Propargyloxyproline Regio- and Stereoisomers for Click-Conjugation of Peptides: Synthesis and Application in Linear and Cyclic Peptides

Article abstract of DOI:10.1071/CH15146

The use of the click reaction for the introduction of conjugate groups, such as affinity or fluorescent labels, to a peptide for the study of peptide biochemistry and pharmacology is widespread. However, the nature and location of substituted 1,2,3-triazoles in peptide sequences may markedly affect conformation or binding as compared with native sequences. We have examined the preparation and application of propargyloxyproline (Pop) residues as a precursor to such peptide conjugates. Pop residues are available in a range of regio- and stereoisomers from hydroxyproline precursors and are readily prepared in Fmoc-protected form. They can be incorporated routinely in peptide synthesis and broadly retain the conformational properties of the parent proline containing peptides. This is exemplified by the preparation of biotin- and fluorophore-labelled peptides derived from linear and cyclic peptides.

Full text of DOI:10.1071/CH15146

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2015, 68, 1365–1372
Full Paper
http://dx.doi.org/10.1071/CH15146
Propargyloxyproline Regio- and Stereoisomers for
Click-Conjugation of Peptides: Synthesis and
Application in Linear and Cyclic Peptides
A
A
A B
,
Susan E. Northfield, Simon J. Mountford, Jerome Wielens,
A
C
C
D
Mengjie Liu, Lei Zhang, Herbert Herzog, Nicholas D. Holliday,
Martin J. Scanlon, Michael W. Parker, David K. Chalmers,
and Philip E. Thompson
A
B E
,
A
A F
,
A
Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences, 381 Royal Parade,
Parkville, Vic. 3052, Australia.
B
ACRF Rational Drug Discovery Centre, St Vincent’s Institute of Medical Research,
Fitzroy, Vic. 3065, Australia.
C
Neuroscience Research Program, Garvan Institute of Medical Research,
St Vincent’s Hospital, Darlinghurst, NSW 2010, Australia.
D
Institute of Cell Signalling, School of Life Sciences, University of Nottingham,
Queen’s Medical Centre, Nottingham NG7 2UH, UK.
E
Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and
Biotechnology Institute, The University of Melbourne, Parkville, Vic. 3010, Australia.
Corresponding author. Email: philip.thompson@monash.edu
F
The use of the click reaction for the introduction of conjugate groups, such as affinity or fluorescent labels, to a peptide
for the study of peptide biochemistry and pharmacology is widespread. However, the nature and location of substituted
1,2,3-triazoles in peptide sequences may markedly affect conformation or binding as compared with native sequences.
We have examined the preparation and application of propargyloxyproline (Pop) residues as a precursor to such peptide
conjugates. Pop residues are available in a range of regio- and stereoisomers from hydroxyproline precursors and are
readily prepared in Fmoc-protected form. They can be incorporated routinely in peptide synthesis and broadly retain
the conformational properties of the parent proline containing peptides. This is exemplified by the preparation of biotin-
and fluorophore-labelled peptides derived from linear and cyclic peptides.
Manuscript received: 26 March 2015.
Manuscript accepted: 15 May 2015.
Published online: 24 June 2015.
Introduction
impact on cyclic peptide conformation. A position in the
peptide must be identified where a conjugate could be incor-
porated without disrupting the binding or activity of the parent
peptide sequence.
The ability to readily label and/or conjugate peptides is an
important facet of biological chemistry research.[1] Such con-
jugates can be used for tagging bioactive peptides with specific
labels to track and identify binding targets, or they can be used
to alter physicochemical properties for improved pharmaco-
logical activity.[2] The positions at which peptides can be
usefully functionalised are critically dependent on what region
of the peptide is necessary for activity. In linear peptides,
conjugates can be added to either the N- or C-terminus, or from
one of the amino acid side-chains, commonly a lysine or
tyrosine, but these regions should not be part of the key phar-
macophore. Conjugates can also be attached using a range of
linking chains, varying in length and/or polarity, to distance
them from the bioactive peptide. When dealing with head-to-
tail cyclic peptides, the choice of where to introduce conjugates
is more limited as there is no free terminus available at which to
functionalise, and local structural changes can have a marked
Proline provides unique conformational restraint as com-
pared with the other natural amino acids, driven by its cyclic
structure and the presence of a tertiary amide bond. Proline is
also frequently reported as a tool to induce reverse-turns in
cyclic peptides.[3] This function can place the proline residue in
a conformation protruding away from the peptide binding/
activity site, making proline an attractive residue at which to
incorporate functionality (Fig. 1). Proline derivitisation, or
‘editing’ as it has recently been termed, has been shown to be
a diverse approach to peptide derivatisation.[4] Zondlo et al.
have described a range of diverse modifications to proline that
allow for further derivatisation. This report on an array of
substituted prolines notes the influence substitution might have
on cis–trans isomerisation and even peptide conformation.
Journal compilation Ó CSIRO 2015
www.publish.csiro.au/journals/ajc

1366
S. E. Northfield et al.
N
N
N
O
O
SPPS
Click
O
N
N
Fmoc
Peptide
CO2H
Peptide
N
O
Peptide
Peptide
and isomers
O
Fig. 1. ‘Exposed’ proline residues in cyclic or linear peptides can be modified to include ‘handles’ for
side-chain conjugation. Proline residues appended with a propargyloxy handle are incorporated in place
of proline in peptides. Labels (star) are introduced using standard conditions for copper-catalysed
azide–alkyne conjugation (CuAAC) reactions (SPPS: solid phase peptide synthesis).
Similarly, we envisaged that introduction of a propargyloxy
group yielding propargyloxyproline (Pop) residues might
fulfil a similar role in the development of peptide conjugates
undergoing click reactions. The click reaction is one of the most
widely used methods of introducing conjugates into azido or
alkyne-containing peptides.[5] The use of click reactions using
Pop derivatives as substrates has included the synthesis of
macrocyclic peptides[6] and also as a way of immobilising
proline as an asymmetric catalyst for aldol reactions.[7] We
anticipated using hydroxyproline stereoisomers as precursors, to
gain access to a range of isomeric Pop-containing peptides. The
resultant triazolylmethoxy conjugates would supply a useful
spacer away from the peptide chain, minimising the impact of
the conjugate on the peptide structure.
