Synthesis and Structure-Affinity Relationship of Small Molecules for Imaging Human CD80 by Positron Emission Tomography

Article abstract of DOI:10.1021/acs.jmedchem.9b00858

The costimulatory molecule CD80 is an early marker for immune activation. It is upregulated on activated antigen-presenting cells. We aimed at developing a tracer for imaging CD80 by positron emission tomography (PET). Novel CD80 ligands were synthesized

Full text of DOI:10.1021/acs.jmedchem.9b00858

Subscriber access provided by Stockholm University Library
Article
SYNTHESIS AND STRUCTURE AFFINITY RELATIONSHIP
OF SMALL MOLECULES FOR IMAGING HUMAN
CD80 BY POSITRON EMISSION TOMOGRAPHY
Marco Francesco Taddio, Linjing Mu, Claudia Adrianna Castro Jaramillo,
Tanja Bollmann, Dominik Schmid, Lukas Muskalla, Tim Gruene, Aristeidis
Chiotellis, Simon Mensah Ametamey, Roger Schibli, and Stefanie D. Krämer
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00858 • Publication Date (Web): 20 Aug 2019
Downloaded from pubs.acs.org on August 21, 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 48
Journal of Medicinal Chemistry
1
2
3
4
5
SYNTHESIS AND STRUCTURE AFFINITY
6
7
8
9
RELATIONSHIP OF SMALL MOLECULES FOR IMAGING
HUMAN CD80 BY POSITRON EMISSION TOMOGRAPHY
Marco F. Taddio1*, Linjing Mu1,2, Claudia A. Castro Jaramillo1, Tanja Bollmann1, Dominik
Schmid1, Lukas Muskalla3, Tim Grüne3,4, Aristeidis Chiotellis1, Simon M. Ametamey1, Roger
Schibli1, Stefanie D. Krämer1*
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 Center for Radiopharmaceutical Sciences ETH, PSI and USZ, Department of Chemistry and
2
Applied Biosciences, ETH Zurich, CH-8093 Zurich, Switzerland; Department of Nuclear
3
Medicine, University Hospital Zurich, CH-8091 Zurich, Switzerland; Laboratory for Catalysis
4
and Sustainable Chemistry, Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland; X-ray
Structure Analysis Centre, Faculty of Chemistry, University of Vienna, A-1090 Vienna, Austria
Keywords: CD80 / imaging / inflammation / PET / small molecules
Abstract
The co-stimulatory molecule CD80 is an early marker for immune activation. It is upregulated on
activated antigen-presenting cells. We aimed at developing a tracer for imaging CD80 by positron
emission tomography (PET). Novel CD80 ligands were synthesized and tested by SPR for affinity
to human CD80 (hCD80) and displacement of endogenous ligands. Several compounds bound
with one-digit nanomolar affinity to hCD80 and displaced CTLA-4 and CD28 at nanomolar
concentrations. A structure-affinity relationship study revealed relevant moieties for strong affinity
to hCD80 and positions for further modifications. Lead compound MT107 (7f) was radiolabelled
with carbon-11. In vitro, [11C]MT107 showed specific binding to hCD80-positive tissue and high
plasma protein binding. In vivo, [11C]MT107 accumulated in liver, gall bladder and intestines, but
only scarcely in hCD80-positive xenografts. The unfavourable in vivo performance may result
from high plasma protein binding and extensive biliary excretion.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 48
1
2
3
4
ABSTRACT
5
6
7
8
9
The co-stimulatory molecule CD80 is an early marker for immune activation. It is upregulated on
activated antigen-presenting cells. We aimed at developing a tracer for imaging CD80 by positron
emission tomography (PET). Novel CD80 ligands were synthesized and tested by SPR for affinity
to human CD80 (hCD80) and displacement of endogenous ligands. Several compounds bound
with one-digit nanomolar affinity to hCD80 and displaced CTLA-4 and CD28 at nanomolar
concentrations. A structure-affinity relationship study revealed relevant moieties for strong affinity
to hCD80 and positions for further modifications. Lead compound MT107 (7f) was radiolabelled
with carbon-11. In vitro, [11C]MT107 showed specific binding to hCD80-positive tissue and high
plasma protein binding. In vivo, [11C]MT107 accumulated in liver, gall bladder and intestines, but
only scarcely in hCD80-positive xenografts. The unfavourable in vivo performance may result
from high plasma protein binding and extensive biliary excretion.
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
2

Page 3 of 48
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
9
The co-stimulatory molecules CD80 (B7-1) and CD86 (B7-2) are expressed on activated antigen-
presenting cells. They interact with immune-modulatory molecules on T cells, among them CTLA-
4 and CD28. The weak-affinity interaction of CD80 or CD86 with CD28 is involved in T-cell
activation, while the strong interaction of CD80/CD86 with CTLA-4 prevents T-cell activation.1,2,3
The balance between T cell activation and inactivation is disturbed in various diseases including
cancer4, atherosclerosis5,6, autoimmune diseases7 and transplant rejection8. Several therapeutic
strategies aim at modulating the levels or functions of the co-stimulatory molecules accordingly.
CD80/CD86 and their receptors are targeted by immunotherapies for various indications.9,10
However, following their dual functionality, finding the right balance in CD80/CD86 modulation
remains challenging. Diagnosis, disease staging and therapy planning may benefit from a non-
invasive method to detect and quantify T-cell activating co-stimulatory molecules in tissue.
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
Several fusion proteins are on the market which bind to CD80 and CD86 with strong affinity
leading to co-stimulation blockade and anergy of T cells. Abatacept and belatacept are successfully
used in rheumatoid arthritis9,10 and for the prophylaxis of organ rejection in transplantation
medicine, respectively.11,12 Both are fusion proteins of the extracellular domain of CTLA-4 and
the IgG Fc part. Belatacept contains two pointmutations in the binding domains which result in a
2-fold stronger affinity to human CD80 and a 4-fold stronger affinity to human CD86 as compared
to abatacept.11
Low molecular weight compounds with strong affinities to CD80 were developed to overcome the
drawback of negligible oral bioavailability of the fusion proteins.13,14,15,16 Erbe et al. discovered a
binding site for small molecules on human CD80 (hCD80) which is adjacent to the binding
domains for CTLA-4 and CD28, respectively.13 The small-molecule binding site is absent on
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 48
1
2
3
4
5
6
7
8
9
rodent CD80 as concluded from both the amino acid sequence and absence of strong-affinity
binding of the low molecular weight ligands of hCD80.13 In the year 2004, Huxley et al. published
structures with strong affinities and high selectivity for hCD80 over hCD86.15
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
Previous work from our group with a carbon-11 labelled radiotracer ([11C]AM7), based on the lead
structure from Huxley et al.15 (1 in Fig. 1), showed promising results in in vitro autoradiography
experiments with hCD80-positive xenograft and human atherosclerotic plaque tissue slices.
However, it had unfavourable pharmacokinetic properties in mice with high plasma protein
binding (human and mouse) and substantial biliary excretion. In vivo accumulation in a hCD80-
positive xenograft was poor.6
In this study, we explored the potential of structurally modified CD80 ligands for the development
of a PET tracer with altered pharmacokinetics. We increased the structural diversity by
hybridisation of the lead structures introduced by Huxley et al.15 and Green et al.14 and
investigated the structure-affinity relationship of the new compounds by surface plasmon
resonance (SPR). A ligand combining core-structure modifications and strong affinity was labelled
with carbon-11 and evaluated by PET/CT in mice with hCD80-positive Raji xenografts.
ACS Paragon Plus Environment
4

