A structure-activity analysis of biased agonism at the dopamine D2 receptor

Article abstract of DOI:10.1021/jm401318w

Biased agonism offers an opportunity for the medicinal chemist to discover pathway-selective ligands for GPCRs. A number of studies have suggested that biased agonism at the dopamine D₂ receptor (D₂R) may be advantageous for the treatment of neuropsychiatric disorders, including schizophrenia. As such, it is of great importance to gain insight into the SAR of biased agonism at this receptor. We have generated SAR based on a novel D₂R partial agonist, tert-butyl (trans-4-(2-(3,4-dihydroisoquinolin- 2(1H)-yl)ethyl)cyclohexyl)carbamate (4). This ligand shares structural similarity to cariprazine (2), a drug awaiting FDA approval for the treatment of schizophrenia, yet displays a distinct bias toward two different signaling end points. We synthesized a number of derivatives of 4 with subtle structural modifications, including incorporation of cariprazine fragments. By combining pharmacological profiling with analytical methodology to identify and to quantify bias, we have demonstrated that efficacy and biased agonism can be finely tuned by minor structural modifications to the head group containing the tertiary amine, a tail group that extends away from this moiety, and the orientation and length of a spacer region between these two moieties.

Full text of DOI:10.1021/jm401318w

Subscriber access provided by FORDHAM UNIVERSITY
Article
A Structure-Activity Analysis of Biased Agonism at the Dopamine D2 Receptor
Jeremy Shonberg, Carmen Klein Herenbrink, Laura Lopez Munoz, Arthur
Christopoulos, Peter J. Scammells, Ben Capuano, and J. Robert Lane
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 19 Oct 2013
Downloaded from http://pubs.acs.org on October 31, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A StructureꢀActivity Analysis of Biased Agonism
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
at the Dopamine D Receptor
2
†,#
‡,#
‡
‡
Jeremy Shonberg, Carmen Klein Herenbrink, Laura López, Arthur Christopoulos,
†
†,*
‡,*
Peter J. Scammells, Ben Capuano, J. Robert Lane.
†
‡
Medicinal Chemistry and Drug Discovery Biology, Monash Institute of Pharmaceutical
Sciences, Monash University (Parkville Campus), 381 Royal Parade, Parkville, Victoria 3052
Australia
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
Biased agonism offers an opportunity for the medicinal chemist to discover pathwayꢀ
selective ligands for GPCRs. A number of studies have suggested that biased agonism at 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
2 2
dopamine D receptor (D R) may be advantageous for the treatment of neuropsychiatric
disorders, including schizophrenia. As such, it is of great importance to gain insight into the
SAR of biased agonism at this receptor. We have generated SAR based around a novel D R
2
partial
agonist,
tertꢀbutyl
(transꢀ4ꢀ(2ꢀ(3,4ꢀdihydroisoquinolinꢀ2(1H)ꢀ
yl)ethyl)cyclohexyl)carbamate (4). This ligand shares structural similarity to cariprazine (2),
a drug awaiting FDA approval for the treatment of schizophrenia, yet displays a distinct bias
towards two different signaling endpoints. We synthesized a number of derivatives of 4 with
subtle structural modifications, including incorporation of cariprazine fragments. By
combining pharmacological profiling with analytical methodology to identify and quantify
bias we have demonstrated that efficacy and biased agonism can be finely tuned by minor
structural modifications to the head group containing the tertiary amine, a tail group that
extends away from this moiety and the orientation and length of a spacer region between
these two moieties.
2
ACS Paragon Plus Environment

Page 3 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
G proteinꢀcoupled receptors (GPCRs), also known as 7ꢀtransmembrane receptors are the
1
single largest class of drug targets, and have more than 800 members in the human
2,3
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
genome. For decades, studies of novel GPCRꢀtargeting compounds have focused almost
exclusively on affinity for the receptor as a single descriptor, yet it is efficacy that ultimately
4
determines the nature of the cellular response. More recently, it has become obvious that
rather than “linear” (sequentiallyꢀcoupled) signaling cascades, GPCRs activate nonꢀlinked
5
ensembles of G proteinꢀdependent and ꢀindependent signaling pathways. Furthermore, there
is increasing evidence that highlights the ability of ligands acting at the same GPCR to
stabilize distinct receptor conformations that, in turn, are linked to different functional
4,6
outcomes. This challenges earlier ideas of linear efficacy, and the emerging paradigm has
been termed biased agonism, stimulus bias, ligand directed signaling, or functional
4
ꢀ8
selectivity.
Biased agonism provides the opportunity to design pathwayꢀselective in
addition to receptor subtype selective ligands. However, understanding the structureꢀactivity
relationships (SAR) around biased ligands is, as yet, a largely unexplored challenge. To
achieve this, standard SAR must be enriched through incorporation of parameters that allow
quantification of bias. This concept represents a challenge in itself.
The dopamine D
2
receptor (D
2
R) represents a significant drug target for the treatment of
9
diseases including schizophrenia and Parkinson’s disease. Currently, all clinically marketed
st
nd
antipsychotics act by targeting the D
2
R either as antagonists/inverse agonists (1 and 2
rd
generation antipsychotics), or partial agonists (3 generation antipsychotics), thereby
modulating the action of the neurotransmitter dopamine. This latter approach is exemplified
by the discovery of aripiprazole (1); a D
2
R partial agonist marketed for the treatment of
R partial agonists have entered clinical trials,
schizophrenia. Although a number of D
2
aripiprazole remains the sole example of this ligand class in the clinic. More recently,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aripiprazole has been shown to display biased agonism, and it has been postulated that this
1
0,11
bias may underlie its clinical efficacy.
biased ligands at the D R given a number of other D
clinical trials for the treatment of schizophrenia, including cariprazine (2).
The 1,2,3,4ꢀtetrahydroisoquinoline (THIQ) moiety has been explored as an isostere for the
As such, it is important to understand the SAR of
2
2
R partial agonists are also currently in
12ꢀ14
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
5ꢀ17
D
2
ꢀlike receptor privileged phenylpiperazine scaffold.
has yielded the D R subtypeꢀselective antagonist SB269652 (3), although more recently this
compound has been suggested to act as a negative allosteric modulator at the D Our
Indeed exploration of this structure
3
15,18
2
R.
own exploration of this scaffold revealed tertꢀbutyl (transꢀ4ꢀ(2ꢀ(3,4ꢀdihydroisoquinolinꢀ
2(1H)ꢀyl)ethyl)cyclohexyl)carbamate (4), a novel D R partial agonist with subꢀmicromolar
2
potency and structural similarities to cariprazine (Figure 1). To further probe this finding, we
generated derivatives focused on three main portions of the lead compound: the tertiary
amineꢀcontaining “head group”; the cyclohexylene “spacer” group, and the tertꢀbutyl
carbamate “tail group” as seen in Figure 2. We combined this approach with novel analytical
1
9
pharmacology methods that allow us to quantify biased agonism and gain novel insight
2
around SAR for the fine control of ligand efficacy and biased signaling at the D R.
Results
Chemical Synthesis
The synthesis of all compounds followed the procedures outlined in Schemes 1ꢀ7. Ethyl 2ꢀ
transꢀ4ꢀaminocyclohexyl)acetate (8) was synthesized by hydrogenation of 4ꢀ
(
nitrophenylacetic acid (5) in a Parr shaker for 3 days at 60 psi, followed by ethyl ester
formation in the presence of concentrated hydrochloric acid and ethanol. The trans
stereoisomer (8) was isolated as the hydrochloride salt by fractional crystallization from
diethyl ether and acetonitrile. This was then protected as the tertꢀbutyl carbamate (12) with
4
ACS Paragon Plus Environment

