Solid-phase parallel synthesis of functionalised medium-to-large cyclic peptidomimetics through three-component coupling driven by aziridine aldehyde dimers

Article abstract of DOI:10.1002/chem.201500068

The first solid-phase parallel synthesis of macrocyclic peptides using three-component coupling driven by aziridine aldehyde dimers is described. The method supports the synthesis of 9- to 18-membered aziridine-containing macrocycles, which are then functionalized by nucleophilic opening of the aziridine ring. This constitutes a robust approach for the rapid parallel synthesis of macrocyclic peptides. Solid-phase macrocycles: The first solid-phase parallel synthesis of macrocyclic peptides using three-component coupling driven by aziridine aldehyde dimers is described. The method supports the synthesis of 9- to 18-membered aziridine-containing macrocycles, which are then functionalised by nucleophilic opening of the aziridine ring (see scheme). This constitutes a robust approach for the rapid parallel synthesis of macrocyclic peptides.

Full text of DOI:10.1002/chem.201500068

DOI: 10.1002/chem.201500068
Full Paper
&
Macrocycles
Solid-Phase Parallel Synthesis of Functionalised Medium-to-Large
Cyclic Peptidomimetics through Three-Component Coupling
Driven by Aziridine Aldehyde Dimers
Adam P. Treder,[a] Jennifer L. Hickey,[b] Marie-Claude J. Tremblay,[a] Serge Zaretsky,[b]
Conor C. G. Scully,[b] John Mancuso,[a] Annie Doucet,[a] Andrei K. Yudin,[b] and Eric Marsault*[a]
Abstract: The first solid-phase parallel synthesis of macrocy-
clic peptides using three-component coupling driven by
aziridine aldehyde dimers is described. The method supports
the synthesis of 9- to 18-membered aziridine-containing
macrocycles, which are then functionalized by nucleophilic
opening of the aziridine ring. This constitutes a robust ap-
proach for the rapid parallel synthesis of macrocyclic
peptides.
Introduction
ence [1f,g]). Macrocycles have also been synthesised by using
multicomponent reactions (for representative examples, see
reference [4]).
Macrocycles are cyclic molecules composed of 12 or more
atoms. The synthesis of chemically and conformationally di-
verse macrocycles has been tackled by using several technolo-
gies, ranging from traditional combinatorial chemistry to meth-
ods allying chemistry and biology (for reviews, see refer-
ence [1]). Some of these approaches have become the plat-
forms of emerging companies (for specific examples, see refer-
ence [2]; for reviews, see reference [1f,g]). One challenge
inherent to the macrocyclisation reaction is the ability to bring
the two termini of the linear precursor in close proximity to
enable ring formation. Indeed, macrocyclisation is a unimolecu-
lar, largely entropy-driven reaction in which backbone flexibili-
ty and transannular interactions represent critical factors that
can favour or disfavour the reaction.[3] As a result, macrocyclisa-
tion is often performed under high or pseudo-high dilution
conditions to reduce the formation of undesired dimers and
higher oligomers. The availability of a broad diversity of natural
and unnatural amino acids has meant that the synthesis of
macrocyclic peptides has received particular attention, and
several methods have been optimised to synthesise these mol-
ecules incorporating variable degrees of nonpeptidic contents,
using macrolactamisation, ring-closing metathesis, SNAr or
[3+2] cycloaddition to name a few (for reviews, see refer-
In 2006, the Yudin group reported a novel class of ampho-
teric reagents in the form of aziridine aldehyde dimers,[5] and
the group later demonstrated the capability of such com-
pounds to disrupt the three-component Ugi reaction. When
aziridine aldehyde dimers reacted with an isocyanide and
a linear peptide possessing a free amino and carboxylate func-
tions, aziridine amide-containing peptide macrocycles were
isolated.[6] One of the distinguishing features of this macrocycli-
sation is the unusually high concentration at which it remains
feasible (typically higher than 0.1m), which constitutes a prepa-
rative advantage compared with other macrocyclisation reac-
tions, which are usually performed at millimolar concentra-
tions.[1g,h] The rationale behind this efficiency has been provi-
sionally attributed to the maintenance, through each inter-
mediate step, of a stabilising ion pair that facilitates chain fold-
ing and keeps the two ends in close proximity.[6d] The resulting
endocyclic aziridine amide can be opened post-macrocyclisa-
tion with a variety of nucleophiles, which adds an extra point
of diversity to introduce exocyclic substituents.[6a,f,7]
To further exploit the potential of this novel method, we
report herein the development of a solid-phase approach to
macrocyclisation with aziridine aldehydes. Given its optimisa-
tion for the synthesis of peptides, the potential of solid-phase
synthesis (SPS) has been amply demonstrated in both high-
throughput parallel synthesis and split-pool combinatorial
chemistry.[8] SPS constitutes a tool of choice to build libraries
that are suitable for hit identification and rapid analogue gen-
eration for hit-to-lead optimisation. Building on our recent ap-
plication of the disrupted Ugi reaction with aziridine aldehyde
dimers to the solid-phase synthesis of piperazinones,[9] we
report herein the development of a solid-phase macrocyclisa-
tion protocol and aziridine ring-opening strategy.
