A Novel Decalin-Based Bicyclic Scaffold for FKBP51-Selective Ligands

Article abstract of DOI:10.1021/acs.jmedchem.9b01157

Selective inhibition of FKBP51 has emerged as possible novel treatment for diseases like major depressive disorder, obesity, chronic pain, and certain cancers. The current FKBP51 inhibitors are rather large, flexible, and have to be further optimized. By using a structure-based rigidification strategy, we hereby report the design and synthesis of a novel promising bicyclic scaffold for FKBP51 ligands. The structure-activity analysis revealed the decalin scaffold as the best moiety for the selectivity-enabling subpocket of FBKP51. The resulting compounds retain high potency for FKBP51 and excellent selectivity over the close homologue FKBP52. With the cocrystal structure of an advanced ligand in this novel series, we show how the decalin locks the key selectivity-inducing cyclohexyl moiety of the ligand in a conformation typical for FKBP51-selective binding. The best compound 29 produces cell death in a HeLa-derived KB cell line, a cellular model of cervical adenocarcinoma, where FKBP51 is highly overexpressed. Our results show how FKBP51 inhibitors can be rigidified and extended while preserving FKBP51 selectivity. Such inhibitors might be novel tools in the treatment of human cancers with deregulated FKBP51.

Full text of DOI:10.1021/acs.jmedchem.9b01157

Subscriber access provided by WESTERN SYDNEY U
Article
A novel decalin-based bicyclic scaffold for FKBP51-selective ligands
Xixi Feng, Claudia Sippel, Fabian Knaup, Andreas Bracher,
Stefania Staibano, Maria Fiammetta Romano, and Felix Hausch
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01157 • Publication Date (Web): 04 Dec 2019
Downloaded from pubs.acs.org on December 5, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A novel decalin-based bicyclic scaffold for
FKBP51-selective ligands
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
Xixi Feng†,‡, Claudia Sippel†,§, Fabian Knaup , Andreas Bracher#, Stefania Staibano‖,
⁑
⁑,*
Maria F. Romanoǂ, Felix Hausch
† Max Planck Institute of Psychiatry, Dept. Translational Research in Psychiatry,
Kraepelinstrasse 2, 80804 Munich, Germany
# Max Planck Institute of Biochemistry, Am Klopferspitz 18, 82152 Martinsried, Germany
ǂDepartment of Molecular Medicine and Medical Biotechnologies, Federico II University, 80131
Naples, Italy
‖Department of Advanced Biomedical Sciences, Federico II University of Naples, 80131
Naples, Italy
⁑
Institute for Organic Chemistry and Biochemistry, Technische Universität Darmstadt,
Alarich-Weiss-Strasse 4, D-64287 Darmstadt, Germany
Present address:
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 35
1
2
3
4
5
6
7
8
9
‡ Merck Holding (China) Co., 201203, Shanghai, P.R.China
§Pieris Pharmaceuticals GmbH, 85354, Freising, Germany
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
*Correspondence: hausch@drugdiscovery.chemie.tu-darmstadt.de; Tel.: +49-6151-16-
21245
KEYWORDS: FKBP51, decalin, FK506, immunophilins, isomerase,
ABSTRACT
Selective inhibition of FKBP51 has emerged as possible novel treatment for diseases like major
depressive disorder, obesity, chronic pain and certain cancers. The current FKBP51 inhibitors are
rather large, flexible and have to be further optimized. Using a structure-based rigidification
strategy, we hereby report the design and synthesis of a novel promising bicyclic scaffold for
FKBP51 ligands. The structure-activity analysis revealed the decalin scaffold as the best moiety
for the selectivity-enabling subpocket of FBKP51. The resulting compounds retain high potency
for FKBP51 and excellent selectivity over the close homolog FKBP52. With the cocrystal structure
of an advanced ligand in this novel series, we show how the decalin locks the key selectivity-
inducing cyclohexyl moiety of the ligand in a conformation typical for FKBP51-selective binding.
The best compound 29 produces cell death in a HeLa-derived KB cell line, a cellular model of
cervical adenocarcinoma, where FKBP51 is highly overexpressed. Our results show how FKBP51
ACS Paragon Plus Environment
2

Page 3 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
inhibitors can be rigidified and extended while preserving FKBP51 selectivity. Such inhibitors
might be novel tools in the treatment of human cancers with deregulated FKBP51.
7
8
9
INTRODUCTION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The FK506-binding protein 51 (FKBP51, encoded by the FKBP5 gene) belongs to the
immunophilin family and acts as a cochaperone in the Hsp90 machinery1, 2. It functions in part by
regulating glucocorticoid receptor and NFκB signaling.3-4 Human genetic studies have repeatedly
associated FKBP51 with stress-related mental disorders such as depression.5,6 FKBP51 knockout
mice displayed antidepressant-like behavior7-9, had an improved stress hormone regulation7-9,
sleep pattern10, and glucose tolerance11,12, and were protected from weight gain after a high fat
die11,12 and several forms of chronic pain13,14. Moreover, FKBP51 is highly overexpressed in
certain cancer types such as gliomas or melanomas where it seems to promote cancer progression
or resistance to chemo-or irradiation therapy via modulating NFκB or PD-L1 signaling.4,15
A key problem in FKBP51 drug discovery is to achieve selectivity over its close homolog
FKBP52. Due to the high structural homology between these proteins, prototypical ligands derived
from FK506 exhibit similar binding affinities to both proteins. Concomitant binding, however, has
to be avoided since both proteins have opposing effects, e.g. in neuronal cells.25 An inhibition of
FKBP52 is even predicted to have toxic effects since a knockout of this protein in mice led to
severe impairment of the reproduction system.26,27
We recently discovered a transient binding pocket in FKBP51, which is highly disfavored for
FKBP52 due to different dynamics in the two proteins. SAFit1 and SAFit2, the most advanced
FKBP51 inhibitors so far, stabilize this transient pocket and are highly selective for FKBP51 over
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 35
1
2
3
4
5
6
7
8
9
FKBP52 (Fig. 1). In mice, SAFit2 had antidepressant-like effects29,32, attenuated weight gain11,
pain hypersensitivity13,14 and inappropriate alcohol consumption33, and facilitated glucose uptake
in human primary adipocytes.34 Most recently, we showed that FKBP51 inhibition can attenuate
tumor progression in a mouse model of glioblastoma.15
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
MeO
MeO
MeO
MeO
MeO
MeO
O
R
OH
OH
O
O
O
N
NH
O
O
O
N
N
OH
O
1 (SAFit 1)
Ki = 4 nM
O
O
O
R =
O
O
O
MeO
OMe
O
MeO
OMe
2 (SAFit 2)
Ki = 6 nM
OMe
N
OMe
3
4
Ki = 43 nM
Ki = 1.2 µM
Figure 1: Chemical structures of the FKBP51-selective ligands SAFit1 (1), SAFit2 (2), the SAFit1
analog 3, analog 4, and their binding affinities towards FKBP51. The cyclohexyl moiety, which
imparts selectivity for FKBP51 over FKBP52, is highlighted in red.
However, SAFit1 and SAFit2 are unsuitable as drug candidates and the development of improved
analogs has been challenging. Initial structure-activity relationship analyses revealed that the
SAFit scaffold generally has poor ligand efficiency. SAFit1 and SAFit2 are highly flexible
molecules that possess >12 rotatable bonds. The aim of this study was to explore rigidified SAFit
analogs as improved starting points for the development of FKBP51-selective compounds.
Results and Discussion
Rigidification for improving binding affinity
ACS Paragon Plus Environment
4

