Cyclic aza-peptide integrin ligand synthesis and biological activity

Article abstract of DOI:10.1021/jo300311q

Aza-peptides are obtained by replacement of the α-C-atom of one or more amino acids by a nitrogen atom in a peptide sequence. Introduction of aza-residues into peptide sequences may result in unique structural and pharmacological properties, such that aza-scanning may be used to probe structure-activity relationships. In this study, a general approach for the synthesis of cyclic aza-peptides was developed by modification of strategies for linear aza-peptide synthesis and applied in the preparation of cyclic aza-pentapeptides containing the RGD (Arg-Gly-Asp) sequence. Aza-amino acid scanning was performed on the cyclic RGD-peptide Cilengitide, cyclo[R-G-D-f-N(Me)V] 1, and its parent peptide cyclo(R-G-D-f-V) 2, potent antagonists of the αvβ3, αvβ5, and α5β1 integrin receptors, which play important roles in human tumor metastasis and tumor-induced angiogenesis. Although incorporation of the aza-residues resulted generally in a loss of binding affinity, cyclic aza-peptides containing aza-glycine retained nanomolar activity toward the αvβ3 receptor.

Full text of DOI:10.1021/jo300311q

Article
pubs.acs.org/joc
Cyclic Aza-peptide Integrin Ligand Synthesis and Biological Activity
Jochen Spiegel,†,‡ Carlos Mas-Moruno,† Horst Kessler,†,§ and William D. Lubell*,‡
†Department Chemie, Institute for Advanced Study, Technische Universitat Munchen, Lichtenbergstrasse 4, 85747 Garching,
̈
̈
Germany
‡Dep
́
artement de Chimie, Universite de Montreal, C.P. 6128, Succursale Centre Ville, Montreal, Quebec H3C 3J7, Canada
́
́
́
́
§Chemistry Department, Faculty of Science, King Abdulaziz University, P.O. Box 80203, Jeddah 21589, Saudi Arabia
S
* Supporting Information
ABSTRACT: Aza-peptides are obtained by replacement of the α-C-atom of one or more amino acids by a nitrogen atom in a
peptide sequence. Introduction of aza-residues into peptide sequences may result in unique structural and pharmacological
properties, such that aza-scanning may be used to probe structure−activity relationships. In this study, a general approach for the
synthesis of cyclic aza-peptides was developed by modification of strategies for linear aza-peptide synthesis and applied in the
preparation of cyclic aza-pentapeptides containing the RGD (Arg-Gly-Asp) sequence. Aza-amino acid scanning was performed
on the cyclic RGD-peptide Cilengitide, cyclo[R-G-D-f-N(Me)V] 1, and its parent peptide cyclo(R-G-D-f-V) 2, potent antagonists
of the αvβ3, αvβ5, and α5β1 integrin receptors, which play important roles in human tumor metastasis and tumor-induced
angiogenesis. Although incorporation of the aza-residues resulted generally in a loss of binding affinity, cyclic aza-peptides
containing aza-glycine retained nanomolar activity toward the αvβ3 receptor.
hydrazine chemistry and expand the diversity of aza-residue
side chains.25−27
INTRODUCTION
■
Aza-peptides are mimics possessing an α-carbon to nitrogen
atom replacement in the peptide sequence. Although a
configurationally labile α-center is installed by this substitution,
enhanced rigidity is also introduced into the aza-peptide
backbone as a result of the planarity of the urea moiety and
electronic repulsion between nitrogen atoms in the N,N′-diacyl
hydrazine constituent.1 Of particular note, aza-amino acids may
induce β-turn conformations within peptides, as has been
shown in a variety of theoretical,1−5 spectroscopic,3,5−7and
crystallographic8−12 studies.
Several different strategies have been explored for the
synthesis of aza-peptides featuring combinations of hydrazine
and peptide chemistry. The aza-residue has been typically
constructed from a hydrazine component and a carbonyl-
donating reagent.13,14 For example, hydrazides have been
reacted, respectively, with resin-bound isocyanate and carba-
moyl halides derived from activation of the amine terminus of a
peptide chain.15−17 Moreover, N′-alkyl carbazates, such as N-
Fmoc,18,19 N-Boc,20−22 and N-Ddz-N′-alkyhydrazines23 have
been activated using phosgene equivalents and the resulting
aza-amino acid building blocks coupled onto resin bound
peptides, with enhanced yields using microwave irradiation.24
In addition, a submonomer approach for the synthesis of aza-
peptides by alkylation of a resin-bound aza-glycine residue has
been used to circumvent the limitations of solution-phase
Aza-amino acids have been introduced into biologically active
peptides to study structure−activity relationships (SARs) as
well as to generate analogues with superior pharmacological
and pharmacokinetic properties. For example, incorporation of
aza-residues has resulted in greater stability to enzymatic
degradation, enhanced receptor affinity, and increased
selectivity compared to the parent peptide.13,14,28 Although
several linear peptides have been scanned by systematic
substitution of aza-residues for each amino acid in the
sequence,18,19,22,24,25,27 to the best of our knowledge, no cyclic
peptide has similarly been systematically scanned. Moreover,
only few aza-cyclopeptides have been reported in the
literature.29,30
Searching for a general strategy for the preparation of cyclic
aza-peptides, we have explored the synthesis of aza-derivatives
of the cyclic RGD-pentapeptide Cilengitide, cyclo[R-G-D-f-
(NMe)V] 1, and its parent counterpart cyclo(R-G-D-f-V) 2
(Figure 1). Cilengitide was discovered by a mono-N-
methylation scan of 2, from which the N-(methyl)valine
analogue exhibited subnanomolar activity and increased
selectivity for the αvβ3 integrin.31,32 Peptides 1 and 2 are
Received: February 11, 2012
Published: May 14, 2012
© 2012 American Chemical Society
5271
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
Figure 1. Cilengitide (cyclo[R-G-D-f-(NMe)V], (1)) and its parent peptide cyclo(R-G-D-f-V) (2).