[a] Dr. A. P. Treder, M.-C. J. Tremblay, Dr. J. Mancuso, A. Doucet,
Prof. E. Marsault
Institut de Pharmacologie de Sherbrooke
UniversitØ de Sherbrooke, 3001, 12e av nord
Sherbrooke (QC) J1H 5N4 (Canada)
E-mail: eric.marsault@usherbrooke.ca
[b] Dr. J. L. Hickey, Dr. S. Zaretsky, Dr. C. C. G. Scully, Prof. A. K. Yudin
Davenport Research Laboratories
Department of Chemistry, University of Toronto
80 St. George Street, Toronto (ON) M5S 3H6 (Canada)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500068.
Chem. Eur. J. 2015, 21, 9249 – 9255
9249
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Results and Discussion
We sought a process in which every synthetic step was imple-
mented on solid support to maximise diversity generation with
a minimal number of operations. Key steps of our design
(Figure 1) included peptide elongation, pivotal three-compo-
Scheme 1. Strategies for side-chain attachment of the C-terminal amino
acid.
resin through its side chain carboxylate in the presence of
N,N’-diisopropyl carbodiimide (DIC) and 4-(N,N-dimethylamino)-
pyridine (DMAP) in dichloromethane, to deliver precursor 5.[10]
For the attachment of Ser and Thr, Merrifield resin was first
functionalised to generate Ellman resin bearing a tetrahydro-
pyranyl (THP) linker by reaction with 3,4-dihydro-2H-pyran-2-
methanol and sodium hydride.[11] Subsequent reaction with the
hydroxyl group of Fmoc-l-Ser-OAll or Fmoc-l-Thr-OAll in the
presence of camphorsulfonic acid in dichloromethane deliv-
ered precursors 6 and 7, respectively.[11] Anchoring of Lys
through its side chain started with reaction of Wang resin with
p-nitrophenyl chloroformate in the presence of pyridine in di-
chloromethane, to generate the corresponding carbonate. Sub-
sequent reaction with Fmoc-l-Lys-OAll in the presence of
Hünig’s base in N,N-dimethylformamide (DMF) delivered pre-
cursor 8.[12] Finally, anchoring of Gln was prepared by amide
bond formation between Fmoc-l-Glu-OAll and Rink amide
resin in the presence of DIC, diisopropylethylamine (DIPEA),
and DMAP in dichloromethane, to deliver 9. Generally, resin
loadings were determined after Fmoc cleavage and found to
be approximately 0.40 mmolgÀ1 for THP-linked resins, or simi-
lar to suppliers’ loadings (0.68 mmol of Fmoc-l-Glu-OAll/g,
0.66 mmol of Fmoc-l-Lys-OAll/g, 0.47 mmol of Fmoc-l-Gln-
OAll/g). Loadings were confirmed by cleavage in trifluoroacetic
acid (TFA)/dichloromethane (1:1). The purity of cleavage prod-
ucts was confirmed by LC-MS analysis and found to be gener-
ally excellent (>95%, UV monitoring).
Figure 1. General synthetic approach (exemplified with all-l-amino acids and
(S)-aziridine aldehyde dimer derived from l-Ser).
nent macrocyclisation, subsequent nucleophilic ring-opening
of the newly formed aziridine, and final resin cleavage with si-
multaneous deprotection of side chains. To address these re-
quirements, we decided to attach the precursor peptide to the
resin through the side chain of a suitably protected C-terminal
amino acid. Accordingly (Figure 1), we reasoned that a suitable
general strategy might involve side chain attachment of the
first amino acid, protected on the amino and the carboxylate
positions, to deliver precursor 1. Subsequent chain elongation,
ideally by using standard Fmoc chemistry, followed by N- and
C-terminal deprotections, would deliver linear precursor 2. The
latter would then undergo macrocyclisation in the presence of
aziridine aldehyde dimer and isocyanide to produce macrocy-
cle 3. Finally, nucleophilic opening of the newly formed acyl
aziridine 3 in situ followed by acid-mediated resin cleavage
and concomitant side chain deprotection(s), would deliver the
desired product 4, which could be purified by reverse-phase
HPLC. This synthetic approach offered a lot of flexibility to in-
troduce diversity by varying ring size and stereochemistry as
well as the substitution of amino aldehyde, nucleophile and
amino acids side chains. Structurally, macrocycles resulting
from this strategy are distinguished from macrocyclic peptides
or depsipeptides by the presence of a 1,2-disubstituted dia-
mine moiety bearing two controlled stereogenic centres (see
below). Here, we report the successful implementation of this
strategy and demonstrate its utility by synthesizing a represen-
tative set of 9- to 18-membered rings bearing diverse amino
acids and nucleophiles.