Page 5 of 48
Journal of Medicinal Chemistry
1
2
3
4
RESULTS
5
6
7
8
Chemistry. A cinnoline-based compound (1)15 and pyrazolopyridine-based compounds (2, 3)14
were previously reported as potent small molecule ligands for hCD80. Based on these core
structures, we designed and synthesized novel cinnoline-, quinoline- and pyrazolopyridine-based
compounds in this work. Replacement of the methyl-pyrazole moiety of compound 2 with the
benzene ring of compound 1 led to novel quinoline-based compounds 7e-g (Figure 1). The acid
intermediates were included for comparison with the acid compound 3.
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
O
O
Cl
NH
NH
COOH
OH
N
N
HN
N
N
HN
N
N
N
O
O
O
N
N
CF3
CF3
N
N
N
N
H
F
1
2
3
O
O
NH
R
O
NH
NH
R
N
N
HN
N
N
HN
N
HN
N
O
O
O
N
CF3
CF3
N
N
N
X
X
1
1
7e
7d
7a
7b
7c
: X=H; R=R
: X=H; R=R
1
2
2
7f
: X=F; R=R
: X=F; R=R
: X=H; R=R
2
7g
: X=F; R=R
R2 =
R1 =
N
OH
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 48
1
2
3
4
5
6
Figure 1. Structures of the synthesized compounds. Boxes indicate structural features from
reported compounds (1 & 2) which were conserved in hybrid compounds 7e – 7g.
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
Three series of cinnoline-, quinoline- and pyrazolopyridine-based compounds were synthesized
based on the modified synthetic procedures as previously described.14,15 Scheme 1 depicts the
synthetic route for cinnoline-based compounds. Compounds 5a and 5b were synthesized by
reacting their corresponding phenylhydrazonomalonic acid with thionylchloride, cyclization and
methyl ester formation in the presence of TiCl4 and methanol. Converting the hydroxyl group to
the chlorinated intermediate, then reacting with 4-hydrazino-benzoic acid formed the key acid
compounds 6a and 6b in more than 75% yields. Subsequently, coupling with 1-methylpiperidin-
4-amine or 1-(3-aminophenyl)ethan-1-ol afforded compounds 1 and 7a-c in moderate yields.
Scheme 1: Synthetic pathway towards cinnoline-based compounds.
O
COOH
NH
R
O
O
X
R1
R2
=
=
OH
N
HO
HN
OH
HN
N
N
N
HN
N
COOMe
iii
ii
N
i
O
O
N
N
N
N
OH
X
X
X
5a
5b
: X = H
: X = F
1
6a
6b
7a
7b
7c
: X = H
: X = F
: Y= H; R=R
2
2
: Y=F; R=R
: Y=H; R=R
Reagents and conditions: (i) 1) SOCl2, 80°C; 2) TiCl4, 110°C; 3) MeOH, 60°C; (ii) 1) SOCl2,
80°C, 1 h; 2) 4-hydrazinylbenzoic acid, 70°C, EtOH, 1 h; (iii) R-NH2, EDC, DIPEA, rt, DMF, 18
h. R1, R2, see Figure 1.
ACS Paragon Plus Environment
6

Page 7 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The pyrazolopyridine- and quinoline-based compounds were synthesized as illustrated in Scheme
2 and 3, respectively. Compound 4 was synthesized by heating ethyl 3-(trifluoromethyl)benzoate
in EtOAc with sodium hydride and subsequent methylation with trimethylsilyldiazomethane at
room temperature with 48% yield. Gould-Jacobs cyclisation with ethyl 3-methoxy-3-(3-
(trifluoromethyl)phenyl)acrylate (4) under strong basic conditions, followed by substitution of the
hydroxy group by chlorine using phosphorous oxychloride, afforded 1-methyl-1H-pyrazolo[3,4-
b]pyridine (5c) or quinolines (5d, 5e) with 55-61% yields. Subsequent Knorr pyrazole ring
formation with hydrazinyl benzoic acid derivatives under basic conditions and heating afforded
compounds 6c-e in 66-85% and compound 3 in 31% yield. In the last step, the carboxylic acid was
activated with a coupling reagent (EDC hydrochloride or HATU) and two different building blocks
(meta-1-hydroxyethyl-phenyl or para-1-methylpiperidinyl) were introduced by amide bond
formation in 13-87% yields (2, 7d-g).
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 2: Synthetic route for pyrazolopyridine-based compound 7d.
O
COOH
NH
Cl
COOEt
NH2
O
N
COOEt
N
N
HN
N
CF3
HN
N
i
iii
N
+
ii
N
O
CF3
N
O
N
N
COOEt
N
N
CF3
CF3
N
N
5c
6c
7d
4
Reagents and conditions: (i) 1) NaH, reflux, THF, 16 h; 2) POCl3, 120°C, 2 h;
(ii) 4-hydrazinylbenzoic acid, NaOtBu, 100°C, ethylene glycol, 18 h; (iii) 1-methylpiperidin-4-
amine, EDC, DIPEA, rt, DMF, 18 h
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 48
1
2
3
4
Scheme 3: Synthetic pathway toward quinoline-based compounds.
5
6
7
8
9
O
COOH
NH
R
Cl
N
O
COOEt
NH2
N
COOEt
N
HN
N
HN
N
CF3
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
i
iii
ii
+
O
O
CF3
COOEt
CF3
CF3
X
X
4
X
X
5d: X = H
5e: X = F
6d: X = H
6e: X = F
7e: X=H; R=R1
7f: X=F; R=R1
7g: X=F; R=R2
Reagents and conditions: (i) 1) NaH, reflux, THF, 16 h; 2) POCl3, 120°C, 2 h;
(ii) 4-hydrazinylbenzoic acid, NaOtBu, 100°C, ethylene glycol, 18 h; (iii) R-NH2, HATU, DIPEA,
rt, DMF, 30 min. R1, R2, see Figure 1.
ACS Paragon Plus Environment
8

