The discovery and characterization of the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor potentiator N-{(3 S,4 S)-4-[4-(5-cyano-2-thienyl)phenoxy]tetrahydrofuran-3-yl}propane-2-sulfonamide (PF-04958242)

Article abstract of DOI:10.1021/acs.jmedchem.5b00300

A unique tetrahydrofuran ether class of highly potent α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor potentiators has been identified using rational and structure-based drug design. An acyclic lead compound, containing an ether-linked isopr

Full text of DOI:10.1021/acs.jmedchem.5b00300

Subscriber access provided by OAKLAND UNIV
Article
The Discovery and Characterization of the #-Amino-3-
hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)
Receptor Potentiator N-{(3S,4S)-4-[4-(5-cyano-2-
thienyl)phenoxy]tetrahydrofuran-3-yl}propane-2-sulfonamide (PF-04958242)
Christopher L. Shaffer, Nandini C. Patel, Jacob Bradley Schwarz, Renato J Scialis, Yunjing Wei,
Xinjun J Hou, Longfei Xie, Kapil Karki, Dianne Bryce, Sarah M Osgood, William E. Hoffmann,
John T Lazzaro, Cheng Chang, Dina F McGinnis, Susan M Lotarski, Jianhua Liu, Ronald
Obach, Mark L Weber, Laigao M Chen, Kenneth R Zasadny, Patricia A Seymour, Christopher
Schmidt, Mihaly Hajós, Raymond S Hurst, Jayvardhan Pandit, and Christopher J. O'Donnell
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00300 • Publication Date (Web): 23 Apr 2015
Downloaded from http://pubs.acs.org on May 3, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The Discovery and Characterization of the α-Amino-3-hydroxy-5-methyl-4-
isoxazolepropionic acid (AMPA) Receptor Potentiator N-{(3S,4S)-4-[4-(5-cyano-2-
thienyl)phenoxy]tetrahydrofuran-3-yl}propane-2-sulfonamide (PF-04958242)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
,†,‡
§
§
†
§
Christopher L. Shaffer* , Nandini C. Patel , Jacob Schwarz , Renato J. Scialis , Yunjing Wei ,
§
§
§
‡
†
Xinjun J. Hou , Longfei Xie , Kapil Karki , Dianne K. Bryce , Sarah M. Osgood , William E.
‡
†
†
‡
‡
Hoffmann , John T. Lazzaro , Cheng Chang , Dina F. McGinnis , Susan M. Lotarski , JianHua
†
†
‡
║
║
Liu , R. Scott Obach , Mark L. Weber , Laigao Chen , Kenneth R. Zasadny , Patricia A.
‡
‡
‡
‡
§
Seymour , Christopher J. Schmidt , Mihály Hajós , Raymond S. Hurst , Jayvardhan Pandit , and
,
§
Christopher J. O’Donnell*
†
‡
Department of Pharmacokinetics, Dynamics and Metabolism, Neuroscience Research Unit,
§
║
Worldwide Medicinal Chemistry and BioImaging Center, Worldwide Research &
Development, Groton Laboratories, Pfizer Inc., Groton, CT 06340, USA
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 75
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
ABSTRACT
A unique tetrahydrofuran ether class of highly potent αꢀaminoꢀ3ꢀhydroxyꢀ5ꢀmethylꢀ4ꢀ
isoxazolepropionic acid receptor potentiators has been identified using rational and structureꢀ
based drug design. An acyclic lead compound, containing an etherꢀlinked isopropylsulfonamide
and biphenyl group, was pharmacologically augmented by converting it to a conformationally
constrained tetrahydrofuran to improve key interactions with the human GluA2 ligandꢀbinding
domain. Subsequent replacement of the distal phenyl motif with 2ꢀcyanothiophene to enhance
its potency, selectivity and metabolic stability afforded Nꢀ{(3S,4S)ꢀ4ꢀ[4ꢀ(5ꢀcyanoꢀ2ꢀ
thienyl)phenoxy]tetrahydrofuranꢀ3ꢀyl}propaneꢀ2ꢀsulfonamide (PFꢀ04958242, 3), whose
preclinical characterization suggests an adequate therapeutic index, aided by low projected
human oral pharmacokinetic variability, for clinical studies exploring its ability to attenuate
cognitive deficits in patients with schizophrenia.
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
ACS Paragon Plus Environment

Page 3 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
αꢀAminoꢀ3ꢀhydroxyꢀ5ꢀmethylꢀ4ꢀisoxazolepropionic acid (AMPA) receptors (AMPAR)
mediate fast glutamatergic excitatory neurotransmission in the central nervous system, and
alterations in their density and subunit composition are involved in dynamically regulating
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
synaptic function. Specifically, AMPAR opening depolarizes neuronal membranes to relieve
2+
the Mg block of colocalized NꢀmethylꢀDꢀaspartate receptors (NMDAR), which increases
2
+
2
NMDARꢀmediated Ca gating that ultimately generates changes in the synaptic morphology
1
a, 3
4
and function
thought to underlie learning and memory. A key mechanism of NMDARꢀ
5
regulated increases in synaptic activity is AMPAR insertion into the synapse. Thus, enhanced
6
AMPAR activation by a positive allosteric modulator (“potentiator”) would be expected to
augment NMDARꢀinduced synaptic potentiation. Accordingly, AMPAR stimulation is an
7
attractive mechanism for treating various neurological and psychiatric diseases , particularly
8
certain cognitive impairments in schizophrenia since they are hypothesized to primarily result
9
from dysfunction in NMDAR glutamatergic neurotransmission. This theory is supported by
9
c, 10
clinical studies
in which healthy volunteers receiving acute subanesthetic doses of
nonselective NMDAR antagonists (e.g. ketamine) present with schizophreniaꢀlike symptoms
including memory deficits. AMPAR potentiators have demonstrated procognitive properties in
6
a, 7a, 11
numerous animal models of hippocampal/cortical function and working memory
and in
1
2
small clinical trials. However, excessive AMPAR excitation, via orthosteric or allosteric
13
14
molecular interactions, can cause motor coordination disruptions , neurotoxicity and/or
1
3, 15
convulsion.
Since the effects of variations in AMPAR subtypes (GluA1–4), splice forms
(
flip (i) and flop (o)) and binding partners on an AMPAR potentiator’s therapeutic index (TI) are
unknown, it is essential to understand potentiator exposures causing desired versus harmful
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 75
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
effects to identify a safe molecule for clinical testing. Thus, unbound brain compound
1
6
concentration (C ), a highꢀconfidence surrogate of interstitial fluid drug concentrations that
b,u
determine AMPAR ligandꢀbinding domain (LBD) interactions, was deemed the most appropriate
exposure value for defining this mechanismꢀbased TI.
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
Based on these concepts, we previously reported our strategy for identifying and profiling the
two structurally unique AMPAR potentiators Nꢀ{(3R,4S)ꢀ3ꢀ[4ꢀ(5ꢀcyanoꢀ2ꢀ
1
3a
thienyl)phenyl]tetrahydroꢀ2Hꢀpyranꢀ4ꢀyl}propaneꢀ2ꢀsulfonamide (PFꢀ4778574, 1) and Nꢀ
((5S)ꢀ(3ꢀ(3ꢀfluoroꢀ4ꢀ(pyrrolidinꢀ1ꢀyl)phenyl)ꢀ4,5ꢀdihydroisoxazolꢀ5ꢀyl)methyl)propaneꢀ2ꢀ
13b
sulfonamide (PFꢀ4701475, 2) (Figure 1). Other AMPAR potentiators of this sulfonamide
class, specifically the biarylpropylsulfonamides (R)ꢀ3,5ꢀdifluoroꢀNꢀ(4ꢀ(1ꢀ((1ꢀ
1
7
methylethyl)sulfonamide)propanꢀ2ꢀyl)phenyl)benzamide (LY450108) and (R)ꢀNꢀ(2ꢀ(4'ꢀ(2ꢀ
(methylsulfonamido)ethyl)ꢀ[1,1'ꢀbiphenyl]ꢀ4ꢀyl)propyl)propaneꢀ2ꢀsulfonamide (LY451395,
1
7ꢀ18
mibampator)
and the amino indane sulfonamide Nꢀ[(2S)ꢀ5ꢀ(6ꢀfluoroꢀ3ꢀpyridinyl)ꢀ2,3ꢀ
1
9
dihydroꢀ1Hꢀindenꢀ2ꢀyl]ꢀ2ꢀpropanesulfonamide (GSK729327) , have demonstrated excellent
safety and tolerability in humans. Compounds 1 and 2, and those within their respective
chemical series, were characterized in select in vitro and singleꢀdose in vivo assays assessing
AMPARꢀdependent activities related to nootropic effects and safety. Due to the aforementioned
complex functional heterogeneity of AMPARs and the lack of a clear understanding of the
optimal AMPARꢀsubtype specificity profile for TI maximization, potentiator binding affinity
and functional potency were determined in rodent brain tissue expected to express a native
mixture of AMPAR subunits and their associated proteins to afford composite pharmacological
values. Functional activity was further evaluated in recombinant cell lines stably expressing
human GluA2i and GluA2o, since other AMPAR potentiators bind within these flip/flop
4
ACS Paragon Plus Environment