a
Scheme 1. Attempted Submonomer Synthesis of an AzaF Peptide Ester
a
Reagents and conditions: (a) p-nitrophenyl chloroformate, DCM, 0 °C → rt, 2 h; (b) DIEA, DCM, rt, 16 h; (c) KOtBu, THF, 30 min; (d) benzyl
bromide, THF, 12 h; (e) NH2OH·HCl, pyridine, 60 °C, sonication.
both potent inhibitors of the αvβ3, αvβ5, and α5β1 integrins,
which play important roles in human tumor metastasis and
tumor-induced angiogenesis. Cilengitide, the first antiangio-
genic small molecule targeting the αvβ3, αvβ5, and α5β1
integrins, is currently in clinical phase II and III trials for the
treatment of glioblastoma and other cancer types.33,34 The role
of these pro-angiogenic integrins in cancer, however, is not yet
fully understood, and new molecules with improved selectivity
profiles (i.e., targeting only one integrin subtype) would
represent very promising tools to study the specific function of
integrins in malignant angiogenesis and for selective tumor
imaging applications.35
The receptor selectivity and activity of Cilengitide and
related RGD-ligands has been shown to be dependent on
backbone conformation and the relative populations of active
conformers.36,37 Aza-amino acids have already been successfully
employed in the synthesis of linear RGD-ligands, which
exhibited modifications in their receptor binding profiles.38 In
an earlier communication, replacement of aza-glycine for
glycine in cyclic aza-analogues of Cilengitide and cyclo(R-G-
D-f-V) provided analogues with nanomolar activity and varying
selectivity for the αvβ3, αvβ5, αvβ6, and αIIbβ3 integrin
receptors.29 Considering these earlier reports, a more detailed
aza-residue scan of cyclic peptides 1 and 2 was merited to
further study influences of aza-substitution on conformation,
activity, and selectivity.
C-terminal residue during the cyclization step. In addition,
biological activity is presented for the cyclic aza-peptides
toward the αvβ3 and α5β1 integrin receptors.
RESULTS AND DISCUSSION
■
Peptide Design. The strategy for the synthesis of the cyclic
aza-peptides was based on the established protocols for the
preparation of their cyclic integrin ligand counterparts.39 Side-
chain-protected linear aza-peptides were assembled on an acid-
labile solid support using the Fmoc/tBu-strategy for solid-phase
peptide synthesis (SPPS) and cleaved for subsequent
cyclization in solution and final removal of side chain
protection to yield the cyclic aza-peptides.
Particular synthetic challenges for the preparation of
cyclic,35,40,41 N-methylated,35,42 and aza-peptides14 were taken
into consideration. For example, glycine was usually positioned
at the carboxylate terminus to minimize steric hindrance and
racemization during cyclization.32 N-Fmoc-N-methyl-L-valine
was prepared by reduction of the corresponding N-Fmoc-
oxazolidinone,43 and its linkage directly to the resin was
avoided to prevent premature cleavage due to diketopiperazine
formation.44 At the N-terminal position of the linear peptide,
N-methyl and aza-amino acid residues were usually avoided,
because their amines are generally less reactive than those of
conventional amino acids and necessitate the use of demanding
coupling conditions.
Insight is now given for the design of cyclic aza-peptides in
general. Moreover, seven aza-analogues of cyclic peptides 1 and
2 were prepared possessing aza-aspartic acid (azaD), aza-glycine
(azaG), aza-phenylalanine (azaF), and aza-valine (azaV)
residues. Aza-arginine (azaR) analogues were not included in
the scope of this study. Two additional diastereomeric
analogues were isolated, resulting from epimerization of the
Aza-peptide Synthesis. In an initial attempt to circumvent
the issues of solution-phase hydrazine chemistry, we introduced
the aza-residue into the resin-bound linear peptide using a
submonomer approach (Scheme 1), which has been successful
for the synthesis of various aza-derivatives of biologically active
linear peptides.25,27 Benzylidene carbazate 4, which was
generated in situ by reaction of benzaldehyde hydrazone 3
5272
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
with p-nitrophenyl chloroformate, was linked to a resin-bound
peptide to yield semicarbazone protected aza-glycine 5.
Alkylations with benzyl bromide, isopropyl iodine, and tert-
butyl bromoacetate were used to introduce the side chains of
the azaF (6), azaV, and azaD residues, respectively. Semi-
carbazone deprotection caused, however, concomitant cleavage
of the aza-peptide from the acid-labile 2-chlorotrityl chloride
(CTC) solid support (Scheme 1). Effective conditions for
removing semicarbazones on rink-amide resin,25 such as
hydroxylamine hydrochloride in pyridine, caused aza-peptide
cleavage from the CTC resin. In the face of several failed efforts
to surmount this obstacle, the submonomer approach on CTC
resin was abandoned.
N-Fmoc-aza-amino acid chlorides were next employed to
introduce the aza-residue onto the resin-bound peptide. N′-
Alkyl-fluoren-9-ylmethyl carbazate precursors were thus
synthesized in solution and subsequently reacted with phosgene
prior to coupling onto the solid-supported peptide sequence
(Scheme 2).18,38 The appropriate carbazate building-blocks for
chloride in the presence of triethylamine to afford N-Fmoc-N-
methylcarbazate 11.38 Condensation of 11 with acetone gave
the corresponding hydrazone, which was reduced using sodium
cyanoborohydride to afford N′-isopropyl-N-fluoren-9-ylmethyl-
N-methyl carbazate 12.
