Discovery of Potent and Orally Bioavailable Macrocyclic Peptide-Peptoid Hybrid CXCR7 Modulators

Article abstract of DOI:10.1021/acs.jmedchem.7b01028

The chemokine receptor CXCR7 is an attractive target for a variety of diseases. While several small-molecule modulators of CXCR7 have been reported, peptidic macrocycles may provide advantages in terms of potency, selectivity, and reduced off-target activity. We produced a series of peptidic macrocycles that incorporate an N-linked peptoid functionality where the peptoid group enabled us to explore side-chain diversity well beyond that of natural amino acids. At the same time, theoretical calculations and experimental assays were used to track and reduce the polarity while closely monitoring the physicochemical properties. This strategy led to the discovery of macrocyclic peptide-peptoid hybrids with high CXCR7 binding affinities (K_i < 100 nM) and measurable passive permeability (P_app > 5 × 10^-6 cm/s). Moreover, bioactive peptide 25 (K_i = 9 nM) achieved oral bioavailability of 18% in rats, which was commensurate with the observed plasma clearance values upon intravenous administration.

Full text of DOI:10.1021/acs.jmedchem.7b01028

Subscriber access provided by University of Florida | Smathers Libraries
Article
Discovery of potent and orally bioavailable macrocyclic
peptide-peptoid hybrid CXCR7 modulators.
Markus Boehm, Kevin Beaumont, Rhys M Jones, Amit S. Kalgutkar, Liying Zhang, Karen Atkinson,
Guoyun Bai, Janice A. Brown, Heather Eng, Gilles H. Goetz, Brian R Holder, Bhagyashree Khunte,
Sarah Lazzaro, Chris Limberakis, Sangwoo Ryu, Michael J Shapiro, Laurie Tylaska, Jiangli
Yan, Rushia Turner, Siegfried S. F. Leung, Mahesh Ramaseshan, David A. Price, Spiros Liras,
Matthew P Jacobson, David J. Earp, R. Scott Lokey, Alan M Mathiowetz, and Elnaz Menhaji-Klotz
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01028 • Publication Date (Web): 18 Oct 2017
Downloaded from http://pubs.acs.org on October 19, 2017
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 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 37
1
2
3
4
5
6
7
8
9
Discovery of potent and orally bioavailable
macrocyclic peptideꢀpeptoid hybrid CXCR7
modulators
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
Markus Boehm,† Kevin Beaumont,† Rhys Jones,† Amit S. Kalgutkar,† Liying Zhang,⊥ Karen
Atkinson,‡ Guoyun Bai,‡ Janice A. Brown,‡ Heather Eng,‡ Gilles H. Goetz,‡ Brian R. Holder,‡
Bhagyashree Khunte,‡ Sarah Lazzaro,‡ Chris Limberakis,‡ Sangwoo Ryu,‡ Michael J. Shapiro,‡
‡
‡
#
§
,∥
Laurie Tylaska, Jiangli Yan, Rushia Turner, Siegfried S. F. Leung, Mahesh Ramaseshan,∥
David A. Price,† Spiros Liras,† Matthew P. Jacobson,§ David J. Earp,*,∥ R. Scott Lokey,*,# Alan
M. Mathiowetz,† and Elnaz Menhaji-Klotz*,†
Pfizer Worldwide Research & Development, †Cambridge, Massachusetts 02139, United States
and ‡Groton, Connecticut 06340, United States
#Chemistry and Biochemistry, University of California, Santa Cruz, California 95064, United
States
§Department of Pharmaceutical Chemistry, University of California, San Francisco, California
94158, United States
∥Circle Pharma, South San Francisco, California 94080, United States
ACS Paragon Plus Environment
1

Page 3 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
KEYWORDS: CXCR7 modulator, peptidic macrocycle, cyclic peptide, peptoid, peptideꢀpeptoid
hybrid, arginine replacement, tryptophan substitution, Nꢀmethylation, passive permeability,
polarity, lipophilicity, oral bioavailability.
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
ABSTRACT: The chemokine receptor CXCR7 is an attractive target for a variety of diseases.
While several small molecule modulators of CXCR7 have been reported, peptidic macrocycles
may provide advantages in terms of potency, selectivity, and reduced offꢀtarget activity. We
produced a series of peptidic macrocycles that incorporate an Nꢀlinked peptoid functionality
where the peptoid group enabled us to explore side chain diversity well beyond that of natural
amino acids. At the same time, computational calculations and experimental assays were used to
track and reduce polarity while closely monitoring physicochemical properties. This strategy led
to the discovery of macrocyclic peptideꢀpeptoid hybrids with high CXCR7 binding affinities (Ki
< 100 nM) and measurable passive permeability (Papp > 5 x 10ꢀ6 cm/sec). Moreover, bioactive
peptide 25 (Ki = 9 nM) achieved oral bioavailability of 18% in rats, which was commensurate
with the observed plasma clearance values upon intravenous administration.
INTRODUCTION
The chemokine receptor CXCR7 is a member of the Gꢀprotein coupled receptor (GPCR)
family that binds to its chemokine ligands CXCL11 and CXCL12 with high affinity. CXCL12,
also known as SDFꢀ1, has many physiological effects including acting as a stimulator of B cell
lymphopoiesis and bone marrow myelopoiesis.1,2 The first identified CXCL12ꢀbinding
chemokine GPCR was CXCR4, which received a great deal of attention due to its role in the
HIV replication cycle, acting as a coꢀreceptor for the entry of HIV into T cells.3 Further research
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 37
1
2
3
4
5
6
has identified pathways activated by CXCR4 including extracellular signalꢀrelated kinase and
protein kinase B leading to pleiotropic effects.
7
8
9
Unlike CXCR4, CXCR7 signals solely through βꢀarrestin in response to CXCL12 binding,
without noticeable activation of Gꢀprotein coupled pathways.4 Nonetheless, the molecular
pharmacology of CXCR7 is complex as it is known to homodimerize as well as heterodimerize
with CXCR4, affecting its internalization.5,6 It has also been reported to act as a scavenging
receptor for CXCL12, resulting in the modulation of CXCR4 chemokine axis.7
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
CXCR7 expression is tightly regulated and is enhanced during pathological processes such as
cancer and inflammation.8 It has been shown that CXCR7 plays an essential role in the
CXCL12/CXCR4ꢀmediated transendothelial migration of cancer cells.9 Recently it was reported
that CXCR7 modulation ameliorates atherosclerosis and promotes vascular remodeling through
increased cholesterol uptake in adipose tissue.10 CXCR7 has also been shown to act as a decoy
receptor for adrenomedullin, a peptide hormone playing a key role in cardiovascular
development.11 A Phase II clinical study evaluated the safety and efficacy of CXCL12 gene
therapy in heart failure patients,12 providing further impetus for the delivery of orally
bioavailable CXCR7 ligands that can modulate levels of CXCL12, as an attractive therapeutic
drug target.
The identification and development of small molecules for CXCR7 have been described in
several publications13ꢀ15 and patent applications.16ꢀ27 Since CXCR4 and CXCR7 share the same
chemokine ligand CXCL12, and peptide derivatives for CXCR4 have been successfully
developed,28ꢀ31 there is also strong interest in identifying peptides for modulating CXCR7
activity.32,33 Peptidic macrocycles in particular may provide advantages over small molecules in
terms of potency, selectivity, and reduced offꢀtarget activity.34,35
ACS Paragon Plus Environment
3