Page 5 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
diꢀtertꢀbutyl dicarbonate in good yield. Other commercially available trans amino acids,
notably transꢀ4ꢀaminocyclohexanecarboxylic acid (6) and tranexamic acid (7), were
transformed to their respective ethyl esters (9 and 10) as in the synthesis of 8, then protected
as the tertꢀbutyl carbamate (13 and 14, respectively). For the synthesis of the cis
stereoisomer, 2ꢀ(cisꢀ4ꢀ((tertꢀbutoxycarbonyl)amino)cyclohexyl)acetic acid (11) was
commercially supplied, and prepared as the ethyl ester (15) following Steglich esterification
conditions using EDCI, shown in Scheme 1. This was applied to prevent deprotection of the
tertꢀbutyl carbamate group under general Fischer esterification conditions. The prepared ethyl
ester materials (12ꢀ15) were then reduced to their respective aldehydes (16ꢀ19) using
DIBALH, generally in nearꢀquantitative yields.
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 aromatic spacer was prepared according to Scheme 2. 4ꢀNitrophenylacetic acid (5)
was initially converted to the ethyl ester using Fischer esterification conditions, then reduced
to the corresponding aniline (20) in the presence of tin and concentrated hydrochloric acid.
Following basic workup, 20 was protected as the tertꢀbutyl carbamate (21), then treated with
DIBALH to furnish the target aldehyde (22) in good overall yield. The head groups were
either commercially supplied (23, 27, 28) or synthetically prepared as described in Scheme 3.
Nitration of 23 in concentrated sulfuric acid with slow addition of sodium nitrate afforded 24
in good yield. The preparation of 1ꢀ(2,3ꢀdichlorophenyl)homopiperazine (2,3ꢀDCPHP, 26)
11
followed a literature procedure by Allen et al. BuchwaldꢀHartwig amination of 1ꢀbromoꢀ
,3ꢀdichlorobenzene (25) with homopiperazine in the presence of BINAP,
2
tris(dibenzylideneacetone)dipalladium(0), potassium tertꢀbutoxide and diꢀtertꢀbutyl
dicarbonate afforded the protected product, which was deprotected in situ (26), in low overall
yield (3%).
Scheme 4 describes the synthesis of target compounds with variations to the head groups
(4, 29ꢀ32) through reductive alkylation of 16 with a respective amine (23, 24, 26ꢀ28) in 1,2ꢀ
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
DCE with sodium triacetoxyborohydride, generally in moderate yields following purification.
The synthesis of THIQꢀcontaining target compounds with variations to the tail group (35ꢀ42),
shown in Scheme 5, required deprotection of the tertꢀbutyl carbamate derivatives (4, 30) by
trifluoroacetic acid, followed by basic workꢀup to give the free amines (33 and 34) generally
in high yields. Treatment of the amine with acetic anhydride in the presence of Hünig's base
furnished the corresponding acetamide derivatives (35 and 36) in good yields. A range of
carbamates (37ꢀ40) was synthesized in good yields using the corresponding chloroformates
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
(isopropyl, isobutyl or phenyl chloroformate) or diethylpyrocarbonate. The synthesis of the
ureaꢀcontaining tail group was achieved with either dimethylcarbamyl chloride to give the
N,Nꢀdimethylurea (41) in good yield, or 1,1'ꢀcarbonyldiimidazoleꢀmediated coupling with
tertꢀbutylamine to furnish the tertꢀbutyl urea derivative (42) in moderate yield. These
modifications afforded a range of target compounds with variations to the tail group.
The synthesis of compounds with variations in length and nature of the spacer group (43ꢀ
49), illustrated in Scheme 6, followed reductive alkylation conditions with a respective
aldehyde (17ꢀ19, 22) in 1,2ꢀDCE with sodium triacetoxyborohydride, generally in moderate
yields following purification. This afforded a range of compounds containing the THIQ (43ꢀ
46), or 2,3ꢀDCPP (47ꢀ49) head groups with various spacers. Finally, the synthesis of
compounds with the N,Nꢀdimethylurea tail group (2, 54ꢀ56), depicted in Scheme 7, followed
the same general procedure as outlined in Scheme 5 for the synthesis of compounds with
variations to the tail group.
Analytical approach to quantifying stimulus bias from concentrationꢀresponse curves
6
ACS Paragon Plus Environment

Page 7 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To confirm onꢀtarget activity, all compounds were tested for their ability to displace the
3
radiolabeled antagonist [ H]spiperone at the human D2LR expressed in FlpIn CHO cell
membranes (Tables 1ꢀ4; Supplementary Tables 1 and 2). In addition, to assess their
functional activity, all compounds were tested for their ability to stimulate D R mediated
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
inhibition of forskolinꢀstimulated cAMP production. This endpoint represents a canonical
D
2LR signaling pathway mediated by coupling to Gi/o proteins that, in turn, inhibit adenylate
cyclase. We then explored whether such molecular determinants were also important for
another D2LR mediated pathway. Compounds were tested in an assay measuring
phosphorylation of ERK1/2 through activation of the D R expressed in FlpIn CHO cells.
2L
The potency (pEC ) and maximal responses (E as a percentage relative to the maximal
max
5
0
effect of dopamine) of the various compounds are displayed in Supplementary Tables 1 and
. Although reversals in orders of potency or maximal effect are indicators of biased
agonism, the comparison of rank orders of potency or efficacy represents a subꢀoptimal
2
20
approach to identify biased agonists. Using potency values to characterize and quantify
agonist activity is inadequate for agonists that produce different maximal responses and using
4
maximal responses alone fails to differentiate between full agonists. Therefore, a
methodology that captures information encoded by the entire concentrationꢀresponse
relationship and relates this to changes in ligand structure will be far more likely to develop
useful SAR around a biased ligand. We have developed an analytical approach that satisfies
21
these criteria based on the ‘Operational Model of Agonism’ first derived by Black and Leff.
Using this approach we can obtain the functional equilibrium dissociation constant, i.e.
affinity, for the receptor (denoted as K ) coupled to a particular effector protein of signaling
A
pathway, and τ, which encompasses both the intrinsic efficacy of the agonist in activating a
particular cellular response pathway and receptor density. The values can be combined to
give a “transduction coefficient” (τ/K ), as an overall measure of the power of an agonist.
A
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Values of log(τ/K
A
) can be normalized to the reference agonist, dopamine, to cancel the
impact of postꢀreceptor “system bias” (which would be common to all ligands), yielding
values of ꢁlog(τ/K
A
). These values can then be compared across pathways for each ligand to
6
give values of ꢁꢁlog(τ/K ), a quantitative measure of bias between two different pathways.
Α
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
Using this methodology we can relate changes in chemical structure to changes in functional
A A
affinity (K ), efficacy (τ) or bias (ꢁꢁlog(τ/K )) between two signaling pathways.
Pharmacological Profiling of Compounds
The lead compound (4) acted as a partial agonist in both the cAMP assay and the pERK1/2
assay with subꢀmicromolar affinity (K = 741 nM) for the D R as determined by ligand
i
2L
binding (Table 1). Compound 4 did not show significant bias towards one signaling pathway
as compared to dopamine (ꢁꢁlog(τ/K ) = 0.51 ± 0.23). Subtle minimization of tail group
Α
branching to give the isopropyl carbamate (38) resulted in no change in affinity relative to 4,
and a loss of detectable agonism in the pERK1/2 assay. In the cAMP assay a significant 4ꢀ
fold decrease in efficacy (τ) was observed (Table 1). Further minimization of branching to
the ethyl carbamate (37) caused no significant change in affinity relative to 4, and again no
agonism in the pERK1/2 assay. However, in the cAMP assay, a 6ꢀfold increase in functional
A
affinity (pK ) was accompanied by a 5ꢀfold decrease in efficacy (τ). This results in a similar
transduction coefficient for 4 and 37 in the cAMP assay. However, subtle modification to the
isobutyl carbamate (39) significantly reduced the transduction coefficient relative to 4 in the
cAMP assay and resulted in a complete loss of efficacy in the pERK1/2 assay. Similarly,
incorporation of aromaticity to the tail group with the phenyl carbamate (40) resulted in a
significant 20ꢀfold decrease in the transduction coefficient in the cAMP assay as compared to
4
and no detectable agonism in the pERK1/2 assay. Modification of the tail group of 4 to the
acetamide (35) resulted in complete loss of agonism in the pERK1/2 assay and a significant
8
ACS Paragon Plus Environment