Aza-amino acid residues were coupled onto the resin-
supported peptide after activation of the appropriate Fmoc-
N′-alkyl carbazate with phosgene. Coupling reactions were
performed overnight and repeated twice in the case of azaG46
to ensure full conversion as monitored by HPLC−MS
analysis.18
Although couplings of aza-residues on primary amines were
generally efficient, LC−MS analysis revealed that acylation of
N-methyl-L-valine with Fmoc-aza-phenylalanine acid chloride
gave only ca. 15% conversion after two couplings. In order to
enhance coupling yield, microwave irradiation was investigated,
because it had previously been demonstrated to minimize
reaction times and improve yields in SPPS.47−49 Furthermore,
microwave irradiation has recently been applied to enhance
coupling of activated aza-amino acids.24 A higher amount of
base was employed to prevent potential cleavage from the acid-
sensitive CTC-resin due to in situ generated HCl, without fear
of racemization during coupling of the achiral Fmoc-aza-amino
acid. Although coupling times could be reduced significantly,
conversions remained modest. Under our best conditions, a
56% overall conversion was achieved using three consecutive
coupling reactions of a 5-fold excess of activated Fmoc-aza-
amino acid in DMF at 60 °C with microwave irradiation.
Unreacted N-methyl-L-valinyl peptide was capped with acetic
anhydride, and SPPS was continued.
Scheme 2. Synthesis of Aza-amino Acid Precursors N-
Fluoren-9-ylmethyl Carbazates 8−10 and N-Fluoren-9-
a
ylmethyl-N-methyl Carbazate 12
Amino groups of aza-amino acid residues are typically less
reactive than those of natural amino acids, requiring more
vigorous coupling conditions and double couplings.14 Coupling
to aza-glycine was achieved using O-(7-azabenzotriazol-1-yl)-
N,N,N′,N′-tetramethyluronium hexafluorophosphate (HATU)
and 1-hydroxy-7-azabenzotriazole (HOAt). Coupling to more
bulky aza-residues required bis(trichloromethyl)carbonate
(BTC) as a more reactive activating reagent.18,19,24 More
basic conditions were needed to prevent peptide cleavage from
the acid-labile CTC resin during BTC couplings.50,51
Furthermore, additional care was required during activation
of the Fmoc-amino acid, and treatment of the BTC reaction
with 2,4,6-collidine was performed slowly at 0 °C to avoid
undesired side products with molecular ions corresponding to
the product mass +26.
In general, the BTC protocol provided clean couplings and
complete conversion in the acylation of the aza-amino-residue,
except for N-methyl-aza-valine. The latter case gave low
conversion with <8% of the Fmoc-phenylalanine coupled
product, even after two treatments with extended reaction
times. The combination of an aza-residue and N-methylation
was expected to require more forcing coupling conditions, due
to difficulties encountered with each moiety in individual
coupling experiments. Attempts failed to effect coupling onto
N-methyl-aza-valine using other reagents, such as PyBOP52 and
BOP-Cl,53 after extended reaction times or microwave
irradiation using excess reactant. Moreover, acetylation of the
N-methyl-aza-valine residue with a mixture of acetic anhydride
and DIEA in DMF resulted in only 15% conversion, confirming
its poor reactivity. To circumvent this issue, the synthesis of a
dipeptide building block in solution and subsequent coupling to
the resin-bound peptide may be necessary; however, such an
approach was beyond the scope of this work.
a
Reagents and conditions: (a) Fmoc-Cl, acetonitrile/water, 0 °C → rt,
14 h; (b) benzaldehyde, EtOH, at reflux, 2 h; (c) H2 (100 psi),
Pd(OH)2/C, THF, rt, 14 h; (d) acetone, at reflux, 1 h; (e) NaBH3CN,
AcOH, THF, rt, 1 h; EtOH, at reflux, 3 h; (f) tert-butyl bromoacetate,
K2CO3, DMF, rt, 20 h; (g) Fmoc-Cl, NEt3, DCM, −78 °C to rt, 16 h.
azaG, azaF and azaV were obtained as described before.18
Acylation of excess hydrazine with fluoren-9-ylmethyl chlor-
oformate afforded the azaG precursor, fluoren-9-ylmethyl
carbazate 7, in quantitative yield. Condensation of 7 with
benzaldehyde and acetone, respectively, and reduction of the
resulting Fmoc-hydrazones yielded the azaF and azaV
precursors, N′-benzyl- and N′-isopropyl-fluoren-9-ylmethyl
carbazates 8 and 9.18 Although the azaD precursor, N′-tert-
butyl-acetate-fluoren-9-ylmethyl carbazate 10, was previously
prepared by a multiple-step synthesis from an aldehyde
precursor,18 a more expedient route was explored based on
the alkylation of 7 with tert-butyl bromoacetate.23,45 Employing
7, tert-butyl bromoacetate, and K2CO3 in DMF, a mixture of
mono- and bisalkylated carbazates were obtained, from which
the former 10 was isolated in 65% yield. In the synthesis of the
N-methyl-aza-valine precursor, N′-isopropyl-N-fluoren-9-yl-
methyl-N-methyl carbazate 12, the differential reactivity of N-
methylhydrazine was exploited during acylation with Fmoc-
5273
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
a
Scheme 3. Synthesis of Cyclic Aza-peptide 21
a
Reagents and conditions: (a) Fmoc-Gly, DIEA, DCM; (b) piperidine in DMF (20%, v/v); (c) Fmoc-Arg(Pbf), HBTU, HOBt, DIEA, DMF; (d)
Fmoc-NMe-Val, HBTU, HOBt, DIEA, DMF; (e) phosgene in toluene (20 wt %), DCM, 0 °C; (f) DIEA, DMF, microwave irradiation; (g) Fmoc-
Asp(OtBu), BTC, 2,4,6-collidine, 0 °C to rt; DIEA, THF, rt; (h) TFA in DCM (1%); (j) HATU, HOAt, 2,4,6-collidine, DMF; (k) TFA/TES/H2O/
DCM (60/2.5/2.5/35, v/v/v/v).