Page 5 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
During the discovery of SAFit2, the cyclohexyl ring was found to be the most critical structural
feature for high selectivity against FKBP52.29 Replacing it with smaller substituents led to a
significant loss of both selectivity and binding affinity. In contrast, the trimethoxyphenyl ring was
shown to be important for high binding affinity but not for selectivity per se. Replacing it by
different aliphatic substituents and alcohols retained an FKBP51-preferring, induced fit-like
binding mode. A simple methyl group as in compound 4 (Fig. 1) was identified as the minimal
selectivity-inducing moiety.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
In order to improve the affinity of the FKBP51-selective ligands, we set out to explore a
preorganization strategy. A closer inspection of the available cocrystal structures of SAFit analogs
in complex with FKBP51 revealed that the trimethoxyphenyl ring as well as the cyclohexyl ring
are positioned in a way that suggested a possible cyclization by an ethylene linker (Fig. 2, compare
with crystal structure in Fig. 4)35.
A
B
N
N
O
O
O
O
MeO
OMe
OMe
Figure 2: A rigidification strategy to increase binding affinity. (A) The bottom part of the SAFit1-
analog compound 3 from Gaali et al.31 is shown as sticks (magenta) in its active conformation
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 35
1
2
3
4
5
6
7
8
9
bound to FKBP51 (pipecolic amide substituent is not shown for clarity). The decalin-derived
scaffold is superimposed (green sticks). The ethylene linker is highlighted in orange. (B) The
chemical structures of the respective stick representations.
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
We first started with a screening of ligands that contained a bicyclic aromatic moiety instead of
the trimethoxyphenyl-cyclohexyl group of SAFit1 (Fig. 3). All compounds were synthesized by
solid phase-assisted synthesis19, using commercially available aromatic carboxylic acids.
Unfortunately, no binding towards FKBP51 was observed in a fluorescence polarization assay36
up to the solubility limits of the compounds (>10µM in all cases). For analogs harboring an
aromatic or unsaturated carbon in the Cα-position such as 5a-e or 5g, this can be explained by the
need to break conjugation with the adjacent carbonyl group as well as an unfavorable fit of the
planar geometry of the aromatic bicycles, which may not fit to the shape of the FKBP51 binding
pocket (Suppl. Fig. S5). However, compound 5f can be modelled to adopt a conformation that
closely mimics the binding mode of SAFit-like ligands (Fig. 3B). These findings suggest that the
selectivity-inducing subpocket in FKBP51 is very sensitive to the precise geometry of the
interacting moieties.
ACS Paragon Plus Environment
6

Page 7 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3: (A) Chemical structures of synthetic ligands that contain an aromatic bicycle instead of
the trimethoxyphenyl-cyclohexyl group of SAFit1. (B) Compound 5f modelled in the binding site
of FKBP51 (PDB-ID: 5DIU). Compound 3 is shown superimposed as magenta lines. Lys121 has
been removed for clarity.
We therefore set out to synthesize exactly the saturated, stereochemically defined bicyclic structure
derived from our modelling studies. Although this decalin structure is substantially more complex,
all three stereocenters can be conceptually generated in a single step by an intramolecular Diels-
Alder cyclization.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 35
1
2
3
4
Synthesis of a novel bicyclic FKBP51 ligand series
5
6
7
8
9
The key step of the decalin synthesis is a dimethylaluminium chloride-promoted (DMAC)
cyclization of trienimides 10a-c, whereby the stereochemistry is directed by an Evans auxiliary
(Scheme 1). To obtain the trienimides for cyclization we developed a synthesis that is significantly
shorter than original synthetic procedure reported by Evans37 (3 steps instead of 6 to synthesize
the methyltrienoates 8a-c). In addition, our synthesis is more adjustable for the introduction of
substituents.
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
KOtBu
MePPh3I (7a) or
EtPPh3Br (7b+c)
O
R
Grubbs 2. gen.
H
7a
R
: n=0, =H (53 %)
HO
+
n
7b
7c:
R
: n=1, =H (95 %)
HO
n
H
DCM, reflux
THF, 0°C - rt
O
HO
R
n=1, =Me (88 %)
n
6a
: n=0 (quant.)
6b
: n=1 (quant.)
1. (COCl)2, DMSO, TEA
2. Methyl(triphenylphosphoranylidene) acetate
DCM, -78 °C - rt
1. (COCl)2,, DCM, rt
2. n-BuLi, (R)-4-Benzyl-2-oxazolidinone
THF, -78 °C - rt
R
NaOH
8a: n=0, R=H (90 %)
R
O
R
O
O
8b
R
: n=1, =H (quant.)
X
THF/H2O, 70 °C
HO
MeO
8c:
R
n=1, =Me (98 %)
n
n
: n=0, =H (86 %)
9b
n
10a
10b
10c:
R
R
R
9a
R
R
: n=0, =H (not isolated)
: n=1, =H (71 %)
n=1, =Me (66 %)
: n=1, =H (94 %)
9c:
R
n=1, =Me (83 %)
DMAC
DCM, -78 °C - rt
X
O
O
X =
O
N
R
n
11a
11b
R
: n=0, =H (25 %, 2 steps)
R
: n=1, =H (82 %)
11c:
R
n=1, =Me (60 %)
(R)-Benzyloxazolidinone
Scheme 1: Reagents and conditions used for the synthesis of 11 a-c. The obtained yields are stated
in parentheses.
The synthesis of the triene-chain starts with a cross metathesis between an alkenol and trans-
crotonaldehyde38 to exclusively yield the (E)-products 6a and 6b. In the following step the α,β-
unsaturated aldehydes are converted by a Wittig reaction to the dienols 7a-c. The dienol 7c was
obtained in an E/Z ratio of 1:4. Due to the wide range of commercially available Wittig reagents,
it is possible to introduce different moieties in this step that will result in a substituent in 2-position
ACS Paragon Plus Environment
8

