Discovery and Characterization of Biased Allosteric Agonists of the Chemokine Receptor CXCR3

Article abstract of DOI:10.1021/acs.jmedchem.5b01965

In this work we report a design, synthesis, and detailed functional characterization of unique strongly biased allosteric agonists of CXCR3 that contain tetrahydroisoquinoline carboxamide cores. Compound 11 (FAUC1036) is the first strongly biased allosteric agonist of CXCR3 that selectively induces weak chemotaxis and leads to receptor internalization and the β-arrestin 2 recruitment with potency comparable to that of the chemokine CXCL11 without any activation of G proteins. A subtle structural change (addition of a methoxy group, 14 (FAUC1104)) led to a contrasting biased allosteric partial agonist that activated solely G proteins, induced chemotaxis, but failed to induce receptor internalization or β-arrestin 2 recruitment. Concomitant structure-activity relationship studies indicated very steep structure-activity relationships, which steer the ligand bias between the β-arrestin 2 and G protein pathway. Overall, the information presented provides a powerful platform for further development and rational design of strongly biased allosteric agonists of CXCR3.

Full text of DOI:10.1021/acs.jmedchem.5b01965

Subscriber access provided by LOCK HAVEN UNIV
Article
Discovery and characterization of biased allosteric
agonists of the chemokine receptor CXCR3
Lampros Milanos, Regine Brox, Theresa Frank, Gasper Poklukar, Ralf Palmisano,
Reiner Waibel, Jürgen Einsiedel, Maximilian Dürr, Ivana Ivanovic-Burmazovic, Olav
Larsen, Gertrud Malene Hjorto, Mette Marie Rosenkilde, and Nuska Tschammer
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01965 • Publication Date (Web): 10 Feb 2016
Downloaded from http://pubs.acs.org on February 11, 2016
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 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery and characterization of biased allosteric
agonists of the chemokine receptor CXCR3
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
†
†
†
†, ‡
§
Lampros Milanos , Regine Brox , Theresa Frank , Gašper Poklukar , Ralf Palmisano ,
†
†
∥
∥
Reiner Waibel , Jürgen Einsiedel , Maximilian Dürr , Ivana IvanovićꢀBurmazović , Olav
⊥
⊥
⊥
*, †,#
Larsen , Gertrud Malene Hjortø , Mette Marie Rosenkilde , Nuska Tschammer
†
Department of Chemistry and Pharmacy, Medicinal Chemistry, Emil Fischer Center,
Friedrich Alexander University, Schuhstraße 19, 91052 Erlangen, Germany.
‡
Faculty of Pharmacy, University of Ljubljana, Aškerčeva 7, 1000 Ljubljana, Slovenia.
§
Optical Imaging Center Erlangen, Friedrich Alexander University, Hartmannstraße 14,
91052 Erlangen, Germany
⊥
Department of Neuroscience and Pharmacology, Laboratory for Molecular Pharmacology,
Faculty of Health and Medical Sciences, University of Copenhagen, Blegdamsvej 3B, DKꢀ
200 Copenhagen N, Denmark
2
∥Department of Chemistry and Pharmacy, Bioorganic Chemistry, Friedrich Alexander
University, Egerlandstraße 1, 91058 Erlangen, Germany
#Current address: NanoTemper Technologies, Flößergasse 4, 81369 München.
Corresponding author: Nuska Tschammer, nuska.tschammer@nanotemper.de
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 87
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
In this work we report a design, synthesis and detailed functional characterization of unique
strongly biased allosteric agonists of CXCR3 that contain tetrahydroisoquinoline
carboxamide cores. The compound 11 (FAUC1036) is the first strongly biased allosteric
agonist of CXCR3 that selectively induces weak chemotaxis, leads to receptor internalization
and the βꢀarrestin 2 recruitment with potency comparable to that of the chemokine CXCL11
without any activation of G proteins. A subtle structural change (addition of a methoxy
group, 14 (FAUC1104)) led to a contrasting biased allosteric partial agonist that activated
solely G proteins, induced chemotaxis but failed to induce receptor internalization or βꢀ
arrestin 2 recruitment. Concomitant structureꢀactivity relationship studies indicated very steep
structureꢀactivity relationships, which steer the ligand bias between the βꢀarrestin 2 and G
protein pathway. Overall, the information presented provides a powerful platform for further
development and rational design of strongly biased allosteric agonists of CXCR3.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
2

Page 3 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Biased signaling is a now well recognized phenomenon observed in the G protein coupled
receptors (GPCRs) that describes the ability of different ligands acting at the same GPCR to
stabilize distinct receptor conformations that, in turn, are linked to different functional
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
–3
outcomes.
In the case of the human chemokine receptor CXCR3, which binds three
endogenous chemokines CXCL9, CXCL10 and CXCL11, biased signaling is used by nature
+
4
to fine tune the responses of CXCR3 cells upon physiological insult. CXCR3 expression is
rapidly induced on naïve T cells following activation, and preferentially remains highly
+
+
expressed on type1 helper (Th1) CD4 T cells, effector CD8 T cells and innateꢀtype
5
,6
lymphocytes, such as natural killer (NK) and natural killer T (NKT) cells. Depending on
the endogenous chemokine bound, the activation of CXCR3 leads to distinct functional
5,7
outcomes, which considerably influence the course of inflammation and healing processes.
The opportunity to mimic this mechanism by small molecule allosteric ligands has remained
largely unexploited, even though CXCR3 has been deemed as a promising therapeutic target
in pathologies like autoimmune diseases (e.g., rheumatoid arthritis, multiple sclerosis), cancer
8–12
+
and metastasis.
To inhibit the inflammatory response driven by the CXCR3 cells,
published research was focused solely on the development of smallꢀmolecules with inhibitory
13–15
effects on the CXCR3, which ultimately failed to deliver therapeutics.
A single study used
metalꢀion site engineering to probe the molecular mechanism of CXCR3 activation and
identified highꢀpotency allosteric agonists based on complexes containing Cu or Zn and
16
phenantroline or bipyridine. Beside this, the development of smallꢀmolecule agonists of
17–19
CXCR3 was largely neglected and only a few agonists were reported (Figure 1)
, despite
rising indications that CXCR3 agonists may have significant therapeutic potential in wound
2
0
21,22
healing and antiꢀinflammatory therapy.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 87
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 general focus of our group is the understanding of the structureꢀactivity relationships
SAR) of allosteric modulators of CXCR3 with probeꢀdependent behavior and biased
(
23–25
signaling.
The unique nature of allosteric ligands (reflected in permissivity, saturability
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
of effect, probe dependence and biased signaling) offers great potential to extend therapeutic
2
,4
effects compared to orthosteric ligands, and is still a poorly explored opportunity. In this
work we focused our research on the allosteric agonists of CXCR3.
The allosteric modulators of CXCR3 based on the tetrahydroisoquinoline carboxamide core
were previously reported as allosteric agonists of CXCR3 (Figure 1) with no noticeable
1
7,19
ligandꢀdirected signaling.
Our own explorations of this promising scaffold resulted in 11
(FAUC1036), a strongly biased allosteric agonists of CXCR3 (Figure 2 and 3). 11 selectively
induces chemotaxis and receptor internalization, and induces βꢀarrestin 2 recruitment without
any stimulation of G proteins. During subsequent structureꢀactivity studies and detailed
pharmacological profiling of these novel compounds we unambiguously determined crucial
structural elements that drive biased signaling of these unique allosteric agonists. We
discovered a molecular switch that steers ligandꢀinduced biased signaling between the βꢀ
arrestin 2 and G protein pathway as shown in 14 (FAUC1104) (Figure 3) that activates G
proteins and induces chemotaxis, but cannot induce receptor internalization and βꢀarrestin 2
recruitment.
Results and Discussion
Based on our initial hit, the strongly biased allosteric agonist 11, we synthesized an
enantiomerically and epimerically pure series (Supplemental Figure S4) of derivatives
focusing on variations of the northern side chain (Figure 2). We combined this approach with
detailed pharmacological characterization of novel compounds that allowed us to gain novel
ACS Paragon Plus Environment
4