Page 9 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
decrease in efficacy (τ) in the cAMP assay. Modification of the tail group from tertꢀbutyl
carbamate to tertꢀbutyl urea (42) showed similar affinity to 4 in a radioligand binding assay,
no agonist activity in the pERK1/2 assay and a significant 6ꢀfold decrease in efficacy (τ) in
the cAMP assay. As such, all of the above modifications to the tail group of 4 abrogated
detectable agonism in the pERK1/2 assay, and either maintained or reduced agonist efficacy
in the cAMP assay. Modification to the N,Nꢀdimethylurea (41) which represents the tail
group of cariprazine (2), caused no significant change in functional affinity or efficacy in the
cAMP assay, indicating that 41 has equivalent activity to 4 in this functional endpoint
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). As this compound showed no agonist activity in the pERK1/2 assay, and is
therefore less active than 4 at this functional endpoint, this modification confers bias towards
the cAMP pathway.
We next focused on the role of the spacer group of 4 (Table 2). Modification from the
trans to the cis stereoisomer (45) resulted in a similar binding affinity relative to 4, but a
complete loss of agonist efficacy in the pERK1/2 assay. Furthermore, in the cAMP assay this
modification resulted in a significant 8ꢀfold increase in functional affinity (pK ) but no
A
change in efficacy (τ). Overall this resulted in a significant increase in the transduction
coefficient in the cAMP assay. Given that 45 was inactive in the pERK1/2 assay, this clearly
demonstrates that this modification caused an increase in bias towards the cAMP pathway
(Figure 4). This effect was not observed with corresponding compounds in which the tertꢀ
butyl carbamate tail group was replaced with N,Nꢀdimethylurea. Indeed, with the N,Nꢀ
dimethylurea tail group, modification from trans (41) to cis stereoisomer (54) resulted in no
significant change in all parameters relating to the cAMP assay (Student’s tꢀtest, p > 0.05)
and 54 maintained no agonist efficacy in the pERK1/2 assay, although a significant loss of
binding affinity (pK ) was observed as compared to 41. We then explored further subtle
i
modifications to the spacer group of 4. Truncation of the spacer group by 1 carbon (43)
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
caused no significant change in overall bias as compared to 4. However, rather than the
transduction coefficients remaining unchanged at both pathways, a significant 5ꢀfold increase
in the transduction coefficient at the cAMP assay was cancelled out by a significant 5ꢀfold
increase in efficacy (τ) observed in the pERK1/2 assay. Shifting the spacer group one carbon
towards the THIQ head group (44) had no significant effect upon binding affinity (pKi) as
compared to 4. In the pERK1/2 assay, this compound behaved as a very weak partial agonist.
Consequently estimates of functional affinity in this assay were associated with significant
error and, as such, changes in bias as compared to 4 were not statistically significant. Finally,
incorporation of aromaticity to the spacer group (46) resulted in similar binding affinity as 4,
but with a significant decrease of transduction coefficient in the cAMP assay, and no
detectable response in the pERK1/2 assay.
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
Next we explored modification of the THIQ head group (Table 3 and 4). Stemp et al.
demonstrated that deactivation of the THIQ ring through 6ꢀ or 7ꢀsubtitutions has little impact
15
on activity at D ꢀlike receptors, and we sought to explore this effect with a focus on agonist
2
efficacy at the D R. Due to synthetic accessibility and commercial availability, we explored
2L
THIQ deactivating groups in the 7ꢀposition. Incorporation of the 7ꢀnitro substituent (29)
resulted in a loss of agonism in both assays, with no change in binding affinity. Modification
to the nitrile substituent (30), which shares structural similarities with SB269652 (3), resulted
in a loss of agonist efficacy in the pERK1/2 assay, and a significant 4ꢀfold decrease in
efficacy in the cAMP assay, with no significant change in binding affinity. Of interest, the 7ꢀ
cyano derivative with the acetamide tail group (36) displayed a significant 32ꢀfold higher
binding affinity than 4, albeit with loss of agonism at both signaling endpoints measured.
Furthermore, direct comparison with 35 (Table 1) reveals that the addition of a cyano group
at position 7 confers a 260ꢀfold increase in the affinity determined by radioligand binding.
10
ACS Paragon Plus Environment

Page 11 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
These results infer that THIQ ring deactivation can have significant effects on binding
affinity, but consistently have a detrimental effect upon agonism.
Mindful of the structural similarities of 4 to cariprazine (2), we next explored modification
of the THIQ head group to 2,3ꢀDCPP (Table 4). As a general observation, this modification
resulted in higher binding affinity than equivalent compounds with the THIQ head group
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
(
Tables 1ꢀ4). Direct replacement of the THIQ moiety of 4 for 2,3ꢀDCPP (32) resulted in a
significantly (19ꢀfold) increased affinity at the D2LR as determined by radioligand binding
Student’s tꢀtest P < 0.05). In contrast, values of τ or pK determined in both pERK1/2 and
(
A
cAMP assays did not differ significantly from that of 4 (Student’s tꢀtest P > 0.05).
Introduction of the N,Nꢀdimethylurea tail group (cariprazine, 2) did not change binding
affinity relative to 32 (Table 4). However, while pK
A
or τ did not change significantly in the
= 8.71) was
pERK1/2 assay, a significant 30ꢀfold increase in functional affinity (pK
A
observed in the cAMP assay relative to that of 32 (Figure 5). This conferred a large increase
in bias towards the cAMP pathway relative to 32, 4 and dopamine (42ꢀfold, 68ꢀfold and 218ꢀ
fold respectively). Of interest, a similar pattern was observed in the corresponding
compounds containing the THIQ core (4 vs 41, Figure 3) although in this case, a lack of
detectable agonism in the pERK1/2 assay for 41 meant that the degree of bias conferred by
this modification could not be quantified.
We then focused on the role of the spacer group with 2,3ꢀDCPP head group in place.
Modification of the spacer configuration from trans (32) to cis stereoisomer (49) resulted in
similar binding affinity to 32, as was observed for the corresponding compounds in the THIQ
series (trans; 4, cis; 45). However, unlike the increase in bias observed for 45 as compared to
4, no change in bias was observed for 49 relative to 32. Modification of cariprazine (2) to the
cis stereoisomer (56) caused a loss of agonism in the pERK1/2 assay. However this was also
accompanied by both a significant (50ꢀfold) decrease in functional affinity (pK ) and a
A
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
significant 8ꢀfold decrease in efficacy (τ) in the cAMP assay relative to cariprazine (2)
(Student’s tꢀtest, P < 0.05). As such, it is not clear whether this chemical modification
influences bias. Further subtle modifications to the spacer group of 32 were then investigated.
Truncation of the spacer group of 32 by one carbon (47) caused a significant 8ꢀfold decrease
in affinity measured by radioligand binding. However, this modification caused markedly
different effects upon efficacy and functional affinity in the two different assays. In 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
A
pERK1/2 assay a significant 120ꢀfold decrease in functional affinity (pK ) was accompanied
by a significant 10ꢀfold increase in τ. In the cAMP assay no significant change in either
parameter was observed. Consequently this modification confers 6ꢀfold bias towards the
cAMP pathway relative to 32. Shifting the spacer group of 32 by 1 carbon towards the 2,3ꢀ
DCPP head group (48) caused no significant change in bias, although this modification did
i
cause a significant 9ꢀfold loss of affinity (pK ). In contrast, shifting the spacer group of
cariprazine (2) towards the 2,3ꢀDCPP head group (55) caused a significant 84ꢀfold decrease
in bias towards the cAMP pathway (Student’s tꢀtest, p > 0.05) and as such a bias profile not
significantly different to that of 32. This effect was predominantly mediated by a significant
A
61ꢀfold decrease in functional affinity (pK ) in the cAMP assay (Student’s tꢀtest, p < 0.05).
However significant 3ꢀfold and 4ꢀfold decreases in intrinsic efficacy (τ) were observed in the
pERK1/2 and cAMP assays respectively (Student’s tꢀtest, p < 0.05). Ring expansion to 2,3ꢀ
DCPHP (31) yielded the compound with the highest binding affinity of the series but with
loss of efficacy in both assays.
Compound 4 and cariprazine (2) display significantly different (60ꢀfold) bias profiles
between an assay measuring cAMP, and that measuring phosphorylation of ERK1/2, with
cariprazine demonstrating 230ꢀfold bias towards the cAMP pathway as compared to
dopamine. Although both compounds have a conserved spacer group they differ in their head
group (4; THIQ, 2; 2,3ꢀDCPP) and in their tail group (4; tertꢀbutyl carbamate, 2; N,Nꢀ
12
ACS Paragon Plus Environment

Page 13 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
dimethylurea). Exchange of the tertꢀbutyl carbamate for N,Nꢀdimethylurea conferred bias
towards the cAMP pathway to the THIQ derivative (41). The opposite modification
decreased bias of the 2,3ꢀDCPP derivative (32). This observation is consistent with N,Nꢀ
dimethylurea substitution of the tail group conferring bias towards cAMP. This raises the
possibility that the tail group extends in to, and interacts with a secondary (or allosteric)
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
pocket. Indeed, a dualsteric or bitopic (orthosteric/allosteric) mode of interaction has been
22,23
suggested to underlie the biased actions of a number of GPCR ligands.
However, this
pattern becomes more complex as we consider further derivatives of these ligands whereby
conserved structural modifications have a differential impact upon biased agonism depending
upon the head group. Focusing upon compounds with the tertꢀbutyl carbamate head group, a
comparison of trans (4; THIQ, 32; 2,3ꢀDCPP) with the corresponding cis stereoisomer
revealed that this modification confers bias towards the cAMP pathway for the compound
containing the THIQ head group (45) but not for the compound containing the 2,3ꢀDCPP
head group (49). Conversely truncation of the spacer region of 4 and 32 by one carbon atom
confers bias towards the cAMP pathway for the compound containing the 2,3ꢀDCPP head
group (47) but not for the corresponding compound with a THIQ head group (43).
Furthermore, comparison of 43 and 47 revealed that exchange of the THIQ head group for
the 2,3ꢀDCPP head group conferred bias towards cAMP underpinned by a large decrease in
functional affinity in the pERK1/2 assay (Figure 6). This is distinct from the lack of bias
observed between 4 and 32, which differ only by an additional carbon within the spacer
group. This then suggests that the nature and relative orientation of both the head and tail
group can impact the bias profile of this series of compounds. It should be noted that the
structural changes that can engender significant changes in bias within this study are often
relatively subtle. Such an observation is not surprising if one considers the relatively modest
changes in the orthosteric binding site of active agonist bound receptor crystal structures, for
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
example those of the β
2
ꢀadrenoceptor, when compared to the corresponding antagonist bound
24ꢀ27
structures.
Such minor movements are coupled to major structural movements at the
cytosolic face of the receptor, allowing signalling proteins such as G proteins to bind. Indeed,
28
the structural difference between GPCR agonists and antagonists can also be subtle.
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
Accordingly, we can expect that structural changes that determine differential agonist action
at one pathway as compared to another will also be subtle. As such the SAR around bias
represents a challenge, but the use of the operational model to identify and quantify changes
in bias profile and relate them to such subtle changes is a valuable approach to meet this
challenge.
In view of the therapeutic importance of the D R, it is not surprising that a number of studies
2
11,29ꢀ34
have explored the molecular determinants of efficacy at this receptor.
Notably, Allen et
al. applied a combinatorial approach to explore the SAR around aripiprazole (1), and
1
1
identified β arrestinꢀbiased agonists based on a similar scaffold. Another study investigated
determinants of efficacy (for activation of Gαi/o proteins) at the D R for a series of
3
3
5
compounds containing the phenylpiperazine moiety. Newman et al. demonstrated a
progressive decrease in efficacy of such compounds with the incremental addition of
methylene units to the phenylpiperazine moiety in a series of 2,3ꢀDCPP derivatives. In
contrast similar modifications to a series of 2ꢀmethoxyphenylpiperazine derivatives were
weak partial agonists. Using molecular docking and modeling approaches, the authors
suggested that as the linker length increased in the 2,3ꢀDCPP series, the orientation of the
2,3ꢀDCPP moiety became similar to that in the 2ꢀmethoxyphenylpiperazine series and could
no longer form hydrogen bonding interactions with the conserved serines within TM5.
We performed molecular docking experiments to gain insight into the binding mode of
such biased agonists at the D R. Following molecular dynamics simulations over 200 ps, we
2
14
ACS Paragon Plus Environment