Table 1. Synthesized Aza-peptides
a
b
c
Peptide purity ascertained by UV absorption at λ = 214 nm; see Experimental Section. After cleavage from resin. After cyclization and side chain
d
e
f
deprotection. After HPLC purification; see Experimental Section. Yields are calculated on the basis of resin loading. Product isolated as a
diastereomeric product after cyclization. After HPLC purification, the LC trace of the product exhibited inseparable overlapping peaks that were
g
attributed to interconverting conformers.
Cyclization and Deprotection. Although incorporation of
turn-inducing residues, such as N-methyl-, D-, and aza-amino
acids into the linear sequence may be expected to facilitate
cyclization due to backbone preorganization,40,41 the cyclization
step proved another challenging point of the synthesis. After
cleavage of the linear, side-chain-protected peptides from the
solid support, formation of the macrocycle was performed at
high dilution via in situ activation using diphenyl phosphorazide
(DPPA) and NaHCO3.54 Some of the peptides required
extended reaction times, up to 5 days, as well as employment of
more powerful activation conditions such as HATU, HOAt,
and 2,4,6-collidine, to ensure full conversion. Linear peptides
possessing valine (i.e, 14 and 15) as well as N-methyl-valine
(i.e., 19 and 20), respectively, at the carboxylate terminus,
epimerized during the cyclization step, providing diastereo-
meric mixtures, despite the use of the relatively mild DPPA
conditions (Table 3). Moreover, conditions for the removal of
side chain protection, such as reaction time and TFA
concentration, needed to be optimized to avoid ring opening
by hydrolysis. As previously observed employing multiple N-
methylations of Cilengitide,35 the incorporation of aza-residues
may enhance backbone rigidity favoring conformers, which may
inhibit head-to-tail cyclization, due in part to added ring strain.
5274
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
Table 2. Integrin Binding Affinity
a
n.d. = not determined.
The overall process for synthesizing a cyclic aza-peptide is
summarized for the preparation of analogue 21 (Scheme 3).
Yields, purities and HRMS characterization of all aza-peptides
are summarized in Table 1.
Biology. The affinity of the cyclic aza-peptide analogues for
the αvβ3 integrin receptor was evaluated using a competitive
solid-phase binding assay (see Experimental section for details).
From the series, only the nonmethylated azaG derivative
(peptide 13) retained the antagonistic activity for αvβ3 in the
low nanomolar range (Table 2). Peptide 18, which contains the
azaG moiety and N-methyl-Val, also showed αvβ3-binding
activity, albeit 1 order of magnitude lower than that of the
control peptide Cilengitide.29
Aza-residue substitution at other positions was not tolerated
and caused reduced or total loss of affinity for the αvβ3 integrin
receptor. Aza-amino acids may induce conformational changes
in the peptide backbone,14 which distort the bioactive αvβ3-
binding conformation32,55 causing loss of activity. In this regard,
variations in the conformation of cyclic RGD peptides have
resulted in dramatic effects in both antagonistic affinity and
selectivity profiles for different integrin subtypes.35
Selectivity profiles of selected cyclic aza-peptides were also
evaluated at the integrin α5β1 receptor. Interest in this integrin
has increased due to its crucial role in angiogenesis.34,56
Peptides 13 and 18 showed a decreased binding-activity for this
integrin receptor (Table 2); however, selectivity ratios with
respect to α5β1 and αvβ3 were similar to those found for
Cilengitide.29 Although additional structural studies are needed,
these observations may suggest a similar binding geometry for
Cilengitide and peptides 13 and 18 on interaction with the
integrin receptors.
selectivity, albeit with decreased affinity compared to
Cilengitide. A more detailed structural analysis is necessary to
elucidate the reason for the limited tolerance of 1 and 2 to the
incorporation of aza-residues. The reported strategy for the
synthesis of cyclic aza-peptides has surmounted considerable
synthetic challenges and offers thus significant potential for
application in the study of other biologically active cyclic
peptides.
EXPERIMENTAL SECTION
■
Hazards. Phosgene solution and triphosgene (BTC) are both
highly toxic and may cause death by inhalation. These substances
should be handled in a well-ventilated hood with extreme caution.
Chemistry. Fmoc-N-methyl valine,43 Fmoc-carbazates 7, 8, and
9,18 and N-Fmoc-N-methyl carbazate 1138 were synthesized as
reported.
N′-1-(tert-Butyloxycarbonylmethylamino)-N-fluoren-9-ylmethyl
Carbazate 10. To a well-stirred solution of fluoren-9-ylmethyl
carbazate 7 (2.29 g, 9.0 mmol, 1.0 equiv) and K2CO3 (1.05 g, 9.9
mmol, 1.1 equiv) in dry DMF (36 mL, 0.25 M) was added tert-butyl
bromoacetate (1.33 mL, 9.0 mmol, 1.0 equiv) dropwise at 0 °C. The
mixture was stirred at 0 °C for 5 min, allowed to warm to room
temperature, and stirred for another 20 h when dialkylation became
significant, as observed by TLC (Rf (monoalkylation) = 0.25, Rf
′
(dialkylation) = 0.58; hexane/EtOAc 1:1). The reaction mixture was
diluted with 250 mL of water and extracted with EtOAc (3 × 200 mL).
The combined organic layers were washed with water (250 mL) and
brine (250 mL), dried over MgSO4, and filtered. The volatiles were
removed under reduced pressure, and the residue was purified by flash
chromatography57 (hexane/EtOAc 1:1) on silica gel to yield 10 as a
white foam (2.12 g, 6.0 mmol, 65%): Rf = 0.25 (hexane/EtOAc 1:1);
mp 84−85 °C. 1H NMR (400 MHz, CDCl3): δ 1.75(s, 9H), 3.80 (brs,
2H), 4.42 (m, 2H), 4.69 (d, J = 6.8 Hz, 2H), 7.55 (t, J = 7.4 Hz, 2H),
7.64 (t, J = 7.4 Hz, 2H), 7.83 (d, J = 7.3 Hz, 2H), 8.00 (d, J = 7.5 Hz,
2H). 13C NMR (100 MHz, CDCl3): δ 28.1 (3C), 47.1, 53.4, 67.1,
82.0, 120.0 (2C), 125.0 (2C), 127.0 (2C), 127.7 (2C), 141.3 (2C),
143.7 (2C), 156.8, 170.7. MS (ESI): m/z = 369.2 [M + H]+.