We report here on the synthesis of a range of Fmoc-protected
Pop regio- and stereoisomers and their incorporation into two
classes of peptide of interest in our laboratories. The first were
derivatives of the Y1 receptor antagonist peptide, BVD15, and
the second were cyclic hexapeptides incorporating a Lys–Ile–
Asp–Asn (KIDN) pharmacophore motif of lens epithelium-
derived growth factor (LEDGF), a key protein for the activity
of HIV integrase (IN).
We and others have had an on-going interest in the develop-
ment of conjugates of peptides that bind to the Y1 G-protein
coupled receptors,[8] and we had reported modification of a
proline residue by Ctp (Fig. 2) in the dimeric Y1 antagonist,
1229U91 but with significant loss of activity.[9] Another promi-
nent starting point has been the 10-residue peptide Y1 antago-
nist, BVD-15,[10] which has been more amenable to
conjugation.[8,9] We decided to investigate the use of conjugates
built around a variety of Pop isomers.
In our second application of the building blocks we
focussed on a series of cyclic hexapeptides based upon the
reverse turn motif of LEDGF, a protein that binds at the
dimer interface of IN.[11] The interaction is essential for
efficient integration of HIV DNA into the host chromosome,
and consequently for successful viral replication.[12] These
hexapeptides contain the tetrapeptide sequence Lys–Ile–
Asp–Asn (KIDN) linked by a dipeptide scaffold comprised
of one or two proline residues to support a reverse-turn
pharmacophore at the tetrapeptide portion. In X-ray struc-
tures of peptide–IN complexes, the dipeptide scaffold pro-
jected away from the protein binding site and was not taking
part in any binding interactions. This presented an opportu-
nity for us to incorporate additional functionality to our
peptides.
Results and Discussion
Synthesis of Fmoc-Pop Stereoisomers
Fmoc-protected Pop derivatives were prepared from their cor-
responding hydroxyproline isomers, including the commonly
occurring amino acid (2S,4R)-hydroxyproline (or trans-4-
hydroxy-L-proline). A selection of isomers were synthesised
using a common synthetic strategy (Scheme 1). For example,
(2S,4R)-hydroxyproline (i) was Boc-protected (v), and then
treated with propargyl bromide under basic conditions to give
the Boc-(2S,4R)-Pop (ix). Deprotection and then reprotection of
Na gives the key building block Fmoc-(2S,4R)-4-propargyl-
oxyproline (xiii) in 25 % overall yield. Note that the preparation
of ix was recently reported and the sensitivity of the alkylation to
selected base and solvent conditions was highlighted.[13] The
three step synthesis could be carried out without purification of
intermediates, carrying through the crude products at each step.
In the final step, Fmoc-OSu was limited to 1.0 molar equivalent,
limiting the formation of Fmoc-b-Alanine-OH,[14] which had
proved difficult to separate from xiii.
In the same manner, we prepared the N-Fmoc-protected
O-propargyl derivatives of (2R,4R)-, [xiv, cis-4-hydroxy-D-
proline] (2S,4S)- (xv, cis-4-hydroxy-L-proline), and (2S,3R)-
hydroxyproline (xvi, cis-3-hydroxy-L-proline) stereoisomers
from the corresponding building blocks giving a collection of
Fmoc-protected amino acids on a gram scale for incorporation
into peptides using standard solid phase peptide synthesis
(SPPS) protocols (see Supplementary Material).
Synthesis of NPY Analogues
Three parent Y1 antagonists Lys4-BVD15 (1), Arg4-BVD15 (2),
and a cyclic peptide, c(Glu2,Dap4)-BVD15, 3 and their Pop-
containing analogues 4–9 were prepared. The synthesis of linear
peptides 1, 2, and 4–7 was achieved using standard Fmoc-based
SPPS protocols using Rink Amide resin. Cleavage using tri-
fluoroacetic acid (TFA) yielded the products in good recovery
and purity.[9] Peptides 3 and 8 were prepared by solid phase
synthesis of the linear Fmoc-precursor and solution phase Glu
to Dap cyclisation. Peptide 9 was prepared similarly but with an
N-terminal 4-fluorbenzoic acid group. Cyclisation was carried
out in DMF (1 mg mLꢀ1 peptide) using PyClock (2 equiv.)
coupling reagent and N-methylmorpholine (NMM) (12 equiv.)
as activating base. We have previously reported the use of
PyClock as a cyclisation reagent for the dimeric forms of the
peptides, 1229U91,[9] but the formation of the dimeric product is
suppressed by using NMM instead of diisopropylethylamine

Propargyloxyproline for Click-Conjugation of Peptides
1367
ClϪ
Nϩ
N
O
ClϪ
Nϩ
O
NH2
N
O
N
N
O
O
O
O
O
N
N
N
N
N
N
N
N
N
O
O
N
O
N
N
O
O
O
Ctp
R1tp
R2tp
(Rhodamine-2-triazole-proline)
(Coumarin-triazole-proline)
(Rhodamine-1-triazole-proline)
O
H
N
HN
HO
O
O
OH
O
S
O
O
H
OH
O
H
HN
OH
N
NH
N
N
N
N
N
O
N
N
O
N
N
O
O
N
N
O
O
O
Atp
Btp
Biotin-PEG2-triazole-proline
Ptp
Acetyl-coumarin-triazole-proline
Pyranose-sugar-triazole-proline
Fig. 2. Structures of labels incorporated in the NPY and LEDGF analogues.
HO
O
O
HO
a
b
CO2H
c
CO2H
CO2H
Fmoc
CO2H
N
N
N
N
H
Boc
Boc
i ϭ 2S,4R
ii ϭ 2R,4R
iii ϭ 2S,4S
iv ϭ 2S,3R
v ϭ 2S,4R
vi ϭ 2R,4R
vii ϭ 2S,4S
viii ϭ 2S,3R
ix ϭ 2S,4R
x ϭ 2R,4R
xi ϭ 2S,4S
xii ϭ 2S,3R
xiii ϭ 2S,4R
xiv ϭ 2R,4R
xv ϭ 2S,4S
xvi ϭ 2S,3R
Scheme 1. Synthetic scheme for the synthesis of Fmoc-propargyloxyproline residues xiii–xvi. (a) Boc-anhydride, triethylamine, MeOH
reflux overnight; (b) NaH, DMF, propargyl bromide, 08C, 2 h; (c) 1 : 1 triflouroacetic acid/dichloromethane, room temp 30 min, then
Fmoc-OSu, dioxan, 08C, pH 10, 60 min, room temperature overnight.