Page 15 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compared the final binding mode of 32 and 2 (Figure 7A); compounds that displayed 60ꢀfold
differentiation in their bias profiles yet differ only in the nature of the tail group. The
predicted binding modes illustrate that while the tail group and spacer regions adopt a similar
position, the 2,3ꢀDCPP moiety adopts a divergent pose (Figure 7B). This in turn results in
distinct differences in receptor structure between the two models with noticeable variation in
the orientations of conserved serines in TM5 (Ser5.42, Ser5.43 & Ser5.46), Phe5.47 and
Tyr7.35. In particular, a Hꢀbond interaction between Ser5.46 and Tyr3.37 in the ligandꢀ
receptor complex of 32 is absent in the corresponding ligandꢀreceptor complex of 2 (Figure
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
7A). It should be noted that neither compound forms Hꢀbond interactions with the TM5
35
serines in agreement with the study by Newman et al. Comparison of trans and cis
stereoisomers with the THIQ head group (4 and 45), respectively, in equivalent dynamics
simulations revealed significant divergence in both the orientation of the head and tail groups
(Figure 7C & D). These isomers display distinct bias profiles, and it is interesting to note the
distinct orientation of residues within TM5 and TM6 between the two complexes. In
particular, we again observe a large movement of Phe5.47, a large movement of Phe6.51 and,
notably, the different position of Trp6.48 between the two complexes. With both examples, it
appears that subtle structural differences within the tail group or spacer region cause distinct
changes in the position of the head group within the orthosteric pocket and a concomitant
change of conformation within this binding pocket. This in turn underlies the bias profiles of
such compounds. Although such modeling and docking experiments are intrinsically
associated with a degree of speculation it is interesting to note that these results are consistent
35
with those of Newman et al. A study by Tschammer et al. also focused on 1,4 disubstituted
3
2
2
phenylpiperazine derivatives at the D R. Within this study it was observed that rigidly
constraining the methoxy group of the 2ꢀmethoxyphenylpiperazine head group through
derivatization to the sterically demanding (2,3ꢀdihydrobenzofuranꢀ7ꢀyl)piperazine abolished
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the activity of this compound in a cAMP assay whilst maintaining partial agonist activity in a
32
pERK1/2 assay. Such a finding is consistent with bias being driven through different
orientations of the head group within the orthosteric pocket.
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
To our knowledge this is the first study to apply a quantitative pharmacological approach
using the operational model of agonism to enrich a comprehensive SAR study focused upon
biased agonism at the D
2
R. A number of methods to quantify bias have been proposed and
6
published in recent times with particular attention given to the relative merits of the
transduction ratio method used in this study, as compared to the sigma method proposed by
20
Rajagopal et al. Although both methods use the operational model of agonism the latter
fixes the values of functional affinity (K ) to that determined by radioligand binding (K ),
whereas the method used in our study derives the K directly from the fitting of the doseꢀ
response curves to the operational model of agonism. Thus this operationally derived K
A
i
A
A
represents the functional affinity of an agonist associated with that particular signaling
pathway. Of note, the affinity determined in our radioligand binding experiments is
significantly different to that determined in one or both of our functional assays 11 out of 26
times (42%) (Tables 1, 2 and 4). In 8 cases this difference is greater than 10ꢀfold, and for 54,
a difference of more than 100ꢀfold is observed. Furthermore, there is no pattern to the
distribution of the values of pK that differ significantly from the corresponding value of pK ,
A
i
with examples in both functional assays. Indeed, we obtained values of pK both smaller and
A
i A
greater than the corresponding pK value. As such fixing the K to that determined in a
radioligandꢀbinding assay would introduce significant error into estimates of transduction
coefficients, and ultimately identification of ligand bias. Furthermore, it should be noted that
the pK determined for a ligand in a radioligand binding assay can depend on the
i
radiolabelled probe (agonist versus antagonist) used or the experimental design (components
16
ACS Paragon Plus Environment

Page 17 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of buffer, whole cells versus membranes, the presence of guanine nucleotide). Taking
dopamine as an example, a study performed using CHO cell membranes expressing the D2L
R
i i i
determined a pK _high of 58 nM, a pK _low of 3420 nM and a pK determined using the agonist
3
36
[
H]NPA as the radioligand as 8 nM. This begs the question, if one were to fix the value of
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
pK to a value of pK determined by binding affinity, which would be most appropriate? We
A
i
observed no pattern in the way in which the functional affinities determined for the agonists
within this series deviated from the corresponding pK values. Thus we conclude that no
i
value of affinity determined in a binding assay will be a good predictor of functional affinity.
Although significant differences in τ were observed across the chemical series, it is
interesting to note that in all cases in which a chemical modification engenders a significant
change in bias, this change is predominantly driven by a change in functional affinity for one
assay (change in pKA_cAMP 32 vs 2, 40ꢀfold; 55 vs 2, 60ꢀfold: change in pKA_pERK1/2: 43 vs 47;
112ꢀfold; 32 vs 47, 120ꢀfold). This is perhaps not surprising given that 4, cariprazine (2), and
derivatives thereof, behave as partial agonists and thus have relatively small values of τ. Thus
changes in τ will have a much smaller impact upon the overall transduction coefficient
A
(log(τ/K )) as compared to changes in functional affinity for low efficacy agonists. Apart
from the identification and quantification of biased agonism, another key challenge is relating
the bias profile of a ligand to a physiological response in vivo. As such, the identification of
pairs of ligands that are structurally similar and have similar binding affinity for the receptor
target, yet display distinct bias profiles (such as compounds 2 and 32), may prove to be useful
tools to investigate the therapeutic relevance of biased agonism.
Conclusion
Biased agonism represents an attractive paradigm that can be exploited in the design of novel
selective GPCR ligands. However, the identification and quantification of bias remains a
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
challenge. In this study, we have addressed this challenge by applying detailed
pharmacological profiling to a series of compounds based on the hit compound tertꢀbutyl
(transꢀ4ꢀ(2ꢀ(3,4ꢀdihydroisoquinolinꢀ2(1H)ꢀyl)ethyl)cyclohexyl)carbamate (4), a D2LR partial
agonist with structural similarity to cariprazine (2). We combined this approach with
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
analytical methods to allow us to identify and quantify biased agonism. We demonstrated that
4
displayed no significant bias between the two signaling endpoints; ERK1/2 phosphorylation
and inhibition of forskolinꢀstimulated cAMP production, as compared to dopamine. In
contrast, cariprazine (2), awaiting FDA approval for the treatment of schizophrenia,
displayed greater than 200ꢀfold bias towards the cAMP pathway. Taking advantage of these
distinct bias profiles, we explored the SAR of biased agonism around these two compounds.
We discovered that the nature of the head group, composition of the tail group, orientation,
length and flexibility of the spacer are all important factors for the control of biased agonism
at the D2LR. Our molecular modeling studies suggest that subtle changes in the spacer and tail
region can influence the orientation of the head group within the orthosteric pocket and these
distinct orientations may underlie patterns of biased agonism. Using the transduction
coefficient method, we demonstrated that the functional affinities of such ligands can differ
significantly from those determined in radioligand binding, and that instances of significant
changes in bias were driven largely by changes in the functional affinity of a ligand at a
particular pathway. In conclusion, this approach has provided an unprecedented insight into
the molecular determinants and SAR of biased agonism at the D R. Given the clinical
2
L
relevance of biased agonism at the D2LR, this study may serve as the precursor to the
development of novel biased ligands for this therapeutically important target.
18
ACS Paragon Plus Environment