Page 5 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
20
domains , and structureꢀbased drug design (SBDD) using Xꢀray crystal structures of the LBD of
1
3b
human GluA2o iteratively informed GluA2ꢀbased structureꢀactivity relationships (SAR). For
memoryꢀrelated effects (i.e. “efficacy”) in animal models of pharmacologically provoked
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13a
13b
NMDAR hypofunction, both 1 and 2 reversed the [5R,10S]ꢀ5ꢀmethylꢀ10,11ꢀdihydroꢀ5Hꢀ
dibenzo[a,d]cycloheptenꢀ5,10ꢀimine (MKꢀ801)ꢀgenerated subiculumꢀmedial prefrontal cortex
(
PFC) dysfunction in rats, and 1 attenuated ketamineꢀmediated spatial working memory
1
1a
impairment in nonhuman primates (NHP). Safety studies with 1 looked for cerebellumꢀbased
AMPAR activation (mouse) and motor coordination disruptions (mouse, dog and NHP), as well
as convulsions (mouse, rat and dog). The resulting C ꢀnormalized exposureꢀresponse
b,u
continuum for 1 showed adequate separation between C linked to efficacy and adverse events
b,u
(AE). Critically, this work with 1 demonstrated an excellent correlation across species of Cb,u
required for procognitive characteristics and specific AE suggesting a common and translatable
exposureꢀresponse relationship in rodents and higherꢀorder species. Progressively higher C led
b,u
sequentially from efficacy to cerebellar 3',5'ꢀcyclic guanosine monophosphate (cGMP) elevation
to motor coordination disruptions to convulsion. These initial studies also underscored the need
for a potentiator with low human pharmacokinetic variability to ensure highly predictable
maximal plasma drug concentrations (Cmax) due to the clinical doseꢀcapping nature of these
2
1
thresholdꢀmediated AE. Appropriately, this strategy, and the interspecies relationships it
1
3, 21ꢀ22
established for four chemically distinct AMPAR potentiators
, aided in the identification of
the current clinical candidate Nꢀ{(3S,4S)ꢀ4ꢀ[4ꢀ(5ꢀcyanoꢀ2ꢀthienyl)phenoxy]tetrahydrofuranꢀ3ꢀ
yl}propaneꢀ2ꢀsulfonamide (PFꢀ04958242, 3) (Figure 1), which is described herein.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 75
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
RESULTS AND DISCUSSION
Discovery of the Novel Tetrahydrofuran Ether Series of AMPAR Potentiators. As
1
3
described previously , an internally developed fluorometric imaging plate reader (FLIPR)ꢀbased
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
23
functional assay utilizing mouse embryonic stem (mES) cellꢀderived neuronal precursors
enabled a high throughput screen (HTS) of our corporate compound file that identified two lead
1
3b, 23
chemotypes.
Subsequent SBDD using Xꢀray crystal structures of the human GluA2o LBD
3
led to the discovery of Nꢀ((S)ꢀ1ꢀ(3,5ꢀ[ H]ꢀ2ꢀfluoroꢀ4ꢀ((S)ꢀ5ꢀ((1ꢀmethylethylsulfonamido)methyl)ꢀ
3
4,5ꢀdihydroisoxazolꢀ3ꢀyl)phenyl)pyrrolidinꢀ3ꢀyl)acetamide ([ H]PFꢀ04725379, 4), a proprietary
radioligand, and other active compounds in its dihydroisoxazole series including prototype 2,
1
3b
which was used to evaluate AMPARꢀmediated pharmacology in vivo.
Simultaneously during
the HTS, we explored and expanded upon the SAR of the published biarylpropylsulfonamide
AMPAR potentiator (R)ꢀNꢀ[2ꢀ(4'ꢀcyano[1,1'ꢀbiphenyl]ꢀ4ꢀyl)propyl]propaneꢀ2ꢀsulfonamide
2
4
(
LY451646, 5). In this report we describe our medicinal chemistry strategy to ascertain a
unique series of AMPAR potentiators with improved efficacy, selectivity, and metabolic stability
compared to 5. This approach was guided by the Xꢀray structure of 5 bound to the GluA2o LBD
and applied known conformational preferences to generate a new lead series of AMPAR
potentiators. Additionally, we describe considerations of the physicochemical properties to
impart desired absorption, distribution, metabolism and excretion (ADME) characteristics into
newly generated molecules. As a result, we identified the novel potentiator 3, whose complete
primary pharmacology and interspecies ADME profiles are revealed.
Prior to achieving an Xꢀray crystal structure of 5 bound to the human GluA2o LBD, a
number of design hypotheses (Figure 2) were explored: investigation of different vectors around
the central aromatic ring (e.g. metaꢀsubstitution and heteroaryl replacement); linker modulation
6
ACS Paragon Plus Environment

Page 7 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
e.g. chain length alteration and heteroatom incorporation); sulfonamide variations/replacements;
and, a variety of ringꢀbuilding strategies to restrict the conformational flexibility of the ethylene
chain. Synthetic tractability was a key component of the chemistry strategy, often taking
advantage of libraryꢀenabled routes to rapidly expand our SAR knowledge. In all cases, the
physicochemical properties of virtual molecules were carefully considered so that only
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
5
compounds having desirable properties of oral CNS drugs were prepared. To facilitate the
optimization of matter from hitꢀtoꢀlead through lead development, various libraries were
designed and screened (32 ꢀM) in our mES cellꢀderived functional assay. For the original
design hypotheses, neither vector modification (e.g. metaꢀsubstitution) nor heteroaliphatic ring
(e.g. piperidine) replacement of the central aromatic ring afforded compounds with EC <32
50
ꢀ
M. Extending the length of the twoꢀcarbon linker between the central aromatic ring and
sulfonamide was also an ineffective strategy. However, linker modulations, specifically the
incorporation of an ether moiety, proved advantageous. As with the allꢀcarbonꢀlinker analogs,
paraꢀsubstitution on the central aromatic ring was favored over metaꢀsubstitution (Table 1; 6 vs.
8 and 7 vs. 9). Interestingly, while the 2ꢀcyanophenylꢀsubstituted analog 7 was 2ꢀfold less potent
than 5 it had ca. 60% greater efficacy suggesting a possible functionally favorable AMPAR
interaction for this distal aromatic substituent.
The Xꢀray coꢀcrystal structure of 5 bound to the human GluA2o LBD (Protein Data Bank ID:
1
3b
4
LZ5) revealed informative potentiatorꢀAMPAR contacts.
The sulfonamide hydrogen of 5
forms a backbone hydrogen bond to the carbonyl oxygen of Proꢀ494, while its biaryl group
projects across the dimeric LBD filling the symmetrical binding interface. The model structure
26
of 7, the most functionally efficacious carbon linker ether analog, was docked onto the crystal
structure of 5. This overlay suggested that the 2ꢀcyano group on the distal aromatic ring of 7
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 75
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
accessed the symmetrical space to the sulfonamide pocket on the other side of the binding motif
(Figure 3). The binding model of 7 shows that the 2ꢀcyano group on the distal aromatic ring is
precisely positioned to access the symmetrical binding pocket that is occupied by the
sulfonamide group and that the ether group is merely serving as a linker between these two
pockets. The acyclic amidoꢀalkoxyethane substructure in 7 prompted us to consider the use of
physical organic principles of conformational preference to preꢀorganize the ligand for binding
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
7
site complementarity to optimize pharmacological activity. Particularly, 1,2ꢀdialkoxyꢀ or 1,2ꢀ
dihaloethanes, similar in nature to 7, are known to prefer a conformation favoring the gauche
28
orientation (Figure 4A). The gauche orientation of a 1,2ꢀdialkoxyethane is favored by ca. 0.5
kcal/mol, and, in this case, it affords an opportunity for an intramolecular hydrogen bond within
7 between its sulfonamide hydrogen and ether oxygen. Therefore, it was hypothesized that
further constraint, via a carboꢀ or heterocycle, should enhance the conformational preference
imparted to the acyclic series by the gauche effect. Using 5ꢀ or 6ꢀmembered rings anchored on
the ethylene linker within 7 (Figure 4A), a number of conformationally restricted analogs were
designed to afford both transꢀ and cisꢀisomers (Figure 4B). Additional analyses of these virtual
compounds by superimposing their structures onto the crystal structure of 5 indicated the cisꢀ
2
9
isomers were optimal to match the bioactive conformations. Consistent with prior findings ,
further consideration of the physicochemical properties of these molecules allowed additional
refinement and selection of the cisꢀtetrahydrofuran (THF) ether analogs as the primary series
(Figure 4B, A = O).
A fiveꢀstep synthetic sequence assembled the cisꢀTHF ethers (Scheme 1). Highlights of the
synthetic strategy include employing classical resolution to avoid chiral chromatography and
lateꢀstage introduction of both the sulfonamide and aromatic substituents to facilitate their
8
ACS Paragon Plus Environment

