3-Alkoxy-pyrrolo[1,2- b ]pyrazolines as selective androgen receptor modulators with ideal physicochemical properties for transdermal administration

Article abstract of DOI:10.1021/jm5009049

We describe the synthesis and characterization of 3-alkoxy-pyrrolo[1,2-b]pyrazolines as novel selective androgen receptor (AR) modulators that possess excellent physicochemical properties for transdermal administration. Compound 26 bound to human AR with an IC₅₀of 0.7 nM with great selectivity over other nuclear hormone receptors and potently activated AR in a C2C12 muscle cell reporter gene assay with an EC₅₀of 0.5 nM. It showed high aqueous solubility of 1.3 g/L at pH 7.4, and an in silico model as well as a customized parallel artificial membrane permeability assay indicated good skin permeation. Indeed, when measuring skin permeation through excised human skin, an excellent flux of 2 μg/(cm²·h) was determined without any permeation enhancers. In a 2 week Hershberger model using castrated rats, the compound showed dose-dependent effects fully restoring skeletal muscle weight at 0.3 mg/kg/day after subcutaneous administration with high selectivity over prostate stimulation.

Full text of DOI:10.1021/jm5009049

Subscriber access provided by Linköping University Library
Article
3-Alkoxy-pyrrolo[1,2-b]pyrazolines as novel selective
androgen receptor modulators (SARMs) with ideal
physicochemical properties for transdermal administration
Thomas Ullrich, Sanjita Sasmal, Venkatesham Boorgu, Srinivasu Pasagadi, Srisailam
Cheera, Sujatha Rajagopalan, Archana Bhumireddy, Dhanya Shashikumar, Shekar Chelur,
Charamanna Belliappa, Chetan Pandit, Narasimharao Krishnamurthy, Subhendu Mukherjee,
Anuradha Ramanathan, Chakshusmathi Ghadiyaram, Murali Ramachandra, Paulo G
Santos, Bharat Lagu, Mark G Bock, Mark H Perrone, Sven Weiler, and Hansjoerg Keller
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm5009049 • Publication Date (Web): 14 Aug 2014
Downloaded from http://pubs.acs.org on August 15, 2014
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 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
-Alkoxy-pyrrolo[1,2-b]pyrazolines as novel selective androgen receptor modulators
(SARMs) with ideal physicochemical properties for transdermal administration
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
2
2
2
Thomas Ullrich, Sanjita Sasmal, Venkatesham Boorgu, Srinivasu Pasagadi, Srisailam
2
3
3
2
3
3
4
Cheera, Sujatha Rajagopalan, Archana Bhumireddy, Dhanya Shashikumar, Shekar Chelur,
3
3
3
Charamanna Belliappa, Chetan Pandit, Narasimharao Krishnamurthy, Subhendu Mukherjee,
3
3
3
Anuradha Ramanathan, Chakshusmathi Ghadiyaram, Murali Ramachandra, Paulo G. Santos,
5
5
5
1
*,6
Bharat Lagu, Mark G. Bock, Mark H. Perrone, Sven Weiler, Hansjoerg Keller
1
6
Global Discovery Chemistry and Musculoskeletal Diseases, Novartis Institutes for BioMedical
Research, CH-4002 Basel, Switzerland
2
Aurigene Discovery Technologies Ltd, Bollaram Road, Miyapur, Hyderabad 500 049, India
3
Aurigene Discovery Technologies Ltd, 39-40, KIADB Industrial Area, Electronic City Phase II,
Hosur Road, Bangalore 560 100, India
4
Technical Research and Development, Novartis Pharma AG, CH-4002 Basel, Switzerland
5
Global Discovery Chemistry, Novartis Institutes for BioMedical Research, Cambridge MA
02139, USA
ABSTRACT: We describe the synthesis and characterization of 3ꢀalkoxyꢀpyrrolo[1,2ꢀ
b]pyrazolines as novel selective androgen receptor (AR) modulators that possess excellent
physicochemical properties for transdermal administration. Compound 26 bound to human AR
with an IC of 0.7 nM with great selectivity over other nuclear hormone receptors and potently
5
0
activated AR in a C2C12 muscle cell reporter gene assay with an EC of 0.5 nM. It showed high
50
aqueous solubility of 1.3 g/L at pH 7.4 and an in silico model as well as a customized parallel
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 50
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
artificial membrane permeability assay indicated good skin permeation. Indeed, when measuring
2
skin permeation through excised human skin an excellent flux of 2 µg/(cm *h) was determined
without any permeation enhancers. In a twoꢀweek Hershberger model using castrated rats, the
compound showed doseꢀdependent effects fully restoring skeletal muscle weight at 0.3
mg/kg/day after subcutaneous administration with high selectivity over prostate stimulation.
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
INTRODUCTION
Androgens are essential for development, maturation and functional maintenance of the male
1
,2
reproductive system. In addition, they establish secondary sex characteristics such as facial hair
growth and vocal cord enlargement. Furthermore, they exert powerful anabolic actions on bone
3ꢀ5
and skeletal muscle, and decrease fat mass. The two main endogenous androgens in men are
testosterone (T) and its more potent metabolite 5αꢀdihydrotestosterone (DHT) generated by 5αꢀ
reductase. Most of their biological effects are mediated by binding to a single androgen receptor
(AR) belonging to the nuclear receptor family of transcription factors. During male aging, T
levels decrease about 1% per year which is associated with a gradual loss of muscle mass and
6
ꢀ9
function, and development of frailty. Moreover, falling T levels often lead to late onset
hypogonadism (LOH) in elderly men that is characterized by low serum T levels, fatigue and
1
0
decreased libido. Current therapies for LOH and other forms of male hypogonadism include
11
different T hormone treatments. However, the potential risks of T replacement such as prostate
1
2
and cardiovascular safety are currently being debated. Such concerns as well as potential
virilizing effects in women have precluded the use of high doses of androgens as anabolics in
muscle wasting diseases such as cancer cachexia and sarcopenia. To avoid this, selective
androgen receptor modulators (SARMs) have been developed that selectively stimulate anabolic
ACS Paragon Plus Environment