N′-Isopropyl-N-fluoren-9-ylmethyl-N-methyl Carbazate 12. A
suspension of N-Fmoc-N-methyl carbazate 1138 (2.28 g, 8.5 mmol,
1 equiv) in acetone (25 mL, 340 mmol, 40 equiv) was heated at reflux
until complete formation of the hydrazone was observed by TLC [(1:1
hexane/EtOAc), Rf (hydrazone) = 0.28]. The mixture was
concentrated under reduced pressure. The hydrazone was dissolved
in dry THF (43 mL, 0.2 M), treated with AcOH (0.54 mL, 9.4 mmol,
1.1 equiv) and NaBH3CN (0.80 g, 12.8 mmol, 1.5 equiv), stirred at
room temperature for 2 h, when completion of the reaction was
observed by TLC [(1:1 hexane/EtOAc), Rf (hydrazone) = 0.28], and
concentrated under reduced pressure to yield a pale yellow oil that was
dissolved in EtOAc (100 mL), washed with 1 M aqueous KHSO4 (2 ×
100 mL) and brine (100 mL), and dried over MgSO4. The volatiles
were removed under reduced pressure, and the residue was dissolved
in EtOH (50 mL, 0.2 M) and heated at reflux for 3 h to hydrolyze the
aminoborane adduct. The mixture was concentrated in vacuo to yield a
yellow oil that was isolated by flash chromatography on silica gel
(hexane/EtOAc 7:3) to obtain compound 12 as a colorless, viscous oil
CONCLUSION
■
A strategy for the synthesis of cyclic aza-peptides has been
developed and employed in the aza-scan of cyclic RGD-
peptides. Linear aza-peptides were assembled on solid support
using aza-amino acid chloride building blocks, cleaved without
the removal of side chain protection and cyclized in solution.
The acid sensitive chlorotrityl solid support necessitated the
employment of mild reaction conditions for the introduction
and modification of the aza-residue. Examining the cyclic
pentapeptides Cilengitide (1) and cyclo(R-G-D-f-V) (2), seven
cyclic aza-peptides and two diastereomeric aza-analogues were
prepared and evaluated for biological activity. Decreased
affinities for the αvβ3 and α5β1 integrin receptors were
generally observed and were consistent with previous findings
demonstrating the importance of the backbone geometry for
the biological activity of 1. Only the aza-glycine analogues
retained nanomolar activity, and relative to the other aza-
analogues, cyclo[R-azaG-D-f-V] (13) exhibited better receptor
5275
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
Table 3. Synthesis of Cyclic Aza-peptides
a
b
The amino acid loaded to the resin is highlighted in gray. Conditions A and B for the introduction of the aza-residue were as described in the text.
c
d
e
The cyclization methods A and B were as described in the text. Deprotection conditions A and B were as described in the text. Product was
isolated as a side product after cylization.
1
(2.20 g, 6.6 mmol, 76%). Rf = 0.63 (hexane/EtOAc 1:1). H NMR
(400 MHz, CDCl3): δ 1.02 (m, 6H), 3.09 (brs, 3H), 3.26 (brs, 1H),
4.25 (t, J = 4.9 Hz, 1H), 4.46 (brs, 1H), 4.65 (brs, 1H), 7.32 (td, J =
7.4 Hz, J = 1.3 Hz, 2H), 7.40 (t, J = 7.3 Hz, 2H), 7.59 (d, J = 7.5 Hz,
2H), 7.77 (d, J = 7.5 Hz, 2H). 13C NMR (100 MHz, CDCl3): δ 20.8
(2C), 39.2, 47.4, 50.1, 67.6, 120.0 (2C), 124.9 (2C), 127.1 (2C), 127.8
(2C), 141.5(2C), 144.0 (2C), 157.4. MS (ESI): m/z = 311.0 [M +
H]+, 333.0 [M + Na]+. HRMS (ESI): calcd m/z for C19H22N2O2 [M +
H]+ = 311.17540, found 311.17682.
with DMF (3×), DCM (3×), and DMF (3×). Unreacted starting
material was capped by treatment with a solution of acetic anhydride
and DIEA in DMF (10/20/70, v/v/v, 1 × 10 min and 1 × 15 min).
BTC Coupling to the Aza-residue. To a well-stirred solution of
Fmoc-amino acid (3 equiv) and BTC (1.15 equiv) in dry THF, 2,4,6-
collidine (10 equiv) was slowly added dropwise at 0 °C. The resulting
suspension was allowed to warm to room temperature and stirred for
another 5 min. The resin was swollen in dry THF, treated with DIEA
(8 equiv) followed immediately by the suspension containing the
activated Fmoc-amino acid, shaken for 3−4 h, and washed with DMF
(3×) and DCM (3×).