Page 19 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Experimental
General Experimental
All reagents were purchased from SigmaꢀAldrich, Alfa Aesar, AK Scientific or
ChemImpex, and used without purification. GR grade ammonium hydroxide solution (28%
aqueous solution), and LR grade methanol, ethyl acetate, chloroform, DCM and acetonitrile
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
13
were purchased from Merck and used without further purification. All H NMR and
C
NMR spectra (DEPTQ) were recorded on a Bruker Avance III 400 Ultrashield Plus
spectrometer at 400.13 and 100.62 MHz respectively. Results were recorded as follows:
chemical shift values are expressed as δ units generally acquired in CDCl
3
; or CD
3
2
OD, D O
1
or d ꢀDMSO where specified, with tetramethylsilane (0.00 ppm) as reference for H NMR
6
13
37
(
residual solvent peak as reference for C NMR), multiplicity (singlet (s), doublet (d),
triplet (t), quartet (q), broad (br), multiplet (m), doublet of doublets (dd), doublet of triplets
(dt), triplet of triplets (tt), quartet of doublets (qd)), coupling constants (J) in Hertz and
integration. Thinꢀlayer chromatography was conducted on 0.2 mm plates using Merck silica
gel 60 F . Flash Chromatography was performed using Merck Silica Gel 60, 230ꢀ400 mesh
2
54
ASTM. High resolution mass spectra (HRMS) were obtained on a Waters LCT Premier XE
TOF) using electrospray ionization (ESI) at a cone voltage of 50 V. LCMS data was
(
obtained on an Agilent 1200 series LC coupled directly to a photodiode array detector and an
Agilent 6100 Quadrupole MS, using a Phenomenex® column (Luna 5 ꢂm C8, 50 mm × 4.60
mm ID). Analytical reverseꢀphase HPLC was performed on a Waters HPLC system coupled
directly to a photodiode array detector and fitted with a Phenomenex® Luna C8 (2) 100 Å
column (150 mm × 4.6 mm, 5 ꢂm) using a binary solvent system; solvent A: 0.1% TFA/H
2
O;
solvent B: 0.1% TFA/80% CH CN/H O. Gradient elution was achieved using 100% solvent
3
2
A to 100% solvent B over 20 min at a flow rate of 1 mL/min. All compounds were >95 %
purity by HPLC (λ = 254, 214 nm) prior to biological testing. Melting point analysis was
performed in duplicate on a Mettler Toledo MP50 Melting Point System.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
8
Ethyl 2ꢀ(transꢀ4ꢀaminocyclohexyl)acetate hydrochloride (8). Following an adapted
38
literature procedure, 10% Pd/C (881 mg, 828 ꢃmol) was carefully added to an orange
suspension of 5 (5.00 g, 27.6 mmol) in H O (150 mL). The reaction mixture was
2
hydrogenated on a Parr shaker at 60 psi at RT for 3 d, until the uptake of hydrogen was
complete, and no starting materials remained by TLC (CHCl /CH OH, 1:1). The mixture was
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
3
3
filtered through a Celite™ pad, washed with water (30 mL), and the filtrate evaporated to
dryness in vacuo to reveal a white solid. The material was taken up in absolute EtOH (70
mL) to which concentrated HCl (10 mL) was added and the mixture heated at reflux for 2 h.
TLC confirmed ethyl ester formation and the solvents were concentrated in vacuo. The
material was basified with 1 M NaOH solution to pH 14, and a white precipitate emerged.
The product was then extracted from the mixture with EtOAc (3 × 30 mL), and the combined
organic extracts washed with brine, then dried over anhydrous Na
2 4
SO . The product was then
converted to the HCl salt by the addition of 1 M HCl in Et O (27.6 mL, 27.6 mmol), and the
2
solvents concentrated to half volume in vacuo. The solution was then cooled to 0 °C, which
resulted in fractional crystallisation of the trans stereoisomer as a white solid which was then
39
3
collected by filtration and washed with cold CH CN (1.34 g, 22%). mp: 164–166 °C (lit.
1
1
62–163 °C). H NMR (MeOD) δ 4.11 (q, 2H, J = 7.1 Hz), 3.05 (tt, 1H, J = 11.8, 3.9 Hz),
2
.24 (d, 2H, J = 7.0 Hz), 2.11 – 2.00 (m, 2H), 1.93 – 1.83 (m, 2H), 1.83 – 1.68 (m, 1H), 1.43
1
3
(qd, 2H, J = 12.8, 3.6 Hz), 1.24 (t, 3H, J = 7.1 Hz), 1.14 (qd, 2H, J = 13.3, 3.3 Hz). C NMR
(
CD OD) δ 174.2 (C), 61.4 (CH ), 51.2 (CH), 41.8 (CH ), 34.7 (CH), 31.50 (CH ), 31.47
3
2
2
2
(
CH
2
3
), 14.6 (CH ).
4
0
transꢀEthyl 4ꢀaminocyclohexanecarboxylate (9). transꢀ4ꢀAminocyclohexanecarboxylic
acid hydrochloride (6, 250 mg, 1.39 mmol) was taken up in absolute EtOH (10 mL) and
concentrated HCl (2 mL), and the pale orange solution was heated at reflux for 3 d. The
solution was concentrated in vacuo, diluted with water (10 mL), and made alkaline with
20
ACS Paragon Plus Environment

