Cyclization of RGD Peptides by Suzuki-Miyaura Cross-Coupling

Article abstract of DOI:10.1021/acs.jmedchem.9b00360

Halogenated l- or d-tryptophan obtained by biocatalytic halogenation was incorporated into RGD peptides together with a variety of alkyl or aryl boronic acids. Suzuki-Miyaura cross-coupling either in solution or on-resin results in side chain-to-tail-cyclized RGD peptides, for example, with biaryl moieties, providing a new dimension of structure-activity relationships. An array of RGD peptides differing in macrocycle size, the presence of d-amino acid, N-methylation, or connectivity between the indole moiety and the boronic acid showed that, in particular, connectivity exhibits a major impact on affinities toward integrins, for example, α_Vβ₃. Structure-activity relationship studies yielded peptides with affinities toward α_Vβ₃ in the low nanomolar range, good selectivity, and high plasma stability. Structural characteristics of representative molecules have been investigated by molecular dynamics simulations, which allowed understanding the observed activity differences.

Full text of DOI:10.1021/acs.jmedchem.9b00360

Subscriber access provided by UNIV OF SOUTHERN INDIANA
Article
Cyclization of RGD peptides by Suzuki-Miyaura cross-coupling
Isabell Kemker, Christian Schnepel, David Christopher C. Schröder, Antoine Marion, and Norbert Sewald
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00360 • Publication Date (Web): 15 Jul 2019
Downloaded from pubs.acs.org on July 16, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Cyclization of RGD peptides by Suzuki-Miyaura
cross-coupling
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
Isabell Kemker,¶ Christian Schnepel,¶ David C. Schröder,¶ Antoine Marion,§ Norbert Sewald¶,*
¶Organic and Bioorganic Chemistry, Bielefeld University, Department of Chemistry,
Universitätsstraße 25, 33615 Bielefeld, Germany
§Department of Chemistry, Middle East Technical University, 06800 Ankara, Turkey
KEYWORDS. halotryptophan, RGD peptides, integrins, on-resin Suzuki-Miyaura cross-
coupling, side chain-to-tail cyclization, conformational analysis
ABSTRACT. Halogenated L- or D-tryptophan obtained by biocatalytic halogenation was
incorporated into RGD peptides together with a variety of alkyl or aryl boronic acids. Suzuki-
Miyaura cross-coupling either in solution or on-resin results in side chain-to-tail-cyclized RGD
peptides, e.g. with biaryl moieties, providing a new dimension of structure-activity relationships.
An array of RGD peptides differing in macrocycle size, presence of D-amino acid, N-
methylation, or connectivity between the indole moiety and the boronic acid showed that in
particular connectivity exhibits a major impact on affinities towards integrins, e.g. αVβ3.
Structure-activity relationship studies yielded peptides with affinities towards αVβ3 in the low
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 49
1
2
3
4
5
6
7
8
9
nanomolar range, good selectivity as wells as high plasma stability. Structural characteristics of
representative molecules have been investigated by molecular dynamics simulations which
allowed understanding the observed activity differences.
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
Introduction
Integrins recognize and bind RGD sequences that naturally occur in extracellular matrix (ECM)
proteins, e.g. vitronectin or fibronectin, which play a crucial role in the bidirectional cell
signaling. Targeting integrin-ECM interaction is an appealing strategy for medical applications
in the field of cancer treatment1, radio therapy2 or surface coating of implants.3 Hence, a
7; 8
9
multitude of ligands binding αVβ3,4; 5 α5β1,6 or more recently αVβ6
and αVβ8 have been
developed. The integrin subtypes αVβ3 and α5β1 are known to be up-regulated in tumor tissues
and are required for angiogenesis although their role remains ambiguous.10 The highly potent
αVβ3 ligand Cilengitide, c(RGDf(NMe)V), had been developed for the treatment of glioblastoma.
Promising preclinical results demonstrated an inhibition of angiogenesis in models,11 but
Cilengitide failed in phase III as the progression-free survival and overall survival did not turn
out to be superior to standard treatment. One additional major concern was that low doses of
Cilengitide can enhance angiogenesis, which actually favors cancer growth.12 This questioned
not only the integrity of preclinical models for rational drug design13 but also demanded
improving pharmacological properties, e.g. oral bioavailability as the short half-life led to greatly
varying plasma concentrations during administration.14; 15 However, the synthesis of active and
specific ligands can help to overcome the lack of knowledge regarding the problems when
targeting integrins. The pioneering work of Kessler et al. established rational drug design based
on conformation-dependent recognition leading to an αVβ3 receptor model.5;
Peptide
16
ACS Paragon Plus Environment
2

Page 3 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cyclization is an efficient approach to increase affinity and selectivity, as the degree of freedom
is reduced compared to linear peptides.17; 18 Further restraints as turn-inducing amino acids or N-
methylation help to favor a discrete conformation, while the latter also increases metabolic
stability and bioavailability.4 Spatial screening of c(RGDFV), where one amino acid is
substituted by its D-enantiomer inducing βII’-turns showcases how affinity depends on the
orientation of side chains. Combined with other biochemical data, a αVβ3 receptor model evolved
stating that the position after the RGD sequence should be occupied by a hydrophobic amino
acid, while the next position can accommodate variable residues without compromising activity.5
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
Results and Discussion
Our interest in RGD peptides19–22 prompted us to utilize RGD peptides as a benchmark system
for side chain-to-tail cyclization by Suzuki-Miyaura cross-coupling (SMC) in peptides, which
gives access to biaryl motifs. SMC relies on the reaction of a halogenated arene with a boronic
acid and has been demonstrated on protected and unprotected peptides in solution23 or on solid-
phase.24 Only recently, the incorporation of bromotryptophan in pentapeptides that were cyclized
by on-resin SMC to give access to cyclic and bicyclic peptides, was published.25 Chemical
halogenation of tryptophan lacks selectivity and requires elemental bromine or chlorine and high
temperatures. Owing to their benign reaction conditions and high selectivity, halogenases have
emerged as useful biocatalysts for C-H functionalization.26 Flavin-dependent tryptophan
halogenases (FDH) only require halide salts, oxygen and FADH2 as cofactor for regioselective
halogenation. The reaction involves hypohalous acid (HO-X) as the halogenating agent.27–29
A
remarkable feature of these enzymes is their strict regioselectivity, which allows for halogenation
at electronically disfavored position C5 (PyrH), C6 (Thal) or C7 (RebH) of the indole moiety.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 49
1
2
3
4
5
6
Enzymatic synthesis of preparative quantities of halogenated tryptophan is mainly hampered by
low enzymatic stability and insufficient kinetic parameters.
7
8
9
It was shown that the immobilization utilizing cross-linked enzyme aggregates (CLEAs) gives
access to L-7-bromotryptophan on gram scale, which paved the way for applications in the field
of peptide chemistry.30 The introduction of bromotryptophan isomers in RGD peptide analogs
allows using the bromoarene together with an intramolecular boronic acid for cyclization by
SMC. The SMC is compatible with many functional groups and accepts a versatility of starting
materials.31 The success of the SMC also comprises a wide range of applications, for instance
bioorthogonal reactions, labeling, site-selective and late-stage modification and, importantly, the
resulting biaryl motifs possess a huge potential to improve biological activity and expand the
portfolio in drug development.32–34 It has previously been shown that cross-coupling of
unprotected halotryptophans in aqueous media gives fluorescent aryl tryptophans in high
yields.35 Encouraged by these findings a sequential one-pot synthesis for the cross-coupling after
biocatalytic halogenation without intermediary purification was developed by us.36 Furthermore,
the formation of fluorescent aryl tryptophans was utilized to establish a halogenase activity assay
capable to monitor halogenase turnover by fluorescence readout in a microtiter plate that is
applicable for high-throughput screening of mutant libraries.37
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
Biochemistry. For preparative synthesis of halotryptophans, combiCLEAs of the halogenase
RebH, flavin reductase PrnF, and alcohol dehydrogenase (ADH) allow gram-scale synthesis of
L-7-bromotryptophan.30 The halogenase Thal giving 6-bromotryptophan provides similar
conversions whereas PyrH is less active (not shown). Previous studies on RGD peptides revealed
that D-amino acids are crucial building blocks to induce turn structures that can increase
activity.17; 18 Hence, the enzymatic synthesis of D-configured bromotryptophan was attempted.
ACS Paragon Plus Environment
4