A typical synthesis is exemplified in Scheme 2. Starting from
Fmoc-l-Glu-OAll (5) attached to Wang resin (Scheme 1), the
Fmoc group was deprotected in the presence of 20% piperi-
dine/DMF, followed by HATU-mediated couplings to assemble
the precursor linear amino acid sequence. Deprotection of the
allyl ester in the presence of palladium tetrakis(triphenylphos-
phane) and phenylsilane in dichloromethane, followed by
Fmoc removal, delivered macrocyclisation precursor 10 anch-
Several strategies were selected to attach different amino
acids through their side chains to the solid support
(Scheme 1). Fmoc-l-Glu-OAll was attached directly to Wang
Chem. Eur. J. 2015, 21, 9249 – 9255
www.chemeurj.org
9250
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
four amide bonds, respectively, the replacement of one amino
acid by the substituted aminoethyl moiety originating from
the three-component reaction, is expected to reduce ring
strain through the removal of one amide bond. Second, mac-
rocyclisation proved tolerant of all side chain functionalities, in-
cluding nonpolar and polar side chains. Likewise, introduction
of non-alpha-amino acids in the chain was well tolerated (e.g.,
23–25). Third, macrocyclisation was tolerant of variations in
stereochemistry at every position, provided that the stereo-
chemistry of the Pro residue on the N-terminus of the chain
was matched to that of the aziridine aldehyde dimer (i.e., l-Pro
reacts cooperatively with (S)-aziridine aldehyde dimer derived
from l-Ser as in 14–26, 28–33, whereas d-Pro matches the (R)-
aziridine aldehyde dimer derived from d-Ser as in 13, 27, 34).
Mismatched stereochemistry between the Pro residue and the
aziridine aldehyde dimer typically gave intractable mixtures.
This observation was rationalised by Zaretsky et al., who dem-
onstrated in solution that a stereochemistry match between
the aziridine aldehyde dimer and amino acid partners was par-
amount to obtaining good diastereoselectivity in the disrupted
Ugi reaction with secondary amino acids.[16] For previous dis-
cussion on the importance of a terminal Pro residue, see refer-
ences [5,14,16]; for the use of alternative isocyanides, see ref-
erences [6f,16].
Scheme 2. Detailed synthetic scheme exemplified for macrocycle 13. Re-
agents and conditions: a) 20% piperidine/DMF, 30 min, rt (”2); b) Fmoc-AA-
OH, HATU, DIPEA, DMF, 4 h, rt; c) [Pd(PPh3)4], PhSiH3, CH2Cl2, rt, 3 h; d) 11,
tBu-NC, DCM:TFE (1:1), rt, 4 h; e) PhSH, DIPEA, DMF, rt, 16 h; f) TFA/CH2Cl2
(1:1), rt, 75 min.
ored on the resin through the Glu side chain.[13] Reactions
were followed by LC-MS and driven to completion by using
three equivalents of reagents. Typically, macrocyclisation pre-
cursors were more than 90% pure, as determined after mini-
cleavage (2–5 mg resin in TFA/CH2Cl2, 1:1) by LC-MS analysis
(UV monitoring).
The aziridine ring of intermediate 3 (Figure 1) was opened
on resin with a variety of nucleophiles, exemplified here with
different thiols (13, 18, 21, 22, 28, 29) and thiobenzoic acid
(14). It should be noted, however, that all attempts to isolate
aziridine-containing macrocycle 3 by acid-mediated resin-cleav-
age failed, which reflects the relative instability of the aziridine
moiety.[17]
The critical macrocyclisation step was implemented in the
form of a disrupted three-component Ugi reaction in the pres-
ence of aziridine aldehyde dimer 11 (synthesised as reported
in Refs. [14] and [9]) and tert-butyl isocyanide. Previous experi-
ence with the solution-phase macrocyclisation revealed that
the reaction performs best in TFE as a solvent. However, this
solvent did not give satisfactory resin swelling, which is critical
for reaction completion on solid support.[8b] A preliminary
screening of solvent conditions led us to choose CH2Cl2/TFE
(1:1) as the best compromise between swelling and reactivity.
