Disulfide cyclized tripeptide analogues of angiotensin IV as potent and selective inhibitors of insulin-regulated aminopeptidase (IRAP)

Article abstract of DOI:10.1021/jm100793t

The insulin-regulated aminopeptidase (IRAP) localized in areas of the brain associated with memory and learning is emerging as a new promising therapeutic target for the treatment of memory dysfunctions. The angiotensin II metabolite angiotensin IV (Ang IV, Val¹-Tyr²-Ile³-His ⁴-Pro⁵-Phe⁶) binds with high affinity to IRAP and inhibits this aminopeptidase (K_i = 62.4 nM). Furthermore, Ang IV has been demonstrated to enhance cognition in animal models and is believed to play an important role in cognitive processes. It is herein reported that displacement of the C-terminal tripeptide His⁴-Pro ⁵-Phe⁶ with a phenylacetic acid functionality combined with a constrained macrocyclic system in the N-terminal affords potent IRAP inhibitors that are less peptidic in character than the hexapeptide Ang IV. Configurational analysis of three pairs of diastereomeric Ang IV analogues was performed using a combination of solution NMR spectroscopic methods, Monte Carlo conformational searches, and NAMFIS calculations. The compounds encompassing l-amino acids only (4, 8, and 12) showed significantly higher bioactivity compared to their lld-epimers (5, 9, and 13). The best inhibitors in the series, compounds 8 and 12, incorporating a 13- and 14-membered disulfide ring system, respectively, and both with a β³-homotyrosine residue (β³hTyr) replacing Tyr², exhibit K_i values of 3.3 and 5.2 nM, respectively.

Full text of DOI:10.1021/jm100793t

J. Med. Chem. 2010, 53, 8059–8071 8059
DOI: 10.1021/jm100793t
Disulfide Cyclized Tripeptide Analogues of Angiotensin IV as Potent and Selective Inhibitors of
Insulin-Regulated Aminopeptidase (IRAP)
†
Hanna Andersson, Heidi Demaegdt, Georges Vauquelin, Gunnar Lindeberg, Anders Karlen, Mathias Hallberg,
‡
‡
†
†
§
ꢀ
,†
ꢀ ꢀ
Mate Erdelyi, and Anders Hallberg*
ꢀ
†
Department of Medicinal Chemistry, Uppsala University, Box 574, SE-751 23 Uppsala, Sweden, ‡Department of Molecular and Biochemical
Pharmacology, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium, Department of Pharmaceutical Biosciences, Uppsala University,
Box 591, SE-751 24 Uppsala, Sweden, and Department of Chemistry, University of Gothenburg, SE-412 96 Go€teborg, Sweden
§
Received June 27, 2010
The insulin-regulated aminopeptidase (IRAP) localized in areas of the brain associated with memory and
learning is emerging as a new promising therapeutic target for the treatment of memory dysfunctions. The
angiotensin II metabolite angiotensin IV (Ang IV, Val1-Tyr2-Ile3-His4-Pro5-Phe6) binds with high affinity
to IRAP and inhibits this aminopeptidase (Ki=62.4 nM). Furthermore, Ang IV has been demonstrated
to enhance cognition in animal models and is believed to play an important role in cognitive processes. It is
herein reported that displacement of the C-terminal tripeptide His4-Pro5-Phe6 with a phenylacetic acid
functionality combined with a constrained macrocyclic system in the N-terminal affords potent IRAP
inhibitors that are less peptidic in character than the hexapeptide Ang IV. Configurational analysis of
three pairs of diastereomeric Ang IV analogues was performed using a combination of solution NMR
spectroscopic methods, Monte Carlo conformational searches, and NAMFIS calculations. The com-
pounds encompassing L-amino acids only (4, 8, and 12) showed significantly higher bioactivity compared
to their LLD-epimers (5, 9, and 13). The best inhibitors in the series, compounds 8 and 12, incorporating
a 13- and 14-membered disulfide ring system, respectively, and both with a β3-homotyrosine residue
(β3hTyr) replacing Tyr2, exhibit Ki values of 3.3 and 5.2 nM, respectively.
Introduction
the outcome of clinical studies on cholinesterase inhibitors
and NMDA antagonists4-7 nor those on nerve growth factors8
and antioxidants9-11 are encouraging. Today, considerable
effort is devoted to studying interference with the amyloid-β
processing enzymes and Aβ aggregation/oligomerization, as
well as the tau protein, various kinases, and the 5-HT6 recep-
tor.12-14 New avenues must be explored in order to enable the
development of significantly improved cognitive enhancers.