Page 9 of 48
Journal of Medicinal Chemistry
1
2
3
4
Table 1: Synthesized compounds (structures and substituents according to Figure 1 and Scheme 1)
5
6
and their affinity to hCD80 determined by surface plasmon resonance.
7
8
9
Compound
X
F
-
R
R1
R2
-
Kd ± SD [nM] (n)
3.8 ± 2.5 (9)
142 ± 61 (3)
90.7, 48.4 (2)
572, 368 (2)
n.d.
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 (AM7)
2
3
-
6a
H
F
-
-
6b
-
6c
-
308, 397 (2)
194 (1)
6d
H
F
H
F
H
-
-
6e
-
n.d.
7a
R1
R2
R2
R1
R1
R1
R2
9.3, 10.7 (2)
120, 72 (2)
117, 310 (2)
109 ± 10 (3)
7.1, 6.5 (2)
4.2 ± 2.3 (9)
113 ± 50 (4)
7b
7c
7d
7e
7f (MT107)
7g
H
F
F
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 48
1
2
3
4
5
6
(n), number of experiments; SD, standard deviation; n.d., not determined. Kd values of individual
experiments are shown for n  2. R1, R2, see Figure 1.
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
Binding to hCD80. After the synthesis of the unlabelled compounds in Table 1, we investigated
their affinity and kinetics for the binding to recombinant hCD80 (rhCD80) by SPR. Typical
binding curves are shown in Supplementary Figure S1. The calculated binding rate constant (kon)
ranged from 9×104 to 7×105 M-1s-1. The dissociation rate constant (koff) ranged from 4×10-4 to 0.4
s-1 and Kd from 3.8 to 572 nM (Table 1). In order to reach high signal ratios between hCD80-
positive and negative tissues in PET imaging, we aimed at a strong affinity (low Kd) with a low
koff. This is the case for compounds in the right upper corner in the plot 1/Kd vs 1/koff in Figure 2
(1, 7a, 7e, 7f).
ACS Paragon Plus Environment
10

Page 11 of 48
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 2. 1/Kd plotted against 1/koff from SPR experiments. Compounds with preferred binding
properties for PET imaging are in the upper right corner (low Kd, low koff).
Displacement of hCTLA-4 and hCD28 from hCD80 binding. We next investigated by SPR the
potential of 1, 2 and 7f for displacing the binding domains of hCTLA-4 (in our study abatacept)
and rhCD28 from the binding to rhCD80. We first confirmed published Kd values of rhCTLA-4
(0.27 nM) and rhCD28 (16.3 nM) to rhCD80.17 In our assay, they were 0.5 nM and 17.8 nM,
respectively (n = 1). For the displacement assay, we used abatacept and rhCD28 at concentrations
3-4-fold above their Kd values, i.e., 2 nM and 50 nM, respectively. According to the Cheng-Prusoff
equation, keeping a constant ratio between concentration and KD of a displaced ligand (abatacept)
should result in similar IC50 values for a particular displacer compound in varying experiments.
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 48
1
2
3
4
5
6
7
8
9
The displacement of abatacept was most effective by 1 and 7f with IC50 values around
15 nM (Figure 3). The IC50 was more than one magnitude higher for 2. The IC50 values for the
displacement of rhCD28 were around 10 nM for 1 and 7f, as shown in Figure 3 (not determined
for 2). The respective Ki values calculated from these data are shown in Figure 3. The Ki
determined in the displacement experiments were in agreement with the Kd values determined in
the direct binding experiments (Table 1). Furthermore, 7f competed only weakly (at
concentrations > 50 µM) with the binding of abatacept to rhCD86 and recombinant murine CD80
(rmCD80), in agreement with its selectivity for hCD80 over hCD86 and mCD80 (Supplementary
Figures S2 and S3).
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 48
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: Representative displacement curves of (A) 2 nM abatacept or (B) 50 nM rhCD28 from
immobilized rhCD80 by small molecules 1, 2 (only in A) or 7f. IC50 values were determined from
the displacement by different concentrations of synthesized small molecules (x axis) assuming 1:1
binding stoichiometries. The respective (averaged) Ki values (± S.D. for n  3) are indicated in the
Figure. n, number of experiments for Ki determination.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 48
1
2
3
4
5
6
7
8
9
Radiosynthesis of [11C]1 and [11C]7f. Precursor compounds for radiolabelling were synthesized
from intermediates 6b and 6e (shown in Scheme 1 and 3, respectively). By activation of the
carboxylic acid with the coupling reagent EDC and following amide bond formation 1-boc-
piperidin-4-amine under basic conditions, a boc-protected piperidine was introduced which was
subsequently deprotected with DCM/TFA to afford compounds 8a and 8b in good yields (details
in experimental section). Our group recently described the radiolabelling procedure of 1 with
carbon-11 (ref. 6). This procedure was applied for the radiosynthesis of both [11C]1 and our new
radiotracer [11C]7f. Details on the one-step radiosynthesis are described in the experimental
section. Molar activities at the end of synthesis were between 200 and 500 GBq/μmol for [11C]1
and between 20 and 80 GBq/μmol for [11C]7f.
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
Lipophilicity, membrane permeability & plasma protein binding. In vivo, [11C]1 scarcely
distributed to the tissue, in agreement with its low log D at pH 7.4 of 0.1 and its low free fraction
in plasma of 0.02.6 We determined log D of [11C]7f at various pH. 7f differs from 1 by the C-[3-
(trifluoromethyl)phenyl] moiety replacing the nitrogen in position 2 of the cinnoline substructure
(Figure 1 and Table 1). As expected, this modification rendered the compound more lipophilic,
with a log D at pH 7.4 of 1.98 ± 0.06. The log D was significantly lower at pH 5.4 than at higher
pH values (Supplementary Figure S4). This is in agreement with the distribution characteristics of
a basic compound (tertiary amine in R2, Figure 1). The fit log P was 2.1 with a fit pKa of 7.1.
We furthermore investigated the potential of the synthetized compounds 1, 2, 7e, 7f and 7g to cross
lipid bilayers at pH 7.4, using a recently introduced liposomal fluorescent permeability assay.18
ACS Paragon Plus Environment
14