Macrocyclisation was run in the presence of aziridine aldehyde
dimer (1 equiv with respect to measured loading), tert-butyl
isocyanide (2 equiv) in CH2Cl2/TFE (1:1) at ambient tempera-
ture, to deliver acyl aziridine-containing macrocycle 12
(Scheme 2). The acyl aziridine was ring-opened with thiophe-
nol in situ in the presence of DIPEA in DMF. Finally, the macro-
cycle was cleaved from the resin by using TFA/CH2Cl2 (1:1). The
crude material was collected by filtration and evaporation, and
macrocycle 13 was isolated by preparative HPLC. This method
was used to generate macrocycles 13–34 (Figure 2).
As summarised in Figure 2, starting from 0.1 mmol resin
(typically ca. 160 mg of a 0.6 mmolgÀ1 nominal loading resin),
macrocycles were isolated after synthesis, preparative HPLC
purification and lyophilisation, with isolated yields ranging
from 7 to 32% (18 and 23, respectively) for the 7-, 9-, 11- or
13-step syntheses of ring systems built on di-, tri-, tetra- and
pentapeptides, respectively.
Once the purity of the peptides was confirmed by LC-MS
1
analysis, the macrocycles were analysed by using 1D H and
1
1
13C NMR, as well as by 2D H-1H and H-13C NMR spectroscopic
techniques. Successful macrocyclisation was noted by the ap-
pearance of a new amide NH peak with 2D TOCSY cross-peaks
to the adjacent linker region (Figure 3). The structure of com-
pound 24 has the linker atoms both annotated and highlight-
ed in red. We observed the corresponding cross-peaks from
the new NH amide to the a-, b-, and g-protons, as shown in
the TOCSY spectrum. The nucleophilic aziridine ring-opening
was confirmed to proceed with the same regioselectivity as
observed in the solution-phase protocol, which was supported
by NMR evidence of a downfield diastereotopic methylene at
the b-position.
Several observations can be made regarding this sequence.
First, the reaction performed well for macrocycles of various
ring sizes, ranging from 9 atoms (14, 15 derived from dipepti-
des) to 18 atoms in the ring (30–34 derived from pentapepti-
des). Notably, the yields for 9- and 12-membered rings were
reasonable, considering that medium-sized homodetic pep-
tides are notoriously difficult to cyclise.[15] Compared with ho-
modetic cyclic tri- and tetrapeptides, which contain three or
The next step was to compare the properties of this new
family of macrocycles with those of reported drugs and clinical
candidates. In terms of physicochemical and biological proper-
ties, macrocycles are considered an intermediate class between
small organic molecules and large biomolecules.[1e] With molec-
Chem. Eur. J. 2015, 21, 9249 – 9255
www.chemeurj.org
9251
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. Macrocycles synthesised on solid phase by three-component macrocyclisation and subsequent aziridine ring-opening. Macrocycles were isolated as
TFA (13–29) or formate (30–34) salts. Yields of 13–29 were calculated considering one TFA equivalent per net positive charge on the macrocycle, and 31–35
using formate as a counter-anion.
ular weights usually ranging between 500 and 2,000 Da, they
combine large surface areas, usually associated with biomole-
cules such as antibodies, with the potential to reach suitable
physicochemical properties conducive to oral bioavailability,
usually associated with small molecules. These properties
mean that macrocycles have attracted an increasing level of at-
tention recently as an underexploited chemical space (for
recent reviews, see references [1f–h,18]). They have been used
as drug candidates on most target classes, with currently
around 70 macrocyclic drugs or clinical candidates issued
mostly from natural products and peptides.[18a]
Accordingly, we sought to compare the chemical space cov-
ered by representative macrocyclic drugs and clinical candi-
dates with the collection reported in this paper. As shown in
Table 1, the newly reported macrocycles cover a broad range
of molecular weights (449–860 gmolÀ1), which is within the
range of known macrocyclic peptidic and peptidomimetic
drug candidates (Ulimorelin, Linopristin, Simeprevir, Vaniprevir,
Chem. Eur. J. 2015, 21, 9249 – 9255
www.chemeurj.org
9252
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Octreotide, Cyclosporine).[18a,20] Likewise, their cLogP range
from very hydrophilic (26, cLogP À1.7, close to Octreotide,
cLogP À1.0) to reasonably lipophilic, taking as a reference Lip-
inski’s parameters (28, cLogP 3.1, similar to Ulimorelin, cLogP
2.8, which possesses a bioavailability of 25%).[21] Their Polar
2
Surface Areas (PSA) cover a broad range (15, 108 Š to 34,
230.4 Š2), which overlaps with values predictive of an accepta-
ble cellular permeability.[18a] These parameters are readily ad-
justed by varying ring size and substitution pattern. Finally, the
analysis of principal moments of inertia (PMI, Figure 4)[22] re-
Figure 3. 1H-1H TOCSY NMR of compound 24. The linker region is annotated
and highlighted in red on the structure. Corresponding cross-peaks from the
new NH to the adjacent linker atoms are highlighted on the TOCSY spec-
trum.