Page 9 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
efficient diversification for rapid SAR development. Opening of epoxide 10 by pꢀbromophenol
11 in the presence of a base, such as cesium carbonate, in refluxing dioxane installed the aryl
ether group. The resulting alcohol was treated with methanesulfonyl chloride and triethylamine
to give diastereomeric 12 (quantitative yield over the two steps). Treatment of 12 with sodium
azide afforded the corresponding azide, which was converted to amine 13 via the Staudinger
reduction. Classical resolution of 13 with camphorsulfonic acid provided enantiopure 14 (30–
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
35% yield, >99% ee). Alternatively, treatment of mesylate 12 with ammonium hydroxide in
isopropanol at 110 °C cleanly afforded racemic cisꢀamine 14 in ca. 80% yield. Next, 14 was
reacted with isopropylsulfonylchloride in the presence of 1,8ꢀdiazabicyclo[5.4.0]undecꢀ7ꢀene
(DBU) to quantitatively afford sulfonamide 15, which underwent Suzuki coupling with either
boronic acid or ester 16 to effectively generate the desired THFꢀethers 17–28. In instances
where standard Suzuki coupling conditions were unsuccessful, specifically with 5ꢀmembered
heteroaryls, either reverse Suzuki coupling or CHꢀactivation enabled the desired product.
a
Reagents and conditions: a) Cs CO , BnNEt Cl, dioxane, reflux, 18 h, quant.; b) MsCl, Et N,
2
3
3
3
DCM, 0 °C to rt, 18 h, quant.; c) NaN , DMF, 110 °C, 66 h, 85%; d) PPh , THF, 0 °C to rt, 18 h,
3
3
7
8%; e) (+)ꢀCSA, IPA/H O, 18 h, 30–35%; f) iPrSO Cl, DBU, DCM, 0 °C to rt, 20 h, quant.; g)
2 2
Boronic acid or boronate ester (16), XPhos, Pd(OAc) , KF, 1:1 MeOH/toluene , 40–75%.
2
The impact of this ringꢀincorporation design strategy was immediately reflected in differing
functional activity improvements between THFꢀether vs. ethylꢀether analogs with constant distal
aromatics, such as unsubstituted phenyl (17 vs. 6, Emax of 125% vs. 32%) or oꢀcyanophenyl (18
vs. 7, EC of 21.9 nM vs. 7350 nM) analogs (Tables 1 and 2). Enantiopurification of 18
50
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 75
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
revealed its functional potency was predominately due to its (3S,4S)ꢀ(19) and not its (3R,4R)ꢀ
enantiomer (20). For all THFꢀether analogs for which both isomers were prepared, the (3S,4S)ꢀ
enantiomer was always most active (data not shown). A model of 19 bound to the human
GluA2o LBD showed that it overlays similarly to 7 (Figure 3) in that the oꢀcyano is optimally
positioned to access a pocket within the binding site that seems to enhance potency. Compound
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
19 also had physicochemical properties that translated to high membrane permeability (Papp,A→B
ꢀ6
=
32.3 × 10 cm/s) and no Pꢀglycoproteinꢀmediated efflux (MDR1 ER = 1.39) suggesting
favorable central penetration in vivo. However, its human liver microsomesꢀderived intrinsic
clearance (HLM CL ) was unacceptably high (138 mL/min/kg) since a molecule with low
int
human pharmacokinetic variability was desired due to the known clinical doseꢀcapping nature of
2
1
AMPAR potentiatorꢀmediated AE. To improve metabolic stability, we reasoned that the facile
metabolism of 19 was due to hydroxylation of the oꢀcyanophenyl. Accordingly, functionally
equipotent analogs, such as 2ꢀcyanoꢀ6ꢀfluoro (21) and 2ꢀcyanoꢀ4ꢀfluoro (22), indeed had lower
HLM CL than 19 (Table 2). However, subsequent modifications that successfully reduced
int
aromatic ring lipophilicity to achieve even lower HLM CL (e.g. 23) concomitantly reduced
int
functional potency while increasing the number of hydrogen bond donors, TPSA and Pꢀ
3
0
glycoprotein liability. Alternatively, replacing the distal phenyl group with aromatic
heterocycles afforded low HLM CL , high permeability and no Pꢀglycoprotein susceptibility
int
while retaining acceptable functional potency. Such substitution of the distal phenyl group with
more polar heteroaromatics (e.g. pyridine (e.g. 24 and 25) or pyrimidine (e.g. 26)) also imparted
3
1
lower clogP, which tends to decrease the likelihood of in vivo toxicity by small molecules.
Several 5ꢀ or 6ꢀmembered aromatic heterocycles were prepared, such as cyanothiophene 3,
cyanothiazole 27 and dimethylpyrazole 28 (Table 2). Compounds 3 and 27 provide a different
10
ACS Paragon Plus Environment