Page 15 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The logarithmic permeability coefficients, log (PFLipP, cm/s), ranged from -6.7 to -5.4, indicating
poor to moderate membrane permeability.18 Permeation kinetics were fastest for 1, 7e and 7f with
log PFLipP above -6.
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 plasma protein binding of [11C]7f was investigated with human plasma and bovine serum
albumin (BSA) using ultracentrifugation filter units. The free fraction was 0.023 ± 0.002 (n = 4)
for human plasma which is similar to the plasma protein binding of [11C]16, and a higher free
fraction for BSA of 0.159 ± 0.002 (n = 3).
In vitro autoradiography. We chose compound 7f for further in vitro and in vivo evaluation based
on its strong binding affinity to hCD80 (low Kd) at low koff (Figure 2 and Supplementary Figure
S1) and as it differed from 1 in at least one physicochemical property (higher lipophilicity).
Structural modifications with higher predicted impact on the physicochemical properties (e.g.,
protonation/deprotonation equilibria) resulted in weakened binding affinity (Figures 1, 2 and
Table 1). In vitro autoradiography with [11C]7f revealed higher accumulation on tissue slices of
hCD80-positive Raji than hCD80-negative NCI-H69 xenografts (Supplementary Figures S5 and
S6). The binding of [11C]7f to hCD80-positive Raji xenograft slices was reduced by an excess of
5 µM 7f, 1, 2 and abatacept, respectively, by around 35% of the total binding. No reduction in
[11C]7f binding was observed with the hCD80-negative tissue slices under these blocking
conditions.
In vivo evaluation in Raji xenograft bearing mice. We further evaluated [11C]7f in vivo by
PET/CT imaging. We have recently compared the in vivo pharmacokinetics of [11C]1 and [11C]7f
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 48
1
2
3
4
5
6
7
8
9
by kinetic modelling with PET data.19 Both tracers showed low distribution to tissue, in agreement
with the high plasma protein binding, but high extraction into the bile relative to the low free
fraction in plasma. This resulted in high abdominal radioactivity with spill-over to peripheral
tissues in the PET images. The biliary excretion of [11C]7f was significantly reduced by pre-
administration of cyclosporine, an inhibitor of several hepatic transport proteins.19 Here, we
evaluated the uptake of [11C]7f into the Raji xenografts from the data sets which were used for
pharmacokinetic modelling in the previous work.19 Figure 4 shows [11C]7f PET images of a Raji
xenograft bearing SCID mouse after administration of cyclosporine ~30 min before the tracer.
Scarce accumulation of [11C]7f in the periphery of the xenograft, as indicated in Figure 4, was
observed in two out of four cyclosporine-administered animals. No xenograft accumulation was
observed in the four animals without cyclosporine. For comparison, Figure 4 shows in addition a
representative [18F]FDG scan of a Raji-xenograft bearing mouse.
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: PET/CT images of a Raji (hCD80+) xenograft bearing mouse after tail vein injection of
11.7 MBq (13 nmol/kg) [11C]7f, 30 min after i.v. injection of 50 mg/kg cyclosporine (A-D). (E-H)
ACS Paragon Plus Environment
16

Page 17 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
For comparison, representative PET/CT images of a Raji (hCD80+) xenograft bearing mouse after
tail vein injection of 8.6 MBq [18F]FDG. Transversal (A,E), sagittal (B,F) and coronal (C,G)
planes, as indicated by the blue cross hairs. (D,H) Maximal intensity projection. CT, grey scale;
PET, color scale. H, blood pool (heart); GB, gall bladder; I, intestines; K, kidney; B, brain; P1 and
P2, peripheral locations of the xenograft with potential tracer accumulation; X, Raji xenograft.
SUV, standardised uptake value. Note that this scan was used for pharmacokinetic modelling in
our previous work19 and that different animals were used for the [11C]7f and the [18F]FDG scan.
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 evaluate the viability of the tumor cells and their general accessibility by radiotracers,
additional mice with Raji xenografts were scanned with the glucose analogue [18F]FDG. [18F]FDG
uptake was substantial in major parts of the xenografts (Figure 4), confirming the presence of
viable tumor tissue. The in vivo expression of hCD80 and its localisation on the cell surface were
confirmed by flow cytometry and RT-qPCR (Supplementary Figures S7 and S8).
The metabolic stability of [11C]7f was investigated in xenograft-free C57BL/6 mice. Only parent
tracer was detected in blood plasma, bile and urine 15 min after tracer i.v. injection (Supplementary
Figure S9).
DISCUSSION
Our SPR data from a set of novel structures revealed that the methylpiperidine moiety is an
essential part for the strong binding to hCD80. Compounds with the N-methyl-piperidine moiety
(1, 7a, 7e, 7f) showed consistently stronger affinities than compounds with the phenylethan-1-ol
moiety. In addition, the cinnoline- and quinoline-based structures showed stronger binding than
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 48
1
2
3
4
5
6
7
8
9
the pyrazolopyridine-based structures. The modification with an additional CF3-substituted phenyl
ring in position 2 of the quinoline ring system did not affect the binding affinity to hCD80. This
suggests that this part of the structure is tolerant for modifications, as was seen for novel structure
7f, which showed similar affinity as 1 despite an additional bulky group (CF3 substituted aromatic
ring). Taken together, the structure affinity relationship study with hybrid structures of 1 and 2
provides new insights in the structural requirements for CD80 binding and indicates that further
modifications are possible without affinity loss.
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
Displacement of rhCTLA-4 and rhCD28 from hCD80 by 1 and 7f showed that this low molecular
weight compound class has the potential to displace large proteins from binding to hCD80. This
ability may be attained by an allosteric mechanism, due to adjacent binding sites on the tertiary
structure of hCD80.13
The experimental log D at pH 7.4 of [11C]7f was two log units higher than that of [11C]1. However,
lipid bilayer permeation of both compounds 1 and 7f was poor to moderate. In addition, both had
similar high plasma protein binding with a free fraction in plasma of ~ 0.02 (Müller et al.6 and this
study). The free fraction of [11C]7f in 4% BSA was ~ 0.15 which indicates that besides albumin
an alternative plasma protein may be involved in the binding (assuming binding is similar to human
and bovine albumin). This may be 1-acid glycoprotein, which binds mainly basic compounds.20
The negligible uptake of both [11C]1 and [11C]7f into the hCD80-positive xenografts may result
from the high plasma protein binding, hampering extravasation of the tracers. The small volume
of distribution, in particular of [11C]7f (14 % of total body volume19) supports this hypothesis. The
pre-administration of cyclosporine reduced the biliary excretion significantly and in addition
increased the volume of distribution and tissue distribution, though without statistical
significance19. Both the reduction in spill-over from abdominal radioactivity and potential increase
ACS Paragon Plus Environment
18