(DIPEA) as base. For peptides 3 and 8, the N-terminal Fmoc
group was removed and the peptides precipitated in diethyl
ether. The recovered products were used directly for click
chemistry and reverse phase (RP)-HPLC purification.
With the precursors in hand, conjugation reactions were
performed. Peptides 4–7 were conjugated with 7-amino-4-
azidomethylcoumarin to yield the products 10–15. Peptide 4
was also conjugated with azido-functionalised rhodamine
fluorophores to give 16 and 17 (Fig. 2 and Table 1).
Click reactions were carried out by one of two methods,
depending on the solubility of the peptide and azide reagents
involved. Reactions were performed at room temperature, in the
presence of copper sulfate, sodium ascorbate, and stabilising
ligand. Using a DMF solvent system, 1 mg mLꢀ1 of peptide was

1368
S. E. Northfield et al.
treated with a 4-fold excess of azide-conjugate including the
copper-stablising ligand TBTA (tris[(1-benzyl-1H-1,2,3-triazol-
4-yl)methyl]amine). Alternately, an aqueous phosphate buffer
system could be applied.[15] In this method peptides at a
concentration of 20 mg mLꢀ1 in phosphate buffer (pH 7) were
additionally treated with aminoguanidine hydrochloride and
stabilising ligand THPTA (tris(3-hydroxypropyltriazolyl-
methyl)amine). Coupling reactions were very efficient, and
the peptides were subsequently purified by RP-HPLC. While
successfully prepared using crude Pop-containing peptides, the
complexity of the product mixture suggests that purification of
the precursor peptides is advisable.
of the rhodamine group. These were assayed in a recombinant
Y1-293TR cell system and compared again to the parent peptide
1. The affinity of these conjugates proved comparable to 1 with
IC50 values of 10 and 18 nM respectively (Table 1, Fig. S2 in the
Supplementary Material). While still a reasonably strong affini-
ty, we have developed superior conjugates through other routes,
which will be reported elsewhere.
Synthesis of Head-to-Tail Cyclic LEDGF Mimics
The second series of peptides we chose to study using Pop-
derived conjugates were designed to mimic the binding loop of
LEDGF that binds to IN. Inhibitors of LEDGF–IN binding are
postulated to be potential inhibitors of HIV DNA integration into
host cells, and the binding loop of LEDGF comprises a tetra-
peptide Lys–Ile–Asp–Asn motif.[12a,16] We had developed cyclic
hexapeptides including turn-inducing dipeptide units of one or
twoproline residues. Crystal structures of these peptides bound to
IN showed preservation of the tetrapeptide pharmacophore and
that the Pro residues protruded away from the protein binding
pocket, providing a potential conjugation point. While showing a
modest binding affinity (Kd ,1 mM as measured by surface
plasmonresonance (SPR) and HSQC NMR), these peptides show
strong conformational mimicry of the native protein.
Pop residues were included in three cyclic peptides based
upon three parent cyclic hexapeptides for which we had deter-
mined crystal structures: cyclo[Asn–D-Pro–Pro–Lys–Ile–Asp]
18, cyclo[Asn–D-Val–Pro–Lys–Nle–Asp] 19, cyclo[Asn–D-
Val–Pro–Lys–D-Ile–Asp], and 20 (PDB ID: 3WNG and 3WNH)
(Northfield et al., in preparation). In the first example, D-proline
of 18 was replaced with cis-4-propargyloxy-D-proline (xiv) to
Analysis of NPY Analogues
The labelled products of these studies were screened for Y1
receptor affinity in competition binding studies using membrane
preparations from Y2Y4 knockout mice as described previ-
ously.[9] The conjugates 10–15 showed dose-dependent com-
petition with radiolabelled NPY for receptor binding and
comparable to the parent sequences in most cases with 50 %
inhibitory concentration (IC50) values of between 0.6 and 6 nM
(Table 1, Fig. S1 in the Supplementary Material). This showed
that the receptor was relatively unaffected by the peptide sub-
stitutions irrespective of the position or chirality of the alkoxy
substituent. Compound 15 was the exception with a relatively
poor affinity, the combination of FBz group and coumarin
(Fig. 2) both proving deleterious to affinity.
Two rhodamine derivatives 16 and 17 were also prepared,
noting that these analogues could be potentially of use in
receptor imaging using the fluorescence excitation properties
Table 1. Conjugated NPY-derived peptides, incorporating functionality using click chemistry
FBz ¼ 4-Fluorobenzoyl. For other abbreviations see Fig. 2
Peptide
Sequence
ESI-MSA [m/z]
IC50 [nM] Y2Y4 KOE
Precursor peptides
1
2
3
Ile–Asn–Pro–Lys–Tyr–Arg–Leu–Arg–Tyr (Lys4-BVD15)
Ile–Asn–Pro–Arg–Tyr–Arg–Leu–Arg–Tyr (Arg4-BVD15)
Ile–cyc[Glu–Pro–Dap]–Tyr–Arg–Leu–Arg–Tyr (c[Glu2,Dap4]–BVD15)
611.7B
625.7B
589.2B
0.9
1.3
0.9
4
Ile–Asn–trans-4-L-Pop–Lys–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–trans-4-L-Pop–Arg–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–cis-4-L-Pop–Lys–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–cis-3-L-Pop–Lys–Tyr–Arg–Leu–Arg–Tyr
Ile–cyc[Glu–trans-4-L-Pop–Dap]–Tyr–Arg–Leu–Arg–Tyr
FBz–Ile–cyc[Glu–trans-4-L-Pop–Dap]–Tyr–Arg–Leu–Arg-Tyr
638.6B
652.7B
638.8B
638.8B
616.3B
677.4B
1.0
5
6
7
8
9
Final products
10
11
12
13
14
15
Ile–Asn–trans-4-L-Ctp–Lys–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–trans-4-L-Ctp–Arg–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–cis-4-L-Ctp–Lys–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–cis-3-L-Ctp–Lys–Tyr–Arg–Leu–Arg–Tyr
Ile–cyc[Glu–trans-4-L-Ctp–Dap]–Tyr–Arg–Leu–Arg–Tyr
FBz–Ile–cyc[Glu–trans-4-L-Ctp–Dap]–Tyr–Arg–Leu–Arg–Tyr
498.3C
507.8C
498.4C
498.3C
724.3B
785.5B
0.6
6.0
1.9
0.9
1.9
41
IC50 [nM] Y1-HEK293F
1
7.9
10
16
17
Ile–Asn–trans-4-L-R1tp–Arg–Tyr–Arg–Leu–Arg–Tyr
Ile–Asn–trans-4-L-R2tp–Arg–Tyr–Arg–Leu–Arg–Tyr
685.7D
649.0D
18
AESI-MS ¼ electrospray ionisation–mass spectrometry.