Recently published clinical data revealed that antihyperten-
sives targeting the renin-angiotensin system (RASa) induce
beneficial effects in patients suffering from cognitive impair-
ment.15 On the basis of these clinical data and a large number
of studies in animal models, the first report disclosed already
in 1988,16,17 it now seems clear that the angiotensin II metab-
olite angiotensin IV (1, Ang IV) plays an important role in
cognitive processes. Thus, the receptor for the hexapeptide
Ang IV (Val1-Tyr2-Ile3-His4-Pro5-Phe6) is emerging as a new
promising therapeutic target, and selective interaction with this
peptide receptor should provide a new approach for the treat-
ment of memory dysfunctions.15,18,19 Ang IV binds with high
affinity to the insulin-regulated aminopeptidase (IRAP, EC
3.4.11.3)20 localized in areas of the brain associated with
memory and learning.21-23 IRAP is a type II transmembrane
protein that belongs to the same family of aminopeptidases as
aminopeptidase N (AP-N)24 and is often colocalized with the
insulin-responsive glucose transporter GLUT4.25
There is an emerging demand for new drugs for the treatment
of the cognitive decline associated with Alzheimer’s disease as
well as pathological conditions such as brain trauma and cere-
bral ischemia. Alzheimer’s disease and other age-related neu-
rological diseases are expected to be even more prevalent in
the future with the increasing age of the population in most
developed countries.1,2 Apart from an NMDA antagonist,3 all
cognitive enhancers are cholinesterase inhibitors that prevent
the degradation of acetylcholine. It is well-established that the
cholinergic system is associated with learning and memory and
that patients with Alzheimer’s disease suffer from degradation
of the cholinergic neurons in the brain. Unfortunately, neither
*To whom correspondence should be addressed. Phone: þ46 18 471
4284. Fax: þ46 18 4714474. E-mail: Anders.Hallberg@orgfarm.uu.se.
a Abbreviations: Aia, 4-amino-1,2,4,5-tetrahydroindolo[2,3-c]azepin-
3-one; AMPA, 2-(aminomethyl)phenylacetic acid; Ang IV, angiotensin
IV; AP-N, aminopeptidase N; Bn, benzyl; DIC, N,N0-diisopropylcarbo-
diimide; DIEA, N,N-diisopropylethylamine; DTT, threo-1,4-dimercap-
tobutane-2,3-diol; EDTA, 2-[2-[bis(carboxymethyl)amino]ethyl(carbo-
xymethyl)amino]acetic acid; Fmoc, 9-fluorenylmethoxycarbonyl; methy-
lindole AM resin, [3-((methyl-Fmoc-amino)methyl)indol-1-yl]acetylami-
nomethyl resin; FMPB AM resin, 4-(4-formyl-3-methoxyphenoxy)-
butyrylaminomethyl resin; HATU, 1-[bis(dimethylamino)methyliumyl]-
1H-1,2,3-triazolo[4,5-b]pyridine-3-oxide hexafluorophosphate; HBTU,
3-[bis(dimethylamino)methyliumyl]-3H-benzotriazol-1-oxide hexafluor-
ophosphate;Hcy, homocysteine;HOBt, 1-hydroxybenzotriazole;β3hTyr,
β3-homotyrosine; IRAP, insulin-regulated aminopeptidase; NAMFIS,
NMR analysis of molecular flexibility in solution; RAS, renin-
angiotensin system; tBu, tert-butyl; TES, triethylsilane; TFA, trifluoro-
acetic acid; TMP, 2,4,6-trimethylpyridine; Trt, trityl.
Recently, Lukaszuk et al. reported several potent, selective,
and stable analogues of Ang IV,26,27 here exemplified by AL-
40 (Figure 1)26 where Val1 was substituted for β2-homovaline
r
2010 American Chemical Society
Published on Web 11/03/2010
pubs.acs.org/jmc

8060 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Andersson et al.
used in the majority of couplings. This method was also fre-
quently applied for coupling of homocystein (Hcy). Fmoc
deprotection was accomplished using 20% piperidine in DMF.
Completeness of the coupling reactions was usually confirmed
by small scale cleavage and LC-MS analysis. The coupling-
deprotection cycle was repeated until the desired amino acid
sequence had been synthesized. TFA/H2O in the presence of
TES and DTT was used to deprotect and cleave the inter-
mediates from the solid supports. The scavenger DTT was
used to prevent the tert-butyl (tBu) cations from reacting
with the free thiol groups of the cysteine and homocysteine
residues. After purification by RP-HPLC, the linear inter-
mediates were subjected to disulfide cyclization in DMSO/
TFA.34-36 Finally, the cyclic Ang IV analogues were purified
by preparative RP-HPLC and isolated as the corresponding
TFA salts by lyophilization.
Analogues 4, 5, 8, 9, 12, and 13 (Scheme 1) were synthe-
sized by initial attachment of 2-(azidomethyl)phenylacetic
acid29 to 2-chlorotrityl chloride resin. Subsequently, the
S-trityl protected Fmoc-Cys(Trt)-OH or Fmoc-Hcy(Trt)-OH
was coupled via a Staudinger/aza-Wittig reaction.37 Tribu-
tylphosphine was added to a mixture of resin bound azide
and activated amino acid in CH2Cl2. The remaining amino
acids were coupled as described above.
Initially, compound 4 was synthesized on a 100 μmol scale.
At scale-up to 300 μmol extensive epimerization was observed
in the cyclization step using DMSO/TFA.34 The chirality of
the LLL-isomers (4, 8, and 12) and the corresponding epimers
(5, 9, and 13), was elucidated on the basis of comparative
1
HPLC and NMR analysis (chromatograms and H NMR
spectra are in the Supporting Information). Further assign-
ment of the absolute configuration of the three diastereomeric
pairs was established by a combined NMR spectroscopic and
computational study.
Figure 1. Previously reported IRAP inhibitors AL-4026 and HFI-
43728 together with a novel inhibitor discussed in this paper.
(β2hVal) and the His4-Pro5 dipeptide was replaced by 4-ami-
no-1,2,4,5-tetrahydroindolo[2,3-c]azepin-3-one (Aia)-Gly. In
2008, a major step toward bioavailable IRAP inhibitors was
taken by Albiston et al.28 A series of potent druglike IRAP
inhibitors, exemplified by HFI-437 (Figure 1), were identified
by the Australian group utilizing homology modeling of the
catalytic domain of the protein and in silico screening meth-
odologies. It was demonstrated in rat that the performance in
both spatial working and recognition memory paradigms
could be improved after administration of the inhibitor into
the lateral ventricles.28 We have applied the opposite strategy,
which relies on introducing conformational constraints in Ang
IV as the first objective.29-31 Subsequent iterative structural
modifications are expected to provide bioavailable nonpep-
tidic inhibitors with enhanced selectivity. We here show that
cyclization of the N-terminal end of Ang IV truncated in the
C-terminal can provide potent IRAP inhibitors that do not
interfere with AP-N, as exemplified by compound 8 (Figure 1).
For the synthesis of analogues 6, 10, and 11 (Scheme 2), the
appropriate amine was attached to the FMPB AM resin through
reductive amination with NaBH(OAc)3 in 1% AcOH/DMF
under microwave heating at 60 °C for 20 min.38 The resulting
resin-bound secondary amine was then acylated with Fmoc-
Cys(Trt)-OH or Fmoc-Hcy(Trt)-OH overnight using HATU/
TMP in CH2Cl2/DMF. The first and second steps were moni-
tored by colorimetric tests using the starting resins as refer-
ences. The presence of CHO groups was evaluated using the
p-anisaldehyde test,39 and the existence of secondary amines
was examined using the chloranil test.40,41 Standard cou-
plings described above were used for the second and third
amino acid residues. Generally, the oxidative cyclizations
were performed under acidic conditions using DMSO/TFA.
However, the cyclization of 11 was successfully performed at
pH 6 in aqueous AcOH/(NH4)2CO3/DMSO.36
Methylindole AM resin was used in the synthesis of 7
(Scheme 3) starting with Fmoc deprotection and reactionwith
Fmoc-Hcy(Trt)-OH overnight. Standard couplings were then
carried out for the Fmoc-Tyr(tBu)-OH and Fmoc-Hcy(Trt)-
OH residues. Disulfide cyclization was accomplished using
DMSO/TFA.