Page 9 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of the cyclized decalins. To further extend the chain at the opposite end the alcohols 7a-c were
first converted under Swern conditions to aldehydes and elonged by adding
Methyl(triphenylphosphoranylidene) acetate directly to the reaction mixture. This step mainly
provided the E-isomers 8a-c. The small amounts of the formed Z-isomers (~ 5 %) can easily be
separated from the desired trienoate by flash chromatography. To obtain the free carboxylic acids
9a-c the methyl esters were cleaved in a mixture of 2 M aqueous NaOH and THF respectively.
The Evans auxiliary was introduced to the trienoic acids by acetylation with the respective acid
chlorides of 9a-c or directly with a coupling reagent. The following DMAC-promoted cyclizations
of the trienimides proceeded smoothly, yielding the desired trans–cycloadducts as single
diastereomers 11a-c.
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
R1
O
R1=
19
(44 %)
MeO
R1
O
LiOH, H2O2, THF/H2O, rt or
1. n-BuLi, EtSH, THF, -78 °C
2. Hg(CF3COO), THF/H2O, 60 °C
1. COMU or HATU,
30, DIPEA, DMF, rt
2. TFA, DCM, rt
HO
O
HO
O
X
O
H2
Pd/C
R
R
R
MeO
n
n
OtBu
n
MeOH, rt
O
O
O
20
(66 %)
12a-c
13a-c
11a-c
N
H
12a:
12b
12c:
R
13a
13b
13c:
R
: n=0, =H (quant.)
O
n=0, =H (52 %)
11a:
11b
11c:
R
n=0, =H
R1
O
R
R
: n=1, =H (83 %)
: n=1, =H (quant.)
R
: n=1, =H
30
R
R
n=1, =Me (52 %)
n=1, =Me (quant.)
R
n=1, =Me
21
(20 %)
Scheme 2: Synthesis of the intermediates 12a-c and their conversion in to their SAFit 1 analogues.
To explore the aliphatic bicyclic substituents in more detail, the decalin and bicyclo[4.3.0]nonan
scaffolds were further derivatized. After cyclization, the chiral auxiliary (Scheme 2) of the
unsubstituted bicycles 11a and 11b was cleaved with LiOOH to obtain the free carboxylic acids
12a and 12b respectively. Bicycle 11c, which is substituted in the 2-position, was first converted
to the respective thioester and then hydrolyzed with mercury (II) trifluoroacetate to yield 12c39.
Hydrogenation catalyzed by Pd/C provided the respective saturated bicycles 13a-c.
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 35
1
2
3
4
5
6
7
8
9
The free carboxylic acids were coupled to 30 or 31 with 1-[Bis(dimethylamino)methylene]-1H-
1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate (HATU) or (1-Cyano-2-ethoxy-2-
oxoethylidenaminooxy)dimethylamino-morpholino-carbenium hexafluorophosphate (COMU) as
coupling reagent to obtain the final compounds as SAFit1 (after tBu deprotection) or SAFit2
derivatives respectively.
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
R1
O
14
(42 %)
O
R1
=
R1
O
R1
HBF4, Pd(OAc)2,
p-benzoquinone
O
R1
MeO
MeO
MeCN/H2O, rt
O
OtBu
O
O
15
22
23
(73 %)
O
(26 %)
O
(65 %)
O
O
N
H
HO
O
1. HATU/COMU/PyBrop,
30 or 31, DIPEA, DMF, rt
2. TFA, DCM, rt (for 30)
R2
O
O
R2
R2
O
R2
HBF4, Pd(OAc)2,
p-benzoquinone
CCl3
NaBH4
R
30
n
R2=
MeCN/H2O, rt
MeOH, rt
O
12a-c
OH
27
O
MeO
MeO
16
24
25
(25 %)
(quant.)
O
(15 %)
(67 %)
dr =3.4 : 1
O
R1
N
O
O
N
H
O
17
(42 %)
O
31
R2
O
HBF4, Pd(OAc)2,
p-benzoquinone
R2
MeCN/H2O, rt
O
18 74 %)
(
26 49 %
( )
Scheme 3: Synthesis of the SAFit analogues 14-18 and further derivatization of those, via Wacker
conditions and Luche reduction (22-27).
15 and 16 were oxidized under Wacker conditions40 to obtain the 3-keto (22, 24) and the 4-keto
(23, 25) compounds in a 1:2.5 ratio, which can be easily separated by flash chromatography. In
contrast, oxidation of 18 only provided the respective 4-keto compound 26. The carbonyl group of
25 was further reduced via Luche reduction41 to yield 27 in a diasteromeric ratio of 3.4:1.
ACS Paragon Plus Environment
10

Page 11 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
MeO
MeO
MeO
1. OsO4, NMO, acetone, rt
2. 2,2-Dimethoxypropane,
PTSA, acetone, rt
HO
O
O
MeO
O
X
O
N
N
O
O
3. LiOH, H2O2, THF/H2O, rt
HATU, 31, DIPEA
DMF, rt
HCl
O
O
N
N
MeOH, rt
O
O
O
O
O
HO
HO
O
O
O
8
11b
12d
28
29
(30 %)
(36 %, 3 steps)
(77 %)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4: Synthesis of the protected dihydroxylated analogue 28 and its dihydroxy compound
29.
To obtain 12d, a dihydroxylation step was conducted with OsO4 and 11b. After protection of the
diol, the minor diastereomer was removed by flash chromatography. Then the auxiliary was
cleaved and the free carboxylic acid was coupled to 31 to obtain 28. The dihydroxy compound 29
was obtained after treatment of 28 under acidic conditions.
Structure-Affinity Relationship (SAR) of the new FKBP51-selective ligand series
Compounds 14-29 were tested in a competitive fluorescence polarization assay for binding to
FKBP51 and FKBP52. All bicycles showed binding towards FKBP51 but almost no binding
towards FKBP52 was observed. This is in agreement with our modelling studies that one of the
bicyclic rings corresponds to the cyclohexyl substituent that was previously reported to be
important for FKBP51 selective ligands.
In general, we observed that the [4.3.0] bicycles, containing octahydro-indene and derivatives,
show lower binding affinities towards FKBP51 than their corresponding decalin-compounds.
Additionally, all decalin-like compounds showed no binding towards FKBP52, whereas one
octahydro-indene derivative (23) showed low binding towards FKBP52. In all cases, a saturation
of the double bond, a leftover of the intramolecular Diels-Alder cyclization, resulted in a
significant increase in binding affinity (compare 14/20, 15/19, 17/21). The addition of a methyl
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 35
1
2
3
4
5
6
group in the 2-position of the decalins was tolerated (26/25, 18/16) or led to a slight increase in
affinity (21 vs 20).
7
8
9
We next explored the appendage of polar groups in the ring systems. Carbonyl groups in the 3-
position had negative effects on the binding affinity (22 vs 19, 24 vs 20)42. However, a carbonyl
group in the 4-position improved affinity (25 vs 20, 26 vs 21), resulting in a Ki= 0.23 µM for 25.
A hydroxyl group in the 4-position was less effective (27 vs 25). The dihydroxylated derivative 29
showed improved binding (Ki = 0.28 µM), suggesting beneficial interactions of the 3-hydroxy
group. The free hydroxyl groups seemed to be important since alkylation of the hydroxy groups
as in 28 significantly decreases the binding affinity.
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
ACS Paragon Plus Environment
12

Page 13 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
MeO
MeO
MeO
MeO
O
OH
N
O
O
O
O
O
N
N
O
R1
R2
O
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
Ki [µM]
Ki [µM]
Ki [µM]
Ki [µM]
Entry
Entry
Entry
Entry
O
O
O
O
0.33
2.6
2.4
28
O
O
14
22
16
26
O
O
O
O
1.9
0.65
7.8
2.0
2.6
(dr = 3.4:1)
OH
15
19
18
27
O
O
O
O
0.49
1.3
8.2
O
O
O
17
21
24
28
O
O
O
O
0.81
3.1
0.28
0.23
HO
O
O
OH
23
20
25
29
Table 1. FKBP binding affinities bicyclic SAFit1 (1) and SAFit2 (2) analogs; Ki values were
determined by a competitive fluorescence polarization assay36; 25 was tested as diastereomeric
mixture (3.4:1).
In an exploratory assessment of the drug-like parameter of the currently most advanced FKBP51
SAFit1 and SAFit2, inhibition of CYP3A4 was identified as an inacceptable liability (IC50 for
SAFit2 <1µM). Gratifyingly, compound 20 showed reduced CYP3A4 inhibition (estimated IC50
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 35
1
2
3
4
5
6
= 8.4µM, Suppl. Fig. S6) suggesting that the rigidified analogs of the decalin series could provide
compounds with an improved ADME/PK profile (Suppl. Table 2).
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
Cocrystal structure of 25
To confirm the postulated binding mode of our decalin-containing analogs, we solved a cocrystal
structure of 25 in complex with the FK506-binding domain of FKBP51 (Fig. 4)43. The X-ray
structure revealed a binding conformation that is in agreement with previously described FKBP
ligands. The pipecolic ester is centered in a hydrophobic pocket. The prominent hydrogen bonds
to Ile87 and Tyr113, which until now have been observed in all FKBP51 complexes with carbonyl
pipecolate-based ligands are also present (Fig. 4A). We were delighted to see that that the
conformational rearrangement of Phe67, which is indicative for FKBP51-selective binding, is
preserved. The terminal cyclohexyl ring of the decalin is deeply buried in the enlarged binding
pocket created by the conformational rearrangement of Phe67, and it superimposes almost
perfectly with cyclohexyl substituent of cocrystallized ligands of the SAFit series (Fig. 4B). Close
contacts of the 4-keto group of 25 with Lys121 of FKBP51 were observed.
ACS Paragon Plus Environment
14

