Development of Fluorine-18 Labeled Metabolically Activated Tracers for Imaging of Drug Efflux Transporters with Positron Emission Tomography

Article abstract of DOI:10.1021/acs.jmedchem.5b00652

Increased activity of efflux transporters, e.g., P-glycoprotein (P-gp) and breast cancer resistance protein (BCRP), at the blood-brain barrier is a pathological hallmark of many neurological diseases, and the resulting multiple drug resistance represents a major clinical challenge. Noninvasive imaging of transporter activity can help to clarify the underlying mechanisms of drug resistance and facilitate diagnosis, patient stratification, and treatment monitoring. We have developed a metabolically activated radiotracer for functional imaging of P-gp/BCRP activity with positron emission tomography (PET). In preclinical studies, the tracer showed excellent initial brain uptake and clean conversion to the desired metabolite, although at a sluggish rate. Blocking with P-gp/BCRP modulators led to increased levels of brain radioactivity; however, dynamic PET did not show differential clearance rates between treatment and control groups. Our results provide proof-of-concept for development of prodrug tracers for imaging of P-gp/BCRP function in vivo but also highlight some challenges associated with this strategy.

Full text of DOI:10.1021/acs.jmedchem.5b00652

Subscriber access provided by University of Otago Library
Article
Development of Fluorine-18 Labeled Metabolically Activated
Tracers for Imaging of Drug Efflux Transporters with PET
Kerstin Sander, Eva Galante, Thibault Gendron, Elena Yiannaki, Niral Patel, Tammy L Kalber,
Adam Badar, Mathew Robson, Sean P Johnson, Florian Bauer, Severin Mairinger, Johann Stanek,
Thomas Wanek, Claudia Kuntner, Tim Kottke, Lilia Weizel, David Dickens, Kjell Erlandsson,
Brian F Hutton, Mark F Lythgoe, Holger Stark, Oliver Langer, Matthias Koepp, and Erik Årstad
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00652 • Publication Date (Web): 10 Jul 2015
Downloaded from http://pubs.acs.org on July 12, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 104
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 104
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
Development of Fluorineꢀ18 Labeled Metabolically
Activated Tracers for Imaging of Drug Efflux
Transporters with PET
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
2
3
Kerstin Sander, Eva Galante, Thibault Gendron, Elena Yiannaki, Niral Patel, Tammy L.
3
3
4
4
5
6
8
9
Kalber, Adam Badar, Mathew Robson, Sean P. Johnson, Florian Bauer, Severin Mairinger,
6
6
6
7
7
Johann Stanek, Thomas Wanek, Claudia Kuntner, Tim Kottke, Lilia Weizel, David Dickens,
1
1
3
7,#
Kjell Erlandsson, Brian F. Hutton, Mark F. Lythgoe, Holger Stark, Oliver Langer,
10
1,2,*
Matthias Koepp, Erik Årstad
1
University College London, Institute of Nuclear Medicine, 235 Euston Road – T5, London
2
NW1 2BU, UK; University College London, Department of Chemistry, 20 Gordon Street,
3
London WC1H 0AJ, UK; University College London, Centre for Advanced Biomedical
4
Imaging, 72 Huntley Street, London WC1E 6DD, UK; University College London, Cancer
5
Institute, 72 Huntley Street, London WC1E 6DD, UK; University of Vienna, Faculty of Life
6
Sciences, Department of Medicinal Chemistry, Althanstrasse 14, Aꢀ1090 Vienna, Austria; AIT
Austrian Institute of Technology GmbH, Health and Environment Department, Aꢀ2444
7
Seibersdorf, Austria;
Johann Wolfgang Goethe University, Institute of Pharmaceutical
8
Chemistry, Biocenter, MaxꢀvonꢀLaueꢀStrasse 9, 60438 Frankfurt am Main, Germany;
University of Liverpool, Institute of Translational Medicine, Department of Molecular and
Clinical Pharmacology, The Wolfson Centre for Personalised Medicine, Block A Waterhouse
ACS Paragon Plus Environment

Page 3 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
9
Buildings, 1–5 Brownlow Street, Liverpool L69 3GL, UK; Medical University of Vienna,
1
0
Department of Clinical Pharmacology, WaehringerꢀGuertel 18ꢀ20, Aꢀ1090 Vienna, Austria;
University College London, Institute of Neurology, Department of Clinical and Experimental
Epilepsy, Queen Square, London WC1N 3BG, UK
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

Journal of Medicinal Chemistry
Page 4 of 104
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Increased activity of efflux transporters, e.g. Pꢀglycoprotein (Pꢀgp) and breast cancer resistance
protein (BCRP), at the bloodꢀbrain barrier is a pathological hallmark of many neurological
diseases, and the resulting multiple drug resistance represents a major clinical challenge. Nonꢀ
invasive imaging of transporter activity can help to clarify the underlying mechanisms of drug
resistance, and facilitate diagnosis, patient stratification and treatment monitoring. We have
developed a metabolically activated radiotracer for functional imaging of Pꢀgp/BCRP activity
with positron emission tomography (PET). In preclinical studies, the tracer showed excellent
initial brain uptake and clean conversion to the desired metabolite, although at a sluggish rate.
Blocking with Pꢀgp/BCRP modulators led to increased levels of brain radioactivity; however,
dynamic PET did not show differential clearance rates between treatment and control groups.
Our results provide proofꢀofꢀconcept for development of proꢀdrug tracers for imaging of Pꢀ
gp/BCRP function in vivo, but also highlight some challenges associated with this strategy.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 5 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Membrane transporters at the bloodꢀbrain barrier (BBB) play an important role in maintaining
this physiological barrier and hence, brain homeostasis: selective uptake transporters ensure the
supply of essential nutrients like glucose, amino acids and nucleotides, whilst efflux transporters
extrude toxic metabolites from the brain and prevent xenobiotics from reaching the central
nervous system (CNS), thus providing a cellular detoxification system. The most prevalent efflux
transporters at the BBB endothelium belong to the superfamily of adenosine triphosphateꢀ
binding cassette (ABC) transporters, and include Pꢀglycoprotein (Pꢀgp; ABCB1) and breast
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
cancer resistance protein (BCRP; ABCG2). Pꢀgp and BCRP are located at the luminal
membrane of the endothelial cells, where they extrude an array of structurally different
compounds. Pharmacophore models developed for Pꢀgp substrates suggest that compounds
exhibiting two hydrogen bond acceptors, two aromatic rings and additional hydrophobic features
2
are readily transported by the efflux pump. Recent research shows considerable overlap in
3
substrate scope of Pꢀgp and BCRP, and an increasing body of evidence suggests that they coꢀ
4
localize and cooperate in transporting compounds at the BBB. Drug entry into the brain is
therefore highly restricted, and is largely limited to compounds that are not recognized by efflux
transporters and are able to cross the BBB by passive diffusion. In the course of many prevalent
CNS disorders, especially refractory diseases like epilepsy, mood disorders and brain cancers,
the delicate balance of transporter activity is distorted. The BBB becomes ‘leaky’, enabling toxic
substances to temporarily reach the CNS, and, as a compensatory mechanism, activity and
5
expression levels of efflux transporters increase. As a consequence of transporter upregulation
at the BBB, CNS drug concentrations in the brain become too low to achieve a therapeutic effect.
The resulting multiple drug resistance (MDR) affects 20–40% of all patients suffering from
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 104
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
epilepsy, depression and schizophrenia, and is even more prevalent in drug refractory brain
6
cancer and following human immunodeficiency virus infections.
Quantitative assessment of efflux pump function by positron emission tomography (PET) can
help to clarify the underlying mechanisms of drug resistance in individual patients, and facilitate
stratification of patients for specific treatment strategies. As Pꢀgp/BCRP mediates extrusion of
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
6
c,7
the most commonly prescribed CNS drugs,
a PET tracer that enables functional imaging of
this transporter system can potentially be applicable to large patient cohorts spanning a range of
diverse neurological diseases.
In recent years, a number of strategies for imaging of drug efflux transporters have been
1
1
evaluated. Tracers that have been designed as inhibitors, e.g. [ C]tariquidar and analogues, often
8
act as substrates when administered in the low doses that are required for imaging with PET.
11
These Pꢀgp/BCRP modulators, and Pꢀgp substrates like [ C]verapamil all have significant
limitations: poor metabolic stability, low brain uptake (substrates), low binding affinity
(inhibitors), and insufficient selectivity make quantitative imaging of drug efflux pumps
9
challenging. A more recent approach is the development of metabolically activated (proꢀdrug)
tracers, as exemplified by carbonꢀ11ꢀ and fluorineꢀ18ꢀlabeled 6ꢀhalopurines for imaging of
1
0
multidrug resistance protein 1. The soꢀcalled metabolite extrusion method (MEM) relies on
passive diffusion of a proꢀdrug tracer into the organ of interest, e.g. the brain, its subsequent
trapping by metabolic conversion into an efflux pump substrate (‘lockꢀin’ mechanism), and
11
transporterꢀmediated extrusion of the metabolite. Provided that the physicochemical properties
of the metabolite preclude passive diffusion across membranes, drug efflux transporter function
can be directly correlated to the clearance rate of radioactivity from the target tissue, and hence
readily be quantified by dynamic imaging. In order to allow MEM to be exploited for
ACS Paragon Plus Environment