Configurational Analysis. Small flexible peptides are com-
monly present in solution as rapidly equilibrating mixtures of
low-energy conformations, which cannot be accurately rep-
resented by a single structure as conventionally derived by
experimentally restrained structure calculations (constrained
simulated annealing or restrained molecular dynamics) for
proteins.42 Methods for the deconvolution of time-averaged
NMR variables, such as NOEs and scalar couplings, into
Results and Discussion
Chemistry. The syntheses of 2 and 3 have been reported
previously.31 Compounds 4-13 in Table 1 were prepared by
manual SPPS using the Fmoc protection strategy32 followed
by oxidative cyclization in solution according to the synthetic
routes outlined in Schemes 1-3.
Generally, the amino acids were single-coupled using
HATU and DIEA in DMF. Under these conditions cysteine
is prone to racemization. Consequently, the coupling reagent,
base, and solvents must be chosen carefully.33 Here, HATU,
TMP, and CH2Cl2/DMF, with a minimum of DMF, were

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8061
Table 1. Stability and Inhibition Activities of Compounds 1-13d

8062 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Table 1. Continued
Andersson et al.
a [3H]AL-11 competition binding in CHO-K1 cell membranes.63 b Evaluated in an enzyme assay comprising recombinant human IRAP, transiently
transfected in HEK293 cells. c Evaluated in an enzyme assay comprising recombinant human AP-N, transiently transfected in HEK293 cells. d SD,
standard deviation.
structural families representing the solution ensemble and
fulfilling all structural restraints are available but still seldom
utilized.43-49 In principle, all such techniques generate theo-
retical ensembles for compounds of known constitution and
deconvolute the thermally averaged NMR data to distinct
families of conformations. When these conformations are
weighted with their probabilities corresponding to their molar
fractions, the ensemble fits all experimentally observed dis-
tances and dihedrals optimally. For structure elucidation of
the diastereomeric pairs 4 and 5, 8 and 9, and 12 and 13,
a combined computational and spectroscopic approach was
applied, which takes conformational averaging into consid-
eration by ensemble analysis.
comparable chemical properties. Whereas one of the products
in each set has the chirality of the starting materials, i.e.,
L-Hcy1, L-Tyr2/L-β3hTyr2, and L-Cys3/L-Hcy3, here denoted
LLL (reflecting the configurations of the amino acids in posi-
tions 1, 2, and 3), stereomutation of one chiral center of the
other isolated compound appears most plausible. Notably,
more extensive isomerization wouldbeexpected toyieldseveral
diastereomeric entities, which was not observed. As described
above, a comparison of HPLC retention times and NMR data
(Supporting Information Figures S1-S3) of the compounds of
unknown chirality with those derived from epimerization-free
synthesis (4, 6, 7, and 10) indicated an LLL-configuration for
compounds 4, 8, and 12, whereas epimerization was predicted
for 5, 9, and13. Assignment of the absolute configuration of the
three diastereomeric pairs, confirming the above-described
simplified preliminary analysis, was established by a thorough
Following the reaction route shown in Scheme 1, the
synthesis of 4, 8, and 12 generated three sets of closely related
product pairs giving highly similar NMR spectra and showing

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8063
Scheme 1. Synthesis of 4, 5, 8, 9, 12, and 13 Using 2-Chlorotrityl Chloride Resina
a Reagents and conditions: (a) (i) 2-(azidomethyl)phenylacetic acid, DIEA, CH2Cl2, room temp, 4 h, (ii) MeOH, 15 min; (b) (i) Fmoc-Cys(Trt)-OH or
Fmoc-Hcy(Trt)-OH, DIC, HOBt, CH2Cl2, (ii) PBu3, room temp, on, (iii) 20% piperidine/DMF; (c) (i) Fmoc-Tyr(tBu)-OH or Fmoc-β3hTyr(tBu)-OH,
HATU, DIEA, DMF, room temp, 2 h, (ii) 20% piperidine/DMF; (d) (i) Fmoc-Hcy(Trt)-OH, HATU, TMP, CH2Cl2/DMF, room temp, 2 h,
(ii) 20% piperidine/DMF; (e) DTT, TFA/TES/H2O (88:5:7), room temp, 2 h; (f) 10% DMSO/TFA, 0 °C to room temp, 6-26 h.
Scheme 2. Synthesis of 6, 10, and 11 Using FMPB AM Resina
a Reagents and conditions: (a) benzylamine or pyridin-3-ylmethanamine (3-picolylamine) (10 equiv), NaBH(OAc)3 (10 equiv), AcOH/DMF (1:99),
microwave, 60 °C, 20 min; (b) (i) Fmoc-Cys(Trt)-OH or Fmoc-Hcy(Trt)-OH, HATU, TMP, CH2Cl2/DMF, room temp, on, (ii) 20% piperidine/DMF;
(c) (i) Fmoc-Tyr(tBu)-OH or Fmoc-β3hTyr(tBu)-OH, HATU, DIEA, DMF, room temp, 2 h, (ii) 20% piperidine/DMF; (d) (i) Fmoc-Hcy(Trt)-OH,
HATU, TMP, CH2Cl2/DMF, room temp, 2 h, (ii) 20% piperidine/DMF; (e) DTT, TFA/TES/H2O (88:5:7), room temp, 2 h; (f) 10% DMSO/TFA, 0 °C
to room temp, 6-26 h; (g) (i) AcOH (5% aq), (NH4)2CO3 (0.5 M aq), pH 6, (ii) DMSO, 0 °C to room temp, on.
NMR spectroscopic study. The applied method is discussed in
detail here for the stereoisomeric pair 12 and 13 and summar-
ized for the pair 4 and 5 and the pair 8 and 9 in the Supporting
Information.
stereoisomers allow their differentiation by applying NOE-
based distance and scalar coupling-derived dihedral angle
measurements.
Configurations were assigned through deconvolution of
the experimentally acquired NOEs and 3JHH coupling constants
into computationally predicted ensembles of the four theoreti-
cally possible diastereomers of 12 and comparison of the quality
of the fits between the experimental and the computed data of
the various stereoisomers. In order to ensure the best possible
coverage of the conformational space,50 we also performed
a second Monte Carlo search using the Amber* force field,
yielding a total set of 700-950 distinct conformations for
each theoretical stereoisomer (LLL, DLL, LDL, LLD).