BESI-MS base peak corresponds to [M þ 2H]2þ
CESI-MS base peak corresponds to [M þ 3H]3þ
.
.
DESI-MS base peak corresponds to [M þ TFA þ 3H]3þ ([M þ 3H]3þ peaks were observed at lower intensity).
EInhibition of 125I-NPY (25 pM) binding to brain membrane homogenates.
FInhibition of 125I-PYY (15 pM) binding to Y1 transfected 293TR cells.

Propargyloxyproline for Click-Conjugation of Peptides
1369
give 21, and in 19 and 20 L-proline was replaced with trans-
4-propargyloxy-L-proline (xiii) to give 22 and 23 respectively
(Table 2).
The synthesis of the peptides was achieved with Fmoc-based
SPPS of side-chain protected linear precursors followed by
solution phase cyclisation and then side-chain deprotection.
The linear hexapeptide chains were prepared from Fmoc–Asp
(OtBu)–chlorotrityl resin, followed by solution head-to-tail
cyclisation of the linear peptides using diphenylphosphorylazide
(DPPA) as the cyclisation reagent and finally side-chain cleav-
age. The sequences showed some propensity for racemisation in
the cyclisation step, but the D-Asp-containing diastereomers
were in general minor components and readily separated from
the desired compounds. The recovered yields for cyclisation of
Pop-containing peptides 21–23 ranged from 30 to 70 %, and
were comparable to those obtained in the synthesis of the parent
proline-containing sequences 18–20. This demonstrated that
substituting L-Pro for trans-4-L-Pop and D-Pro for cis-4-D-Pop
did not have a detrimental effect on peptide cyclisation.
The Pop-substituted cyclic peptides were purified by semi-
preparative RP-HPLC, then used as substrates for click reactions
with a variety of azido-compounds: 7-amino-4-azidomethyl-
coumarin, 7-acetylamino-4-azidomethylcoumarin, Biotin-
PEG2-azide, and azido-6-deoxy-a-D-galactopyranose (Fig. 2,
Table 2). All of the selected conjugates were successfully
coupled to one or more of the cyclic peptides 21–23, under
standard conditions within 8 h using the same DMF click
chemistry method described for the NPY analogue conjugation
above. An interesting feature of the coupling of the 4-azido-
methyl-7-acetamidocoumarin in the synthesis of peptide 25 was
the increase in fluorescence with progression of the reaction
over a 48 h period, while the spectrum of the same mixture in the
absence of the catalyst was unchanged (Fig. 3). While modest,
the ability to distinguish substrates from products by fluores-
cence might be useful in performing click reactions in more
complex media.
Table 2. Conjugated LEDGF-derived peptides, incorporating func-
tionality using click chemistry
Peptide Sequence
Precursors
ESI-MSA [Da]
18
19
20
21
22
23
cyclic[Asn–D-Pro–Pro–Lys–Ile–Asp]
cyclic[Asn–D-Val–Pro–Lys–Nle–Asp]
cyclic[Asn–D-Val–Pro–Lys–D-Ile–Asp]
665.6
667.5
667.5
719.5
721.6
721.6
cyclic[Asn–cis-4-D-Pop–Pro–Lys–Ile–Asp]
cyclic[Asn–D-Val–trans-4-L-Pop–Lys–Nle–Asp]
cyclic[Asn–D-Val–trans-4-L-Pop–Lys–D-Ile–Asp]
Final products
24
25
26
27
28
29
30
cyclic[Asn–cis-4-D-Ctp–Pro–Lys–Ile–Asp]
935.4
977.4
cyclic[Asn–cis-4-D-Atp–Pro–Lys–Ile–Asp]
cyclic[Asn–cis-4-D-Btp–Pro–Lys–Ile–Asp]
cyclic[Asn–D-Val–trans–4-L-Actp–Lys–Nle–Asp]
cyclic[Asn–D-Val–trans-4-L-Btp–Lys–Nle–Asp]
cyclic[Asn–D-Val–trans-4-L-Ptp–Lys–Nle–Asp]
cyclic[Asn–D-Val–trans-4-L-Btp–Lys–D-Ile–Asp]
1119.6
979.5
1121.6
926.6
1121.6
AESI-MS peaks correspond to m/z: [M þ H]þ.
Analysis of Cyclic LEDGF Mimics
With the conjugated peptides in hand we examined the effect the
substitution had on the peptide conformation at the LEDGF
binding site of IN. The peptides showed comparable, albeit
weak affinity for IN to the parent hexapeptides, 18 and 19, and
we obtained crystal structures of respective labelled derivatives
peptide 24 and peptide 28, and the propargyloxy derivatives 21
and 22 bound to the core domain of IN.
The crystal structures all show well resolved density at the
tetrapeptide sequence, allowing for comparison of the homolo-
gous series at the key pharmacophore (Fig. 4). The density of the
prosthetic structures was poorly resolved suggesting that the
labels are flexible and do not interact with the IN protein.