Page 7 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
quantitative imaging of Pꢀgp/BCRP function in vivo, a proꢀdrug tracer should have a high initial
brain uptake and be rapidly transformed into a single radioactive metabolite that acts as a dual Pꢀ
gp/BCRP substrate, but does not interact with other transporters or targets in the brain. The
radioactive metabolite should be unable to cross membranes by passive diffusion, so that in the
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
9
b
absence of Pꢀgp and BCRP, it is effectively trapped.
Histamine H receptor (H R) antagonists are widely used for the treatment of allergies, and are
1
1
often characterized by their ability to cross the BBB: compounds of the first generation readily
enter the brain causing H Rꢀmediated side effects, whereas substances of the second generation
1
12
are avid Pꢀgp substrates and hence give minimal exposure to the CNS. Interestingly, several
H R antagonists currently in clinical use are active metabolites of early generation drugs, e.g.
1
cetirizine and fexofenadine are zwitterions formed by enzymatic oxidation of hydroxyzine and
13
terfenadine, respectively. These wellꢀcharacterized metabolic pairs of H R antagonists were
1
used as templates for the design of radiolabeled proꢀdrug tracers and metabolically activated dual
Pꢀgp/BCRP substrates (Figure 1).
RESULTS & DISCUSSION
Chemistry. Using the H R antagonists fexofenadine and cetirizine as templates, a library of
1
compounds was prepared (Scheme 1 and Scheme 2) in order to identify (i) structural motifs
required for Pꢀgp substrate activity, (ii) functionalities that could be exploited as proꢀdrug
moieties for metabolic activation, and (iii) potential positions for labeling with fluorineꢀ18
(
Figure 1). With the aim to circumvent the need for complex synthetic routes, 1ꢀ
benzhydrylpiperazine (scaffold A, derived from cetirizine) and diphenyl(piperidinꢀ4ꢀyl)methanol
(
scaffold B, derived from fexofenadine) derivatives decorated with benzylic and aliphatic side
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 104
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
chains were initially evaluated. Scaffold B as well as its monoꢀ and difluorinated analogues C
and D, which were obtained by Grignard reactions with paraꢀsubstituted piperidine derivatives
(vide infra), were reacted with suitably functionalized benzylic halides by means of Nꢀalkylation
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
14
to give compounds 1, 3, 6, 9, 12, 14, and 16 (Scheme 1). The alkylating reagents were either
commercially available, or prepared from the respective tolyl derivatives by radical bromination
with Nꢀbromosuccinimide. The alcohol 1 was further modified by Swern oxidation to obtain
1
5
aldehyde 2, whereas hydrolysis of the esters 3, 6, 9, 12, 14, and 16 under basic conditions gave
the corresponding carboxylic acids 4, 7, 10, 13, 15, and 17.
A fluorinated aliphatic side chain was prepared as depicted in Scheme 2. Treatment of αꢀ
1
6
hydroxyꢀγꢀbutyrolactone with DAST provided the fluorinated lactone 18, which after
17
delactonization with methanol gave alcohol 19, albeit in moderate yield (20%). The alcohol 19
was converted into the corresponding tosylate 20, which was used for subsequent Nꢀalkylation of
1
8
scaffolds A and B. The resulting esters 21 and 25 were either reduced to give alcohols 22 and
19
20
2
6, or hydrolysed to yield the acids 24 and 28. Finally, treatment of the alcohols 22 and 26
with acetic anhydride provided the acyl esters 23 and 27.
18
18
As biodistribution studies with [ F]23 and [ F]27 pointed to rapid defluorination of the
aliphatic side chain in vivo (vide infra), a second set of compounds was designed consisting of
fluorinated aromatic scaffolds and aliphatic side chains carrying proꢀdrug groups (alcohol,
aldehyde and ester) as putative tracer candidates. To enable tuning of lipophilicity, additional
scaffolds (CꢀF) were prepared (Scheme 3). Treatment of Bocꢀprotected piperidineꢀ4ꢀcarboxylic
2
1
acid ethyl ester with 4ꢀfluorophenylmagnesium bromide gave the difluorinated compound 33.
The corresponding monoꢀfluorinated analogue 32 was prepared from Bocꢀprotected 4ꢀ
2
2
hydroxymethylꢀpiperidine using sequential oxidation and Grignard reactions. Deprotection of
ACS Paragon Plus Environment

Page 9 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds 31ꢀ33 provided the monoꢀ and difluorinated scaffolds 34, 35 and 37 (= scaffold C, D
and F), whereas treatment of 32 with trifluoroacetic acid resulted in elimination of the tertiary
alcohol to give compound 36 (= scaffold E). Nꢀalkylation of scaffolds C, E and F with
appropriately functionalized alkyl bromides provided the putative proꢀdrug tracers and nonꢀ
radioactive reference compounds 38–47 (Figure 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to enable labeling of the putative tracers 23 and 27 with fluorineꢀ18, the corresponding
mesylate precursors 52 and 56 were prepared from a dioxolane building block (Scheme 4). The
tosylate 48 was coupled to scaffolds A and B, and the resulting protected diols 49 and 53 were
subsequently hydrolyzed under acidic conditions to give vicinal diols 50 and 54. Selective
acetylation of the primary alcohol in compounds 50 and 54 was achieved with acetic anhydride
to give the acetates 51 and 55, which upon mesylation gave the desired precursors 52 and 56,
respectively.
Whereas labeling of aliphatic groups with fluorineꢀ18 is well established, aromatic
18
23
[
F]fluorination of small molecules bearing complex functional groups remains a challenge.
24
Limitations of existing labeling methods prompted us to explore the use of sulfonium salts as
1
8
25
leaving groups for aromatic [ F]fluorination of the scaffolds described above. For the
synthesis of the sulfonium salts 62 and 64 bearing an electron deficient scaffold, the ketone 59
was prepared by a lithiumꢀmediated reaction of bromoarene 57 and Weinreb amide 58 (Scheme
5). After Boc deprotection of 59, the desired side chains were introduced by Nꢀalkylation, and
the ammonium salts of the resulting diaryl thioethers 61 and 63 were treated with
diphenyliodonium triflate in the presence of copper(II) benzoate to give the triarylsulfonium salts
62 and 64, respectively. Labeling of compounds with the electron neutral scaffold E required
decoration of the sulfonium salts with anisole moieties in order to achieve regioselective
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 104
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
18
[
F]fluorination. The thioether 65 was converted into ketone 66 as described above. Addition of
phenylmagnesium bromide to 66 provided alcohol 67, which upon treatment with trifluoroacetic
acid yielded the deprotected 4ꢀ(methylene)piperidine 68. Following the route outlined above, Nꢀ
alkylation of 68 gave compounds 69, 71 and 73, which were converted into the corresponding
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
6
triarylsulfonium salts 70, 72 and 74 using bisꢀ(4ꢀmethoxyphenyl)iodonium triflate 75.
Radiochemistry. Radiolabeling of the mesylate precursors 52 and 56 (Scheme 6) with
18
[
F]fluoride worked well in dimethyl sulfoxide under heating (20 min at 90 °C). Acetyl esters
18
18
[ F]23 and [ F]27 were obtained in 40–50% analytical radiochemical yield (RCY) as
determined by radioꢀHPLC. The isolated products were radiochemically pure but the specific
activity was low (0.5 GBq/µmol) due to extensive decomposition of the respective precursors.
1
8
18
18
18
As attempted hydrolysis of [ F]23 and [ F]27 to the corresponding alcohols [ F]22 and [ F]26
18
18
further reduced the specific activity, the acetyl esters [ F]23 and [ F]27 were used for an initial
18
18
biological evaluation of compounds [ F]22 and [ F]26 assuming that the esters would be
rapidly cleaved by esterases in the blood after intravenous injection (Scheme 6).
18
[
F]Fluorination of triarylsulfonium salts gave the highest radiochemical yields when using a
combination of potassium bicarbonate, kryptofix and DMSO with heating to 110 °C for 15
25
min. In the case of substrates containing electron withdrawing groups on the aromatic moiety,
i.e. the ketones 62 and 64 (Scheme 7), the radiochemical reaction was exceptionally clean and
1
8
18
yielded [ F]45 (36–40% decayꢀcorrected isolated RCY, n = 3) and [ F]47 (22% decayꢀ
corrected isolated RCY), respectively, without formation of any radioactive side products.
18
18
Hydrolysis of [ F]45 afforded the alcohol [ F]46 (18% decayꢀcorrected isolated RCY after two
1
8
steps), whereas treatment of [ F]45 with phenylmagnesium bromide led to formation of the
fexofenadineꢀlike scaffold C with concurrent cleavage of the ester group to give the alcohol
ACS Paragon Plus Environment