Peptide Cleavage and Cyclization. After coupling of the last
amino acid and Fmoc-deprotection as described above, the side-chain-
protected linear peptides were cleaved off solid support using a
solution of 1% TFA in DCM (10 × 2 min). Two different methods
were used for the backbone cyclization of the linear peptides. Method
A: The linear, side-chain-protected peptide was diluted with DMF to
10−3 M. After addition of DPPA (3 equiv) and NaHCO3 (5 equiv), the
mixture was stirred until LC−MS monitoring indicated complete
consumption of the starting material. The solution was concentrated
under reduced pressure, and the cyclic peptide was precipitated by
dropwise addition into a 1:1 mixture of water/brine (v/v). The
peptide was spun down in a centrifuge, washed twice with water, and
dried under vacuum. Method B: The linear, side-chain-protected
peptide was diluted with DMF to 10−3 M. After addition of HATU
(1.5 equiv), HOAt (1.5 equiv), and collidine (1.5 equiv) the mixture
was stirred until LC−MS monitoring indicated complete consumption
of starting material. The solution was concentrated under reduced
pressure, taken up in saturated aqueous NaHCO3 solution, and
extracted with EtOAc (3×). The combined organic layers were washed
with brine (1×), dried over Na2SO4, and concentrated under reduced
pressure.
Side Chain Deprotection. Condition A: Cyclic aza-peptides were
treated with a mixture of TFA/H2O/TES/DCM (60/2.5/2.5/35, v/v/
v/v) at room temperature until complete removal of side chain
protection was indicated by ESI-MS monitoring. Condition B: In case
the monitoring indicated ring-opening hydrolysis, side chains
protection was removed using a mixture of TFA/H2O/TES/DCM
(40/2.5/2.5/55, v/v/v/v) and stirred at room temperature until ESI-
MS monitoring indicated that ring-opening hydrolysis became the
dominant reaction over removal of side chain protection.
The resulting crude peptide was precipitated into chilled diethyl
ether, spun down in a centrifuge, and washed with chilled diethyl ether
twice. After final centrifugation the peptide was dissolved in an
acetonitrile/water (1:1, v/v) solution and lyophilized. HPLC
purification yielded a white powder (Table 3).
General Procedures for Peptide Synthesis. Peptides were
prepared by Fmoc SPPS methods in an automated shaker using 2-
chlorotrityl chloride resin58 (manufacturer’s reported loading 0.8
mmol/g). The resin was swollen in dry DCM, treated with the first
amino acid (1.2 equiv) and DIEA (3.5 equiv), and shaken for 1 h. The
loaded resin was washed with DCM (3×) and DMF (3×). Fmoc
deprotection was achieved by treatment with a 20% solution of
piperidine in DMF (1 × 10 min, 1 × 15 min). Couplings of Fmoc
amino acids (3 equiv) were performed in DMF using HBTU (3 equiv)
and HOBt (3 equiv) as coupling reagent and DIEA (3 equiv) as base
(usually for 2 h). Coupling to N-methyl valine was achieved by
treatment with Fmoc amino acid (3 equiv), DIEA (6 equiv), HATU (3
equiv), and HOAt (3 equiv) (usually for 1 h). The resin was washed
after each coupling and deprotection step with alternatively DMF
(3×), DCM (3×), and DMF (3×). Complete coupling reactions were
confirmed by LC−MS monitoring. In case of unreacted starting
material, coupling reactions were repeated, and remaining free amino
groups were capped by treatment with a solution of acetic anhydride
and DIEA in DMF (10/20/70, v/v/v) (1 × 10 min and 1 × 15 min) if
necessary. Finally, the resin was washed with DMF (3×), DCM (3×),
and DMF (3×).
Introduction of Aza-residues. Fmoc-aza-amino acid chlorides
were prepared and coupled to the solid-supported peptides as
previously described.18 Briefly, a solution of phosgene in toluene (20
wt %, 2 equiv) was added dropwise to a stirred 0.1 M solution of N′-
alkyl Fmoc-carbazate 7−10 or N′-isopropyl-N-Fmoc-N-methyl carba-
zate 12 (1 equiv) in dry DCM under argon at 0 °C. The mixture was
stirred at 0 °C for 5 min and then allowed to warm up to room
temperature. After TLC indicated complete consumption of starting
material (usually after 30 min), the reaction mixture was concentrated
under reduced pressure to yield the Fmoc-aza-amino acid chloride.
Coupling method A: The Fmoc-aza-amino acid chloride (3 equiv
relative to resin loading) was dissolved in dry DCM (0.15 M), treated
with DIEA (6 equiv), stirred for 1−2 min, and transferred to the resin
swollen in dry DCM. The resin mixture was shaken overnight at room
temperature. To ensure full conversion, the resin was washed with dry
DCM (3×) and treated under the same coupling conditions. Finally,
the resin was washed with DCM (3×) and DMF (3×). Coupling
method B: The resin was transferred into a glass microwave reaction
vessel equipped with a magnetic stir bar and swollen in dry DMF.
After the addition of DIEA (15 equiv) followed by a solution of Fmoc-
aza-amino acid chloride (5 equiv) in dry DMF (0.3 M), the vessel was
sealed with a septum cap and irradiated in the microwave at constant-
temperature mode at 60 °C for 30 min. Resin was washed with dry
DMF, and coupling was repeated twice. Finally, the resin was washed
Analysis and Purification of Aza-peptides. Analytical LC−MS
analyses were performed on a Gemini 5μ C18 110A reverse-phase
column (Phenomenex Inc., Torrance, CA, USA, 150 mm × 4.6 mm, 5
μm) with a flow rate of 0.5 mL/min using linear gradients of methanol
(0.1% formic acid) in water (0.1% formic acid) or acetonitrile (0.1%
formic acid) in water (0.1% formic acid). Purification of the aza-
peptides was performed on a semipreparative column (250 mm × 21.2
mm, 5 μm, C18 Gemini column) using specified linear gradients of
5276
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
methanol (0.1% formic acid) in water (0.1% formic acid) with a flow
rate of 10.6 mL/min.
(13) Gante, J. Synthesis 1989, 405−413.
(14) Proulx, C.; Sabatino, D.; Hopewell, R.; Spiegel, J.; Garcia
Ramos, Y.; Lubell, W. D. Future Med. Chem. 2011, 3, 1139−1164.
(15) Gray, C. J.; Quibell, M.; Baggett, N.; Hammerle, T. Int. J. Pept.