Page 5 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
insight into the structureꢀactivity relationships for the fine control of allosteric modulator
potency, cooperativity and biased signaling at CXCR3.
Synthesis of allosteric ligands
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The preparation of the 11 began with an acylation step of diphenyl ethylamine using Bocꢀ
Lys(Z)ꢀOH in presence of EDC hydrochloride (compound 1), followed by hydrogenolysis to
afford the primary amine 2. Then, Cbz protection of the commercially available (S)ꢀ2ꢀ
((benzyloxy)carbonyl)ꢀ1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxylic acid
using benzyl
26
chloroformate in basic medium to give compound 3 followed by an amide coupling step
using 2 resulted in the formation of compound 4. Hydrogenolysis yielded compound 5, a
precursor ready to be coupled to the northern side chain in one step (Scheme 1).
For the variation of the northern side chain several nonꢀaromatic heterocyclic acids 6ꢀ8
were synthesized by acylating piperidine, morpholine and Nꢀmethyl piperazine, respectively
27,28
with succinic anhydride at room temperature.
The aromatic heterocyclic acids 9, 10 were
obtained by a Friedel Crafts acylation of furan and thiophene with succinic anhydride as
2
9,30
well
were coupled to 5 using EDC hydrochloride and DIPEA to afford the final compounds 11 ꢀ
4 (Scheme 3). Compound 27 was produced by acylating of 5 with succinic anhydride. The
(Scheme 2). The remaining carboxylic acids were commercially available. All acids
2
Boc group of compounds 11, 24 and 5 was removed with TFA to yield the protonated
compounds 25, 26, 28 respectively.
Furthermore, the (S)ꢀconfiguredꢀlysine moiety of 11 and 25 was replaced with the (R)ꢀlysine
resulting in the compounds 34 and 35. The synthetic strategy of 29ꢀ35 was identical to that of
11 (supporting information, Supplemental Scheme 1).
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 87
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
All compounds mentioned above are free of contaminants, which could have arisen due to
racemization during the activation process of the amino acids. This has been demonstrated by
3
additionally recording spectra in CDCl , since they unambiguously allowed to discriminate
key proton resonances of both the epimers 11 (S,S) and 34 (S,R) (see Supporting Figure S4).
In neither recorded sample of these two compounds, the resonances of the corresponding
epimer could be detected. This allows the conclusion, that also the precursor 5 leading to the
other target compounds 12ꢀ28 does not contain the corresponding epimer either. Due to the
missing tendency of enantiomerization during the peptide coupling process, we are also able
to exclude the potential formation of the (R,R) enantiomer and the (R,S) epimer (which
should give the same set of signals like the (S,R) isomer 34) of target compound 11.
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
17
To enable appropriate comparison of reference compound 46 (VUF10661, see Figure 1)
with our novel ligands, we developed analogues by inverting the lysine’s center of chirality to
give the epimer 47. Furthermore, we decided to test the Bocꢀprotected intermediate 44 and
45, in order to investigate the influence of the bulky, uncharged side chains compared to the
protonated aminoalkyl side chains of the compounds 46 and 47. The positively charged
18,19
amine groups were reported to be crucial for receptor activation.
For the synthesis of
compound 46 and its analogues, we followed a strategy similar to that eported in Scheme 1
(supporting information, Supplemental Scheme 2), which gave significantly higher yields
19
than those found in the literature.
Pharmacological characterization
The binding of endogenous agonists CXCL11, CXCL10 and CXCL9 to CXCR3 results in
the activation of Gα proteins, recruitment of βꢀarrestins, receptor internalization and
i
chemotaxis. To elucidate the pharmacological properties of derivatives of 11, we focused on
the ability of derivatives to modulate G protein activation and βꢀarrestin 2 recruitment. For
ACS Paragon Plus Environment
6

Page 7 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the most prominent compounds 11 and 14 their capacity to induce CXCR3 receptor
internalization and chemotaxis was investigated in detail. Prior to functional characterization
the potential cytotoxicity of derivatives was determined with ApoToxꢀGlo Triplex assay
(Promega), which easily assesses viability, cytotoxicity and apoptosis events in the same cellꢀ
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
based assay. None of derivatives exhibited nonspecific effects on the cells (supplemental
information, Supplementary Figure S1). The activation of G proteins was explored in terms
35
of CXCR3 mediated [ S]GTPγS incorporation.
35
The activation of G proteins was monitored as the incorporation of [ S]GTPγS and thus
increase in the incorporated radioactivity. The potency and efficacy of our novel compounds
was first compared to the reference compound 46. The compound 46 produced a doseꢀ
35
dependent increase in the [ S]GTPγS incorporation with a potency of 0.6 µM (pEC50= 6.2 ±
19
0
.1), which is in accordance with the published data (Table 1). The epimer of 46 compound
47 as well as the Boc protected derivatives 44 and 45 showed no significant changes in
potency and efficacy. Although it was reported that the charge at the amine on the side chain
18,19
of the lysine moiety was crucial for the biological activity of 46,
we clearly demonstrate
that this charge is fully dispensable for the activation of G proteins by our allosteric CXCR3
agonists.
Strikingly, 11 failed to induce CXCR3 mediated G protein activation, just as its
unprotected form 25 and its epimer 34 (Table 2). The epimer of 25 compound 35 showed
some weak partial agonism at pEC50 of 8.01±0.31. Compound 22, where the keto group of
the benzoyl moiety was removed, showed very low potency (pEC50 = 5.07±0.50) as well. To
determine the influence of substitution pattern using polar groups at the para position of the
phenyl ring on the northern side chain, fluorine (13), methoxy (14) and methyl sulfonyl (15)
groups were introduced. This modification resulted in an increase of the intrinsic agonist
properties of compounds. The introduction of methoxy group (14) caused the most significant
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 87
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
increase in the efficacy (50%; the pEC50 value of 5.22±0.31). An increase of the electron
density near the receptor by an addition of a second methoxy group at the meta position (21)
caused a surprising reduction of the efficacy (18%), which indicated that the activity is not
dependent on the electron density of the phenyl ring. Therefore, we removed the aromaticity
by replacing the phenyl ring of the northern side chain with a piperidine moiety (16),
however no reactivity was observed. Knowing that the polar substitution of the phenyl ring
increased the reactivity of the ligands in the G protein activation pathway, we introduced a
polar morpholine (17) and a charged Nꢀmethyl piperazine moiety (18), in order to improve
the reactivity of the compound 16. The results showed that nonꢀaromatic heterocycles have
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
5
no effect in the [ S]GTPγS assay. Similar results were obtained for the aromatic heterocycles
9 and 20, which might indicate that the length of the northern side chain is too short for
1
creating a possible hydrogen bond/s with the receptor compared to 14. We therefore
generated analogues 5, 23 and 24 in which we varied the length of a spacer in the northern
side chain. Compound 5 showed the most interesting results, where the efficacy in the
35
[
S]GTPγS incorporation increased to 34% with the EC50 value of 2.2 µM (pEC = 5.7 ±
50
0.34) indicating the importance of the specific side chain for keeping the G protein inactive.
Finally, Bocꢀfree compounds 26 and 28 with modified spacer of the northern side chain
showed no significant changes compared to their Boc analogues 24 and 5, respectively,
showing that the Boc group or the positive charge are not crucial for the G protein activation.
Importantly, all novel ligands that showed no intrinsic agonist activity were tested for the
ability to inhibit the CXCL11 or CXCL10 mediated activation of G proteins. None of the
ligands were able to inhibit the activation of CXCR3 by its endogenous chemokine ligand.
ACS Paragon Plus Environment
8

Page 9 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The ability of our novel ligands to induce CXCR3ꢀmediated βꢀarrestin 2 recruitment was
investigated in the assay system PathHunter (DiscoveRx), which detects the recruitment of βꢀ
arrestin 2 to the activated receptor using βꢀgalactosidase (βꢀgal) enzyme fragment
complementation. We used two different ProLink tags, ProLink1 (PK1) and ProLink2 (PK2),
of which PK2 has higher affinity for the enzyme acceptor to detect weak proteinꢀprotein
interactions. The CXCR3ꢀPK2 construct is readily able to detect the differences in the
propensity of CXCL11 and CXCL10 to induce βꢀarrestin 2 recruitment at CXCR3
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
(supplemental information, Supplementary Figure S2). As reported, the reference compound
19
46 showed the pEC50 value 5.89 ± 0.07 and nearly 40% higher efficacy than CXCL11.
Contrary to the G protein activation, the enantiomer of 46 47 and the Boc protected
derivatives 44 and 45 had a reduced ability to induce recruitment of βꢀarrestin 2 to CXCR3
(Table 3). The most striking difference was observed for the Boc protected enantiomer of 46,
compound 45, which showed the reduced efficacy of about 36% compared to 46. The
presence of the positive charge at the amine on the southern side chain seems to significantly
contribute to the activation of the βꢀarrestin 2 pathway.
11 emerged as a strongly biased allosteric partial agonist as it activated the recruitment of
βꢀarrestin 2 to CXCR3 with a pEC50 of 6.64 ± 0.38 and the efficacy of 30% (measured with
PK1 construct, Table 4), whereas no G protein activation could be observed for this
compound. Every substitution on the phenyl ring of the northern side chain, fluorine (13),
methoxy (14), sulfone (15), dimethoxy (21) groups resulted in complete loss of intrinsic
agonistic properties in the βꢀarrestin 2 recruitment assay, results that suggest phenyl ring
polar substitutions enhance biased agonism by preferentially activating G protein over
recruitment of βꢀarrestin 2. Replacement of the benzoyl group with an acetyl group
(compound 12) or increase of the spacer length (compound 27) behaved as strongly biased
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 87
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
allosteric agonists with efficacies and potencies comparable to 11 (Table 4). For the
compounds 22 and 23, where the keto group of the benzoyl moiety was removed, a
significant loss of efficacy is observed indicating the importance of that group with respect to
βꢀarrestin 2 recruitment. The S,Rꢀepimer 34 reached the levels of full agonism in assays using
PK2 construct. The charged analogues 25, 28 and 35 showed a decrease of efficacy.
Compound 5, which is devoid of the entire northern fragment, retained its ability to recruit βꢀ
arrestin 2 to CXCR3 (the pEC50 value 6.50±0.17, efficacy 31%), and showed no bias between
the investigated pathways. Finally, the rest of the compounds, 16ꢀ20 and 27, proved to have
no agonistic effect in the given assays.
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
GPCRs are generally rapidly desensitized, internalized and degraded upon agonist
31–33
stimulation.
The βꢀarrestin proteins are generally involved in the internalization of
34
GPCRs. Although it has been shown that βꢀarrestin is involved in CXCR3 internalization to
some extent, it is not the sole determinant of this process, and the exact mechanism (both βꢀ
32,33,35
arrestin dependent and independent) of CXCR3 internalization is still unclear.
For the
chemokine receptors it is known that also a βꢀarrestin independent type of internalization
3
6
plays an important role. E.g., for the viral chemokine receptor US28 it was reported that the
receptor endocytosis does not require βꢀarrestins and is dependent on the clathrinꢀmediated
37
pathway. For the chemokine receptor CCR5 it was reported that caveolae are involved in
3
8
receptor internalization. Overall, it seems that in chemokine receptors total receptor
internalization is the sum of βꢀarrestin dependent and –independent events. To investigate the
effect of 11 (βꢀarrestin 2 biased allosteric agonist) and 14 (G protein biased allosteric agonist)
on internalization, we employed scanning laser confocal microscopy and an ELISAꢀbased
assay. The HEK293T cells were transiently transfected by CXCR3ꢀeGFP and either left
untreated or stimulated with CXCL11 and 11 or 14 for one hour. As evident in Figure 4A, the
treatment with either CXCL11 or 11 induced the CXCR3 internalization. Untreated cells also
ACS Paragon Plus Environment
1
0