Page 11 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
18
18
[
F]39 (16% decayꢀcorrected isolated RCY after two steps). Subsequent reaction of [ F]39 with
1
8
acetyl chloride provided the acyl ester [ F]38.
Compounds containing the nonꢀactivated ritanserinꢀlike scaffold E were prepared by direct
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
18
labeling of the respective sulfonium salts 70, 72 and 74 with [ F]fluoride (Scheme 8). The
1
8
18
18
resulting esters [ F]40a and [ F]43 as well as the acetal [ F]41 were hydrolyzed to give the
1
8
18
18
alcohol [ F]40, the carboxylic acid [ F]44 and the aldehyde [ F]43, respectively. The
18
18
18
18
18
radiochemical purity of the desired radiotracers ([ F]38ꢀ[ F]40, [ F]42ꢀ[ F]44, [ F]46 and
18
[ F]47) was > 96%, however, the use of multiꢀstep synthesis resulted in moderate overall yields
(
2–18% decayꢀcorrected isolated RCY after formulation in saline).
In our hands, preparation of sulfonium salts from the corresponding thioethers provided a
practical and versatile synthetic method that gave rapid access to a number of structurally diverse
precursors for labeling. Conveniently, triarylsulfonium salts are highly stable, easy to handle, and
can readily be purified using conventional techniques including extractive aqueous workup and
column chromatography on silica gel. In labeling studies, the precursors showed a remarkable
18
reactivity with [ F]fluoride: activated substrates reacted at room temperature, and with heating
the reaction proceeded in the presence of hydrogen bond donors. Furthermore, efficient and
1
8
regioselective [ F]fluorination of nonꢀactivated aromatic moieties was achieved using sulfonium
salts decorated with electron rich spectator ligands. The ability to label a spectrum of electron
18
neutral to electron deficient aromatic groups with F in the presence of basic moieties, as well as
hydrogen bond donors, makes the method broadly applicable to drugꢀlike compounds, and hence
opens up extensive pharmacological space for the design of small molecule PET tracers.
In vitro compound screening. A luminescent Pꢀgp ATPase assay was initially used to
establish if the simplified fexofenadine derivatives retained the Pꢀgp substrate activity of the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 104
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
parent compounds, and to identify putative metabolic pairs that could be exploited for the design
of a proꢀdrug tracer. The assay allows rapid screening of Pꢀgp substrate affinity, but the accuracy
2
7
is insufficient to determine IC50 values. In order to determine structural elements that are
prerequisite to efflux pump substrate activity, ATP turnover triggered by the carboxylic acids 4,
7 and 10 was compared to that of parent compound fexofenadine (Figure 3A). Of the test
compounds, the benzoic acid 4 evoked the highest ATPase activity, which was similar to that of
fexofenadine, whereas ketone 7 and isobutyric acid 10 appeared to be poor Pꢀgp substrates. To
evaluate functional groups for the design of proꢀdrug tracers, the benzylic alcohol 1,
benzaldehyde 2, and benzoic ester 3 were tested (Figure 3B). The results pointed to a gradual
increase in Pꢀgp affinity with increasing oxidation state, with alcohol 1 resulting in modest ATP
consumption, whereas the aldehyde 2 and ester 3 appeared to be more avid substrates.
Remarkably, incorporation of fluorine into the benzoic moiety of 4 to give compound 13
practically abolished the Pꢀgp substrate affinity (Figure 3C). However, fluorine substituents on
scaffold B were well tolerated, and the monoꢀ and difluorinated derivatives 15 and 17 evoked
ATP turnover comparable to that of compound 4. The use of a fluorinated aliphatic side chain, as
for alcohol 22 and carboxylic acid 24, also appeared to be tolerated, and evoked ATP
consumption to the same degree as their parent compounds hydroxyzine, and its metabolite
cetirizine, respectively (Figure 3D). The results suggest that the zwitterions 4 and 24 are good Pꢀ
gp substrates, whereas the corresponding alcohols 1 and 22 are likely to evade active efflux, and
hence are suitable leads for proꢀdrug tracer development.
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
Tracer Evaluation in Vivo. The putative proꢀdrug tracers were evaluated using biodistribution
studies in mice. The tissue distribution of radioactivity at 5 and 30 min postꢀinjection (p.i.) was
used to assess brain uptake, brain/blood ratios (Table 1), and the rate of clearance. Two
ACS Paragon Plus Environment

Page 13 of 104
Journal of Medicinal Chemistry
1
2
3
18
derivatives labeled at the aliphatic side chain, namely [ F]23 (cetirizineꢀlike scaffold A) and
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
8
[
F]27 (fexofenadineꢀlike scaffold B), were initially investigated to guide the choice of scaffold.
As discussed above, we envisaged that ester hydrolysis would lead to rapid formation of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
18
18
18
18
resulting alcohols [ F]22 and [ F]26 in vivo. Administration of [ F]23 and [ F]27 resulted in
good initial uptake of radioactivity in the brain (5.8 and 7.2% injected dose per gram tissue
(
ID/g) at 5 min p.i., respectively), however, high uptake (3–6% ID/g at 30 min p.i.) was also
observed in bone tissue (femur and skull) suggesting rapid enzymatic defluorination in vivo.
Hence, we abandoned further evaluation of derivatives containing fluorinated aliphatic side
chains, and instead evaluated proꢀdrug tracers bearing fluorineꢀ18 at aromatic moieties. Indeed,
1
8
18
administration of the acetate ester [ F]38, a structurally closely related analogue to [ F]27,
exhibited far superior CꢀF bond stability as judged by the low level of radioactivity in bone tissue
18
(
< 2% ID/g at 30 min p.i.). However, the brain uptake and brain/blood ratio of alcohol [ F]39, as
1
8
well as the corresponding acetyl ester [ F]38, were low. To improve the initial uptake of the proꢀ
1
8
drug tracer in the brain, we prepared an analogue of [ F]39 for which the hydroxyl group in
18
scaffold B was replaced by an alkene to give scaffold E. The resulting proꢀdrug tracer [ F]40
exhibited excellent initial brain uptake (9.4% ID/g at 5 min p.i.) with high brain/blood ratios
(
10.4 and 7.3 at 5 and 30 min p.i., respectively). In contrast, administration of the corresponding
18
carboxylic acid [ F]44 resulted in an initial brain uptake of 1.3% ID/g at 5 min p.i. with a
brain/blood ratio of 0.4. In accordance with the results from the in vitro assay, replacement of the
18
18
alcohol with an aldehyde ([ F]42) or ester ([ F]43) as the proꢀdrug group proved futile, and
1
8
resulted in moderate brain uptake (3.1 and 2.9 ID/g at 5 min p.i., respectively). Alcohol [ F]46,
derived from scaffold F, showed good initial brain uptake (5.0% ID/g at 5 min p.i.), but brain
18
clearance was rapid (1.8% ID/g at 30 min p.i.). The structurally related ester [ F]47 hardly
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 104
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
penetrated the brain, most likely due to the rapid hydrolysis in vivo to give the corresponding
carboxylic acid (70% conversion at 5 min p.i.).
1
8
The 4ꢀhydroxybutyl proꢀdrug tracer [ F]40 showed high initial brain uptake and a favorable
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
clearance from the brain and was thus investigated further (Figure 4). The brain metabolism was
1
8
exceptionally clean and gave the desired carboxylic acid [ F]44 as the only radioactive
18
18
metabolite. The ratio of [ F]40 to [ F]44 in the brain decreased over time and was determined
to be 6.1 at 30 min p.i., 1.5 at 60 min p.i., 0.6 at 90 min p.i. and 0.4 at 120 min p.i., respectively.
1
8
However, as [ F]44 is extruded from the brain into the periphery, this ratio does not reflect the
actual rate of metabolism. In the blood, oxidation of the alcohol was faster: whilst low amounts
18
18
of [ F]40 were still detectable, [ F]44 constituted between 70 and 85% of total amount of
radioactivity detected between 60 and 120 min p.i. (Figure 4C/D). Development of a MEM
tracer requires that the radioactive metabolite does not cross the BBB to any significant extent by
other mechanisms than extrusion by the targeted efflux pump. As the brain/blood ratios observed
1
8
after administration of the carboxylic acid [ F]44 remained at or below 1.1 over the course of
18
two hours after administration (Figure 4B), our results suggest that brain clearance from [ F]44
1
8
is largely mediated by active efflux. However, a faster metabolism of the proꢀdrug tracer [ F]40
1
8
18
would be desirable. The aldehyde [ F]42 and ester [ F]43 were designed as to accelerate
18
18
formation of the carboxylic acid [ F]44 in vivo. Although metabolism to the zwitterions [ F]44
was completed within five (ester) and 60 (aldehyde) minutes, respectively, their low brain/blood
ratios rendered them unsuitable as proꢀdrug tracers (cf. Table 1).
1
8
The proꢀdrug tracer [ F]40 was investigated further to determine if the corresponding
18
metabolite [ F]44 acted as an efflux pump substrate in vivo. By injecting the dual Pꢀgp/BCRP
inhibitor tariquidar one to two hours prior to tracer administration at a dose of 15 mg/kg
ACS Paragon Plus Environment