Page 19 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
in tissue distribution may have contributed to the observed increased radioactivity in peripheral
parts of the hCD80-rich xenografts in two out of four cyclosporine treated animals.
7
8
9
Biliary excretion as well as strong binding to plasma proteins may depend on the cationic
characteristics of the tracers21,22,23,24. We currently investigate the effect of the charge and
ionization state on the binding affinity to hCD80, biliary excretion, plasma protein binding and
finally in vivo imaging.
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
A structure-affinity relationship study identified structural features which promote low nanomolar
affinity to hCD80. Despite strong in vitro binding of 7f to hCD80, observed by SPR and
autoradiography, PET-tracer [11C]7f hardly accumulated in hCD80-rich xenografts in vivo. High
plasma protein binding and high spill-over due to biliary excretion may have hampered observable
accumulation. Further structural modifications towards pharmacokinetic optimisation will be
based on these results. Besides imaging, our findings may be of interest for the development of
therapeutic CD80 inhibitors.
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 48
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
9
Materials & methods. All chemicals, unless otherwise stated, were purchased from Sigma
Aldrich GmbH (Taufkirchen, Germany), ABCR GmbH (Karlsruhe, Germany), Merck (Darmstadt,
Germany), Fluorochem Ltd. (Hadfield, GB), Tokyo Chemical Industry Co. Ltd (Tokyo, Japan),
Apollo Scientific Ltd (Cheshire, GB) or Fisher Scientific International Inc. (Hampton, US) and
were used without further purification.
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
Solvents for thin layer chromatography (TLC), column chromatography and extractions were
commercial grade. Organic chemistry reactions were monitored by TLC using Sigma-Aldrich
silica gel 60 plates (UV light at 254 nm). Nuclear magnetic resonance (NMR) spectra (1H and 13C
NMR) were obtained on a Bruker Avance FT-NMR spectrometer (400 MHz). Chemical shifts are
given in delta (δ) units, in ppm relative to tetramethylsilane (TMS, 0 ppm). Multiplicities in the 1H
NMR spectra are described as: s = singlet, d = doublet, t = triplet and m = multiplet. Coupling
constants (J) are reported in Hz. High resolution mass spectrometry (HR-MS) was performed by
the MS service of the Laboratory for Organic Chemistry at the ETH Zürich. The MALDI spectra
were recorded on a Bruker solariX (MALDI-FTICR-MS) or Bruker UltraFlex II (MALDI-TOF-
MS), while the ESI were recorded on a Bruker maXis (ESI-Qq-TOF-MS) or Bruker solariX (ESI-
FTICR-MS). High-performance liquid chromatography (HPLC) analysis was performed using a
reversed phase column (ACE column, C18, 3 μm) with a gradient system of acetonitrile and
50 mM sodium acetate. Semi-preparative HPLC purifications were carried out on a reversed phase
column (ACE column, Symmetry C8 5 μm; 7.8 × 50 mm) under following conditions: 4.1 g/L
sodium acetate in H2O (solvent A), MeCN (solvent B); isocratic with 25% B; flow rate
4.5 mL/min. The UV signal was measured at a wavelength of 320 nm.
ACS Paragon Plus Environment
20

Page 21 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The purity of all biologically tested compounds was  95%, except of intermediate 6d with a purity
of 92%. Purity was determined by HPLC using an analytical reverse phase column (Xbridge, C18,
5µm; 4.6 x 150mm) with H2O/MeCN (both containing 0.1% TFA) as mobile phase. Method: 2 min
isocratic 5% MeCN followed by a gradient from 5% to 90% MeCN in 12 min. Peaks
corresponding to the desired product are described, including retention time (rt) and purity by
integration.
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 recombinant human CTLA-4 fusion protein abatacept (Orencia®, Bristol-Myers-Squibb) and
cyclosporine (Sandimmun®, Novartis) were purchased from a local pharmacy.
All animal experiments were carried out in accordance with the Swiss Animal Welfare legislation
and approved by the Veterinary Office of the Canton Zurich (ZH17/2015 and ZH18/2018). Female
C.B.17 SCID mice (xenografts) and male C57BL/6 mice (tracer metabolism) were purchased from
Charles River Laboratories (Sulzfeld, Germany) and housed under standard conditions (12 h
light/12 h dark, free access to chow and water). [18F]FDG was generated in a routine production
by the University Hospital Zurich.
Chemistry
4-(6-fluoro-1,3-dihydro-3-oxo-2H-pyrazolo[4,3-c]cinnolin-2-yl)-N-(1-methyl-4-piperidinyl)-
benzamide (1). Compound 1 was synthesised from 6b following the same experimental procedure
+
like the synthesis of 2 and 1-methylpiperidine-4-amine. MS (ESI) calculated for C22H22FN6O2 :
421.2; m/z found was 421.2 [M+H]+; rt 9.36 min, 95%.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 48
1
2
3
4
5
6
7
8
9
4-[3,6-dihydro-6-methyl-3-oxo-4-[3-(trifluoromethyl)phenyl]dipyrazolo[3,4-b:3',4'-d]pyri-
din-2(1H)-yl]-N-[3-(1-hydroxyethyl)phenyl]-benzamide (2). 100 mg (0.22 mmol) of 6c were
dissolved in 1.1 mL DMF and stirred in a round-bottom flask. 93 mg (0.66 mmol) 1-(3-
aminophenyl)ethan-1-ol, 70 L (0.40 mmol) DIPEA and 85 mg (0.44 mmol) EDC hydrochloride
were added and the mixture was stirred for 7 h, before EtOAc and 0.25-M HCl was added, and
extraction was performed. The combined organic layers were washed with brine, dried over
sodium sulfate, filtered and concentrated. The crude was purified by flash chromatography (9:1
EtOAc:EtOH saturated with ammonium bicarbonate) and relevant fractions were collected and
concentrated, which resulted in a product weight of 78 mg (62% yield). HR-MS (ESI) calculated
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
+
for C30H24F3N6NaO3 : 595.1676; m/z found was 595.1675 [M+Na]+; rt 13.58 min, 95%.
2-chloro-5-[3,6-dihydro-6-methyl-3-oxo-4-[3-(trifluoromethyl)phenyl]dipyrazolo[3,4-b:3',
4'-d]pyridin-2(1H)-yl]-benzoic acid (3). 64 mg (0.28 mmol) of 2-chloro-5-hydrazinylbenzoic
acid hydrochloride were dispersed in 480 μL ethylene glycol, before 47.4 mg (0.47 mmol) of
sodium tert-butoxide were added and the mixture was stirred for 1.5 h at 75°C. It was cooled to rt
and 36 mg (0.09 mmol) of 5c were added with 1 mL of tert-butanol. The mixture was heated at
reflux (100°C) for 30 h. After 23 h additional sodium tert-butoxide (85 mg, 0.84 mmol) and
another 10 mg (0.05 mmol) of 2-chloro-5-hydrazinylbenzoic acid hydrochloride were added. The
reaction was quenched with water and EtOAc. The pH was set to 3 with 0.5-M HCl. The combined
organic layers were washed with brine, dried over sodium sulfate, filtered and concentrated. The
crude was dissolved in 2-3 mL EtOAc and kept at 0°C for 72 h. Orange precipitation formed,
which was filtered and resulted in a product weight of 14.3 mg (31% yield). MS (ESI) calculated
+
for C22H14ClF3N5O3 : 488.1; m/z found 488.1 [M+H]+; rt 13.64 min, 95%.
ACS Paragon Plus Environment
22