Page 11 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
vector for the cyano group, which potentially occupies a different binding pocket that could
3
2
impact flip/flop selectivity. It has been speculated that the smaller serineꢀ754ꢀcontaining side
chain within the flip splice variant creates a pocket having a more favorable binding interaction
with the requisite cyano moiety that fills it, resulting in a flipꢀfavoring selectivity for such
potentiators. An Xꢀray coꢀcrystal structure of 3 bound to the human GluA2o LBD (Figure 5)
revealed the common hydrogen bond between the sulfonamide and Proꢀ494, and also showed
that the cyano group occupied the same space in the vicinity of Asnꢀ754 as the cyano group in 5.
Of these exemplified compounds, 3, whose absolute stereochemistry was unambiguously
elucidated as the (3S,4S)ꢀenantiomer via single crystal Xꢀray analysis (see Supporting
Information), was selected for further profiling since it provided the best overall balance of
functional potency and efficacy with desired in vitro ADME properties.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Characterization and Preclinical Profile of 3. Potentiator 3 was carefully profiled in a
series of singleꢀdose rodent assays to determine its neuropharmacokinetics and mechanismꢀ
mediated exposureꢀresponse continuum using the previously described interspecies knowledge
13a, 21ꢀ22
for select AMPAR potentiators.
Observed animal effects were most likely solely
AMPARꢀdependent since 3 (10 ꢀM) only inhibited three receptors/transporters and one enzyme
within a broad humanꢀbased receptor/enzyme selectivity panel (Cerep, 118 targets), with the
serotonin transporter (K of 530 nM) having the most potent interaction. For each animal dose,
i
C was calculated to define a C ꢀnormalized AMPARꢀdependent exposureꢀresponse
b,u
b,u
continuum for 3 to understand separations between C enhancing cognition and disturbing
b,u
motor coordination, as well as in vitroꢀin vivo pharmacological correlations. Additionally,
animal and human ADME data were generated and used to project human pharmacokinetic
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 75
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
parameters, an efficacious dose and the expected Cmax variability at this dose. Collectively, these
datasets afforded an acute TI that supported the progression of 3 into chronic toxicology studies.
In Vitro Pharmacology. The in vitro pharmacological characterization of 3 used a primary
strategy of quantifying its properties in rodent tissues expected to express native populations of
AMPARs and their associated proteins; a good correlation was found for AMPAR potentiator
functional potencies in mouse and human ES cellꢀderived neurons suggesting interspecies
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
3
3
13b
consistency. Using the proprietary radioligand [ H]4 , 3 had a mean K of 170 nM in a rat
i
cortical homogenate. Functionally, 3 showed a glutamateꢀdependent AMPAR positive allosteric
modulation of SꢀAMPAꢀevoked current in rat primary cortical neurons (EC of 43±10 nM, ca.
50
5ꢀfold increase of the maximal SꢀAMPAꢀinduced current; Supporting Information) and Sꢀ
2
+
AMPAꢀmediated Ca influx measured in mouse ES cellꢀderived neuronal precursors (EC of
5
0
3
10 nM, Emax of 110%) and recombinant human GluA2i (EC of 24 nM, E of 124%) and
50 max
GluA2o (EC of 880 nM, Emax of 108%) (Table 3).
50
Neuropharmacokinetics. Singleꢀ (CDꢀ1 mouse) or multipleꢀ (rat) time point
neuropharmacokinetics studies were performed with 3 in rodents (Supporting Information) prior
to respective in vivo pharmacology tests to optimize both dose selection and the time of
pharmacodynamic assessment(s). Furthermore, doses within the respective speciesꢀspecific
pharmacology models (except the mouse rotarod assay) were assumed to have the same doseꢀ
C relationship as in the neuropharmacokinetics studies and, thus, their C were extrapolated
b,u
b,u
linearly from the neuropharmacokinetics data to define the exposureꢀresponse relationships of 3
at a specific time point postꢀdose. Thirty minutes after a single dose (1.78 mg/kg, SC) of 3, CDꢀ
1
mice had a mean ± SD total plasma compound concentration (C ) of 242±106 ng/mL and total
p
brain compound concentration (C ) of 247±26 ng/mL (equating to a C of 12.0±1.3 nM using
b
b,u
12
ACS Paragon Plus Environment

Page 13 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the rat unbound fraction of compound in brain homogenate (f ) of 0.019), which afforded a
u,b
C :C of 1.1±0.3. After a single dose (1 mg/kg, SC) to SpragueꢀDawley rats, 3 demonstrated a
b
p
similar terminal halfꢀlife (t ) in plasma (0.76 h), cerebrospinal fluid (0.71 h) and brain (0.53 h),
1
/2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
and a Tmax of 0.75 h within each neuromatrix. Consistent with the in vitro data predicting it to be
a highly permeable nonꢀPꢀglycoprotein substrate (Table 2), 3 afforded AUC0–Tlastꢀderived ratios
suggesting it is readily brain penetrant (C :C of 1.5) with net equilibrium at the bloodꢀbrain
b
p
barrier (C :C of 0.59) (Table 4). Similarly, an IV bolus (0.01 mg/kg) neuropharmacokinetics
b,u p,u
study, performed in urethaneꢀanesthetized rats under pairedꢀpulse facilitation (PPF) assay
conditions, showed 3 had parallel t1/2 in each neurocompartment, a C :C of 2.6 and C :C of
b
p
b,u p,u
1
.0. The data suggest urethane anesthesia has no significant effect on the neuropharmacokinetics
of 3 in rats.
In Vivo Pharmacology. Compound 3 was evaluated in specific rodent pharmacology models
to define the relationship between its C and AMPARꢀmediated effects. Collectively, this
b,u
allowed a C ꢀnormalized exposureꢀresponse continuum for 3 across multiple models of
b,u
AMPARꢀdependent mechanism, safety and procognitive properties (“efficacy”) to determine its
singleꢀdose TI (Table 3). From a mechanism perspective, 3 was first assessed for its ability to
elevate cerebellar cGMP in CDꢀ1 mice, where it showed statistically significant increases (P
<0.05 vs. vehicleꢀtreated animals) 0.5 h after receiving a dose of 1.0 mg/kg, SC (Figure 6A).
Based on the satellite CDꢀ1 mouse neuropharmacokinetics study (1.78 mg/kg, SC) and assuming
stationary pharmacokinetics at all tested doses, the projected mean C at this cGMPꢀincreasing
b,u
dose was 6.7 nM. Subsequently, the ability of 3 to produce motor deficits was assessed in
C57BL/6J mice undergoing an accelerating rotarod test since prior data with 1 defined a very
close relationship between C decreasing fall latency in this mouse assay to those causing
b,u
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 75
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
13a
movementꢀrelated tremors/ataxia in NHP. Potentiator 3 caused statistically significant (P
<
0.05) motor deficits in mice 0.5 h after 0.32 mg/kg, SC (Figure 6B), which corresponded to a
mean C of 10.8 nM (Table 3).
b,u
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
For efficacy, 3 was tested in two contemporary rat models of pharmacologically provoked
3
3
11b
NMDAR hypofunction : MKꢀ801ꢀinduced subiculumꢀmedial PFC dysfunction and ketamineꢀ
3
4
disrupted spatial working memory in the radial arm maze (RAM). In the electrophysiological
model, an initial cumulative IV doseꢀresponse study (0.001 to 0.01 mg/kg, N=5) was conducted
with 3, and a significant (P <0.01) reversal of the disruptive effects of MKꢀ801 on both cortical
electroencephalography (EEG) and PPF were observed following 0.01 mg/kg, IV (Figure 7A).
For each dose, C was projected linearly from the C (0.77 nM) observed at 5 min postꢀdose in
b,u
b,u
the 0.01 mg/kg, IV neuropharmacokinetics study performed under identical assay conditions. A
SpearmanꢀKarber analysis revealed an effective dose for 50% of animals (ED ) of 0.003 mg/kg
5
0
(
C of 0.23 nM). Subsequently, a single (i.e. nonꢀcumulative) 0.01 mg/kg, IV time course study
b,u
was conducted for pharmacokinetic/pharmacodynamic (PK/PD) modeling (Figure 7B); mean
onset of the reversal by 3 of the MKꢀ801ꢀdependent PPF disruption was 8±3 min postꢀdose (Cb,u
of ca. 0.77 nM). Using a link model, the estimated slope of 3 to reverse the MKꢀ801ꢀinduced
PPF disruption was 14.2 %•mL/ng, the population mean baseline (i.e. MKꢀ801ꢀmediated
ꢀ1
disruption) was 79.2% and the K was estimated to be 1.93 h . Before the MKꢀ801 disruption,
e0
the percent of low frequency delta power for untreated rats was 38%. The PD response observed
in the PPF experiment ranged from 0% to 73% of the reversal of the MKꢀ801ꢀinduced
disruption, hence the model may reliably predict the extent of this reversibility up to an
inhibitory concentration maximum of 73%. The modeling results indicate that in rats 3 is able to
reverse 50% of the MKꢀ801ꢀinduced PPF disruption at a C of 1.5 ng/mL, which, based on the
p
14
ACS Paragon Plus Environment