Figure 4. Distribution of Principal Moments of Inertia (PMI) of newly report-
ed macrocycles, separated by ring size. Macrocyclic drugs and development
candidates (purple) include cyclosporin, ulimorelin, linopristin, simeprevir, va-
niprevir and octreotide.
Table 1. Analysis of properties of newly reported macrocycles and com-
parison with macrocyclic drugs and development candidates.
Compound
Ring size
Mw [gmolÀ1
]
cLogP[a]
PSA [Š2][a]
veals that this class of macrocycles covers a shape ensemble
that overlaps with that of the aforementioned macrocyclic
drugs and development candidates, independently of ring size.
Overall, this class of macrocycles appears suitable to support
drug discovery. The first biologically active molecules in this
class were reported recently.[23]
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
12
9
9
637.8
518.6
448.6
617.8
636.9
620.8
560.8
519.7
672.8
608.8
560.8
602.8
630.9
731.9
751.0
778.1
728.8
787.1
821.1
837.1
782.0
860.1
1202.6
538.7
950.1
749.9
755.9
1019.2
0.3
À1.2
0.5
À1.0
2.0
0.0
0.4
À0.1
2.5
À0.1
0.5
0.9
1.8
À1.7
À0.3
3.1
1.3
2.0
2.0
1.4
0.9
1.2
0.9
2.8
1.9
5.4
4.9
À1.0
137.1
139.9
108.0
169.4
128.9
153.2
139.6
137.2
119.9
152.2
144.0
132.4
130.5
179.5
148.1
147.1
155.0
171.8
176.1
214.9
208.9
230.3
282.2
97.8
12
12
12
12
12
12
12
13
15
18
15
15
15
15
18
18
18
18
18
11
18
19
14
20
20
Conclusion
The methodology reported herein provides a versatile tool for
the solid-phase parallel synthesis of diversified libraries of mac-
rocyclic peptidomimetics. The reaction sequence is completely
implemented on solid-phase, which minimises transfers and
user-intensive operational steps. The method was used to gen-
erate a library of several hundred macrocycles, 22 of which are
reported herein. The synthesis was achieved by using 24-well
Mettler–Toledo parallel synthesis blocks.[24] Typically, the syn-
thesis of a subset of 48 macrocycles is achievable in two weeks
by a single chemist (excluding purification and lyophilisation).
The method is tolerant of a broad diversity in terms of ring
size (9- to 18-membered rings) as well as the nature and ste-
reochemistry of amino acids contained in the ring, providing
molecules that also possess a nonpeptidic exocyclic element
as an additional point of diversity or subsequent functionalisa-
tion. At a time when macrocycles are generating tremendous
29
30
31
32
33
34
cyclosporin
ulimorelin
linopristin
simeprivir
vaniprevir
octreotide
229.7
157.9
193.4
336.3
[a] Physicochemical properties were calculated by using QikProp.[19]
Chem. Eur. J. 2015, 21, 9249 – 9255
www.chemeurj.org
9253
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(4”8 mL, 4 min cycles). The resin was further washed with DMF
(4”8 mL, 4 min cycles), then DMF (2.5 mL) was added, followed by
PhSH (113 mL, 1.1 mmol, 11 equiv) and DIPEA (174 mL, 1.0 mmol,
10 equiv). The reaction mixture was agitated overnight then
drained and washed with DMF, CH2Cl2, MeOH, CH2Cl2, MeOH,
CH2Cl2 (3”8 mL each, 4 min cycles).
interest in the drug discovery community, we anticipate that
both the versatility of the method and the properties of the re-
sulting macrocycles will make this technology broadly
applicable.
NB: The (R)-aziridine aldehyde dimer derived from d-Ser was used
in the synthesis of macrocycles 13, 27 and 34, whereas the other
macrocycles were obtained by reaction with (S)-aziridine aldehyde
dimer derived from l-Ser.
Experimental Section
Representative experimental procedures: synthesis of
macrocycle 13
Resin cleavage: The final macrocycle was cleaved with TFA/CH2Cl2
(1:1, 6 mL) for 75 min. The resin was then washed (4”) with TFA/
CH2Cl2 (1:1). Combined filtrates were evaporated and purified by
preparative HPLC (13–29) or MS-triggered preparative HPLC (31–
34). Fractions containing the product were re-analysed by UPLC-
MS, pooled and the product was isolated by lyophilisation.