Page 15 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4: X-ray structures of 25 (green) in complex with the FK506-binding domain of FKBP51
(grey surface); Lys121 has been omitted for clarity. (A) Hydrogen bonds to Tyr113 (yellow) and
Ile87 (yellow) are indicated as black dotted lines. The displaced Phe67 is shown as purple sticks.
(B) The SAFit1 analog 3 (magenta, from 5DIU)) has been superimposed on the 25-FKBP51
complex
Pro-apoptotic effects of 29
Because FKBP51 sustains cancer resistance, we evaluated the anti-cancer effect of the advanced
compound 29 using the HeLa-derived KB cell line as a model system, based on the finding that,
in this cell line, FKBP51 transcript is at least ten-fold higher compared with mRNA level of two
different melanoma cell lines (Suppl. Fig. 1). Cells were cultured at different doses with the
compound and cell death was determined after 72h. As shown in Fig. 5a, the cytotoxic effect of
the compound increased in a dose-dependent manner. To address whether cells died by apoptosis,
cells, stimulated with 29 were harvested after 40h and stained with annexin V and PI. The
proportion of died cells with an intact membrane (annexin V+/PI-), consistent with early apoptosis,
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 35
1
2
3
4
5
6
7
8
9
was measured. Fig. 5b shows the proportion of annexin V+/PI- progressively increased in the
presence of 29 in a dose-dependent manner, suggesting the cells died by apoptosis and confirming
the anti-cancer effect 29. The FKBP51 inhibitors SAFit1 and SAFit2 similarly induced apoptosis
in KB cells at 1µM, in line with their enhanced affinity to FKBP51 (see Suppl. Fig. 2 for a full
dose-response relationship for SAFit1 and SAFit2). The apoptotic effect of compound 29 as well
as of SAFit2 were higher in KB cells (expressing high levels of FKBP51) compared with the
progenitor HeLa cells (expressing low levels of FKBP51, Suppl. Fig. 3). This suggests an addiction
of KB cells to FKBP51, which in turn leads to a higher sensitivity to FKBP51 inhibitors. To further
confirm the specificity of 29, we tested the close analog 5a, which does not bind to FKBP51, in
KB cells. As expected, 5a did not induce cell death up to a concentration of 60µM (Suppl. Fig.
S4). Collectively, these data strongly suggest FKBP51 as a relevant target for the apoptotic
activities of 29.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(A)
(B)
Figure 5: Compound 29 dose-dependently enhances cell death of KB cells. (A) Cells were
incubated with compound 29 at different doses. Cell death was measured after 72h by PI
incorporation and measurement of hypodiploid cells. Representative histograms of KB cell DNA
content are shown on top. Boxplots below represent values of hypodiploid (dead) cells from four
experiments, each performed in triplicate. (b) Cells were incubated with 29 at different doses and
with SAFit 1 and 2, which were used as positive control for FKBP51 inhibition. Cell death was
ACS Paragon Plus Environment
16

Page 17 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
measured after 40h by annexin V/PI double staining to determine apoptosis. Representative
histograms of KB cell apoptosis are shown on top. Boxplots represent values of annexinV+/PI-
cells (corresponding to early apoptosis) from four experiments each performed in triplicate.
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
Conclusion
In order to explore the binding of selective FKBP51 antagonists, we used a rigidification strategy
to design a novel scaffold for new FKBP51 ligands. The resulting decalin analogs bound with
exactly the same binding mode as noncyclic cyclohexyl derivatives. The direct comparison of the
underivatized decalin analog 20 with the methyl-containing compound 4 (Fig. 1) shows that the
cyclization per se only minimally enhances binding affinity. However, the constrained decalin
scaffold provides several new rigid exit vectors, which are not possible for the parent cyclohexyl-
containing SAFit analogs. This can be exploited to further enhance affinity, as shown for
compounds 25 and 29. These analogs are rapidly accessible using the improved synthesis
developed for the decalin scaffold. Using these analogs, we show that selective FKBP51 inhibitors
impair the viability of cancer cells. Collectively, our findings substantially improve our
understanding of FKBP51 inhibitors and provide a new rigidified scaffold for future optimization
of FKBP51-directed drugs.
Experimental Section:
All dry reactions were performed under argon atmosphere and with commercially available dry
solvents. Chromatographic separations were performed either by manual flash chromatography or
automated flash chromatography using an Interchim Puriflash 430 with an UV detector. Merck
F-254 (thickness 0.25 mm) commercial plates were used for analytical TLC. 1H NMR spectra, 13C
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 35
1
2
3
4
5
6
7
8
9
NMR spectra, 2D HSQC, HMBC, and COSY of all intermediates were obtained from the
Department of Chemistry and Pharmacy, Ludwig-Maximilians-University Munich, on a Bruker
Avance III HD 400/800 or a Varian NMR system 300/400/600 at room temperature. Chemical
shifts for 1H or 13C are given in ppm (δ) downfield from tetramethylsilane using residual protio
solvent as an internal standard. Mass spectra (m/z) were recorded on a Thermo Finnigan LCQ
DECA XP Plus mass spectrometer at the Max Planck Institute of Psychiatry, while the high
resolution mass spectrometry was carried out at MPI for Biochemistry (Microchemistry Core
Facility) on Bruker Daltonics MicrOTOF. Purities of all tested compounds were greater than 95%
based on HPLC chromatograms. HPLC analyses were performed on a Beckman System Gold
125S Solvent Module fitted with a Phenomenex Jupiter 4µm Proteo 90Å column (250 mm x 4.6
mm) and a Beckman System Gold Diode Array Detector Module 168 (220 nm). The following
buffers were used: buffer A: 0.1% TFA in H2O and buffer B: 0.1% TFA in MeCN. Unless
otherwise indicated, in the supporting information, samples were run at a gradient of 0 - 100% B
over 20 min at a flow rate of 1.5 mL/min.
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
(E)-8-hydroxyoct-2-enal (6b). To a solution of Grubbs 2nd generation catalyst (85.0 mg, 100
µmol) in 50 mL DCM was added hept-6-en-1-ol (1.35 mL, 10.0 mmol) and crotonaldehyde (4.14
mL, 50.0 mmol). The reaction mixture was heated to reflux for 2 h (oil bath was preheated to 50
°C). After completion of the reaction the mixture was cooled to rt. Some silica gel was added to
the flask and the resulting slurry was stirred open to air for 30 min. The solvent was removed and
the crude product was purified by flash chromatography (0-40 % EtOAc in cyclohexane) to obtain
the title compound as light brown oil (quantitative yield). 1H NMR (599 MHz, CDCl3) δ 9.49 (d,
J = 7.9 Hz, 1H), 6.83 (dt, J = 15.6, 6.8 Hz, 1H), 6.10 (ddt, J = 15.6, 7.9, 1.5 Hz, 1H), 3.64 (t, J =
6.5 Hz, 2H), 2.38 – 2.30 (m, 2H), 1.61 – 1.56 (m, 2H), 1.55 – 1.50 (m, 2H), 1.44 – 1.38 (m, 2H).13C
ACS Paragon Plus Environment
18