Page 11 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
show minor internalization of CXCR3. This observation is in accordance with the published
39
observation that CXCR3 is constitutively internalized to some extent. 14 did not induce
receptor internalization beyond the internalization observed in the untreated cells. The
quantification of receptor internalization upon the treatment with the endogenous agonists
and with either βꢀarrestin 2 biased 11 or Gꢀprotein biased 14 was performed with an ELISA
based assay. The HEK293T cells transiently expressing FLAGꢀCXCR3 were stimulated with
compounds and chemokines for one hour, washed, fixed and the amount of FLAGꢀCXCR3
present on the surface was detected by antiꢀFLAG antibody. As evident from Figure 4B, only
CXCL11 and 11, the βꢀarrestin 2 biased allosteric agonist, induced receptor internalization.
These experiments also demonstrated the unique value of our ligands to describe signaling
events because they remove one of the signals from an otherwise normal activation event.
The ligands 11 and 14 clearly displayed multiple dimensions of biased signaling with 11
stimulation the βꢀarrestin 2 recruitment and receptor internalization, and 14 the activation of
G proteins only. To estimate potential cooperativity between those ligands and endogenous
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
chemokine ligand CXCL11, we performed radioligand binding assay using [ I]CXCL11
(Figure 5A and B). The data showed that neither 11 nor 14 were able to suppress the binding
of CXCL11 to CXCR3 (Figure 5A). Our ligands thus bind to an allosteric binding site of
CXCR3 and do not exert any cooperativity toward CXCL11. This observation was further
confirmed by the experiment in which the fixed concentration (1 µM) of 11 and 14 was used
and increasing concentrations of CXCL11. It is evident that the presence of either 11 or 14
has no influence on the binding of CXCL11 (Figure 5B). Our ligands are thus allosteric
19
agonists. Similar observation was previously reported also for 46. The lack of cooperativity
of between the 11 and CXCL11 was observed also in the βꢀarrestin 2 recruitment assay
(Figure 5C). We decided to determine the potential for CXCL11ꢀmediated beta arrestin
recruitment to CXCR3 receptors already responding to 11, and thus extended the ligand
ACS Paragon Plus Environment
1
1

Journal of Medicinal Chemistry
Page 12 of 87
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
response time. During this experiment we observed some minor reduction in the basal βꢀ
arrestin 2 recruitment. The data in Figure 5C are normalized in a manner that makes the basal
βꢀarrestin 2 recruitment visible. The maximal stimulation of 100% is defined by the maximal
response exerted by CXCL11. Because our ligand do not show any toxicity (Supplemental
Figure 1), we assume that this could be a result of a long treatment period (of four hours) in
the combination with high basal βꢀarrestin 2 recruitment to CXCR3 that results in the
receptor internalization and thus causes desensitation, which is defined as the fading of a
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
40
receptor response to persistent agonist stimuli. The demonstrated internalization caused by
11 (Figure 4) would inhibit constitutive CXCR3 activity, but being allosteric, would not
affect the potency of CXCL11 on those receptors still available on the cell surface.
Chemotaxis represents the crucial and defining functional response induced by the
12,41–43
activation of chemokine receptors.
i
Signaling through Gα has been shown to activate
CXCR3 induced migration through multiple pathways, including p38 and PI3K in human
44
45
epithelial airway cells and PI3K and PTEN in human neutrophils . No direct studies on
CXCR3 induced βꢀarrestin mediated migration has been published, but it is possible that
CXCR3 chemokines convey some of their migratory stimulation through βꢀarrestin induced
signaling, as reported for CXCL11 at the decoy receptor ACKR3/CXCR7 and for angiotensin
46,47
II at the angiotensin II type 1a receptor.
To investigate the influence of 11 (βꢀarrestin 2
biased allosteric agonist) and 14 (G protein biased allosteric agonist) on CXCR3 induced
migration, we used these compounds alongside CXCL11 in a standard transwell assay with
HEK293 cells transiently transfected with CXCR3. When applied to the source reservoir,
both compounds on their own induced migration approximately 0.25 fold compared to buffer
control, which is less than the maximal effect observed with CXCL11 (0.7 fold), but still
significant, indicating that both compounds positively influence migration (Figure 6A). This
means that in CXCR3, neither G protein activation, nor βꢀarrestin recruitment, can adequately
ACS Paragon Plus Environment
1
2

Page 13 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
account for the final functional chemotaxis response. To test whether 11 and 14 influence
CXCL11 induced migration, we preꢀincubated the cells with these compounds for 15 minutes
and then set up the migration assay in the presence of either compound in both the top
chamber (sink) containing the cells and in the lower chamber (source) containing CXCL11 as
chemoattractant. We observed no antagonizing effect of these compounds on CXCL11
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
induced migration (Figure 6B). Previously reported CXCR3 agonists (Figure 1) were shown
17,19,22
to induce migration in the absence of endogenous chemokines.
Only 48 (PS372424)
was tested for its influence on CXCL11 induced migration and it was, interestingly, found to
2
2
antagonize CXCL11 induced migration in activated Tꢀcells.
Overall, detailed pharmacological characterization of the most prominent derivatives 11
and 14 demonstrated the multiple dimensions of biased agonism induced by these compounds
and importantly contributed to the dissection of the CXCR3 signaling involved in receptor
internalization and chemotaxis. With the aid of 11, the βꢀarrestin 2 biased allosteric agonist,
and 14, the G protein biased allosteric agonist, we: (1) identified molecular switches which
steer the ligand bias between the βꢀarrestin 2 and G protein pathway (Figure 7); (2)
demonstrated that the CXCR3 internalization is largely βꢀarrestin 2 dependent or at least G
protein independent; and, (3) showed in CXCR3, neither G protein activation, nor βꢀarrestin
recruitment, can adequately account for the final functional chemotaxis response.
Conclusions
Selective regulation of CXCR3 signaling by an appropriate allosteric ligand could be an
attractive therapeutic strategy in inflammatory diseases. However, the complex
pharmacology of the CXCR3 signaling system poses a nontrivial challenge for the design of
biased ligands that selectively activate desired pathways. In this work we report unique
strongly biased allosteric agonists of CXCR3 that contain a tetrahydroisoquinoline core. 11
ACS Paragon Plus Environment
1
3

Journal of Medicinal Chemistry
Page 14 of 87
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
is the first strongly biased allosteric agonist of CXCR3 that exclusively stimulates receptor
internalization and βꢀarrestin 2 recruitment with a potency comparable to the endogenous
chemokine CXCL11 and an efficacy of 66% (using CXCR3ꢀPK2) without any activation of
G proteins (Figure 3). An additional methoxy group led to the strongly biased allosteric
agonist 14 that activated solely G proteins (a pEC50 value 5.22, efficacy 50%). Concomitant
structureꢀactivity studies identified molecular switches which steer the ligand bias between
the βꢀarrestin 2 and G protein pathway (Figure 7). Detailed analysis of the influence of 11
and 14 on receptor internalization and chemotaxis showed that 11 induces βꢀarrestin 2
dependent receptor internalization and chemotaxis; 14 induces only G protein dependent
chemotaxis.
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
With good reason, it is broadly acknowledged that biased ligands represent an opportunity
to induce precisely defined GPCR signaling, without casually activating irrelevant pathways,
which many speculate to be a major source of side effects in current medical drugs. For
practical reasons, the most effective benchmark assays to characterize receptor signaling
provide very narrow information, but it is vital to always interpret those as a fragment of a
larger range of signaling pathways. One of the most important lessons of 11 and 14 is that βꢀ
arrestin 2 and large Gα protein subunit signaling are dispensable to produce an important
i
functional response ꢀ chemotaxis. It is possible that a crossꢀtalk occurs between βꢀarrestin 2
i
and Gα signaling as a late pathway event, perhaps as a backup mechanism, which would
make it very difficult to produce true biased signaling. However, it seems more likely that
another, distinct, pathway has a stronger connection to chemotaxis, such as Gα12/13, which has
48
already been connected to T cell adhesion . 11 and 14 are rare examples of fully biased
synthetic agonists, and will be valuable tools to dissect and describe CXCR3 responses in a
variety of physiologically applicable contexts.
ACS Paragon Plus Environment
1
4