Page 5 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RebH is known to accept D-Trp in low concentrations but without reaching full conversion
(Figure 1).30 By reducing the amount of iso-propanol (i-PrOH) to 0.5% the conversion for D-7Br-
Trp (1) reached 90% which made synthetic applications feasible. The lower amount of i-PrOH
may lead to enhanced enzymatic stability. With Thal almost quantitative conversion for D-6Br-
Trp (2) was attained, whereas with PyrH preparative scales for D-5-BrTrp (3) could not be
achieved. A simple desalting step using a C18-column and subsequent purification by RP-HPLC
gives the desired product. Just recently, an alternative approach combining biocatalytic
halogenation with Trp synthase and L-amino acid oxidase RebO-induced dynamic
stereoinversion extended the preparative scope of D-configured halotryptopthans enabling the
synthesis of D-fluoro- or iodotryptophan, which are not accessible with halogenases.38
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. Optimized conditions for the halogenation of D-Trp using combiCLEAs.a
D
-6Br-Trp
combiCLEAs
halogenase PrnF ADH
(
,
,
)
D
-Trp
Br,
30 mM Na
O2, i-PrOH,
H
H
Br
pH 7.4, 25 °C
N
N
NH2
OH
NH2
OH
FADH2
FAD
4 d
PrnF
3 d
O
O
2 d
NAD+
OH
NADH+H+
ADH
1 d
0 d
0
2
4
6
8
10
O
tr / min
Enzyme D-Trp [mM] i-PrOH [%] d Conv. RP-HPLC [%] Yield [%]
Product
RebH
RebH
RebH
Thal
0.5
0.75
1
5
5
5
5
4
63
49
90
99
30 b
37 b
D-7Br-Trp (1)
D-7Br-Trp (1)
D-7Br-Trp (1)
D-6Br-Trp (2)
5
0.5
0.5
72 b
1
quant.c
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 49
1
2
3
4
PyrH
1
0.5
4
9
n.i. d
D-5Br-Trp (3)
5
6
a
Reaction conditions: 100 µM NAD+, 1 µM FAD, 15 mM Na2HPO4, 30 mM NaBr, i-PrOH in a
7
8
9
b
final volume of 1 L H2O, pH adjusted with H3PO4; after desalting over C18-column and
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
preparative RP-HPLC; c after desalting over C18-column; d n.i.: not isolated.
Chemistry. Peptide Synthesis. The Fmoc/tBu strategy was applied to obtain the linear peptides
by solid-phase peptide synthesis (SPPS). N-Terminal protection of bromotryptophan was
introduced using FmocOSu to give Fmoc-D-Br-Trp-OH (4a-b) and Fmoc-L-Br-Trp-OH,25 resp.,
which were loaded on Rink Amide resin. Alkyl or aryl boronic acids (5a-f) for SMC on-resin or
in solution were attached as the final residue to the N-terminus. N-Methylation of amino acids
was performed on-resin by activating the N-terminus with 2-nitrobenzenesulfonyl chloride (o-
NBS) prior to alkylating with dimethyl sulfate (DMS).39 Peptides were cleaved from the resin
with a mixture of trifluoroacetic acid (TFA), triisopropylsilane (TIPS) and water with removal of
protecting groups. The reversed peptide was synthesized on 2-chlorotrityl chloride resin (CTC)
and was cyclized between Gly and Asp to avoid epimerization. To retain protecting groups the
peptide was cleaved ten times with 1% TFA in DCM. The peptide was cyclized under high
dilution using HATU and HOAt and the protecting groups were removed with a mixture of TFA,
TIPS and water. The crude peptides were purified using RP-HPLC.
Side chain-to-tail cyclization by Suzuki-Miyaura cross-coupling. The SMC was carried out to
effect side chain-to-tail cyclization between bromotryptophan and the boronic acid. At first,
intramolecular cross-couplings in solution starting from linear precursor peptides that were
cleaved from resin and purified by RP-HPLC (peptides 6 - 10) were investigated (Scheme 1).
Optimization studies with peptide 6 were conducted on analytical scales (1 µmol). As Na2PdCl4
ACS Paragon Plus Environment
6

Page 7 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
together with water soluble sSPhos (sodium 2’-dicyclohexylphosphino-2,6-dimethoxy-1,1’-
biphenyl-3-sulfonate hydrate) as phosphine ligand yielded excellent conversions for cross-
couplings of free bromotryptophan with different arylboronic acids in solutions, this
methodology was employed for peptide cyclization.36 Under high dilution conditions (0.2 mM),
which is favorable to avoid intermolecular SMC, no product was detectable by MALDI-TOF-
MS analysis. Increasing the substrate concentration to 2.0 mM led to significant conversion
without formation of undesired dimer (see Supporting Information, Figure S1). Further
improvements were obtained by increasing the temperature to 100 °C and the addition of a
chaotropic salt like guanidinium hydrochloride (GdnHCl, see Supporting Information, Figure S2-
3). It is supposed that the latter disrupts secondary structures detrimental to cross-coupling and
thus allowing a smoother cyclization.40 Unfortunately, some starting material decomposes by
deborylation. This common side reaction could, however, not be suppressed by using degassed
solvents under inert conditions.41 Although the optimized reaction conditions showed only minor
formation of side products, on-resin SMC offers important advantages, e.g. a yield-lowering
purification step can be avoided. Therefore, conditions by Afonso et al. were adapted for on-
resin cross-coupling with Pd2(dba)3, potassium fluoride and a mixture of 1,2-dimethoxyethane
(DME)/EtOH/H2O.42 The reaction was monitored by MALDI-TOF-MS or LC-MS for all cyclic
peptides (11 - 24) and finally no starting material or side product could be detected. In general,
the on-resin reaction proceeded smoothly with full conversion of the linear peptide and without
significant side products (see Supporting Information, Figure S4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of RGD peptides and side chain-to-tail cyclization on-resin and in solution
by SMC.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 49
1
2
3
4
Biohalogenation
Solid-phase peptide synthesis
5
6
Br
Br
7
8
9
NH
NH
a)
NH
OH
O
H
R1
N
OH
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
*
N
H2N
H2N
*
*
*
O
R2 R3
O
O
O
d-e)
d-e)
b-c)
Xaa3
Xaa2
Xaa3
Xaa1 Xaa2
O
Xaa1
Xaa4
O
R3
N
Xaa4
2
Xaa5
NH2
R3
N
O
2
3
n
3
NH2
HN
*
NH2
*
4
N
O
*
R3
O
4
6
O
6
N
RGDF
7
H
N
7
H
11 - 20
6 - 10
21 - 24
SMC in solution
SMC on-resin
No. Xaa1-5
*
L
R3
H
H
H
H
H
H
H
H
Connection No. Xaa1-5/ n
*
L
R3
Connection
6
7
8
9
RGDfV
RGDf
RGD
RGDf
RGD
RGDF
RGD
RGDF
RGDF
RGD
7-4
7-4
7-4
7-3
7-3
7-4
7-4
6-4
16
17
18
19
20
21
22
23
24
RGDf
RGD
RGDF
RGDF
RGDF
5
4
3
2
Me 7-4
Me 7-4
Me 6-4
Me 7-3
Me 7-2
Me
L
L
L
D
D
D
D
D
D
D
L
10
11
12
13
14
15
L
D
D
D
D
D
Me
Me
Me
Me 7-4
Me 7-4
aReagents and conditions: (a) combiCLEAs (halogenase-PrnF-ADH), NAD+, FAD, Na2HPO4,
NaBr, i-PrOH, 1 L water, pH 7.4, O2,; (b) TFA/H2O/TIPS; (c) Na2PdCl4, sSPhos, K3PO4,
GdnHCl, water/dioxane (3:1), 100 °C, µwave, 0.5 h; (d) Pd2(dba)3, sSPhos, KF,
DME/EtOH/H2O (9:9:1), 120 °C, µwave, 0.5 h; (e) TFA/H2O/TIPS. R1 = alkyl or aryl boronic
acid, R3 = H or CH3.
1
The LC-MS and H-NMR analyses revealed for most peptides broadened signals or double
peaks, which was observed for both biaryl- and alkyl-aryl-cyclized peptides and was attributed to
1
conformers or diastereomers (see Supporting Information). In many H-NMR spectra more than
ACS Paragon Plus Environment
8