Page 3 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
actions of AR in muscle and bone without inducing androgenic action in tissues such as prostate
1
3,14
and skin.
Several steroidꢀderived and nonꢀsteroidal SARMs, all oral formulations, have entered clinical
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
5ꢀ18
evaluation (Figure 1).
1 (enobosarm), 2 (MKꢀ0773) and LGDꢀ4033 (structure not disclosed)
increased lean body mass in volunteers, and 1 improved physical function in healthy elderly men
1
5,16
and women, and in cancer patients with muscle wasting
However, the clinical potential of
oral SARMs appears to be limited by the occurrence of adverse events at highly efficacious
1
5,18
doses, e.g. induction of alanine aminotransferase (ALT)
and reduction of highꢀdensity
1
5,17
lipoprotein (HDL) in plasma.
It is therefore apparent that successful prevention of androgenic
tissue stimulation by a SARM does not eliminate other putatively onꢀtarget adverse effects.
Induction of signals generally associated with hepatotoxicity and unfavorable changes of plasma
1
9ꢀ21
lipoproteins are well known adverse events of oral androgen therapies.
Current T replacement
therapies often use parenteral treatments such as the most frequently applied transdermal patch
and gel delivery systems, which generally do not show adverse lipid changes and hepatic side
2
0,22
effects.
With the objective to minimize high drug exposure in the liver and to provide
constant drug levels to the body we decided to fully exploit the potential of SARMs to be
administered by transcutaneous routes of administration.
Figure 1. Exemplary steroidal and nonꢀsteroidal SARMs that have been evaluated in clinical trials
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 50
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
The ideal characteristics for passive diffusion of a compound through skin are not necessarily the
same as those promoting gastrointestinal absorption. In general, the drugs currently administered
across skin fall within more stringent limits than predicted by the Rule of Five, which is
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
commonly used to guide optimization for oral absorption. Good guidelines have, however,
emerged in recent years based on the wealth of drugs that have been considered – with or without
success – for transdermal delivery. Ideal molecules should be small (molecular weight MW ≤400
Da), have a low melting point, high aqueous solubility, but at the same time display a balanced
lipophilicity profile (partition coefficient logPoctanol/water ~2). Furthermore, the number of
hydrogen bond donors/acceptors as well as ionizable groups which impact permeation through
the cutaneous lipids at the pH of skin (it is estimated that the natural skin surface pH is on
2
4
average 4.7) need to be kept at a minimum. Finally, as the permeationꢀlimited flux through skin
will only allow a certain drug load to reach the systemic circulation over time, the optimal
transdermal drug also has to be highly potent and efficacious in vivo to warrant low doses
triggering the desired pharmacodynamic effect. Most of the drugs currently available in patch
25
formulation share a very low required daily dose (<2 mg).
Taking into consideration the physicochemical property and in vivo efficacy requirements for
transdermal candidates as mentioned above, we reviewed well described clinical and preꢀclinical
SARMs. 3 (BMSꢀ564929, Figure 1) caught our particular attention because of its small size (MW
2
6
=
305) and high potency in vitro and in vivo. However, the low calculated logP (ClogP = 0.11)
2
7
and low reported aqueous solubility (0.019 mg/mL) of 3 would predict only limited
transcutaneous permeation for the molecule. Many hydantoinꢀbased drugs suffer from low
28
aqueous solubility, especially when the hydantoin is attached to aromatic rings. Li et al. argued
that removal of one of the carbonyl groups from the hydantoin core would eliminate interꢀ
ACS Paragon Plus Environment

Page 5 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
27
molecular hydrogen bonding between two individual molecules in the crystal lattice. Further
elaborating on this concept we conceived a scaffold which would retain many of the advantages
of 3 (low MW, high potency) and provide additional benefits with respect to exploring novel
chemical space while not being susceptible to lactam or imide ring opening giving rise to
potentially mutagenic aniline fragments. Pyrazolines were thought to be appropriate hydantoin
replacements, as they can contribute to polar interactions by virtue of their nitrogen atoms
without being charged at physiological pH. The privileged substitution pattern on the phenyl ring
described by Li et al. remained unchanged in the course of this study, based on its beneficial
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
7
impact on biological activity. In this paper, we describe the synthesis and biological evaluation
of 3ꢀalkoxyꢀpyrrolo[1,2ꢀb]pyrazolines, a novel bicyclic SARM scaffold devoid of the hydantoin
subꢀstructure and with physicochemical properties suited for transdermal administration (Figure
2
).
Figure 2. Restructuring the hydantoin scaffold
CHEMISTRY
The first prototype was synthesized from transꢀ3ꢀhydroxyꢀLꢀproline methyl ester (4) (Scheme 1).
Nꢀnitrosylation to 5, followed by protection of the alcohol and chemoselective ester reduction
with Superhydride, led to Nꢀnitrosoꢀprolinol 7, which then underwent zinc mediated reduction to
hydrazine intermediate 8.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 50
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 1: Synthesis of hydrazine intermediate 8 (2R,3S)
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
Reagents and conditions: (a) NaNO , acetic acid, H O, 0 °C, 4 h, 68%; (b) TBSꢀchloride, imidazole, 25 °C, 4 h,
2
2
9
3%; (c) LiEt BH, THF, ꢀ78 °C, 5 h, 58%; (d) Zinc, NH Cl, methanol, H O, 50 °C, 90 min, 8%.
3 4 2
In a similar fashion, hydrazine intermediate 13 (2S,4R) was synthesized from Lꢀhydroxyproline
methyl ester (9) (Scheme 2):
Scheme 2: Synthesis of hydrazine intermediate 13 (2S,4R)
Reagents and conditions: (a) NaNO , acetic acid, H O, 0 °C, 4 h, quant.; (b) TBSꢀchloride, imidazole, 25 °C, 4h,
2
2
87%; (c) LiEt BH, THF, ꢀ78 °C, 5 h, 67%; (d) Zinc, NH Cl, methanol, H O, 50 °C, 90 min, 97%.
3
4
2
2
7,29
Intermediate 8 was treated with 2ꢀchloroꢀ4ꢀformylꢀ3ꢀmethylbenzonitrile (14)
to give
hydrazone 15, which was then subjected to Swern oxidation of the alcohol to furnish aldehyde
3
0
1
6. Lewisꢀacid catalyzed stereospecific cyclization to intermediate 17 was then accomplished in
good yield (Scheme 3).
Scheme 3: Synthesis of pyrazoline intermediate 17
ACS Paragon Plus Environment

Page 7 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reagents and conditions: (a) acetic acid, 25 °C, 3 h, 66%; (b) oxalyl chloride, DMSO, DCM, ꢀ78 °C, 30 min, 94%;
(
c) BF .OEt , DCM, 0 °C, 3 h, 60%.
3 2
Intermediate 13 was treated in an analogous manner to give annulated pyrazoline 20 (Scheme 4).
Scheme 4: Synthesis of pyrazoline intermediate 20
Reagents and conditions: (a) acetic acid, 25 °C, 3 h, 44%; (b) oxalyl chloride, DMSO, DCM, ꢀ78 °C, 30 min, quant.;
(
c) BF .OEt , DCM, 0 °C, 3 h, 50%.
3 2
Deprotection of 17 yielded diol 21, and analogues were synthesized by 3ꢀOꢀalkylation of 17,
followed by cleavage of the silyl protective group (Scheme 5). The absolute configuration of
methyl derivative 26 was unequivocally proven by coꢀcrystallization with the ligand binding
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 50
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
domain of the AR (Fig. 3), whereby (S)ꢀconfiguration could be assigned to each of the
stereocenters.
Scheme 5: Alkylation and deprotection of the pyrazoline intermediate 17
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
Reagents and conditions: (a) TBAF, THF, 25 °C, 3 h, 28%; (b) exemplary conditions for compound 22: NaH,
iodomethane, THF, 0 °C, 3 h, 82%; (c) exemplary conditions for compound 26: TBAF, THF, 0 °C, 3 h, 62%.
ACS Paragon Plus Environment