Protein Res. 1992, 40, 351−362.
Competitive Solid-Phase Integrin Binding Assay. The in vitro
inhibition of integrin−extracellular matrix (ECM) protein binding was
measured in a solid-phase binding assay using soluble integrins and
coated ECM proteins, following previously described methods.59
Briefly, for the αvβ3-binding activity, flat-bottom 96-well plates were
coated at 4 °C overnight with vitronectin (1.0 μg/mL). After a
blocking step, soluble integrin αvβ3 (1.0 μg/mL) was next incubated
with a serial dilution of integrin inhibitors (10 to 0.0032 μM) for 1 h at
room temperature. Then, primary (CD51/CD61, 2.0 μg/mL) and
secondary antibodies (anti-mouse-horseradish peroxidase (HRP)
conjugate, 1.0 μg/mL) were applied each for 1 h at room temperature,
respectively. The detection of HRP was performed using a HRP
substrate (3.3.5.5′-tetramethylethylenediamine, TMB), adding 2 M
H2SO4 to stop the reaction. The absorbance of each well was recorded
at 450 nm with a POLARstar Galaxy plate reader. Every concentration
was analyzed in duplicate, and the resulting data points were fitted to a
sigmoidal curve using OriginPro 7.5G software. The turning point
describes the IC50 value. Alternatively, for α5β1, the plates were coated
with fibronectin (0.50 μg/mL), and the inhibitors were incubated with
soluble α5β1 (1.0 μg/mL). After treatment with primary (CD49e, 1.0
μg/mL) and secondary (anti-mouse-HRP, 2.0 μg/mL) antibodies, the
binding was visualized as described above. In all assays, each plate
contained Cilengitide32,34 as internal control.
(16) Liley, M.; Johnson, T. Tetrahedron Lett. 2000, 41, 3983−3985.
(17) Verhelst, S. H. L.; Witte, M. D.; Arastu-Kapur, S.; Fonovic, M.;
Bogyo, M. ChemBioChem 2006, 7, 943−950.
(18) Boeglin, D.; Lubell, W. D. J. Comb. Chem. 2005, 7, 864−878.
(19) Boeglin, D.; Xiang, Z.; Sorenson, N. B.; Wood, M. S.; Haskell-
Luevano, C.; Lubell, W. D. Chem. Biol. Drug Des. 2006, 67, 275−283.
(20) Han, H.; Janda, K. D. J. Am. Chem. Soc. 1996, 118, 2539−2544.
(21) Hansen, T. K. Tetrahedron Lett. 1999, 40, 9119−9120.
(22) Melendez, R. E.; Lubell, W. D. J. Am. Chem. Soc. 2004, 126,
6759−6764.
(23) Freeman, N. S.; Hurevich, M.; Gilon, C. Tetrahedron 2009, 65,
1737−1745.
(24) Freeman, N. S.; Tal-Gan, Y.; Klein, S.; Levitzki, A.; Gilon, C. J.
Org. Chem. 2011, 76, 3078−3085.
(25) Sabatino, D.; Proulx, C.; Klocek, S.; Bourguet, C. B.; Boeglin,
D.; Ong, H.; Lubell, W. D. Org. Lett. 2009, 11, 3650−3653.
(26) Bourguet, C. B.; Proulx, C.; Klocek, S.; Sabatino, D.; Lubell, W.
D. J. Pept. Sci. 2010, 16, 284−296.
(27) Sabatino, D.; Proulx, C.; Pohankova, P.; Ong, H.; Lubell, W. D.
J. Am. Chem. Soc. 2011, 133, 12493−12506.
(28) Zega, A. Curr. Med. Chem. 2005, 12, 589−597.
ASSOCIATED CONTENT
* Supporting Information
Characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
■
(29) Goodman, S. L.; Holzemann, G.; Sulyok, G. A. G.; Kessler, H. J.
̈
S
Med. Chem. 2002, 45, 1045−1051.
(30) Dyker, H.; Scherkenbeck, J.; Gondol, D.; Goehrt, A.; Harder, A.
J. Org. Chem. 2001, 66, 3760−3766.
(31) Aumailley, M.; Gurrath, M.; Muller, G.; Calvete, R.; Timpl, R.;
̈
AUTHOR INFORMATION
Corresponding Author
*E-mail: lubell@chimie.umontreal.ca.
Kessler, H. FEBS Lett. 1991, 291, 50−54.
■
(32) Dechantsreiter, M. A.; Planker, E.; Matha, B.; Lohof, E.;
Holzemann, G.; Jonczyk, A.; Goodman, S. L.; Kessler, H. J. Med. Chem.
̈
1999, 42, 3033−3040.
Notes
(33) Reardon, D. A.; Nabors, L. B.; Stupp, R.; Mikkelsen, T. Expert
Opin. Invest. Drugs 2008, 17, 1225−1235.
The authors declare no competing financial interest.
(34) Mas-Moruno, C.; Rechenmacher, F.; Kessler, H. Anti-Cancer
Agents Med. Chem. 2010, 10, 753−768.
ACKNOWLEDGMENTS
■
(35) Mas-Moruno, C.; Beck, J. G.; Doedens, L.; Frank, A. O.;
Marinelli, L.; Cosconati, S.; Novellino, E.; Kessler, H. Angew. Chem.,
Int. Ed. 2011, 50, 9496−9500.
This research was supported by the Natural Sciences and
Engineering Research Council of Canada (NSERC). J.S.
gratefully thanks the German Academic Alliance for his
fellowship.
(36) Gottschalk, K.-E.; Kessler, H. Angew. Chem., Int. Ed. 2002, 41,
3767−3774.
(37) Laufer, B.; Frank, A. O.; Chatterjee, J.; Neubauer, T.; Mas-
REFERENCES
■
Moruno, C.; Kummerlowe, G.; Kessler, H. Chem.Eur. J. 2010, 16,
̈
(1) Thormann, M.; Hoffmann, H. -J. J. Mol. Struct.: THEOCHEM
1999, 469, 63−76.