Peptides 18, 21, and 24 differ only in the presence of the
substituent at the cis-4-position of the D-proline residue. Peptide
18 and the coumarin conjugate 24 show a close overlay of the
tetrapetide pharmacophore. In peptide 21 however the lysine
residue has moved substantially and cannot be said to be
mimicking the native structure. In both cases, the proline motif
has undergone some conformational change, although the
∗∗
250
200
150
100
50
∗
11.5 12.0
Unlabelled peptide with
coumarin azide
Clicked peptide (95 %)
0
350
400
450
500
550
600
Wavelength [nm]
Fig. 3. Formation of labelled peptide 25 from 21. (a) Reaction mixture in
absence (blue) or presence (red) of catalyst after 48 h reaction time.
(b) Reverse phase-HPLC trace of mixture showing 21 (*) and 25 (**).
Fig. 4. Crystal structures of peptides 18, 21, and 24 (left to right) in complex with core domain of HIV integrase (IN).

1370
S. E. Northfield et al.
resolution at those residues is not sufficient to identify the cause.
In the case of the cyclic peptide structures 19, 28, and 22 again
the conformation of the conjugate 28 more closely resembled
the native Pro-containing peptide 19 than the Pop counterpart 22
(Fig. 3 in the Supplementary Material).
acetonitrile (ACN) in 0.05 % aqueous TFA over 10 min or
0–100 % B over 15 min (Buffer B is 100 % ACN þ 0.1 %
TFA) at a flow rate of 0.2 mL minꢀ1 unless stated otherwise.
Mass spectra were obtained in positive mode with a scan range
of m/z 200–2000. Semi-preparative RP-HPLC was performed
using a Waters Associates liquid chromatography system (Mod-
el 600 controller and Waters 486 Tuneable Absorbance Detec-
tor) using a gradient of 0–64 % ACN in 0.1 % TFA over 20 min
or 30 min at a flow rate of 10 mL minꢀ1 on a Phenomenex Luna
Conclusions
In commencing this work, we reasoned that the incorporation of
(1H-1,2,3-triazol-4-yl)methoxy substituents on proline would
be a successful strategy in order to produce conjugated versions
of proline-containing peptides, especially as compared with the
corresponding products from commonly used propargylglycine.
Proline is a structurally rigid amino acid and so an additional
substituent might not be expected to impact the native peptide
conformation, and the linker group is also relatively remote from
the peptide backbone. Except in the case of proline isomerases
and proline specific proteases, proline does not generally play a
direct role in ligand binding events and so can be a benign place
to make a residue replacement. Furthermore, the use of varied
stereochemistry or regiochemistry of substituents projecting
from the proline ring allows conjugates to be directed away from
the peptide pharmacophore or the target protein binding site, but
may also influence levels of cis or trans amide conformers. The
Pop stereoisomers can be readily obtained from commercially
available starting materials and incorporated in peptide
sequences using standard Fmoc-SPPS protocols. The pharma-
cological and biophysical data we have obtained in these two
examples supports the approach for both small linear and cyclic
peptides. Given the absence of N- or C-terminal residues, the
ability to link through proline seems prospectively valuable in
head-to-tail cyclic peptides especially. However, the data also
provide the caveat that irrespective of the linking handle, the
nature of the prosthetic group can have a marked effect on the
binding affinity or conformation adopted by these conjugates.
˚
C8 100 A, 10 mm (50 ꢁ 21.2 mm I.D.) or a Phenomenex Luna C8
˚
100 A, 10 mm (250 ꢁ 21.2 mm I.D.) column.
Chemical Synthesis
(2S,4R)-1-(Tert-butoxycarbonyl)-4-hydroxypyrrolidine-
2-carboxylic Acid (Boc-trans-L-4-hydroxyproline, v)
To a stirred solution of trans-L-4-hydroxyproline i (2.0 g,
15.3 mmol) in MeOH (36.0 mL) was added Et3N (4.0 mL,
28.7 mmol) and Boc anhydride (6.7 g, 30.5 mmol) and the
reaction was refluxed for 3.5 h, cooled to room temperature,
and stirred for 20 h. Solvent was removed under vacuum and the
residue cooled to 08C. Following the addition of NaH2PO4
(150 mg), the solution was acidified to pH 2 with 0.5 M HCl.
The mixture was stirred at 08C for 30 min before extracting
the product with EtOAc (4 ꢁ 20 mL). The combined organic
layers were dried with MgSO4 and filtered. The solvent was
removed under vacuum yielding v as a white foam (3.23 g,
14 mmol, 92 %)
dH (CD3OD, 400 MHz) 4.40 (dd, J 5.5, 3.4, CH, 1H), 4.32 (dt,
J 12.9, 8.0, CH, 1H), 3.54 (dt, J 11.4, 4.0, 0.5 ꢁ CH2, 1H), 3.44
(dt, J 11.4, 1.9, 0.5 ꢁ CH2, 1H), 2.27 (dddd, J 12.3, 7.7, 2.8, 1.8,
0.5 ꢁ CH2, 1H), 2.06 (ddd, J 13.2, 8.6, 4.5, 0.5 ꢁ CH2, 1H), 1.45
(s, Boc, 9H). dC (CD3OD, 101 MHz) 176.75 and 176.37 (pair of
rotamers, Cq), 156.54 and 156.02 (pair of rotamers, Cq), 81.72
and 81.42 (pair of rotamers, Cq), 70.68 and 70.06 (pair of
rotamers, CH), 59.39 and 58.91 (pair of rotamers, CH), 55.85
and 55.51 (pair of rotamers, CH2), 40.07 and 39.4 (pair of
rotamers, CH2), 28.71 and 28.53 (pair of rotamers, 3 ꢁ CH3).
m/z (LC-MS) 277.35 (100 %, [M þ 2Na]þ).
Experimental
Na-Fmoc-protected amino acids were purchased from Auspep
and ChemImpex. Rink amide resin and O-(1H-6-chlorobenzo-
triazol-1-y1)-N,N,N0,N0-tetramethyluronium hexafluorophos-
phate (HCTU) were purchased from ChemImpex. Piperidine
and TFA were purchased from Auspep. N,N-DIPEA, DMF, and
dichloromethane (DCM), were purchased from Merck. Diphe-
nylphosphorylazide (DPPA) and triisopropylsilane (TIPS) were
purchased from Sigma–Aldrich. Fluorobenzoic acid was pur-
chased from Alfa Aesar. 7-Amino-4-(azidomethyl)-2H-chromen-
2-one[17] was a gift from Dr Bim Graham (Monash Institute of
Pharmaceutical Sciences). The Rhodamine B derivatives were
prepared in-house. All chemicals were used without further
purification.