Page 15 of 104
Journal of Medicinal Chemistry
1
2
3
bodyweight it was ensured that Pꢀgp was fully blocked over the course of the biodistribution
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
8
studies (longest time point 120 min p.i.). The tariquidar dose is, however, not sufficiently high
to fully inhibit BCRP. In a separate experiment, injection of elacridar (10 mg/kg bodyweight)
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
8c
was used to completely block both Pꢀgp and BCRP. The comparison of the organ distribution
without and with concomitant administration of tariquidar and elacridar, respectively (Figure
5
A), showed a similar brain uptake at 30 min p.i. (ca. 8% ID/g) in all animals, but revealed a
significant effect on efflux pump inhibition at 60–120 min p.i.: when blocking with tariquidar,
tracer brain uptake (in %ID/g) at 60 and 90 min p.i. was 30 and 50% higher than in untreated
animals. Using elacridar, the amount of tracer in the brain was doubled between 60 and 120 min
p.i. Tracer levels in the blood remained constant between 1 and 2% ID/g in all animals. The
1
8
results provide evidence for active transport of the metabolically activated tracer [ F]44 across
the BBB by Pꢀgp as well as BCRP.
Dynamic PET scans were performed in untreated mice, mice treated with either tariquidar or
elacridar, and Pꢀgp knockout mice (Figure 5B/C). Whole brain timeꢀactivity curves were fitted
with the sum of a constant (irreversible compartment, representing the amount of tracer trapped
in the brain) and a monoꢀexponential function (reversible compartment, representing the amount
2
9
of tracer washed out from the brain and the efflux rate). For analysis and determination of
efflux rates, data from 10 min p.i. onwards was used as this gave the best match between the
fitted curves and the experimental results. Focusing on the time from 60 min p.i. onwards, i.e.
18
when increasing formation of the metabolite [ F]44 was observed, did not change the
quantitative outcome measure.
In a first set of experiments (performed at AIT), untreated, tariquidar treated and Pꢀgp
knockout mice (Figure 5B) showed comparable initial brain uptake. The ratio of tracer that was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 104
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
washed out from the brain to tracer that was trapped (washout/trapped ratio) was similar in
untreated and knockout mice (2.9 and 2.8, respectively), whereas in tariquidar treated mice, the
washout/trapped ratio was lower (1.5). This suggests that more tracer was trapped in the brains
of the treated mice. In a second set of experiments (performed at UCL), elacridar treated mice
were compared to untreated animals and showed a lower washout/trapped ratio (2.3 versus 4.1).
The marginally higher peak brain uptake in elacridar treated mice compared to the control group
may have contributed to the increased amounts of radioactivity at later time points (Figure 5C). It
should be noted that discrepancies between the two control groups (washout/trapped ratio of 1.5
and 2.3, respectively) are most likely due to variations in the experimental setup. The difference
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
18
in washout/trapped ratios provides evidence for [ F]44 acting as a dual Pꢀgp/BCRP substrate,
and the experiments in Pꢀgp knockout mice suggest that BCRP can compensate for Pꢀgp when
18
extruding [ F]44. However, minor differences in efflux rates when comparing untreated and
1
8
18
treated mice raise the possibility that not only tracer [ F]44 but also the proꢀdrug tracer [ F]40
was washed out of the brain, either by passive diffusion or by active transport as an efflux pump
substrate.
Due to the complimentary and partially overlapping substrate scope of efflux pumps belonging
to the large family of ABC transporters it is possible that transporters other than Pꢀgp/BCRP
18
18
30
contribute to the efflux of [ F]40/[ F]44. The tracer kinetics can also be affected by other
factors such as offꢀtarget binding. Indeed, both the alcohol 40 and the carboxylic acid 44 showed
high affinity to the human H R (2.3 ± 0.4 and 18 ± 4 nM, respectively) as determined in an in
1
vitro ligand displacement assay. The affinity to H R may contribute to the retention of the proꢀ
1
18
drug tracer [ F]40 in the brain and may affect the rate of oxidative metabolism. Moreover, the
18
18
metabolic activation, i.e. the conversion of [ F]40 to [ F]44, appeared to be slower than the
ACS Paragon Plus Environment

Page 17 of 104
Journal of Medicinal Chemistry
1
2
3
18
extrusion of the metabolite [ F]44, and as a result there was no significant difference between
the brain clearance rates of the control group and the two treatment groups. However, differences
in the oxidizing rate of enzymes, as well as in efflux transporter distribution and function, may
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
18
31
make [ F]40 more suitable as a proꢀdrug tracer in other species.
SUMMARY & CONCLUSIONS
With the aim to develop a proꢀdrug tracer to enable dynamic imaging of Pꢀgp/BCRP function
in vivo, we used H R antagonists as lead compounds to synthesize a library of putative proꢀdrug
1
tracers and their corresponding metabolites. In biodistribution studies, derivatives containing a
fluorinated aliphatic side chain were found to undergo rapid defluorination in vivo. The results
prompted us to develop triarylsulfonium salts as novel precursors for aromatic labeling of drugꢀ
like molecules with fluorineꢀ18. This labeling strategy provided access to a range of putative
1
8
proꢀdrug tracers and corresponding metabolites, and allowed identification of the alcohol [ F]40
1
8
as a lead proꢀdrug tracer candidate. [ F]40 fulfills many of the requirements of MEM: it has
excellent initial brain uptake (9.4% ID/g 5 min p.i.) and undergoes clean metabolic conversion to
18
the desired carboxylate [ F]44 in the brain. Blocking experiments with the dual Pꢀgp/BCRP
inhibitors tariquidar and elacridar led to significantly increased levels of radioactivity in the brain
in a period between 60 and 120 minutes after tracer injection, suggesting that the metabolite
18
[ F]44 is a substrate for both Pꢀgp and BCRP. However, there was no significant difference in
the brain efflux rates between control animals and treatment groups as measured by dynamic
PET. Because of the sluggish metabolic conversion and crossꢀreactivity with H R we were not
1
1
8
18
able to use [ F]40/[ F]44 for quantification of the efflux rate; however, the tracer is still likely
to provide a qualitative depiction of regional brain efflux rates.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 104
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
Our results provide proofꢀofꢀconcept for development of metabolically activated proꢀdrug
tracers for imaging of Pꢀgp/BCRP activity with PET. However, the results also highlight some of
the challenges in achieving appropriate rates of enzymatic conversion and brain clearance.
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. Reagents were purchased from SigmaꢀAldrich, Acros Organics or Fluorochem and
were used without further purification. Purification of nonꢀlabeled compounds by column
chromatography was performed on silica unless otherwise specified. For characterization of nonꢀ
1
13
labeled compounds, H and C NMR spectra were recorded at room temperature unless
otherwise specified. The respective instruments, Bruker Avance 400, 500 or 600 were operated
1
13
at a frequency of 400, 500 or 600 MHz for H and 125 or 150 MHz for C, respectively. Proton
19
decoupled F NMR spectra were recorded on a Bruker Avance 300 instrument at a frequency of
282 MHz. All spectra were internally referenced to the respective deuterated solvents. Chemical
shifts are reported in ppm and coupling constants (J) are given in Hertz (Hz). Full NMR
assignment was performed with the aid of multidimensional and long range experiments. High
resolution mass data were recorded on either a Thermo Finnigan MAT900xp (CI, EI) or a
MALDI microMX (TOF) mass spectrometer. Melting points were taken on a Gallenkamp
heating block and are uncorrected. Elemental analysis (C, H, N) was performed on a CE440
analyser (Exeter Analytical).
We have recently described the synthesis of compounds 29–32, 36, 37, 40, 43, 45, 46, 57–62,
25
16
32
32
33
6
5–72 and 75. Compounds 18, 34, 35 and 48 have been described by others, and
analytical data are in accordance with the published data.
ACS Paragon Plus Environment

Page 19 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
(
1-(4-(Hydroxymethyl)benzyl)piperidin-4-yl)diphenylmethanol
(1).
Under
inert
atmosphere, diphenyl(piperidinꢀ4ꢀyl)methanol hydrochloride (2.1 g, 7.0 mmol), (4ꢀ
chloromethyl)phenyl)methanol (1.0 g, 6.4 mmol), potassium carbonate (2.9 g, 21.0 mmol) and
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
potassium iodide (0.1 g, 0.64 mmol) were heated in acetone (25 ml) for 6 h. After cooling the
inorganic components were filtered off and the filtrate was purified by column chromatography
(
TEA 0.5%; DCM: methanol = 100: 0 ꢀ 98: 2). The resulting colorless oil was triturated with
1
diethyl ether to give a white solid (0.3 g, 13%). Mp: 119 °C. H NMR (DMSOꢀd , 500 MHz):
δ
6
7
.48 (d, J = 7.4 Hz, 4H, ph’ꢀ2,6H), 7.29ꢀ7.22 (m, 8H, phꢀ2,3,5,6H, ph’ꢀ3,5H), 7.10 (t, J = 7.3 Hz,
2
H, ph’ꢀ4H), 5.24 (s, 1H, COH), 5.11 (t, J = 4.9 Hz, 1H, CH OH), 4.46 (d, J = 5.5 Hz, 2H,
2
eq
CH OH), 3.41 (br s, 2H, NCH ), 2.81 (br s, 2H, pipꢀ2,6H ), 2.53ꢀ2.49 (m, 1H, pipꢀ4H), 2.06 (br
2
2
ax
ax
eq 13
s, 2H, pipꢀ2,6H ), 1.49ꢀ1.47 (m, 2H, pipꢀ3,5H ), 1.24 (d, J = 11.1 Hz, 2H, pipꢀ3,5H ). C
NMR (DMSOꢀd , 125 MHz): 147.3 (ph’ꢀ1C), 128.7 (phꢀ1C), 127.8 (ph’ꢀ2,6C), 126.3 (ph’ꢀ4C),
25.8 (phꢀ4C), 125.7 (ph’ꢀ3,5C, phꢀ2,3,5,6C), 78.5 (COH), 62.7 (CH OH), 62.3 (NCH ), 53.3
δ
6
1
2
2
+
(
pipꢀ2,6C), 43.3 (pipꢀ4C), 25.9 (pipꢀ3,5C). HRMS (m/z): [M+H] calcd. for C H NO ,
26 29 2
.
386.2120; found, 386.2129. Elemental analysis: calcd. for C H NO 0.75 H O, C 77.87, H
2
6
29
2
2
7
.67, N 3.49; found, C 77.94, H 7.39, N 3.49.
4
-((4-(Hydroxydiphenylmethyl)piperidin-1-yl)methyl)benzaldehyde (2). Oxalyl chloride
(0.22 ml, 2.6 mmol) was dissolved in DCM (5 ml) and the solution was cooled to < ꢀ50 °C.
Under inert conditions, DMSO (0.22 ml, 3.1 mmol) dissolved in DCM (2 ml) was added
dropwise and the solution was stirred for 15 min. Alcohol 1 (0.25 g, 0.65 mmol) dissolved in
DCM (5 ml) was added and the solution was stirred for further 15 min. The reaction was
quenched with TEA (0.8 ml, 5.7 mmol) and allowed to warm to room temperature. Water (30
ml) was added and the solution extracted with DCM. The organic phase was dried (MgSO4),
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 104
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
filtered, concentrated and crystallized using a mixture of petroleum ether: ethyl acetate = 3: 2 to
1
give the product as a white solid (0.09 g, 34%). Mp: 149 °C. H NMR (DMSOꢀd , 500 MHz):
δ
6
9.98 (s, 1H, CHO), 7.82 (d, J = 8.2 Hz, 4H, phꢀ2,6H), 7.53ꢀ7.48 (m, 6H, phꢀ3,5H, ph’ꢀ2,6H),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7.23 (t, J = 7.8 Hz, 4H, ph’ꢀ3,5H), 7.10 (t, J = 7.3 Hz, 2H, ph’ꢀ4H), 5.23 (s, 1H, COH), 3.51 (s,
eq
2H, NCH ), 2.76 (d, J = 11.2 Hz, 2H, pipꢀ2,6H ), 2.53ꢀ2.45 (m, 1H, pipꢀ4H), 1.97 (t, J = 11.0
2
ax
ax
eq 13
Hz, 2H, pipꢀ2,6H ), 1.53ꢀ1.44 (m, 2H, pipꢀ3,5H ), 1.23 (d, J = 12.8 Hz, 2H, pipꢀ3,5H ). C
NMR (DMSOꢀd , 150 MHz): 192.8 (CHO), 147.3 (ph’ꢀ1C), 146.2 (phꢀ4C), 135.1 (phꢀ1C),
29.5 (phꢀ2,6C), 129.2 (phꢀ3,5C), 127.8 (ph’ꢀ3,5C), 125.8 (ph’ꢀ4C), 125.8 (ph’ꢀ2,6C), 78.5
δ
6
1
+
(
COH), 62.0 (NCH ), 53.6 (pipꢀ2,6C), 43.3 (pipꢀ4C), 26.0 (pipꢀ3,5C). HRMS (m/z): [M+H]
2
.
calcd. for C H NO , 386.2120; found, 386.2106. Elemental analysis: calcd. for C H NO 0.5
2
6
27
2
26 27
2
H O, C 79.16, H 7.15, N 3.55; found, C 78.95, H 6.90, N 3.48.
2
Methyl 4-((4-(hydroxydiphenylmethyl)piperidin-1-yl)methyl)benzoate (3). Under inert
atmosphere, diphenyl(piperidinꢀ4ꢀyl)methanol hydrochloride (2.0 g, 6.6 mmol), methyl 4ꢀ
(bromomethyl)benzoate (1.26 g, 5.5 mmol), potassium carbonate (2.73 g, 19.8 mmol) and
potassium iodide were suspended in acetonitrile (50 ml) and refluxed overnight. After cooling to
room temperature the inorganic components were filtered off and the filtrate was purified by
column chromatography (TEA; DCM: methanol = 100: 0 ꢀ 98: 2). The product was obtained as
1
a white solid (1.4 g, 61%). Mp: 116 °C. H NMR (DMSOꢀd , 300 MHz):
δ 7.87 (d, J = 8.2 Hz,
6
2H, phꢀ2,6H), 7.48 (d, J = 7.6 Hz, 4H, ph’ꢀ2,6H), 7.41 (d, J = 8.2 Hz, 2H, phꢀ3,5H), 7.23 (t, J =
7
.7 Hz, 4H, ph’ꢀ3,5H), 7.09 (t, J = 7.3, 2H, ph’ꢀ4H), 5.25 (s, 1H, OH), 3.81 (s, 3H, OCH ), 3.47
3
eq
(
s, 2H, NCH ), 2.74 (d, J = 10.8 Hz, 2H, pipꢀ2,6H ), 2.50ꢀ2.43 (m, 1H, pipꢀ4H), 1.94 (t, J = 11.0
2
ax
ax
eq 13
Hz, 2H, pipꢀ2,6H ), 1.54ꢀ1.43 (m, 2H, pipꢀ3,5H ), 1.23 (d, J = 12.1 Hz, 2H, pipꢀ3,5H ). C
NMR (DMSOꢀd , 75 MHz):
δ
166.1 (C=O), 147.2 (ph’ꢀ1C), 144.6 (phꢀ4C), 129.0 (phꢀ2,6C),
6
ACS Paragon Plus Environment

Page 21 of 104
Journal of Medicinal Chemistry
1
2
3
1
4
28.7 (phꢀ3,5C), 128.1 (phꢀ1C), 127.7 (ph’ꢀ3,5C), 125.7 (ph’ꢀ4C), 125.7 (ph’ꢀ2,6C), 78.4
5
(COH), 61.8 (NCH ), 53.5 (pipꢀ2,6C), 52.0 (OCH ), 43.3 (pipꢀ4C), 25.8 (pipꢀ3,5C). HRMS
2 3
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/z): [M+H] calcd. for C H NO , 415.2142; found, 415.2131. Elemental analysis: calcd. for
27
29
3
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
C H NO , C 78.04, H 7.03, N 3.37; found, C 77.79, H 7.04, N 3.32.
27
29
3
4-((4-(Hydroxydiphenylmethyl)piperidin-1-yl)methyl)benzoic acid (4). Compound 3 (1.0 g,
2
.4 mmol) was dissolved in methanol (2 ml) and THF (2 ml). Aqueous potassium hydroxide (2
M, 4 ml) was added and the mixture was heated in the microwave oven for 15 min at 70 °C.
After cooling the organic solvents were removed under reduced pressure and the aqueous
solution neutralized with HCl. The product precipitated as white solid that was recrystallized
1
from 2ꢀpropanol under addition of a drop of water (0.71 g, 74%). Mp: > 250 °C. H NMR
(
DMSOꢀd , 500 MHz):
δ
7.91 (d, J = 8.1 Hz, 2H, phꢀ2,6H), 7.53ꢀ7.48 (m, 6H, phꢀ3,5H, ph’ꢀ
6
2
,6H), 7.24 (t, J = 7.7 Hz, 4H, ph’ꢀ3,5H), 7.11 (t, J = 7.3 Hz, 2H, ph’ꢀ4H), 5.54 (s, 1H, OH), 3.77
eq
ax
(br s, 2H, NCH ), 2.95 (br s, 2H, pipꢀ2,6H ), 2.58 (br s, 1H, pipꢀ4H), 2.30 (br s, 2H, pipꢀ2,6H ),
2
ax
ax 13
1
.63ꢀ1.59 (m, 2H, pipꢀ3,5H ), 1.28 (d, J = 12.6 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀd , 125
6
MHz):
δ
167.2 (C=O), 147.0 (ph’ꢀ1C), 139.8 (phꢀ4C), 130.3 (phꢀ1C), 129.8 (phꢀ3,5C), 129.3
(
(
phꢀ2,6C), 127.8 (ph’ꢀ3,5C), 125.9 (ph’ꢀ4C), 125.7 (ph’ꢀ2,6C), 78.4 (COH), 60.5 (NCH ), 52.7
2
+
pipꢀ2,6C), 43.4 (pipꢀ4C), 25.0 (pipꢀ3,5C). HRMS (m/z): [M+H] calcd. for C H NO ,
2
6
27
3
.
402.2069; found, 402.2085. Elemental analysis: calcd. for C H NO H O, C 74.44, H 6.97, N
2
6
27
3
2
3
.34; found, C 74.28, H 6.64, N 3.34.
Methyl 4-(2-bromoacetyl)benzoate (5). Methyl 4ꢀacetylbenzoate (2 g, 0.011 mol) and pꢀ
toluenesulfonic acid (0.2 g, 1.2 mmol) were suspended in anhydrous acetonitrile (10 ml). Nꢀ
Bromosuccinimide (2.2 g, 11.2 mmol) was added under inert conditions. The mixture was stirred
for 15 min at room temperature and subsequently heated to 60 °C for 2 h. After cooling, the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 104
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
mixture was stirred in a mixture of water (50 ml) and toluene (50 ml) for 15 min. The organic
layer was washed with brine, dried (MgSO ) and concentrated to dryness resulting in a light
4
yellow solid, which was again washed with toluene to remove remaining bromine (67–85%).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
Mp: 78–80 °C. H NMR (DMSOꢀd , 300 MHz):
δ
8.02 (s, 4H, phꢀ2,3,5,6H), 3.82 (s, 3H, OCH3),
195.8 (C=O ketone), 165.6 (C=O ester),
37.2 (phꢀ4C), 133.1 (phꢀ1C), 129.6 (phꢀ2,6C), 129.0 (phꢀ3,5C), 70.9 (phꢀCH ), 57.7 (OCH ).
6
1
3
2
.60 (s, 2H, CH ). C NMR (DMSOꢀd , 75 MHz): δ
2 6
1
2
3
+
HRMS (m/z): [M+H] calcd. for C H BrO , 256.9813; found, 256.9818.
1
0
9
3
Methyl 4-(2-(4-(hydroxydiphenylmethyl)piperidin-1-yl)acetyl)benzoate (6). Diphenylꢀ
piperidinꢀ4ꢀylꢀmethanol hydrochloride (1.3 g, 4.3 mmol), compound 5 (1.0 g, 3.9 mmol) and
potassium carbonate (1.2 g, 8.7 mmol) were stirred in acetone (30 ml) at room temperature. After
12 h the inorganic salts were filtered off and the filtrate concentrated under reduced pressure.
The product was purified by column chromatography (TEA 0.5%; DCM: methanol = 100: 0 ꢀ
1
9
8: 2). Reꢀcrystallization from acetonitrile gave a yellow solid (0.5 g, 30%). Mp: 140 °C. H
NMR (DMSOꢀd , 500 MHz):
δ 8.08ꢀ8.03 (m, 4H, phꢀ2,3,5,6H), 7.48 (d, J = 8.1 Hz, 4H, phꢀ
6
2,6H), 7.24 (t, J = 7.8 Hz, 4H, ph’ꢀ3,5H), 7.10 (t, J = 7.3 Hz, 2H, ph’ꢀ4H), 5.23 (s, 1H, OH), 3.87
eq
(
s, 3H, OCH ), 3.77 (s, 2H, NCH ), 2.84 (d, J = 10.7 Hz, 2H, pipꢀ2,6H ), 2.49 (overlay with
3
2
ax
ax
DMSO peak, pipꢀ4H), 2.10 (t, J = 11.1 Hz, 2H, pipꢀ2,6H ), 1.50ꢀ1.43 (m, 2H, pipꢀ3,5H ), 1.21
eq 13
(
d, J = 12.3 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀd , 125 MHz): δ 197.2 (C=O ketone), 165.6
6
(C=O ester), 147.3 (ph’ꢀ1C), 139.4 (phꢀ4C), 133.1 (phꢀ1C), 129.3 (phꢀ2,6C), 128.5 (phꢀ3,5C),
127.8 (ph’ꢀ3,5C), 125.7 (ph’ꢀ2,4,6C), 78.5 (COH), 64.6 (NCH ), 53.7 (pipꢀ2,6C), 52.5 (OCH ),
2
3
+
4
3.0 (pipꢀ4C), 25.9 (pipꢀ3,5C). HRMS (m/z): [M+H] calcd. for C H NO , 444.2175; found,
28 29 4
444.2172.
ACS Paragon Plus Environment

Page 23 of 104
Journal of Medicinal Chemistry
1
2
3
4
5
4
-(2-(4-(Hydroxydiphenylmethyl)piperidin-1-yl)acetyl)benzoic acid (7). Compound 6 (0.3
g, 0.7 mmol) was dissolved in methanol (1 ml) and THF (1 ml). Aqueous potassium hydroxide
2 M, 2 ml) was added and the mixture was heated in the microwave oven for 15 min at 70 °C.
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
After cooling the organic solvents were removed under reduced pressure and the aqueous
solution neutralized with HCl. The product precipitated as an offꢀwhite solid that was
1
recrystallized from acetonitrile under addition of a drop of water (0.1 g, 35%). Mp: 187 °C. H
NMR (DMSOꢀd , 500 MHz):
δ
8.04ꢀ8.00 (m, 4H, phꢀ2,3,5,6H), 7.48 (d, J = 8.0 Hz, 4H, ph’ꢀ
6
2
,6H), 7.24 (t, J = 7.7 Hz, 4H, ph’ꢀ3,5H), 7.10 (t, J = 7.2 Hz, 2H, ph’ꢀ4H), 5.24 (s, 1H, OH), 3.80
eq
(
s, 2H, NCH ), 2.86 (d, J = 9.2 Hz, 2H, pipꢀ2,6H ), 2.49 (overlay with DMSO peak, pipꢀ4H),
2
ax
ax
eq
2.14ꢀ2.10 (m, 2H, pipꢀ2,6H ), 1.52ꢀ1.47 (m, 2H, pipꢀ3,5H ), 1.22 (d, J = 11.6, 2H, pipꢀ3,5H ).
13
C NMR (DMSOꢀd , 125 MHz):
δ 197.0 (C=O ketone), 166.7 (C=O ester), 147.2 (ph’ꢀ1C),
6
1
4
3
38.8 (phꢀ4C), 135.1 (phꢀ1C), 129.4 (phꢀ3,5C), 128.3 (phꢀ2,6C), 127.8 (ph’ꢀ3,5C), 125.8 (ph’ꢀ
C), 125.7 (ph’ꢀ2,6C), 78.5 (COH), 64.4 (NCH ), 53.6 (pipꢀ2,6C), 43.9 (pipꢀ4C), 25.8 (pipꢀ
2
+
,5C). HRMS (m/z): [MꢀH] calcd. for C H NO , 428.1862; found, 428.1843. Elemental
2
7
27
4
.
analysis: calcd. for C H NO 0.5 H O, C 73.95, H 6.44, N 3.19; found, C 74.10, H 6.23, N
2
7
27
4
2
3
.29.
Methyl 2-(4-(bromomethyl)phenyl)-2-methylpropanoate (8). Under inert atmosphere, pꢀ
tolylacetic acid (5 g, 0.033 mol) was dissolved in methanol (60 ml) and under cooling (ice bath)
chlorotrimethyl silane (6.5 ml) was added dropwise. The solution was stirred overnight at room
temperature and subsequently concentrated to dryness. The resulting colorless oil was dissolved
in THF (25 ml) and added to a solution of sodium hydride (4 g, 60% dispersion in mineral oil,
0
.1 mol) in THF (100 ml). Methyl iodide (5 ml, 0.08 mol) was added dropwise and under
cooling. The resulting suspension was stirred overnight at room temperature. After concentration
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 104
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
to dryness the crude was taken up in water (50 ml) and extracted with ethyl acetate (3 x 50 ml).
The organic layer was dried (MgSO ) and concentrated to give methyl 2ꢀmethylꢀ2ꢀ(pꢀ
4
1
tolyl)propanoate (8a) as a yellow oil (4.6 g; 73%). H NMR (CDCl , 300 MHz):
δ 7.32 (d, J =
3
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
8
.3 Hz, 2H, phꢀ2,6H), 7.17 (d, J = 8.1 Hz, 2H, phꢀ3,5H), 3.66 (s, 3H, OCH ), 2.34 (s, 3H, phꢀ
3
1
3
CH ), 1.58 (s, 6H, C(CH ) ). C NMR (CDCl , 75 MHz):
δ
177.4 (C=O), 141.8 (phꢀ4C), 136.3
3
3 2
3
(
phꢀ1C), 129.1 (phꢀ3,5C), 125.5 (phꢀ2,6C), 52.2 (OCH ), 46.2 (C(CH ) ), 26.6 (C(CH ) ), 21.0
3 3 2 3 2
+
(phꢀCH ). HRMS (m/z): [M+H] calcd. for C H O , 193.1229; found, 193.1236. A solution of
3
12 16
2
2,2′ꢀazobis(2ꢀmethylpropionitrile) in toluene (5 ml, 0.2 M) was concentrated to dryness without
heating and under exclusion of light. 2,2′ꢀAzobis(2ꢀmethylpropionitrile), Nꢀbromosuccinimide
(1.0 g, 5 mmol) and compound 8a (1.0 g, 5 mmol) were dissolved in bromobenzene or
chlorobenzene (5 ml). The solution was heated in an oil bath kept at 110 °C. When the reaction
mixture reached 80ꢀ90 °C, an exothermic reaction started with the development of bromine
(orange color). Subsequently, the mixture turned dark brown and was quenched by cooling with
an ice bath and adding ice water. The product was extracted with petrol. The organic layer was
washed with brine, dried (MgSO ) and concentrated to dryness. The product was purified by
4
column chromatography (petrol: ethyl acetate = 100: 0 ꢀ 98: 2) giving a yellow oil (50 – 61%).
1
H NMR (DMSOꢀd , 300 MHz): δ 7.38 (d, J = 8.3 Hz, 2H, phꢀ2,6H), 7.29 (d, J = 8.4 Hz, 2H,
6
1
3
phꢀ3,5H), 4.67 (s, 2H, CH Br), 3.57 (s, 3H, OCH ), 1.48 (s, 6H, C(CH ) ). C NMR (DMSOꢀd ,
2
3
3 2
6
125 MHz):
δ
176.4 (C=O), 142.9 (phꢀ4C), 140.9 (phꢀ1C), 126.6 (phꢀ2,6C), 125.1 (phꢀ3,5C), 62.5
+
(
CH Br), 52.0 (OCH ), 46.0 (C(CH ) ), 26.4 (C(CH ) ). HRMS (m/z): [MꢀH] calcd. for
2
3
3 2
3 2
C H BrO , 269.0172; found, 269.0163.
12
15
2
Methyl
2-(4-((4-(hydroxydiphenylmethyl)piperidin-1-yl)methyl)phenyl)-2-methyl-
propanoate (9). Diphenyl(piperidinꢀ4ꢀyl)methanol hydrochloride (1.0 g, 3.3 mmol), compound
ACS Paragon Plus Environment

Page 25 of 104
Journal of Medicinal Chemistry
1
2
3
8
4
(0.7 g, 2.6 mmol), potassium carbonate (1.46 g, 10.6 mmol) and potassium iodide (0.05 g, 0.33
5
mmol) were suspended in acetonitrile (25 ml). Under inert atmosphere the mixture was refluxed
overnight. After cooling the inorganic components were removed by filtration. The filtrate was
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
concentrated to dryness and the crude purified by column chromatography (TEA 0.5%; DCM:
1
methanol = 100: 0 ꢀ 98: 2). The product was isolated as colorless oil. H NMR (DMSOꢀd , 500
6
MHz):
δ 7.48 (d, J = 7.6 Hz, 4H, ph’ꢀ2,6H), 7.29ꢀ7.18 (m, 8H, phꢀ2,3,5,6H, ph’ꢀ3,5H), 7.10 (t, J
=
7.1 Hz, 2H, ph’ꢀ4H), 5.22 (s, 1H, OH), 3.56 (s, 3H, OCH ), 3.37 (s, 2H, NCH ), 2.75 (d, J =
3
2
eq
ax
1
0.8 Hz, 2H, pipꢀ2,6H ), 2.49ꢀ2.44 (m, 1H, pipꢀ4H), 1.91 (t, J = 11.3 Hz, 2H, pipꢀ2,6H ), 1.50ꢀ
ax
eq 13
1
.43 (m, 8H, pipꢀ3,5H , C(CH ) ), 1.23 (d, J = 12.2 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀd ,
3
2
6
125 MHz):
δ 167.4 (C=O), 147.3 (ph’ꢀ1C), 143.0 (phꢀ1C), 137.0 (phꢀ4C), 128.7 (phꢀ3,5C),
1
27.7 (ph’ꢀ3,5C), 125.8 (ph’ꢀ2,6C, phꢀ2,6C), 125.2 (ph’ꢀ4C), 78.5 (COH), 61.9 (NCH ), 53.5
2
(pipꢀ2,6C), 52.0 (C(CH ) ), 45.8 (OCH ), 43.4 (pipꢀ4C), 26.4 (C(CH ) ), 25.5 (pipꢀ3,5C). HRMS
3 2 3 3 2
+
(
m/z): [M+H] calcd. for C H NO , 458.2695; found, 458.2696.
30
35
3
2-(4-((4-(Hydroxydiphenylmethyl)piperidin-1-yl)methyl)phenyl)-2-methylpropanoic acid
(10). Compound 9 (0.5 g, 1.0 mmol) was dissolved in methanol (1 ml) and THF (1 ml). Aqueous
potassium hydroxide (2 M, 2 ml) was added and the mixture was heated in the microwave oven
for 15 min at 70 °C. After cooling the organic solvents were evaporated and the aqueous solution
neutralized with HCl. The product precipitated as white solid that was recrystallized from
1
acetonitrile under addition of a drop of water (0.23 g, 48%). Mp: 246 °C. H NMR (DMSOꢀd ,
6
500 MHz):
δ 7.48 (d, J = 7.7 Hz, 4H, ph’ꢀ2,6H), 7.30ꢀ7.19 (m, 8H, phꢀ2,3,5,6H, ph’ꢀ3,5H), 7.09
(
t, J = 7.3 Hz, 2H, ph’ꢀ4H), 5.22 (s, 1H, OH), 3.39 (s, 2H, NCH ), 2.77 (d, J = 10.6 Hz, 2H, pipꢀ
2
eq
ax
2
,6H ), 2.49ꢀ2.44 (m, 1H, pipꢀ4H), 1.94 (t, J = 11.2 Hz, 2H, pipꢀ2,6H ), 1.50ꢀ1.43 (m, 8H, pipꢀ
ax
eq 13
3,5H , C(CH ) ), 1.20 (d, J = 12.4 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀd , 125 MHz):
δ
3
2
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 104
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
77.6 (C=O), 147.3 (ph’ꢀ1C), 143.6 (phꢀ4C), 136.2 (phꢀ1C), 128.6 (phꢀ2,6C), 127.7 (ph’ꢀ3,5C),
1
25.8 (ph’ꢀ2,6C, phꢀ3,5C), 125.3 (ph’ꢀ4C), 78.5 (COH), 61.9 (NCH ), 53.5 (pipꢀ2,6C), 45.5
2
+
(C(CH ) ), 43.3 (pipꢀ4C), 26.4 (C(CH ) ), 25.9 (pipꢀ3,5C). HRMS (m/z): [M+H] calcd. for
3
2
3 2
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
.
C H NO , 444.2539; found, 444.2520. Elemental analysis: calcd. for C H NO 0.5 H O, C
29
33
3
29 33
3
2
76.96, H 7.57, N 3.09; found, C 77.19, H 7.38, N 3.09.
Methyl 4-(bromomethyl)-3-fluorobenzoate (11). 3ꢀFluoroꢀ4ꢀmethylbenzoic acid (2.0 g, 13.0
mmol) was dissolved in methanol (20 ml). Chlorotrimethyl silane (2.6 ml) was added dropwise
and under cooling and the mixture was stirred overnight at room temperature. The resulting
crude was concentrated on silica and the product was purified by column chromatography
1
(DCM). Methyl 3ꢀfluoroꢀ4ꢀmethylbenzoate (11a) was isolated as a colorless oil (1.4 g, 64%). H
NMR (CDCl , 500 MHz):
δ 7.70 (d, J = 7.9 Hz, 1H, phꢀ6H), 7.64 (d, J = 10.2 Hz, 1H, phꢀ2H),
3
1
3
7.23 (t, J = 7.7 Hz, 1H, phꢀ5H), 3.89 (s, 3H, OCH ), 2.29 (s, 3H, phꢀCH ). C NMR (CDCl ,
3 3 3
1
3
125 MHz):
δ 166.2 (C=O), 162.2 (d, J = 118.8 Hz, phꢀ3C), 131.4 (d, J = 5.0, phꢀ5C),
C,F C,F
2
3
4
130.6 (d, J = 17.3 Hz, phꢀ4C), 129.7 (d, J = 7.5 Hz, phꢀ1C), 125.1 (d, J = 3.6 Hz, phꢀ
C,F
C,F
C,F
2
3
19
6C), 116.1 (d, J = 23.9 Hz, phꢀ2C), 52.2 (OCH ), 14.9 (d, J = 3.5 Hz, phꢀCH ). F NMR
C,F
3
C,F
3
+
(CDCl , 282 MHz):
δ
–117.16 (phꢀF). HRMS (m/z): [M+H] calcd. for C H FO , 169.0665;
3
9
9
2
found, 169.0662. A solution of 2,2’ꢀazobis(2ꢀmethylpropionitrile) in toluene (10 ml, 0.2 M) was
concentrated to dryness without heating and under exclusion of light. 2,2’ꢀAzobis(2ꢀ
methylpropionitrile), Nꢀbromosuccinimide (1.9 g, 10.0 mmol) and compound 11a (1.4 g, 8.3
mmol) were dissolved in chlorobenzene and treated as described for compound 8, giving the
1
desired product as yellow oil (1.5 g, quantitative conversion). H NMR (CDCl , 500 MHz):
δ
3
7.80 (d, J = 8.0 Hz, 1H, phꢀ6H), 7.71 (d, J = 10.1 Hz, 1H, phꢀ2H), 7.46 (t, J = 7.7 Hz, 1H, phꢀ
1
3
5
H), 4.51 (s, 2H, CH Br), 3.92 (s, 3H, CH ). C NMR (CDCl , 125 MHz):
δ
165.6 (C=O), 160.3
2
3
3
ACS Paragon Plus Environment