Page 9 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
one signal was found for the indole-NH, which is a good probe due to a pronounced downfield
shift compared to other signals. For peptide 17 no isomers were visible in the 1H-NMR spectrum
whereas peptide 16 that is elongated by one D-Phe shows two signals for the indole-NH.
Temperature studies in DMF-d7 from 248 – 328 K did not show any change of integrals which
hints toward the presence of diastereomers (see Supporting Information Figure, S5). It has been
reported previously that tryptophans involved in cyclic biaryls, e.g. in complestatin, are known to
form atropisomers caused by the indole moiety.43 Accordingly, molecular modeling studies were
performed to clarify these observations (vide infra). Additionally, peptide 19 was synthesized in
a reverse manner starting from a suitably protected biaryl with a final backbone cyclization to
examine whether this phenomenon is only caused by SMC cyclization (Scheme 2). For this
approach, SMC between compound 25 and 3-(tert-butoxycarbonyl)phenylboronic acid was
initially performed in solution. Bromotryptophan had to be converted into the corresponding
amide to maintain structural analogy as the peptides were synthesized on Rink Amide resin.
After removal of the Boc and tert-butyl protecting groups the N-terminus was Fmoc-protected
(27) and this biaryl building block was coupled to the peptide precursor. After cleavage from the
resin, cyclization of the linear peptide was carried out using HATU and HOAt under high
dilution conditions leading to quantitative conversion. Purified peptide 28 led to the same
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
features in the H-NMR spectrum concerning the indole-NH shift as peptide 19 that had been
obtained by on-resin cyclization (see Supporting Information Figure, S6). This indicates that the
formation of isomers is not conditioned by SMC but a consequence of different, stable
isomers/conformers. For further validation molecular dynamics simulations were applied to
elucidate the role of conformers and atropisomers. It is further important to note that peptides 19
and 28 do not significantly differ in affinity (see Table 1).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 49
1
2
3
4
Scheme 2. Reversed synthesis of peptide 19.a
5
6
7
8
O
O
Br
tBu
Br
H
H
H
N
N
N
b)
c)
d-e)
a)
H
9
N
HN Boc
NH2
NH2
OH
NH2
OH
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
H
N
Boc
NH2
O
O
O
26
1
25
O
H
N
O
OH
O
O
R(Pbf)G-OH
NH
H
f-g)
H
O
h-i)
N
NH2
NH
H
N
N
NH
NH
O
O
NH
Fmoc
N
H2N
HN
HN
N
FD(tBu)-H
NH2
O
NH2
O
O
O
O
OH
27
28
: 19%
aReagents and conditions: (a) combiCLEA (RebH, ADH, PrnF), NAD+, FAD, Na2HPO4,
NaBr, i-PrOH, 1 L water, pH 7.4, O2; (b) (1) SOCl2, MeOH 0 °C  rt, 36 h; (2) Et3N, Boc2O,
MeOH, rt, 24 h; (3) NH4OH, MeOH, rt, 36 h; (c) Na2PdCl4, sSPhos, K3PO4, water/dioxane,
100 °C, 1 h, Ar; (d) TFA/DCM/thioanisole/TIPS/EDT/H2O, rt, 3 h; (e) FmocCl, Na2CO3,
water/dioxane, 0 °C  rt, 1 h, f) HATU, HOAt, DIEA, rt; g) TBTU, DIEA, rt; h) HATU, HOAt,
DIEA, rt; i) TFA/H2O/TIPS.
Biological Evaluation
The binding affinity towards αVβ3 and α5β1 was determined in an ELISA-like assay with isolated
integrins44;
and cell-adhesion19 assays with human melanoma cancer cells (WM115)
45
overexpressing αVβ3 to establish structure-activity relationships. An array of side chain-to-tail
cyclized RGD peptides was synthesized which differ in length, presence and position of D-amino
acid or N-methylation, connectivity and alkyl or aryl moieties attached to the indole. In contrast
to lead peptide c(RGDfV), Phe and Val were substituted by the hydrophobic biaryl moiety
ACS Paragon Plus Environment
10

Page 11 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
obtained by cross-coupling between Br-Trp and a boronic acid. Among the first generation of
peptides (7 - 13) the best affinity towards αVβ3 is observed with hexapeptide 11 (counting the
aryl moiety as an amino acid). Spatial screening studies revealed a favorable stabilization of
turn-structures upon incorporation of D-amino acids for the lead peptide.17 Although the
conformation for the cross-coupled peptides could not be elucidated, D-configured tryptophan
resulted in overall increased affinities. Reports about the conformation of c(RGDfV) demonstrate
a tight γ turn around Gly whereas for c(RGDFv) the position of the γ turn was shifted to Asp
resulting in an opposing orientation of the Arg and Asp side chains and decreased affinity.17; 18
However, in our case switching the position of the D-amino acid from Trp (11) to Phe (7)
decreases 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
Interestingly, the connectivity of the indole and phenyl moiety seems to have a major impact on
the affinity as observed by the low affinity of peptides 9 and 10. The connectivity can be
influenced by the choice of the boronic acid (ortho, meta, para) and the position of the bromide
in the indole using different halogenases (C6, C7). The second generation of peptides was N-
methylated because this has been reported to increase oral bioavailability, conformational
restriction and metabolic stability.4 With peptide 11 as the first lead structure a small array
containing N-methylated RGD peptides was synthesized. For comparison, peptide 11 was N-
methylated to give peptide 14 which shows αVβ3 affinity comparable to 11 but lacks selectivity
over α5β1. However, hexapeptide 14 displays higher affinity towards αVβ3 than pentapeptide 15.
Regarding the connectivity it was concluded from the first results that a 7-3 connection might be
promising. Indeed, peptide 19 displays the highest affinity towards αVβ3 in this set as shown by
both cell adhesion assay (8.9 µM) and ELISA (5.4 nM). Especially the selectivity over α5β1 for
peptide 19 or 11 is a major improvement compared to Cilengitide and meets the request of a
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 49
1
2
3
4
5
6
7
8
9
highly discriminating ligand. Another interesting approach was to create a small set of
compounds with alkyl boronic acids with a chain length of three to six C-atoms. The affinity for
αVβ3 increases with a shorter alkyl chain, most likely due to the decreased conformational
flexibility. The lack of affinity in cell-adhesion assays, e.g. for peptide 24, might derive from
inhomogeneous integrin expression levels which is a drawback in cell-based tests. For a selection
of important derivatives affinity towards platelet integrin αIIbβ3 revealed no or very low affinity
avoiding potential bleeding issues due to deficient platelet aggregation.
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. IC50 values determined by a cell-free method using isolated integrins and WM115 cells
overexpressing αVβ3.
Xaa3
Xaa4
Xaa1 Xaa2
O
R3
N
O
2
n
HN
*
R3
NH2
3
NH2
N
*
O
4
O
RGDF
6
N
7
H
7 - 20
21 - 24
IC50 αVβ3
IC50 α5β1 IC50 αIIbβ3
IC50 WM115
No.
Xaa1-4 / n
* R3 Connection
[nM]
579
491
65
[nM]
5669
3660
983
[nM]
n.d.b
n.d.b
n.d.b
n.d.b
[µM]
> 500
> 500
> 500
> 500
51±9.4
> 500
310±37
7
RGDf
RGD
L
L
L
L
D
D
D
H
H
H
H
H
H
H
7-4
7-4
7-3
7-3
7-4
7-4
6-4
8
9
RGDf
RGD
10
11
12
13
19
8029
RGDF
RGD
17.7±2.6 9287±2577 > 10000
41.3±6.9 5944±2253 > 10000
RGDF
59.2±5.5
> 10000
n.d.b
ACS Paragon Plus Environment
12