Page 23 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Ethyl-3-methoxy-3-(3-(trifluoromethyl)phenyl)acrylate (4). 5 mL (28 mmol) ethyl 3-
(trifluoro-methyl)benzoate were dissolved in 200 mL EtOAc. 1.13 mg (28 mmol) sodium hydride
was added under stirring and heated to 40°C for 5 min, before another 1.13 mg (28 mmol) sodium
hydride was added and the reaction mixture was heated under reflux (90°C) for 18 h. The reaction
was cooled to rt and DCM/water was added. Extraction was performed and the combined organic
layers were washed with brine, dried over sodium sulfate, filtered and concentrated. The crude was
further dried with benzene under reduced pressure. In a second step, the crude intermediate was
dissolved in 100 mL 4:1 MeCN:MeOH and 28 mL (56 mmol) 2-M (diazomethyl)trimethylsilane
in hexane was added. The reaction was stirred for 36 h at rt, before it was quenched with 8 mL
aqueous 1-M HCl and extraction was performed with EtOAc. The combined organic layers were
washed with brine, dried over sodium sulfate, filtered and concentrated, which resulted in 9.63 g
crude. Flash chromatography was performed with 95:5 hexane:EtOAc, resulting in a product
weight of 3.74 g (48% yield). 1H NMR (400 MHz, Chloroform-d) δ 7.81 (s, 1H), 7.74 (d, J = 7.5
Hz, 1H), 7.68 (d, J = 7.5 Hz, 1H), 7.53 (t, J = 7.5 Hz, 1H), 5.63 (s, 1H), 4.22 (q, J = 7.1 Hz, 2H),
3.90 (s, 3H), 1.31 (t, J = 7.1 Hz, 4H). 19F NMR (376 MHz, Chloroform-d) δ -62.80. HR-MS (ESI)
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
+
calculated for C13H14F3O3 : 275.0890; m/z found was 275.0895 [M+H]+.
Methyl 4-hydroxylcinnoline-3-carboxylate (5a). To a mixture of 6 mL (84 mmol)
thionylchloride and 10 mL chloroform, 4 g (19 mmol) of compound 2-(2-
phenylhydrazono)malonic acid was added. The reaction mixture was stirred at 80°C for two hours
under nitrogen atmosphere. The solvents were removed under vacuum and the brown residue was
re-dissolved in dichloroethane and dried again. The resulting acid chloride was dissolved in 10 mL
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 48
1
2
3
4
5
6
7
8
9
dichloroethane, then 1.2 mL TiCl4 were added and the dark brown solution was stirred at 110°C
for 4 hours under nitrogen. The reaction mixture was cooled to 10°C, 20 mL methanol were added
slowly and the resulting red-brown solution was stirred at 60°C for 30 min. The solvents were
removed under reduced pressure and the residues were crystalized from 10 mL acetonitrile to yield
0.6 g of compound 5a (15% yield). 1H NMR (400 MHz, DMSO-d6) δ 13.94 (s, 1H), 8.10 (d, J =
8.2 Hz, 1H), 7.84 (t, J = 7.7 Hz, 1H), 7.68 (d, J = 8.2 Hz, 1H), 7.52 (t, J = 7.7 Hz, 1H), 3.83 (s,
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
+
3H). MS (ESI) calculated for C10H9N2O3 : 205.1; m/z found was 204.7 [M+H]+.
Methyl 4-hydroxy-8-fluorocinnoline-3-carboxylate (5b). Compound 5b was synthesized in an
analogous way to the procedure for compound 5a, but with 2-(2-(2-
fluorophenyl)hydrazono)malonic acid. 1H NMR (400 MHz, DMSO-d6) δ 14.19 (s, 1H), 7.90 (d, J
= 8.2 Hz, 1H), 7.77 (ddd, J = 11.2, 8.0, 1.2 Hz, 1H), 7.49 (td, J = 8.1, 4.8 Hz, 1H), 3.85 (s, 3H).
+
MS (ESI) calculated for C10H8FN2O3 : 223.0; m/z found was 222.9 [M+H]+.
Ethyl 4-chloro-1-methyl-6-(3-(trifluoromethyl)phenyl)-1H-pyrazolo[3,4-b]pyridine-5-car-
boxylate (5c). This intermediate was synthesised from 4 following the same experimental
procedure like the synthesis of 5e and ethyl 5-amino-1-methyl-1H-pyrazole-4-carboxylate,
resulting in a product weight of 830 mg (60% yield). 1H NMR (400 MHz, Chloroform-d) δ 8.18
(s, 1H), 7.96 (s, 1H), 7.88 (d, J = 7.8 Hz, 1H), 7.73 (d, J = 7.8 Hz, 1H), 7.60 (t, J = 7.8 Hz, 1H),
4.22 (q, J = 7.1 Hz, 2H), 4.20 (s, 3H), 1.12 (t, J = 7.1 Hz, 3H). MS (ESI) calculated for
+
C17H14ClF3N3O2 : 384.1; m/z found was 384.2 [M+H]+.
ACS Paragon Plus Environment
24