Preparation of Fmoc-l-Glu-OAll anchored to Wang resin through
the side chain: Wang resin (0.95 mmolgÀ1
, 9 g, 8.55 mmol,
1 equiv) was swollen in CH2Cl2 (150 mL) in a 250 mL glass solid-
phase peptide synthesis reactor equipped with a fritted glass
funnel. The resin was then further washed with CH2Cl2 (2”150 mL,
4 min cycle on an orbital shaker). Fmoc-l-Glu-OAll (8.75 g,
21.375 mmol, 2.5 equiv) and DMAP (157 mg, 1.28 mmol,
0.15 equiv) in CH2Cl2 (150 mL) was added to the resin, followed by
DIC (3.31 mL, 21.375 mmol, 2.5 equiv). The reaction mixture was
shaken overnight on an orbital shaker then drained and washed
with CH2Cl2, MeOH, CH2Cl2, MeOH, CH2Cl2 (3”150 mL each, 4 min
cycles). The resin was capped with acetic anhydride (3.2 mL,
34.2 mmol, 3 equiv) in the presence of pyridine (2.75 mL,
34.2 mmol, 3 equiv) in anhydrous CH2Cl2 (150 mL). Finally the resin
was drained, washed (CH2Cl2, MeOH, CH2Cl2, MeOH, CH2Cl2; 3”
150 mL each, 4 min cycles) and dried in vacuo. Loading was con-
firmed by mini-cleavage (TFA/CH2Cl2 (1:1), 1.5 h) and was close to
Computational modelling and PMI analysis
Structures were built in Maestro 9.9 and energy minimised by
using 10000 steps of the Polak–Ribier Conjugate Gradient
method.[25] A Monte-Carlo molecular mechanics conformational
search for each macrocycle was carried out by using the OPLS_
2005 forcefield and a GB/SA implicit water model as implemented
in Macromodel.[26] The electrostatic interaction cutoff distance was
set to 20 Š, the van der Waals interaction cutoff distance to 8 Š,
and the hydrogen-bonding interaction cutoff distance to 4 Š. Tor-
sional sampling of amide bonds was performed and a 5 kcalmolÀ1
energy window was employed to discard high-energy structures.
Mirror-image conformers were retained. Redundant conformers
were discarded if they were found to be within an RMSD of 5 Š of
existing conformers. All conformers were energy minimised by
using 10000 steps of the Polak–Ribier Conjugate Gradient method.
the
manufacturer’s
specifications
(0.68 mmol
of
Fmoc-l-Glu-OAll/g).
Fmoc deprotection: Resin (0.147 mg, 0.1 mmol) was transferred
into a polypropylene disposable reactor (10 mL). Piperidine/DMF
(20%, 8 mL) was added and the resin was shaken for 30 min. The
resin was then drained and the above operation was repeated. Fi-
nally, the resin was drained and washed with DMF (5”8 mL).
The lowest energy conformation for each macrocycle was selected
for PMI analysis. Principal axes of inertia (I1, I2, and I3) were calculat-
ed by using a built-in script in Maestro.
Attachment of Fmoc-d-Phe-OH followed by Fmoc-d-Pro-OH:
Fmoc-d-Phe-OH (116 mg, 0.3 mmol, 3 equiv) and HATU (97 mg,
0.255 mmol, 2.55 equiv) were dissolved in DMF (4 mL). DIPEA
(89 mL, 0.51 mmol, 5.1 equiv) was added and the mixture was
transferred to the resin. After 4 h agitation, the resin was drained
and washed with DMF (2”), IPA, DMF, CH2Cl2 (3”), and diethyl
ether (8 mL, 4 min cycles). The Fmoc group was deprotected by
using 20% piperidine/DMF as described above, then Fmoc-d-Pro-
OH was introduced similarly to Fmoc-d-Phe-OH.
Acknowledgements
This work was financially supported by the Consortium QuØbØ-
cois de Recherche sur le MØdicament (CQDM). Dr. Andrew
Roughton is gratefully acknowledged for helpful comments.
Allyl ester deprotection: Resin (0.1 mmol, 1 equiv) was washed
with anhydrous CH2Cl2 (8 mL). Phenylsilane (123 mL, 1 mmol,
10 equiv) in anhydrous CH2Cl2 (2 mL) was added, followed by
Pd(PPh3)4 (35 mg, 0.03 mmol, 0.3 equiv) in anhydrous CH2Cl2
(2 mL). The resin was agitated for 3 h then the dark-coloured solu-
tion was drained. The resin was washed with 0.5% sodium diethyl
dithiocarbamate/DMF (3”), isopropanol, DMF, CH2Cl2 (3”) and di-
ethyl ether (8 mL, 3–4 min cycles), then dried in vacuo.