Page 15 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Experimental section
Chemistry
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
General methods
All reactions requiring anhydrous conditions were carried out under nitrogen and the
glassware were dried in flame temperature under vaccum before use. All reactions were
monitored by TLC using Kieselgel 60 F254 plates. Visualization of the reaction components
4
was achieved using UV fluorescence (254 nm) and KMnO stain. Silica gel chromatography
1
13
was carried out on Kieselgel 60. The yields are reported after purification. H, C NMR
spectra were recorded in deuterated solvents and chemical shifts (δ) are quoted in parts per
1
13
million (ppm) calibrated to TMS ( H and C δ = 0). Coupling constants (J) are measured in
Hertz (Hz). The following abbreviations are used to describe multiplicities: s = singlet, d =
doublet, dd = double doublet, dt = double triplet, t = triplet, q = quartet, br s = broad singlet,
m = multiplet. Electronꢀionization (EI) MS measurements were performed on a JEOL JMSꢀ
GC Mate II spectrometer, capable of 5.000 FWHM. Detection was in positive ion mode and
source voltage of 70eV. The machine was calibrated prior to every experiment via direct
infusion of the low concentration tuning mixture, which provided an m/z range of singly
charged peaks. Electrosprayꢀionization MS (ESIꢀMS) measurements were performed on a
UHRꢀTOF Bruker Daltonik (Bremen, Germany) maXis plus 5G, an ESIꢀToF MS capable of
resolution of at least 60.000 FWHM. Detection was in positive ion mode, the source voltage
2
was 2.8 kV. The flow rates were 180 µL/hour. The drying gas (N ), to aid solvent removal,
was held at 180 °C and the spray gas was held at 20 °C. The machine was calibrated prior to
every experiment via direct infusion of the Agilent ESIꢀTOF low concentration tuning
mixture, which provided an m/z range of singly charged peaks up to 2700 Da in both ion
modes. Optical rotation was measured using a JASCO Pꢀ2000 polarimeter, with Na lamp at
ACS Paragon Plus Environment
1
5

Journal of Medicinal Chemistry
Page 16 of 87
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
589 nm. Purity and identity were assessed by analytical RPꢀHPLC (Agilent 1100 analytical
series, column: Zorbax Eclipse XDBꢀC8 , 4.6 × 150 mm, 5 ꢁm, flow rate: 0.5 ml/min,
detection wavelength: 254 nm) coupled to a Bruker Esquire 2000 mass detector equipped
with an ESIꢀioniser. Purities of the products were assessed from HPLCꢀMS using MeOH /
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H O 0.1% HCOOH as the solvent system and employing the following gradient systems:
2
LCꢀMS:
Gradient A: 0ꢀ3 min: 10% MeOH, 3ꢀ18 min: 10ꢀ100% MeOH, 18ꢀ24 min: 100% MeOH, 24ꢀ
30 min: 100ꢀ10% MeOH.
Gradient B: 0ꢀ2 min: 10% MeOH, 2ꢀ10 min: 10ꢀ100% MeOH, 10ꢀ14 min: 100% MeOH, 14ꢀ
6 min: 100ꢀ10% MeOH, 16ꢀ18 min: 10% MeOH.
1
The purity of final products was ≥95%.
NMR performance at high temperature.
NMR analysis at high temperature was performed for compounds with multiple isomers, in
order to avoid as much as possible multiple peaks due to isomerism. (or similar).
Preparative HPLC
Preparative RPꢀHPLC was performed using Agilent 1100 preparative series. Column:
Nucleodur C18 HTec, 32 × 250 mm, 5 ꢁm particles, flow rate 32 mL/min, detection
wavelengths: 210, 254 nm. The solvent system, that was used for the preparative HPLC, was
acetonitrile (ACN), water with 0.1% trifluoroacetic acid (TFA). The methods differ only in
gradient.
Method A
ACS Paragon Plus Environment
1
6

Page 17 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
20ꢀ80% ACN in 0ꢀ20 min, 80ꢀ95% ACN in 20ꢀ22 min, 95ꢀ95% ACN in 22ꢀ25 min, 95ꢀ20%
ACN in 25ꢀ30 min.
Method B
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
0ꢀ70% ACN in 0ꢀ20 min, 70ꢀ95% ACN in 20ꢀ22 min, 95ꢀ95% ACN in 22ꢀ25 min, 95ꢀ30%
ACN in 25ꢀ30 min.
Method C
20ꢀ70% ACN in 0ꢀ20 min, 70ꢀ95% ACN in 20ꢀ22 min, 95ꢀ95% ACN in 22ꢀ25 min, 95ꢀ20%
ACN in 25ꢀ30 min.
Method D
40ꢀ90% ACN in 0ꢀ20 min, 90ꢀ95% ACN in 20ꢀ22 min, 95ꢀ95% ACN in 22ꢀ25 min, 95ꢀ40%
ACN in 25ꢀ30 min.
Method E
20ꢀ60% ACN in 0ꢀ20 min, 60ꢀ95% ACN in 20ꢀ22 min, 95ꢀ95% ACN in 22ꢀ25 min, 95ꢀ20%
ACN in 25ꢀ30 min.
General procedure 1 (GP1) for amide couplings: Representative experimental
procedure for the amide formation using primary amines.
To solution of acid (1 eq.) in anhydrous DCM was added Nꢀ(3ꢀdimethylaminopropyl)ꢀN′ꢀ
ethylcarbodiimide (EDC) hydrochloride (1.1 eq.) at 0 °C and stirred for 1h. The primary
amine (1.1 eq.) was added to the reaction mixture, and then it was warmed up to room
temperature and stirred for 14ꢀ18 h. A saturated solution of NaHCO
aqueous phase was extracted with DCM (3 times). The combined organic extracts were
washed with brine, dried over Na SO , filtered and evaporated in vacuo. The crude mixture
3
was added and the
2
4
ACS Paragon Plus Environment
1
7

Journal of Medicinal Chemistry
Page 18 of 87
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
was purified by flash column chromatography to afford the compounds 1, 4, 30, 32, 36, 37,
40, 41.
General procedure 2 (GP2) for amide couplings: Representative experimental
procedure for the amide formation using secondary amines.
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 a solution of acid (1.1 eq.) in anhydrous DCM were added N’ꢀ(3ꢀDimethylaminopropyl)ꢀ
N′ꢀethylcarbodiimide (EDC) hydrochloride (1.1 eq.) and N,Nꢀdiisopropylethylamine (1.5 eq.)
at 0 °C. The reaction mixture stirred for 30 min. The secondary amine (1 eq) was added, then
the reaction mixture warmed up to room temperature and stirred for 16ꢀ20 h. A saturated
solution of NaHCO
3
was added and the aqueous phase was extracted with DCM (3 times).
SO , filtered and
The combined organic extracts were washed with brine, dried over Na
2
4
evaporated in vacuo. The crude mixture was purified by flash column chromatography to
afford the compounds 11, 13ꢀ22, 24, 34, 35, 44, 45.
General procedure 3 (GP3) for hydrogenation: Representative experimental procedure
for the Cbz deprotection.
The protected amine (1 eq.) was dissolved in anhydrous methanol and the solution was
transferred to a dry schlenk flask containg Pd/C (0.2 eq.) under counter current at room
2
temperature. The reaction mixture stirred for 6 h under H atmosphere at room temperature.
Then the catalyst was removed by filtration (celite pad) and the solvent was evaporated in
vacuo. To afford the final compounds 2, 5, 31, 33, 28, 38, 39, 42, 43. No further purification
was required, however samples were purified by flash column chromatography for their
analysis.
General procedure 4 (GP4) for Boc deprotection: Representative experimental
procedure for the TFA deprotection.
ACS Paragon Plus Environment
1
8

Page 19 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The Boc protected amine was added in a solution of DCM/TFA (4:1). The reaction mixture
stirred for 3 h at room temperature. Then the reaction mixture evaporated in vacuo. The
3
precipitate was neutralized with a saturated solution of NaHCO and the aqueous phase was
extracted with DCM (3 times). The combined organic extracts were washed with brine, dried
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
over Na SO , filtered and evaporated in vacuo. The crude mixture was purified by
2
4
preparative HPLC to afford the compounds 25, 26, 28, 36, 46, 47 as TFA salts.
General procedure 5 (GP5) for Acylation: Representative experimental procedure for
the Acylation using heterocycles with the anhydride.
To solution of succinic anhydride (1 eq.) in anhydrous DCM was added the heterocycle base
(1 eq) and the reaction mixture was stirred for 2 h at room temperature. The DCM was
evaporated and the crude compound was recrystallized from Hex/EtOAc to afford the
compounds 6, 7, 8.
General procedure 6 (GP6) for Friedel Crafts reaction: Representative experimental
procedure for the Friedel Crafts using heterocycles with the anhydride.
3
To solution of succinic anhydride (1 eq.) in anhydrous DCM was added anhydrous AlCl (1.2
°
eq.) at 0 C and the reaction mixture was stirred for 2 h at 0 °C. Then the aromatic heterocycle
was added dropwise over a period of 1 h. The reaction mixture was stirred for 4 h at room
temperature. 50 ml of iceꢀcold H
to pH 2. The aqueous phase was extracted with DCM (3 times). The combined organic
extracts were dried over Na SO , filtered and evaporated in vacuo. The crude mixture was
2
O was added at 0 °C, followed by acidification with 2N HCl
2
4
purified by recrystallization or flash column chromatography to afford the compounds 9, 10.
Benzyl tertꢀbutyl {6ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ6ꢀoxohexaneꢀ1,5ꢀdiyl}(S)ꢀdicarbamate (1)
ACS Paragon Plus Environment
1
9