Subsequently, for each diastereomer, the set of conforma-
tions best fitting the experimental data (NOE and 3JHH) was
derived using the NAMFIS program,47 and the probabilities
of the conformations were predicted. The root-mean-square
deviations (rmsd) of the back-calculated intramolecular dis-
tances and J couplings of the probability-weighted averages
of the selected conformational ensembles, compared to
the observed experimental NMR data, were evaluated for
The feasible conformational populations of the theoretically
possible LLL, DLL, LDL, and LLD diastereomers of cyclo[Hcy-
β3hTyr-Hcy]-2-(aminomethyl)phenylacetic acid (AMPA)
were predicted by restraint-free systematic Monte Carlo con-
formational searches (SPMC, 10 000 steps). Throughout these
calculations the OPLS-2005 force field and the GB/SA water
solvent model were employed, as implemented in the program
MacroModel (version 9.0.211), yielding sets of ∼250 con-
formations for each configuration within 10 kcal/mol of the
global minimum. The complete set of conformations of c[Hcy-
β3hTyr-Hcy]-AMPA resulting from the unrestrained con-
formational search revealed an exceptionally high degree
of rigidity of the 14-membered macrocycle (Figure 2). The low
flexibility of the cyclic peptide results in a very limited number
of possible steric arrangements of the central core of its
four theoretically obtainable epimers (Figure 3). Significant
differences in the CHR hydrogen orientation of the possible

8064 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Scheme 3. Synthesis of 7 Using Methylindole AM Resina
Andersson et al.
a Reagents and conditions: (a) (i) 20% piperidine/DMF, (ii) Fmoc-Hcy(Trt)-OH, HATU, TMP, CH2Cl2/DMF, room temp, on, (iii) 20% piperidine/
DMF; (b) (i) Fmoc-Tyr(tBu)-OH, HATU, DIEA, DMF, room temp, 2 h, (ii) 20% piperidine/DMF; (c) (i) Fmoc-Hcy(Trt)-OH, HATU, TMP, CH2Cl2/
DMF, room temp, 2 h, (ii) 20% piperidine/DMF; (d) DTT, TFA/TES/H2O (88:5:7), room temp, 2 h; (e) 10% DMSO/TFA, 0 °C to room temp, 23 h.
Figure 2. An overlay of the backbones of all (228) conformations
of c[Hcy-β3hTyr-Hcy]-AMPA, within 10 kcal/mol (42 kJ/mol) of
the global minimum generated by a restraint-free systematic Monte
Carlo conformational search, is shown above, whereas a magnifica-
tion of the backbone atoms of Hcy-β3hTyr-Hcy is depicted below.
Side chain atoms are omitted for clarity. A remarkable rigidity of
the central cyclic core is indicated, as well as well-defined orienta-
tions of the CHR and amide NH protons, despite some flexibility in
Figure 3. Overlay of the backbones of the native LLL-configuration
the disulfide region.
and the lowest energy conformations of the possible epimers of
c[Hcy-β3hTyr-Hcy]-AMPA (compounds 12 and 13, as depicted in
identification of the best fitting configuration. The goodness
of fit was expressed as the sum of the square differences (SSD)
between the measured and modeled variables (a lower SSD
reflects a better fit), as previously described by Snyder et al.42
Sets of 37 and 18 critical NOEs were used, together with 16
Table 1). The color coding is as follows: LLL pink, DLL orange, LDL
blue, and LLD green. The overlaps indicate a highlysimilar orientation
of the central core of the diastereomers. Discrimination between the
isomers is only possible by differentiation of the orientation of Hcy1-
HR for the DLL-LLL pair, β3hTyr2-Hβ for the LDL-LLL pair, and Hcy3-
HR for the LLD-LLL pair.
3
and 8 JHH values, for compounds 12 and 13, respectively
(Supporting Information pp S13-S14). The different num-
bers of experimental parameters included in the calculations
reflect the larger degree of signal overlaps for compound 13,
allowing the determination of a larger number of precise
distance and dihedral constraints for 12 compared to 13. To
minimize the uncertainty of the structure determination of the
main ring, NOEs and 3JHH values originating from the com-
parably more flexible AMPA residue and the side chain of
β3hTyr2 (Figure 2) were excluded from this analysis. The best fit
between the NMR data and the ensemble selected by NAMFIS
was obtained for the LLL-configuration of compound 12 (SSD
values as follows: LLL=1.28; LLD = 7.24; LDL = 1.84; DLL =
3.33) and the LLD-configuration of 13 (LLL = 20.7; LLD = 14.0;
LDL = 26.6; DLL = 42.4),86 which fulfilled the experimental re-
straints without any significant deviation (Supporting Informa-
tion pp S13-S14). Further investigations using the NAMFIS
program allowed the determination of the configuration of
compounds 4 and 5 and compounds 8 and 9 (4, LLL; 5, LLD; 8,
LLL; 9, LLD). This finding was further confirmed by small scale,
epimerization-free preparation of analogue 4.

Article
We emphasize that NAMFIS analysis was originally devel-
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8065
oped for the prediction of the available conformational space
of compounds of known configuration. Its application to the
elucidation of the configuration of Ang IV analogues is novel,
and was possible because of the unusual rigidity of the central
tripeptidic core of the investigated peptides, and cannot be
generalized. The applied method, similar to other available
NMR techniques, enables only the determination of the abso-
lute configuration of a stereocenter in case the chirality of at
least one additional chiral center in the proximity is known.
Previously published NMR methods for the determination
of the relative configuration of organic molecules use the
combination of H-H scalar coupling constants and local
NOEs; however, they require at least one additional param-
eter such as two- or three-bond H-C scalar couplings51,52
and/or residual dipolar couplings.53 Methods for the determi-
nation of configurations based on residual dipolar couplings
only are also available.54-56 A common feature of these
techniques is the requirement of comparably large amounts
(preferably tenths of milligram) of investigated substance,
since they necessitate the observation of heteronuclear cou-
pling constants. To us only 0.4-4 mg samples of the discussed
compounds were available for NMR analysis and such meth-
ods were therefore unsuitable.