Page 21 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH
4
OH solution to pH 10. The product was extracted with EtOAc (2 × 30 mL) and the
2 4
combined organic extracts washed with brine (15 mL), dried over anhydrous Na SO and
evaporated to dryness to give the product as a pale orange oil which required no further
1
purification (194 mg, 81%). H NMR δ 4.11 (q, J = 7.1 Hz, 2H), 2.66 (tt, J = 11.0, 3.9 Hz,
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
1H), 2.21 (tt, J = 12.3, 3.6 Hz, 1H), 2.03–1.85 (m, 4H), 1.56–1.39 (m, 4H), 1.25 (t, J = 7.1
13
2
Hz, 3H), 1.11 (qd, J = 13.0, 3.2 Hz, 2H). C NMR δ 176.0 (C), 60.3 (CH ), 50.0 (CH), 42.8
2 2 3
(CH), 35.8 (CH ), 28.1 (CH ), 14.3 (CH ).
41
transꢀEthyl 4ꢀ(aminomethyl)cyclohexanecarboxylate (10). Tranexamic acid (7, 1.00 g,
6.36 mmol) was taken up in absolute EtOH (30 mL) and concentrated HCl (5 mL). The
colourless solution was heated at reflux for 2 h. The solution was then cooled to 0 °C on ice,
then basified with 1 M NaOH solution (30 mL) to pH 12. The product was extracted with
EtOAc (3 × 20 mL), the combined organic extracts washed with brine (20 mL), dried over
anhydrous Na
2
SO
4
and evaporated to dryness to reveal the product as a colourless oil which
1
solidified on drying (775 mg, 66%). H NMR δ 4.11 (q, J = 7.1 Hz, 2H), 3.32 (s, 2H), 2.61
(d, J = 6.5 Hz, 2H), 2.22 (tt, J = 12.2, 3.4 Hz, 1H), 2.07–1.95 (m, 2H), 1.95–1.82 (m, 2H),
13
1
.51–1.33 (m, 3H), 1.25 (t, J = 7.1 Hz, 3H), 0.97 (qd, J = 13.0, 3.1 Hz, 2H). C NMR δ
2 2 2 2 3
176.0 (C), 60.2 (CH ), 47.9 (CH ), 43.5 (CH), 39.4 (CH), 29.8 (CH ) 28.6 (CH ), 14.3 (CH ).
General Procedure A (tertꢀButyl Carbamate Protection of Primary Amine)
The amine (occasionally as the HCl salt, 1–1.2 equiv) was taken up in DCM (15–30 mL),
and Et N (1.2 equiv, or 2.4 equiv if using hydrochloride salt) added. To the stirred solution at
3
RT was added a solution of diꢀtertꢀbutyl dicarbonate (1 equiv) in DCM (5 mL). The solution
was stirred for 2–24 h, then diluted with DCM (20 mL), washed with 1 M KHSO
4
(2 × 20
mL), and brine (10 mL). The organic layer was dried over anhydrous Na SO and evaporated
2
4
to dryness in vacuo to reveal the title compound, which was purified by flash column
chromatography if required.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
2
Ethyl 2ꢀ(transꢀ4ꢀ((tertꢀbutoxycarbonyl)amino)cyclohexyl)acetate (12). Using 8 (682 mg,
3
.07 mmol) as the starting material, following General Procedure A, gave the product as
white needles (746 mg, 94%). Determination of diastereomeric purity (> 95% trans) was
1
achieved by HꢀNMR analysis. The trans stereoisomer (12) exhibited a characteristic
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
resonance at δ 2.18 ppm, whilst the cis stereoisomer (15) exhibited the equivalent resonance
1
at δ 2.24 ppm. H NMR δ 4.52 (br s, 1H), 4.12 (q, J = 7.1 Hz, 2H), 3.37 (br s, 1H), 2.18 (d, J
=
6.9 Hz, 2H), 2.04–1.95 (m, 2H), 1.84–1.66 (m, 3H), 1.43 (s, 9H), 1.25 (t, J = 7.1 Hz, 3H),
13
1
.20–1.01 (m, 4H). C NMR δ 172.8 (C), 155.2 (C), 78.9 (C), 60.1 (CH
2
), 49.4 (CH), 41.4
(
CH ), 33.4 (CH), 33.1 (CH ), 31.5 (CH ), 28.4 (CH ), 14.2 (CH ).
2
2
2
3
3
43
transꢀEthyl 4ꢀ((tertꢀbutoxycarbonyl)amino)cyclohexanecarboxylate (13). Using 9 (190
mg, 1.11 mmol) as the starting material, following General Procedure A. The product was
purified by flash column chromatography (Petroleum spirits/EtOAc, 10:1) to give the title
1
compound as white needles (174 mg, 58%). H NMR δ 4.40 (br s, 1H), 4.11 (q, J = 7.1 Hz,
2H), 3.41 (br s, 1H), 2.20 (tt, J = 12.1, 3.5 Hz, 1H), 2.12–1.93 (m, 4H), 1.52 (qd, J = 12.6, 3.2
13
Hz, 2H), 1.44 (s, 9H), 1.24 (t, J = 7.1 Hz, 3H), 1.11 (qd, J = 12.7, 3.3 Hz, 2H). C NMR δ
1
75.6 (C), 155.3 (C), 79.4 (C), 60.4 (CH
7.9 (CH ), 14.3 (CH ).
transꢀEthyl 4ꢀ(((tertꢀbutoxycarbonyl)amino)methyl)cyclohexanecarboxylate (14). Using 10
2 2 3
), 49.1 (CH), 42.6 (CH), 32.7 (CH ), 28.5 (CH ),
2
2
3
(713 mg, 3.85 mmol) as the starting material, following General Procedure A, gave the title
1
compound as a pale yellow oil (880 mg, 96%). H NMR δ 4.63 (br s, 1H), 4.11 (q, J = 7.1
Hz, 2H), 2.98 (t, J = 6.4 Hz, 2H), 2.21 (tt, J = 12.2, 3.6 Hz, 1H), 2.06–1.92 (m, 2H), 1.89–
1
.74 (m, 2H), 1.50–1.33 (m, 12H), 1.24 (t, J = 7.1 Hz, 3H), 0.96 (qd, J = 13.1, 3.4 Hz, 2H).
1
3
C NMR δ 176.0 (C), 156.2 (C), 79.2 (C), 60.3 (CH
2 2
), 46.7 (CH ), 43.4 (CH), 37.9 (CH),
2
9.8 (CH ), 28.6 (CH ), 28.5 (CH ), 14.3 (CH ).
2
2
3
3
22
ACS Paragon Plus Environment

Page 23 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Ethyl 2ꢀ(cisꢀ4ꢀ((tertꢀbutoxycarbonyl)amino)cyclohexyl)acetate (15). To a solution of
commercially available 2ꢀ(cisꢀ4ꢀ((tertꢀbutoxycarbonyl)amino)cyclohexyl)acetic acid (11, 500
mg, 1.94 mmol) in DCM (10 mL) at RT was added 1ꢀethylꢀ3ꢀ(3ꢀ
dimethylaminopropyl)carbodiimide hydrochloride (410 mg, 2.14 mmol), followed by a 4ꢀ
(dimethylamino)pyridine (5 mol%, 12 mg, 97.2 ꢃmol). After 15 min, absolute EtOH (15 mL)
was added, and the mixture stirred overnight. TLC confirmed reaction completion (petroleum
spirits/EtOAc, 3:1) and the mixture was concentrated in vacuo, then taken up in EtOAc (30
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
mL) causing a precipitate to emerge. The mixture was then washed with 1 M KHSO
4
(2 × 20
mL), brine (20 mL), dried over anhydrous Na SO and evaporated to dryness to reveal a
2
4
1
colourless oil (354 mg, 64%) which required no further purification. H NMR δ 4.65 (br s,
1
1
7
H), 4.13 (q, J = 7.1 Hz, 2H), 3.72 (s, 1H), 2.24 (d, J = 7.3 Hz, 2H), 2.00–1.84 (m, 1H),
13
.70–1.53 (m, 6H), 1.50–1.37 (m, 9H), 1.33–1.16 (m, 5H). C NMR δ 172.8 (C), 155.2 (C),
2 2 2 3 2
9.0 (C), 60.2 (CH ), 46.2 (CH), 40.4 (CH ), 32.8 (CH), 29.5 (CH ), 28.4 (CH ), 27.6 (CH ),
14.3 (CH3).
General Procedure B (Reduction of Ester to Aldehyde)
The ethyl ester (1 equiv) was taken up in toluene (15ꢀ30 mL), degassed with nitrogen
bubbling for 15 min, then cooled to ꢀ78 °C on a dry ice/acetone bath for a further 10 min. To
the stirring colourless solution under nitrogen, was slowly added DIBALH (1 M in toluene, 2
equiv) dropwise over 15 min. The mixture stirred at ꢀ78 °C until foaming of the reaction
mixture stopped (30–60 min). The mixture was then quenched with CH OH (10–20 mL) in
3
toluene (10 mL), and warmed to RT with stirring for 15 min. Saturated potassium sodium
tartrate solution (30 mL) was added and the mixture stirred vigorously for 30 min. The
product was then extracted with Et
2
O (3 × 30 mL), and the combined organic extracts dried
over anhydrous Na SO and evaporated to dryness to give the crude compound. If required,
2
4
the product was purified by flash column chromatography.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
5
tertꢀButyl (transꢀ4ꢀ(2ꢀoxoethyl)cyclohexyl)carbamate (16). Using 12 (1.25 g, 4.38 mmol)
as the starting material, following General Procedure B, the material was purified by column
chromatography (petroleum spirits/EtOAc, gradient 6:1 to 4:1) gave the title compound as a
1
5
1
white wax (944 mg, 89%, lit. 53%). H NMR δ 9.75 (t, J = 2.0 Hz, 1H), 4.47 (br s, 1H),
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
3.37 (br s, 1H), 2.32 (dd, J = 6.6, 2.0 Hz, 2H), 2.04–1.97 (m, 2H), 1.89–1.75 (m, 3H), 1.45 (s,
1
3
9
H), 1.21–1.03 (m, 4H). C NMR δ 202.2 (CH), 155.3 (C), 79.2 (C), 50.7 (CH
3.2 (CH ), 31.8 (CH ), 31.7 (CH), 28.5 (CH ).
tertꢀButyl (transꢀ4ꢀformylcyclohexyl)carbamate (17). Using 13 (160 mg, 590 ꢃmol) as the
2
), 49.5 (CH),
3
2
2
3
44
starting material, following Procedure B, gave the title compound as a white wax (134 mg,
1
99%). H NMR δ 9.62 (s, 1H), 4.44 (br s, 1H), 3.40 (br s, 1H), 2.21–1.97 (m, 5H), 1.51–1.27
1
3
(
m, 11H), 1.22–1.10 (m, 2H). C NMR δ 203.9 (CH), 155.5 (C), 81.6 (C), 49.4 (CH), 32.2
CH ), 28.5 (CH ), 24.9 (CH ), 19.5 (CH).
tertꢀButyl ((transꢀ4ꢀformylcyclohexyl)methyl)carbamate (18). Using 14 (850 mg, 2.98
(
2
3
2
45
mmol) as the starting material, following Procedure B. The crude material was used
immediately in subsequent reactions to avoid rapid degradation. To confirm aldehyde
formation the crude material (150 mg) was taken up in absolute ethanol (15 mL), and was
2 4
added a solution of 2,4ꢀDNP (50 mg) in EtOH (5 mL) and concentrated H SO (1 mL). The
solution was heated at reflux for 1 h, then allowed to cool revealing a yellow precipitate. The
product was collected by filtration and washed with ethanol to give (transꢀ4ꢀ((2ꢀ(2,4ꢀ
dinitrophenyl)hydrazono)methyl)cyclohexyl)methanaminium sulfate as a yellow solid.
+
Formation of the 2,4ꢀDNP derivative was confirmed by LCMS. ESIꢀMS (m/z): 322.2 (MH ).
46
tertꢀButyl (cisꢀ4ꢀ(2ꢀoxoethyl)cyclohexyl)carbamate (19). Using 15 (400 mg, 1.40 mmol)
as the starting material, following Procedure B, gave the title compound as a pale yellow wax
1
(
325 mg, 96%). H NMR δ 9.76 (t, J = 2.1 Hz, 1H), 4.63 (s, 1H), 3.78–3.64 (m, 1H), 2.37
dd, J = 6.9, 2.1 Hz, 2H), 2.10–1.95 (m, 1H), 1.68–1.58 (m, 6H), 1.44 (s, 9H), 1.35–1.19 (m,
(
24
ACS Paragon Plus Environment