Page 9 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Xꢀray coꢀcrystal structure of 26 bound to the AR ligand binding domain at 2.1 Å resolution. Key
interactions between 26 and residues in the active site are highlighted by dashed lines indicating polar interactions.
Numerical values shown for the distances are in Å. The cyano group of 26 interacts with the side chain of R752,
while the 5ꢀhydroxyl group presumably forms a bifurcated hydrogen bond with N705 and T877. The aryl ring with
the chloro and methyl substituents being surrounded by hydrophobic residues makes an edgeꢀtoꢀface πꢀinteraction
with F764.
The free 4ꢀhydroxyl group of 26 could be manipulated in various ways (Scheme 6): inversion of
the steric configuration to 30 was accomplished under Mitsunobu conditions, 4ꢀOꢀalkylation to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 50
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
3
1 could be done by deprotonation and treatment with methyl iodide. Fluorination of 30 and 26,
respectively, with DAST afforded 32 and 33.
Scheme 6: Inversion of configuration, fluorination and methylation of 26
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
Reagents and conditions: (a) Triphenylphosphine, diisopropyl azodicarboxylate, benzoic acid, THF, 25 °C, 3 h, 50%;
(
b) NaH, iodomethane, THF, 25 °C, 2 h, 20%, (c) DAST, DCM, 0 °C, 2 h, 33%; (d) DAST, DCM, 0 °C, 2 h, 40%.
Intermediate 20 was methylated (34) and treated with TBAF to furnish the 5ꢀhydroxy derivative
35, which was subjected to inversion of the 5ꢀOH group to give 36, fluorination to 37, and
oxidation to ketone 38, which was used to introduce a gemꢀdifluoro group into the molecule (39)
(Scheme 7).
Scheme 7: Derivatization of Intermediate 20
ACS Paragon Plus Environment

Page 11 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reagents and conditions: (a) NaH, iodomethane, THF, 25 °C, 2 h, 77%, (b) TBAF, THF, 25 °C, 3 h, 69%; (c)
Triphenylphosphine, benzoic acid, THF, 25 °C, 3 h, 28%; (d) DessꢀMartin periodinane, DCM, 25 °C, 24 h, 86%; (e)
DAST, DCM, 0 °C, 2 h, 20%; (f) DAST, DCM, 0 °C, 16 h, 32%.
RESULTS AND DISCUSSION
In Vitro Biological Data. Primary agonism of the 3ꢀalkoxyꢀpyrrolo[1,2ꢀb]pyrazolines was
measured in an AR reporter gene assay using the murine myoblast cell line C2C12 transiently
transfected with human AR and an androgenꢀspecific response element driven luciferase reporter
gene. Table 1 summarizes potency (EC ) and maximal efficacy relative to DHT (Emax) of
5
0
compounds observed in this assay.
a
Table 1: AR agonist activity of derivatives 21ꢀ39
Compound
DHT
R1
R2
R3
C2C12 EC [nM]
C2C12 Emax [%]
50
0.13 ± 0.02
100
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 50
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
2
2
2
2
2
3
3
3
3
3
3
3
3
3
1
6
7
8
9
0
1
2
3
5
6
7
8
9
H
Me
Et
(S)ꢀOH
(S)ꢀOH
(S)ꢀOH
(S)ꢀOH
(S)ꢀOH
(R)ꢀOH
(S)ꢀOMe
(S)ꢀF
(R)ꢀF
H
H
H
978 ± 221
0.5 ± 0.4
1 ± 0.5
55 ± 21
97 ± 13
98.3 ± 2.9
102 ± 2.8
82 ± 10
86.7 ± 11
82 ± 4.2
93.5 ± 2.1
98 ± 3
H
iꢀPr
MeOEt
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
4.5 ± 0.1
0.93 ± 0.3
0.4 ± 0.2
7.1 ± 1.1
1.5 ± 0.3
0.16 ± 0.06
9.7 ± 4.2
inactive
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
H
H
H
H
H
(R)ꢀOH
(S)ꢀOH
(S)ꢀF
O
75 ± 9.5
ꢀ
H
H
1.4 ± 0.4
26 ± 2.4
0.3 ± 0.2
86 ± 19
93 ± 7
H
H
diꢀF
100 ± 0
a
EC50 and Emax values are given as the means ± sem of at least two independent experiments. Emax values were
determined at 10 µM and were normalized by arbitrarily taking maximal efficacy of DHT as 100%.
3
,4ꢀDiol 21 proved to be a weak, partial AR agonist at low micromolar concentrations only. 3ꢀOꢀ
Alkylation however, rendered the compounds highly potent (EC <5 nM) and fully efficacious
50
(Emax ~100%) as shown for 26-28. Linear alkyl chains were preferred over isopropyl; the
solubilizing methoxyethyl chain (29) was well tolerated. Inversion of the hydroxyl group in
position 4 (30) retained potency (albeit with lower intrinsic efficacy compared to 26), in contrast
3
1
to the hydantoin SAR demonstrated previously, but in agreement with the imidazolinꢀ2ꢀone
27
SAR reported by Li et al. Docking 30 into the AR ligand binding domain reveals that its 4ꢀOH
group is expected to interact only with Asn705 and not Thr877 (Figure 4ꢀa), suggesting that the
Thr877 interaction may not be essential for biological activity. Oꢀmethylation in that position
(
31) led to roughly a 13ꢀfold loss of potency. Conversion of the 4ꢀOH group into a fluoro
ACS Paragon Plus Environment