Figure 4. Overlaid backbones of the low energy conformations of
the bioactive 4 (orange), 8 (purple), and 12 (green) indicate compa-
rable orientation of the functionalities that may be involved in the
molecular recognition process responsible for their IRAP inhibitory
activity. Enlargement of the carbon ring by one atom does not
significantly alter the location of the amino acid side chains or that
of the attached AMPA residue.
itself (1, Ki=62.4 nM). Thus, a bioactive conformation can
be adopted after cyclization, but a ring system of greater
conformational flexibility than Cys1/Cys3 seems to be re-
quired for appropriate spatial orientation of the side chains
in the N-terminal of Ang IV.31
The fact that extensive epimerization was observed only
for the compounds containing AMPA in the C-terminal
suggests that the carboxyl group might be involved in inter-
actions favoring epimerization in position 3.
Biochemical Evaluation. The ability of 1-13 to inhibit the
catalytic activity of recombinant human IRAP and amino-
peptidase N (AP-N) transiently transfected in HEK293 cells
was examined and compared. Furthermore, the stability of
1-6 and 8-13 toward degradation by metallopeptidases in
membrane homogenates was investigated. To this end, mem-
branes of endogenous IRAP-containing CHO-K1 cells were
preincubated with different concentrations of compounds in
either the absence or presence of chelators and then further
incubated with a radioligand in the presence of chelators.
Binding affinities in the continuous presence of chelators
refer to the IRAP apo-enzyme.57,58 When preincubation
proceeds in the absence of chelators, the competition curve
will shift to the right if the compound is rapidly degraded by
IRAP and/or other metallopeptidases. The binding affinities
and IRAP activity inhibition data of the compounds are
presented in Table 1.
It is well accepted that wide structural variations can be
performed in the C-terminal tripeptide of Ang IV with re-
tained biological activity.59-62 We previously reported that
replacement of His4-Pro5-Phe6 in Ang IV with an AMPA
moiety rendered a ligand with an equipotent binding affinity
and IRAP inhibitory activity to Ang IV.30 Conversely, small
variations in the N-terminal region had considerable effects
on the activity. We imposed steric constraints at the N-term-
inals via disulfide cyclizations in order to identify bioactive
conformations of Ang IV.31 An 11-membered ring system
was created by the cyclization of Cys1 with Cys3 in an Ang
IV analogue where Val1 and Ile3 had been displaced. This
maneuver was found to be deleterious to the activity (2, Ki =
16 900 nM). Incorporation of a simpler one-residue back-
bone mimetic, i.e., replacement of the Tyr2 residue in Ang IV
by a 4-hydroxydiphenylmethane scaffold, was also nonpro-
ductive.29 However, formation of a 13-membered ring sys-
tem via an Hcy1/Hcy3 cyclization afforded an IRAP inhibi-
tor (3, Ki=303 nM), albeit 5 times less potent than Ang IV
As can be seen in Table 1, compound 4, comprising a
13-membered disulfide ring system and a C-terminal 2-(amino-
methyl)phenylacetic acid group, is a more efficient IRAP
inhibitor (Ki =22.5 nM) than Ang IV and exhibits higher
selectivity over AP-N. The large difference between the Ki
values obtained from the binding assays, with and without
chelators, suggests that 4 is rapidly degraded in the CHO-K1
cell membranes just as Ang IV. The chelators serve to block
the proteolytic activity of IRAP, AP-N, and other metallo-
peptidases.58,64 Epimer 5 is 10 times weaker as an inhibitor
of IRAP than 4 and exhibits lower selectivity over AP-N. As
shown in previous studies, a C-terminal carboxyl group is
preferred (cf. 4 and 6).30 Also, removal of the aromatic
moiety in the C-terminal part is deleterious to the activity
(cf. 4 and 7). Retaining the ring size but replacing Hcy3 by
Cys and Tyr2 by a β-amino acid residue renders an alter-
native 13-membered disulfide ring system. This modification
led to a very potent IRAP inhibitor, compound 8 (Ki =
3.3 nM), which was 10 times more potent than 4 and 20 times
more potent than Ang IV itself. The epimer 9 is considerably
less active. As was the case with 4, the removal of the car-
boxyl group from 8 results in a less active inhibitor (10, Ki =
18.5 nM), but this macrocyclic disulfide is still 3 times more
potent than the parent compound Ang IV. Contrary to the
carboxyl analogue 8, compound 10 is essentially devoid of all
AP-N inhibitory activity. There are previous reports of linear
Ang IV analogues containing nitrogen heteroaromatics (e.g.,
2-quinolyl and 4-pyridyl) in the C-terminal end, with good
affinity to the Ang IV binding sites (AT4 receptor/IRAP).59,60
However, as reported here, the replacement of the phenyl
moiety in 10 by 3-pyridyl in the C-terminal, giving 11, resulted
in a less effective inhibitor of IRAP which exhibited low
selectivity. A comparison of 4 or 8 with inhibitor 12 reveals
the possibility of ring enlargement without losing IRAP inhib-
iting capacity (Figure 4). Epimer 13 possess lower activity

8066 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Andersson et al.
also in this case. The NMR-based configuration analysis indi-
cated that 5, 9, and 13 were epimerized at the amino acid
residue in position 3, revealing the high importance of the
configuration in this position for IRAP activity because the
LLD epimers are considerably less active thantheircorrespond-
ing LLL epimers.
Experimental Section
General Information. Microwave heated reactions were per-
formed in a Smith Synthesizer (Biotage AB, Uppsala, Sweden)
producing controlled irradiation at 2450 MHz. Analytical RP-
HPLC-MS was performed on a Gilson HPLC system with a
Finnigan AQA quadrupol mass spectrometer using an Onyx mono-
lithic C18 column (4.6 mm ꢀ 50 mm) with gradients of CH3CN/
H2O (0.05% HCOOH) or MeOH/H2O (0.05% HCOOH) at a flow
rate of 4 mL/min. Both UV (DAD, 214 and 254 nm) and MS (ESI)
detection were utilized. Preparative RP-HPLC was performed using
a Zorbax SB-C8 column (5 μm, 21.2 mm ꢀ 150 mm) with gra-
dients of CH3CN/H2O (0.1% TFA) at a flow rate of 5 mL/min
and UV detection at 230 nm. The purity of the compounds was
determined on a Zorbax SB-C8 (5 μm, 4.6 mm ꢀ 50 mm) and an
Allure Biphenyl column (5 μm, 4.6 mm ꢀ 50 mm) using the same
buffer systems at a flow rate of 2 mL/min and with UV detection
at 220 nm. Opticalrotationswere obtained on a Perkin-Elmer 241
polarimeter. Specific rotations ([R]D) are reported in deg/dm, and
the concentration (c) is given as g/100 mL in the specified solvent.