Page 25 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
3
2H). C NMR δ 202.3 (CH), 155.7 (C), 79.3 (C), 49.7 (CH
CH ), 28.6 (CH ), 28.0 (CH ).
Ethyl 2ꢀ(4ꢀaminophenyl)acetate (20). 4ꢀNitrophenylacetic acid (5, 1.00 g, 5.52 mmol)
2
), 46.3 (CH), 30.6 (CH), 29.6
(
2
3
2
47
was taken up in EtOH (20 mL) and concentrated HCl (2 mL). The solution was stirred at
reflux for 2 h. The mixture was left open to evaporate slowly over 3 d causing crystallisation
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
of the product, which was filtered and washed with 5% NH OH solution (30 mL). The
crystals were then dried overnight, then taken up in absolute EtOH (30 mL) and concentrated
HCl (5 mL), and to the pale yellow solution was added finely granulated tin (2.98 g, 25.1
mmol). The mixture was heated at reflux for 16 h, then filtered to remove any remaining tin,
then concentrated in vacuo. Water (20 mL) and NH OH solution were added until the
4
mixture became alkaline, which caused a white precipitate to form. The product was
extracted with EtOAc (3 × 30 mL) and the combined organic extracts washed with brine (20
2 4
mL), dried over anhydrous Na SO , filtered through a sinter funnel to remove any remaining
precipitates, then evaporated to dryness to give the product as a pale yellow oil (764 mg,
1
8
2
1
5%). H NMR δ 7.09–7.02 (m, 2H), 6.66–6.59 (m, 2H), 4.12 (q, J = 7.1 Hz, 2H), 3.58 (br s,
13
H), 3.48 (s, 2H), 1.24 (t, J = 7.1 Hz, 3H). C NMR δ 172.3 (C), 145.5 (C), 130.2 (CH),
24.1 (C), 115.3 (CH), 60.8 (CH
2 2 3
), 40.7 (CH ), 14.3 (CH ).
48
Ethyl 2ꢀ(4ꢀ((tertꢀbutoxycarbonyl)amino)phenyl)acetate (21). Using 20 (700 mg, 3.91
mmol) as the starting material, following General Procedure A. The product was purified by
flash column chromatography (petroleum spirits/EtOAc, 6:1) to give the title compound as a
1
colourless oil which slowly crystallised (833 mg, 76%). H NMR δ 7.31 (d, J = 8.4 Hz, 2H),
7
.20 (d, J = 8.5 Hz, 2H), 6.50 (br s, 1H) 4.13 (q, J = 7.1 Hz, 2H), 3.55 (s, 2H), 1.51 (s, 9H),
1
3
1.24 (t, J = 7.1 Hz, 3H). C NMR δ 171.2 (C), 151.4 (C), 137.5 (C), 129.9 (CH), 128.8 (C),
18.8 (CH), 80.9 (C), 61.0 (CH ), 40.9 (CH ), 28.5 (CH ), 14.3 (CH ).
1
2
2
3
3
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
5
tertꢀButyl (4ꢀ(2ꢀoxoethyl)phenyl)carbamate (22). Using 21 (414 mg, 1.48 mmol) as the
starting material, following General Procedure B, gave the title compound as a colourless oil
1
which slowly solidified (347 mg, 99%). H NMR δ 9.71 (t, J = 2.4 Hz, 1H), 7.37 (d, J = 8.4
13
Hz, 2H), 7.13 (d, J = 8.5 Hz, 2H), 6.53 (s, 1H), 3.63 (d, J = 2.4 Hz, 2H), 1.52 (s, 9H). C
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
NMR δ 199.6 (CH), 152.6 (C), 137.9 (C), 130.3 (CH), 126.3 (C), 119.2 (CH), 80.2 (C), 50.0
(
CH
2 3
), 28.5 (CH ).
49
7ꢀNitroꢀ1,2,3,4ꢀtetrahydroisoquinoline
hydrochloride
(24).
1,2,3,4ꢀ
Tetrahydroisoquinoline (23, 10.0 g, 75.1 mmol) was added dropwise to ice cold concentrated
sulfuric acid (40 mL), and sodium nitrate (7.66 g, 90.1 mmol) was carefully added so as not
to reach above 5 °C. The solution was allowed to warm to RT, and was stirred for 20 h. The
mixture was then cooled on ice, diluted with water (80 mL), and conc. NH
4
OH solution was
(3 × 40 mL).
added to pH 10, causing a brown precipitate which was extracted with CHCl
3
The combined organic extracts were washed with brine (40 mL), dried over anhydrous
Na SO and evaporated to dryness. The orange oil was taken up in EtOH (100 mL), cooled
2
4
on ice and concentrated HCl (20 mL) added to force the slow precipitation of the HCl salt as
a yellow solid. The product was filtered, then recrystallised from CH OH and washed with
3
acetone to reveal the title compound as pale brown needles. A second recrystallisation in
1
aqueous acetone revealed the product as white needles (4.82 g, 30%) H NMR (d
6
ꢀDMSO) δ
10.08 (br s, 2H), 8.22 (d, J = 2.3 Hz, 1H), 8.09 (dd, J = 8.5, 2.5 Hz, 1H), 7.53 (d, J = 8.5 Hz,
1
3
1
H), 4.38 (s, 2H), 3.38 (t, J = 6.2 Hz, 2H), 3.17 (t, J = 6.1 Hz, 2H). C NMR (d ꢀDMSO) δ
6
1
45.9 (C), 140.4 (C), 131.1 (C), 130.2 (CH), 122.0 (CH), 121.9 (CH), 43.1 (CH
), 24.9 (CH ).
2
), 39.8
(
CH
2
2
11
1ꢀ(2,3ꢀDichlorophenyl)ꢀ1,4ꢀdiazepane (26). To a solution of 2,2'ꢀbis(diphenylphosphino)ꢀ
1
,1'ꢀbinaphthalene (110 mg, 177 ꢃmol) in degassed toluene (45 mL) was added
tris(dibenzylideneacetone)dipalladium (162 mg, 177 ꢃmol) and potassium tertꢀbutoxide (2.98
26
ACS Paragon Plus Environment