Journal of Medicinal Chemistry
Page 20 of 87
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
Compound 1 was prepared according to general procedure 1 (GP1). The crude mixture was
purified by flash column chromatography (DCM/MeOH: 99/1) to afford compound 1 (2.37 g,
1
4
.24 mmol, 80%) as a colorless solid. HꢀNMR (360 MHz, CDCl
3
): δ (ppm) 7.28ꢀ7.14 (m,
15H), 5.94 (br s, 1H), 5.01 (s, 2H), 4.92 (s, 1H), 4.71 (s, 1H), 4.10 (t, J = 8.0 Hz, 1H), 3.90ꢀ
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
13
3.82 (m, 3H), 3.05ꢀ3.02 (m, 2H), 1.63ꢀ1.56 (m, 1H), 1.38ꢀ1.13 (m, 14H). CꢀNMR (90 MHz,
CDCl , rotamers were observed): δ (ppm) 170.8 (C=O), 155.5 (C=O), 154.5 (C=O), 140.7 (2
x C ), 135.6 (C ), 127.73 and 127.73 (4 x CH, rotamers), 127.5 (2 x CH), 127.1 (2 x CH),
27.0 (4 x CH), 125.9 (CH), 125.5 (CH), 79.0 (C ), 65.7 (CH ), 53.3 (CH), 49.6 (CH), 42.7
CH ), 39.4 (CH ), 31.0 (CH ), 28.4 (CH ), 27.3 (3 x CH ), 21.3 (CH ). MS (EI) m/z (%):
3
q
q
1
q
2
(
2
2
2
2
3
2
+
559 (3, M ), 307 (24), 263(10), 218 (15), 180 (35), 174 (13), 167 (29), 108 (10), 91 (100), 84
(19), 79 (13), 59 (12), 57 (43.4), 56 (14), 44 (12), 41 (31), 39 (12), 30 (26), 28 (11) . MS (EI)
+
m/z calcd for C33
H
41
N
O
3 5
[M] , 559.3; found, 559.3. HPLC purity, 98%, t
R
= 22.3 min,
25
gradient A. [α]
D
ꢀ19.0 (c = 0.97, MeOH).
tertꢀButyl (S)ꢀ{6ꢀaminoꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ1ꢀoxohexanꢀ2ꢀyl}carbamate (2)
Compound 2 was prepared according to general procedure 3 (GP3). Compound 2 (1.82 g,
1
4
.28 mmol, 99%) was afforded as a colorless solid. HꢀNMR (600 MHz, CDCl
3
): δ (ppm)
7.23ꢀ7.17 (m, 4H), 7.16ꢀ7.13 (m, 6H), 6.29 (s, 1H), 4.96 (s, 1H), 4.11 (t, J = 8.0 Hz, 1H),
3.91ꢀ3.75 (m, 3H), 2.59ꢀ2.53 (m, 2H), 1.65ꢀ1.61 (m, 1H), 1.40ꢀ1.27 (m, 12H), 1.18ꢀ1.14 (m,
1
3
2
H). CꢀNMR (150 MHz, CDCl , rotamers were observed): δ (ppm) 170.9 (C=O), 154.5
3
(
C=O), 140.74 (C
q
), 140.72 (C
25.8 (CH), 78.9 (C ), 53.3 (CH), 49.6 (CH), 42.7 (CH
), 27.3 (3 x CH ), 21.5 (CH
q
), 127.71 and 127.68 (4 x CH, rotamers), 127.0 (4 x CH),
1
q
2
), 40.5 (CH
2
2
), 31.5 (CH ), 28.7
+
+
(CH
2
3
2
). MS (EI) m/z (%): 425 (12, M ), 258 (14, M ꢀC13
H11), 202
(18), 181 (11), 180 (25), 173 (14), 167 (27), 165 (18), 145 (13), 139 (15), 129 (12), 84 (100),
+
8
2 (18), 57 (50), 56 (18), 41 (26), 30 (36). MS (EI) m/z calcd for C H N O [M+H] , 426.2;
25 36 3 3
ACS Paragon Plus Environment
2
0

Page 21 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
found, 426.2. LCꢀMS (ESI) m/z calcd. for C25
36
H N
3
O
3
[M+H] , 426.3; found, 426.3. HPLC
24
purity, 95%, t = 11.0 min, gradient B. [α] ꢀ22.5 (c = 0.67, MeOH).
R D
(
S)ꢀ2ꢀ[(benzyloxy)carbonyl]ꢀ1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxylic acid (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
To a 1 M aqueous solution of NaOH (3.20 g, 80.0 mmol in 80 mL of H O) at room
2
temeparture were added (S)ꢀ1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxylic acid (5.00 g, 28.2)
and dioxane slowly until the suspension dissolves. Then benzyl chloroformate was added
(6.25 g, 5.59 ꢁL, 36.7 mmol) dropwise over a period of 30 min at 0 °C. The reaction mixture
was stirred for 16 h at room temperature. Dioxane was removed by evaporation and reaction
mixture was acidified to pH 2 with 1M HCl at 0 °C. The aqueous layer was extracted with
EtOAc. The combined organic layers were dried over Na
2
SO
4
, filtered and evaporated in
1
vacuo to afford the title compound 3 (8.68, 27.9 mmol, 99%) as a colorless foam. HꢀNMR
(
360 MHz, CDCl
4.91 (t, J = 4.9 Hz, 3H), 4.70 (d, J = 16.3 Hz), 4.50 (m, 1H), 3.22ꢀ3.07 (m, 2H). CꢀNMR
150 MHz, CDCl , rotamers were observed): δ (ppm) 175.4 and 175.0 (C=O, rotamers),
3
, rotamers were observed): δ (ppm) 7.32ꢀ6.99 (m, 9H), 5.19ꢀ5.10 (m) and
1
3
(
3
1
55.4 and 154.4 (C=O, rotamers), 135.2 (C
rotamers), 127.5 (2 x CH), 127.2 and 127.1 (CH, rotamers), 127.0 and 126.9 (2 x CH,
rotamers), 126.1 (CH), 126.0 (CH), 125.2 (CH), 66.8 and 66.6 (CH , rotamers), 52.4 and 52.0
(CH, rotamers), 43.5 and 43.3 (CH , rotamers), 30.2 and 29.8 (CH , rotamers). MS (EI) m/z
q q q
), 131.9 (C ), 131.2 (C ), 130.4 and 130.3 (CH,
2
2
2
+
+
(
%): 222 (14, [M+H] ꢀ C H NO ), 220 (11, M ꢀC H ), 176 (53), 130 (23), 104 (11), 91
11
11
4
7
7
+
(100). MS (EI) m/z calcd for C18
H
17NO
4
[M] , 311.1; found, 311.1. LCꢀMS (ESI) m/z calcd.
+
for C18
H
17NO
4
Na [M+Na] , 334.1; found, 334.6. HPLC purity, 96%, t
R
= 20.5 min, Gradient
24
A. [α]
D
+23.6 (c = 1.03, MeOH).
ACS Paragon Plus Environment
2
1