(2) Lee, H. J.; Ahn, I. A.; Ro, S.; Choi, K. H.; Choi, Y. S.; Lee, K. B. J.
Peptide Res. 2000, 56, 35−46.
5385−5390.
(38) Gibson, C.; Goodman, S. L.; Hahn, D.; Holzemann, G.; Kessler,
̈
H. J. Org. Chem. 1999, 64, 7388−7394.
(39) Heckmann, D.; Kessler, H. Design and Chemical Synthesis of
Integrin Ligands. In Methods in Enzymology: Integrins; Cheresh, D. A.,
Ed.; Academic Press: New York, 2007; pp 463−503.
(40) Demmer, O.; Frank, A. O.; Kessler, H. Design of Cyclic
Peptides. In Peptide and Protein Design for Biopharmaceutical
Applications; John Wiley & Sons, Ltd.: New York, 2009; pp 133−176.
(41) White, C. J.; Yudin, A. K. Nat. Chem. 2011, 3, 509−524.
(42) Chatterjee, J.; Gilon, C.; Hoffman, A.; Kessler, H. Acc. Chem.
Res. 2008, 41, 1331−1342.
(3) Lee, H. J.; Choi, K. H.; Ahn, I. A.; Ro, S.; Jang, H. G.; Choi, Y. S.;
Lee, K. B. J. Mol. Struct.: THEOCHEM 2001, 569, 43−54.
(4) Lee, H. J.; Song, J. W.; Choi, Y. S.; Park, H. M.; Lee, K. B. J. Am.
Chem. Soc. 2002, 124, 11881−11893.
(5) Lee, H. J.; Park, H. M.; Lee, K. B. Biophys. Chem. 2007, 125,
117−126.
(6) Andre, F.; Vicherat, A.; Boussard, G.; Aubry, A.; Marraud, M. J.
Pept. Res. 1997, 50, 372−381.
(7) Zouikri, M.; Vicherat, A.; Aubry, A.; Marraud, M.; Boussard, G. J.
Pept. Res. 1998, 52, 19−26.
(43) Freidinger, R. M.; Hinkle, J. S.; Perlow, D. S.; Arison, B. H. J.
Org. Chem. 1983, 48, 77−81.
(44) Teixido, M.; Albericio, F.; Giralt, E. J. Pept. Res. 2005, 65, 153−
66.
(45) Busnel, O.; Baudy-Floc’h, M. Tetrahedron Lett. 2007, 48, 5767−
5770.
(46) Bourguet, C. B.; Sabatino, D.; Lubell, W. D. Biopolymers 2008,
90, 824−831.
(47) Yu, H. M.; Chen, S. T.; Wang, K. T. J. Org. Chem. 1992, 57,
4781−4784.
(8) Benatalah, Z.; Aubry, A.; Boussard, G.; Marraud, M. Int. J. Pept.
Protein Res. 1991, 38, 603−605.
(9) Lecoq, A.; Boussard, G.; Marraud, M.; Aubry, A. Biopolymers
1993, 33, 1051−1059.
(10) Gante, J.; Krug, M.; Lauterbach, G.; Weitzel, R.; Hiller, W. J.
Pept. Sci. 1995, 1, 201−206.
(11) Marraud, M.; Aubry, A. Pept. Sci. 1996, 40, 45−83.
(12) Andre, F.; Boussard, G.; Bayeul, D.; Didierjean, C.; Aubry, A.;
Marraud, M. J. Pept. Res. 1997, 49, 556−562.
5277
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278

The Journal of Organic Chemistry
Article
(48) Loffredo, C.; Assunca̧ o, N. A.; Gerhardt, J.; Miranda, M. T. M. J.
̃
Pept. Sci. 2009, 15, 808−817.
(49) Pedersen, S. L.; Sørensen, K. K.; Jensen, K. J. Biopolymers 2010,
94, 206−212.
(50) Thern, B.; Rudolph, J.; Jung, G. Angew. Chem., Int. Ed. 2002, 41,
2307−2309.
(51) Thern, B.; Rudolph, J.; Jung, G. Tetrahedron Lett. 2002, 43,
5013−5016.
(52) Coste, J.; Le-Nguyen, D.; Castro, B. Tetrahedron Lett. 1990, 31,
205−208.
(53) Tung, R. D.; Rich, D. H. J. Am. Chem. Soc. 1985, 107, 4342−
4343.
(54) Brady, S. F.; Varga, S. L.; Freidinger, R. M.; Schwenk, D. A.;
Mendlowski, M.; Holly, F. W.; Veber, D. F. J. Org. Chem. 1979, 44,
3101−3105.
(55) Xiong, J.-P.; Stehle, T.; Zhang, R.; Joachimiak, A.; Frech, M.;
Goodman, S. L.; Arnaout, M. A. Science 2002, 296, 151−155.
(56) Kim, S.; Bell, K.; Mousa, S. A.; Varner, J. A. Am. J. Pathol. 2000,
156, 1345−1362.
(57) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923−
2925.
(58) Barlos, K.; Gatos, D.; Kallitsis, J.; Papaphotiu, G.; Sotiriu, P.;
Wenqing, Y.; Schafer, W. Tetrahedron Lett. 1989, 30, 3943−3946.
̈
(59) Frank, A. O.; Otto, E.; Mas-Moruno, C.; Schiller, H. B.;
Marinelli, L.; Cosconati, S.; Bochen, A.; Vossmeyer, D.; Zahn, G.;
Stragies, R.; Novellino, E.; Kessler, H. Angew. Chem., Int. Ed. 2010, 49,
9278−9281.
5278
dx.doi.org/10.1021/jo300311q | J. Org. Chem. 2012, 77, 5271−5278