Dynamic Chirality in the Mechanism of Action of Allosteric CD36 Modulators of Macrophage-Driven Inflammation

Article abstract of DOI:10.1021/acs.jmedchem.9b00918

Dynamic chirality influences numerous processes in nature from protein folding to catalysis. Azapeptides are peptidomimetics possessing semicarbazide residues that can interconvert between sp² and sp³ hybridization, resulting in ster

Full text of DOI:10.1021/acs.jmedchem.9b00918

Subscriber access provided by University of Winnipeg Library
Article
Dynamic Chirality in the Mechanism of Action of Allosteric
CD36 Modulators of Macrophage-Driven Inflammation
Emma Danelius, Ragnhild Gaard Ohm, Ahsanullah Ahsanullah, Mukandila
Mulumba, Huy Ong, Sylvain Chemtob, Máté Erdélyi, and William D. Lubell
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00918 • Publication Date (Web): 27 Nov 2019
Downloaded from pubs.acs.org on November 28, 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 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Dynamic Chirality in the Mechanism of Action of
Allosteric CD36 Modulators of Macrophage-Driven
Inflammation
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
║
,†
‡
‡,£
§
§
Emma Danelius, Ragnhild G. Ohm, Ahsanullah, Mukandila Mulumba, Huy Ong,
║
,
‡,*
Sylvain Chemtob, Mate Erdelyi ° and William D. Lubell
‡
◊
§
Département de Chimie, Département de Pédiatrie, et Faculté de Pharmacie,
Universite de Montreal, CP 6128, Station Centre-ville, Montreal, Quebec H3C 3J7,
Canada
§Faculty of Pharmacy, University of Montreal, Montréal, Québec, H3T 1J4 Canada
║Department of Pediatrics, Centre Hospitalier Universitaire (CHU) Sainte ‐ Justine
Research Center, Montréal, Québec, H3T 1C5 Canada
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Department of Chemistry – BMC, Uppsala University, Husargatan 3, SE-752 37
Uppsala, Sweden
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
°
The Swedish NMR Centre, Medicinaregatan 5, SE-413 90 Gothenburg, Sweden.
£Current address: Department of Chemistry, Quaid-i-Azam University, Islamabad
45320, Pakistan
Dynamic chirality, CD36, macrocycles, NMR solution ensemble, bioactive conformation.
ABSTRACT Dynamic chirality influences numerous processes in nature from protein
folding to catalysis. Azapeptides are peptidomimetics possessing semicarbazide
2
3
residues that can interconvert between sp and sp hybridization, resulting in stereo-
dynamic interconversions of pseudo-R and -S-configurations by way of a planar
intermediate. Cyclic azapeptides have shown unprecedented binding affinity to the
Cluster of Differentiation 36 receptor (CD36) and ability to mitigate macrophage-driven
inflammation by modulation of the Toll-like receptor 2/6 pathway. A novel approach to
3
synthesize cyclic peptides via A -macrocyclization has been used to make R- and S-
ACS Paragon Plus Environment
2

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
configuration controls to study the relevance of semicarbazide hybridization for
modulator activity. NMR spectroscopic analysis of potent cyclic azapeptide CD36
modulators (e.g., 1 and 2) and related cyclic peptides demonstrated that binding affinity
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
correlated with conformational rigidity, and a hybridization preference for sp > S- > R-
3
sp semicarbazide nitrogen configuration. Evidence of the active conformation and the
relevance for dynamic chirality serve as insight for creating cyclic (aza)peptide CD36
modulators to curb inflammation.
INTRODUCTION
The world is chiral from anatomical arrangements to the configurations of the
building blocks of life. Dynamic chirality occurs when a surmountable barrier exists for
interconversion between configurational isomers. Stereochemical-dynamics is common
in natural phenomena. For example, light induced olefin isomerization of retinal is
1
essential for vision. Prolyl isomerization is required for peptide folding, and protein
regulation, motion and function.2–5 Stereochemical-dynamics play similarly important
6
7
8–10
roles in catalysis, medicine, and molecular machine design.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Azapeptides employ a semicarbazide as an amino amide surrogate in a peptide.
Replacement of the amino amide -carbon by nitrogen in the azapeptide introduces
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
potential for stereochemical-dynamics through interconversion between sp and sp
hybridizations giving rise to planar, and pseudo-R and -S configurations, respectively.
Semicarbazide introduction has been shown to restrict the overall peptide conformation
by favoring turn conformation.11,12 The dynamic chirality of azapeptides has not been
13
well studied, but may enhance molecular recognition, especially in cases in which the
receptor itself undergoes conformational dynamics.
The Cluster of Differentiation 36 (CD36) receptor is a class B scavenger receptor
found in the membrane of several cell types.14–16 A multifunctional membrane
glycoprotein, CD36 is a signaling hub for major metabolic functions and immunological
responses, and binds various endogenous and exogenous ligands including damage-
associated molecular patterns, oxidized LDL, oxidized phospholipids, and fibrillary A
amyloid.14 Composed of a single 472-amino acid chain which is bridged by three
disulfide bonds and heavily post-translationally modified by glycosylation and
phosphorylation, CD36 features a large extracellular loop that binds diverse ligands,
ACS Paragon Plus Environment
4

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and two transmembrane spanning regions at the N- and C-termini with short cytosolic
domains that are anchored into the membrane by four palmitoylation sites.15 As co-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
receptor of the Toll-like receptor (TLR)-2/6 heterodimer assembly, CD36 has been
17
shown to sustain TLR2/6 response elicited by diacylglycerols, and to promote TLR2-
dependent macrophage driven inflammation, involving downstream activation of NF-

B.18 Targeting CD36 to modulate immune response mediated by activation of the
19
CD36-TLR2/6 complex, we have reported azacyclopeptides exhibiting unprecedented
CD36 binding affinity (e.g., 1, IC50 0.08M) that cause downstream suppression of
release of nitric oxide and pro-inflammatory cytokines and chemokines in TLR2-agonist
activated macrophages.20
NH
NH
H
H
N
NH
N
NH
O
Me
O
N
O
H
O
Me
Y
N
O
O
H
N
O
N
O
H
H
NH
O
N
O
H
O
NH2
NH2
N
NH2
Me
N
N
H
O
Y
NHR
1
1
, Y = N, R = Ala
0, Y = N, R = H
2, Y = N
R-9, Y = (R)-CH
S-9, Y = (S)-CH
11, Y = N, R = Ala-Ac
R-8, Y = (R)-CH, R = Ala
S-8, Y = (S)-CH, R = Ala
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 1. Structures of azacyclopeptides 1, 2, 10 and 11, and cyclic peptides R- and S-8
and 9.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The geometrical requirements for binding to various endogenous ligands demands
that CD36 exists in a dynamic conformational equilibrium. Dynamic chirality may thus
be involved in the remarkable CD36 binding affinity and activity of azacyclopeptides,
such as 1 and 2 (Figure 1). To study their conformation and dynamics, a three-pronged
approach has now been pursued featuring synthesis of cyclic peptide counterparts
possessing respectively -carbons of R- and S-configuration in replace of the -
nitrogen, biological analysis in various assays and conformational analysis using a
combined NMR spectroscopic and computational technique. A new method for
3
synthesizing cyclic peptides was developed featuring A -macrocyclization. Biological
assessment of the cyclic peptides in head-to-head comparisons with their
azacyclopeptide counterparts has demonstrated the relevance of nitrogen hybridization
for activity in different assays. Moreover, conformational analysis has revealed that
specific conformers are likely responsible for CD36 binding and immunomodulatory
ACS Paragon Plus Environment
6

Page 7 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
activity. The relevance of dynamic chirality for modulator activity has been addressed by
the collective data, which supports a rigid biologically active conformation offering vital
information for the design of improved therapeutics to curb macrophage-driven
inflammation.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Scheme 1. Solid-phase synthesis of cyclic peptides.
R
o-NBS
R
N
N
peptide elongation
H
N
H
Fmoc
*
N
N
H
Boc-Ala N
Xaa
N
H
H
O
O
O
HOCH2CH=CH2,
DIAD, PPh3, THF
4: R = H
3: R = CH2CH=CH2
HSCH2CH2OH,
DBU, DMF
5: R = o-NBS
6: R = H
a: = R, Xaa = D-Trp(Boc)-Ala-Trp(Boc)-D-Phe
*
b: = S, Xaa = D-Trp(Boc)-Ala-Trp(Boc)-D-Phe
*
c: = R, Xaa = Ala-Trp(Boc)-D-Phe
*
d: = S, Xaa = Ala-Trp(Boc)-D-Phe
*
CH2O,
CuI,
N
N
(95:2.5:2.5)
DMSO
TFA:TES:H2O
H
Boc-Ala N
H
*
N
H-Ala
N
*
NH2
H
Yaa-Trp-D-Phe N
Xaa
a-d
N
H
H
O
O
O
O
7
R-8: * = R, Yaa = D-Trp-Ala
S-8: * = S, Yaa = D-Trp-Ala
R-9: * = R, Yaa = Ala
S-9: * = S, Yaa = Ala
Cyclic peptides were synthesized by a novel route employing respectively R- and S-
3
propargylglycine (Pra) in an A -macrocycliztion protocol (Scheme 1), in which aldehyde,
alkyne and amine components reacted together in a copper-catalyzed reaction to form
the macrocycle.20
N -Allyl-N -(o-NBS)-N -(Fmoc)-lysine Rink-amide resin 3 was



ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
prepared using the Mitsunobu reaction on o-nitrobenzenesulfonamide 4. The linear
peptide was assembled using standard Fmoc-based solid phase chemistry conditions
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
employing DIC and HOBt for amino acid couplings. The final residue was introduced
as an N-Boc derivative. The o-NBS group was removed from linear azapeptide 5 using
3
mercaptoethanol and DBU in DMF to give peptide 6. A -macrocyclization was
performed by treating resin 6 and aqueous formaldehyde with CuI in DMSO to yield
macrocycles 7.20 Resin cleavage using a TFA/H O/TES cocktail and purification by
2
reversed phase HPLC gave cyclic peptides 8 and 9.
Azacyclopeptides 1, 2, 10 and 11 as well as their cyclic peptide counterparts 8
and 9 were examined for their binding affinity to CD36. The cyclic (aza)peptides were
also tested for ability to modulate nitric oxide (NO) overproduction elicited by the
specific TLR-2 agonist fibroblast-stimulating lipopeptide (R-FSL-1) and to promote
cholesterol efflux in macrophage cells. In a competitive assay on purified membranes
from rat heart as a source of CD36, the binding affinity of cyclic (aza)peptides was
125
measured in competition against the photo-activated agonist [ I]-Tyr-Bpa-Ala-
hexarelin. As previously described, azacyclopeptide 1 exhibited the highest CD36
ACS Paragon Plus Environment
8

Page 9 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
binding affinity of any known analog. Moreover, the N-acetyl derivative of 1 and its
smaller ring sized analog in which the D-Trp residue was removed, azacyclopeptides 11
and 2, both exhibited sub-micromolar binding affinity for CD36. Cyclic peptide 8 and 9
exhibited weaker binding affinities than their aza-counterparts with the S-analog
exhibiting relatively enhanced affinity compared to its R-diastereomer. For example,
relative to azacyclopeptide 1, R- and S-8 exhibited respectively 13- and 8-fold
reductions in binding affinity. Compared to the aza-analog 2, R- and S-9 exhibited
respectively 14- and 6-fold decreases in CD36 binding affinity. Notably, azacyclopeptide
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
10, which possesses no alanine residue at its N-terminal, exhibited no binding affinity
for CD36.
The ability of the cyclic analogs to inhibit NO over-production induced by the
TLR-2 agonist R-FSL-1 was performed in RAW macrophage cell lines. In macrophages,
NO production is mediated by pro-inflammatory stimuli, and used to fight against
invading pathogens, but may damage host tissue by reacting with oxygen and
superoxide anions to produce reactive nitrogen species.22 Effectively measured as
nitrite, NO is a distinct marker of activation of inflammation. In earlier studies of cyclic
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
0
azapeptides, aside from 10, ability to reduce nitrite production correlated with binding
affinity. Moreover, in the cases of 1, 2 and 11, but not 10, NO reduction was linked to
ability to decrease production of pro-inflammatory cytokines and chemokines in
macrophages. In the case of 10, which does not bind to CD36, NO reduction may likely
be due to an off-target pathway. Examination of the NO reduction that was caused by
cyclic peptides 8 and 9 demonstrated their weaker activity relative to their aza-
counterparts 1 and 2. Moreover, the S-analogs gave higher modulatory effects than the
R-counterparts in correlation with their relative CD36 binding affinities.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Modulators of CD36 have been shown to promote cholesterol efflux from
macrophages by inducing an increase of intracellular 15-deoxy-(12,14)-prostaglandin
J2 resulting in the activation of the peroxisome proliferator-activated receptor--liver X
receptor--ATP-binding cassette (ABC) transporter pathway.23 Considering a potential
protective effect against atherosclerosis, the capacity of macrocycles 1, 2, and 8–11 to
promote cholesterol efflux was measured in J774 murine macrophages using a
fluorescent-labeled cholesterol kit.24 Briefly, the J774 cells were first loaded with
fluorescent-labeled cholesterol for 24 h and then washed and treated with or without
ACS Paragon Plus Environment
10

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
modulator at 100 nM. After subsequent incubation for 17 h, cholesterol efflux was
induced by adding apolipoprotein A-I as cholesterol acceptor for 4 h. Cell-culture
medium containing extracellular cholesterol was then collected. The intracellular
cholesterol fraction was extracted from lysed cells. Percentage of efflux was calculated
as the ratio of fluorescence measured in the cell-culture medium alone / (fluorescence
of the cell lysates and cell-culture medium combined) x 100 (Table 1).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In the case of azacyclopeptide 1 and cyclic peptide counterparts R- and S-8, the
former enhanced more effectively cholesterol efflux in macrophages. The S-
diastereomer of 8 exhibits a slight improvement relative to the R-isomer. On the other
hand, in the case of macrocycles 2 and R- and S-9, S-diastereomer S-9 exhibited
greater activity in enhancing the efflux of cholesterol relative to azacyclopeptide 2 and
the R-cyclic peptide R-9, which exhibited similar activity.
Table 1. Biological data.
Entry Peptid NO
Affinity Cholester
ol efflux
(%)
e
inhibition (µM)
(%)
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
1
45
18
20
10
17
29
9
0.08
11
-
10
NDa
0.67
1.04
0.66
0.24
3.32
1.47
11
17
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R-8
S-8
7
2
R-9
6
6
S-9
16
13
aNot detected.
Macrocycles 1, 2 and 8–11 possess similar core structures, yet exhibited notably
different binding affinities and modulator potencies. For example, the binding affinities of
1, 10 and 11 varied significantly due to the subtle differences of their aza-residue N-
terminal substituent. The variance of a potentially dynamic -nitrogen versus a chiral -
carbon influenced considerably the activity of macrocycles 2 and 9. Peptide bioactivity is
generally dependent on conformation and the preferred conformer of the cyclic
(aza)peptide in solution may reflect that which initially binds the receptor.25 Restraint
that pre-organizes a favorable binding conformer may facilitate molecular recognition by
diminishing the entropy costs for folding. Based on the presumption that the bioactive
ACS Paragon Plus Environment
12

Page 13 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conformation of a flexible ligand is represented to a measurable extent in the ensemble
of conformers observed in solution,26–28 the NMR Analysis of Molecular Flexibility in
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Solution (NAMFIS) algorithm was used to identify the conformational families adopted in
solution by macrocycles 1, 2, S-9, 10 and 11.29
An X-ray structure has been recently reported for the complex of the human
CD36 ectodomain (residues 35–439) and the cysteine-rich interdomain region- domain
of the Plasmodium adhesive protein PfEMP1;30 however, no structures of CD36 alone
nor in complex with small molecules have yet been reported. The site tagged by the
125
photoactivable agonist [ I]-Tyr-Bpa-Ala-hexarelin that was used in the CD36 binding
affinity studies described above overlaps with that for oxidized low-density lipoprotein
155
183
and corresponds to residues Gln –Lys
on an -helical ‘platform’ at the membrane
distal tip of the protein.31,32 Considering the protein-bound ligand conformation is often
difficult, or even impossible to obtain, the conformational ensembles of the macrocycles
were determined in solution based on the logic iterated above.
NAMFIS is a combined NMR spectroscopic and computational technique that
deconvolutes time averaged spectroscopic data into the data of the solution
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conformational ensemble, providing the molar fraction of the conformers. The NAMFIS
protocol has been previously utilized in successful investigations of peptides,33–35
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
macrocycles27,36 and small molecules.37,38 Macrocycles are often flexible and exist as
ensembles of rapidly equilibrating low-energy conformations in solution. Considering
that a single structure does not represent the overall molecular geometry, following
previous procedures,34,35,39–42 NAMFIS was employed to give probability weighted
conformations based on experimentally determined NOE distances. Aqueous solutions
of cyclic azapeptides 1, 2, 10 and 11 and cyclic peptide S-9 were measured on a 900
MHz NMR spectrometer and assigned using a NOESY-TOCSY backbone walk
supported by HSQC, HMBC and COSY spectra. Inter-proton distances were derived
from NOE build-up rates according to the initial rate approximation;43,44 NOESY spectra
were recorded at 100, 200, 300, 400, 500, 600 and 700 ms without solvent suppression,
and distances were calculated from the initial linear part of the build-ups (Supporting
Information). The theoretical ensembles used as input for NAMFIS were obtained from
Monte Carlo Molecular Mechanics (MCMM) calculations (Supporting Information), using
the OPLS and the Amber force fields. Ensembles covering the entire available
ACS Paragon Plus Environment
14

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conformational space of the cyclic (aza)peptides were calculated, in agreement with the
literature,34,35,39–42 using an energy window of 42 kJ mol and 50000 Monte Carlo steps.
The window was made twice that of the energy available for molecules at room
temperature (~21 kJ/mol) to ensure all possible conformations were examined in the
-1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
conformational search. The generated MCMM-ensembles fulfilled the equation 1−(1−
M
(
1/N)) (N = total number of conformers, M = number of search steps) indicating that the
available conformational space was fully covered and hence any existing conformation
was not missed in the input ensemble.37
The solution ensembles were established by fitting the measured distances from
the NMR spectra to the computationally calculated conformations using established
protocols.29,34,35,45,46 The considerable line broadening of the proton signals in the
segment extending from the Lys -hydrogen to the D-Trp amine proton indicated that
this part of the macrocycle is dynamic with an exchange rate comparable to the NMR
time scale. Comparing the line broadening of macrocycles 1, 10 and 11,
azacyclopeptide 1 was found to be the least flexible. The larger number of NOE-
correlations observed in the NOESY spectrum of 1 (34 distances) as compared to those
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of 10 and 11 (23 and 24 distances respectively) agreed with the lower flexibility of the
former. The ensemble of conformers for 1 was identified by the NAMFIS algorithm to
have six plausible conformers, as compared to nine in the ensemble of 10 (Supporting
Information and Figure 2). The ensemble of peptide 11 also possessed six major
conformer families, however, these were more disparate relative to those of 1, yet less
diverse than those of 10, as indicated by the RMSD of the overlaid backbone atoms (1,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1.83; 10, 2.79; 11, 2.32). The flexibility order (10 > 11 > 1) was thus inversely
proportional to the observed CD36 binding affinity.
Figure 2. Overlaid backbone conformations (superposition of α-carbons of Phe-Trp-Ala-
Trp) of the solution ensembles of peptides 1, 10 and 11. Amino acid side chains are
omitted for clarity. RMSD overlaying backbone atoms: 1 = 1.83, 10 = 2.79, 11 = 2.32.
The conformations are given in the Table S10 in the Supporting Information and are
shown individually in Figures S2-S4.
ACS Paragon Plus Environment
16

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The conformations identified by NAMFIS were grouped into conformational families
according to the similarity of their backbone geometry (RMSD). Four groups of
backbone conformers dominated among 1, 10 and 11 (Figure 3). Macrocycle 1 is the
least flexible with only two major conformations: family A with molar fraction of 31%, and
family B with 66%. The most flexible macrocycle 10 adopted all four dominant
conformer families: 2% A, 41% B, 21% C and 2% D. Macrocycle 11 was found to
populate three of the dominant conformers: 22% A, 8% B, and 21% D. The complete
conformational ensembles with all molar fractions are given in Supporting Information
Table S10 (Supporting Information).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Major conformational groups A–D for macrocycles 1, 10 and 11. Amino acid
side chains are omitted for clarity. RMSD overlaying backbone atoms: A = 0.90, B =
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
.32, C = 0.52, D = 1.22. Pink/purple conformations represent peptide 1, green
conformations peptide 10, and blue conformations peptide 11.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The biologically active conformation may be approximated by comparing the
relative populations of the conformations of 1, 10 and 11 to their CD36 binding affinity
and activity. Conformers greatly populated by macrocycle 10 (e.g., B and C) are most
likely inactive, because this azacyclopeptide exhibits no CD36 binding affinity.
Moreover, conformational group D which is not populated by macrocycle 1 is likely
insignificant for binding and activity. In conformational group A, the population in the
ensembles of 1, 10 and 11 correlates with the observed CD36 binding affinity and
activity: 1 > 11 >> 10. The protein bound conformation of flexible ligands is generally
accepted to be represented to a measurable extent in the ensemble of conformers
observed in solution.26–28 Conformational group A, populated in 32% probability by 1, is
therefore expected to be responsible for the notable CD36 binding affinity and activity.
This conformation has a -II’ turn centered at the D-Trp-Ala residues, with a
characteristic intramolecular hydrogen bond between the carbonyl of the aza-amino-
ACS Paragon Plus Environment
18

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acid and the amide proton of Trp forming a pseudo-ten-membered ring (Figure 4). No
additional typical peptide secondary structure was observed within the conformational
families A−D of the studied peptides.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Overlaid solution conformers of 1, 10 and 11, belonging to conformational
family A. This conformational family possesses a β-II’-turn and represents the likely
bioactive conformer. The Trp side chains and hydrogens are not shown for clarity. The
hydrogen bond resulting in a pseudo-10-membered ring, typical for a type-II’-β-turn is
indicated with a red line. Peptide 1 is shown in purple, 10 in green, and 11 in blue.
The conformation of macrocycles 2 and S-9 were subsequently investigated to
gain further insight into the biologically active geometry of smaller macrocycles. We
have particularly given attention to the aza-nitrogen’s hybridization state in compound 2.
Based on NOE build-up analysis, 19 and 20 distance constraints were respectively
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
obtained for 2 and S-9. Both macrocycles adopt five conformational families (Figures S5
and S6, and Table S11 in the Supporting Information), of which two families, shown in
Figure 5, are similar. Overall, 2 and S-9 have similar flexibility as reflected by their
comparable overall backbone RMSD, 2.5 and 2.1, respectively [RMSD was calculated
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
in comparison to the first conformer, as 1/N×(RMSD ), in which N is the number of
N,1
conformers in an ensemble]. The T proton relaxation times indicated comparable
2
mobility of all methylene groups of 2, and suggested no increased mobility close to the
aza-nitrogen, due to a possible dynamic conformational exchange process.
The hybridization of the aza-residue of 2 was evaluated by comparison of the
goodness of the fit of the theoretical conformational ensembles having the
2
semicarbazide -nitrogen defined as an sp configured atom, and as an R- or S-
3
configured sp atom within the experimental data (NOE). The Akaike information
criterion statistical analysis was applied to estimate the accuracy of the three models in
describing the geometry, i.e., the hybridization and the configuration, of the aza-
residue.47 Applying this technique, the model best fitting to the experimental data has
the lowest criterion value (AIC), the evidence ratio (ER) of 1 (all other less likely models
i
ACS Paragon Plus Environment
20

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
have higher ER) and a logarithmic evidence ratio (LER) value of 0. The difference in
i
i
LER expresses the degree by which the best proposed model is better than all other
i
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
anticipated models. Accordingly, a LER difference of 0−0.5 is interpreted as ‘weak’,
i
44
0
.5−1 as ‘substantial’, 1−2 as ‘strong’ and >2 as ‘decisive’ evidence. Our analyses of 2
2
revealed its aza-region to possess an sp -hybridized nitrogen (Table 2). The S-
3
configured sp -hybridized model showed a significantly worse fit to the experimental
data (LER = 2.6), however, the Akaike analysis indicated this ensemble to be still a
i
significantly better description of the geometry of the aza region than the analogous R-
3
configured sp -hybridized model. This may indicate that the aza-geometry might be
closer to the S- than to the R-configuration (Table S14). A full description of the
statistical analysis is given in the Supporting Information.
Figure 5. The overlaid backbone of the similar conformers of 9 and 2, the structures
being superimposed at their -carbons. Here, turquoise represents a conformation of S-
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
9
whereas blue of 2, pink another conformation of S-9 whereas yellow and blue two
further conformers of 2.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 2. The results of the Akaike information criterion analyses of the fit of theoretical
ensembles with different hybridization and configuration at the N4/C4 atom to the
experimental data of 2.
sp2-
S-(sp )-
3
R-(sp )-
3
Statistical
parametera ensemble ensemble ensemble
AIC
17.0
18.1
0.0
28.9
35.6
11.9
391.8
2.6
49.7
AICc
ΔAICc
ERi
57.7
32.7
1.0
13109584.6
7.1
LERi
0.0
aAIC is the Akaike information criterion, AICc is the small sample size corrected AIC
that has a compensation for overfitting for systems with small sample size, ΔAICc is the
difference in AIC of an individual model as compared to the best model, ER is the
i
evidence ratio describing the relative likelihood of a pair of models, representing the
evidence about fitted models as to which is better in an information criteria sense, and
LER is the logarithmic evidence ratio. LER = 0–0,5: minimal; 0,5–1: substantial; 1–2:
i
i
strong; >2 decisive evidence.
ACS Paragon Plus Environment
22

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Conformation 1 of compound 2 (See supporting Information for xyz coordinates and
further details) is populated in 57% molar fraction in solution and shows great
similarities to conformational family A of cyclic azapeptides 1 and 11 (conformations 1, 2
and 16 Table S10 and Figure 6), which were concluded to be likely responsible for the
CD36 binding affinity and activity of the latter macrocycles. Conformation 22 of 2 has
similarity to conformation 28 of S-9 (Table S11), as shown in Figure 5, with similar
orientation of all amino acid side chains for binding to CD36. This is notably the only
mutual conformer for the bioactive compounds 1, 2, S-9, 10 and 11 reported here, and
correlates with the population predicted to explain the experimental CD36 binding
affinity. The common structural features of these conformers may provide useful
information for future drug development (Figures 5 and 6).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Conformer 2 of 1 (violet) and conformer 22 of 2 (blue), the structures being
superimposed at their -carbons and shown from two different perspectives. This
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conformational family is populated for all compounds with CD36 binding affinity and
activity studied here.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CONCLUSIONS
3
Employing a novel A -macrocyclization strategy to synthesize cyclic peptides
related to potent azacyclopeptide CD36 modulators (e.g., 1 and 2), a set of
configurationally stable analogs were prepared to assess the importance of dynamic
chirality for binding affinity and activity. Examination of binding affinity to CD36
demonstrated that the cyclic azapeptides 1 and 2 were superior to their R- and S-
peptide counterparts 8 and 9. Moreover, the cyclic azapeptides were more active than
their R- and S-peptide counterparts in reducing pro-inflammatory nitric oxide release
from macrophages stimulated by the TLR-2 agonist R-FSL-1. Although a similar trend
was observed between 1 and R- and S-8 in their ability to promote cholesterol efflux in
macrophage cells, a higher cholesterol efflux was observed using cyclic peptide S-9
compared to cyclic azapeptide 2 and peptide R-9.
ACS Paragon Plus Environment
24

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The features of the biologically active conformation of azacyclopeptide 1 were
identified by the combined computational and NMR spectroscopic analysis (NAMFIS) of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the cyclic azapeptides 1, 10 and 11. Line broadening (i.e., T -relaxation) indicated the
2
flexibility order 10 > 11 > 1 that was inversely proportional to their CD36 binding affinity.
The least flexible macrocycle 1 was identified to have two major conformational families
in a 7:3 molar ratio in solution, whereas 11 populated three major conformer families (A,
B and D) in a 22:8:21 ratio and the most flexible cyclic azapeptide 10 adopted four
conformer families (2:41:21:2, A:B:C:D). Correlation of the conformational populations
of these cyclic azapeptides with their binding affinities suggested that conformational
family A (Figure 3) is likely responsible for CD36 binding affinity of this substance group.
This conformational family has a -II-type turn with a characteristic pseudo-ten-
membered intramolecular hydrogen bonded ring. The smaller ring analogs 2 and S-9
also populate conformers, with 57% and 11% probability, that orient amino acid side
chains in similar positions for CD36 binding (Figure 6). Compound 2 adopted five
conformers in 16:11:6:40:27 ratio, with two of its plausible conformers having high
similarity to conformers of S-9. The latter compound had four plausible conformers in
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
solution, in a 57:11:21:11 molar ratio. The comparable overall backbone RMSD and the
similar T relaxation of the methylene groups suggest comparable overall flexibility of 2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and 9.
The hybridization and configurational preference of the aza-residue of
macrocycle 2 was evaluated using the Akaike information criterion (AIC) to analyze the
fit of the NMR data for conformational ensembles corresponding to theoretical models
2
3
with -nitrogen in sp , or S- or R-configured sp -geometry. Among the three models,
2
AIC analysis identified the sp configured conformational ensembles as the best
3
description of the aza-residue geometry of macrocycle 2. The sp -hybridized ensembles
provided poor descriptions; however, the model with S-configuration fit significantly
better than the R-configured one, which is in excellent agreement with the observation
that S-9 bound to CD36, inhibited NO and enhanced cholesterol efflux, all better than R-
9. Macrocycle 2 showed higher NO inhibition (Table 1) and higher CD36 binding affinity
than S-9, which only differs in an N-to-CH substitution, but S-9 promoted cholesterol
efflux better than 2 (Table 1). Although the semicarbazide nitrogen may adopt pseudo-
3
3
sp configuration, the degree of hybridization is significantly smaller than that of a sp
ACS Paragon Plus Environment
26

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
carbon.12,13 In contrast to the carbon analogs, the sp aza-nitrogen may in principle
3
engage in a hydrogen-bond, which may alter conformation and binding affinity. The
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
activities of 2 and S-9 indicate the subtle differences between sp and sp hybridizations
may influence modulatory activity on CD36 in NO production and cholesterol efflux from
macrophages. Our results highlight the vast impact of hybridization and configuration of
a single atom in the bioactivity of complex molecules.
EXPERIMENTAL SECTION
General methods. Polystyrene Rink Amide resin (0.5 mmol/g) was purchased
from Advanced Chemtech and the resin loading reported by the manufacture was used
to calculate the yields of the final products. Commercially available Boc-Ala-OH, Fmoc-
D- and L-Pra-OH, Fmoc-D- and L-Trp(Boc)-OH, Fmoc-Ala-OH, Fmoc-D-Phe-OH, Fmoc-
Lys-OH, Fmoc-chloride, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), diisopropyl
azodicarboxylate (DIAD) and triphenylphosphine, all were purchased from ChemImpex
and used as received. Acetic anhydride, allyl alcohol, copper iodide,
diisopropylcarbodiimide (DIC), diisopropylethylamine (DIPEA), 37% aqueous
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
formaldehyde, 65% hydrazine hydrate solution, 80% propargyl bromide solution in
toluene and 2-mercaptoethanol were purchased from Sigma Aldrich and used as
received. 1-Hydroxybenzotriazole (HOBt) was purchased from Advanced Chemtech. O-
NBSCl was purchased from Acros Organics and used as purchased. All solvents were
obtained from VWR international and dried by passage through a solvent filtration
system (Glass-Contour, Irvine, CA). Aza-peptide 1, 2, 10 and 11 were synthesized as
previously described.20,44
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Fmoc-Lys(allyl, o-NBS) resin (3): Vacuum dried Fmoc-Lys(o-NBS)-resin 4 (0.44
mmol) was placed in a syringe fitted with a Teflon filter, suspended in THF (dry, 5
mL), and treated sequentially with solutions of allyl alcohol (206 L, 3.03 mmol) in THF
(
dry, 1 mL), PPh (397 mg, 1.51 mmol) in THF (dry, 1 mL) and DIAD (298 L, 1.51
3
mmol) in THF (dry, 1 mL). The mixture in the syringe was shaken for 3 h. The resin was
filtered and sequentially washed with DMF (3), MeOH (3) and CH Cl (3).
2
2
Examination by LCMS of a cleaved resin sample (5 mg) showed complete allylation:
LCMS [30–95% MeOH [0.1% formic acid (FA)] in water (0.1% FA) over 10 min] R =
t
+
+
8.65 min. ESI-MS m/z: calcd for C H N O S [M+H] 593.2, found 593.2.
30
33
4
7
ACS Paragon Plus Environment
28

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Fmoc-based peptide synthesis of (aza)peptide 1,2, 8−11: Fmoc-based peptide
synthesis was performed using standard conditions21 on an automated shaker using
polystyrene Rink amide resin. Couplings of amino acids (3 equiv.) were performed in
DMF using DIC (3 equiv.) and HOBt (3 equiv.) for 3–6 hours. Fmoc group removal was
performed by treating the resin with 20% piperidine in DMF for 30 min. The resin was
washed after each coupling and deprotection step sequentially with DMF (3), MeOH
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
3) and CH Cl (3).
2
2
Representative protocol for o-NBS-deprotection. O-NBS-protected resin-bound
heptapeptide Boc-Ala-D-Pra-D-Trp(Boc)-Ala-Trp(Boc)-D-Phe-Lys(allyl, o-NBS)-NH-Rink
amide resin 5a (300 mg, 0.1 mmol) in a syringe fitted with a Teflon filter was swollen
in DMF (6 mL) and treated with DBU (150 L, 1.00 mmol) and 2-mercaptoethanol (35
L, 0.50 mmol). The mixture in the syringe was shaken for 3 h. The resin was filtered
and sequentially washed with DMF (3), MeOH (3) and CH Cl (3). Examination of
2
2
the cleaved resin aliquot by LCMS (5 mg) showed complete o-NBS removal: LCMS [20–
8
0% MeOH (0.1% FA) in water (0.1% FA) over 10 min] R = 4.79 min. ESI-MS m/z:
t
+
+
calcd for C H N O [M+H] 942.5, found 942.5.
51
64 11
7
ACS Paragon Plus Environment
29