Journal of Medicinal Chemistry
Page 22 of 87
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
Benzyl (S)ꢀ3ꢀ{[(S)ꢀ5ꢀ((tertꢀbutoxycarbonyl)amino)ꢀ6ꢀ((2,2ꢀdiphenylethyl)amino)ꢀ6ꢀoxohexyl]
carbamoyl}ꢀ3,4ꢀdihydroisoquinolineꢀ2(1H)ꢀcarboxylate (4)
Compound 4 was prepared according to general procedure 1 (GP1). The crude solid mixture
was purified by flash column chromatography (DCM/MeOH: 97/3) to afford compound 4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
(1.2 g, 1.8 mmol, 51%) as a colorless solid. HꢀNMR (600 MHz, DMSOꢀd , 400 K): δ (ppm)
6
7
.34ꢀ7.11 (several m, 19H), 5.96 (br s, 1H), 5.15ꢀ5.10 (m, 2H), 4.60ꢀ4.65 (m, 2H), 4.50 (d, J
15.8 Hz, 1H), 4.20 (t, J = 8.0 Hz, 1H), 3.77ꢀ3.74 (m, 2H), 3.69ꢀ3.64 (m, 1H), 3.08 (d, J =
.5 Hz, 2H), 2.90ꢀ2.86 (m, 2H), 1.44ꢀ1.42 (m, 1H), 1.26ꢀ1.23 (m, 10H), 1.20ꢀ1.13 (2H), 1.05ꢀ
=
5
0
13
.99 (2H). CꢀNMR (150 MHz, CDCl , rotamers were observed): δ (ppm) 171.0 (C=O),
3
169.9 (C=O), 154.6 (C=O), 140.7 (2 x C ), 135.1 (C ), 132.7 (C ), 127.69 and 127.66 (4 x
q
q
q
CH, rotames), 127.6 (2 x CH), 127.3 (2 x CH), 127.1 (2 x CH), 127.0 (4 x CH), 125.8 (2 x
CH), 79.0 (C ), 66.9 (CH ), 55.4 (CH) 53.3 (CH), 49.6 (CH), 44.2 (CH ), 42.7 (CH ), 37.4
CH ), 30.3 (CH ). LCꢀMS (ESI) m/z calcd. for
q
2
2
2
(
2
2
2 3 2
), 27.8 (CH ), 27.3 (3 x CH ), 21.1 (CH
+
C H N O [M+H] , 719.4; found, 720.2. HPLC purity, 96%, t = 22.2 min, gradient A.
43
51
4
6
R
2
4
[
α] ꢀ18.7 (c = 0.51, MeOH).
D
tertꢀButyl {(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ1ꢀoxoꢀ6ꢀ[(S)ꢀ1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀ
carboxamido]hexanꢀ2ꢀyl}carbamate (5)
Compound 5 was prepared according to general procedure 3 (GP3). Compound 5 (2.0 g, 3.4
1
mmol, 98%) was afforded as a colorless solid. HꢀNMR (360 MHz, CDCl ): δ (ppm) 7.24ꢀ
3
7.07 (m, 15H + CDCl
3
, H1ꢀ3, 6, 34ꢀ43), 6.97 (br s, 1H, H21), 6.08 (br s, 1H, H26), 4.97 (s, 1H),
4.12 (t, J = 8.0 Hz, 1H, H23), 3.93ꢀ3.75 (m, 5H, H8, 10, 22), 3.46 (dd, J = 5.1 Hz, J = 10.6 Hz,
1H, H18), 3.17ꢀ3.12 (m, 3H, H7, 14a), 2.74 (dd, J = 10.4 Hz, J = 16.3 Hz, 1H, H14b), 1.59 (m,
13
1
H, H ), 1.41ꢀ1.35 (m, 3H, H15, 17b), 1.31 (s, 9H, H31ꢀ33), 1.20ꢀ1.12 (m, 2H, H ). CꢀNMR
1
7
a
1
6
(
150 MHz, CDCl ): δ (ppm) 172.1 (C=O), 170.9 (C=O), 155.9 (C=O), 140.7 (2 x C ), 134.8
3
q
ACS Paragon Plus Environment
2
2

Page 23 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(C
q
), 133.3 (C
q
), 128.1 (CH), 127.7 (4 x CH), 127.0 (4 x CH), 125.8 (2 x CH), 125.6 (CH),
125.1 (CH), 124.5 (CH), 78.9 (C
q
), 55.5 (CH), 53.3 (CH), 49.6 (CH
2
), 46.4 (CH), 42.7
(CH
2
), 37.2 (CH
2
), 30.8 (CH
2
), 30.0 (CH
2
), 28.0 (CH
2
), 27.2 (3 x CH
3
), 21.4 (CH ). HRMS
2
+
(ESI) m/z calcd for C H N O [M+H] , 585.3435; found, 585.3436. MS (EI) m/z (%): 584
35
45
4
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
+
(3, M ), 133 (14, C H N), 132 (100, C H N), 131 (20), 130 (11), 84 (11), 59 (32), 57 (12),
9
11
9
10
+
4
44 4 4 R
1 (17). MS (EI) m/z calcd for C35H N O [M] , 584.3; found, 584.3. HPLC purity, 97%, t
24
=
19.5 min, gradient A. [α]
D
ꢀ38.0 (c = 0.74, MeOH).
2
7
4
ꢀoxoꢀ4ꢀ(piperidinꢀ1ꢀyl)butanoic acid (6)
Compound 6 was prepared according to general procedure 5 (GP5). The crude solid mixture
was purified by recrystallization (Hex/EtOAc: 1/2) to afford compound 6 (2.6 g, mmol,
1
9
4%) as a colorless solid. HꢀNMR (600 MHz, CDCl
3
): δ (ppm) 3.50 (t, J = 6.0 Hz, 2H), 3.36
13
(
t, J = 6.0 Hz, 2H), 2.62ꢀ2.61 (m, 4H), 1.60ꢀ1.48 (m, 6H). CꢀNMR (90 MHz, CDCl
3
): δ
(ppm) 176.4 (C=O), 170.7 (C=O), 46.0 (CH ), 43.4 (CH ), 30.4 (CH ), 28.5 (CH ), 26.6
2
2
2
2
(
CH ), 25.8 (CH ), 24.7 (CH ).
2 2 2
2
8
4
ꢀmorpholinoꢀ4ꢀoxobutanoic acid (7)
Compound 7 was prepared according to general procedure 5 (GP5). The crude solid mixture
was purified by recrystallization (Hex/EtOAc: 1/3) to afford compound 7 (2.05 g, mmol,
1
9
3
1
5%) as light yellow crystals. HꢀNMR (360 MHz, CDCl ): δ (ppm) 3.64ꢀ56 (m, 6H), 3.44ꢀ
3
13
.42 (m, 2H), 2.67ꢀ2.62 (m, 2H), 2.59ꢀ2.55 (m, 2H). CꢀNMR (90 MHz, CDCl
3
): δ (ppm)
77.0 (C=O), 170.5 (=CO), 66.9 (CH
).
2
), 66.6 (CH
2
), 46.0 (CH
2
), 42.4 (CH
2
), 29.4 (CH ), 27.9
2
(CH
2
27x
4
ꢀ(4ꢀmethylpiperazinꢀ1ꢀyl)ꢀ4ꢀoxobutanoic acid (8)
ACS Paragon Plus Environment
2
3

Journal of Medicinal Chemistry
Page 24 of 87
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
Compound 8 was prepared according to general procedure 5 (GP5). The crude product was
purified by recrystallization (Hex/EtOAc/EtOH: 2/4/1) to afford the titled compound 8 (1.71
1
g, 8.54 mmol, 86%) as yellowish solid. HꢀNMR (360 MHz, CDCl
3
): δ (ppm) 10.69 (br s,
13
1
H), 3.62 (t, J = 6.0 Hz, 2H), 3.52 (t, J = 6.0 Hz, 2H), 2.57ꢀ2.44 (m, 8H), 2.30 (s, 3H). Cꢀ
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 (90 MHz, CDCl ): δ (ppm) 175.2 (C=O), 169.5 (C=O), 66.9 (CH ), 53.2 (CH ), 52.3
3
2
2
(CH
2
), 44.0 (CH
3
), 43.4 (CH
2
), 39.8 (CH
2
2 2
), 28.9 (CH ), 27.2 (CH ).
2
9
4ꢀoxoꢀ4ꢀ(thiophenꢀ2ꢀyl)butanoic acid (9)
Compound 9 was prepared according to general procedure 9 (GP6). The crude product was
purified by recrystallization (H O) to afford compound 9 (1.38 mg, 7.5 mmol, 63%) as a light
2
1
yellow solid. HꢀNMR (600 MHz, CDCl
3
): δ (ppm) 7.70 (dd, J = 1.1, J = 3.8, 1H), 7.58 (dd, J
=
1.1, J = 4.9, 1H), 7.07 (dd, J = 3.8, J = 4.9, 1H), 3.19 (t, J = 6.8 Hz, 2H), 2.74 (t, J = 6.7
1
3
3 q
Hz, 2H). CꢀNMR (150 MHz, CDCl ): δ (ppm) 190.9 (C=O), 178.5 (C=O), 143.7 (C ),
134.0 (CH), 132.3 (CH), 128.4 (CH), 33.9 (CH), 28.2 (CH).
30
4ꢀ(furanꢀ2ꢀyl)ꢀ4ꢀoxobutanoic acid (10)
Compound 10 was prepared according to general procedure 6 (GP6). The crude product was
purified by flash column chromatography (DCM/MeOH/AcOH: 96/4/0.1) to afford
1
compound 10 (263 mg, 1.6 mmol, 57%) as a light yellow solid. HꢀNMR (600 MHz, CDCl ):
3
δ (ppm) 7.52 (dd, J = 0.8, J = 1.7, 1H), 7.16 (dd, J = 0.8, J = 3.6, 1H), 6.48 (dd, J = 1.7, J =
13
3
.6, 1H), 3.11 (t, J = 6,7 Hz, 2H), 2.73 (t, J = 6.7 Hz, 2H). CꢀNMR (150 MHz, CDCl
3
): δ
q
(ppm) 187.2 (C=O), 178.3 (C=O), 152.5 (C ), 146.6 (CH), 117.3 (CH), 112.5 (CH), 33.0
(CH), 27.7 (CH).
ACS Paragon Plus Environment
2
4