Page 25 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Ethyl 4-chloro-2-(3-(trifluoromethyl)phenyl)quinoline-3-carboxylate (5d). This intermediate
was synthesised from 4 following the same experimental procedure like the synthesis of 5e and
ethyl-2-aminobenzoate, resulting in a product weight of 418 mg (61% yield). 1H NMR (400 MHz,
Chloroform-d) δ 8.32 (d, J = 8.4 Hz, 1H), 8.19 (d, J = 8.4 Hz, 1H), 8.03 (s, 1H), 7.95 (d, J = 7.8
Hz, 1H), 7.86 (t, J = 8.4 Hz, 1H), 7.77 – 7.69 (m, 2H), 7.62 (t, J = 7.8 Hz, 1H), 4.29 (q, J = 7.2
Hz, 2H), 1.18 (t, J = 7.2 Hz, 3H). 13C NMR (101 MHz, CDCl3) δ 166.2, 154.7, 148.0, 141.2, 140.0,
132.0, 131.0, 130.1, 129.3, 128.8, 127.1, 126.9, 126.2, 125.6, 125.4, 124.7, 122.7, 62.6, 13.8. 19F
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
+
NMR (376 MHz, Chloroform-d) δ -62.68. HR-MS (ESI) was calculated for C19H13ClF3NNaO2 :
380.0660; m/z found was 380.0663 [M+H]+.
Ethyl 4-chloro-8-fluoro-2-(3-(trifluoromethyl)phenyl)quinoline-3-carboxylate (5e). 740 mg
(4.4 mmol) of methyl 2-amino-3-fluorobenzoate were dissolved in 8 mL THF, before 193 mg
(4.8 mmol) sodium hydride was added and the mixture was stirred at room temperature. After
30 min, 1.20 g (4.4 mmol) of 4 with 1.7 mL THF was transferred to the mixture. The reaction was
run under reflux for 16 h, before it was cooled to room temperature and quenched with water. The
pH was set to 5 with aqueous HCl. Extraction was performed with EtOAc and the combined
organic fractions were dried, filtered and concentrated, which resulted in a crude weight of 1.86 g.
The crude was dissolved in 19 mL phosphorus oxychloride and stirred at reflux for 2 h. After
cooling to room temperature, it was concentrated under reduced pressure. The residues were
dissolved in EtOAc, cooled to 0°C and neutralised with aqueous 5% NaHCO3. Extraction was
performed with EtOAc and the collected organic layers were dried over sodium sulfate and
concentrated, resulting in a crude weight of 1.81 g. Flash chromatography (9:1 Hexane:EtOAc)
was performed, giving a product weight of 955 mg (55% yield). 1H NMR (400 MHz, Chloroform-
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 48
1
2
3
4
d) δ 8.12 (d, J = 8.5 Hz, 1H), 8.02 (s, 1H), 7.98 (d, J = 7.7 Hz, 1H), 7.75 (d, J = 7.7 Hz, 1H), 7.72
5
6
13
– 7.54 (m, 3H), 4.30 (q, J = 7.1 Hz, 1H), 3.84 (s, 2H), 1.19 (t, J = 7.1 Hz, 1H). C NMR (101
7
8
9
MHz, CDCl3) δ 165.9, 159.5, 156.9, 154.8, 139.7, 132.2, 129.5, 128.6, 128.5, 128.1, 126.4, 126.3,
19
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
125.6, 125.4, 122.7, 120.5, 116.2, 62.8, 13.8. F NMR (376 MHz, Chloroform-d) δ -62.66, -
+
122.57. HR-MS (ESI) calculated for C19H13ClF4NO2 : 398.0565; m/z found was 398.0565
[M+H]+.
4-(3-oxo-1,3-dihydro-2H-pyrazolo[4,3-c]cinnolin-2-yl)benzoic acid (6a). 0.5 g (2.4 mmol) of
compound 5a were added in 2 mL (27 mmol) thionyl chloride and the reaction mixture was stirred
at 80°C for 1 h. The excess of thionyl chloride was removed under vacuum. The residue was
dissolved in toluene and the solvent was removed under vacuum to remove the traces of thionyl
chloride. Then 0.73 g (4.8 mmol) 4-hydrazinylbenzoic acid and 7 mL ethanol were added and the
reaction mixture was stirred at 70°C for 1 h. The resulting precipitates were collected and washed
with methanol:37% aqueous HCl mixture (9:1, 10 mL). The crude product was washed
additionally 3 times with water, methanol, and methylene chloride and dried to yield 0.57 g of
compound 6a (78% yield). 1H NMR (400 MHz, DMSO-d6) δ 14.62 (s, 1H), 12.84 (s, 1H), 8.33 (d,
J = 8.9 Hz, 2H), 8.21 (d, J = 7.8 Hz, 1H), 8.06 (d, J = 8.9 Hz, 2H), 7.86 – 7.78 (m, 2H), 7.73 –
7.66 (m, 1H). MS (ESI) calculated for C16H11N4O3 : 307.1; m/z found was 306.7 [M+H]+; rt
+
10.78 min, >99%.
4-(6-fluoro-3-oxo-1,3-dihydro-2H-pyrazolo[4,3-c]cinnolin-2-yl)benzoic acid (6b). The
structure was synthesised in an analogous way to the procedure of compound 6a with compound
5b as starting material. 1H NMR (400 MHz, DMSO-d6) δ 14.71 (s, 1H), 12.87 (s, 1H), 8.26 (d, J
ACS Paragon Plus Environment
26