Page 13 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
substituent led to the most potent compound of the series (33) with an EC of 0.16 nM, whereas
5
0
its epimer 32 showed approximately 10ꢀfold lower activity. The docked binding mode of 33
(Figure 4ꢀb) demonstrates that its fluorine atom is likely to form a halogen bond with Asn705 and
Thr877, thereby allowing a binding pocket orientation very similar to that of 26. A switch of the
hydroxyl group from position 4 to 5, conserving the steric configuration, led to a more than 20ꢀ
fold decrease in potency (35), again accompanied by a decrease in Emax. AR activity was
completely lost after inversion of the OHꢀconfiguration (36) but fully restored by replacing the
hydroxyl group with a fluorine substituent (37). Ketone 38 was moderately active (EC = 26
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
5
0
nM), and geminal difluoroꢀsubstitution in position 5 led to the highly active agonist 39. The
docked binding mode of 39 (Figure 4ꢀc) suggests that this compound may not interact with the
Asn705 side chain but that the geminal difluoro substituent makes several vanꢀderꢀWaals
contacts with different hydrophobic residues like Leu701, Leu704, or Phe876. This possibly
compensates for the missing Asn705 interaction. Overall, these results suggest that the hydroxyl
group of the pyrrolidine ring, which we installed based on the published hydantoin SAR, is not
critical for AR agonist activity either in position 4 or 5. The pivotal role of the hydroxyl group,
however, in modulating the physicoꢀchemical properties in this series will be discussed in the
following section.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 50
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
Figure 4. Crystallographic binding mode of 26 (pink) compared with docked models of (a) 30,
(b) 33 and (c) 39. Arg752 pocket interactions are not shown.
Skin Permeation Assessment. Some of the most potent and efficacious compounds were
selected for assessing their transdermal permeation. First, we determined their thermodynamic
aqueous solubility at pH 7.4 and logP (Table 2). Using these two parameters together with
molecular weight (MW), skin permeability (K ) and transdermal flux (calcd. J) were predicted
p
32
according to the Potts and Guy relationship:
2
3
J [ꢀg/(h*cm )] = SolpH7.4 [ꢀg/cm ] * K [cm/h], and log K [cm/h] = ꢀ2.7 + 0.71logP –
p
p
0.0061MW
ACS Paragon Plus Environment

Page 15 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
However, it should be noted that these estimations are based on aqueous formulations. Thus,
changing formulation will change the permeability and solubility of a drug. Furthermore, there is
a tendency to overestimate the permeability of ionized and lipohilic compounds. Recently, an in
vitro assay was established with artificial membranes consisting of silicone and isopropyl
myristate (70:30) that mimic the barrier properties of human stratum corneum, the highest barrier
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
3,34
for transdermal drugs.
This assay is from here onwards referred to as “skinꢀPAMPA” and
permits an improved estimation of the passive permeation and distribution of a drug in human
skin. In addition, the skinꢀPAMPA assay allows to discriminate molecules with low membrane
retention and low permeation (category A), low membrane retention and high permeation
(category B), and high membrane retention and low permeation (category C).
Finally, the most predictive in vitro assay for human skin flux is to measure skin permeation
3
5
using human cadaver skin in a Franz diffusion cell system. It is the standard in vitro test
adopted by the Organization for Economic CoꢀOperation and Development (OECD) for
3
6
examining the skin absorption of chemicals. Franz diffusion cells consisting of donor and
receptor compartments separated by the skin samples are commonly used for determining the
drug permeation in vitro by measuring the amount of drug permeated over time from the donor
chamber through the skin and into the acceptor compartment. All studies were performed with
saturated aqueous donor solutions in order to maximize drug permeation and allow comparison
between compounds.
Below, we are describing the selection of our best SARM candidates using all three methods, as
summarized in Table 2.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 50
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
Table 2: Physicoꢀchemical profiling and in vitro skin permeation
Sol pH 7.4
[mg/mL]
logP
skin PAMPA
category
calcd. Jss
measured Jss
MW
Cpd
2
2
(oct/water)
[ꢀg/(cm *h)]
[ꢀg/(cm *h)]
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
T
288.4
0.04
1.30
0.25
3.3
2.0
2.4
B
A/B
B
0.27
0.85
0.26
0.19 ± 0.05
2.0 ± 0.2
0.9 ± 0.1
2
6
305.8
2
7
319.8
2
9
349.8
2.3
2.0
n.d.
0.81
1.1 ± 0.3
3
0
305.8
0.25
1.9
n.d.
0.10
0.2 ± 0.0
3
3
307.8
0.010
2.7
B/C
0.02
1.2 ± 0.1
ACS Paragon Plus Environment

Page 17 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
39
325.8
0.010
3.2
C
0.04
0.06 ± 0.04
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Skin flux values at steadyꢀstate (J ) are given as the means ± sem of at least three independent experiments.
SS
Solubility and logP values were determined in single experiments, respectively.
A thorough analysis of the results shown in Table 2 reveals that compounds 33 and 39, both
devoid of the 4ꢀhydroxy group, are only poorly soluble in water at neutral pH (Sol = 0.01 mg/mL,
respectively). At the same time, they display the highest lipophilicity, as expressed by their
measured logP values (2.7 and 3.2, respectively). Both parameters are subꢀoptimal for skin
permeation (see above) and for the difluoro derivative 39 this is clearly reflected by the outcome
in the skinꢀPAMPA assay (category C), indicating high membrane retention (due to the highly
lipophilic character) and low permeation into the acceptor compartment (due to the lack of
solubility). Both compounds rank lowest in the in silico prediction of skin flux (calcd. J <0.05
2
ꢀ
g/(cm *h)), and while this was confirmed for 39 in the in vitro Franz cell assay, surprisingly 33
had a 60ꢀfold higher skin flux in the in vitro assay than predicted. For 33, the skinꢀPAMPA assay
appeared to be more predictive, as its classification delivered results between B and C, indicating
at least moderate permeation.
Gratifyingly, a pyrrolo[1,2ꢀb]pyrazoline which combined high aqueous solubility (26) (Sol = 1.30
mg/mL) with a very favorable lipophilicity coefficient (logP = 2.0) emerged as the molecule with
2
the highest in vitro skin flux at steady state (J = 2.0 ꢀg/(cm *h)) of all tested compounds,
ss
2
exceeding predictions by the in silico method (J = 0.85 ꢀg/(cm *h)) and the skinꢀPAMPA
ss
(category A/B). The exceptionally high aqueous solubility, which exceeds that of 3 by 70ꢀfold,
could be attributed to a distinct crystalline form as a mixture of two polymorphs melting at 83 °C
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 50
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 89 °C, respectively. Figure 5 demonstrates skin permeation of 26 over time, resulting in a
linear cumulation. The observed skin flux at steady state resembles that of known testosterone
preparations, where skin permeation enhancers have been used to enhance the systemic
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
37
bioavailability. Notably, the aqueous formulation of 26 in this experiment was not treated with
any permeation enhancer, underlining the high potential of the compound to achieve even higher
flux values when using appropriate additives.
Figure 5: Compound 26 penetrates through human skin from a saturated aqueous solution
2
buffered at pH 7.4. The observed flux at steady state (J ) is 2.0 ± 0.2 ꢀg/(cm *h), based on three
ss
different donor skin samples. The dotted line represents the calculated fitted curve using diffusion
modeling simulations.
Its closest analogue, epimer 30, was inferior in terms of solubility (Sol = 0.25 mg/mL) and in
2
vitro skin flux (J = 0.2 ꢀg/(cm *h)). 3ꢀOꢀalkyl chain extension by one carbon (27) reduced both
ss
solubility and in vitro skin flux to some extent, while the solubilizing methoxyethyl group of 29
resulted in the highest aqueous solubility measured in this series (Sol = 2.30 mg/mL) but not
adequately superior skin flux. It could be clearly concluded that both models of skin permeation
prediction, the Guy & Potts algorithm as well as the artificial membrane assay, had their
ACS Paragon Plus Environment