Keywords: drug discovery · macrocycles · peptides · small ring
systems · solid-phase synthesis
[1] a) N. K. Bashiruddin, H. Suga, Curr. Opin. Chem. Biol. 2015, 24C, 131–
138; b) S. E. Northfield, C. K. Wang, C. I. Schroeder, T. Durek, M. W. Kan,
J. E. Swedberg, D. J. Craik, Eur. J. Med. Chem. 2014, 77, 248–257; c) J. R.
Frost, J. M. Smith, R. Fasan, Curr. Opin. Struct. Biol. 2013, 23, 571–580;
d) M. Gao, K. Cheng, H. Yin, Biopolymers 2015, DOI: 10.1002/bip.22625;
e) E. M. Driggers, S. P. Hale, J. Lee, N. K. Terrett, Nat. Rev. Drug Discovery
2008, 7, 608–624; f) E. Marsault, M. L. Peterson, J. Med. Chem. 2011, 54,
1961–2004; g) C. J. White, A. K. Yudin, Nat. Chem. 2011, 3, 509–524;
h) A. K. Yudin, Chem. Sci. 2015, 6, 30–49.
[2] a) E. A. Jefferson, S. Arakawa, L. B. Blyn, A. Miyaji, S. A. Osgood, R.
Ranken, L. M. Risen, E. E. Swayze, J. Med. Chem. 2002, 45, 3430–3439;
b) E. Marsault, H. R. Hoveyda, R. Gagnon, M. L. Peterson, M. Vezina, C.
Saint-Louis, A. Landry, J. F. Pinault, L. Ouellet, S. Beauchemin, S. Beau-
bien, A. Mathieu, K. Benakli, Z. Wang, M. Brassard, D. Lonergan, F. Bilo-
deau, M. Ramaseshan, N. Fortin, R. Lan, S. Li, F. Galaud, V. Plourde, M.
Champagne, A. Doucet, P. Bherer, M. Gauthier, G. Olsen, G. Villeneuve, S.
Bhat, L. Foucher, D. Fortin, X. Peng, S. Bernard, A. Drouin, R. Deziel, G.
Fmoc deprotection, macrocyclisation and nucleophilic ring-
opening of aziridine: The above resin (0.1 mmol) was swollen in
DMF (30 min), drained and then piperidine/DMF (20%, 8 mL) was
added. After 30 min agitation, the resin was drained and again
20% piperidine/DMF was added. The resin was agitated for
30 min, drained and washed with DMF (5”) then CH2Cl2 (5”8 mL,
4 min cycles) and finally with anhydrous CH2Cl2/TFE (1:1, 3”3 mL,
4 min cycles). Next, anhydrous CH2Cl2 (1 mL) was added, followed
by a 0.1m solution of (R)-aziridine aldehyde dimer 11 and tert-butyl
isocyanide (22.6 mL, 0.2 mmol, 2 equiv) in TFE (1 mL). The reaction
mixture was agitated for 4 h, then drained and washed with CH2Cl2
Chem. Eur. J. 2015, 21, 9249 – 9255
www.chemeurj.org
9254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Berthiaume, Y. L. Dory, G. L. Fraser, P. Deslongchamps, Bioorg. Med.
Chem. Lett. 2008, 18, 4731–4735; c) B. N. Tse, T. M. Snyder, Y. Shen, D. R.
Liu, J. Am. Chem. Soc. 2008, 130, 15611–15626.
[15] W. D. Meutermans, G. T. Bourne, S. W. Golding, D. A. Horton, M. R. Cam-
pitelli, D. Craik, M. Scanlon, M. L. Smythe, Org. Lett. 2003, 5, 2711–2714.
[16] S. Zaretsky, S. Adachi, B. Rotstein, J. L. Hickey, C. C. G. Scully, J. D. St De-
nis, R. Courtemanche, J. C. Y. Yu, B. K. W. Chung, A. K. Yudin, J. Org.
Chem. 2014, 79, 9948–9956.
[17] Free aziridine-containing macrocycles have been observed to be stable
to column chromatography, see for example references [6f,7] and, in
certain cases, stability was also observed at high temperatures (658C in
[D6]DMSO) during variable-temperature NMR studies (Jennifer L. Hickey,
Sai Chakka Kumar, Megan A. St.Denis, Monzur Morshed, Serge Zaretsky,
Conor C. G. Scully, Andrew L. Roughton, and Andrei K. Yudin; unpub-
lished results).
[3] a) A. R. Bogdan, S. V. Jerome, K. N. Houk, K. James, J. Am. Chem. Soc.
2012, 134, 2127–2138; b) J. E. Delorbe, J. H. Clements, B. B. Whiddon,
S. F. Martin, ACS Med. Chem. Lett. 2010, 1, 448–452; c) F. Enck, M. A.
Marahiel, A. Geyer, ChemBioChem 2008, 9, 2597–2601.
[4] a) D. G. Rivera, A. V. Vasco, R. Echemendia, O. Concepcion, C. S. Perez,
J. A. Gavin, L. A. Wessjohann, Chem. Eur. J. 2014, 20, 13150–13161; b) G.
Koopmanschap, E. Ruijter, R. V. Orru, Beilstein J. Org. Chem. 2014, 10,
544–598; c) T. F. Niu, M. Sun, M. F. Lv, W. B. Yi, C. Cai, Org. Biomol. Chem.
