOH; ii. SOCl , PhCH , 91%; iii. NaCN, DMF, 93%; (c) i. H SO , EtOH, 80 °C, 82%; ii.
2
3
2
4
Pd(OH) /C, H (50 psi), 50 °C, 91%; (d) i.AcCl, MeOH, 0–65 °C, 82%; ii. HCl, MeOH, 88%; iii.
2
2
Pd(OH) , H (100 psi), MeOH, 70 °C, 99%.
2
2
The synthesis of amino ethyl esters 16 and 26 are shown in Scheme 4. Starting from benzyl
protected amino alcohol 22,22 activation of the alcohol as the mesylate and S 2 displacement
N
with sodium cyanide led to formation of nitrile 23. Acid-mediated solvolysis led to formation of
the ester and hydrogenolysis of the benzyl protecting group led to amino ester 16. Methyl ester
2
6 was employed in larger scale syntheses of 8 since it could be stored as a stable solid
intermediate in contrast to ethyl ester 16 which was an oil that partially decomposed upon long-
term storage. Improvements were also made during homologation step: (1) conversion of the
alcohol to the chloride with thionyl chloride and formation of the HCl salt led to a more stable
intermediate than the analogous mesylate and (2) after free-basing the salt, the chloride was
displaced with sodium cyanide and the HCl salt of 23 was formed. Solvolysis with methanol
rather than ethanol also led to a solid amino ester HCl salt rather than an oil. Benzyl deprotection
then led to amino methyl ester hydrochloride 26 that was used directly in the S Ar chemistry.
N
ACS Paragon Plus Environment