1H NMR spectra were routinely recorded at 300 MHz using
a 300 MHz Bruker Advance DPX-300 spectrometer or at
400 MHz using a 400 MHz Bruker Ultrashield–Advance III
NMR spectrometer, with TOPSPIN v2.1 software, at 298 K.
13C NMR spectra were recorded at 101 MHz using a 400 MHz
Bruker Ultrashield–Advance III NMR spectrometer, with
TOPSPIN v2.1 software, at 298 K. Liquid chromatography mass
spectra were acquired on a Shimadzu 2020 LCMS system
incorporating a photodiode array detector coupled directly into
an electrospray ionisation source and a single quadrupole mass
analyser. Standard RP-HPLC was carried out at room tempera-
ture employing a Phenomenex Luna C8 (100 ꢁ 2.0 mm internal
diameter, I.D.) column eluting with a gradient of either 0–64 %
(2S,4R)-1-(Tert-Butoxycarbonyl)-4-(prop-2-yn-1-yloxy)
pyrrolidine-2-carboxylic Acid (Boc-trans-L-4-
propargyloxyproline, ix)
A
solution of Boc-trans-L-hydroxyproline, v (2.80 g,
12.13 mmol) in dry DMF (30 mL) was added to a suspension
of NaH (0.93 g, 38.75 mmol) in dry DMF (10 mL) under nitro-
gen at 08C. After 15 min, 1.5 equivalents of propargyl bromide
(80 % in toluene) was added dropwise to the reaction (1.68 mL,
18.85 mmol). The reaction was stirred at 08C for 2 h and then
quenched with H2O and lyophilised in H2O/ACN. The reaction
was taken up in EtOAc and the pH adjusted to 2 with 10 % citric
acid. The aqueous layer was extracted with EtOAc (3 ꢁ 20 mL).
The combined organic layers were dried with MgSO4 and
filtered. Solvent was removed under vacuum to yield ix as a
brown solid (2.97 g, 11.0 mmol, 91 %) which was directly
carried on to the next step.
dH (CD3OD, 400 MHz) 4.37–4.31 (m, CH, 1H), 4.31–4.21
(m, CH, 1H), 4.19 (d, J 2.4, CH2, 2H), 3.64–3.57 (m, 0.5 ꢁ CH2,
1H), 3.57–3.50 (m, 0.5 ꢁ CH2, 1H), 2.94–2.77 (m, CH, 1H),
2.44 (tddd, J 14.3, 11.5, 3.0, 1.6, 0.5 ꢁ CH2, 1H), 2.13– 2.04 (m,
0.5 ꢁ CH2, 1H), 1.45 (s, Boc, 9H). dC (CD3OD, 101 MHz)
178.33 and 175.54 (pair of rotamers, Cq), 156.04 and 155.93
(pair of rotamers, Cq), 81.64 and 80.9 (pair of rotamers, Cq),

Propargyloxyproline for Click-Conjugation of Peptides
1371
79.31 (CH), 76.2 and 75.85 (pair of rotamers, CH), 75.04 (Cq),
57.90 and 57.87 (pair of rotamers, CH), 56.58 and 56.51 (pair of
rotamers, CH2), 51.93 and 51.18 (pair of rotamers, CH2), 36.57
and 34.57 (pair of rotamers, CH2), 28.45 and 28.33 (pair of
rotamers, 3 ꢁ CH3). m/z (LCMS) 315.35 (80 %, [M þ 2Na]þ).
were then added and the reaction mixed for 3 h. Peptides 16 and
17 were prepared in the same fashion but using the appropriate
azido-substituted rhodamine B derivatives. Peptides were puri-
fied by reverse-phase preparative HPLC. Purity of fractions
was assessed using electrospray ionisation-mass spectrometry
(ESI-MS) (Table 1) and analytical HPLC (Fig. S5 in the
Supplementary Material).
(2S,4R)-1-(((9H-Fluoren-9-yl)methoxy)carbonyl)-4-
(prop-2-yn-1-yloxy)pyrrolidine-2-carboxylic Acid
(Fmoc-trans-L-4-Propargyloxyproline-OH, xiii)
Synthesis of Cyclic LEDGF Analogues
Cyclic peptides 18–23 were synthesised on 2-chlorotrityl
chloride (2CTC) resin on a 0.1 mequiv. scale. Couplings were
performed using 3 equiv. of Fmoc-protected amino acid,
3 equiv. of HCTU, and 6 equiv. of DIPEA in DMF (0.1 M in
amino acid) for 50 min. Fmoc deprotection was carried out with
30 % (v/v) piperidine in DMF (2 ꢁ 5 min). After each coupling
and deprotection step, the resin was washed six times with DMF.
Peptides were cleaved from the resin using 1 % (v/v) TFA in
DCM. Head-to-tail cyclisation of side-chain protected peptide
was performed in DMF (4 mM final concentration of peptide)
with 3 equiv. of DPPA and 4 equiv. of DIPEA. Following
removal of the solvent, side-chain protecting groups were
removed in 95: 5 TFA/TIPS. After cyclisation, the peptides were
purified by reverse-phase preparative HPLC. Purity of fractions
was assessed using ESI-MS (Table 2) and analytical HPLC.