Page 15 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
11b
IV bolus neuropharmacokinetics data, equates to a C of 0.18 nM (Table 3). As reported for
b,u
another AMPAR potentiator in this electrophysiological model, the restorative effects of 3 were
fully blocked by pretreatment (1 mg/kg, IV; 10 min after MKꢀ801 administration and 20 min
before dosing 3) with the selective AMPAR antagonist (S)ꢀ3ꢀ(2ꢀchlorophenyl)ꢀ2ꢀ[2ꢀ(6ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
5
diethylaminomethylꢀpyridinꢀ2ꢀyl)ꢀvinyl]ꢀ6ꢀfluoroꢀ3Hꢀquinazolinꢀ4ꢀone (CPꢀ465022) (data not
shown).
Results in this electrophysiological assay provided the theoretical basis for studying the
ability of 3 to curb the disruptive effects of the NMDAR antagonist ketamine on spatial working
memory in rats since monosynaptic projection from the hippocampus/subiculum to the medial
3
6
PFC contributes to working memory and such dysfunction is implicated in the decreased
3
7
cognitive performance of patients with schizophrenia. Thus, employing a spatial working
34
memory version of the RAM , 3 significantly attenuated the ketamineꢀevoked working memory
errors at doses of 0.0032 to 0.032 mg/kg, SC (Figure 8), which equated to projected C of 0.029
b,u
to 0.29 nM (Table 3) per the subcutaneous rat neuropharmacokinetics study. As compared to the
ketamineꢀtreated group, the 0.032 mg/kg dose also significantly improved several other
performance indices (data not shown): it decreased the average time to maze completion; it
increased the average number of choices prior to first error; and, it increased the percentage of
correct choices on the maze. The data suggest increased glutamatergic neurotransmission
produced by 3 afforded significant protection from the ketamineꢀdependent deficits in spatial
11a, 38
working memory in a hormetic exposureꢀresponse fashion typical of procognitive agents.
Once demonstrating both electrophysiological and behavioral effects in rats, 3 was further
1
8
studied in rats using 2ꢀdeoxyꢀ2ꢀ[ F]fluoroꢀ
Dꢀglucose (FDG) positron emission tomography
(
PET) as a potential methodology for determining a clinically translatable central biomarker of
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 75
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
39
40
AMPAR potentiation. Prior FDGꢀPET studies with 5 in rats and NHP revealed increased
cerebral glucose metabolism rotating from cortical regions to the cerebellum at C paralleling
b,u
the behavioral transition from procognitive effects (rat and NHP) to elevation in cerebellar
cGMP (rat) and motor coordination disruptions (NHP). Thus, FDGꢀPET studies strategically
used doses of 3 (0.027, 0.08 and 0.60 mg/kg, SC) affording at the time of FDG administration
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
maximal C of 0.3 to 6.7 nM, which showed activity in the rat RAM and mouse cerebellar
b,u
cGMP models, respectively (Table 3). In rats, 3 produced significant increases in FDG uptake
(P <0.05 vs. vehicleꢀtreated animals) in cortex and cerebellum (Figure 9A). As observed with 5,
the magnitude of increased FDG uptake, as measured by the number of significantly increased
voxels within a brain region of interest, caused by 3 had an opposite pattern in cortex and
cerebellum. Specifically, the 3ꢀdependent FDGꢀuptake increase at the low dose, which afforded
a C of 0.3 nM that showed the most robust effects in the rat RAM, was almost exclusively
b,u
within the cortex vs. the cerebellum (210 vs. 6 significantly increased voxels). Conversely,
FDGꢀuptake at the high dose, which yielded the minimally effective cGMPꢀelevating C of 6.7
b,u
nM, was almost exclusively within the cerebellum vs. the cortex (1182 vs. 58 significantly
increased voxels) (Figure 9B). Since small molecule concentrations are essentially
1
6b
homogeneous across brain regions , these data suggest that different concentrations of 3 are
required to selectively (and optimally) activate the respective brain regions associated with
procognitive effects (e.g. cortex) and motor coordination (e.g. cerebellum), which may be due to
intrinsically different biophysical properties of circuitꢀspecific AMPAR subpopulations. As in
the rat PPF model, these increases in FDG uptake were completely blocked by pretreatment (3.2
mg/kg, SC; 30 min before dosing 3) with the AMPAR antagonist CPꢀ465022 (data not shown).
Subsequently, a similar FDGꢀPETꢀbased exposureꢀresponse relationship was determined for 3 in
16
ACS Paragon Plus Environment

Page 17 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
41
NHP suggesting this methodology could be considered as a potential translatable central
biomarker for both efficacyꢀ and safetyꢀrelated effects of AMPAR potentiation by 3 in humans.
Interestingly, the C of 3 generating physiological effects in all in vivo assays were just
b,u
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
fractions of its concentrations required to produce halfꢀmaximal responses in the receptorꢀbased
in vitro assays (Table 3). Using the rat K and respective C , in vivo measures of efficacy and
i
b,u
motor deficits occur at projected AMPAR occupancies of <1% and 6%, respectively, and 10 nM
was the lowest tested concentration of 3 that clearly enhanced SꢀAMPAꢀevoked currents in rat
13a
cortical neurons (Supporting Information). It is hypothesized that this phenomenon is due to
animals, unlike the employed in vitro systems, having completely intact synergistic neural
networks in which only minimal AMPAR potentiation produces enough excitatory
neurotransmission for the nonlinear activation of downstream pathways.
Pharmacokinetics and Metabolism. Consistently across all species tested and a 10ꢀfold
concentration range, 3 showed a moderate unbound fraction in liver microsomes (fu,LM; 0.42 to
0.52), nearꢀunity bloodꢀtoꢀplasma concentration ratio (B/P; 0.82 to 0.90) and low unbound
fraction in plasma (f ; 0.047 to 0.110) (Tables 4 and 5). Singleꢀdose pharmacokinetic studies
u,p
(Table 4) indicate 3 is rapidly absorbed in rats and dogs (Tmax ≤0.8 h) following oral (PO)
administration. In rats, 3 had very high systemic plasma clearance (CL, 169 mL/min/kg, <1%
renal), moderate steadyꢀstate volume of distribution (V , 3.5 L/kg) and moderate oral
ss
bioavailability (F, 58%). In dogs, its apparent CL (11.2 mL/min/kg) and apparent V (6.0 L/kg)
ss
were moderate. Mean t1/2 of 3 were short in rats (≤0.8 h) and moderateꢀtoꢀlong in dogs (6.2 h)
and NHP (12.3 h).
Compound 3 demonstrated low and moderate scaled clearance in HLM (CL ≤3.8
H
mL/min/kg) and hepatocytes (CL ≤7.1 mL/min/kg), respectively (Table 5). Equivalent studies
H
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 75
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
in rat (CL ≤63.6 mL/min/kg) and dog (CL ≤7.4 mL/min/kg) hepatic microsomes were
H
H
consistent with their respective in vivo CL (Table 4). Preliminary in vitro interspecies (rat, dog
and human) biotransformation studies with nonꢀradiolabeled 3 (10 ꢀM) found it underwent
similar metabolism in dog and human systems where tetrahydrofuran hydroxylation
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
predominates; Oꢀdealkylation occurred in all species (Supporting Information). Initial
metabolite profiling of plasma (rat and dog) and brain (rat) samples identified both 3 and its Oꢀ
dealkylated metabolite in rat plasma, but only 3 in both dog plasma and rat brain. These studies
coupled with those employing major recombinant human cytochrome P450 (rhCYP) isoforms
suggest the predominant human clearance pathway for 3 is hepatic CYPꢀmediated metabolism
via CYP2C19 and CYP3A4, both at minimal rates (Table 5). Although a CYP3A4 substrate, no
turnover (i.e. t1/2 >93 min) of 3 (1 ꢀM) was detected in human intestinal microsomes, consistent
with its very low rhCYP3A4 CLint,app (Table 5). The ability of 3 to inhibit seven major CYPs
was investigated using HLM and probe substrates, and its lowest IC was 18 ꢀM (CYP2D6).
5
0
For clinical pharmacokinetic parameter projections, both more traditional and newer in silico
4
2
methods were employed. Primarily for the latter methodology, Simcyp software used select
experimentally determined data and predicted human pharmacokinetic values to simulate Cpꢀ
time profiles for a forecasted efficacious dose of 3 to estimate the clinical variability in Cmax
relative to both C and motor coordination disruptions, an exposure thresholdꢀdictated AE.
eff
th
Particularly, this clarified the adequacy of exposure separation between the 95 percentile Cmax
,
which corresponds to the lowest Cmax for the five virtual subjects exhibiting the least extensive
metabolism of 3, and the C for movementꢀrelated deficits (C ) at the putative efficacious oral
p
AE
dose. For this exercise, projected clinical C (2.4 ng/mL) and CAE (89 ng/mL) were derived
eff
using Eq. 2 (Experimental Section) from rodent C corresponding to the highest effective
b,u
18
ACS Paragon Plus Environment