Page 25 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
tertꢀButyl
{(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ1ꢀoxoꢀ6ꢀ[(S)ꢀ2ꢀ(4ꢀoxoꢀ4ꢀphenylbutanoyl)ꢀ
1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxamido]hexanꢀ2ꢀyl}carbamate (11)
Compound 11 was prepared according to general procedure 2 (GP2) by using benzoyl
propionic acid. The crude solid mixture was purified by flash column chromatography
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
(Hex/EtOAc/EtOH: 8/3/1) to afford Compound 11 (89 mg, 0.12 mmol, 70%) as a colorless
1
solid. HꢀNMR (600 MHz, DMSOꢀd
6
, 400 K): δ (ppm) 7.94ꢀ7.92 (m, 2H, H44 and H48), 7.63ꢀ
ꢀ H , H
), 4.80 (d, J = 15.9 Hz, 1H, H10a), 4.57
7
.58 (m, 1H, H46), 7.50ꢀ7.47 (m, 2H, H45 and H47), 7.24ꢀ7.13 (m, 15H, H
1
3
6
, H27, H33ꢀ
H
42), 5.96ꢀ5.95 (br s, 1H, H25, H19), 4.91 (br s, 1H, H
8
(
br s, 1H, H10b), 4.19 (t, J = 7.7 Hz, 1H, H ), 3.76ꢀ3.68 (m, 2H, H and H ), 3.66ꢀ3.64 (m,
30 29a 24
1H, H29b), 3.27 (t, J = 6.4 Hz, 2H, H ), 3.18 (dd, J = 4.0 Hz, J =15.7 Hz, 1H, H ), 3.06 (br s,
16
7a
1H, H7b), 2.96ꢀ2.75 (m, 4H + H
2
O, H15 and H20), 1.44ꢀ1.39 (m, 1H, H23a), 1.38 (s, 9H, H53ꢀ
1
H
55), 1.36ꢀ1.29 (m, 1H, H23b), 1.23ꢀ1.16 (m, 2H, H21), 1.10ꢀ1.00 (m, 2H, H22). HꢀNMR (600
3
MHz, CDCl ): δ (ppm) 7.93ꢀ7.90 (m, 2H), 7.52ꢀ7.51 (m, 1H, ), 7.42ꢀ7.39 (m, 2H), 7.21ꢀ7.10
(m, 15H), 6.64 (t, J = 5.9 Hz, rotamer 1) and 6.13 (t, J = 5.9 Hz, 1H, rotamer 2), 6.45 (br s,
rotamer 1) and 6.25 (br s, 1H, rotamer 2), 5.15 (br s, 1H), 5.00ꢀ4.53 (several m, 3H), 4.09ꢀ
4.04 (m, 1H), 3.85ꢀ3.60 (m, 4H), 3.49ꢀ3.41 (m, 1H), 3.29ꢀ2.71 (several m, 5H), 2.56ꢀ2.50 (m,
1
3
1
H), 1.49ꢀ1.41 (m, 1H), 1.32 (s, 9H), 1.28ꢀ0.93 (several m, 5H). CꢀNMR (150 MHz,
, 350 K, rotamers were observed): δ (ppm) 172.5 (C=O), 143.3 (2 x C
137.6 (C ), 133.4 (C , C ), 128.8 (4 x CH), 128.42 and 128.40 (4 x CH, rotamers), 128.3
DMSOꢀd
6
q
, C31ꢀ32),
q
q
46
(
CH), 126.8 (2 x CH), 126.5 (CH), 78.6 (C ), 50.7 (CH, C ), 43.8 (CH , C ), 34.0 (CH ,
q
30
2
29
2
C
C
16), 32.3 (CH
2
, C21), 29.2 (CH
2
, C23), 28.7 (3 x CH
3
, C53ꢀ55), 28.2 (CH
2
, C15), 23.0 (CH
2
,
22). Four C=O, One C
q
2
, two CH, four aromatic CH and three CH groups were not detected
due to peak broadening or overlapping caused by the high temperature. All protons and
carbons have been assigned by performing 2D NMR analysis [COSY (400 K), HSQC (350
K) and HMBC (350 K)], where the assigned structure and 2D spectra of compound 11 can be
ACS Paragon Plus Environment
2
5

Journal of Medicinal Chemistry
Page 26 of 87
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
+
52 4 6
found in supporting information. HRMS (ESI) m/z calcd for C45H N O Na [M+Na] ,
25
7
67.3779; found, 767.3789. HPLC purity, 100%, t
.69, MeOH).
R D
= 22.7min, gradient A. [α] ꢀ23.5 (c =
0
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
tertꢀButyl{(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ1ꢀoxoꢀ6ꢀ[(S)ꢀ2ꢀ(4ꢀoxopentanoyl)ꢀ1,2,3,4ꢀ
tetrahydroisoquinolineꢀ3ꢀcarboxamido]hexanꢀ2ꢀyl}carbamate (12)
Compound 12 was prepared according to general procedure 2 (GP2) by using 4ꢀ
oxopentanoic acid. The crude solid mixture was purified by flash column chromatography
(
Hex/EtOAc/EtOH: 6/3/1) to afford compound 12 (45 mg, 0.066 mmol, 78%) as a colorless
1
solid. HꢀNMR (600 MHz, DMSOꢀd
6
, 350 K): δ (ppm) 7.33ꢀ7.19 (m, 15H, H1ꢀ3, 6, 21, 34ꢀ43),
), 4.80 (d, J = 16.0 Hz, 1H, H10a), 4.60 (br s, 1H,
10b), 4.26 (t, J = 8.0 Hz, 1H, H23), 3.84ꢀ3.79 (m, 2H, H18, 22a), 3.75ꢀ3.71 (m, 1H, H22b), 3.27ꢀ
.16 (m, 2H, H47), 3.08ꢀ2.96 (m, 2H, H ), 2.76ꢀ2.57 (m, 4H, H14, 45), 2.16 (s, 3H, H49), 1.51ꢀ
.41 (m, 10H, H31ꢀ33, 17a), 1.37ꢀ1.30 (m, 1H, H17b), 1.27ꢀ1.21 (m, 2H, H ), 1.11ꢀ1.04 (m, 2H,
6.02 (br s, 1H, H12, 26), 4.93 (br s, 1H, H
8
H
3
7
1
15
13
H
16). CꢀNMR (150 MHz, DMSOꢀd
6
, 350 K, rotamers were observed): δ (ppm) 172.4
), 128.76 and 128.73 (4 x CH,
rotamers), 128.70 (CH), 128.33 and 128.31 (4 x CH, rotamers), 126.7 (2 x CH), 126.4 (CH),
(C=O), 170.5 (C=O), 167.9 (C=O), 157.3 (C=O), 143.2 (2 x C
q
7
q 2 2 2 2 2
8.5 (C ), 50.6 (CH), 50.0 (CH), 47.9 (CH ), 43.7 (CH ), 38.3 (CH ), 32.2 (CH ), 30.4 (CH ),
2
9.1 (CH ), 28.7 (3 x CH ), 22.8 (CH ). One C=O, two C , one CH, two aromatic CH and
3
3
2
q
three CH
2
groups were not detected due to peak broadening or overlapping caused by the
+
high temperature. HRMS (ESI) m/z calcd for C40
05.3625. LCꢀMS (ESI) m/z calcd. for C40
purity, 95%, t = 12.9 min, gradient B.
H
50
N
4
O
6
Na [M+Na] , 705.3623; found,
+
7
H
51
N
4
O
6
[M+H] , 683.4; found, 683.7. HPLC
R
ACS Paragon Plus Environment
2
6

Page 27 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
tertꢀButyl
{(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ6ꢀ[(S)ꢀ2ꢀ(4ꢀ(4ꢀfluorophenyl)ꢀ4ꢀoxobutanoyl)ꢀ
1,2,3,4 ꢀtetrahydroisoquinolineꢀ3ꢀcarboxamido]ꢀ1ꢀoxohexanꢀ2ꢀyl}carbamate (13)
Compound 13 was prepared according to general procedure 2 (GP2) by using 3ꢀ(4ꢀ
fluorobenzoyl)propionic acid. The crude solid mixture was purified by flash column
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
chromatography (Hex/EtOAc/EtOH: 6/3/1) to afford compound 13 (39 mg, 0.05 mmol, 60%)
1
as a colorless solid. HꢀNMR (600 MHz, DMSOꢀd
6
, 400 K): δ (ppm) 8.09 ꢀ8.07 (m, 2H, H51,
), 4.86
d, J = 15.9 Hz, 1H, H10a), 4.64 (br s, 1H, H10b), 4.26 (t, J = 7.7 Hz, 1H, H23), 3.84ꢀ3.80 (m,
H, H18, 22a), 3.72 (m, 1H, H22b), 3.33 (t, J = 7.2 Hz, 2H, H ), 3.25 (dd, J = 4.0 Hz, J =15.7
5
5 8
), 7.29ꢀ7.20 (m, 16H, H1ꢀ3, 6, 21, 34ꢀ43, 52, 54), 6.03 (br s, 1H, H12, 26), 4.97 (br s, 1H, H
(
2
47
Hz, 1H, H ), 3.12 (br s, 1H, H ), 2.95ꢀ2.88 (m, 4H + H O, H14, 45), 1.44ꢀ1.31 (m, 11H, H17,
7a
7b
2
1
3
3
1ꢀ33), 1.29ꢀ1.20 (m, 2H, H15), 1.11ꢀ1.05 (m, 2H, H16). CꢀNMR (150 MHz, DMSOꢀd
K): δ (ppm) 158.9 (C=O) 143.3 (2 x C ), 140.0 (C ), 131.3 (2 x C ), 128.8 (4 x CH), 128.4 (4
x CH), 126.8 (2 x CH), 126.5 (CH), 116.0 (2 x CH), 55.1 (CH), 50.7 (CH), 43.8 (CH ), 33.9
(CH ), 32.3 (CH ), 31.6 (CH ), 29.2 (CH ), 28.7 (3 x CH ), 23.0 (CH ). Four C=O, One CF,
6
, 350
q
q
q
2
2
2
2
2
3
2
One CH, for aromatic CH and three CH groups were not detected due to peak broadening or
2
4 6
overlapping caused by the high temperature. HRMS (ESI) m/z calcd for C45H53FN O
+
24
[
M+H] , 763.3865; found, 763.3865. HPLC purity, 99%, t
R
= 21.8 min, gradient A. [α]
D
–
34.0 (c = 0.5, MeOH).
tertꢀButyl {(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ6ꢀ[(S)ꢀ2ꢀ(4ꢀ(4ꢀmethoxyphenyl)ꢀ4ꢀoxobutanoyl)ꢀ
1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxamido]ꢀ1ꢀoxohexanꢀ2ꢀyl}carbamate (14)
14
was prepared according to general procedure (GP2) by using 3ꢀ(4ꢀ
2
methoxybenzoyl)propionic acid. The crude solid mixture was purified by flash column
chromatography (Hex/EtOAc/EtOH: 6/3/1) to afford 14 (39 mg, 0.05 mmol, 58%) as a
1
colorless solid. HꢀNMR (600 MHz, DMSOꢀd , 400 K): δ (ppm) 7.95ꢀ7.94 (m, 2H, H51, 55),
6
ACS Paragon Plus Environment
2
7