Page 19 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
limitations when looking at relatively similar molecules. Only highly lipophilic and poorly
soluble compounds such as 39 were correctly classified as weak permeants. Based on the
favorable potency and predicted skin permeation potential, 26 emerged as an attractive candidate
for additional profiling.
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
Selectivity of 26 against other nuclear hormone receptors was determined by measuring binding
affinity for AR, the glucocorticoid receptor (GR), progesterone receptor (PR), and estrogen
receptor (ER). The compound showed high selectivity for AR, as shown in Table 3.
a
Table 3: Binding selectivity for 26 in the nuclear hormone receptor panel
AR
GR
PR
ER
IC [nM]
0.7±0.0
>1000
700
>1000
5
0
a
The IC50 value for AR is given as the mean ± sem of two independent experiments, while affinities for GR, PR and
ER were determined in single experiments, respectively.
Pharmacokinetics. The pharmacokinetics of the compound with the best predicted and measured
flux through human skin, 26, was analyzed in male Wistar rats using subcutaneous (s.c.) bolus
injection as a surrogate for transdermal systemic delivery. Drug exposure over 24 hours was
compared to the AUC after intravenous (i.v.) injection. Results are summarized in Table 4. 26
showed moderateꢀtoꢀhigh clearance but high subcutaneous bioavailability, slow absorption from
subꢀdermal compartments, and a long elimination halfꢀlife after reaching peak plasma
concentrations after 1.5 h. This profile was considered highly advantageous, as one could expect
a similar exposureꢀoverꢀtime profile from a transdermal patch.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 50
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 4: Pharmacokinetics of 26 in male Wistar rats
i.v. injection (1.0 mg/kg)
s.c. injection (0.5 mg/kg)
CL [L/h.kg]
1.7
5.7
4.6
708
tmax [h]
1.5
79
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
t1/2 [h]
Cmax [ng/mL]
AUC0ꢀ48 [ng.h/mL]
BAV [%]
V [L/kg]
D
670
75
AUC0ꢀinf [ng.h/mL]
3
8
In Vivo Efficacy. 26 was evaluated for efficacy and tissue selectivity in the Hershberger assay.
This is a wellꢀestablished animal model to assess the anabolic versus androgenic effects of
compounds by analyzing weight stimulations of bulbocavernosus/levator ani (BLA) muscle
versus prostate in immature orchidectomized (ORX) male rats, respectively. Two weeks after
castration, rats were treated with subcutaneous injections of ascending doses of 26 for 14 days.
As a control for nonꢀselective androgen action, rats were also treated with a subcutaneous dose
range of testosterone propionate (TP). Figure 6 shows dose response curves of 26 and TP on
BLA muscle and prostate related to vehicleꢀtreated ORX and sham animals. TP stimulated BLA
and prostate weights comparably with very similar ED50 values of 0.29 and 0.31 mg/kg,
respectively. In contrast, 26 recovered BLA weight much more potently than prostate weight with
ED values of 0.07 and 2.73 mg/kg, respectively. Thus, a window of selective stimulation of
5
0
muscle versus prostate was observed over two log units of doses from 0.03 to 3 mg/kg with
maximal selectivity at 0.3 mg/kg, where muscle was fully restored and prostate only minimally
stimulated. Analysis of plasma levels of 26 three hours after the last administration on day 15
revealed a doseꢀproportional drug exposure indicating no saturation or accumulation during the
2ꢀweek treatment period.
ACS Paragon Plus Environment

Page 21 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
26
b
TP
Prostate
BLA
Prostate
BLA
150
150
100
50
100
50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
0
1
0-3
10-2
10-1
100
101
102
10-3
10-2
10-1
100
101
102
Dose (mg/kg)
Dose (mg/kg)
Figure 6: Hershberger in vivo assay of 26. Doseꢀresponse curves of 26 (a) and TP (b) as control
for the stimulation of BLA muscle and prostate weights in ORX rats. Relative weights are shown
by normalizing to vehicleꢀtreated ORX rats taken as 0% and vehicleꢀtreated Sham rats taken as
100%. (c) Plasma levels of 26 3 h after the last dosing on day 15.
2
6
The in vivo efficacy and selectivity of 26 show some resemblance to the profile described for 3.
This may be due to the similarity of the binding modes suggested by the respective Xꢀray coꢀ
crystal structures. It can been speculated whether certain molecular features, e.g. the peripheral
hydroxyl group which apparently forms a hydrogen bond with T877, may impact the
conformation of the activator functionꢀ2 (AFꢀ2) pocket. Such conformational changes are thought
to influence the ability of the liganded AR to recruit coꢀactivator and coꢀrepressor proteins that
show a differential expression pattern in reproductive and anabolic tissues, resulting in different
degrees of tissue selectivity in vivo. However, a clear and uniformly accepted link between
molecular structure and phenotypic observations has been elusive to date. Continuing efforts in
our group are addressing the role of such molecular features, in particular the 4ꢀOH group, in the
class of pyrrolo[1,2ꢀb]pyrazolines.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 50
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
CONCLUSIONS
This study validated the concept of transdermal drugꢀlikeness applied on a novel class of SARMs
using appropriate design parameters, prediction models and assays. Based on its high potency,
selectivity and efficacy in the Hershberger model, and its ability to permeate through human skin
in vitro, 26 holds promise as a novel transdermal AR modulator for the treatment of various
muscle wasting disorders potentially avoiding essential limitations of oral SARMs.
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
EXPERIMENTAL SECTION
Chemistry. Unless otherwise stated, all procedures were performed under an argon atmosphere
using anhydrous solvents purchased from commercial sources. Commercially available starting
materials and reagents were used as received without further purification. Intermediates were
characterized by electrospray mass spectroscopy (ESꢀMS) and 1H NMR except when not
purified due to the lack of chemical stability and used as crude materials in the subsequent steps
(ESꢀMS only). Final compounds were characterized by melting point, 1H NMR, liquid
chromatography/mass spectroscopy (LC/MS) and chiral HPLC with respective retention times
(tR) as well as diasteromeric excess values given. Thinꢀlayer chromatography was performed
using preꢀcoated Merck silica gel 60 F254 plates, visualized under UV light. Preparative
chromatography was conducted over Merck silica gel (0.040ꢀ0.063 mm). All 1H NMR spectra
were recorded using a Varian Mercury (400 MHz) spectrometer, and samples were dissolved in
the deuterated solvents as specified. Chemical shifts are expressed in parts per million as a δ
value relative to the shift of tetramethylsilane. Infrared spectra were recorded on a Shimadzu, FTꢀ
IR Prestige 21 spectrometer. LC/MS was performed on an Applied Biosciences APIꢀ3000
instrument with a photodiode array detector coupled to a mass spectrometer with electrospray
ionization. The following HPLC method was chosen: Waters Xterra RP18 (3.5 ꢀm, 50 x 2.1 mm
ACS Paragon Plus Environment