NMR spectra were recorded on a Varian Mercury Plus spectrom-
eter (1H at 400 MHz and 13C at 101 MHz) at ambient tempera-
ture. Chemical shifts (δ) are reported in ppm referenced indirectly
to TMS via the 2H lock signal. Assignment was made by gCOSY,
gHSQC, and gHMBC experiments. Exact molecular masses were
determined on a Micromass Q-Tof2 mass spectrometer equipped
with an electrospray ion source at the Department of Pharma-
ceutical Biosciences, Uppsala University, Sweden. Elemental
analysis was performed at Mikrokemi AB, Uppsala, Sweden.
The manual solid-phase synthesis was performed in disposable
syringes fitted with porous polyethylene filters, and a Stuart SB2
tube rotator was used for agitation. The resins were purchased
from Novabiochem. Dichloromethane used in the reactions was
distilled over calcium hydride. Other chemicals were purchased
and used without further purification. Final product yields were
determinedon the basisof the initial loading of the resin. The final
compounds were more than 97% pure as assessed from analytical
RP-HPLC analysis. The synthesis of 2-(azidomethyl)phenyl-
acetic acid29 and analogues 1-331 has been described previously.
General Procedure A: Attachment of 2-(Azidomethyl)phenyla-
cetic Acid to 2-Chlorotrityl Chloride Resin. 2-(Azidomethyl)-
phenylacetic acid29 (1.0 equiv) was reacted with 2-chlorotrityl
chloride resin (1.5 equiv) in CH2Cl2 in the presence of DIEA
(4.0 equiv). After 5-7 h, MeOH was added and the mixture
was stirred for at least 15 min before the resin was transferred to
a 5 mL disposable syringe fitted with a porous polyethylene
filter, washed with several portions of, in turn, CH2Cl2, DMF,
MeOH, and CH2Cl2, and dried in vacuo.
General Procedure B: Reductive Amination of FMPB AM
Resin. FMPB AM resin, NaBH(OAc)3 (10 equiv) and DMF
were added to a 10-20 mL Biotage microwave vial containing
a stirring bar. After being sealed, the vial was flushed with nitro-
gen gas and the appropriate amine (10 equiv) and AcOH (cat.)
were added. The reaction mixture was heated to 60 °C by
microwave irradiation for 20 min, cooled to room temperature,
and transferred to a 5 mL disposable syringe fitted with a porous
polyethylene filter. The resin was washed with several portions
of, in turn, DMF, MeOH, and CH2Cl2 and dried in vacuo. Two
colorimetric tests were employed to examine the reaction: the
p-anisaldehyde test to detect the presence of CHO groups39 and
the chloranil test40,41 to detect the presence of secondary amines.
The FMPB AM resin was used as a reference.
General Procedure C: Fmoc Deprotection and Coupling to
Methylindole AM Resin. Methylindole AM resin was weighed
into a 5 mL disposable syringe fitted with a porous polyethylene
filter and swollen in CH2Cl2 for approximately 1 h. The solvent
was replaced by DMF, and the Fmoc group was removed by
treatment with 20% piperidine in DMF (2 ꢀ 3 mL) followed by
repeated rinsing with DMF. The liberated secondary amine was
coupled with the appropriate amino acid (2.0 equiv) overnight in
All of the IRAP inhibitors 2-4, 8, 10, and 12 exhibited low
metabolic stability. Thus, cyclization did not make the
N-terminal peptide bond resistant to proteolysis by IRAP
and/or other metallopeptidases present in the CHO-K1 cell
membranes. This observation is not unexpected, since vaso-
pressin and oxytocin, both substrates for IRAP, contain an
N-terminal disulfide-bridged cyclic hexapeptide and both
peptides are subjected to N-terminal sequential degradation
by IRAP.65-67 The incorporation of a β-amino acid residue
to replace Tyr2 (i.e., β3hTyr) did not improved the proteo-
lytic stability. However, as shown in the example of the
metabolically stable β2hVal1-Ang IV by Lukaszuk et al.,27
modification of the N-terminal amino acid residue signifi-
cantly improves proteolytic stability. As an IRAP inhibitor,
β2hVal1-Ang IV was slightly less efficient (Ki = 100 nM). The
hexapeptide Val-β3hTyr-Ile-His-Pro-Phe was also found to be
less potent than Ang IV (Ki = 173 nM).27 The replacement of
Tyr2 in this Ang IV analogue27 with a β-homoamino acid
residue and most other modifications centered around Tyr2
provide less potent peptides/pseudopeptides.29,30 It is therefore
notable that the cyclized disulfides 8, 10, and 12, all comprising
β3hTyr2, are among the most potent IRAP inhibitors reported
to date. As previously deduced from theoretical conforma-
tional analysis and 1H NMR spectroscopy, oxidative cycliza-
tion of Cys3-Tyr4-Cys5-Ang II creates an 11-membered ring
system prone to adopt γ-turns.68 In contrast, Hcy3-Tyr4-Hcy5-
Ang II forms a 13-membered ring system comprising a disul-
fide unit with a large number of low-energy conformations
including one γ-turn and one β-turn conformation, but no
preferred well-defined conformation was identified.68 Since the
corresponding Cys1-Tyr2-Cys3-Ang IV 2 is essentially inactive
as IRAP inhibitor,31 it thus appears that a γ-turn is not
adopted upon interaction with IRAP. Further data acquisition
is needed in order to elucidate the conformation optimal for
IRAP inhibition. Coordination to the catalytic zinc ion in
IRAP by elements in the N-terminal is expected to have a
pronounced effect on the conformation and the binding mode
of conformationally flexible inhibitors.67,69-73
Conclusion
We have demonstrated that oxidative cyclization of the Ang
IV analogue Hcy-Tyr-Hcy-His-Pro-Phe provides an IRAP
inhibitor 3 (Ki = 303 nM) 5 times less potent than Ang IV,
whereas replacement of the C-terminal tripeptide His-Pro-Phe
of 3 by a phenylacetic acid moiety resulted in a 2-fold increase
in inhibitory activity (4, Ki = 22.5 nM). Moreover, further
increased potency was obtained by replacing Tyr2 in 4 with
β3hTyr and Hcy3 with Cys, followed by cyclization; compound
12 with a 14-membered ring system exhibits a Ki of 5.2 nM, and
compound 8 encompassing a 13-membered ring system ex-
hibits a Ki of 3.3 nM. A pronounced impact of the spatial
arrangement of the functional groups at the C-terminal end on
IRAP inhibitory activity was revealed by the lower potency of
the corresponding LLD epimers 13 and 9 compared to Ang IV.