Page 19 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
exposure in the rat efficacy models (C of 0.29 nM) and to the mean value inducing statistically
b,u
meaningful motor deficits in the mouse rotarod assay (C of 10.8 nM), respectively (Table 3).
b,u
2
1
Using previously described methodology , integration of the interspecies in vitro and in vivo
metabolism datasets predicts 3 to have in humans a lowꢀtoꢀmoderate CL (3.12 mL/min/kg) and
Vss (0.63 L/kg), resulting in a shortꢀtoꢀmoderate t1/2 (2.3 h) and a moderateꢀtoꢀhigh F (81%)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(Table 5). The human efficacious dose (0.3 mg) was calculated from these pharmacokinetic
parameters, a dosing interval (τ) of 12 h (BID), an assumed 70 kg subject and the desire to
maintain C as the average C at steadyꢀstate (C ). Using these same C and
eff
p
ss,av
eff
pharmacokinetic parameters, but substituting respective rhCYPꢀgenerated CLint,app, Simcyp
simulations support this projected dosing regimen after a single or multiple (BID for 7 d) dose(s)
of 0.3 mg, with low intersubject pharmacokinetic variability and ≤2ꢀfold accumulation at steadyꢀ
state (Figure 10). Importantly, the modeling of this clinically efficacious dosing regimen of 3
th
expects no violation of its CAE by its 95 percentile Cmax. Assuming that both rodent efficacy
and motor coordination disruptions are directly translatable, the described animal data suggest
that 3 has a singleꢀdoseꢀdefined TI of 37 in humans for selfꢀlimiting intention tremors, a readily
clinically monitorable AE. This, coupled with the projected low variability in human C ꢀtime
p
profiles, supported the progression of 3 to preclinical safety studies.
On the basis of the described holistically favorable dataset, 3 underwent 90ꢀd regulatory
toxicology studies in rats and dogs in which adequate TIs (both Cmaxꢀ and AUCꢀbased) were
achieved to advance it to human testing. To date, 3 has been evaluated in eleven clinical trials
(see www.clinicaltrials.gov) where it has been safe and well tolerated in healthy volunteers (18–
43
5
5 y old), in patients (55–70 y old) with ageꢀrelated sensorineural hearing loss and as
adjunctive treatment in subjects with schizophrenia on stable doses of antipsychotics and other
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 75
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
psychotropic medications. Current clinical studies are evaluating 3 for procognitive effects in
select neurodisease populations, but principally for its ability to overcome cognitive impairments
associated with schizophrenia.
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
20
ACS Paragon Plus Environment

Page 21 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CONCLUSIONS
Rational drug design optimizing both ADMEꢀrelated physicochemical properties and
molecular interactions with the human GluA2 LBD led to the discovery of a new tetrahydrofuran
ether class of highly potent AMPAR potentiators. This novel chemical series ultimately afforded
Nꢀ{(3S,4S)ꢀ4ꢀ[4ꢀ(5ꢀcyanoꢀ2ꢀthienyl)phenoxy]tetrahydrofuranꢀ3ꢀyl}propaneꢀ2ꢀsulfonamide (PFꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
04958242, 3), whose preclinical characterization predicted in humans both a singleꢀdoseꢀdefined
TI of 37 for selfꢀlimiting tremors and low intersubject pharmacokinetic variability. This profile
is expected to greatly minimize the potential of eliciting thresholdꢀmediated AEs at a clinically
efficacious dosing regimen. Consistent with favorable regulatory toxicology studies that
included 90ꢀd studies in rats and dogs, 3 has been safe and wellꢀtolerated in both healthy
volunteers (young and elderly) and targeted disease populations, and it is currently being
evaluated for its ability to attenuate cognitive deficits in schizophrenia.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 75
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
EXPERIMENTAL SECTION
General. All compounds were synthesized and fully characterized at Pfizer Worldwide
Research & Development (WRD, Groton, CT). Chemicals and solvents of reagent or HPLC
grade were supplied by SigmaꢀAldrich Corporation, Thermo Fisher Scientific (Waltham, MA)
and Mallinckrodt Baker, Inc. (Phillipsburg, NJ). For control matrices, speciesꢀspecific plasma
was acquired from Bioreclamation, Inc. (Hicksville, NY) and rat brain tissue was obtained at
WRD. Male CDꢀ1 mice (20 to 25 g) and SpragueꢀDawley rats (200 to 275 g) were bought from
Harlan Laboratories (Indianapolis, IN) or Charles River Laboratories, Inc. (Wilmington, MA),
male C57BL/6J mice (20 to 25 g) were procured from The Jackson Laboratory (Bar Harbor, ME)
and beagle dogs (9.4 to 10.3 kg) were sourced from Marshall BioResources (North Rose, New
York). Nonꢀnaïve NHP (Macaca fascicularis, 6.8 to 8.5 kg) resided within the WRD NHP
colony (Groton, CT). All animal studies were performed in accordance with the Guide for the
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
4
4
Care and Use of Laboratory Animals using protocols reviewed and approved by the WRD
Institutional Animal Care and Use Committee. Under a protocol approved by Schulman
Associates IRB (Cinicnnati, OH), human blood was collected at WRD from three anonymous
medicationꢀfree male donors on the assay day and a portion of it was used to harvest plasma.
Liver microsomes were either prepared and characterized by WRD (Groton, CT) or purchased
from BD Gentest (Woburn, MA), cryopreserved human hepatocytes were acquired from
Invitrogen (Carlsbad, CA), intestinal microsomes were procured from Xenotech LLC (Lenexa,
KS) and recombinant Baculosomes expressing human cytochrome P450 isozymes (CYP1A2,
CYP2C8, CYP2C9, CYP2C19, CYP2D6 and CYP3A4) were obtained from PanVera
Corporation (Madison, WI). All blood samples were collected into EDTAꢀcontaining tubes and
processed immediately to obtain plasma, and the collection of neuromatrices from rodents
22
ACS Paragon Plus Environment

Page 23 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
16b
followed published techniques.
All biomatrices collected for drug quantification were stored
at ꢀ70 °C until processing for bioanalysis.
Chemistry. General Chemistry Procedure. All reagents and solvents were used as
purchased from commercial sources. Reactions were performed under nitrogen. Mass spectral
data were collected on a Micromass ADM atmospheric pressure chemical ionization instrument
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
LRMS APCI) (Micromass UK Ltd., Manchester, UK). NMR spectra were generated on a
Varian 300, 400 or 500 MHz (Varian Medical Systems, Inc., Palo Alto, CA) or Bruker 400 MHz
Bruker Corporation, Coventry, UK) instrument. Chemical shifts were recorded in ppm relative
(
to TMS with multiplicities given as s (singlet), d (doublet), t (triplet), q (quartet), sep (septet), dd
(doublet of doublets), dt (double of triplets), ddd (doublet of doublet of doublets), m (multiplet)
and br (broad signal). Accurate MS analyses (HRMS) were conducted on an Agilent 6220 TOF
mass spectrometer (Agilent Technologies, Wilmington, DE) in electrospray mode in either
positive ion mode using an Agilent ZORBAX column (1.8 µ, 3.0 × 50 mm, 60 °C) or negative
ion mode using a Waters XBridge column (2.5 µ, 3.0 × 50 mm, 60 °C) (Waters Corporation,
Milford, MA). The system was calibrated to greater than 1 ppm accuracy across the mass range
prior to analyses according to manufacturer’s specifications. The samples were separated using
UHPLC on an Agilent 1200 system prior to mass spectrometric analysis. The resulting spectra
were automatically lock massꢀcorrected, and the target mass ions and any confirming adducts
+
+
(
Na , NH ) were extracted and combined as a chromatogram. The mass accuracy was
4
calculated for all observed isotopes against the theoretical mass ions derived from the chemical
formula using MassHunter software (Agilent Technologies). Lowꢀresolution mass spectra were
determined on a Waters/Micromass system. The Xꢀray diffraction measurements were
performed at 298 K on a Bruker SMART APEX CCD Area Detector System equipped with a
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 75
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
graphite monochromator and sealedꢀtube Cu radiation (1.54178 Å) source. Compound purity
was determined by either combustion analysis (Intertek Pharmaceutical Services, Whitehouse,
NJ) where the analyses, indicated by the symbols of the elements, were within ±0.4% of the
theoretical values or HPLC, confirming it to be ≥95% pure, or by a combination of these two
methods. For HPLC (A ), analytes were eluted on a Waters XBridge C18 analytical column (5
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
54
ꢁ
, 4.6 × 150 mm) at 1.5 mL/min with acetonitrile (solvent A) and water containing 0.1% TFA
(solvent B) using the following gradient: 0 to 1.5 min, 5% solvent A in solvent B; 1.5 to 10 min,
5% to 100% A in B; 10 to 11 min, 100% A; 11 to 12.5 min, 100% to 5% A in B. Retention
times (t ) are reported in min and purity is calculated as the percent of total area.
R
trans-4-(4-bromophenoxy)tetrahydrofuran-3-yl methanesulfonate (12). 3,6ꢀ
Dioxabicyclo[3.1.0]hexane (10, 100 g, 1.16 mol), 4ꢀbromophenol (11, 241.1 g, 1.39 mol),
cesium carbonate (492 g, 1.51 mol), and benzyltriethylammonium chloride (52.9 g, 0.23 mol)
were suspended in dioxane (1 L) and heated at reflux for 18 h. The reaction was cooled to rt,
diluted with ethyl acetate and then washed with saturated aqueous sodium carbonate solution.
The aqueous layer was extracted with ethyl acetate, the combined organic portions were dried
over sodium sulfate and filtered, and the solvent was removed under reduced pressure to provide
crude product (354 g, >100%, assumed quantitative), which solidified on standing; this was used
without purification in the next step. Material from a smallerꢀscale experiment carried out in
similar fashion was purified by silica gel chromatography for characterization (gradient of 10%
to 35% ethyl acetate in heptane) to afford transꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀol as a
1
solid. H NMR (400 MHz, CDCl ) δ 7.40 (d, J = 9.1 Hz, 2H), 6.81 (d, J = 9.0 Hz, 2H), 4.67 (m,
3
1H), 4.41 (br s, 1H), 4.26 (dd, J = 10.4, 4.9 Hz, 1H), 4.05 (dd, J = 10.1, 4.0 Hz, 1H), 3.91 (dd, J
=
10.3, 2.1 Hz, 1H), 3.83 (m, 1H), 2.09 (br d, J = 4.5 Hz, 1H).
24
ACS Paragon Plus Environment