Page 27 of 48
Journal of Medicinal Chemistry
1
2
3
4
= 8.9 Hz, 2H), 8.02 (d, J = 8.9 Hz, 2H), 7.96 (d, J = 7.8 Hz, 1H), 7.73 – 7.58 (m, 2H). MS (ESI)
5
6
+
calculated for C16H11FN4O3 : 325.1; m/z found was 324.7 [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
4-(6-methyl-3-oxo-4-(3-(trifluoromethyl)phenyl)-3,6-dihydrodipyrazolo[3,4-b:3',4'-
d]pyridin-2(1H)-yl)benzoic acid (6c). This intermediate was synthesised from 5c following the
experimental procedure like the synthesis of 6e, resulting in a product weight of 825 mg (85%
yield). 1H NMR (400 MHz, Acetone-d6) δ 12.24 (s, 1H), 8.45 (s, 1H), 8.39 (d, J = 7.8 Hz, 1H),
8.22 (s, 1H), 8.12 (s, 4H), 7.87 (d, J = 7.8 Hz, 1H), 7.76 (t, J = 7.8 Hz, 1H), 4.16 (s, 3H). 19F NMR
+
(376 MHz, Acetone-d6) δ -62.88. HR-MS (ESI) calculated for C22H15F3N5O3 : 454.1122; m/z
found was 454.1128 [M+H]+; rt 13.31 min, >99%.
4-(3-oxo-4-(3-(trifluoromethyl)phenyl)-1,3-dihydro-2H-pyrazolo[4,3-c]quinolin-2-
yl)benzoic acid (6d). This intermediate was synthesised from 5d following the same experimental
procedure like the synthesis of 6e, resulting in a product weight of 120 mg (68% yield). 1H NMR
(400 MHz, DMSO-d6) δ 12.72 (s, 1H), 8.39 – 8.32 (m, 3H), 8.30 (d, J = 7.6 Hz, 1H), 8.25 (d, J =
7.8 Hz, 1H), 8.05 (d, J = 7.8 Hz, 1H), 8.01 (d, J = 8.9 Hz, 2H), 7.90 – 7.84 (m, 2H), 7.73 (t, J =
13
7.6 Hz, 1H), 7.60 (t, J = 7.6 Hz, 1H). C NMR (101 MHz, DMSO-d6) δ 167.5, 162.0, 151.3,
144.7, 143.9, 136.6, 134.7, 131.0, 130.9, 130.7, 129.7, 128.5, 127.4, 127.1, 126.1, 125.8, 123.1,
122.4, 120.3, 118.7 118.2, 102.6. 19F NMR (376 MHz, DMSO-d6) δ -60.96. MS (ESI) calculated
+
for C24H15F3N3O3 : 450.1; m/z found was 450.1 [M+H]+; rt 13.35 min, 92%.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 48
1
2
3
4
5
6
7
8
9
4-(6-fluoro-3-oxo-4-(3-(trifluoromethyl)phenyl)-1,3-dihydro-2H-pyrazolo[4,3-c]quinolin-2-
yl)benzoic acid (6e). 429 mg (2.82 mmol) of 4-hydrazinylbenzoic acid were dissolved in 4.7 mL
ethylene glycol, 279 mg (2.82 mmol) sodium tert-butoxide were added and stirred at 75°C for 1 h.
After cooling, 374 mg (0.94 mmol) of 5e dissolved in 2.0 mL THF were added and the reaction
mixture was heated to 100°C. After 16 h, another 100 mg of sodium tert-butoxide was added and
stirred at 100°C for another 3 h. The reaction was quenched with H2O and EtOAc. The crude
solution was acidified to pH 3 using 1-M HCl and extracted with EtOAc. The organic layers were
dried with sodium sulfate and concentrated, giving 54.4 mg crude. For purification, trituration was
performed as described in the synthesis for 7g. In short, the crude was dissolved in aceton, few
1-M HCl was added and the solution was concentrated at 100 mbar at 40°C until precipitate
formed. The suspension was stored in the fridge for 1 h. Filtration, washing with 1-M HCl and
cold H2O, and subsequent drying, resulted in a product weight of 288 mg (66% yield). 1H NMR
(400 MHz, DMSO-d6) δ 12.79 (s, 1H), 12.58 (s, 1H), 8.30 (d, J = 8.9 Hz, 2H), 8.24 (s, 1H), 8.17
(d, J = 7.8 Hz, 1H), 8.08 (d, J = 7.4 Hz, 1H), 8.04 – 7.95 (m, 3H), 7.81 (t, J = 7.8 Hz, 1H), 7.71 –
7.43 (m, 2H). 13C NMR (101 MHz, DMSO-d6) δ 167.0, 161.3, 153.7, 151.9, 151.2, 143.5, 143.3,
134.6, 130.3, 128.8, 128.5, 128.2, 127.7, 127.2, 126.9 (d, J = 7.6 Hz), 125.7, 125.5, 122.8, 120.4,
117.7, 115.8 (d, J = 17.1 Hz), 103.1.19F NMR (376 MHz, DMSO-d6) δ -60.92, -124.46. HR-MS
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
+
(ESI) calculated for C24H14F4N3O3 : 468.0966; m/z found was 468.0966 [M+H]+.
N-(1-methylpiperidin-4-yl)-4-(3-oxo-1,3-dihydro-2H-pyrazolo[4,3-c]cinnolin-2-
yl)benzamide (7a). 24.7 mg (0.08 mmol) of 6a were dissolved in 850 L 2:1 DCM/THF and
240 L (3.3 mmol) thionylchloride was added. The mixture was stirred at reflux for 3 h. The
reaction was cooled down and solvents were evaporated under reduced pressure. Benzene was
ACS Paragon Plus Environment
28

Page 29 of 48
Journal of Medicinal Chemistry
1
2
3
4
added two times to remove solvent residues, resulting in a crude weight of 32.1 mg. This crude
5
6
was then resuspended in 200 L DMF. 12.7 L (0.10 mmol) 1-methylpiperidin-4-amine and
7
8
9
22.8 L (0.16 mmol) triethylamine were mixed with 210 L DMF in a separate flask and then
added to the reaction. After 20 hours, the reaction was stopped by adding aqueous HCl solution
and EtOAc, which led to the formation of precipitates. The precipitates were washed with aqueous
HCl solution and cold water. The precipitates were dissolved with ethanol/water and concentrated
to half the volume and stored in the fridge overnight. Orange precipitates formed and were filtered
and washed. The precipitates were treated with toluene and dried under high-vacuum to give a
final product weight of 6.4 mg (19% yield). 1H NMR (400 MHz, DMSO-d6) δ 8.35 (d, J = 8.9 Hz,
2H), 8.33 – 8.29 (m, 1H), 8.21 (d, J = 8.0, 1H), 7.97 (d, J = 8.9 Hz, 2H), 7.90 (d, J = 8.0 Hz, 1H),
7.78 (t, J = 7.7 Hz, 1H), 7.66 (t, J = 7.7 Hz, 1H), 3.96 – 3.83 (m, 1H), 3.14 – 3.03 (m, 2H), 2.57 –
2.51 (m, 2H), 2.45 (s, 3H), 1.94 – 1.86 (m, 2H), 1.79 – 1.66 (m, 2H). HR-MS (ESI) calculated for
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
+
C22H23N6O2 : 403.1877; m/z found was 403.1875 [M+H]+; rt 9.32 min, 97%.
4-(6-fluoro-3-oxo-1,3-dihydro-2H-pyrazolo[4,3-c]cinnolin-2-yl)-N-(3-(1-
hydroxyethyl)phenyl)benzamide (7b). This structure was synthesised from 20 mg of 6b
following the same experimental procedure like the synthesis of 7c, resulting in a product weight
of 11 mg (40% yield). 1H NMR (400 MHz, DMSO-d6) δ 10.23 (s, 1H), 8.34 (d, J = 8.9 Hz, 2H),
8.13 (d, J = 8.9 Hz, 2H), 8.05 (d, J = 7.6 Hz, 1H), 7.77 (s, 1H), 7.77 (s, 1H), 7.76 – 7.64 (m, 3H),
7.29 (t, J = 7.8 Hz, 1H), 7.08 (d, J = 7.8 Hz, 1H), 5.19 (s, 1H), 4.72 (q, J = 6.5 Hz, 1H), 1.34 (d,
J = 6.5 Hz, 3H). 19F NMR (376 MHz, DMSO-d6) δ -125.47. HR-MS (MALDI/ESI) calculated for
+
C24H18FN5NaO3 : 466.1286; m/z found was 466.1282 [M+Na]+; rt 11.64 min, 98%.
ACS Paragon Plus Environment
29