Page 23 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
i.d.); (A) water + 0.01% of formic acid and (B) acetonitrile + 0.01% of formic acid; gradient from
9
5/5 A/B at time 0 to 100% B after 4 min, then 100% B for 3 min. The purity of the described
test compounds was at least 95% according to HPLC analysis. Chiral HPLC was performed on a
ChiralpakꢀIA (250*4.6*5.0ꢁ) using the following conditions: mobile phase A: nꢀhexane; mobile
phase B: ethanol, isocratic: 70:30 (A:B), flow rate: 1.0 mL/min, column temperature: ambient,
diluent: ethanol. Monitoring wavelengths are indicated for each compound, and diastereomeric
excess (de) values are given. Melting points were measured on a Buchi melting point apparatus
Bꢀ540, except for 26, where the melting point was determined by differential scan calorimetry
using a TA Instruments Q2000 calorimeter calibrated for temperature and enthalpy with indium
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
(>99.9999% pure).
(2R,3S)-1-Amino-3-(tert-butyldimethylsilyloxy)pyrrolidin-2-yl)methanol (8).
2
6
Step 1: (2S,3S)-Methyl 3-hydroxy-1-nitrosopyrrolidine-2-carboxylate (5). To a solution of 4 (16
g, 0.110 moles) in water (60 mL) was added NaNO (16 g, 0.220 moles) in water (30 mL)
2
followed by the addition of glacial acetic acid (9.43 mL, 0.165 moles) at 0 °C, and the reaction
mixture was stirred for 4 h. Once the starting material had disappeared (monitored by TLC), the
reaction mixture was diluted with ethyl acetate. The organic layer was washed with water, brine,
dried over Na SO and concentrated to get 13 g (68%) the title compound which was used in the
2
4
+
next step without purification. ESꢀMS: m/z 175 (M + H) .
Step 2: (2S,3S)-Methyl 3-(tert-butyldimethylsilyloxy)-1-nitrosopyrrolidine-2-carboxylate (6). To a
solution of 5 (13 g, 0.0742 moles) in DCM (150 mL) at room temperature were added imidazole
(15.2 g, 0.224 moles) and TBDMSꢀCl (22.5 g, 0.149 moles) and the mixture was stirred at room
temperature for 3 h. The reaction mixture was diluted with DCM and washed with water, brine,
dried over Na SO and concentrated. Purification by column chromatography over silica gel
2
4
1
using 10% ethyl acetate in hexane provided 20 g (93%) of the title compound. H NMR (400
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 50
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
MHz, DMSOꢀd ) (mixture of rotamers): δ 5.19 – 5.17 (m) and 4.72 – 4.70 (m, 1H), 4.58 – 4.54
6
(m) and 4.17 – 4.16 (m, 1H), 4.57 – 4.41 (m, 1H), 3.75(s) and 3.65 (s, 3H), 3.63 – 3.58 (m, 1H),
2.28 – 2.23 (m, 1H), 2.10 – 2.00 (m, 1H), 0.84 (s, 9H), 0.11 (d, J = 4.4 Hz, 3H), 0.07 (d, J = 4.7
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
+
Hz, 3H): ESꢀMS: m/z 289 (M + H) .
Step 3: (2R,3S)-3-(tert-Butyldimethylsilyloxy)-1-nitrosopyrrolidin-2-yl)methanol (7). To a
solution of 6 (20 g, 0.0692 moles) in THF (350 mL) at ꢀ78 °C was added Superhydride (207 mL,
0.207 moles, 1M Solution in THF) and the reaction mixture was stirred at room temperature for 5
h. Then it was poured over ice cold water and extracted with ethyl acetate. The organic layer was
washed with water, brine, dried over Na SO and concentrated. Purification by column
2
4
chromatography over silica gel using 30% ethyl acetate in hexane provided 10.5 g (58%) of the
1
title compound. H NMR (400 MHz, DMSOꢀd ): δ 5.08 (t, J = 5.9 Hz, 1H), 4.43 – 4.42 (m, 1H),
6
4.21 – 4.09 (m, 1H), 3.66 – 3.59 (m, 2H), 3.57 – 3.40 (m, 2H), 2.16 – 2.07 (m, 1H), 1.75 – 1.72
(m, 1H), 0.76 (s, 9H), 0.03 (s, 3H), 0.02 (s, 3H) ; IR (KBr): 3309, 2953, 2929, 2858, 1469 cmꢀ1;
+
ESꢀMS: m/z 261 (M + H) .
Step 4: (2R,3S)-1-Amino-3-(tert-butyldimethylsilyloxy)pyrrolidin-2-yl)methanol (8). To a solution
of 7 (9.0 g, 0.034 moles) in methanol (140 mL) at room temperature were added zinc dust (23 g,
0
.353 moles) and NH Cl (28 g, 0.529 moles). After the addition was over, water (70 mL) was
4
added at 0 °C and the reaction mixture was stirred at room temperature for 30 min and then at 50
C for 1 h. Once the starting material had disappeared (monitored by TLC), the reaction mixture
°
was cooled to room temperature, filtered and the filtrate was concentrated to give a liquid which
was diluted with ethyl acetate. Organic layer was washed with water, brine, dried over Na SO
2
4
and concentrated to get 7 g (83%) of the title compound, which was not stable enough to conduct
NMR or LC data. The material was used with MS characterization only and without further
+
purification to prepare compound 15. ESꢀMS: m/z 247 (M + H) .
ACS Paragon Plus Environment