Page 25 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Triethylamine (181.9 mL, 1.31 mol) was added to a solution of transꢀ4ꢀ(4ꢀ
bromophenoxy)tetrahydrofuranꢀ3ꢀol (354 g, assumed 300.6 g, 1.16 mol) in methylene chloride
(
2 L), and the reaction was cooled to 0 °C in an ice bath. Methanesulfonyl chloride (101.3 mL,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
.31 mol) was then added dropꢀwise, while keeping the reaction temperature <5 °C, and the
reaction was stirred at rt for 18 h. Water (1.5 L) was added and the aqueous layer was extracted
with methylene chloride. The organics were combined, dried over sodium sulfate and filtered,
and the solvent was removed under reduced pressure to yield the brownꢀoil product (399.6 g,
quantitative). Material from a smallerꢀscale experiment performed similarly was triturated with
ethanol for characterization to afford transꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀyl
1
methanesulfonate (12). H NMR (400 MHz, CDCl ) δ 7.42 (d, J = 9.0 Hz, 2H), 6.85 (d, J = 9.1
3
Hz, 2H), 5.20 (m, 1H), 4.98 (br d, J = 4.6 Hz, 1H), 4.23 (dd, J = 10.5, 4.6 Hz, 1H), 4.16 (dd, J =
11.1, 3.9 Hz, 1H), 4.07 (m, 1H), 4.00 (br dd, J = 10.4, 1.9 Hz, 1H), 3.10 (s, 3H); Elemental
ꢀ
analysis calcd for C H BrO S C: 39.18, H: 3.89, found C: 39.31, H: 3.71; HRMS (ESI ) m/z [M
1
1
13
5
+
+
H] calcd for C H BrO S 360.9539, found 360.9543.
11 13 5
cis-4-(4-bromophenoxy)tetrahydrofuran-3-amine (13). Sodium azide (192.6 g, 2.96 mol)
was added to a solution of transꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀyl methanesulfonate (12,
o
133.2 g, 0.395 mol) in DMF (3 L) and the reaction was heated at 110 C for 66 h. The reaction
was cooled to rt and water (12 L) was added. This reaction was conducted a total of three times
on the same scale and the combined batches were extracted with MTBE. The organic layers
were then dried over sodium sulfate and the solvent removed under reduced pressure to afford
cisꢀ3ꢀazidoꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuran as a redꢀbrown oil contaminated with 18%
1
DMF (corrected yield of 286.7 g, 85%). H NMR (400 MHz, CDCl ) δ 7.41 (d, J = 9.0 Hz, 2H),
3
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 75
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
6
.83 (d, J = 9.1 Hz, 2H), 4.90 (ddd, J = 5.4, 5.4, 4.4 Hz, 1H), 4.17 (dd, J = 10.0, 5.6 Hz, 1H),
.09 (dd, J = 8.7, 5.8 Hz, 1H), 4.00 (m, 1H), 3.93–3.97 (m, 2H).
4
A solution of cisꢀ3ꢀazidoꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuran (286.7 g, 1.01 mol) in THF
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
(1.25 L) was cooled to 0 °C and treated with triphenylphosphine (278 g, 1.06 mol). The reaction
was allowed to warm to rt and stirred for 18 h, after which water (53 mL) was added and the
reaction was further stirred at rt for 66 h. Solvent was removed under reduced pressure and the
residue was mixed with diethyl ether. The supernatant was decanted and the solvent was
removed under reduced pressure to provide a residue, which was filtered through a short plug of
silica gel (gradient of 0% to 5% methanol in methylene chloride) to afford cisꢀ4ꢀ(4ꢀ
bromophenoxy)tetrahydrofuranꢀ3ꢀamine (13, 155 g) and impure material (366 g). An acid/base
extraction of the impure material yielded additional 13 (combined yield of 203.5 g, 78%).
Resolution of cis-4-(4-bromophenoxy)tetrahydrofuran-3-amine (13). A mixture of cisꢀ4ꢀ
(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀamine (13, 191 g, 0.74 mol) and (1S)ꢀ(+)ꢀ10ꢀ
camphorsulfonic acid (154.2 g, 0.66 mol) was dissolved in 2ꢀpropanol (2.4 L) and water (100
mL) at reflux. The clear solution was cooled to rt over 18 h, and the resulting crystals were
isolated, washed and dried to afford the (+)ꢀcamphorsulfonic acid salt of (3S,4S)ꢀ4ꢀ(4ꢀ
bromophenoxy)tetrahydrofuranꢀ3ꢀamine (139.2 g, 0.284 mol) with an 85% ee. Recrystallization
from 2ꢀpropanol (1.2 L) and water (70 mL) provided the (+)ꢀcamphorsulfonic acid salt of
(3S,4S)ꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀamine (113.2 g, 31% yield). Chiral HPLC, using
a Chiralpak IA column (4.6 × 250 mm) at 0.7 mL/min with heptanes:isopropanol:diethylamine
(v/v/v, 80:20:0.2), determined 98.7% ee at 220 nm and 99.4% ee at 260 nm with a t of 14.7 min.
R
1
H NMR (300 MHz, DMSOꢀd ) δ 8.23 (br s, 3H), 7.52 (d, J = 9.0 Hz, 2H), 6.99 (d, J = 8.8 Hz,
6
2H), 5.04 (m, 1H), 3.96–4.10 (m, 3H), 3.78–3.84 (m, 2H), 2.85 (d, J = 14.7 Hz, 1H), 2.66–2.73
26
ACS Paragon Plus Environment