Page 13 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
14
15
16
17
18
19
20
21
22
23
24
28a
RGDF
RGD
RGDf
RGD
RGDF
RGDF
RGDF
5
D Me
D Me
L Me
L Me
D Me
D Me
D Me
D Me
D Me
D Me
D Me
D Me
7-4
7-4
7-4
7-4
6-4
7-3
7-2
23.0±4.6
537±229
835±263
>10000
> 10000
n.d.b
11.1±2.2
> 500
95.2±28.9 2161±530
239±52 >10000
14.9±3.3 2722±475
n.d.b
> 500
n.d.b
> 500
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
> 10000
5804±541
n.d.b
244±48
9.4±1.3
158±36
> 500
5.4±0.6
547±87
64.6±7.7
1.92±0.4
38.9±10.2 419±122
36.2±10.0 563±159
n.d.b
4
n.d.b
> 500
3
34.0±2.4
12.8±0.8
13.9±1.4
154±10
325±21
n.d.b
n.d.b
> 500
2
5028±644
n.d.b
250±39
n.d.b
RGDF
7-3
29 c(RGDf(NMe)V)
0.54±0.02 15.4±4.0
n.d.b
n.d.b
30
31
c(RGDfV)
n.d.b
n.d.b
n.d.b
n.d.b
n.d.b
0.75±0.05
n.d.b
Tirofiban
1.3±0.2
asynthesized in reversed order compared to peptide 19; b n.d.: not determined. When sufficient
amounts of peptide were available cell-free ELISA was performed twice in duplicates otherwise
only in duplicates and cell-adhesion assays were performed twice with four technical replicates
for active compounds.
Plasma stability in human plasma was investigated with one representative peptide to evaluate
degradation and metabolic stability (see Supporting Information, Figure S7). For peptide 19 good
stability in human plasma (t1/2 > 24 h) is observed, whereas procaine as a positive control (t1/2
=
0.5 h) is rapidly degraded.
Molecular Modeling. Further examination of spatial structures by molecular modeling was
necessary as experimental data suggested the formation of isomers or conformers. While the
spectrum of compound 17 allows peak assignment corresponding to a single conformation,
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 49
1
2
3
4
5
6
7
8
9
multiple sets of signals for compounds 14 and 19 point towards the coexistence of different
conformations. Extensive molecular dynamics (MD, Amber16/AmberTools17) simulations in
explicit solvent (i.e., DMSO and water) at 300 K as well as at 350 K were performed to assess
the convergence of conformational sampling, to gain insights into the structural preferences of
these three compounds, and to further rationalize their observed activity towards integrins. The
last 100 ns of four parallel MD simulations were gathered for each system and analyzed by
means of a two-step structural clustering based i) on the root-mean-square-deviation (RMSD) of
backbone atoms only and ii) on the RMSD of all non-hydrogen atoms in the molecule (see
Supporting Information, Figure S8-10).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the representative structure of all three compounds obtained from MD simulations in DMSO
at 300 K the polar side chains fold towards the core of the cyclic peptide to minimize their
contact with the apolar solvent (Figure 2). This results in an interaction between the carboxylate
group of Asp and the guanidinium group of Arg that remains stable along the dynamics. Only
one conformation of compound 17 was significantly populated (i.e., greater than 10%) in the
simulation in DMSO. Clustering based on backbone RMSD for this molecule resulted in two
clusters with populations of 43% and 35%, for which the representative structures differed only
by a 180˚ rotation of the phenyl group of the linker about its principal axis (see Supporting
Information, Figure S11). Since the carbon atoms of this phenyl ring are pairwise non-
distinguishable, the two geometries are strictly identical and result in a single conformation with
population of 78%. In contrast, compound 14 is characterized by two conformations with
populations of 60% as well as 39% and differing in the orientation of the peptide bond between
the Arg and Gly amino acids of the RGD motif. This difference results in the loss of a hydrogen
bond interaction between the amide NH group of the mentioned peptide bond and the
ACS Paragon Plus Environment
14

Page 15 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
carboxylate group of the Asp residue in the second conformation. Similarly, three conformations
of compound 19 were found to be stable throughout the trajectory in DMSO, with populations of
64%, 18% and 14%, respectively. The former two conformations are atropisomers and differ in
the relative orientation of the phenyl and indol moieties in the linker of the cyclic molecule
(Figure 2). The third conformation did not show any significant difference compared to the main
one (see Supporting Information, Figure S12). For all three compounds, the simulations at 350 K
yielded a clustering analysis nearly identical to that performed at 300 K, with slightly different
populations of the clusters (see 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
Even though the loss of hydrogen bond in compound 14 and the presence of atropisomers in
compound 19 contribute to minor structural changes, the relative population of the two resulting
conformers for these molecules is significant enough to be also detected by NMR spectroscopy
thus explain multiple signal sets. In contrast, the single conformation identified for compound 17
correlates with the correspondingly neat spectrum recorded experimentally.
1
Figure 2. H-NMR spectra (left panel) and predicted representative structures as obtained from
molecular dynamics simulations at 300 K (right panel) of compounds 17, 14 and 19 in DMSO.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 49
1
2
3
4
5
6
7
8
9
The first and second most populated conformations of 14 and 19 are depicted with carbon atoms
shown in gray and pink, respectively. Relative populations are reported for each conformation
based on backbone RMSD clustering. All other conformations of compound 17 have populations
lower than 8%. Backbone RMSD of the two structures are reported for compound 14 and 19.
The dotted-delimited areas indicate the main differences between the conformations. Only
relevant hydrogen atoms are represented for clarity.
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
Similar MD studies were also performed in water to get a closer insight into the conformational
properties in a more biologically relevant context. This would be comparable to the conditions of
the affinity assays as well as the desired application. As illustrated in Figure 3, the polar side
chains of Arg and Asp amino acids in the RGD moiety of the three cyclic peptides are separated
in water and are individually well solvated. The average distance between the carbon atoms of
the guanidinium and carboxylate groups (R-D distance), as obtained from the last 100 ns of four
independent MD runs for each molecule at 300 K, is: 7.5  2.5 Å, 9.4  1.9 Å, and 9.3  2.8 Å,
for compounds 17, 14 and 19, respectively. Structural clustering revealed only one, highly
populated conformation for compound 17 in water, with a population of 99%. In contrast,
compounds 14 and 19 are characterized by one particular main conformation with a population
of 42% and 52%, respectively, and more than 10 other low-occupied conformations (1-8%).
In the crystal structure of αVβ3 by Xiong et al. (PDB: 1L5G),46 the co-crystallized cyclic RGD
peptide presents an extended conformation of the polar side chains with an R-D distance of
13.7 Å. This spatial separation between the guanidinium and carboxylate groups is crucial for
receptor binding since both groups interact with distinct binding pockets. Analysis of the average
R-D distance as obtained from MD simulations in water indicates that such a separation between
the Arg and Asp side chains is rather unlikely for compound 17, while the target distance falls
ACS Paragon Plus Environment
16