Journal of Medicinal Chemistry
Page 28 of 87
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
7.28ꢀ7.16 (m, 15H, H1ꢀ3, 6, 21, 34ꢀ43), 7.04ꢀ7.02 (m, 2H, H52, 54), 5.98 (br s, 1H, H12, 26), 4.94 (br
s, 1H, H
.87 (s, 3H, H57), 3.80ꢀ3.77 (m, 2H, H18, 22a), 3.71ꢀ3.68 (m, 1H, H22b), 3.25 (t, J = 7.2 Hz, 2H,
H ), 3.21 (dd, J = 4.0 Hz, J =15.7 Hz, 1H, H ), 3.08 (br s, 1H, H ), 2.94ꢀ2.75 (m, 4H +
8
), 4.83 (d, J = 16.0 Hz, 1H, H10a), 4.60 (br s, 1H, H10b), 4.22 (t, J = 8.0 Hz, 1H, H23),
3
4
7
7a
7b
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H O, H14, 45), 1.44ꢀ1.27 (m, 11H, H17, 31ꢀ33), 1.26ꢀ1.19 (m, 2H, H ), 1.08ꢀ1.01(m, 2H, H ).
2
15
16
1
3
CꢀNMR (150 MHz, DMSOꢀd
6
, 350 K, rotamers were observed): δ (ppm) 171.6 (C=O),
), 133.0 (C ), 129.7 (2 x CH), 129.5
), 127.9 and 127.87 (4 x CH, rotamers), 127.47 and 127.45 (4 x CH, rotamers), 125.8 (2 x
CH), 125.6 (CH), 113.6 (2 x CH), 77.7 (C ), 55.1 (CH ), 49.8 (CH), 42.8 (CH ), 32.7 (CH ),
1
69.6 (C=O), 162.8 (CꢀOMe), 154.6 (C=O), 142.3 (2 x C
q
q
(C
q
q
3
2
2
31.4 (CH ), 28.2 (CH ), 27.8 (3 x CH ), 27.7 (CH ), 22.0 (CH ). Two C=O, one C , four
2
2
3
2
2
q
2
aromatic CH, two CH and three CH groups were not detected due to peak broadening or
54 4 7
overlapping caused by the high temperature. HRMS (ESI) m/z calcd for C46H N O Na
+
22
[
M+Na] , 797.3885; found, 797.3888. HPLC purity, 99%, t
R
= 21.8 min, gradient A. [α]
D
ꢀ
23. 5 (c = 0.46, MeOH).
tertꢀButyl
oxobutanoyl)ꢀ1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxamido]ꢀ1ꢀoxohexanꢀ2ꢀyl}carbamate
15)
{(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ6ꢀ[(S)ꢀ2ꢀ(4ꢀ(4ꢀ(methylsulfonyl)phenyl)ꢀ4ꢀ
(
Compound 15 was prepared according to general procedure 2 (GP2) by using 3ꢀ(4ꢀ
methylsulfonylbenzoyl)propionic acid. The crude solid mixture was purified by flash column
chromatography (Hex/EtOAc/EtOH: 3/3/1) to afford compound 15 (56 mg, 0.068 mmol,
1
8
0%) as a colorless solid. HꢀNMR (600 MHz, DMSOꢀd
6
, 400 K): δ (ppm) 8.21ꢀ8.19 (m, 2H,
H
51, 55), 8.11ꢀ8.09 (m, 2H, H52ꢀ54), 7.29ꢀ7.06 (m, 15H, H1ꢀ3, 6, 21, 34ꢀ43), 6.03 (br s, 1H, H12, 26),
4.96 (br s, 1H, H ), 4.86 (d, J = 16.0 Hz, 1H, H ), 4.64 (br s, 1H, H10b), 4.26 (t, J = 8.0 Hz,
8 10a
1
H, H ), 3.86ꢀ3.81 (m, 2H, H18, 22a), 3.74ꢀ3.71 (m, 1H, H22b), 3.38 (t, J = 7.1 Hz, 2H, H47),
23
ACS Paragon Plus Environment
2
8

Page 29 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
3.23 (m, 4H, H7a, 57), 3.10 (br s, 1H, H7b), 2.98ꢀ2.86 (m, 4H + H O, H14, 45), 1.40ꢀ1.30 (m,
13
1
1H, H17, 31ꢀ33), 1.28ꢀ1.22 (m, 2H, H15), 1.11ꢀ1.06 (m, 2H, H16). CꢀNMR (150 MHz,
6
DMSOꢀd , 350 K, rotamers were observed): δ (ppm) 199.3 (C=O), 172.5 (C=O), 171.9,
1
71.6, 155.6 (C=O), 145.0 (CS), 143.3 (2 x C ), 141.3 (C ), 129.3 (2 x CH), 128.87 and
q q
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
128.84 (4 x CH, rotamers), 128.43 and 128.41 (4 x CH, rotamers), 127.8 (2 x CH), 126.8 (2 x
CH), 126.5 (CH) , 126.3 (CH), 126.1 (CH), 78.7 (C ), 56.7 (CH), 55.1 (CH), 50.7 (CH), 47.3
CH ), 44.0 (CH ) 43.8 (CH ), 38.8 (CH ), 34.4 (CH ) , 32.4 (CH ), 31.7 (CH ), 29.3 (CH ),
8.8 (3 x CH ), 28.4 (CH ), 23.1 (CH ). Two C and two aromatic CH groups were not
q
(
2
3
2
2
2
2
2
2
2
3
2
2
q
detected due to peak broadening or overlapping caused by the high temperature. HRMS (ESI)
+
m/z calcd for C H N O SNa [M+Na] , 845.3555; found, 845.3557. LCꢀMS (ESI) m/z
46
54
4
8
+
calcd. for C46
H
55
N
4
O
8 R
S [M+H] , 824.0; found, 824.4. HPLC purity, 96%, t = 21.0 min,
24
gradient A. [α]
D
ꢀ19.8 (c = 0.20, MeOH).
tertꢀButyl
{(S)ꢀ1ꢀ[(2,2ꢀdiphenylethyl)amino]ꢀ1ꢀoxoꢀ6ꢀ[(S)ꢀ2ꢀ(4ꢀoxoꢀ4ꢀ(piperidinꢀ1ꢀ
yl)butanoyl)ꢀ1,2,3,4ꢀtetrahydroisoquinolineꢀ3ꢀcarboxamido]hexanꢀ2ꢀyl}carbamate (16)
Compound 16 was prepared according to general procedure 2 (GP2) by using compound 6.
The crude product was purified by flash column chromatography (DCM/MeOH: 97/3) to
1
afford compound 16 (58 mg, 0.077 mmol, 91%) as a colorless solid. HꢀNMR (600 MHz,
DMSOꢀd , 400 K): δ (ppm) 7.26ꢀ7.12 (m, 15H, H1ꢀ3, 6, 21, 34ꢀ43), 5.94 (br s, 1H, H12, 26), 4.95
6
(
br s, 1H, H ), 4.79 (d, J = 15.9 Hz, 1H, H ), 4.55 (br s, 1H, H10b), 4.19 (t, J = 7.7 Hz, 1H,
8 10a
H
23), 3.80ꢀ3.72 (m, 2H, H18, 22a), 3.68ꢀ3.64 (m, 1H, H22b), 3.41 (m, 4H, H51, 55), 3.22 (dd, J =
4.0 Hz, J =15.7 Hz, 1H, H7a), 2.99 (br s, 1H, H7b), 2.92ꢀ2.83 (m, 2H, H47), 2.78ꢀ2.58 (m, 4H,
H
14, 45), 1.61ꢀ1.57 (m, 2H, H53), 1.49ꢀ1.44 (m, 4H, H52, 54), 1.41ꢀ1.34 (m, 10H, H17a, 31ꢀ33),
1
3
1
.31ꢀ1.23 (m, 1H, H17b), 1.20ꢀ1.15 (m, 2H, H ), 1.03ꢀ0.97 (m, 2H, H ). CꢀNMR (150
15 16
MHz, DMSOꢀd , 350 K, rotamers were observed): δ (ppm) 171.6 (C=O), 169.4 (C=O), 154.5
6
ACS Paragon Plus Environment
2
9