Compounds 8 and 12 are among the most potent IRAP
inhibitor reported to date.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8067
CH2Cl2/DMF 4:1 using HATU (2.0 equiv) in the presence of
TMP (4.0 equiv). After the resin was washed with DMF, the
Fmoc group was removed as described above. Subsequent to
DMF, several portions of MeOH and CH2Cl2 were used to rinse
the resin which was then dried in vacuo or immediately used in
the following step. The chloranil test was used to monitor the
reaction.40,41 The Fmoc deprotected methylindole AM resin was
used as a reference.
at 0 °C for 10 min and then at room temperature until HPLC
analysis showed oxidation to be complete. The reaction mixture
was diluted 2-fold with CH3CN/H2O (0.1% TFA), filtered
through a 0.45 μm nylon membrane, and purified with RP-
HPLC. Selected fractions were analyzed by RP-HPLC and RP-
HPLC-MS, and those containing pure product were pooled
and lyophilized to give the product as the TFA salt.
c[Hcy-Tyr-Hcy]AMPA (4) and c[Hcy-Tyr-D-Hcy]AMPA (5).
Solid-phase attachment was performed according to general
procedureA using 2-(azidomethyl)phenylaceticacid (122 mg, 636
μmol), 2-chlorotrityl chloride resin (683 mg, 962 μmol), DIEA
(443 μL, 2.54 mmol), and CH2Cl2 (6.0 mL). MeOH (1.0 mL) was
added after 7 h, and stirring was continued for 40 min. Yield: 738
mg. Part of the resin (352 mg, 303 μmol) was then reacted with
Fmoc-Hcy(Trt)-OH (268 mg, 447 μmol), as described in proce-
dure D, using HOBt (63.4 mg, 469 μmol), DIC (67 μL, 430 μmol),
85% PBu3 (75 μL, 430 μmol), and CH2Cl2 (10 mL). The sub-
sequent coupling of Fmoc-Tyr(tBu)-OH (401 mg, 873 μmol) was
performed with HATU (338 mg, 889 μmol) and DIEA (300 μL,
1.72 mmol) in DMF (3 mL) following general method F. General
procedure E was applied for the coupling of the N-terminal
Fmoc-Hcy(Trt)-OH (250 mg, 418 μmol) residue using HATU
(165 mg, 433 μmol) and TMP (114 μL, 861 μmol) in CH2Cl2
(2.4 mL) andDMF(1.0 mL). Deprotection andcleavage from the
resin following general procedure G yielded the linear sequence
(46.3 mg, 68.4 μmol), which was subjected to disulfide bridge
formation according to general method H using TFA (15.4 mL)
and DMSO (2.44 mL). The last step gave a mixture of two
epimers, 4 and 5, which could be separated by RP-HPLC.
c[Hcy-Tyr-Hcy]AMPA (4). The product was isolated as the
General Procedure D: Coupling to 2-(Azidomethyl)pheny-
lacetyl-2-chlorotrityl Resin. A round-bottomed flask was charged
with the appropriate amino acid (3.0 equiv), HOBt (3.0 equiv),
and DIC (3.0 equiv) in dry CH2Cl2. After preactivation under
nitrogen for 10 min, 2-(azidomethyl)phenylacetyl-2-chlorotrityl
resin (1.0 equiv) was added, followed by PBu3 (2.0 equiv) 10 min
later. The reaction mixture was stirred under nitrogen for 6 h and
then transferred to a 5 mL disposable syringe fitted with a porous
polyethylene filter. The resin was washed with several portions of,
in turn, CH2Cl2, DMF, MeOH, and CH2Cl2 and dried in vacuo.
General Procedure E: Coupling and Deprotection of Cysteine
and Homocysteine Residues. Unless otherwise stated, a solution
of Fmoc-Cys(Trt)-OH (1.5 equiv), HATU (1.5 equiv), and TMP
(3.0 equiv) in CH2Cl2/DMF or Fmoc-Hcy(Trt)-OH (1.1 equiv),
HATU (1.1 equiv), and either TMP (2.2 equiv) or DIEA (2.2
equiv) in CH2Cl2/DMF or DMF was added to the resin, and the
reaction mixture was agitated by rotation for 2 h (primary
amines) or overnight (secondary amines). To remove the Fmoc
group, the resin was treated with 20% piperidine in DMF
(2 ꢀ 3 mL) and then washed with several portions of DMF.
After completion of the coupling cycle the resin was also washed
with several portions of, in turn, CH2Cl2, MeOH, and CH2Cl2
and dried in vacuo. In some cases DIEA was used insted of TMP
in the last coupling of Fmoc-Hcy(Trt)-OH.
TFA salt (11.5 mg, 6%). [R]21 -38.5° (c 0.21, 0.1% TFA in
D
CD3OD/D2O 9:1). 1H NMR (400 MHz, CD3OD) δ 7.28-7.22
(m, 4H, ArH), 7.08 (m, 2H, Tyr-H2, Tyr-H6), 6.65 (m, 2H, Tyr-
H3, Tyr-H5), 4.75 (m, 1H, Tyr-HR), 4.49 (dd, J = 10.9, 4.0 Hz,
1H, Hcy3-HR), 4.40 (d, J = 15.2 Hz, 1H, NCH2a), 4.35 (d, J =
15.2 Hz, 1H, NCH2b), 3.94 (dd, J = 6.0, 3.4 Hz, 1H, Hcy1-HR),
General Procedure F: Coupling and Deprotection of Tyrosine
and β-Homotyrosine Residues. Unless otherwise stated, the
amino acids were coupled according to general method E above,
using Fmoc-Tyr(tBu)-OH (3.0 equiv), HATU (3.0 equiv), DIEA
(6.0 equiv) or Fmoc-β3hTyr(tBu)-OH (1.1 equiv), HATU (1.1
equiv), and DIEA (2.2 equiv) in DMF.