Page 17 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
near the range of observed values for compounds 14 and 19. Comparison of the representative
structures for the three molecules showed that none of the geometries obtained adopt the exact
conformation that Xiong et al. identified for their active RGD containing ligand (see Supporting
Information, Figure S13). However, the rigidity of compound 17 compared to the relative
flexibility of compounds 14 and 19 indicates that the latter two are more prone adopting a
conformation that favorably interacts with the ligand-binding cavity of the integrin. The
structural features, as revealed from MD simulations, correlate with the measured activity of the
three compounds and explain the poor performance of compound 17 as a ligand for αVβ3. As the
intrinsic flexibility of 17 is limited because of its rigid structure, a favorable interaction with the
integrin binding cavity is therefore not possible.
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
17

Journal of Medicinal Chemistry
Page 18 of 49
1
2
3
4
5
6
7
8
9
Figure 3. Representative structure for the main conformation of compounds 17, 14, and 19 in
water as obtained from molecular dynamics simulations at 300 K. The relative population of the
conformation based on backbone RMSD is reported for all three molecules. The average R-D
distances are measured from the carbon atom of the Arg guanidinium groups to the carbon atom
of the Asp carboxylate group, and reported with their corresponding standard deviations. Only
relevant hydrogen atoms are represented for clarity as well as water molecules within 3 Å of the
Asp or Arg side chain with corresponding hydrogen bond interactions depicted with dashed
lines.
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
Selective and active ligands for αVβ3 remain a crucial target in medicinal chemistry and represent
a good benchmark system to explore the influence of an intramolecular side chain-to-tail
cyclization of RGD peptides. Enzymatic halogenation delivers bromotryptophan which provides
a powerful handle for late-stage modifications by means of cross-coupling. For this approach
preparative amounts of L- and D-bromotryptophan are required that was initially hampered by
low conversion of D-tryptophan and optimized to nearly full conversion by reducing the excess
of co-substrate i-PrOH for halogenases RebH and Thal.
Robust methods for cyclization between an N-terminal boronic acid and the indole ring of
bromotryptophan are established, capable of cross-coupling in solution and on-resin.
Particularly, cyclization on-resin simplifies the reaction setup; giving higher overall yields while
typical side products, as deborylation, are diminished. Experimental data and extensive MD
simulations for three representative molecules reveal the occurrence of stable, distinct
conformers or atropisomers.
ACS Paragon Plus Environment
18

Page 19 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In vitro affinity assays using isolated integrins and cell adhesion assays show improved αVβ3
affinities for peptides containing a D-amino acid at the position of the former bromotryptophan
whereas N-methylation plays an ambiguous role. In particular, the connectivity between the
indole and aromatic moiety has a major impact on both affinity and selectivity. Notably, a 7-3
connection is superior leading to peptide 19 with low nanomolar αVβ3 affinity in this array.
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
Introducing non-canonical halotryptophans, generated by biocatalysis, into RGD peptides as a
benchmark system delivers not only active ligands but a powerful platform for side chain-to-tail
cyclizations utilizing cross-coupling reactions. This further expands Nature’s toolkit of unusual
biaryl systems which gain considerable attention as pharmaceutical building blocks.
Experimental Section
General Procedures. All solvents and chemicals were used as purchased without further
purification. Fmoc-Arg(Pbf)-OH, Fmoc-Asp(tBu)-OH, Fmoc-Gly-OH and Fmoc-Phe-OH were
purchased from Iris Biotech GmbH. Aromatic boronic acids were purchased from Sigma Aldrich
or Alfa Aesar. 2-Boronatobenzoic acid (Sigma Aldrich) led to insufficient coupling and was
synthesized as described elsewhere.47 Moisture- and air-sensitive reactions were conducted in
oven-dried glassware under argon. The water used for reactions, work-ups and purifications was
purified by a Millipore Milli-Q biocel device. If necessary, DCM was freshly distilled from CaH2
and THF from sodium/benzophenone. All final compounds were analyzed by analytical RP-
HPLC to confirm a purity of 95% purity (220 nm).
≤
Analytical RP-HPLC was performed on a Thermo Scientific Accela 600 equipped with a UV-
6000 LP detector, a P-4000 pump, a Hypersil Gold C18 column (3 µm column, 150×2.1 mm).
Analytical LC-MS and determination of ESI-HRMS was performed on an Agilent 6220 TOF-
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 49
1
2
3
4
5
6
7
8
9
MS (Agilent Technologies) with a dual ESI-source operating with a spray voltage of 2.5 kV,
1200 HPLC system with autosampler, degasser, binary pump, column oven, diode array detector
and a Hypersil Gold C18 column (3 µm, 150×2.1 mm). Nitrogen was generated by a nitrogen
generator NGM 11 and served as nebulizer and dry gas. External calibration was performed with
ESI-L Tuning Mix (Agilent Technologies). MALDI-TOF/TOF-MS was conducted with an
Ultraflex (Bruker Daltonik) operated in reflectron positive mode using an LTB nitrogen laser
MNL200 (337 nm, 50 Hz, 1000 shots/spectrum, maximum resolution: 20000) using DHB (2,5-
dihydroxybenzoic acid) or CHCA (α-cyano-4-hydroxycinnamic acid) as matrix. Samples were
dissolved in mixtures of acetonitrile and water and calibration was conducted with PEG 400-
1200. FlexControl 3.0 and FlexAnalysis 3.4 (Bruker Daltonik) were used for recording and
processing. Preparative HPLC was performed using a Merck-Hitachi LaChrom HPLC consisting
of interface D-7000, pump L-7150, detector L-7420 and a Hypersil Gold C18 column (1.9 µm,
250×21.2 mm) or a Hypersil Gold C18 column (7 µm, 250×10.0 mm). NMR spectra were
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
13
recorded on a Bruker DRX-500, Avance III 500 (1H: 500 MHz, C: 126 MHz) or Avance 600
(1H: 600 MHz, 13C: 151 MHz) spectrometer. Chemical shifts are given in parts per million (ppm)
1
relative to tetramethylsilane, residual solvent peaks for H and 13C were used as internal
standard. For monitoring reaction progress with TLC, silica gel 60 coated aluminum sheets with
F254 fluorescence indicator were used with solvent mixtures specified in the corresponding
experiment. Spots were visualized using UV light (254 or 366 nm) or KMnO4 staining solution.
Column chromatography was performed with silica gel 60 (Merck, 40-63 µm).
Enzymatic halogenation. Expression and purification of halogenases RebH, Thal and PyrH and
cofactor regenerating enzymes PrnF and ADH were conducted according to our previously
published procedures48; 26 as well as the formation of combiCLEAs.30 E. coli cells from 1.5 L
ACS Paragon Plus Environment
20