Page 25 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
(2S,4R)-1-Amino-4-(tert-butyldimethylsilyloxy)pyrrolidin-2-yl)methanol (13).
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
3
9
Step 1: (2S,4R)-Methyl 4-hydroxy-1-nitrosopyrrolidine-2-carboxylate (10). To a solution of 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
2
3
4
5
6
7
8
9
0
(
55.2 g, 0.381 moles) in water (350 mL) was added NaNO (52.5 g, 0.762 moles) in water (150
2
mL) followed by the addition of glacial acetic acid (34.3 mL, 0.571 moles) at 0 °C and the
reaction mixture was stirred for 4 h. Once the starting material disappeared (monitored by TLC),
the reaction mixture was diluted with ethyl acetate. The organic layer was washed with water,
brine, dried over Na SO and concentrated to get 66.3 g (100%) the title compound which was
2
4
+
used in the next step without purification. ESꢀMS: m/z 175 (M+H) .
Step 2: (2S,4R)-Methyl 4-(tert-butyldimethylsilyloxy)-1-nitrosopyrrolidine-2-carboxylate (11). To
a solution of 10 (66 g, 0.381 moles) in DCM (1.3 L) at room temperature were added imidazole
(
77.7 g, 1.143 moles) and TBDMSꢀCl (115 g, 0.762 moles) and the mixture was stirred at room
temperature for 3 h. The reaction mixture was diluted with DCM and washed with water, brine,
dried over Na SO and concentrated. Purification by column chromatography over silica gel
2
4
1
using 10% ethyl acetate in hexane provided 95 g (87%) of the title compound. H NMR (400
MHz, DMSOꢀd ): δ 4.64 ꢀ 4.54 (m, 1H), 4.43 (t, J= 8.83 Hz, 1H), 4.30 ꢀ 4.38 (m, 2H), 3.62 (s,
6
+
3
H), 2.23 ꢀ 2.17 (m, 2H), 0.84 (s, 9H), 0.07(s, 6H); ESꢀMS: m/z 289 (M+H) .
Step 3: (2S,4R)-4-(tert-Butyldimethylsilyloxy)-1-nitrosopyrrolidin-2-yl)methanol (12). To a
solution of 11 (20g, 0.069 moles) in THF (400 mL) at ꢀ78 °C was added Superhydride (207 mL,
0
.207 moles, 1M Solution in THF) and stirred at room temperature for 5 h. The reaction mixture
was poured over ice cold water and extracted with ethyl acetate. The organic layer was washed
with water, brine, dried over Na SO and concentrated. Purification by column chromatography
2
4
1
over silica gel using 30% ethyl acetate in hexane provided 12g (67%) of the title compound. H
NMR (400 MHz, DMSOꢀd ): δ 5.0 (m, 1H), 4.59 ꢀ 4.55 (m, 2H), 3.89 ꢀ 3.88 (m, 1H), 3.81 ꢀ 3.79
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 50
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
(
m, 1H), 3.55 ꢀ 3.49 (m, 2H), 2.24ꢀ2.19 (m, 1H), 2.05 ꢀ 2.02 (m, 1H), 0.84 (s, 9H), 0.07 (s, 6H);
+
ESꢀMS: m/z 261 (M+H) .
Step 4: (2S,4R)-1-Amino-4-(tert-butyldimethylsilyloxy)pyrrolidin-2-yl)methanol (13). To a
solution of 12 (12 g, 0.046 moles) in methanol (120 mL) at room temperature were added zinc
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
dust (30.5 g, 0.46 moles) and NH Cl (36.4 g, 0.69 moles). After the addition was over, water (60
4
mL) was added at 0 °C and the reaction mixture was stirred at room temperature for 30 min and
then at 50 °C for 1 h. Once the starting material disappeared (monitored by TLC), the reaction
mixture was cooled to room temperature, filtered and the filtrate was concentrated to give a liquid
which was diluted with ethyl acetate. The organic layer was washed with water, brine, dried over
Na SO and concentrated to get 11 g (97%) of the title compound, which was not stable enough
2
4
to conduct NMR or LC data. The material was used with MS characterization only and without
+
further purification to prepare compound 18. ESꢀMS: m/z 247 (M+H) .
2
6,28
2-Chloro-4-formyl-3-methylbenzonitrile (14).
Modified procedure: To a stirred solution of
2
ꢀchloroꢀ4ꢀiodoꢀ3ꢀmethylbenzonitrile in dry THF at 0 °C was added isopropyl magnesium
chloride (14.5 mL, 129.7 mmol, 2 M solution in ether) dropwise, and the reaction mixture was
stirred for 2 h at 0 °C. To this, 1ꢀformyl piperidine (64.8 mL, 129.7 mmol) was added at 0 °C and
stirred at same temperature for 2 h. The reaction mixture was quenched with saturated NH Cl
4
solution and extracted with ethyl acetate. The organic layer was washed with water, brine, dried
over Na SO and concentrated. The crude product was purified by column chromatography using
2
4
1
8
% ethyl acetate in hexane as eluent to give 14.5 g (75%) of the title compound. H NMR (400
MHz, DMSOꢀd ): δ 10.32 (s, 1H), 8.03 (d, J = 8.2 Hz, 1H), 7.92 (d, J = 8.2 Hz, 1H), 2.71 (s, 3H);
6
ꢀ1
IR (KBr): 3072, 2962, 2927, 2856, 2237, 1707 cm .
ACS Paragon Plus Environment