Page 27 of 104
Journal of Medicinal Chemistry
1
2
3
4
1
3
2
(
d, J = 249.3, phꢀ3C), 132.5 (d, J = 7.9, phꢀ5C), 131.4 (d, J = 22.0, phꢀ4C), 130.1 (phꢀ
C,F
C,F
C,F
5
2
1
C), 125.7 (phꢀ6C), 117.0 (d, J = 23.1, phꢀ2C), 52.6 (OCH ), 24.6 (CH Br).
C,F 3 2
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
Methyl 3-fluoro-4-((4-(hydroxydiphenylmethyl)piperidin-1-yl)methyl)benzoate hydro-
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
chloride (12). Under inert atmosphere, diphenyl(piperidinꢀ4ꢀyl)methanol hydrochloride (2.0 g,
6.6 mmol), compound 11 (1.5 g, 6.0 mmol) and potassium carbonate (2.7 g, 19.6 mmol) were
suspended in acetonitrile (50 ml) and refluxed overnight. After cooling, the inorganic
components were filtered off and the filtrate concentrated under reduced pressure. The product
was purified by column chromatography (TEA 0.5%; DCM: methanol = 100: 0 ꢀ 98: 2). The
oily product was crystallized with HCl in ethanol to give a white solid (1.7 g, 61%). Mp: 151–
1
+
1
7
3
54 °C. H NMR (DMSOꢀd , 500 MHz): δ 10.72 (br s, 1H, NH ), 7.91ꢀ7.84 (m, 2H, phꢀ5,6H),
6
.78 (d, J = 10.0, 1H, phꢀ2H), 7.45 (d, J = 7.5 Hz, 4H, ph’ꢀ2,6H), 7.26 (t, J = 7.7 Hz, 4H, ph’ꢀ
,5H), 7.13 (t, J = 7.3 Hz, 2H, ph’ꢀ4H), 5.54 (s, 1H, OH), 4.32 (s, 2H, NCH ), 3.88 (s, 3H,
2
eq
ax
OCH ), 3.36 (d, J = 11.1 Hz, 2H, pipꢀ2,6H ), 3.02ꢀ2.95 (m, 2H, pipꢀ2,6H ), 2.81 (t, J = 11.7
3
ax
eq 13
Hz, 1H, pipꢀ4H), 1.83ꢀ1.76 (m, 2H, pipꢀ3,5H ), 1.42 (d, J = 13.7 Hz, 2H, pipꢀ3,5H ). C NMR
1
(
DMSOꢀd , 125 MHz):
δ
164.8 (C=O), 161.0 (d, J = 219.8 Hz, phꢀ3C), 164.5 (ph’ꢀ1C), 134.9
6
C,F
(
phꢀ5C), 133.1 (phꢀ4C), 127.9 (ph‘ꢀ3,5C), 126.1 (ph’ꢀ4C), 125.7 (ph’ꢀ2,6C), 125.2 (phꢀ6C),
2
1
21.6 (phꢀ1C), 116.2 (d, J = 23.8 Hz, phꢀ2C), 78.1 (COH), 56.0 (NCH ), 52.7 (OCH ), 51.4
C,F
2
3
1
9
(pipꢀ2,6C), 40.4 (pipꢀ4C), 23.8 (pipꢀ3,5C). F NMR (DMSOꢀd , 282 MHz):
δ
–113.23 (phꢀF).
6
+
HRMS (m/z): [M+H] calcd. for C H FNO , 434.2131; found, 434.2131.
2
7
28
3
3-Fluoro-4-((4-(hydroxydiphenylmethyl)piperidin-1-yl)methyl)benzoic
acid
(13).
Compound 12 (0.5 g, 1.0 mmol) was dissolved in methanol (1 ml) and THF (1 ml) and treated
with aqueous potassium hydroxide (2 M, 2 ml). The mixture was heated in the microwave oven
for 15 min at 70 °C. After cooling the organic solvents were removed under reduced pressure
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 104
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 the aqueous solution neutralized with HCl. The product precipitated as white solid that was
1
recrystallized from acetonitrile under addition of a drop of water (0.15 g, 33%). Mp: 247 °C. H
NMR (DMSOꢀd , 500 MHz):
δ
7.72 (d, J = 7.9 Hz, 1H, phꢀ6H), 7.58 (d, J = 10.4 Hz, 1H, phꢀ
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
H), 7.50ꢀ7.48 (m, 5H, phꢀ5H, ph’ꢀ2,6H), 7.23 (t, J = 7.8 Hz, 4H, ph’ꢀ3,5H), 7.10 (t, J = 7.3 Hz,
eq
2H, ph’ꢀ4H), 5.24 (s, 1H, OH), 3.55 (s, 2H, NCH ), 2.79 (d, J = 11.0 Hz, 2H, pipꢀ2,6H ), 2.50ꢀ
2
ax
ax
2
.45 (m, 1H, pipꢀ4H), 2.02 (t, J = 11.1 Hz, 2H, pipꢀ2,6H ), 1.52ꢀ1.44 (m, 2H, pipꢀ3,5H ), 1.23
eq 13
4
(d, J = 12.6 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀd , 150 MHz):
δ
166.3 (d, J = 2.3 Hz,
6
C,F
1
3
C=O), 160.4 (d, J = 243.9 Hz, phꢀ3C), 147.3 (ph’ꢀ1C), 132.2 (d, J = 7.2 Hz, phꢀ5C), 131.6
C,F
C,F
3
2
(
d, J = 4.4 Hz, phꢀ1C), 129.8 (d, JC,F = 13.1 Hz, phꢀ4C), 127.8 (ph‘ꢀ3,5C), 125.8 (ph’ꢀ4C),
C,F
4
2
125.7 (ph’ꢀ2,6C), 125.1 (d, J = 2.6 Hz, phꢀ6C), 115.7 (d, J = 23.6 Hz, phꢀ2C), 78.5
C,F
C,F
19
(
COH), 54.5 (NCH ), 53.3 (pipꢀ2,6C), 43.1 (pipꢀ4C), 25.9 (pipꢀ3,5C). F NMR (DMSOꢀd , 282
2
6
+
MHz):
δ –117.833 (phꢀF). HRMS (m/z): [M+H] calcd. for C H FNO , 420.1975; found,
26 26 3
.
4
20.1961. Elemental analysis: calcd. for C H FNO 0.5 H O, C 79.16, H 7.15, N 3.55; found,
26 26 3 2
C 78.95, H 6.90, N 3.48.
Methyl 4-((4-((4-fluorophenyl)(hydroxy)(phenyl)methyl)piperidin-1-yl)methyl)benzoate
14). Under inert atmosphere, compound 34 (0.4 g, 1.24 mmol), methyl 4ꢀ
bromomethyl)benzoate (0.43 g, 1.86 mmol) and potassium carbonate (0.43 g, 3.1 mmol) were
(
(
suspended in acetone (15 ml) and stirred for 2 h at room temperature. The inorganic components
were subsequently filtered off and the filtrate was concentrated under reduced pressure.
Purification by column chromatography (TEA 0.5%; DCM: methanol = 100: 0 ꢀ 99: 1) gave the
1
product as a colorless oil (0.19 g, 35%). H NMR (DMSOꢀd , 600 MHz):
δ 7.87 (d, J = 8.2 Hz,
6
2H, phꢀ2,6H), 7.53ꢀ7.48 (m, 4H, ph’ꢀ2,6H, ph’’ꢀ2,6H), 7.42 (d, J = 8.2 Hz, 2H, phꢀ3,5H), 7.25 (t,
J = 7.7 Hz, 2H, ph’ꢀ3,5H), 7.12 (t, J = 7.3 Hz, 1H, ph’ꢀ4H), 7.06 (t, J = 8.9 Hz, 2H, ph’’ꢀ3,5H),
ACS Paragon Plus Environment

Page 29 of 104
Journal of Medicinal Chemistry
1
2
3
5
4
eq
.33 (s, 1H, COH), 3.83 (s, 3H, OCH ), 3.49 (s, 2H, NCH ), 2.77 (d, J = 8.1 Hz, 2H, pipꢀ2,6H ),
3
2
5
ax
2
3
1
.48 (t, J = 11.8 Hz, 1H, pipꢀ4H), 1.95 (t, J = 11.8 Hz, 2H, pipꢀ2,6H ), 1.52ꢀ1.44 (m, 2H, pipꢀ
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
ax
eq 13
,5H ), 1.21 (t, J = 11.5 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀd , 150 MHz):
δ
166.2 (C=O),
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
1
4
60. 5 (d, J = 240.6 Hz, ph’’ꢀ4C), 147.1 (ph’ꢀ1C), 144.7 (phꢀ4C), 143.5 (d, J = 2.7 Hz,
C,F
C,F
3
ph’’ꢀ1C), 129.1 (phꢀ2,6C), 128.8 (ph’ꢀ3,5C), 128.2 (phꢀ1C), 127.9 (phꢀ3,5C), 127.7 (d, JC,F
=
2
7
.7 Hz, ph’’ꢀ2,6C), 125.9 (ph’ꢀ4C), 125.7 (ph’ꢀ2,6C), 114.4 (d, J = 20.7 Hz, ph’’ꢀ3,5C), 78.3
C,F
1
9
(
COH), 61.9 (NCH ), 53.5 (pipꢀ2,6C), 52.1 (OCH ), 43.3 (pipꢀ4C), 26.0 (pipꢀ3,5C). F NMR
2
3
+
(DMSOꢀd , 282 MHz):
δ –118.08 (phꢀF). HRMS (m/z): [M+H] calcd. for C H FNO ,
27 28 3
6
434.2131; found, 434.2122.
4-((4-((4-Fluorophenyl)(hydroxy)(phenyl)methyl)piperidin-1-yl)methyl)benzoic acid (15).
Compound 14 (0.18 g, 0.4 mmol) was dissolved in methanol (0.5 ml) and THF (0.5 ml).
Aqueous potassium hydroxide (2 M, 1 ml) was added and the mixture was stirred overnight at
room temperature. The organic solvents were evaporated and the aqueous solution neutralized
with hydrogen chloride solution. The product precipitated as a white solid that was recrystallized
1
from acetonitrile under addition of a drop of water (0.06 g, 35%). Mp:
DMSOꢀd , 600 MHz):
>
250 °C. H NMR
(
δ
7.86 (d, J = 8.1 Hz, 2H, phꢀ2,6H), 7.53ꢀ7.48 (m, 4H, ph’ꢀ2,6H, ph’’ꢀ
6
2
1
,6H), 7.39 (d, J = 8.1 Hz, 2H, phꢀ3,5H), 7.25 (t, J = 7.7 Hz, 2H, ph’ꢀ3,5H), 7.12 (t, J = 7.3 Hz,
H, ph’ꢀ4H), 7.06 (t, J = 8.9 Hz, 2H, ph’’ꢀ3,5H), 5.34 (s, 1H, OH), 3.49 (s, 2H, NCH ), 2.78 (d, J
2
eq
=
8.0 Hz, 2H, pipꢀ2,6H ), 2.48 (t, J = 11.9 Hz, 1H, pipꢀ4H), 1.96 (t, J = 11.7 Hz, 2H, pipꢀ
ax
ax
eq 13
2
,6H ), 1.51ꢀ1.45 (m, 2H, pipꢀ3,5H ), 1.21 (d, J = 11.1 Hz, 2H, pipꢀ3,5H ). C NMR (DMSOꢀ
1
4
d6, 150 MHz):
δ 167.3 (C=O), 160.5 (d, J = 240.8 Hz, ph’’ꢀ4C), 143.9 (phꢀ4C), 143.5 (d, J
C,F C,F
=
2.6 Hz, ph’’ꢀ1C), 129.5 (phꢀ1C), 129.2 (phꢀ2,6C), 128.7 (ph’ꢀ3,5C), 127.9 (phꢀ3,5C), 127.7 (d,
3
2
JC,F = 7.8 Hz, ph’’ꢀ2,6C), 125.9 (ph’ꢀ4C), 125.7 (ph’ꢀ2,6C), 114.4 (d, J = 20.7 Hz, ph’’ꢀ
C,F
ACS Paragon Plus Environment