Page 27 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
m, 1H), 2.35 (d, J = 14.7 Hz, 1H), 2.19–2.28 (m, 2H), 1.76–1.94 (m, 2H), 1.29 (d, half of AB q,
J = 10 Hz, 1H), 1.23 (d, half of AB q J = 10 Hz, 1H), 1.05 (s, 3H), 0.74 (s, 3H); Elemental
analysis calcd for C H BrNO •C H O S C: 48.98, H: 5.75, N: 2.86, found C: 49.15, H: 5.40,
1
0
12
2
10 16
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
+
N: 2.80; HRMS (ESI ) m/z [M + H] calcd for C H BrNO 258.0124, found 258.0125.
1
0
12
2
The (3S,4S)ꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀamine (+)ꢀcamphorsulfonic acid salt
(
126.8 g, 0.258 mol) was washed with 2 N aqueous sodium hydroxide solution and extracted
three times with methylene chloride. The organic layers were combined and the solvent was
removed under reduced pressure to provide (3S,4S)ꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀ
amine (14a, 65.6 g, 98% yield, 99% ee) as a white solid.
Subsequently, the combined mother liquors, enriched in (3R,4R)ꢀ4ꢀ(4ꢀ
bromophenoxy)tetrahydrofuranꢀ3ꢀamine, were washed with 2 N sodium hydroxide and extracted
with methylene chloride. The combined organic layers were dried over sodium sulfate and the
solvent was removed under reduced pressure to give a residue (156.3 g, 0.606 mol of product
and its enantiomer). This material was combined with (1R)ꢀ(ꢀ)ꢀ10ꢀcamphorsulfonic acid (126.2
g, 0.54 mol) and dissolved in 2ꢀpropanol (1.65 L) and water (100 mL) at reflux. The clear
solution cooled to rt over 18 h and the resulting crystals were isolated, washed and dried to
generate the (ꢀ)ꢀcamphorsulfonic acid salt of (3R,4R)ꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀ
amine (152.6 g, 90% ee). The previously described recrystallization afforded the (ꢀ)ꢀ
camphorsulfonic acid salt of (3R,4R)ꢀ4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀamine as a white
solid (132 g, 36% yield). Chiral HPLC, using a Chiralpak IA column (4.6 × 250 mm) at 0.7
mL/min with heptanes:isopropanol:diethylamine (v/v/v, 80:20:0.2), determined 99.0% ee at 220
nm and 99.7% ee at 260 nm with a t of 12.6 min.
R
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 75
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
cis-N-[4-(4-bromophenoxy)tetrahydrofuran-3-yl]propane-2-sulfonamide (15). 1,8ꢀ
Diazabicyclo[5.4.0]undecꢀ7ꢀene (53 mL, 0.35 mol) was added to a solution of cisꢀ4ꢀ(4ꢀ
bromophenoxy)tetrahydrofuranꢀ3ꢀamine (13, 65.6 g, 0.254 mol) in methylene chloride (500 mL).
The reaction mixture was cooled to 0 °C and propaneꢀ2ꢀsulfonyl chloride (31.2 mL, 0.28 mol)
was added dropꢀwise. The mixture was then stirred at rt for 18 h. Over the course of 20 h, the
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
reaction was monitored by H NMR and additional propaneꢀ2ꢀsulfonyl chloride (2 × 2.8 mL, 25
mmol) was added. Water (500 mL) was added and the layers were separated. The aqueous layer
was extracted with methylene chloride, the combined organic layers were dried over sodium
sulfate and filtered, and the solvent was removed under reduced pressure. The residue was
dissolved in methylene chloride and washed with aqueous 1 N hydrochloric acid (2 × 300 mL),
dried over sodium sulfate and filtered, and the solvent was removed under reduced pressure to
provide cisꢀNꢀ[4ꢀ(4ꢀbromophenoxy)tetrahydrofuranꢀ3ꢀyl]propaneꢀ2ꢀsulfonamide (15, 92.5 g,
1
quantitative) as an orange/brown oil. H NMR (500 MHz, CDCl ) δ 7.41 (d, J = 9.0 Hz, 2H),
3
6.78 (d, J = 9.0 Hz, 2H), 4.81 (d, J = 9.5 Hz, 1H), 4.75 (ddd, J = 5.9, 4.4, 1.8 Hz, 1H), 4.26–4.18
(m, 1H), 4.17–4.11 (m, 2H), 3.93 (dd, J = 10.7, 1.7 Hz, 1H), 3.69 (t, J = 8.7 Hz, 1H), 3.15 (sep, J
1
3
=
6.8 Hz, 1H), 1.38 (d, J = 6.8 Hz, 3H), 1.35 (d, J = 6.8 Hz, 3H); C NMR (125 MHz, CDCl ) δ
3
155.6, 132.7, 117.5, 114.5, 76.3, 71.9, 70.2, 55.5, 54.5, 16.6; Elemental analysis calcd for
C H BrNO S C: 42.87, H: 4.98, N: 3.85, found C: 43.38, H: 4.55, N: 3.49; HRMS (ESIꢀ) m/z
13
18
4
+
[
M + H] calcd for C H BrNO S 364.0213, found 364.0206.
13 18 4
N-[(3S,4S)-4-(4-bromophenoxy)tetrahydrofuran-3-yl]propane-2-sulfonamide (15a).
Compound 15a was analogously prepared as 15. Chiral HPLC, using a Chiralpak IA column
(4.6 × 250 mm) at 0.7 mL/min with heptanes:isopropanol:diethylamine (v/v/v, 80:20:0.2),
determined 99.1% ee at 220 nm and 98.7% ee at 260 nm with a t of 15.2 min. Elemental
R
28
ACS Paragon Plus Environment

Page 29 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
analysis calcd for C H BrNO S C: 42.87, H: 4.98, N: 3.85, found C: 42.93, H: 4.85, N: 3.85;
1
3
18
4
+
HRMS (ESIꢀ) m/z [M + H] calc. for C H BrNO S 364.0213, found 364.0217.
13 18
4
N-{(3S,4S)-4-[4-(5-cyano-2-thienyl)phenoxy]tetrahydrofuran-3-yl}propane-2-
sulfonamide (3). Sulfonamide 15a (0.340 mmol) and (5ꢀcyanoꢀ2ꢀthienyl)boronic acid (0.521
mmol), dicyclohexyl(2',4',6'ꢀtriisopropylbiphenylꢀ2ꢀyl)phosphine (XPhos, 0.034 mmol),
palladium(II) acetate (0.023 mmol) and potassium fluoride (1.71 mmol) were added to a
microwave vial, which was capped and purged three times with nitrogen/vacuum. A 1:1 mixture
of degassed methanol:toluene (10 mL) was added and the reaction was subjected to microwave
irradiation at 130 °C for 30 min. The solvent was removed under reduced pressure and the
residue was partitioned between ethyl acetate and saturated aqueous sodium chloride solution.
The aqueous layer was extracted with ethyl acetate (2 × 20 mL), and the organic layers were
combined, dried over sodium sulfate and filtered. The filtrate was concentrated under reduced
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
pressure to a residue, which was purified by silica gel chromatography to afford 3 (58%) as a
1
solid. LCꢀMS m/z 393.5 (M+1); H NMR (500 MHz, CDCl ) δ 7.58 (d, J = 3.9 Hz, 1H), 7.55 (d,
3
J = 8.8 Hz, 2H), 7.19 (d, J = 3.9 Hz, 1H), 6.95 (d, J = 8.8 Hz, 2H), 4.86 (ddd, J = 5.8, 4.4, 1.7
Hz, 1H), 4.76 (d, J = 9.8 Hz, 1H), 4.30–4.22 (m, 1H), 4.20 (dd, J = 10.7, 4.2 Hz, 1H), 4.17 (t, J =
8.0 Hz, 1H), 3.98 (dd, J = 10.7, 1.7 Hz, 1H), 3.72 (t, J = 8.8 Hz, 1H), 3.17 (sep, J = 6.8 Hz, 1H),
13
1
.40 (d, J = 6.8 Hz, 3H), 1.37 (d, J = 6.8 Hz, 3H); C NMR (125 MHz, CDCl ) δ 157.5, 151.3,
3
138.4, 128.1, 126.5, 122.7, 116.3, 114.4, 107.7, 76.2, 72.0, 70.2, 55.6, 54.7, 16.7, 16.6;
Elemental analysis calcd for C H N O S C: 55.08, H: 5.14, N: 7.14, found C: 55.14, H: 4.99,
1
8
20
2
4 2
ꢀ
+
N: 7.05; HRMS (ESI ) m/z [M + H] calcd for C H N O S 393.0937, found 393.0931; mp
1
8
20
2
4 2
2
0
102–104 °C; [α] = ꢀ89.1° (c = 0.8, CHCl3).
D
29
ACS Paragon Plus Environment