2013, 11, 7232–7238.
[5] R. Hili, A. K. Yudin, J. Am. Chem. Soc. 2006, 128, 14772–14773.
[6] a) B. K. W. H. Chung, J. L. Hickey, C. C. G. Scully, S. Zaretsky, A. K. Yudin,
Med. Chem. Commun. 2013, 4, 1124; b) R. Hili, V. Rai, A. K. Yudin, J. Am.
Chem. Soc. 2010, 132, 2889–2891; c) X. Li, A. K. Yudin, J. Am. Chem. Soc.
2007, 129, 14152–14153; d) C. C. G. Scully, V. Rai, G. Poda, S. Zaretsky,
D. C. Burns, R. S. Houliston, T. Lou, A. K. Yudin, Chem. Eur. J. 2012, 18,
15612–15617; e) A. K. Yudin, R. Hili, Chem. Eur. J. 2007, 13, 6538–6542;
f) S. Zaretsky, C. C. G. Scully, A. J. Lough, A. K. Yudin, Chem. Eur. J. 2013,
19, 17668–17672.
[7] S. Zaretsky, J. Tan, J. L. Hickey, A. K. Yudin Peptide Libraries in Methods in
Molecular Biology (Ed.: R. Derda), Springer Science Business Media, New
York, 2015, pp. 67–80.
[8] a) B. A. Bunin, J. A. Ellman, J. Am. Chem. Soc. 1992, 114, 10997; b) S. A.
Kates, F. Albericio, Solid Phase Synthesis a Practical Guide, CRC Press,
Boca Raton, 2000.
[18] a) F. Giordanetto, J. Kihlberg, J. Med. Chem. 2014, 57, 278–295; b) A. K.
Oyelere, Curr. Top. Med. Chem. 2010, 10, 1359–1360; c) E. A. Villar, D.
Beglov, S. Chennamadhavuni, J. A. Porco, Jr. , D. Kozakov, S. Vajda, A.
Whitty, Nat. Chem. Biol. 2014, 10, 723–731.
[19] N. QikProp v4.2; Schrçdinger Inc.; New York, 2014.
[20] H. R. Hoveyda, E. Marsault, R. Gagnon, A. P. Mathieu, M. Vezina, A.
Landry, Z. Wang, K. Benakli, S. Beaubien, C. Saint-Louis, M. Brassard, J. F.
Pinault, L. Ouellet, S. Bhat, M. Ramaseshan, X. Peng, L. Foucher, S. Beau-
chemin, P. Bherer, D. F. Veber, M. L. Peterson, G. L. Fraser, J. Med. Chem.
2011, 54, 8305–8320.
[21] a) C. A. Lipinski, J. Pharmacol. Toxicol. Methods 2000, 44, 235–249;
b) D. F. Veber, S. R. Johnson, H. Y. Cheng, B. R. Smith, K. W. Ward, K. D.
Kopple, J. Med. Chem. 2002, 45, 2615–2623.
[22] H. S. Beckmann, F. Nie, C. E. Hagerman, H. Johansson, Y. S. Tan, D.
Wilcke, D. R. Spring, Nat. Chem. 2013, 5, 861–867.
[9] A. P. Treder, M. C. Tremblay, A. K. Yudin, E. Marsault, Org. Lett. 2014, 16,
4674.
[23] A. Roxin, J. Chen, C. C. Scully, B. H. Rotstein, A. K. Yudin, G. Zheng, Bio-
conjugate Chem. 2012, 23, 1387–1395.
[10] D. C. Kapeller, S. Bräse, Synlett 2011, 161–164.
[11] T. F. Nishiguchi, S. Kuroda, M. Kajisaki, K. Saitoh, J. Org. Chem. 1998, 63,
8183–8187.
[24] http://www.silicycle.com/ca/products/silicycle-miniblock-xt.
[25] S. I. Maestro v9.9, New York, NY, 2014.
[26] S. I. N. Y. Macromodel v10.3, NY, 2014.
[12] G. RossØ, F. Ouertani, H. Schrçder, J. Comb. Chem. 1999, 1, 397–401.
[13] a) T. Conroy, K. A. Jolliffe, R. J. Payne, Org. Biomol. Chem. 2009, 7, 2255–
2258; b) D. T. Rijkers, J. A. den Hartog, R. M. Liskamp, Bioorg. Med. Chem.
2004, 12, 5099–5106.
Received: January 7, 2015
[14] B. H. Rotstein, V. Rai, R. Hili, A. K. Yudin, Nat. Protoc. 2010, 5, 1813–
1822.
Published online on May 26, 2015
Chem. Eur. J. 2015, 21, 9249 – 9255
www.chemeurj.org
9255
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim