Exploring side-chain diversity by Submonomer solid-phase aza-peptide synthesis

Article abstract of DOI:10.1021/ol901423c

Image Presented Submonomer synthesis of aza-peptides featuring regioselective alkylation of peptide-bound aza-Gly residues provided ten aza-analogues of the Growth Hormone Releasing Peptide-6 (GHRP-6) in 15-42% yield and purity generally g90%. Circular dichroism demonstrated that azaPhe-peptide 7a induced a β-turn conformation which may be responsible for its 1000-fold improvement in GHRP-6 selectivity for the CD36 receptor. This versatile method for making aza-peptides avoids solution-phase hydrazine synthesis and is well suited for studying side-chain-activity relationships of biologically active peptides.

Full text of DOI:10.1021/ol901423c

ORGANIC
LETTERS
2009
Vol. 11, No. 16
3650-3653
Exploring Side-Chain Diversity by
Submonomer Solid-Phase Aza-Peptide
Synthesis
David Sabatino, Caroline Proulx, Sophie Klocek, Carine B. Bourguet,
Damien Boeglin, Huy Ong, and William D. Lubell*
UniVersité de Montréal, Department of Chemistry, C.P. 6128 Succursale Centre-Ville,
Montre´al, Que´bec, Canada H3C 3J7
william.lubell@umontreal.ca
Received June 23, 2009
ABSTRACT
Submonomer synthesis of aza-peptides featuring regioselective alkylation of peptide-bound aza-Gly residues provided ten aza-analogues of
the Growth Hormone Releasing Peptide-6 (GHRP-6) in 15-42% yield and purity generally g90%. Circular dichroism demonstrated that azaPhe-
peptide 7a induced a ꢀ-turn conformation which may be responsible for its 1000-fold improvement in GHRP-6 selectivity for the CD36 receptor.
This versatile method for making aza-peptides avoids solution-phase hydrazine synthesis and is well suited for studying side-chain-activity
relationships of biologically active peptides.
The ability to introduce systematically various side-chain
functionalities at different regions along a peptide or mimic
offers power for investigating structure-activity relationships
responsible for biological activity.1 Ideally, different side
chains could be attached directly to the growing peptide by
a combinatorial method using solid-phase synthesis. For
example, regioselective alkylation of supported glycine Schiff
bases has allowed a variety of side-chain functional groups
to be introduced onto amino esters as well as di- and
tripeptide fragments.2,3 Similarly, copper-catalyzed cross-
coupling reactions in solution have been used to add vinyl,
alkynyl, and aryl side chains onto N-p-methoxyphenyl
glycine residues in simple di- and tripeptides.4 Limited to
the synthesis of amino acids and short peptides, these
procedures are also generally not stereoselective, such that
solution-phase chemistry using chiral phase-transfer catalysis
has been employed to improve isomeric purity during glycine
alkylation.5,6 The issues of stereoselective C-H activation
and modification of glycine residues have thus inhibited the
(1) (a) Kiviranta, P. H.; Suuronen, T.; Walle´n, E. A. A.; Leppa¨nen, J.;
Tervonen, J.; Kyrylenko, S.; Salminen, A.; Poso, A.; Jarho, E. M. J. Med.
Chem. 2009, 52, 2153–2156. (b) Huseby, E. S.; Crawford, F.; White, J.;
Marrack, P.; Kappler, J. W. Nat. Immunol. 2006, 7, 1191–1199. (c) Kumar,
A.; Ye, G.; Wang, Y.; Lin, X.; Sun, G.; Parang, K. J. Med. Chem. 2006,
49, 3395–3401. (d) Zavala-Ruiz, Z.; Strug, I.; Anderson, M. W.; Gorski,
J.; Stern, L. J. Chem. Biol. 2004, 11, 1395–1402.
(2) O’Donnell, M. J.; Zhou, C.; Scott, W. L. J. Am. Chem. Soc. 1996,
118, 6070–6071.
(3) Maruoka, K.; Tayama, E.; Ooi, T. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 5824–5829.
(4) (a) Zhao, L.; Li, C.-J. Angew. Chem., Int. Ed. 2008, 47, 7075–7078.
(b) Zhao, L.; Basle, O.; Li, C.-J. Proc. Natl. Acad. Sci. U.S.A. 2009, 106,
4106–4111.
10.1021/ol901423c CCC: $40.75
Published on Web 07/16/2009
 2009 American Chemical Society

general applicability of these routes for studying biologically
relevant peptide sequences.3,5,6
their endocrine activity, GHRP-6 analogues possess periph-
eral cardiovascular protective effects upon binding with the
multifunctional CD36 scavenging receptor. CD36 plays a
key role in the development of atherosclerosis by binding
oxidized low-density lipoproteins (oxLDL)13 and is involved
in the down-regulation of angiogenesis in binding throm-
bospondin.14a
By modulating CD36 scanvenger receptor function,
GHRP-6 analogues offer therapeutic potential in angiogen-
esis-related diseases such as age-related macular degeneration
and diabetic retinopathy.14 Moreover, the bioactive confor-
mation of GHRP-6 has been suggested to adopt a turn-motif
based on molecular modeling studies,15 making aza-peptide
analogues of GHRP-6 interesting targets for increasing
potency and selectivity for binding to the CD36 target
receptor.
A three-step process has been developed for aza-residue
construction onto the peptide chain within a conventional
Fmoc-based solid-phase peptide synthesis (SPPS):16 (a)
acylation of the supported peptide with a hydrazone-derived
activated carbazate, (b) regioselective semicarbazone alky-
lation, and (c) chemoselective semicarbazone deprotection.
Subsequent acylation of the aza-amino acid residue, comple-
tion of the SPPS sequence, deprotection, and cleavage
provide the aza-peptide (Figure 1).
Circumventing the issues related to Gly modification, we
have explored adding side chains onto aza-Gly residues to
provide aza-peptides. Aza-peptides possess one or more aza-
amino acid residues, in which the R-carbon is substituted
for nitrogen.7 Insertion of aza-residues systematically along
a sequence can identify the location and importance of ꢀ-turn
conformations for peptide activity8 and improve pharmaco-
kinetic properties.9
The introduction of aza-amino acids into peptides involves
a combination of hydrazine and peptide chemistry.7 Typi-
cally, solid-phase aza-peptide synthesis strategies have
necessitated the preformation of hydrazine precursors in
solution prior to their incorporation on solid phase. For
example, N-Fmoc protected N′-alkyl hydrazine8e and N-Boc
aza-dipeptide fragments,8c as well as N-Fmoc-8a and N-2-
(3,5-dimethoxyphenyl)propan-2-yloxy-carbonyl (Ddz)8d-
protected aza-amino acid chlorides have been coupled to
supported peptides to introduce the aza-residue. The scope
of side-chain diversity is limited by these methods for aza-
peptide synthesis because of the inherent difficulties of
selectively differentiating the two nitrogens of the hydrazine
moiety.10 The regioselective alkylation of a supported aza-
glycine moiety is specifically designed to surmount issues
of solution-phase synthesis to introduce broader diversity of
aza-residue side chains.
As a model peptide to demonstrate the utility of our
method, we focused on modifying systematically each
residue in the D-Trp-Ala-Trp tripeptide region of the growth
hormone releasing peptide GHRP-6 (His-D-Trp-Ala-Trp-D-
Phe-Lys-NH2, 10),11 a challenging target because of poten-
tially nucleophilic side chains at the His, Trp, and Lys
residues. This hexapeptide stimulates growth hormone (GH)
release by a pathway related to the G-protein coupled ghrelin
receptor (GHS-R1a) and has been considered a target for
developing treatments for GH secretory deficiency related
to conditions such as cachexia and aging.12 In addition to
(5) O’Donnell, M. J.; Delgado, F.; Pottorf, R. S. Tetrahedron 1999, 55,
6347–6362.
(6) (a) Ooi, T.; Kameda, M.; Maruoka, K. J. Am. Chem. Soc. 1999,
121, 6519–6520. (b) Ooi, T.; Tayama, E.; Maruoka, K. Angew. Chem., Int.
Ed. Engl. 2003, 42, 579–582.
(7) Gante, J. Synthesis 1989, 405–413.
(8) (a) Boeglin, D.; Lubell, W. D. J. Comb. Chem. 2005, 7, 864–878.
(b) Boeglin, D.; Xiang, Z.; Sorenson, N. B.; Wood, M. S.; Haskell-Luevano,
C.; Lubell, W. D. Chem. Biol. Drug Des. 2006, 67, 275–283. (c) Melendez,
R. E.; Lubell, W. D. J. Am. Chem. Soc. 2004, 126, 6759–6764. (d) Freeman,
N. S.; Hurevich, M.; Gilon, C. Tetrahedron 2009, 65, 1737–1745. (e)
Quibell, M.; Turnell, W. G.; Johnson, T. J. Chem. Soc., Perkin Trans. 1
1993, 2843–2849.
(9) (a) Zega, A. Curr. Med. Chem. 2005, 12, 589–597. (b) Weber, D.;
Berger, C.; Eickerlmann, P.; Antel, J.; Kessler, H. J. Med. Chem. 2003,
46, 1918–1930.
Figure 1. Submonomer synthesis and chromatogram at 254 nm of
[aza-Phe4]-GHRP-6 (7a) after resin cleavage.
(10) Ragnarsson, U. Chem. Soc. ReV. 2001, 30, 205–213.
(11) For a recent review on growth hormone secretagogues, see: (a)
Korbonits, M.; Goldstone, A. P.; Guerguiev, M.; Grossma, A. B. Neuroen-
docrinology 2004, 25, 27–68. (b) Fehrentz, J. A.; Martinez, J.; Boeglin,
D.; Guerlavais, V.; Deghenghi, R. I. Drugs 2002, 5, 804–814.
(12) (a) Bowers, C. Y.; Sartor, A. O.; Reynolds, G. A.; Badger, T. M.
Endocrinology 1991, 128, 2027–2035. (b) Cordido, F.; Penalva, A.; Dieguez,
C.; Casanueva, F. F. J. Clin. Endocrinol. Metab. 1993, 76, 819–823. (c)
Frutos, M.G.-S.; Cacicedo, L.; Fernandez, C.; Vicent, D.; Velasco, B.;
Zapatero, H.; Sanchez-Franco, F. Am. J. Physiol. Endocrinol. Metab. 2007,
293, E1140-E1152.
Rink amide polystyrene-divinyl benzene (PS-DVB) resin
was used as a conventional support for SPPS, and the
dipeptide D-Phe-Lys(Boc) constituted our starting sequence
for aza-modifications at Trp4. [Aza-Phe4]-GHRP-6 (7a) was
synthesized to demonstrate poof-of-concept. Aza-Gly was
Org. Lett., Vol. 11, No. 16, 2009
3651

coupled to supported D-Phe-Lys(Boc) using methylidene
carbazate 2 which was prepared in situ on mixing the
respective hydrazone and p-nitrophenyl chloroformate. Better
coupling was detected by LCMS analysis of product from
benzylidene carbazate relative to its diphenylmethylidene
counterpart (3a and 3b, 95% and 89%, respectively, Figure
1). Similarly, in the regioselective alkylation17 of the
semicarbazone, benzylidene 3a was more reactive than its
diphenylmethylidene counterpart 3b; i.e., treatment of 3 with
KOtBu (300 mol %) and BnBr (300 mol %) gave resin
samples, which after cleavage and LCMS analysis exhibited,
respectively, 89% and 66% alkylated products 4a and 4b
without any diastereomeric products from epimerization
during alkylation.
Table 1. GHRP-6 aza-Peptides (7-9): Yields and
Characterization Data
Chemoselective semicarbazone deprotection necessitated
mild conditions to avoid removal of side-chain protection
and cleavage of peptide from resin. Commonly employed
solution-phase acid-catalyzed hydrolysis conditions used for
unmasking semicarbazones to aza-peptides were thus unsuc-
cessful on the solid phase.18 Considerable experimentation
demonstrated that selective semicarbazone deprotection was
effectively accomplished using trans-imination conditions,19
employing NH2OH·HCl in pyridine at 60 °C for 12 h. Again,
benzylidene 4a out-performed diphenylmethylidine 4b;
LCMS analyses showed, respectively, 81% and 64% conver-
sions to semicarbazide 5.
a Crude purity by LCMS at 254 nm using H2O (0.1% FA) and MeOH
(0.1% FA) or MeCN (0.1% FA) as eluent. b Isolated purity by LCMS at
254 nm using H2O (0.1% FA) and MeOH (0.1% FA) as eluent. c Calculated
d
from resin loading. Observed mass (expected mass) as [M + H]+ by
LCMS. e Retention times using 2-40% MeOH/H2O as eluent. f Retention
times using 2-40% MeCN/H2O as eluent. g 63:37 mixture of aza-Val3 and
aza-Gly3 GHRP-6. h 77:23 mixture of His isomers.
In our hands, efficient amino acid coupling to semicarba-
zide 5 was achieved using the symmetric anhydride
method,20 featuring activation of the Fmoc-amino acid with
diisopropylcarbodiimide, DIC. Subsequent termination of the
peptide sequence, deprotection, and cleavage gave [aza-
Phe4]-GHRP-6 (7a) in 73% crude purity. The major impuri-
ties in the chromatogram consisted of sequences generated
from incomplete semicarbazone deprotection and failed
acylation onto the aza-amino acid residue (Figure 1). With
this method in hand, we investigated next further diversifica-
tion at the peptide 4-position using para-substituted benzyl
halides in the same sequence and produced respectively [aza-
p-methoxy-Phe4]- and [aza-p-trifluoromethyl-Phe4]-GHRP-
6 (7b and 7c) in yields and purities similar to 7a (Table 1).
Shifting to the 3-position, submonomer aza-residue syn-
thesis was performed on the Trp(Boc)-D-Phe-Lys(Boc)
sequence linked to Rink amide resin. Four aliphatic side
chains were introduced by alkylation of the aza-Gly residue
using iodomethane, allyl bromide, propargyl bromide, and
iso-propyl iodide. Not surprisingly, lower conversion to
alkylated product was detected using the latter secondary
alkyl halide. Subsequent completion of the aza-peptide
sequences gave the respective [aza-Ala3]-, [aza-allylGly3]-,
[aza-propargylGly3]-, and [aza-Val3]-GHRP-6 analogues
8a-d. Although the crude purity of aza-peptides 8 possessing
various aliphatic side chains at the 3-position were lower
than the benzyl analogues made at the 4-position, product
was isolated in acceptable yield (14-28%) and high final
purity (g90%). The final [aza-Val3]-GHRP-6 8d featured a
63:37 mixture of product contaminated with [aza-Gly3]-
GHRP-6.
(13) Endemann, G.; Stanton, L. W.; Madden, S. K.; Bryant, C. M.;
White, R. T.; Protter, A. A. J. Biol. Chem. 1993, 268 (16), 11811–11816.
(14) (a) Collot-Teixeira, S.; Martin, J.; McDermott-Roe, C.; Poston, R.;
McGregor, J. L. CardioVasc. Res. 2007, 75, 468–477. (b) Febbraio, M.;
Silverstein, R. L. Int. J. Biochem. Cell Biol. 2007, 39, 2012–2030.
(15) Momany, F. A.; Bowers, C. Y. In Growth Hormone Secretagogues;
Bercu, B. B., Walker, R. F., Ed.; Springer-Verlag: New York, 1996.
(16) (a) Fields, C. G.; Lloyd, D. H.; Macdonald, R. L.; Otteson, K. M.;
Noble, R. L. Pept. Res. 1991, 4, 95–101. (b) Lubell, W. D.; Blankenship,
J. W.; Fridkin, G.; Kaul, R. Sci. Synth. 2005, 21, 713–809.
(17) Brondani, D. J.; de Magalhaes Moreira, D. R.; de Farias, M. P. A.;
da S. Souza, F. R.; Barbosa, F. F.; Lima Leite, C. A. Tetrahedron Lett.
2007, 48, 3919–3923.
(18) Bourguet, C. B.; Sabatino, D.; Lubell, W. D. Biopolymers 2008,
90, 824–831.
(19) Koehler, K.; Sandstrom, W.; Cordes, E. H. J. Am. Chem. Soc. 1964,
86, 2413–2419.
(20) (a) Chen, F. M. F.; Kuroda, K.; Benoiton, N. L. Synthesis 1978,
928–929. (b) Heimer, E. P.; Chang, C. D.; Lambros, T. L.; Meienhofer, J.
Int. J. Pept. Protein Res. 1981, 18, 237–241.
Figure 2
. Circular dichroism spectra in water for GHRP-6 parent
peptide 10 (black line) and aza-peptide 7a (red line).
(21) Falb, E.; Yechezkel, T.; Salitra, Y.; Gilon, C. J. Pept. Res. 1999,
53, 507–517.
3652
Org. Lett., Vol. 11, No. 16, 2009

Table 2. IC50 Binding Values for GHS-R1a and CD36 Receptors
entry
sequence
IC50 binding GHS-R1a
IC50 binding CD36
7a
10
11
His-D-Trp-Ala-azaPhe-D-Phe-Lys-NH2
His-D-Trp-Ala-Trp-D-Phe-Lys-NH2
His-D-2-Me-Trp-Ala-Trp-D-Phe-Lys-NH2
2.77 × 10-6
3.65 × 10-9
M
M
1.34 × 10-6
2.97 × 10-6
M
M
At the 2-position, D-Trp was replaced by a series of aryl
aza-residues, using the general protocol. Aza-Gly was
coupled to supported Ala-Trp(Boc)-D-Phe-Lys(Boc) and
alkylated, respectively, with 2-(chloromethyl)pyridine, p-
trifluoromethyl benzyl bromide, and cinnamoyl bromide.
Completion of the SPPS protocol provided aza-peptides
9a-c in isolated yields of 20-42% and purities of g90%.
Coupling of Fmoc-His(Trt) onto the aza-residue was best
completed with triphosgene and collidine,21 albeit, 9c was
isolated as a mixture of diastereomers due to epimerization
during coupling.8a
To examine peptide conformation in water, the CD
spectrum of the parent peptide, 10, was evaluated next to
that of aza-peptide 7a (Figure 2). The insertion of an aza-
Phe motif into the GHRP-6 core had a pronounced effect
on the CD curve indicative of a change in conformation
relative to the native sequence. The CD curve for the parent
peptide was characteristic of a random coil or disordered
structure as suggested by the negative maximum band
observed at 190 nm.22 Alternatively, the CD signature for
aza-peptide 7a was indicative of an ordered ꢀ-turn conformer
with characteristic negative maximum values located at
around 190 and 230 nm and a positive maximum band near
215 nm.23
The binding affinity (IC50 values) was next evaluated for
aza-peptide 7a on the GHS-R1a and CD36 receptors. Aza-
peptide 7a was found to selectively bind to the CD36 receptor
and lost affinity for the GHS-R1a receptor in competition
binding studies (see Supporting Information for binding
curves). More specifically, aza-peptide 7a exhibited a 1000-
fold drop in binding affinity to the GHS-R1a receptor relative
to GHRP-6 (2.77 µM vs 3.65 nM) and retained comparable
binding activity with the GHRP prototype ligand, hexarelin24
(11), for the CD36 receptor (Table 2). Considering that aza-
peptide 7a induces a ꢀ-turn type conformation, then this
geometry may be responsible for binding and differentiating
the CD36 receptor from the GHS-1Ra receptor.25
An effective method for making aza-peptides without
preformation of a hydrazine moiety in solution has been
demonstrated by the solid-phase synthesis of 10 aza-
analogues of GHRP-6. Preliminary conformational and
biological analysis of [aza-Phe4]-GHRP-6 (7a) by CD
spectroscopy and receptor binding studies demonstrated a
preorganized ꢀ-turn geometry that bound discriminately for
the CD36 receptor. In light of the antiangiogenic activity of
GHRP-6 ligands that bind to the CD36 receptor, the
selectivity of 7a may be a promising lead for developing
treatments for disorders such as age-related macular degen-
eration. Considering the scope of electrophiles that may be
added to aza-Gly by this method, considerable opportunity
now exists for studying side-chain-activity relationships at
biologically active turn regions located along the peptide
backbone.
Acknowledgment. Financial Support from the Natural
Sciences and Engineering Research Council (NSERC) of
Canada, le Fond Que´be´cois de la Recherche sur la Nature
et les Technologies (FQRNT), the Canadian Institutes of
Health Research (CIHR) program for operating grants and
the CIHR Team Grant Program (funding no: CTP79848) in
G-Protein Coupled Receptor Allosteric Regulation (CTI-
GAR) is gratefully acknowledged. We are grateful to Petra
Pohankova (Université de Montréal) for performing binding
studies on GHS-R1a and CD36 receptors. The authors thank
Dr. Alexandra Fu¨rto¨s and Marie-Christine Tang (Université
de Montréal) for assistance in LCMS analyses, Dr. Luisa
Ronga (Université de Montréal) for help with CD spectros-
copy, and Dr. Tarek Kassem (Université de Montréal) for
aid in GHRP-6 synthesis.
Supporting Information Available: Experimental pro-
cedures and characterization data of compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(22) Kelly, S. M.; Jess, T. J.; Price, N. C. Biochim. Biophys. Acta 2005,
1751, 119–139.
OL901423C
(23) Bush, C. A.; Sarkar, S. K.; Kopple, K. D. Biochemistry 1978, 17,
4951–4954.
(25) Jamieson, A. G.; Boutard, N.; Beauregard, K.; Bodas, M. S.; Ong,
H.; Quiniou, C.; Chemtob, S.; Lubell, W. D. J. Am. Chem. Soc. 2009, 131,
7917–7927.
(24) Demers, A.; Mc Nicoll, N; Febbraio, M.; Servant, M.; Marleau,
S.; Silverstein, R.; Ong, H. Biochem. J. 2004, 382, 417–424.
Org. Lett., Vol. 11, No. 16, 2009
3653