Page 19 of 35
Journal of Medicinal Chemistry
1
2
3
4
NMR (151 MHz, CDCl3) δ 194.07, 158.54, 133.04, 62.63, 32.63, 32.36, 27.61, 25.32. MS (ESI+):
5
6
calc. 165.09 [C8H14O2+Na]+, found 164.08 [M+Na]+.
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
(E)-nona-6,8-dien-1-ol (7b). To a suspension of ethyltriphenylphosphonium bromide (6.68 g, 18
mmol) in 90 mL THF, KOtBu (19.5 mL, 19.50 mmol, 1 M in THF) was added dropwise at 0 °C.
The solution was stirred at rt for 1 h and then again cooled to 0 °C. The aldehyde 6b (1.07 g, 7.50
mmol) dissolved in 8 mL THF was then added dropwise and the mixture was stirred at 4 °C for 16
h. After warming to rt the reaction was quenched by the addition of a saturated NH4Cl solution
and the product was extracted several times with Et2O. The combined organic layers were dried
over MgSO4, filtered and the solvent was removed under reduced pressure. The crude product was
purified by flash chromatography (0-40 % EtOAc in cyclohexane) as a yellow oil (1.0g, 7.14
mmol, 95.1 %).
1H NMR (300 MHz, CDCl3) δ 6.31 (dt, J = 16.9, 10.2 Hz, 1H), 6.12 – 5.99 (m, 1H), 5.70 (dt, J =
14.6, 6.9 Hz, 1H), 5.14 – 5.04 (m, 1H), 4.99 – 4.92 (m, 1H), 3.64 (t, J = 6.6 Hz, 2H), 2.18 – 2.00
(m, 2H), 1.68 – 1.49 (m, 2H), 1.47 – 1.32 (m, 5H). 13C NMR (75 MHz, CDCl3) δ 137.23, 135.12,
131.07, 114.73, 62.92, 32.61, 32.44, 28.94, 25.29. MS (ESI+): calc. 141.23 [C9H16O+H]+, found
140.90 [M+H]+.
(2E,8E)-methyl undeca-2,8,10-trienoate (8b). DMSO (1.06 mL, 14.97 mmol) was added to a
solution of oxalyl chloride (3.74 mL, 7.49 mmol, 2 M in DCM) in 50 mL DCM at -78 °C and
stirred at that temperature for 30 min. Then the alcohol 7b (700 mg, 4.99 mmol), dissolved in 5
mL DCM was added and the reaction mixture was stirred for further 60 min at -78 °C. In the
following step TEA (6.96 mL, 49.90 mmol) was added and the slurry was allowed to warm to rt.
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 35
1
2
3
4
5
6
7
8
9
Then methyl 2-(triphenylphosphoranylidene)acetate (1.84 g, 5.49 mmol) was added in one portion
and the slurry was stirred at rt for additional 16 h. The reaction mixture was then washed with H2O
and saturated NH4Cl solution. The organic layer was then dried over MgSO4, filtered and the
solvent was removed under reduced pressure. The product 8b was obtained after purification by
flash chromatography (0-5 % EtOAc in cyclohexane) as a yellow oil (quantitative yield). 1H NMR
(400 MHz, CDCl3) δ 6.95 (dt, J = 15.7, 7.0 Hz, 1H), 6.42 – 6.19 (m, 1H), 6.13 – 5.97 (m, 1H),
5.81 (dt, J = 15.6, 1.6 Hz, 1H), 5.75 – 5.58 (m, 1H), 5.15 – 5.03 (m, 1H), 5.03 – 4.86 (m, 1H), 3.72
(s, 3H), 2.20 (qd, J = 7.0, 1.6 Hz, 2H), 2.12 – 2.01 (m, 2H), 1.52 – 1.36 (m, 4H). 13C NMR (101
MHz, CDCl3) δ 167.11, 149.41, 137.13, 134.71, 131.25, 120.97, 114.92, 51.37, 32.20, 32.00,
28.57, 27.48, 26.89. MS (ESI+), calc. 195.14 [C12H18O2+H]+, found 195.01 [M+H]+.
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
(2E, 8E)- undeca-2,8,10-trienoic acid (9b). The methyl ester 8b (680 mg, 3.50 mmol) was
dissolved in 35 mL THF/2 M aq. NaOH (1.5:1) and stirred at 70 °C until the reaction was finished.
After cooling to rt the reaction mixture was extracted with Et2O. The organic layer was removed
and the aqueous layer acidified with concentrated HCl. Then free carboxylic acid was extracted
several times with DCM. The combined organic layers were dried over MgSO4, filtered and the
solvent was removed under reduced pressure to obtain the product 9b as a light yellow oil (595
mg, 3.30 mmol, 94.3 %) without further purification.
1H NMR (400 MHz, CDCl3) δ 7.07 (dt, J = 15.6, 7.0 Hz, 1H), 6.30 (dtd, J = 17.0, 10.2, 0.7 Hz,
1H), 6.21 – 5.92 (m, 1H), 5.82 (dt, J = 15.6, 1.6 Hz, 1H), 5.78 – 5.55 (m, 1H), 5.15 – 5.05 (m, 1H),
5.05 – 4.85 (m, 1H), 2.23 (qd, J = 6.9, 1.6 Hz, 2H), 2.18 – 2.03 (m, 2H), 1.62 – 1.30 (m, 4H).13C
ACS Paragon Plus Environment
20

Page 21 of 35
Journal of Medicinal Chemistry
1
2
3
4
NMR (101 MHz, CDCl3) δ 171.79, 152.04, 137.11, 134.63, 131.30, 120.71, 114.96, 32.19, 32.11,
5
6
28.59, 27.34. MS (ESI+), calc. 181.12 [C11H16O2+H]+, found 181.00 [M+H]+.
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
(R)-4-benzyl-3-((2E,8E)-undeca-2,8,10-trienoyl)oxazolidin-2-one (10b). The carboxylic acid
9b (900 mg, 4.99 mmol) was dissolved in 5 mL DCM and oxalyl chloride (4.99 mL, 9.98 mmol,
2 M in DCM) with a catalytic amount of DMF (a few drops) was added to convert the acid into an
acid chloride. The reaction mixture was stirred at rt for 2 hours. Then the solvent was removed and
the carboxylic acid chloride was used without further purification. In the meantime, n-BuLi (2.19
mL, 5.49 mmol, 2.5 M in hexane) was added to (R)-4-benzyloxazolidin-2-one (0.88 g, 4.99 mmol)
in 25 mL THF at -78 °C and the reaction mixture was stirred at that temperature for 1 h. Then the
acid chloride, dissolved in 5 mL DCM, was added to the deprotonated auxiliary and the mixture
was allowed to warm to rt. The reaction was quenched by the addition of 15 mL saturated NH4Cl
and the product was extracted three times with 40 mL Et2O. The combined organics were then
dried over MgSO4, filtered and the solvent was removed under reduced pressure. The product was
obtained after purification by flash chromatography (0-10 % EtOAc in cyclohexane) as a yellow
oil (1.21 g, 3.55 mmol, 71.1 %). 1H NMR (599 MHz, CDCl3) δ 7.34 – 7.30 (m, 2H), 7.26 – 7.16
(m, 5H), 6.30 (dtd, J = 17.1, 10.3, 0.8 Hz, 1H), 6.15 – 5.98 (m, 1H), 5.68 (dtd, J = 15.0, 7.0, 0.8
Hz, 1H), 5.15 – 5.03 (m, 1H), 5.02 – 4.88 (m, 1H), 4.82 – 4.64 (m, 1H), 4.26 – 4.11 (m, 2H), 3.33
(dd, J = 13.4, 3.3 Hz, 1H), 2.78 (dd, J = 13.4, 9.6 Hz, 1H), 2.30 (qd, J = 6.9, 1.2 Hz, 2H), 2.10 (qd,
13
J = 7.2, 1.4 Hz, 2H), 1.55 – 1.49 (m, 2H), 1.48 – 1.43 (m, 2H). C NMR (101 MHz, CDCl3) δ
165.06, 153.42, 151.64, 137.17, 135.37, 134.80, 131.24, 129.44, 128.93, 127.29, 120.45, 114.92,
66.10, 55.32, 37.89, 32.53, 32.24, 28.67, 27.59, 26.90. MS (ESI+), calc. 340.19 [C21H25NO3+H]+,
found 340.28 [M+H]+.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 35
1
2
3
4
5
6
7
8
9
(R)-4-benzyl-3-((1R,4aS,8aR)-1,2,4a,5,6,7,8,8a-octahydronaphthalene-1-carbonyl)oxa-
zolidin-2-one (11b). The acylated Evans auxiliary 10b (140 mg, 0.41 mmol) was dissolved in 16.5
mL DCM and cooled to -78 °C. DMAC (0.62 mL, 0.62 mmol, 1 M in hexane) was added and the
reaction mixture was stirred for 16 h, whereby it slowly warmed to rt. Then the mixture was
quenched with 1 N HCl and the product extracted with DCM. The combined organics were then
dried over MgSO4, filtered and the solvent was removed under reduced pressure. The product was
obtained after purification by flash chromatography (0-10 % EtOAc in cyclohexane) as a colorless
oil (115 mg, 0.34 mmol, 82.1 %). 1H NMR (599 MHz, CDCl3) δ 7.35 – 7.30 (m, 2H), 7.28 – 7.25
(m, 1H), 7.22 – 7.18 (m, 2H), 5.64 – 5.60 (m, 1H), 5.49 – 5.41 (m, 1H), 4.72 – 4.67 (m, 1H), 4.21
– 4.14 (m, 2H), 3.87 (td, J = 10.8, 5.7 Hz, 1H), 3.25 (dd, J = 13.4, 3.3 Hz, 1H), 2.77 (dd, J = 13.4,
9.5 Hz, 1H), 2.41 – 2.35 (m, 1H), 2.32 – 2.24 (m, 1H), 1.88 – 1.81 (m, 1H), 1.79 – 1.68 (m, 3H),
1.69 – 1.62 (m, 1H), 1.61 – 1.49 (m, 1H), 1.34 – 1.28 (m, 2H), 1.12 – 0.99 (m, 2H). 13C NMR (151
MHz, CDCl3) δ 176.37, 153.12, 135.24, 132.09, 129.41, 128.90, 127.32, 123.96, 65.96, 55.29,
43.13, 42.46, 41.09, 37.96, 32.99, 29.95, 29.83, 26.62, 26.44. MS (ESI+), calc. 340.19
[C21H25NO3+H]+, found 340.11 [M+H]+.
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
(1R,4aS,8aR)-1,2,4a,5,6,7,8,8a-octahydronaphthalene-1-carboxylic acid (12b). To cleave the
Evans auxiliary 11b (43 mg, 0.13 mmol) was dissolved in 2 mL THF/H2O (8:5) followed by the
addition of LiOH (14 mg, 0.59 mmol) and H2O2 (66 µL, 0.65 mmol, 30 wt.-% in H2O). The
reaction mixture was stirred at rt for 2.5 h and then quenched by adding a half-saturated Na2SO3
solution. It was extracted with DCM and the organic layers were removed. Then the aqueous layer
ACS Paragon Plus Environment
22

Page 23 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was acidified with concentrated HCl and the product was extracted with DCM. The combined
organic layers were dried over MgSO4, filtered and the solvent was removed under reduced
pressure to obtain the title compound without further purification as a colorless solid (19 mg, 0.11
mmol, 83.2 %). 1H NMR (300 MHz, CDCl3) δ 5.65 – 5.57 (m, 1H), 5.45 (dq, J = 9.9, 1.9 Hz, 1H),
2.48 – 2.39 (m, 1H), 2.38 – 2.29 (m, 2H), 1.87 – 1.71 (m, 4H), 1.68 – 1.21 (m, 4H), 1.19 – 0.99
(m, 2H). 13C NMR (75 MHz, CDCl 3) δ 181.95, 132.09, 123.79, 46.17, 42.12, 41.33, 32.94, 30.31,
29.47, 26.58, 26.36.
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
(3aR,5R,5aR,9aR,9bS)-2,2-dimethyldecahydronaphtho[1,2-d][1,3]dioxole-5-carboxylic acid
(12d). 11b (80 mg, 0.26 mmol) was dissolved in 1.0 mL acetone and 0.5 mL H2O. Then 4-
methylmorpholine-4-oxide (55 mg, 0.47 mmol) and osmium tetroxide (118 µL, 9.4 µmol, 2.5 wt.-
% in tBuOH) were added and the mixture was stirred at rt for 16 h. The crude product was directly
loaded on silica and purified by flash chromatography (0-50 % [EtOAc + 2 % MeOH] in
cyclohexane) to obtain the intermediate as a diastereomeric mixture (dr = 4:1, quant. yield). The
intermediate was then dissolved in 1.4 mL acetone and 2,2-dimethoxypropane (1:1) and PTSA
(0.02 eq) was added. The reaction was stirred for 24 h and then purified by flash chromatography
(0-5 % EtOAc in cyclohexane), whereby the minor diasteroemer was separated, to obtain the
protected diol. Then general procedure F (Supporting Information) was used to obtain 12d without
1
further purification as colorless oil (24 mg, 9.4 µmol, 36.3 % from 11b). H NMR (400 MHz,
CDCl3) δ 4.28 – 4.20 (m, 1H), 3.60 (dd, J = 8.6, 4.9 Hz, 1H), 2.46 – 2.36 (m, 1H), 2.32 (ddd, J =
14.9, 4.0, 2.6 Hz, 1H), 2.13 – 2.04 (m, 1H), 1.96 (ddd, J = 14.8, 12.1, 4.0 Hz, 1H), 1.84 – 1.64 (m,
3H), 1.48 (s, 3H), 1.34 (s, 3H), 1.26 – 1.14 (m, 4H), 1.02 (dt, J = 10.9, 6.1 Hz, 1H), 0.97 – 0.87
(m, 1H). 13C NMR (101 MHz, CDCl3) δ 181.15, 108.31, 79.53, 72.31, 44.28, 43.73, 40.16, 30.73,
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 35
1
2
3
4
5
6
30.12, 29.67, 28.53, 26.38, 25.77, 25.59. MS (ESI+), calculated 255.16 [C14H22O4+H]+, found
255.02 [M+H]+.
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
(S)-(R)-3-(3,4-dimethoxyphenyl)-1-(3-(2-morpholinoethoxy)phenyl)propyl 1-((1S,4aS,8aR)-
1,2,4a,5,6,7,8,8a-octahydronaphthalene-1-carbonyl)piperidine-2-carboxylate (16). 12b (40.0
mg, 0.22 mmol), 31 (125 mg, 0.24 mmol), HATU (127 mg, 0.33 mmol) and DIPEA (86.0 mg,
0.67 mmol) were stirred in 4 mL DMF at rt for 16 h. The reaction mixture was loaded on silica
and purified by flash chromatography (0-100 % [EtOAc + 2 % MeOH + 0.1 % TEA] in
cyclohexane) to obtain the title compound as light yellow oil (quantitative yield). 1H NMR (599
MHz, DMSO-d6) δ 7.26 – 7.21 (m, 1H), 6.88 – 6.80 (m, 4H), 6.75 – 6.70 (m, 1H), 6.67 – 6.62 (m,
1H), 5.59 (dd, J = 8.8, 4.7 Hz, 1H), 5.56 – 5.50 (m, 2H), 5.41 – 5.33 (m, 2H), 5.27 – 5.23 (m, 1H),
3.71 – 3.66 (m, 6H), 3.59 – 3.52 (m, 4H), 2.53 – 2.48 (m, 2H), 2.44 – 2.38 (m, 3H), 2.20 – 2.14
(m, 2H), 2.11 – 2.05 (m, 3H), 1.83 – 1.76 (m, 1H), 1.72 – 1.56 (m, 7H), 1.54 – 1.48 (m, 3H), 1.36
13
– 1.20 (m, 5H), 1.06 – 1.01 (m, 2H), 0.98 – 0.93 (m, 2H), 0.91 – 0.85 (m, 1H). C NMR (151
MHz, CDCl3) δ 179.56, 177.17, 175.45, 167.31, 163.49, 153.81, 152.20, 147.22, 138.31, 136.98,
134.77, 129.69, 125.10, 123.27, 118.96, 117.35, 116.96, 80.59, 71.20, 70.25, 64.78, 62.12, 60.68,
60.40, 58.75, 56.65, 48.13, 47.26, 45.54, 42.66, 40.94, 37.71, 35.78, 34.45, 31.29, 30.14, 26.21,
19.23. HRMS, calc. 675.4009 [C40H54N2O7+H]+, found 675.3999 [M+H]+. HPLC [0-100%
Solvent B, 20 min]: Rt = 13.60 min, 100 %.
(S)-(R)-3-(3,4-dimethoxyphenyl)-1-(3-(2-morpholinoethoxy)phenyl)propyl 1-((1R,4aS,8aR)-
3-oxodecahydronaphthalene-1-carbonyl)piperidine-2-carboxylate (24) and (S)-(R)-3-(3,4-
ACS Paragon Plus Environment
24

Page 25 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
dimethoxyphenyl)-1-(3-(2-morpholinoethoxy)phenyl)propyl
1-((1R,4aR,8aR)-4-
solution of
oxodecahydronaphthalene-1-carbonyl)piperidine-2-carboxylate (25).
A
7
8
9
tetrafluoroboric acid (16 µL, 0.10 mmol, 48 wt.-% in H2O) and 16 were added to a solution of
palladium(II) acetate (0.83 mg, 3.70 µmol) and p-benzoquinone (8.01 mg, 74 µmol) in 0.45 mL
MeCN and 0.6 mL H2O. The reaction mixture was stirred at rt for 16 h and then directly loaded
on silica for purificaion by flash chromatography (0-50 % [EtOAc + 2 % MeOH + 0.1 % TEA] in
hexane), whereby both products can easily be separated, to obtain title compounds 24 (11.4 mg,
20.9 µmol, 14.8 %, analytics shown in Supplementary Information) and 25 (19.0 mg, 51.1 µmol,
24.7 %) as colorless oils. (25): 1H NMR (599 MHz, CDCl3) δ 7.24 – 7.21 (m, 1H), 6.95 – 6.74 (m,
4H), 6.68 – 6.62 (m, 2H), 5.74 – 5.68 (m, 1H), 5.48 (d, J = 4.9 Hz, 1H), 4.15 – 4.04 (m, 2H), 3.97
(d, J = 13.5 Hz, 1H), 3.89 – 3.79 (m, 6H), 3.79 – 3.67 (m, 4H), 3.21 (td, J = 13.1, 2.8 Hz, 1H),
2.85 (dt, J = 11.9, 5.3 Hz, 1H), 2.82 – 2.75 (m, 2H), 2.65 – 2.52 (m, 5H), 2.52 – 2.45 (m, 1H), 2.42
– 2.33 (m, 3H), 2.24 – 2.16 (m, 1H), 2.13 – 1.59 (m, 11H), 1.51 – 1.42 (m, 1H), 1.41 – 1.32 (m,
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
13
1H), 1.32 – 1.09 (m, 4H), 0.99 (qd, J = 12.3, 3.6 Hz, 1H). (25): C NMR (151 MHz, CDCl3) δ
210.44, 173.52, 170.47, 158.80, 148.84, 147.32, 141.40, 133.40, 129.60, 120.10, 118.92, 113.84,
113.20, 111.66, 111.27, 76.24, 66.87, 57.64, 55.90, 55.82, 54.08, 53.16, 52.06, 45.67, 44.50, 43.51,
40.36, 37.93, 32.29, 31.22, 29.40, 26.82, 25.61, 25.42, 25.30, 25.14, 21.11. HRMS, calc. 691.3958
[C40H54N2O8+H]+, found 691.3991 [M(25)+H]+. HPLC [0-100% Solvent B, 20 min]: Rt = 15.48
min, 99 %.
(2S)-(R)-3-(3,4-dimethoxyphenyl)-1-(3-(2-morpholinoethoxy)phenyl)propyl-1-
((5R,5aR,9aR)-2,2-dimethyldecahydronaphtho[1,2-d][1,3]dioxole-5-carbonyl)piperidine-2-
carboxylate (28). 12d (20.0 mg, 0.08 mmol), 31 (44.3 mg, 0.09 mmol), HATU (44.9 mg, 0.12
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 35
1
2
3
4
5
6
7
8
9
mmol) and DIPEA (30.5 mg, 0.236 mmol) were stirred in 2 mL DMF at rt for 18 h. The orange
reaction mixture was loaded on silica and purified by flash chromatography (0-100 % [EtOAc + 2
% MeOH + 0.1 % TEA] in cyclohexane) to obtain the title compound as light yellow oil (45.5 mg,
0.06 mmol, 77.2 %). 1H NMR (599 MHz, DMSO-d6) δ 7.28 – 7.20 (m, 1H), 6.88 – 6.79 (m, 4H),
6.75 – 6.71 (m, 1H), 6.67 – 6.60 (m, 1H), 5.58 (dd, J = 8.8, 4.7 Hz, 1H), 5.21 (d, J = 6.0 Hz, 0H),
4.20 (d, J = 6.7 Hz, 1H), 4.14 – 4.02 (m, 3H), 3.93 (d, J = 13.4 Hz, 1H), 3.72 – 3.64 (m, 6H), 3.60
– 3.52 (m, 4H), 3.09 – 2.99 (m, 1H), 2.98 – 2.91 (m, 1H), 2.87 – 2.75 (m, 0H), 2.71 – 2.66 (m,
2H), 2.43 (ddd, J = 13.7, 8.3, 4.6 Hz, 1H), 2.19 – 2.12 (m, 2H), 2.09 – 2.01 (m, 2H), 1.98 – 1.95
(m, 2H), 1.77 – 1.48 (m, 8H), 1.44 – 1.31 (m, 5H), 1.29 – 1.04 (m, 8H), 1.02 – 0.66 (m, 4H), 0.30
– 0.20 (m, 1H). 13C NMR (151 MHz, DMSO-d6) δ 174.96, 170.87, 158.85, 149.02, 147.46, 142.43,
133.64, 130.04, 120.38, 120.33, 118.68, 114.34, 112.59, 112.32, 75.86, 74.86, 68.72, 66.47, 65.58,
57.36, 55.93, 55.78, 53.98, 51.88 38.45, 38.06, 34.59, 31.05, 30.85, 29.45, 29.37, 26.90, 26.77,
26.26, 26.13, 25.39, 22.53, 21.18. HRMS, calc. 749.4377 [C43H60N2O9+H]+, found 749.4380
[M+H]+. HPLC [0-100% Solvent B, 20 min]: Rt = 16.87 min, 97 %.
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
(S)-(R)-3-(3,4-dimethoxyphenyl)-1-(3-(2-morpholinoethoxy)phenyl)propyl
1-
((1R,3R,4S,4aR,8aR)-3,4-dihydroxydecahydronaphthalene-1-carbonyl)piperidine-2-
carboxylate (29). 28 (40 mg, 53 µmol) was stirred in 1.86 mL MeOH and concentrated 0.14 mL
HCl for 1.5 h at rt. Then the solvent was removed under reduced pressure. Compound 29 (11.5
mg, 16.2 µmol, 30.3 %) was obtained after purification by preparative HPLC (40-50 % solvent B).
1H NMR (599 MHz, DMSO-d6) δ 7.26 – 7.22 (m, 1H), 6.88 – 6.83 (m, 3H), 6.82 – 6.79 (m, 1H),
6.74 – 6.71 (m, 1H), 6.66 – 6.62 (m, 1H), 5.58 (dd, J = 8.8, 4.7 Hz, 1H), 5.23 (d, J = 6.1 Hz, 1H),
4.11 – 4.00 (m, 3H), 3.95 (d, J = 13.2 Hz, 1H), 3.69 (d, J = 11.7 Hz, 6H), 3.57 – 3.52 (m, 4H),
ACS Paragon Plus Environment
26

Page 27 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3.49 (td, J = 8.8, 4.8 Hz, 1H), 3.06 (td, J = 13.1, 2.9 Hz, 1H), 2.76 – 2.69 (m, 1H), 2.65 (q, J = 7.6,
6.5 Hz, 2H), 2.43 (tt, J = 9.2, 4.3 Hz, 4H), 2.18 – 2.12 (m, 1H), 2.11 – 2.05 (m, 1H), 2.02 – 1.93
(m, 2H), 1.92 – 1.86 (m, 1H), 1.76 – 1.69 (m, 1H), 1.68 – 1.60 (m, 3H), 1.59 – 1.46 (m, 3H), 1.36
13
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
– 1.28 (m, 2H), 1.24 – 1.16 (m, 3H), 1.14 – 1.03 (m, 3H), 0.88 – 0.76 (m, 2H). C NMR (151
MHz, DMSO-d6) δ 174.49, 170.79, 158.87, 149.04, 147.49, 142.36, 133.60, 130.00, 120.38,
118.76, 114.33, 112.57, 112.31, 107.78, 79.10, 75.90, 72.55, 66.60, 57.42, 55.92, 55.78, 54.05,
52.01, 43.48, 43.35, 37.91, 31.03, 30.34, 30.19, 29.60, 29.05, 26.87, 25.96, 25.85, 25.38, 21.06.
HRMS, calc. 709.4064 [C40H56N2O9+H]+, found 709.4089 [M+H]+. HPLC [0-100% Solvent B,
20 min]: Rt = 14.28 min, 96 %.
Analysis of apoptosis
Analysis of DNA content by propidium iodide incorporation was performed in permeabilized cells
by flow cytometry. Cells (1×105) were harvested following incubation with the drug, washed in
PBS and resuspended in 500 L of a solution containing 0.1% sodium citrate w/v, 0.1% Triton X-
100 v/v and 50 g/mL propidium iodide (Sigma Chemical Co, Gallarate, Italy). Following
incubation at 4°C for 30 min in the dark, cell nuclei were analyzed with a Becton Dickinson
FACScan flow-cytometer. Cellular debris were excluded from analysis by raising the forward
scatter threshold and the DNA content of the nuclei was registered on a logarithmic scale. The
percentage of the elements in the hypodiploid region was calculated.
Apoptosis was measured using propidium iodide in double staining with annexin V-FITC (BD
Biosciences, NJ, USA). Annexin V binds to phosphatidylserine, which, in dying cells, is exposed
on the outer leaflet of plasma membrane. Propidium iodide does not stain cells with intact
membrane, as it occurs in early apoptosis. Briefly, 1×105 cells were resuspended in 100 L of
binding buffer (10 mM Hepes/NaOH pH 7.5, 140 mM NaCl, 2.5 mM CaCl2) containing 5 μL of
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 35
1
2
3
4
5
6
7
8
9
annexin V-FITC (Pharmingen/Becton Dickinson, San Diego, CA, USA) and 10 μl of PI solution
without Triton. Cells were incubated for 15 min at room temperature in the dark and analyzed with
C6 BDAccuri flow cytometer (Becton Dickinson).
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
Statistical analysis
Student’s t-test was used to assess differences between means of values. A p-value of ≤ 0.05 was
considered significant.
ACS Paragon Plus Environment
28

Page 29 of 35
Journal of Medicinal Chemistry
1
2
3
4
5
6
ANCILLIARY INFORMATION
Supporting Information
7
8
9
Further experimental details (Suppl. Fig. 1 - 6; Suppl. Tab. 1 & 2, crystallograhic data collection
and refinement statistics, NMR spectra, biological, biochemical and synthetic procedures of all
remaining compounds, list of SI items incl. Molecular Formula Strings) are provided in the
Supporting Information which is available free of charge on the ACS Publications website at DOI:
XXX
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
Accession Codes
Atomic coordinates for the X-ray structure of compound 25 in complex with FKBP51 (16-140)-
A19T (PDB code 6SAF) are available from the RCSB Protein Data Bank (www. rcsb.org).
Authors will release the atomic coordinates and experimental data upon article publication.
AUTHOR INFORMATION
Corresponding Author
*Email: hausch@drugdiscovery.chemie.tu-darmstadt.de
ORCID
Felix Hausch: 0000-0002-3710-8838
Present Address
Technical University Darmstadt, Alarich-Weiss-Str. 4, 64287 Darmstadt, Germany.
ACS Paragon Plus Environment
29