Page 27 of 50
Journal of Medicinal Chemistry
1
2
3
4-((E)-((2R,3S)-3-(tert-Butyldimethylsilyloxy)-2-(hydroxymethyl)pyrrolidin-1-
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
ylimino)methyl)-2-chloro-3-methylbenzonitrile (15). A mixture of 8 (7 g, 0.028 moles) and 14
(5.6 g, 0.0313 moles) in glacial acetic acid (60 mL) was stirred at room temperature for 3 h. Once
both the starting materials disappeared (monitored by TLC), the reaction mixture was poured into
ice water and extracted with ethyl acetate. The organic layer was washed with water, brine, dried
over Na SO and concentrated to give crude product which was purified by column
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
4
chromatography over silica gel using 15% ethyl acetate in hexane as a solvent to provide 7.5 g
1
(66%) of the title compound. H NMR (400 MHz, DMSOꢀd ): δ 7.77 (d, J = 8.3 Hz, 1H), 7.65 (d,
6
J = 8.8 Hz, 1H), 7.20 (s, 1H), 4.87 ꢀ 4.85 (m, 1H), 4.38 (s, 1H), 3.61 ꢀ 3.59 (m, 1H), 3.58 – 3.41
(m, 3H), 3.29 – 3.24 (m, 1H), 2.44 (s, 3H), 2.23 – 2.08 (m, 1H), 1.89 – 1.85 (m, 1H), 0.85 (s,
ꢀ1
9
H), 0.09 (s, 3H), 0.08 (s, 3H) ; IR (KBr): 3473, 2945, 2854, 2233, 1522, 1523 cm ; ESꢀMS: m/z
+
4
08 (M + H) .
4-((E)-((2S,3S)-3-(tert-Butyldimethylsilyloxy)-2-formylpyrrolidin-1-ylimino)methyl)-2-
chloro-3-methylbenzonitrile (16). To a solution of oxalyl chloride (1.92 mL, 0.022 moles) in
dry DCM (25 mL) at ꢀ78 °C was added DMSO (3.12 mL, 0.044 moles) and the mixture was
stirred for 30 min. Then a solution of 15 (7.5 g, 0.0184 moles) in 100 mL DCM was added at the
same temperature and stirring continued for 1 h. The reaction mixture was quenched with
triethylamine (13 mL) at ꢀ30 °C and stirred for 1 h. Then it was diluted with DCM and washed
with water, brine, dried over Na SO and concentrated to get 7.0 g (94%) of the title compound
2
4
1
which was used in the next step without further purification. H NMR (400 MHz, DMSOꢀd ): δ
6
9.53 (d, J = 1.9 Hz, 1H), 7.62 (d, J = 8.3 Hz, 1H), 7.57 (d, J = 8.3 Hz, 1H), 7.29 (s, 1H), 4.52 ꢀ
4.51 (m, 1H), 4.01 (s, 1H), 3.60 ꢀ 3.53 (m, 1H), 3.37 – 3.17 (m, 1H), 2.36 (s, 3H), 2.10 – 2.05 (m,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 50
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
1
H), 1.89 – 1.85 (m, 1H), 0.76 (s, 9H), 0.009 (s, 3H), 0.00 (s, 3H); IR (KBr): 2953, 2927, 2856,
ꢀ1
+
2227, 1730, 1556 cm ; ECꢀMS: m/z 406 (M + H) .
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
-((3S,3aR,4S)-4-(tert-Butyldimethylsilyloxy)-3-hydroxy-3a,4,5,6-tetrahydro-3H-
pyrrolo[1,2-b]pyrazol-2-yl)-2-chloro-3-methylbenzonitrile (17). To a solution of 16 (7.0 g,
0
.0172 moles) in DCM (100 mL) at 0 °C was added borontrifluoride diethyl etherate (6.2 mL,
.0207 moles) and the reaction mixture was stirred at room temperature for 3 h. Once the starting
0
material had disappeared (monitored by TLC), the reaction mixture was diluted with DCM, water
and washed with saturated NaHCO aqueous solution, water, brine, dried over Na SO and
3
2
4
concentrated to get the crude product. Purification of the crude by column chromatography over
silica gel using 10% ethyl acetate in hexane as a solvent provided 4.2 g (60%) of the title
1
compound. H NMR (400 MHz, DMSOꢀd ): δ 7.87 (d, J = 8.3 Hz, 1H), 7.71 (d, J = 8.3 Hz, 1H),
6
5.94 (d, J = 6.8 Hz, 1H), 5.27 (d, J = 7.3 Hz, 1H), 4.17 ꢀ 4.15 (m, 1H), 3.55 ꢀ 3.50 (m, 1H), 3.45 –
3.42 (m,1H), 3.41 – 3.40 (m, 1H), 2.54 (s, 3H), 1.81 – 1.76 (m, 1H), 1.62 – 1.57 (m, 1H), 0.89 (s,
+
9H), 0.10 (s, 3H), 0.08 (s, 3H); ECꢀMS: m/z 406 (M + H) .
4-((E)-((2S,4R)-4-(tert-Butyldimethylsilyloxy)-2-(hydroxymethyl)pyrrolidin-1-
ylimino)methyl)-2-chloro-3-methylbenzonitrile (18). A mixture of 13 (11 g, 0.044 moles) and
14 (4 g, 0.022 moles) in glacial acetic acid (60 mL) was stirred at room temperature for 3 h. Once
both starting materials disappeared (monitored by TLC), the reaction mixture was poured into ice
water and extracted with ethyl acetate. The organic layer was washed with water, brine, dried
over Na SO and concentrated to give crude product which was purified by column
2
4
chromatography over silica gel using 15% ethyl acetate in hexane as a solvent to provide 8 g (44
1
%) of the title compound. H NMR (400 MHz, DMSOꢀd ): δ 7.77 (d, J= 8.3 Hz, 1H), 7.64 (d, J=
6
ACS Paragon Plus Environment

Page 29 of 50
Journal of Medicinal Chemistry
1
2
3
9
Hz, 1H), 7.17 (s, 1H), 4.71 ꢀ 4.63 (m, 1H), 4.59 (bs, 1H), 3.82 (bs, 1H), 3.68 ꢀ 3.65 (m, 1H),
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
3
.64 ꢀ 3.61 (m, 2H), 3.11 (d, J= 3.4 Hz, 1H), 2.49 (s, 3H), 2.10 ꢀ 2.06 (m, 1H), 1.90 ꢀ 1.86 (m,
+
1H), 0.86 (s, 9H), 0.09 (s, 6H); ECꢀMS: m/z 408 (M + H) .
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-((E)-((2S,4R)-4-(tert-Butyldimethylsilyloxy)-2-formylpyrrolidin-1-ylimino)methyl)-2-
chloro-3-methylbenzonitrile (19). To a solution of oxalyl chloride (1.99 mL, 0.022 moles) in
dry DCM (60 mL) at ꢀ78 °C was added DMSO (3.33 mL, 0.047 moles) and stirred for 30 min.
Then a solution of 18 (8 g, 0.019 moles) in 100 mL DCM was added to the reaction mixture at
the same temperature, and stirring continued for 1 h. The reaction mixture was quenched with
triethylamine (13. mL 0.095 moles) at ꢀ30 °C and stirred for 1 h. Then it was diluted with DCM
and washed with water, brine, dried over Na SO and concentrated to get 8 g (100 %) of the title
2
4
compound which was used in the next step without further purification. Due to the low stability
of the product no analytical data was recorded.
4
-((3S,3aS,5R)-5-(tert-Butyldimethylsilyloxy)-3-hydroxy-3a,4,5,6-tetrahydro-3H-
pyrrolo[1,2-b]pyrazol-2-yl)-2-chloro-3-methylbenzonitrile (20). To a solution of 19 (8 g,
0
.019 moles) in DCM (160 mL) at 0 °C was added borontrifluoride diethyl etherate (2.9 mL,
.024 moles) and the reaction mixture was stirred at room temperature for 3 h. Once the starting
0
material disappeared (monitored by TLC), the reaction mixture was diluted with DCM, water and
washed with saturated aqueous NaHCO solution, water, brine, dried over Na SO and
3
2
4
concentrated to get the crude product. Purification of the crude material was achieved by column
chromatography over silica gel using 10% ethyl acetate in hexane as a solvent provided 4 g (50
1
%
) of the title compound. H NMR (400 MHz, DMSOꢀd ): δ 7.85 (d, J= 8.3 Hz, 1H), 7.70 (d, J=
6
8.3 Hz,1H), 5.78 (d, J= 6.3 Hz, 1H), 5.10 (d, J= 6.3 Hz, 1H), 4.39 ꢀ 4.37 (m, 1H), 3.84 – 3.80 (m,
ACS Paragon Plus Environment