Page 27 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
g, 26.6 mmol). 1ꢀBromoꢀ2,3ꢀdichlorobenzene (25, 2.00 g, 8.85 mmol) and homopiperazine
(1.77 g, 17.7 mmol) were added and the brown mixture heated to 80 °C for 24 h. After
cooling to RT, a solution of diꢀtertꢀbutyl dicarbonate (5.80 g, 26.6 mmol) in toluene (5 mL)
was slowly added to the mixture and it was stirred at RT for 2 h. Water (20 mL) was added to
quench the reaction, and the mixture was extracted with EtOAc (3 × 25 mL). The combined
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
organic extracts were washed with brine (20 mL), dried over anhydrous Na
2 4
SO and
evaporated to dryness to give an orange oil. tertꢀButyl 4ꢀ(2,3ꢀdichlorophenyl)ꢀ1,4ꢀdiazepaneꢀ
1
ꢀcarboxylate was then purified by flash column chromatography (petroleum spirits/EtOAc,
:1). The yellow oil was taken up in DCM (5 mL) and TFA (1 mL) and stirred at RT. After
9
90 min, 10% NH OH solution (2 mL) was added, then the product extracted with DCM (2 ×
4
20 mL), and the combined organic extracts washed with brine (10 mL), dried over anhydrous
2 4
Na SO and evaporated to dryness to give the title compound as an orange oil (54 mg, 3%).
1
H NMR δ 7.11–7.06 (m, 2H), 7.04–6.99 (m, 1H), 3.33–3.20 (m, 4H), 3.12–2.99 (m, 4H),
13
2.09 (br s, 1H), 1.99–1.89 (m, 2H). C NMR δ 153.6 (C), 134.0 (C), 127.5 (C), 127.2 (CH),
23.9 (CH), 120.1 (CH), 57.9 (CH ), 54.6 (CH ), 49.9 (CH ), 47.9 (CH ), 31.4 (CH ). LCMS
1
2
2
2
2
2
+
(
m/z): 245.1 [M+H] .
General Procedure C (Reductive Alkylation)
The aldehyde (1 equiv) and amine (1 equiv) were taken up in 1,2ꢀDCE (10 mL).
NaBH(OAc) (1.5 equiv) was added to the stirred solution at RT under nitrogen. After 16–24
3
h LCMS confirmed reaction completion. The mixture was diluted with DCM (15 mL),
2 3
washed with 1 M K CO solution (3 × 20 mL), brine (15 mL), then dried over anhydrous
Na SO and evaporated to dryness. The crude material was purified by flash column
2
4
chromatography to give the title compound.
tertꢀButyl (transꢀ4ꢀ(2ꢀ(3,4ꢀdihydroisoquinolinꢀ2(1H)ꢀyl)ethyl)cyclohexyl)carbamate (4).
Using 16 (150 mg, 622 ꢃmol) as the aldehyde, and 23 (82.8 mg, 622 ꢃmol) as the amine
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
following General Procedure C. Eluted with EtOAc to give the title compound as a white
1
wax (190 mg, 85%). mp: 110ꢀ111 °C. H NMR δ 7.16–7.05 (m, 3H), 7.04–6.97 (m, 1H), 4.39
(br s, 1H), 3.60 (s, 2H), 3.38 (br s, 1H), 2.89 (t, J = 5.9 Hz, 2H), 2.71 (t, J = 5.9 Hz, 2H),
2.53–2.50 (m, 2H), 2.02–1.94 (m, 2H), 1.84–1.72 (m, 2H), 1.55–1.45 (m, 2H), 1.44 (s, 9H),
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
3
1.33–1.20 (m, 1H), 1.13–0.97 (m, 4H). C NMR δ 155.4 (C), 134.9 (C), 134.4 (C), 128.7
CH), 126.7 (CH), 126.2 (CH), 125.7 (CH), 79.1 (C), 56.4 (CH ), 56.3 (CH ), 51.1 (CH ),
), 29.2 (CH ), 28.6 (CH ). HPLC t
, 359.2693; found
(
2
2
2
50.0 (CH), 35.4 (CH), 34.3 (CH
2
), 33.6 (CH
2
), 32.1 (CH
2
2
3
R
+
=
8.93 min, 99% purity. HRMS (m/z): [MH] calcd. for C22
H
34
N
2
O
2
3
59.2711.
tertꢀButyl (transꢀ4ꢀ(2ꢀ(7ꢀnitroꢀ3,4ꢀdihydroisoquinolinꢀ2(1H)ꢀyl)ethyl)cyclohexyl)carbamate
29). Using 16 (162 mg, 671 ꢃmol) as the aldehyde, and 24 (119 mg, 671 ꢃmol) as the amine
(
following General Procedure C. Eluted with EtOAc, to give the title compound as pale
1
yellow flakes (213 mg, 79%). mp: 137ꢀ138 °C. H NMR δ 7.97 (dd, J = 8.4, 2.4 Hz, 1H),
7.92 (d, J = 2.2 Hz, 1H), 7.24 (d, J = 8.4 Hz, 1H), 4.38 (br s, 1H), 3.68 (s, 2H), 3.38 (br s,
1
H), 2.98 (t, J = 5.8 Hz, 2H), 2.75 (t, J = 5.9 Hz, 2H), 2.61–2.50 (m, 2H), 2.04–1.92 (m, 2H),
.85–1.72 (m, 2H), 1.53–1.46 (m, 2H), 1.45 (s, 9H), 1.36–1.22 (m, 1H), 1.17–0.97 (m, 4H).
C NMR δ 155.37 (C), 146.2 (C), 142.6 (C), 136.6 (C), 129.7 (CH), 121.9 (CH), 121.3
1
1
3
2 2 2 2
(CH), 79.1 (C), 56.1 (CH ), 55.9 (CH ), 50.3 (CH ), 50.0 (CH), 35.3 (CH), 34.1 (CH ), 33.5
(CH ), 32.1 (CH ), 29.5 (CH ), 28.6 (CH ). HPLC t = 9.01 min, 95% purity. HRMS (m/z):
2
2
2
3
R
+
[
MH] calcd. for C H N O , 404.2544; found 404.2549.
22 33 3 4
tertꢀButyl
(transꢀ4ꢀ(2ꢀ(7ꢀcyanoꢀ3,4ꢀdihydroisoquinolinꢀ2(1H)ꢀ
1
5
yl)ethyl)cyclohexyl)carbamate (30). Using 16 (215 mg, 891 ꢃmol) as the aldehyde, and 27
(141 mg, 891 ꢃmol) as the amine following General Procedure C. Eluted with EtOAc to give
1
the title compound as white needles (193 mg, 56%). mp: 294ꢀ296 °C. H NMR δ 7.39 (dd, J
=
7.9, 1.6 Hz, 1H), 7.32 (s, 1H), 7.18 (d, J = 7.9 Hz, 1H), 4.37 (br s, 1H), 3.60 (s, 2H), 3.37
28
ACS Paragon Plus Environment

Page 29 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(br s, 1H), 2.94 (t, J = 5.8 Hz, 2H), 2.72 (t, J = 5.9 Hz, 2H), 2.57–2.49 (m, 2H), 2.05–1.94 (m,
1
3
2
H), 1.82–1.74 (m, 2H), 1.53–1.40 (m, 11H), 1.32–1.22 (m, 1H), 1.14–0.98 (m, 4H).
NMR δ 155.2 (C), 140.3 (C), 136.3 (C), 130.3 (CH), 129.5 (CH), 129.4 (CH), 119.0 (C),
09.2 (C), 78.8 (C), 56.0 (CH ), 55.5 (CH ), 50.2 (CH ), 49.8 (CH), 35.1 (CH), 34.0 (CH ),
C
1
2
2
2
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
33.3 (CH ), 31.9 (CH ), 29.4 (CH ), 28.4 (CH ). HPLC t = 8.76 min, 98% purity. HRMS
2
2
2
3
R
+
(
m/z): [MH] calcd. for C22
H
33
N
3
2
O , 384.2646; found 384.2645.
tertꢀButyl
(transꢀ4ꢀ(2ꢀ(4ꢀ(2,3ꢀdichlorophenyl)ꢀ1,4ꢀdiazepanꢀ1ꢀ
50
yl)ethyl)cyclohexyl)carbamate (31). Using 16 (50 mg, 207 ꢃmol) and 26 (56 mg, 228 ꢃmol)
as the amine following General Procedure C. Purification by flash column chromatography
1
(CHCl /CH OH, 50:1 to 20:1) gave the title compound as a pale yellow oil (48 mg, 49%). H
3
3
NMR δ 7.12–7.05 (m, 2H), 7.03–6.96 (m, 1H), 4.37 (br s, 1H), 3.43–3.20 (m, 5H), 2.87 (br s,
4H), 2.66–2.53 (m, 2H), 2.08–1.90 (m, 4H), 1.83–1.70 (m, 2H), 1.50–1.39 (m, 11H), 1.30–
1
3
1
.18 (m, 1H), 1.13–0.96 (m, 4H). C NMR δ 155.3 (C), 153.0 (C), 134.1 (C), 127.1 (CH),
127.0 (C), 123.9 (CH), 119.9 (CH), 79.1 (C), 56.3 (CH ), 56.2 (CH ), 54.6 (CH ), 54.2
2
2
2
(
CH ), 53.4 (CH ), 50.0 (CH), 35.5 (CH), 34.2 (CH ), 33.6 (CH ), 32.1 (CH ), 28.6 (CH ),
2
2
2
2
2
3
+
27.8 (CH
2
R
). HPLC t = 10.18 min, >99% purity. HRMS (m/z): [MH] calcd. for
C
24
H
37Cl
2
N
3
2
O , 470.2336; found 470.2345.
tertꢀButyl
(transꢀ4ꢀ(2ꢀ(4ꢀ(2,3ꢀdichlorophenyl)piperazinꢀ1ꢀyl)ethyl)cyclohexyl)carbamate
12
(32). Using 16 (200 mg, 829 ꢃmol) as the aldehyde and 28 (230 mg, 995 ꢃmol) as the
amine following General Procedure C. Purification by flash column chromatography
(petroleum spirits/EtOAc, 5:1) gave the title compound as a white wax (262 mg, 69%). mp:
1
1
43ꢀ145 °C. H NMR δ 7.17–7.10 (m, 2H), 6.99–6.92 (m, 1H), 4.37 (br s, 1H), 3.37 (br s,
1H), 3.07 (br s, 4H), 2.62 (br s, 4H), 2.48–2.38 (m, 2H), 2.04–1.92 (m, 2H), 1.82–1.73 (m,
13
2
H), 1.49–1.37 (m, 11H), 1.30–1.17 (m, 1H), 1.15–0.97 (m, 4H). C NMR δ 155.3 (C),
151.5 (C), 134.2 (C), 127.64 (C), 127.56 (CH), 124.7 (CH), 118.7 (CH), 79.2 (C), 56.7
29
ACS Paragon Plus Environment