Boc-trans-L-propargyloxyproline ix (2.97 g, 11.01 mmol)
was treated with 1 : 1 TFA/DCM (10 mL) at room temperature
over 45 min and solvent removed under vacuum. The reaction
was diluted with H2O (10 mL) and adjusted to pH 9 with
Na2CO3. To the reaction solution 1.4 equiv. of Fmoc-OSu
(5.20 g, 16.18 mmol) in dioxane (22 mL) was added at 08C and
stirred for 1 h. The reaction was then brought to room tempera-
ture and stirred overnight. Dioxane was removed under vacuum
and the reaction acidified to pH 3 with 1 M HCl. Product was
extracted with EtOAc (3 ꢁ 20 mL), washed with brine, and dried
with MgSO4. Solvent was removed under vacuum to yield a
yellow foam. Purification was achieved by flash chromato-
graphy (0–2 % MeOH in chloroform) yielding xiii as a white
powder (1.31 g, 3.35 mmol, 30 %).
dH (CD3OD, 400 MHz) 7.8 (t, J 7.5, 2H), 7.63 (td, J 7.5, 2.4,
2H), 7.39 (td, J 7.4, 4.0, 2H), 7.35–7.28 (m, 2H), 4.46–4.18 (m,
6H), 4.15 (dd, J 4.7, 2.4, 1H), 3.64–3.5 (m, 2H), 2.89 (t, J 2.4,
1H), 2.57–2.4 (m, 1H), 2.22–2.05 (m, 1H). dC (CD3OD,
101 MHz) 175.98 and 175.75 (pair of rotamers, Cq), 156.71
and 156.62 (pair of rotamers, Cq), 145.31, 145.29, 145.12,
145.05 (rotamers, Cq), 142.64, 142.61, 142.56, 142.49 (rota-
mers, Cq), 128.88 (CH), 128.25 (CH), 126.28, 125.25, 126.16,
126.15 (rotamers, CH), 121.03 and 120.98 (pair of rotamers,
CH), 80.58 and 80.57 (pair of rotamers, Cq), 77.92 and 77.15
(pair of rotamers, CH), 76.27 and 77.26 (pair of rotamers, CH),
69.32 and 68.75 (pair of rotamers, CH2), 59.26 and 59.01 (pair of
rotamers, CH), 57.25 and 57.21 (pair of rotamers, CH2), 53.21
and 52.78 (pair of rotamers, CH2), 48.39 and 48.33 (pair of
rotamers, CH), 37.6 and 36.6 (pair of rotamers, CH2). m/z
(LCMS) 392.30 (100 %, [M þ H]þ). HRMS m/z 392.1494;
C23H22NOþ5 [M þ H]þ requires 392.1492.
Synthesis of Peptides 24–30
A solution of the Pop-containing peptide (1 mg mLꢀ1 in
H2O) was treated with a 4-fold excess of the azido derivative
(1 mg mLꢀ1 in DMF). One equivalent of sodium ascorbate
(1 mg mLꢀ1 in H2O), one equivalent of TBTA (1 mg mLꢀ1 in
DMF), and one equivalent of copper sulfate (1 mg mLꢀ1 in H2O)
were subsequently added to the reaction. The reaction was left at
room temperature and progression monitored by LCMS. When
no remaining unlabelled peptide was observed, the reaction was
diluted in 1 : 1 ACN/H2O and lyophilised before purification
by RP-HPLC. Purity of fractions was assessed using ESI-MS
(Table 2) and analytical HPLC (Fig. S5 in the Supplementary
Material).
In the case of peptide 29, the click reaction was performed
using
1,2 : 3,4-di-O-isopropylidne-6-azido-6-deoxy-a-D-
galactopyranose.[19] The resultant acetonide (m/z 1006.7,
[M þ H]þ) was deprotected by treatment with 90 % TFA
overnight, diluted in 1 : 1 ACN/H2O, and lyophilised before
purification by RP-HPLC to yield the free galactopyranose 29.
Compounds xiv, xv, and xvi were prepared in the same
manner. Full details are provided in the Supplementary Material.
BVD15 analogues 1–9 were prepared as previously reported.
Peptide syntheses were performed on Rink amide resin (0.3–0.7
mequiv. gꢀ1, 100–200 mesh, 0.1 mmol scale) using conven-
tional Fmoc-based solid phase peptide synthesis. Fmoc-
protected amino acids in 3-fold molar excess were coupled
using DMF as solvent, a 6-fold molar excess of DIPEA in DMF
(70 mL Lꢀ1) with a 3-fold molar excess of HCTU as the
activating agent for 50 min. Fmoc deprotection was carried
out by treatment with 20 % piperidine in DMF for 10 min.
Peptide cleavage from the resin was performed using a
cocktail containing TFA/TIPS/DMB (92.5 : 2.5 : 5 %; DMB ¼
1,3-dimethoxybenzene) for 3 h.[18] The cleavage mixture was
filtered, concentrated by a stream of nitrogen, precipitated by
cold diethyl ether, and centrifuged. The resulting crude product
was dissolved by water/ACN (1 : 1) and lyophilised overnight.
The click reactions to prepare peptides 10–15 involved
dissolving the corresponding peptide–alkyne 5–9 (1 equiv.) in
H2O and adding a solution of the azidocoumarin[17] (4 equiv.)
in DMF to give a 1 : 3 ratio of H2O to DMF. Copper sulfate
(10 equiv.), TBTA (10 equiv.), and sodium ascorbate (10 equiv.)
Competition Binding Studies
Competition binding assays were carried out as described pre-
viously.[9] In brief, receptor binding assays to measure Y1R
affinity of the ligands 10–15 (described below) were performed
on crude membranes prepared from the brains of Y2R- and Y4R-
deficient mice (Y2ꢀ/ꢀY4ꢀ/ꢀ), where Y1R accounts for the
majority of remaining Y receptors. Peptides 16 and 17 were
assayed using 293TR Y1 receptor GFP membranes.
For mouse brain preparations, equal volumes (25 mL) of
non-radioactive ligands and 125I-human polypeptide YY
125I-hPYY, 2200 Ci mmolꢀ1; PerkinElmer Life Science Pro-
ducts, Boston, MA, USA) were added into each assay. The final
concentration of 125I-hPYY in the assay was 25 pM. The binding
of 125I-hPYY competed with Y1R ligands of interest at increas-
ing concentrations ranging from 10ꢀ12 to 10ꢀ6 M over 2 h. Non-
radioactive human PYY (Auspep, Parkville, Vic., Australia) at
10ꢀ6 M was used as the non-specific binding control.