Page 5 of 37
Journal of Medicinal Chemistry
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
8
9
Compound Synthesis. All peptides were synthesized via standard Fmoc solid phase peptide
synthesis methods using the 2ꢀchloroꢀtritylchloride (CTC) resin.36 After resin cleavage, the
resulting acyclic peptideꢀpeptoid hybrids underwent macrolactamization under dilute solution
phase conditions to deliver either the cyclic sideꢀchain protected peptideꢀpeptoids or the final
cyclic peptideꢀpeptoids.37 The cyclic sideꢀchain protected peptideꢀpeptoids were globally
deprotected to afford the target compounds. All final compounds were purified using reversedꢀ
phase HPLC and analyzed to confirm > 95% purity.
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
CXCR7 Pharmacology. The binding affinity of peptides (given as Ki value) against CXCR7
was determined by their ability to displace radiolabeled 125IꢀCXCL12 from membranes obtained
from a cell line overexpressing the human CXCR7 receptor.38 The functional agonist activity of
peptides (given as EC50 value) against CXCR7 was determined by their ability to induce βꢀ
arrestin recruitment in a cell line overexpressing the human CXCR7 receptor.39
Passive Cell Permeability and EPSA. The rate of peptides to passively diffuse through cell
membranes was determined in a permeability assay using a lowꢀefflux MadinꢀDarby canine
kidney (
MDCKIIꢀLE) cell line.40 To guide the improvement of passive permeability, the polarity
of peptides was measured by EPSA, a supercritical fluid chromatography method.41,42 EPSA
values are analogous to calculated polar surface area (e.g. TPSA) values in medicinal chemistry
programs. It has been shown that EPSA values of less than 90 are a necessary requirement for
peptides to achieve measurable passive permeability.43 In contrast, while criteria for calculated
PSA to achieve oral bioavailability have been described in literature for small molecules (TPSA
< 140), consistent guidelines for the prospective design of peptides or macrocycles have not been
established.
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 37
1
2
3
4
5
6
7
8
9
Macrocyclic hexapeptide 1 (Figure 1) has been reported as a CXCR7 modulator with
functional activity in the low nanomolar range.44 Upon resynthesis its functional activity in the
CXCR7 βꢀarrestin assay was confirmed (EC50 = 46 nM), however, the binding affinity against
CXCR7 was relatively moderate (Ki = 2.1 µM). Furthermore, due to the presence of a basic
arginine moiety, the compound showed poor permeability (Papp < 1 x 10ꢀ6 cm/s) in the MDCKIIꢀ
LE permeability assay. In this work, we synthesized novel macrocyclic peptideꢀpeptoid hybrids
as an alternative to peptide 1, in order to identify compounds with improved binding affinity and
lower overall polarity, with the goal of increasing passive permeability toward the development
of orally bioavailable CXCR7 modulators.
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 1. Structures of peptide 1 and peptoid 2.
To help guide the optimization of binding affinity, a homology model of CXCR7 was
developed based on the crystal structure of CXCR4 in complex with a macrocyclic peptide
ligand (see Experimental Section).45 Induced fit docking of peptide 1 provided a hypothesis for
its bioactive conformation in the CXCR7 binding pocket (Figure 2) that suggested strategies to
improve binding affinity. The proposed docking model shows the tetrahydroꢀisoquinolineꢀ3ꢀ
carboxylic acid (TIC) group residing in a large hydrophobic pocket formed by Trp100, Phe124,
ACS Paragon Plus Environment
5

Page 7 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Leu128, Trp265, and Leu305, but without a tight contact for the TIC phenyl ring, indicating
additional space for enhancing hydrophobic complementarity and improving potency.
The implementation of peptoid residues offers the potential for modular diversification of
macrocyclic peptides due to broad availability of synthetic building blocks.46ꢀ48 Peptide to
peptoid substitutions have been also shown to improve both metabolic stability and cell
permeability.49 Since the TIC group in 1 contains an Nꢀbenzyl substructure, we hypothesized that
it could be replaced by an Nꢀlinked peptoid residue, allowing the efficient exploration of peptoid
side chains at this position using primary amine reagents. This was confirmed by the synthesis of
peptideꢀpeptoid hybrid 2 (Figure 1, Table 1) with a binding affinity (Ki = 4.5 µM) comparable to
that of peptide 1 (Ki = 2.1 µM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Computational docking of peptide 1 in a homology model of the CXCR7 receptor. The
TIC group is located in a large hydrophobic pocket (white) with room for further expansion. The
arginine side chain points into a pocket containing acidic residues, forming favorable hydrogen
bonding and electrostatic interactions with negatively charged regions (red).
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 37
1
2
3
4
5
6
7
8
9
In order to test the hypothesis that the Nꢀbenzyl peptoid group binds in a hydrophobic pocket
with additional space, a series of analogs with larger side chains at the peptoid position were
synthesized, guided in part by docking a larger series of synthetically available building blocks
into the CXCR7 receptor homology model (Table 1). Analog 4 with an Nꢀphenylpropyl side
chain (Ki = 96 nM) showed a 50ꢀfold improved binding affinity over the shorter Nꢀbenzyl (2)
and Nꢀphenethyl (3) moieties. Further elongation of the side chain did not lead to additional
affinity gain (data not shown). Optimization of the side chain phenyl ring of 4 with F and Cl
substituents further increased affinity by a factor of 15, culminating in 5 and 6 with singleꢀdigit
nanomolar binding affinities, approximately 750ꢀfold more potent than the initial peptoid 2. The
relative docking scores, expressing free energy of binding, correlated well with the experimental
binding affinities (Table 1), which indicates an improved complementary fit of the side chain
into the hydrophobic pocket, resulting in an enhanced binding of the peptoids. Notably, one of
the most potent analogs (6) also showed the best relative docking score in our computational
modeling. Albeit not the most potent analog, peptoid 4 was selected as the reference compound
for the next set of structural modifications, knowing that the ring substitutions of 5 and 6 could
be reintroduced in the final design stage to regain binding affinity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Peptoid variations with improved binding affinity.
ACS Paragon Plus Environment
7

Page 9 of 37
Journal of Medicinal Chemistry
1
2
3
4
Compd.
R (peptoid)
CXCR7 Ki [nM]a Docking Scoreb ClogP
5
6
7
4500 (5.3±0.3)
ꢀ14.4
5.2
2
8
9
7700 (5.1±0.1)
96 (7.0±0.2)
6 (8.3±0.3)
ꢀ14.1
ꢀ15.3
ꢀ16.2
ꢀ16.4
5.7
6.1
6.8
6.3
3
4
5
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8 (8.1±0.2)
aValues were determined by a radioligand binding assay and are reported as geometric mean of
at least two independent experiments with pKi ± SD in parentheses. Glide XP docking score,
where more negative values indicate stronger binding affinity (see Experimental Section).
In order to evaluate whether the backbone conformation of the peptideꢀpeptoid hybrid 4
remained unchanged compared to peptide 1, we employed NMR spectroscopy to gain
b
information on the compound conformation in DMSO. The resulting data showed that 1 and 4
have similar conformations, both with two backbone transannular hydrogen bonds between 1ꢀ
tryptophan and 4ꢀarginine according to NMR temperature coefficient data (Table S1). The
slightly larger 3JNHꢀαH value of 4 at 4ꢀarginine (9.1 Hz) indicated that the dihedral angle HꢀNꢀCαꢀ
H is closer to 180 degrees. The overall backbone conformation is similar to that of previously
reported orally bioavailable macrocyclic peptides,50,51 indicating the potential for this scaffold to
provide reasonable passive cell membrane permeability and oral bioavailability when side chain
polarity is limited.
After obtaining peptoid analogs with good binding affinity, we shifted our attention towards
improving passive permeability. Our focus was on replacing the arginine side chain, which is
positively charged at physiological pH and likely contributes to the poor passive permeability of
parent peptide 1. The EPSA values for peptides 1ꢀ6 were determined to be in the range of 135
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 37
1
2
and 150, and not surprisingly, they exhibited poor passive permeability (Papp < 1 x 10ꢀ6 cm/s).
The docking model of peptide 1 in the CXCR7 receptor homology model indicated that the
arginine side chain points into a polar pocket containing aspartate and glutamate residues,
forming favorable hydrogen bonding and electrostatic interactions with Asp275 and Glu290
(Figure 2). We anticipated that replacement of the arginine side chain with nonꢀbasic or less
polar residues, while retaining binding affinity, would be challenging. Nevertheless, arginine
replacements with improved membrane permeability have been described in the literature,52,53
and we set out to explore the effects of unnatural amino acids containing heterocyclic ring
systems (Table 2).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Arginine replacements with decreased polarity.
Compd.
R
CXCR7 Ki [nM]a EPSAb ClogP
96 (7.0±0.2)
30 (7.5±0.1)
410 (6.4±1.1)
32 (7.5±0.3)
63 (7.2±0.1)
145
127
120
123
120
6.1
6.7
6.8
7.5
7.6
4
7
8
9
10
ACS Paragon Plus Environment
9

Page 11 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
71 (7.1±0.1)
189 (6.7±0.3)
774 (6.1±0.3)
116
116
120
7.6
7.6
9.1
11
12
13
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
aValues were determined by a radioligand binding assay and are reported as geometric mean of
at least two independent experiments with pKi ± SD in parentheses. bValues were measured by a
chromatographic method.41,43
Replacing the arginine side chain of 4 with various fiveꢀ and sixꢀmembered heteroaryl rings
led to a pronounced decrease in polarity, with EPSA values ranging from 116 to 127 for peptoid
analogs 7ꢀ12, approximately 20ꢀ30 units lower than that of 4 (EPSA = 145). Encouragingly,
several arginine replacements (7, 9ꢀ11) were equipotent or showed an even greater binding
affinity compared to 4. To explore the lowest achievable polarity for this side chain we also
prepared a nonꢀheteroaryl phenylalanine analog 13 which had an EPSA value similar to those of
analogs 7ꢀ12. However, presumably due to the lack of functional groups capable of hydrogen
bonding or electrostatic interactions with the negatively charged regions in the binding pocket
(Figure 2), its binding affinity decreased 8ꢀfold compared to 4. Since peptoid 11 exhibited the
best balance of good potency, matching that of 4, and low polarity (EPSA = 116), it was selected
as a starting point for further structural optimization.
After successful replacement of the arginine side chain with heterocyclic ring systems, we
tested the effect of replacing the two tryptophan residues with alternate aryl side chains (Table
3). Substitution of one tryptophan residue in 11 by hhPhe resulted in 14 that showed an EPSA
decrease from 116 to 104, but permeability in the MDCKIIꢀLE assay remained low (Papp = 0.9 x
10ꢀ6 cm/s). Additional replacement of the second tryptophan residue with hhPhe (15) and hPhe
(16) further reduced EPSA below the aforementioned threshold of less than 90, resulting for the
first time in measurable passive permeability (Papp > 1 x 10ꢀ6 cm/s). Since 15 and 16 had slightly
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 37
1
2
3
4
5
6
reduced binding affinity, we were particularly intrigued by peptoid 18 with a nonꢀaromatic tꢀ
butyl substitution, which retained the binding affinity of the tryptophan analog (14). With its
7
8
9
further reduced polarity (EPSA = 79), 18 also exhibited the highest passive permeability (Papp
=
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
4.8 x 10ꢀ6 cm/s) measured thus far in this series. Interestingly, a comparably low EPSA value for
the equipotent peptoid 17 did not translate to measurable permeability (Papp = 0.7 x 10ꢀ6 cm/s).
Table 3. Tryptophan substitutions with reduced polarity and increased passive permeability.
Compd.
R
CXCR7 Ki [nM]a EPSAb ClogP Papp [10ꢀ6 cm/s]c
23 (7.6±0.2)
71 (7.2±0.2)
88 (7.1±0.3)
24 (7.6±0.1)
34 (7.5±0.1)
104
90
89
89
79
8.7
9.8
9.3
7.1
9.2
0.9
1.2
1.8
0.7
4.8
14
15
16
17
18
aValues were determined by a radioligand binding assay and are reported as geometric mean of
at least two independent experiments with pKi ± SD in parentheses. bValues were measured by a
chromatographic method.41,43 cValues were determined by a permeability assay in a lowꢀefflux
MadinꢀDarby canine kidney cell line.40
ACS Paragon Plus Environment
11

Page 13 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
With 18 emerging as the most promising analog, we embarked on further optimization of
various side chains to obtain peptides with an improved balance between potency and passive
permeability (Table 4). First, at the original arginine position, we replaced the 3ꢀpyridyl group
with a thiazole ring for which we expected a modest potency boost from known SAR (Table 2).
This was confirmed with peptide 19 (Ki = 9 nM). The pyridyl to thiazole substitution resulted
only in a negligible change in EPSA and permeability. In the next step, shortening of the side
chain from hhPhe (19) to phenylalanine (20) improved permeability (Papp = 6.6 x 10ꢀ6 cm/s) but
potency was decreased by 50ꢀfold. Again taking advantage of existing SAR (Table 1), we
anticipated a potency boost by switching to a 2,4ꢀdifluoroꢀsubstituted Nꢀphenylpropyl side chain
at the peptoid position. Indeed, 21 showed a 20ꢀfold increase in potency (Ki = 25 nM) in
comparison to 20. The introduction of an additional 4ꢀfluoro substituent on the phenylalanine
side chain (22) further increased permeability (Papp = 7.6 x 10ꢀ6 cm/s). In an attempt to eliminate
another aromatic group from the peptide, we also investigated aliphatic side chains (23, 24).
Interestingly, the introduction of a second tꢀbutyl group in 24 showed the highest measured
permeability in this series (Papp = 8.3 x 10ꢀ6 cm/s). Peptide 25, however, provided the best overall
balance between potency (Ki = 9 nM) and passive permeability (Papp = 6.2 x 10ꢀ6 cm/s), and was
progressed towards examination of its in vivo oral absorption potential in rats. Measured in the
CXCR7 βꢀarrestin functional assay, the activity of 25 (EC50 = 15 nM) showed a 3ꢀfold
improvement over the initial peptide 1 (EC50 = 46 nM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 37
1
2
3
4
Table 4. Side chain optimization with improved balance between potency and permeability.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compd.
R
X CXCR7 Ki [nM]a EPSAb ClogP Papp [10ꢀ6 cm/s]c
H
H
F
F
F
F
F
9 (8.1±0.1)
537 (6.3±0.1)
25 (7.6±0.1)
42 (7.4±0.1)
34 (7.5±0.1)
95 (7.0±0.1)
9 (8.0±0.4)
80
79
74
74
74
70
77
9.0
7.9
8.2
8.4
9.5
8.7
8.3
4.0
6.6
5.2
7.6
3.8
8.3
6.2
19
20
21
22
23
24
25
O
*
aValues were determined by a radioligand binding assay and are reported as geometric mean of
at least two independent experiments with pKi ± SD in parentheses. bValues were measured by a
chromatographic method.41,43 cValues were determined by a permeability assay in a lowꢀefflux
MadinꢀDarby canine kidney cell line.40
To summarize, our design strategy was to employ peptoids capable of probing a larger
diversity of side chains to efficiently explore a large hydrophobic pocket in the CXCR7 binding
site. Simultaneously, we monitored physicochemical properties such as polarity measured by
EPSA, with the aim of identifying peptideꢀpeptoid hybrids that were both potent and passively
permeable across cell membranes. The overall progression of peptides synthesized in these
iterative design cycles is shown in Figure 3. The initial peptides showed only weak binding
ACS Paragon Plus Environment
13

Page 15 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
affinity and high polarity with no measurable permeability. In the first design step,
implementation of Nꢀlinked peptoids led to peptideꢀpeptoid hybrids with improved binding
affinity, but relatively high EPSA values (Table 1). Next, replacement of the basic arginine side
chain yielded compounds with lower polarity while retaining moderate activity (Table 2).
Substitution of the two tryptophan residues yielded analogs with further reduced polarity (Table
3), and ultimately resulted in peptides with improved potency and passive permeability (Table
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
Figure 3. Optimization of peptides 1-25 as measured by binding affinity (CXCR7 Ki), polarity
(EPSA), and in vitro passive permeability (Papp). Compounds exemplified in Tables 1ꢀ4 are
grouped by shape. Permeability values of peptides were binned and are highlighted by color.
Peptide 25 (*) was selected for further characterization.
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 37
1
2
3
4
5
6
7
8
9
The peptides described herein are only a subset of a larger collection of analogs that were
synthesized and tested during the iterative design cycle process. Having a large data set of related
analogs with experimental data in hand, we were interested to learn of any relationships between
physicochemical properties and experimental data to guide prospective peptide design efforts.
The correlation between passive permeability (Papp), polarity (EPSA), and lipophilicity (ClogP)
is shown in Figure 4. In agreement with previous findings,43 our data confirms that EPSA values
of less than 90 are indeed necessary for peptides to achieve measurable passive permeability.
However, satisfying this rule alone did not always result in permeable peptides (Figure 4, left).
Our data suggests that peptides also need to fulfill a certain lipophilicity requirement, and with
this particular scaffold they must have ClogP values between 7.5 and 10 in order to yield
measureable passive permeability. Peptides outside this ClogP range did not lead to substantial
permeability in the MDCKIIꢀLE assay (Figure 4, right). Apparently, not only is high polarity
detrimental to passive permeability, but high lipophilicity seems to impede passive perfusion
through the membrane. Similar findings of a parabolic correlation between a variety of
calculated logP and experimental Papp values were observed by other groups as well, claiming
‘islands of permeability’ where only peptides with the right balance of polarity, lipophilicity and
solubility are likely to have measurable permeability.54,55 The equivalent analysis correlating
passive permeability with calculated PSA (TPSA) instead of EPSA revealed that the cutoff
criteria established in literature (TPSA < 140) would have not been applicable for this series of
macrocyclic peptides in a prospective design (Figure S1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
15

Page 17 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Correlation between passive permeability (Papp), polarity (EPSA) and lipophilicity
(ClogP). Dashed lines indicate the EPSA threshold of less than 90 and a ClogP range from 7.5 to
10, necessary to achieve measurable permeability for peptides with this particular scaffold.
In addition to the optimization of side chains, Nꢀmethylation of backbone amide groups has
been shown to have a beneficial effect on permeability in cyclic peptide scaffolds.51,56ꢀ60 We
examined the effect of Nꢀmethylation on selected peptides developed in this study (Table 5). We
focused mainly on Nꢀmethylation of the R2 and R3 backbone amides since analogs methylated at
these positions are expected to have a similar lowꢀdielectric backbone conformation compared to
cyclic peptides that were previously shown to have good permeability.48,55 We found that Nꢀ
methylation was generally beneficial to permeability when ClogP was in a favorable range, as
described above. For example, peptide 17 (Table 3) had shown poor permeability (Papp = 0.7 x
10ꢀ6 cm/s) despite its low polarity value (EPSA = 89). The likely reason for that is the ClogP
value of 7.1 which is below the optimal range of 7.5 and 10 (Figure 4). Nꢀmethylation of the R3
(26) and R2/R3 (27) backbone amides not only further decreased polarity (EPSA = 86 and 83,
respectively), but also slightly raised the ClogP values (ClogP = 7.8 and 8.5, respectively)
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 37
1
2
3
4
towards the optimal range necessary to achieve passive permeability. As a result (Table 5), we
5
6
improved the low permeability of 17 with both its monoꢀNꢀmethyl derivative 26 (Papp > 2.7 x 10ꢀ
7
8
9
6
cm/s) and bisꢀNꢀmethyl derivative 27 (Papp = 4.0 x 10ꢀ6 cm/s). Ultimately, we decided not to
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
pursue the Nꢀmethylation strategy further, because in most cases Nꢀmethylation at the R2
position also had a deleterious effect on CXCR7 binding affinity.
Table 5. Nꢀmethylation of cyclic peptide scaffold to improve permeability.
R2
N
O
N
H
O
N
O
O
O
H
R3
N
N
S
N
O
N
N
Compd. R2 R3 CXCR7 Ki [nM]a EPSAb ClogP Papp [10ꢀ6 cm/s]c
H
H
H
24 (7.6±0.1)
20 (7.7±0.3)
211 (6.7±0.2)
89
86
83
7.1
7.8
8.5
0.7
2.7
4.0
17
26
27
Me
Me Me
aValues were determined by a radioligand binding assay and are reported as geometric mean of
at least two independent experiments with pKi ± SD in parentheses. bValues were measured by a
chromatographic method.41,43 cValues were determined by a permeability assay in a lowꢀefflux
MadinꢀDarby canine kidney cell line.40
To assess whether oral bioavailability can be achieved for peptides with measurable passive
permeability as determined in the MDCKIIꢀLE assay, the pharmacokinetics of 25 was examined
in rats. The intravenous (IV) and oral (PO) pharmacokinetic results are shown in Table 6. After
IV dosing, peptide 25 exhibited a high plasma clearance (CLp) that exceeded the rat hepatic
blood flow value (70 mL/min/kg).61 Consistent with this observation, incubation of 25 (1 ꢀM)
with NADPHꢀsupplemented rat liver microsomes resulted in facile metabolic turnover leading to
ACS Paragon Plus Environment
17

Page 19 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a short halfꢀlife (45 min) and a high apparent intrinsic clearance value (610 ꢀL/min/mg). The
finding that 25 was resistant towards metabolism (halfꢀlife > 2 h) in rat liver microsomes lacking
NADPH cofactor strongly suggests a role for rat cytochrome P450 enzymes in the oxidative
metabolism of 25 as a cause of its metabolic instability, which leads to observed high in vivo
plasma clearance. The volume of distribution (Vd,ss) for 25 was considerably above the total body
water volume (0.6 L/kg) indicating extensive distribution into tissues. The corresponding halfꢀ
life after IV administration was relatively short (< 1 h).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 6. Pharmacokinetic profile of compound 25 in ratsa.
b
b
b
c
CLp
Vd,ss
t1/2
[h]
AUCb
Cmax
AUCc
F
Compd.
fu,p
[mL/min/kg] [L/kg]
[ng/mL×h] [ng/mL] [ng/mL×h] [%]
136±9 4.4±0.3 0.95±0.01
124±9 97±18 224±30 18 0.011
25
aPharmacokinetic parameters were calculated from plasma concentrationꢀtime data in male
b
Wistar Han rats (2 animals per route) and are reported as mean ± SD values. Following IV
c
administration (1 mg/kg) formulated in DMSO: Cremophor: Saline (5:5:90 v/v). Following PO
administration (10 mg/kg) as an emulsion in 10% Miglyol 812: Cremophor RH40: Capmul
MCM (30:40:30 v/v) and 90% water (SEDDS formulation).
Following PO administration of 25 in the lipophilic selfꢀemulsifying drug delivery system
(SEDDS) formulation at 10 mg/kg, the systemic exposure measured by maximal plasma
concentrations (Cmax) and area under the plasma concentrationꢀtime curve (AUC) was 97 ng/mL
and 224 ng/mL x h, respectively. The corresponding oral bioavailability (F) of 25 was 18%. The
reason for achieving oral systemic exposure despite the hepatic blood flowꢀlimited intrinsic
clearance of 25 remains unclear at the present time. One possibility is the saturation of firstꢀpass
metabolic extraction by rat cytochrome P450 enzymes in the small intestine and the liver62 due to
the 10ꢀfold higher PO dose administered relative to the IV dose.
When normalized against the rat plasma unbound fraction (fu,p),63 the free Cmax value (~ 1.3
nM) of 25 after PO dosing was approximately 7ꢀfold lower than the corresponding CXCR7
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 37
1
2
3
4
5
6
7
8
9
binding affinity (9 nM). In order to probe in vivo CXCR7 pharmacology, sustained unbound
plasma concentrations several fold over the Ki value are desired, which currently would be
unattainable with a reasonable PO dosing regimen of 25.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSION
In this study, we demonstrated the successful optimization of low affinity, highly polar and
nonꢀpermeable peptides towards macrocyclic peptideꢀpeptoid hybrids with improved CXCR7
binding affinity (Ki < 100 nM) and in vitro passive permeability in the MDCKIIꢀLE assay (Papp
>
5 x 10ꢀ6 cm/sec), thereby achieving our goal to identify potent and cellꢀpermeable CXCR7
modulators. Moreover, bioactive peptide 25 (Ki = 9 nM, EC50 = 15 nM) with measurable passive
permeability (Papp = 6.2 x 10ꢀ6 cm/sec) also demonstrated reasonable oral bioavailability in rats
(F = 18%). A pragmatic next step to further increase the oral absorption of 25 would be to
improve its metabolic stability whilst maintaining CXCR7 potency and the favorable passive
permeability characteristics.
EXPERIMENTAL SECTION
All procedures performed on animals in this study were in accordance with established
guidelines and regulations, and were reviewed and approved by Pfizer Institutional Animal Care
and Use Committee. Pfizer animal care facilities that supported this work are fully accredited by
AAALAC International.
Peptide Synthesis. Standard solid phase peptide synthesis (SPPS) was used to synthesize the
peptideꢀpeptoid hybrids. Unless otherwise stated, the resin (100ꢀ200 mesh or 200ꢀ400 mesh,
cross linking degree: 1% DVB), Fmoc amino acids, and other starting materials were
commercially available and used without further purification. After cleaving from the resin, the
ACS Paragon Plus Environment
19

Page 21 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
resulting acyclic peptides were then cyclized under dilute solution phase conditions: (a) BOP,
HOBt, DIPEA, DCM ([substrate] ~6 mM), 16 h, and (b) EDC, HOAt, DIPEA, ACN:DCM (1:1,
0.5 mg of linear peptideꢀpeptoid compound/mL), 16 h. Full synthetic details for individual
peptideꢀpeptoid hybrids are provided in the Supporting Information.
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
Purification of Peptides. Crude material was dissolved in a minimal amount of water and
purified using multiple conditions: (a) Waters 4000 system connected to a Phenomenex LunaTM
C18, reversed phase HPLC column (10 microns, 100 Å, 25 mm x 200 mm) eluting with a solvent
gradient A:B where A = 0.1% TFA in water and B = 0.1% TFA in ACN:water (4:1), (b)
Shimadzu LCꢀ8A preparative HPLC system coupled to a Higgins C18 analytical column (5
microns, 100 Å, 20 mm x 250 mm) eluting with 0.1% TFA in water:0.1% TFA in ACN (40:60 to
0:100) over 30 min at a flow rate of 10 mL/min, (c) Shimadzu LCꢀ8A preparative HPLC system
coupled to a Higgins C18 analytical column (5 micron, 100 Å, 20 mm x 250 mm) eluting with
0.1% TFA in water:0.1% TFA in ACN (10:90 to 0:100) over 30 min at a flow rate of 10 mL/min,
and (d) Waters system coupled to a SunFire C18 column (5 microns, 100 Å, 19 mm x 100 mm)
eluting with 0.05% TFA in water:0.05% TFA in ACN (90:10 to 50:50 for 10.5 min, then 50:50
to 0:100 for 0.5 min, then 5:95 for 1 min) at a flow rate of 25 mL/min. Specific conditions for
individual peptideꢀpeptoid hybrids are provided in the Supporting Information.
Purity Analysis of Peptides. Purified material was analyzed to confirm > 95% purity under a
variety of conditions: (a) HP1090 system coupled to a Phenomenex C18 (2) (5 microns, 100 Å,
4.6 mm x 150 mm) reversed phase HPLC column eluting with a solvent gradient of A:C, where
A = 0.1% TFA in water and C = 0.09% TFA in ACN:water (4:1) over 20 minutes at a flow rate
of 1.0 mL/min, (b) Waters LC/MS/ELSD (single quad MS) system coupled to an Acquity HSS
T3 UPLC reversed phase column (1.8 microns, 100 Å, 2.1 x 50 mm) eluting with 0.05% TFA in
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 37
1
2
3
4
5
6
7
8
9
water: 0.05% TFA in ACN (95:5 to 0:100 over 5 min then 0:100 for 2.5 min) at flow rate of 0.7
mL/min, and (c) Pure peptides were analyzed using a Waters LC/MS/ELSD (single quad MS)
system coupled to a SunFire C18 column (5 microns, 100 Å, 4.6 mm x 50 mm) reversed phase
column eluting with 0.05% TFA in water:0.05% TFA in ACN (90:10 to 50:50 for 3.75 min, then
50:50 to 5:95 for 0.5 min, then 5:95 for 0.5 min) at a flow rate of 2.0 mL/min. Specific
conditions, retention times, and UV purities (215 nm or 220 nm) for individual peptideꢀpeptoid
hybrids are provided in the Supporting Information.
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
CXCR7 radiolabeled ligand binding assay. The binding affinity of test compounds for the
human CXCR7 chemokine receptor was determined by their ability to displace 125IꢀCXCL12
(Perkin Elmer) from membranes obtained from CHOꢀK1 cells overexpressing the human
CXCR7 receptor (DiscoveRx). Ki values are reported as the geometric mean of at least two
independent experiments each of which was run in duplicate.
Test compounds were serialized in 100% DMSO and spotted into 96 well plates (NBS). Total
binding wells were spotted with diluent. Nonꢀspecific wells were defined by the addition of
potent CXCR7 agonist. 10 ꢀL of 125IꢀCXCL12 with a final concentration of 100 pM was added
to each well of the plate, followed by the addition of 90 ꢀL of membranes. Both the membranes
and 125IꢀCXCL12 were diluted to the appropriate concentration in assay buffer (HBSS containing
10 mM HEPES and 0.2% BSA). After the addition membranes plates were incubated at room
temperature for 2 h (shaking). Reactions were terminated by rapid filtration through
poly(ethyleneimine) (0.3%) treated 96 well GF/C Unifilter plates (Perkin Elmer). Unbound
ligand was removed by washing the filters with ice cold wash buffer. Filters were allowed to dry
thoroughly prior to the addition of Ready Safe scintillation fluid (Perkin Elmer). The amount of
bound 125IꢀCXCL12 was quantitated by reading plates on the Trilux (Perkin Elmer).
ACS Paragon Plus Environment
21

Page 23 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CXCR7 β-arrestin functional assay. The agonist activity of test compounds was determined
by their ability to induce βꢀarrestin recruitment in CHOꢀK1 cells overexpressing the human
CXCR7 receptor (DiscoveRx). EC50 values are reported as the geometric mean of at least two
independent experiments each of which was run in duplicate.
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 receptor is fused with the small enzyme fragment ProLink™ and coꢀexpressed in the cells
stably expressing a fusion protein of βꢀarrestin and the larger, Nꢀterminal deletion mutant of βꢀ
gal (called enzyme acceptor or EA). Agonist activation of the receptor by test compounds causes
binding of βꢀarrestin (EA) to the ProLinkꢀtagged CXCR7 resulting in the formation of active βꢀ
gal enzyme. Enzyme activity is measured using chemiluminescent PathHunter® detection
reagents (DiscoveRx).
Prior to the assay, cells are removed from culture flasks and plated at a density of 6K viable
cells per well in white solid bottom 384 well plates. After the addition of cells, plates were
placed in humidified 37 ℃ 5% CO2 incubator for 18ꢀ24 h. Test compounds were serialized in
100% DMSO and diluted in assay buffer (HBSS containing 10 mM HEPES and 0.1% BSA)
prior to their addition to the cell plates. Media was removed from the cell plates and 20 ꢀL of
compound or diluent was added to the appropriate wells of the plate. Following compound
addition, cell plates were incubated at 37 ℃ in 5% CO2 for 30 min. After incubation 20 ꢀL of
PathHunter® detection reagents (DiscoveRx) were added to each well of the plate. Plates were
incubated at 23 ℃ in the dark for 1 h prior to reading on the Envision (Perkin Elmer).
MDCKII-LE Permeability Assay. Passive permeability was determined using a lowꢀefflux
MadinꢀDarby canine kidney (MDCKIIꢀLE) cell line (Pfizer, Groton, CT). Cells were grown in
minimum essential medium αꢀnucleosides (Life Technologies, Grand Island, NY) supplemented
with 10% fetal bovine serum, 1% minimum nonꢀessential amino acids solution, 1%
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 37
1
2
3
4
5
6
7
8
9
GlutaMAX™ and 1% penicillinꢀstreptomycin prior to seeding into Millipore 96ꢀwell cell culture
insert plates (EMD Millipore Corporation). MDCKIIꢀLE cells were cultured on the inserts with
100ꢁµL medium per well on the apical side and 36ꢁmL for all 96 wells on the basolateral side.
Donor solutions were prepared from HBSS containing 20 mM HEPES, pH 6.5. Stock solutions
of test compounds, prepared at 5 mM in DMSO, were used to prepare donor solutions of 6 µM
compound. Receiver solutions were prepared from HBSS containing 20 mM HEPES and 0.4 %
BSA (w/v), pH 7.4. Prior to the assay, cell culture medium was removed and cells were preꢀ
incubated with HBSS for 10ꢁmin. To start the assay, 100 µL of donor solution and 300 µL of
receiver solution were added to the apical and basolateral chambers, respectively. After 90ꢁmin
incubations, aliquots were taken from the receiver chambers to determine the translocated
amount of compound. Samples were taken from the donor chambers before and after incubation
to determine the initial concentration (C0) and recovery values. An internal standard solution,
0.5ꢁµg/mL CPꢀ628374 (MWꢁ=ꢁ687) in 100% methanol, was added to the receiver and donor
samples. The samples were analyzed by LCꢀMS/MS to determine the peak area for test
compound and the internal standard. All incubations were conducted in triplicate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Apparent permeability values (Papp) were calculated according to the following equation:
dx
dt
Papp =
C0 × A
where dx is the amount of compound in the receiver compartment, dt is the incubation time, C0
is the initial concentration in the donor compartment, and A is the area of the filter of the
transwell plate.
Pharmacokinetics. Rat pharmacokinetics were determined in male Wistar Han rats sourced
from Charles River, weighing between 250ꢀ281 g, (n=2/route) following single IV bolus (1
ACS Paragon Plus Environment
23

Page 25 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mg/kg) and PO (10 mg/kg) administration. For the IV dose, 25 was formulated in DMSO:
Cremophor: Saline (5:5:90 v/v) and administered at 2 mL/kg via the jugular vein cannula. In
order to avoid poor aqueous solubility confounding oral bioavailability determination, the PO
doses of 25 were administered in a selfꢀemulsifying (SEDDS) formulation. Peptide 25 was
presented as an emulsion in 10% Miglyol 812: Cremophor RH40: Capmul MCM (30:40:30 v/v)
and 90% water (10 mL/kg). Doses were administered by gavage tube. At 2 min (IV only), 5 min,
15 min, 30 min, 1 h, 2 h, 4 h, 7 h and 24 h post dose, blood samples were taken from previously
implanted inꢀdwelling carotid artery catheters. Blood was centrifuged to produce plasma which
was stored frozen until bioanalysis. Bioanalysis of plasma samples was completed using specific
HPLCꢀMS/MS methods with lower limits of detection for 25 equal to or below 1 ng/mL.
Pharmacokinetic parameters were calculated using standard nonꢀcompartmental analysis.
Molecular Modeling. The crystal structure of CXCR4 in complex with the cyclic peptide
antagonist CVX15 (PDB code 3OE0) was used to develop a CXCR7 homology model. The
protein sequence of CXCR7 was aligned to the CXCR4 structure using ClustalW. The homology
model of CXCR7 was generated using Prime version 3.1 from Schrödinger. Conformational
modeling of cyclic peptides was performed using PLOP with robust backbone sampling, and the
predicted conformations were used as input for subsequent docking calculations. Docking of
compounds into the CXCR7 homology model was conducted using Glide XP from Schrödinger
(release 2016ꢀ2). The Induced Fit Docking (IFD) protocol, which uses Prime and Glide to induce
adjustments in the receptor structure, was employed to investigate binding models of key
compounds, such as peptide 1 as well as peptideꢀpeptoid hybrids 2 and 4. To support
optimization of the peptoid position for CXCR7 binding, CombiGlide from Schrödinger was
utilized to enumerate virtual libraries of peptideꢀpeptoid hybrids based on commercially
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 37
1
2
3
4
5
6
available primary amines and to dock against the CXCR7 homology model to identify peptoid
substitutions favorable for binding.
7
8
9
ASSOCIATED CONTENT
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
Supporting Information. The Supporting Information is available free of charge on the ACS
Publication website at DOI:
3
NMR temperature coefficients and chemical shifts of NHs, JNHꢀαH couplings; EPSA
method; correlation plot of passive permeability vs. TPSA; synthetic procedures and
analytical data of peptides 1-27; IV and oral PK concentration vs. time curves for peptide
25 (PDF)
Atomic coordinates of the CXCR7 homology model with computational docking of
peptide 1 (PDB)
AUTHOR INFORMATION
Corresponding Authors
*Phone: +1 617 551 3517. Eꢀmail: elnaz.menhajiꢀklotz@pfizer.com. *Phone: +1 831 459 1307.
Eꢀmail: slokey@ucsc.edu. *Phone: +1 650 392 0363. Eꢀmail: davidjearp@circlepharma.com.
Present Addresses
‡B.R.H.: Middletown, Connecticut 06457. ‡M.J.S.: Niantic, Connecticut 06357. ‡J.Y.:
#
MassBiologics, Boston, Massachusetts 02124. R.T.: Monterey Peninsula College, Monterey,
California 93940.
ACS Paragon Plus Environment
25

Page 27 of 37
Journal of Medicinal Chemistry
1
2
3
4
Author Contributions
5
6
7
All authors have given approval to the final version of the manuscript.
8
9
Notes
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 authors declare the following competing financial interest: M.P.J. and R.S.L. are coꢀ
founders of, and advisors to, Circle Pharma and each holds an equity interest in the company.
ACKNOWLEDGMENT
We thank the Chinese Peptide Company for peptide synthesis support.
ABBREVIATIONS USED
CXCR7, CXC chemokine receptor 7; CXCR4, CXC chemokine receptor 4; CXCL12, CXC
chemokine ligand 12; EC50, half maximal effective concentration; ELSD, evaporative light
scattering detector; HBSS, Hanks’ balanced salt solution; HPLC, highꢀperformance liquid
chromatography; IV, intravenous; Ki, binding affinity; MDCKIIꢀLE, lowꢀefflux MadinꢀDarby
Canine Kidney cell line; MS, mass spectrometry; NMR, nuclear magnetic resonance; Papp,
absorptive permeability; PO, oral; SAR, structureꢀactivity relationship; SEDDS, selfꢀemulsifying
drug delivery system; UPLC, ultraꢀperformance liquid chromatography; UV, ultraviolet; ACN,
acetonitrile;
DCM,
dichloromethane;
DMSO,
dimethyl
sulfoxide;
Fmoc,
fluorenylmethyloxycarbonyl; hPhe, homophenylalanine; hhPhe, homohomophenylalanine; TFA,
trifluoroacetic acid; TIC, tetrahydroisoquinolineꢀ3ꢀcarboxylic acid.
REFERENCES
(1) Nagasawa, T.; Hirota, S.; Tachibana, K.; Takakura, N.; Nishikawa, S.; Kitamura, Y.;
Yoshida, N.; Kikutani, H.; Kishimoto, T. Defects of Bꢀcell lymphopoiesis and boneꢀmarrow
myelopoiesis in mice lacking the CXC chemokine PBSF/SDFꢀ1. Nature 1996, 382 (6592), 635ꢀ
638.
(2) Ma, Q.; Jones, D.; Borghesani, P. R.; Segal, R. A.; Nagasawa, T.; Kishimoto, T.; Bronson,
R. T.; Springer, T. A. Impaired Bꢀlymphopoiesis, myelopoiesis, and derailed cerebellar neuron
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 37
1
2
3
4
migration in CXCR4ꢀ and SDFꢀ1ꢀdeficient mice. Proc. Natl. Acad. Sci. U. S. A. 1998, 95 (16),
5
6
9448ꢀ9453.
7
8
9
(3) Bleul, C. C.; Wu, L.; Hoxie, J. A.; Springer, T. A.; Mackay, C. R. The HIV coreceptors
CXCR4 and CCR5 are differentially expressed and regulated on human T lymphocytes. Proc.
Natl. Acad. Sci. U. S. A. 1997, 94 (5), 1925ꢀ1930.
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) Rajagopal, S.; Kim, J.; Ahn, S.; Craig, S.; Lam, C. M.; Gerard, N. P.; Gerard, C.;
Lefkowitz, R. J. Betaꢀarrestinꢀ but not G proteinꢀmediated signaling by the "decoy" receptor
CXCR7. Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (2), 628ꢀ632.
(5) Levoye, A.; Balabanian, K.; Baleux, F.; Bachelerie, F.; Lagane, B. CXCR7 heterodimerizes
with CXCR4 and regulates CXCL12ꢀmediated G protein signaling. Blood 2009, 113 (24), 6085ꢀ
6093.
(6) SanchezꢀAlcaniz, J. A.; Haege, S.; Mueller, W.; Pla, R.; Mackay, F.; Schulz, S.; Lopezꢀ
Bendito, G.; Stumm, R.; Marin, O. Cxcr7 controls neuronal migration by regulating chemokine
responsiveness. Neuron 2011, 69 (1), 77ꢀ90.
(7) Naumann, U.; Cameroni, E.; Pruenster, M.; Mahabaleshwar, H.; Raz, E.; Zerwes, H. G.;
Rot, A.; Thelen, M. CXCR7 functions as a scavenger for CXCL12 and CXCL11. PLoS One
2010, 5 (2), e9175.
(8) SanchezꢀMartin, L.; SanchezꢀMateos, P.; Cabanas, C. CXCR7 impact on CXCL12 biology
and disease. Trends Mol. Med. 2013, 19 (1), 12ꢀ22.
(9) Zabel, B. A.; Lewen, S.; Berahovich, R. D.; Jaen, J. C.; Schall, T. J. The novel chemokine
receptor CXCR7 regulates transꢀendothelial migration of cancer cells. Mol. Cancer 2011, 10, 73.
ACS Paragon Plus Environment
27

Page 29 of 37
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(10) Li, X.; Zhu, M.; Penfold, M. E.; Koenen, R. R.; Thiemann, A.; Heyll, K.; Akhtar, S.;
Koyadan, S.; Wu, Z.; Gremse, F.; Kiessling, F.; van Zandvoort, M.; Schall, T. J.; Weber, C.;
Schober, A. Activation of CXCR7 limits atherosclerosis and improves hyperlipidemia by
increasing cholesterol uptake in adipose tissue. Circulation 2014, 129 (11), 1244ꢀ1253.
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
(11) Klein, K. R.; Karpinich, N. O.; Espenschied, S. T.; Willcockson, H. H.; Dunworth, W. P.;
Hoopes, S. L.; Kushner, E. J.; Bautch, V. L.; Caron, K. M. Decoy receptor CXCR7 modulates
adrenomedullinꢀmediated cardiac and lymphatic vascular development. Dev. Cell 2014, 30 (5),
528ꢀ540.
(12) Chung, E. S.; Miller, L.; Patel, A. N.; Anderson, R. D.; Mendelsohn, F. O.; Traverse, J.;
Silver, K. H.; Shin, J.; Ewald, G.; Farr, M. J.; Anwaruddin, S.; Plat, F.; Fisher, S. J.; AuWerter,
A. T.; Pastore, J. M.; Aras, R.; Penn, M. S. Changes in ventricular remodelling and clinical status
during the year following a single administration of stromal cellꢀderived factorꢀ1 nonꢀviral gene
therapy in chronic ischaemic heart failure patients: the STOPꢀHF randomized Phase II trial. Eur.
Heart J. 2015, 36 (33), 2228ꢀ2238.
(13) Wijtmans, M.; Maussang, D.; Sirci, F.; Scholten, D. J.; Canals, M.; MujicꢀDelic, A.;
Chong, M.; Chatalic, K. L.; Custers, H.; Janssen, E.; de Graaf, C.; Smit, M. J.; de Esch, I. J.;
Leurs, R. Synthesis, modeling and functional activity of substituted styreneꢀamides as smallꢀ
molecule CXCR7 agonists. Eur. J. Med. Chem. 2012, 51, 184ꢀ192.
(14) UtoꢀKonomi, A.; McKibben, B.; Wirtz, J.; Sato, Y.; Takano, A.; Nanki, T.; Suzuki, S.
CXCR7 agonists inhibit the function of CXCL12 by downꢀregulation of CXCR4. Biochem.
Biophys. Res. Commun. 2013, 431 (4), 772ꢀ776.
ACS Paragon Plus Environment
28