Page 21 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
culture medium containing overexpressed halogenases (RebH, Uniprot ID: Q8HKZ8; Thal,
UniprotID: A1E280 or PyrH, Uniprot: A4D0H5) were suspended in Na2HPO4 buffer (100 mM,
pH 7.4) and lysed three times by French press. The lysate was spun down (4200×g, 4 °C,
30 min) and the crude supernatant was transferred into a fresh tube. After addition of PrnF
(2.5 U/mL), ADH (1 U/mL) and finely ground ammonium sulfate (16.2 g) the mixture was
incubated for 1 h at 4 °C in a tube rotator. For crosslinking glutardialdehyde (RebH: 0.5%, Thal:
1.0%, PyrH:1.25% w/v) was added and the mixture was further incubated for 2 h at 4 °C. Next,
combiCLEAs were centrifuged (4200×g, 4 °C, 30 min) and washed three times thoroughly with
Na2HPO4 buffer (100 mM, pH 7.4, 30 mL; 4200×g, 4 °C, 30 min). To a solution of Trp (1 mM),
NaBr (30 mM), Na2HPO4 (15 mM), NAD+ (100 µM), FADH (1 µM) and i-PrOH (0.5% v/v) in
water (995 mL) with a pH of 7.4 adjusted by H3PO4 the combiCLEAs were added. For optimal
conversion the combiCLEAs must be finely suspended. The combiCLEAs were shaken for 3 to 5
days until the reaction was complete or conversion stopped. The combiCLEAs were removed by
centrifugation (4200×g, 4 °C, 30 min) and the solvent was evaporated. The residue was dissolved
in water/0.1% TFA and desalted using a manual C18 reversed-phase column (25 g, 400-220
mesh). For equilibration and desalting water/0.1% TFA (200 mL) was used whereas for eluting
CH3CN/0.1% TFA was used (300 mL). Subsequently, the solvent was evaporated, lyophilized
and if necessary purified using RP-HPLC (method 1).
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
D-7Br-Trp (1). The TFA-salt was isolated as a slightly yellowish solid after desalting and
1
purification by RP-HPLC (285 mg, 0.72 mmol, 72%). C11H11BrN2O2; H-NMR (500 MHz,
3
DMSO-d6): δ = 11.28 (s, 1H, Ind-NH), 8.17 (s, 2H, CαNH2), 7.58 (d, 1H, J = 7.9 Hz, Ind-H4),
3
3
3
7.32 (d, 1H, J = 7.5 Hz, Ind-H6), 7.30 (d, 1H, J = 2.3 Hz, Ind-H2), 6.97 (t, 1H, J = 7.7 Hz,
Ind-H5), 4.14 (dd, 1H, 3J = 6.1 Hz, 3J = 6.1 Hz, CαH), 3.30 - 3.21 (m, 2H, CβH).
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 49
1
2
3
4
5
6
D-6Br-Trp (2). The TFA-salt was isolated as a slightly yellowish solid after desalting (397 mg,
1.00 mmol, quant.). C11H11BrN2O2; 1H-NMR (500 MHz, DMSO-d6): δ = 11.22 (s, 1H, Ind-NH),
7
8
9
4
3
8.20 (s, 2H, CαNH2), 7.56 (d, 1H, J = 1.7 Hz, Ind-H7), 7.52 (d, 1H, J = 8.4 Hz, Ind-H4), 7.25
(d, 1H, 3J = 2.4 Hz, Ind-H2), 7.14 (dd, 1H, 3J = 8.4 Hz, 4J = 1.7 Hz, Ind-H5), 4.15 (dd, 1H, 3J =
6.1 Hz, 3J = 6.1 Hz, CαH), 3.26 - 3.24 (m, 2H, CβH).
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
Fmoc-protection of halotryptophan. FmocOSu (1.1 equiv, 1.29 mmol) was added to a solution
of the brominated Trp derivative (1 equiv, 1.17 mmol) and NaHCO3 (3 equiv, 3.51 mmol) in
CH3CN/H2O (8 mL, 6:4) and stirred at rt. The pH was kept between 8-9 with 1 M NaOH and the
reaction progress was monitored by RP-HPLC until completion. The solvent was evaporated
under reduced pressure, the residue dissolved in EtOAc/H2O (1:1, 20 mL) and extracted with
EtOAc (3×20 mL). The combined organic layers were washed with 5% aq. KHSO4 (20 mL), the
organic layer dried over MgSO4 and the solvent evaporated. The crude was purified by RP-
HPLC (method 2) and lyophilized yielding Fmoc-Br-Trp as a white solid.
Fmoc-D-7Br-Trp (4a). The compound was isolated as a colorless solid (458 mg, 0.91 mmol,
78%). 1H-NMR (600 MHz, DMSO-d6): δ = 12.74 (bs, 1H, COOH), 11.11 (s, 1H, Ind-NH), 7.88
(d, 2H, J = 7.6 Hz, Fmoc-H4, Fmoc-H5), 7.73 (d, 1H, J = 8.2 Hz, CαNH), 7.66 - 7.59 (m, 3H,
3
3
3
Fmoc-H1, Fmoc-H8, Ind-H6), 7.40 (q, 2H, J = 7.0 Hz, Fmoc-H3, Fmoc-H6), 7.34 - 7.24 (m,
4H, Fmoc-H2, Fmoc-H7, Ind-H2, Ind-H4), 6.95 (t, 1H, 3J = 7.7 Hz, Ind-H5), 4.27 - 4.15 (m, 4H,
CαH, Fmoc-H9, Fmoc-CH2), 3.20 (dd, 1H, J = 14.7 Hz, J = 4.6 Hz, CβH1), 3.03 (dd, 1H, J =
2
3
2
14.7 Hz, J = 9.8 Hz, CβH2). 13C-NMR (126 MHz, DMSO-d6): δ = 173.5, 155.9, 143.7, 143.7,
3
140.6, 134.3, 128.9, 127.6, 127.0, 125.2, 125.1, 123.4, 120.0, 119.9, 117.8, 111.7, 65.6, 54.8,
46.5, 26.8. HRMS: m/z calculated for C26H21BrN2O4+H+ [M+H+] = 505.07574. Found:
505.07510.
ACS Paragon Plus Environment
22

Page 23 of 49
Journal of Medicinal Chemistry
1
2
3
4
Fmoc-D-6Br-Trp (4b). The compound was isolated as a colorless solid (scale: 1.00 mmol,
5
6
1
359 mg, 0.71 mmol, 71%). H-NMR (600 MHz, DMSO-d6): δ = 12.71 (bs, 1H, COOH), 11.00
7
8
9
3
3
(s, 1H, Ind-NH), 7.88 (d, 2H, J = 7.6 Hz, Fmoc-H4, Fmoc-H5), 7.70 (d, 1H, J = 8.3 Hz,
3
3
CαNH), 7.64 (t, 2H, J = 8.3 Hz, Fmoc-H1, Fmoc-H8), 7.57 - 7.50 (m, 2H), 7.41 (q, 2H, J =
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
7.1 Hz, Fmoc-H3, Fmoc-H6), 7.32 - 7.26 (m, 2H), 7.20 (d, 1H, J = 2.2 Hz, Ind-H2), 7.10 (dd,
1H, 3J = 8.5 Hz, 5J = 1.6 Hz), 4.23 - 4.15 (m, 4H, CαH, Fmoc-H9, Fmoc-CH2), 3.17 (dd, 1H, 2J =
14.6 Hz, J = 4.4 Hz, CβH1), 3.01 (dd, 1H, J = 14.6 Hz, J = 9.9 Hz, CβH2). C-NMR (126
MHz, DMSO-d6): δ = 173.5, 155.9, 143.7, 143.7, 140.6, 136.9, 127.6, 127.0, 126.2, 125.2,
124.7, 121.2, 120.0, 113.9, 113.7, 110.6, 65.6, 54.9, 46.5, 26.6. HRMS: m/z calculated for
C26H21BrN2O4+H+ [M+H+] = 505.07574. Found: 505.07748.
3
3
2
13
Peptide Syntheses: General Procedures. Peptide synthesis was conducted according to the
Fmoc/tBu-strategy in a plastic syringe fitted with a polypropylene porous disk at rt. Washing
steps were performed after each reaction with DMF (5×) or NMP (5×) with 10 mL/g resin.
Solvents and soluble reagents were removed by suction. During couplings the resin was shaken
with a horizontal shaker. SPPS was conducted on scales between 0.1-0.5 mmol using 10 mL/g
resin for reactions. Reaction control was performed by MALDI-TOF-MS or LC-MS.
Loading of Rink Amide resin. For on-resin side chain-to-tail cyclization Rink Amide resin was
used. The resin (0.5 mmol/g) was swollen in DMF (10 mL/g) for 15 min. The resin was
deprotected twice with 20% piperidine/DMF, 0.1 M 1-hydroxybenzotriazole (HOBt) for 20 min
and washed. Fmoc-Br-Trp-OH (1 equiv) and O-(7-azabenzotriazol-1-yl)-N,N,N′,N′-
tetramethyluronium-hexafluorophosphate (HATU) (1.3 equiv) were dissolved in DMF,
combined, diisopropylethylamine (DIEA) (3 equiv) was added and the mixture was shaken for
30 sec. This solution was added to the resin and incubated for 3 h at rt. Afterwards the resin was
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 49
1
2
3
4
5
6
7
8
9
filtered, washed and capping was carried out twice using Ac2O (10 equiv) and pyridine
(10 equiv) in DMF. The resin was dried in a desiccator over night or washed with Et2O and dried
under vacuum (50 mbar). The loading capacity ranged between 95-100%.
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
Loading of CTC resin. For peptides cyclized via the backbone CTC resin was used. CTC
(1.0 mmol/g) was swollen under argon in dry DCM (10 g/mL) for 15 min. Fmoc-Gly-OH (1
equiv) dissolved in dry DCM and DIEA (10 equiv) was added within 5 min to the resin. After
overlaying the resin with argon, it was shaken for 3 h at rt. For capping MeOH (2 mL/g) was
added to the resin and incubated for further 10 min. After filtration, washing and drying the resin
loading was estimated 96%.
Manual SPPS. For Fmoc-deprotection the resin was treated twice with a solution of 20%
piperidine/DMF, 0.1 M HOBt for 20 min. Coupling steps were repeated twice. Fmoc-Xaa-OH (5
equiv) and O-(benzotriazol-1-yl)-N,N,N',N'-tetramethyluronium-tetrafluoroborate (TBTU) (5
equiv) were dissolved in DMF, combined and DIEA (10 equiv) was added. After inverting the
solution for 0.5 min, it was added to the resin and incubated for 15 min. For Fmoc-Br-Trp-OH
and compound 26 (1.1 equiv) coupling was carried out only once in NMP with HATU (1.3
equiv), 1-hydroxy-7-azabenzotriazole (HOAt) (1.3 equiv) and DIEA (3 equiv) for 3 h. A
subsequent capping was conducted twice with Ac2O (10 equiv) and pyridine (10 equiv) in DMF
for 20 min. Coupling of aromatic boronobenzoic acid (5 equiv) was performed twice using
TBTU (5 equiv) and DIEA (10 equiv) in NMP for 20 min. Coupling of alkyl boronobenzoid acid
(1.3 equiv) was performed once using TBTU (1.3 equiv) and DIEA (3 equiv) in DMF for 3 h.
On-resin N-Methylation. The Fmoc-group was deprotected and the resin washed with NMP.
All steps were repeated twice. For the NBS-protection a freshly prepared solution of o-NBS (4
ACS Paragon Plus Environment
24

Page 25 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
equiv) with 2,4,6-collidine (10 equiv) in NMP was incubated with the resin for 15 min. Then a
solution of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) (3 equiv) and DMS (10 equiv) in NMP
was incubated with the resin for 15 min. The completion of the N-methylation was monitored by
RP-HPLC. To cleave the o-NBS group 2-mercaptoethanol (10 equiv) and DBU (5 equiv) in
NMP were incubated with the resin for 15 min. The next Fmoc-Xaa-OH (3 equiv) was attached
using HATU (3 equiv), HOAt (3 equiv) and DIEA (6 equiv) in NMP for 3 h.39
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
Cleavage of peptides from CTC resin. Peptides to be cyclized in solution were cleaved by
treating 10× for 5 min with a mixture of 1% TFA in DCM. The cleavage solution was filtered
into excess of i-PrOH (100 mL) and the combined fractions were evaporated under reduce
pressure. The crude peptide was purified by RP-HPLC (method 1).
General synthesis of alkyl boronic acids.
General procedure for borylation. A Schlenk tube charged with CuI (0.2 equiv),
bis(pinacolato)diboron (2 equiv) and MeOLi (3 equiv) was evacuated and filled with argon (3×).
DMF (0.5 mL) was added under argon and the suspension was heated at 60 °C for 15 min and
cooled to 40 °C for 20 min. A solution of tert-butyl 5-bromopentanoate (1 equiv, 0.56 mmol) or
tert-butyl 6-bromohexanoate (1 equiv, 1.54 mmol) in DMF (0.5 mL) was added under argon and
was stirred at 40 °C for 24 h. The suspension was diluted with EtOAc, filtered through celite,
concentrated and purified by silica column chromatography (EtOAc/petroleum spirit, 10:90)
giving a brown oil. 49
tert-Butyl 5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pentanoate (5a). Yield: 48%,
76 mg, 0.27 mmol. C15H29BO4.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 49
1
2
3
4
tert-Butyl 6-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)hexanoate (5b). Yield: 53%,
5
6
243 mg, 0.82 mmol. C16H31BO4.
7
8
9
General procedure fo tBu-deprotection. For the tBu-deprotection compound 5a or 5b (0.42-
0.24 mmol) was taken up in a mixture of DCM/TFA (4 mL, 1:1, v/v) and stirred at rt for 2 h. The
solvent was concentrated in vacuo to give the deprotected compounds 5c and 5d as a colorless
oil.
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
5-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)pentanoic acid (5c). Yield: 88%, 48.2 mg,
3
0.21 mmol. C11H21BO4; 1H-NMR (500 MHz, CDCl3): δ = 2.32 (t, 2H, J = 7.5 Hz,
3
3
BCH2CH2CH2CH2C), 1.62 (p, 2H, J = 7.5 Hz, BCH2CH2CH2CH2C), 1.44 (p, 2H, J = 7.8 Hz,
3
BCH2CH2CH2CH2C), 1.22 (s, 12H, (C(CH3)2), 0.77 (t, 2H, J = 7.9 Hz, BCH2CH2CH2CH2C).
13C-NMR (126 MHz, CDCl3): δ = 180.1, 83.1, 34.3, 27.3, 24.8, 23.6.
6-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)hexanoic acid (5d). Yield: 99%, 197 mg,
3
0.81 mmol. C12H23BO4; 1H-NMR (500 MHz, CDCl3): δ = 2.34 (t, 2H, J = 7.5 Hz,
3
3
BCH2CH2CH2CH2CH2C), 1.63 (p, 2H, J = 7.5 Hz, BCH2CH2CH2CH2CH2C), 1.43 (p, 2H, J
= 7.6 Hz, BCH2CH2CH2CH2CH2C), 1.39 - 1.29 (m, 2H, BCH2CH2CH2CH2CH2C), 1.24 (s, 12H,
(C(CH3)2), 0.78 (t, 2H, 3J = 7.7 Hz, BCH2CH2CH2CH2CH2C). 13C-NMR (126 MHz, CDCl3): δ =
179.4, 83.1, 33.9, 31.8, 24.9, 24.6, 23.7.
3-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)propanoic acid (5e). An oven dried flask was
charged with CuCl (0.03 equiv, 0.10 mmol), NaOtBu (0.09 equiv, 0.30 mmol) and bis[(2-
diphenylphosphino)phenyl] ether (DPEPhos, 0.03 equiv, 0.10 mmol), evacuated and flushed
with argon (3×). Dry THF (5 mL) was added and the suspension was stirred for 0.5 h at rt. Then
bis(pinacolato)diboron (1.05 equiv, 3.62 mmol) in dry THF (4 mL) was added dropwise under
ACS Paragon Plus Environment
26

Page 27 of 49
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
argon and the suspension was stirred for another 0.5 h at rt. Finally, tert-butyl acrylate (1.0
equiv, 3.44 mmol) and MeOH (0.5 mL) were added and the suspension was stirred for 3 h at rt.
After filtrating over celite and washing with EtOAc, the solvent was evaporated under reduced
pressure and the crude product purified using silica column chromatography (EtOAc/petroleum
spirit, 10:90). The compound was obtained as a colorless oil (705 mg, 3.37 mmol, 98%).50 For
the tBu-deprotecting the ester (1 equiv, 3.37 mmol) was stirred in a mixture of TFA/DCM (1:1,
v/v, 4 mL) for 2 h at rt and subsequently concentrated to give the final compound as a colorless
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
oil (593 mg, 3.19 mmol, 95%). C9H17BO4; 1H-NMR (500 MHz, CDCl3): δ = 2.51 (t, 2H, J =
3
13
7.4 Hz, BCH2CH2C), 1.24 (s, 12H, (C(CH3)2), 1.03 (t, 2H, J = 7.4 Hz, BCH2CH2C). C-NMR
(126 MHz, CDCl3): δ = 180.1, 83.5, 28.6, 24.8.
4-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)butanoic acid (5f). A flask was charged with
tert-butanol (1.5 equiv, 15.6 mmol), triethylamine (1 equiv, 10.4 mmol) and pentan (20 mL) and
cooled to 0 °C. trans-Crotonyl chloride (1 equiv, 10.4 mmol) was added dropwise within 10 min
giving a yellow suspension with brown precipitate. The cooling was removed and the suspension
stirred for 3 h at rt. Subsequently, saturated aq. NaHCO3 (10 mL) followed by water (25 mL)
was added and the mixture extracted with Et2O/pentane (10:90, 3×20 mL). The combined
organic layers were dried with MgSO4, concentrated carefully and the crude purified using silica
column chromatography (pentane/EtOAc, 98:2) giving tert-butyl but-3-enoate as a colorless oil
(860 mg, 6.05 mmol, 58%).51 For the borylation, an oven dried flask equipped with 1,2-
bis(diphenylphosphino)ethane (0.03 equiv, 50 µmol) and bis(1,5-cyclooctadiene)diiridium(I)
dichloride (0.008 equiv, 13 µmol) was evacuated and flushed with argon (3×). After the addition
of dry DCM (3 mL) the suspension was cooled to 0 °C and tert-butyl but-3-enoate (1.0 equiv,
1.65 mmol) was added. Then pinacolborane (1.5 equiv, 2.48 mmol) was added within 5 min and
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 49
1
2
3
4
5
6
7
8
9
the suspension was stirred over night at rt. Upon the addition of water (5 mL) gas evolution was
visible and the layers were extracted with EtOAc (3×30 mL). The combined organic layers were
dried with MgSO4, concentrated and purified using silica column chromatography (petroleum
spirit/EtOAc, 95:5) giving tert-butyl 4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)butanoate as
a colorless oil (338 mg, 1.25 mmol, 76%).52 For the tBu-deprotecting the latter (1 equiv,
1.25 mmol) was stirred in a mixture of TFA/DCM (4 mL, 1:1, v/v) for 3 h at rt and subsequently
concentrated to give the final compound as a colorless oil (248 mg, 1.16 mmol, 93%).
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
C10H19BO4; H-NMR (500 MHz, CDCl3): δ = 2.40 (t, 2H, 3J = 7.5 Hz, BCH2CH2CH2C), 1.76 (p,
3
3
2H, J = 7.6 Hz, BCH2CH2CH2C), 1.26 (s, 12H, (C(CH3)2), 0.86 (t, 2H, J = 7.8 Hz,
BCH2CH2CH2C). 13C-NMR (126 MHz, CDCl3): δ = 180.4, 83.6, 36.1, 24.8, 19.3.
General procedure for the side chain-to-tail cyclization in solution. A glass tube was charged
with the linear peptide precursor on-resin (1 equiv), K3PO4 (5 equiv), GdnHCl (1 mM) and
solved in water/dioxane (3:1, v/v) with a final peptide concentration of 2 mM. sSPhos (15 mol%)
and Na2PdCl4 (5 mol%) were preactivated at 40 °C for 10 min, then added to the tube and the
sealed tube was heated in a microwave at 100 °C for 0.5 h. The crude was purified by preparative
RP-HPLC (method 1) and lyophilized to give a white solid. The cyclization in solution was
conducted on scales between 1-6 µmol.
Peptide Formula
m/zcalc
m/zfound
Yielda [%]
6
C44H52N10O10 880.4100 880.4100 10
C39H44N10O8 781.3416 781.3416 24
7
8
C30H35N9O7
C39H44N10O8 781.3416 781.3416 10
C30H35N9O7 634.2732 634.2732 3
634.2732 634.2732 22
9
10
ACS Paragon Plus Environment
28

Page 29 of 49
Journal of Medicinal Chemistry
1
2
3
4
aoverall yield after peptide synthesis and SMC.
5
6
7
8
9
General procedure for the side chain-to-tail cyclization on-resin. A Schlenk tube was
charged with the linear peptide precursor of 11 - 24 (1 equiv), Pd2(dba)3 (20 mol%), sSPhos
(40 mol%) and KF (4 equiv), evacuated and flushed with argon (3×). A degassed solution of
DME/EtOH/H2O (9:9:1, v/v/v, 2-3 mL) was added under argon and the mixture was heated using
a microwave at 120 °C for 0.5 h. Subsequently, the resin was washed with DMF (6×), EtOH (6×)
and Et2O (6×). The cyclic peptide was cleaved from the resin with a mixture of TFA/H2O/TIPS
(95:2.5:2.5 v/v/v) for 2×3 h each. The solvent was evaporated under reduced pressure, the
residue dissolved in CH3CN/H2O and lyophilized. The crude peptide was purified using
preparative RP-HPLC (method 1, 4). The on-resin cyclization was conducted in scales between
20-150 µmol.42
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
Peptide Formula
m/zcalc
m/zfound
Yielda [%]
11
12
13
14
15
16
17
18
19
20
21
22
C39H44N10O8 781.3416 781.3416 22
C30H35N9O7 634.2732 634.2736 11
C39H44N10O8 781.3416 781.3415 8
C40H46N10O8 795.3572 795.3577 36
C31H37N9O7 648.2888 648.2876 18
C40H46N10O8 795.3572 795.3577 77
C31H37N9O7 648.2888 648.2912 84
C40H46N10O8 795.3572 795.3555 26
C40H46N10O8 795.3572 795.3587 41
C40H46N10O8 795.3572 795.3568 53
C39H52N10O8 789.4042 789.4063 38
C38H50N10O8 775.3885 775.3900 21
ACS Paragon Plus Environment
29