(

1372
S. E. Northfield et al.
Using membranes from the 293TR Y1 receptor-sfGFP cell
[3] (a) M. MacDonald, J. Aube, Curr. Org. Chem. 2001, 5, 417.
doi:10.2174/1385272013375517
competition binding assays were performed for 90 min at 218C
in buffer (25 mM HEPES, 2.5 mM CaCl2, 1.0 mM MgCl2, 0.1 %
bovine serum albumin, 0.1 mg mLꢀ1 bacitracin; pH 7.4), increas-
ing concentrations of unlabelled ligands (10ꢀ12 to 10ꢀ6 M,
duplicate), and [125I]PYY (15 pM). Non-specific binding in
these experiments comprised less than 5 % of total counts, and
was subtracted from the data.
In both sets of data, IC50 values were calculated from
displacement curves (repeated 2–4 times for each peptide, fitted
using non-linear least-squares regression in GraphPad Prism
5.01 (Graphpad software, San Diego, CA, USA).
(b) C. M. Deber, B. Brodsky, A. Rath, in Encyclopedia of Life Sciences
2010 (John Wiley & Sons, Ltd: Chichester).
[4] A. K. Pandey, D. Naduthambi, K. M. Thomas, N. J. Zondlo, J. Am.
Chem. Soc. 2013, 135, 4333. doi:10.1021/JA3109664
[5] M. Meldal, C. W. Tornøe, Chem. Rev. 2008, 108, 2952. doi:10.1021/
CR0783479
[6] G. Chouhan, K. James, Org. Lett. 2011, 13, 2754. doi:10.1021/
OL200861F
[7] D. Font, C. Jimeno, M. A. Perica`s, Org. Lett. 2006, 8, 4653.
doi:10.1021/OL061964J
[8] B. Gue´rin, V. Dumulon-Perreault, M.-C. Tremblay, S. Ait-Mohand,
P. Fournier, C. Dubuc, S. Authiera, F. Be´nard, Bioorg. Med. Chem.
Lett. 2010, 20, 950. doi:10.1016/J.BMCL.2009.12.068
[9] S. J. Mountford, M. Liu, L. Zhang, M. Groenen, H. Herzog, N. D.
Holliday, P. E. Thompson, Org. Biomol. Chem. 2014, 12, 3271.
doi:10.1039/C4OB00176A
[10] M. Liu, S. J. Mountford, L. Zhang, I. C. Lee, H. Herzog, P. E.
Thompson, Int. J. Pept. Res. Ther. 2013, 19, 33. doi:10.1007/
S10989-012-9330-Z
X-Ray Crystallography
Crystal structures of the cyclic hexapeptides bound to IN were
determined as previously described.[20] The coordinates of the
four INCORE4H123/cyclic LEDGF peptide complexes have been
deposited in the protein database (PDB) with the accession
numbers 4Y1C and 4Y1D.
[11] A. Engelman, P. Cherepanov, PLoS Pathog. 2008, 4, e1000046.
doi:10.1371/JOURNAL.PPAT.1000046
Supplementary Material
[12] (a) W. Thys, K. Busschots, M. McNeely, A. Voet, F. Christ,
Z. Debyser, HIV Ther. 2009, 3, 171. doi:10.2217/17584310.3.2.171
(b) S. Hare, P. Cherepanov, Viruses 2009, 1, 780. doi:10.3390/
V1030780
Detailed synthesis procedures as well as additional supple-
mentary figures showing dose–response curves for Y1R binding
by peptides 1–4 and 10–17, structures of peptides 19, 22, and 28
complexed with IN, NMR data for Pop derivatives v–xvi, and
RP-HPLC traces of peptides 10–17 and 24–30 are available on
the Journal’s website.
[13] V. Mihali, F. Foschi, M. Penso, G. Pozzi, Eur. J. Org. Chem. 2014,
2014, 5351. doi:10.1002/EJOC.201402429
[14] M. Obkircher, C. Sta¨helin, F. Dick, J. Pept. Sci. 2008, 14, 763.
doi:10.1002/PSC.1001
[15] V. Hong, S. I. Presolski, C. Ma, M. G. Finn, Angew. Chem. Int. Ed.
Acknowledgements
2009, 48, 9879. doi:10.1002/ANIE.200905087
This work was supported by CRC for Biomedical Imaging Development
(CRC-BID), Australia, and the Australian Research Council (ARC) Linkage
project (LP0775192). This research was partly undertaken on the MX2
beamline at the Australian Synchrotron, Victoria, Australia and the authors
thank the beamline staff for their assistance. Funding from the Victorian
Government Operational Infrastructure Support Scheme to St Vincent’s
Institute is gratefully acknowledged. M.W.P. is a National Health and
Medical Research Council of Australia Research Fellow. SEN and ML were
supported by an Australian Post-graduate scholarship.
[16] Z. Hayouka, M. Hurevich, A. Levin, H. Benyamini, A. Iosub, M. Maes,
D. E. Shalev, A. Loyter, C. Gilon, A. Friedler, Bioorg. Med. Chem.
2010, 18, 8388. doi:10.1016/J.BMC.2010.09.046
[17] M. A. Kamaruddin, P. Ung, M. I. Hossain, B. Jarasrassamee,
W. O’Malley, P. Thompson, D. Scanlon, H.-C. Cheng, B. Graham,
Bioorg. Med. Chem. Lett. 2011, 21, 329. doi:10.1016/J.BMCL.2010.
11.005
[18] P. Stathopoulos, S. Papas, V. Tsikaris, J. Pept. Sci. 2006, 12, 227.
doi:10.1002/PSC.706
[19] S. B. Ferreira, A. C. R. Sodero, M. F. C. Cardoso, E. S. Lima,
C. R. Kaiser, F. P. Silva Jr, V. F. Ferreira, J. Med. Chem. 2010, 53,
2364. doi:10.1021/JM901265H
[20] J. Wielens, S. J. Headey, J. J. Deadman, D. I. Rhodes, G. T. Le, M. W.
Parker, D. K. Chalmers, M. J. Scanlon, ChemMedChem 2011, 6, 258.
doi:10.1002/CMDC.201000483
References
[1] R. J. Pieters, D. T. S. Rijkers, R. M. J. Liskamp, QSAR Comb. Sci.
2007, 26, 1181. doi:10.1002/QSAR.200740075
[2] C. Hein, X.-M. Liu, D. Wang, Pharm. Res. 2008, 25, 2216.
doi:10.1007/S11095-008-9616-1