3.72 (app s, 2H, CH2CO), 3.09 (dd, J = 13.9, 7.6 Hz, 01H, Tyr-
0
Hβ), 2.98-2.79 (m, 3H, Hcy3-Hγ, Hcy3-Hγ , Tyr-Hβ ), 2.77-
0
2.64 (m, 02H, Hcy1-Hγ, Hcy1-Hγ ), 2.29-2.100 (m, 3H, Hcy1-Hβ,
Hcy1-Hβ , Hcy3-Hβ), 1.99 (m, 1H, Hcy3-Hβ ). 13C NMR (101
MHz, CD3OD) δ 175.4, 173.2, 172.3, 169.0, 157.4, 137.9, 134.3,
131.9, 131.4, 129.5, 128.8, 128.7, 128.6, 116.4, 56.1, 53.4, 52.6,
41.9, 39.1, 36.6, 35.0, 31.5, 31.3, 31.2. HPLC purity: C8 column
99.5%, biphenyl column 99.8%. HRMS (M þ Hþ): 561.1849,
General Procedure G: Resin Cleavage, Side Chain Deprotec-
tion, and Purification. The resin was transferred from the dispos-
able syringe to a 50 mL Falcon tube containing a stirring bar.
The resin was treated with a mixture of 95% TFA (2.1 mL), TES
(150 μL), and DTT (0.01 g, 64.8 μmol) under gentle stirring for
2 h. CH3CN was added to dissolve any precipitate formed before
the resin was filtered off and further washed with CH3CN. The
filtrate was evaporated, and the residue was dissolved in CH3CN.
Water was added, resulting in a precipitate which was filtered off.
The filtrate was freeze-dried, and the resulting product was dis-
solved in CH3CN/H2O (0.1% TFA), filtered through a 0.45 μm
nylon membrane, and purified with RP-HPLC. Selected fractions
were analyzed by RP-HPLC and RP-HPLC-MS, and those con-
taining pure product were pooled and lyophilized to give the pure
product as the TFA salt.
General Procedure H: Disulfide Bridge Formation and Puri-
fication.34 S-S oxidation was achieved by dissolving the linear
peptide in TFA at 3 mg/mL, adding DMSO at 0 °C to give a final
concentration of 10-15% (v/v) and then maintaining the reaction
mixture at room temperature until HPLC analysis showed oxida-
tion to be complete. After concentration in a stream of nitrogen
and coevaporation with toluene, the residue was diluted with
CH3CN/H2O (0.1% TFA), filtered through a 0.45 μm nylon
membrane, and purified with RP-HPLC. Selected fractions were
analyzed by RP-HPLC and RP-HPLC-MS, and those contain-
ing pure product were pooled and lyophilized to give the product
as the TFA salt.
C26H33N4O6S2 requires 561.1842. Anal. Calcd for C26H32N4O6S2
3
CF3COOH 0.8H2O: C, 48.80; H, 5.06; N, 8.13; S, 9.31. Found:
C, 48.70; H, 5.10; N, 7.90; S, 9.40.
3
c[Hcy-Tyr-Hcy]NHMe (7). Solid-phase attachment was
done according to general procedure C using methylindole AM
resin (300 mg, 189 μmol), Fmoc-Hcy(Trt)-OH (231 mg, 385
μmol), HATU (165 mg, 433 μmol), and TMP (100 μL, 756 μmol)
in CH2Cl2 (3.2 mL) and DMF (0.8 mL). Subsequently, the resin
was reacted with Fmoc-Tyr(tBu)-OH (274 mg, 596 μmol) as
described in procedure F using HATU (221 mg, 560 μmol),
DIEA (1.14 mmol, 199 μL), DMF (3 mL). Procedure E was
applied for the coupling of the N-terminal Fmoc-Hcy(Trt)-OH
(153 mg, 255 μmol) residue using HATU (149 mg, 393 μmol),
TMP (100 μL, 760 μmol), CH2Cl2 (2.4 mL), and DMF (2.0 mL).
Deprotection and cleavage from the resin following the general
procedure G yielded the linear sequence (48.7 mg, 89.8 μmol),
which was subjected to disulfide bridge formation according to
general method H using TFA (16.2 mL) and DMSO (1.62 mL).
The product was isolated as the TFA salt (23.1 mg, 23%).
c[Hcy-β3hTyr-Cys]NH(3-picolyl) (11). Solid-phase attach-
ment was done according to general procedure B using FMPB
AM resin (265 mg, 259 μmol), pyridin-3-ylmethanamine
(260 μL, 2.55 mmol), NaBH(OAc)3 (548 mg, 2.58 mmol), AcOH
(0.05 mL), DMF (4.95 mL). The resin was then reacted with
Fmoc-Cys(Trt)-OH (457 mg, 780μmol) as described in procedure E
using HATU (297 mg, 780 μmol), TMP (205 μL, 1.55 mmol),
General Procedure I: Disulfide Bridge Formation and Puri-
fication.36 The linear peptide was dissolved in 5% AcOH at
1.5 mM. The pH was adjusted to 6 using (NH4)2CO3 (0.5 M, aq).
DMSO was added to the solution at 0 °C to give a final con-
centration of 10-15% (v/v). The reaction mixture was stirred

8068 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Andersson et al.
CH2Cl2 (2.4 mL), and DMF (0.6 mL). The subsequent coupling
of Fmoc-β3hTyr(tBu)-OH (137 mg, 288 μmol) was performed
with HATU (109 mg, 285 μmol, 1.1 equiv) and DIEA (99 μL,
569 μmol) in DMF (3.0 mL) following general method F.
Procedure E was applied for the coupling of the N-terminal
Fmoc-Hcy(Trt)-OH (179 mg, 298 μmol) residue using HATU
(109 mg, 286 μmol) and TMP (75 μL, 570 μmol) in CH2Cl2
(2.4 mL) and DMF (0.6 mL). Deprotection and cleavage from
the resin following the general procedure G yielded the linear
sequence (2.6 mg, 3.5 μmol), which was subjected to disulfide
bridge formation according to general method I using 5% AcOH
(2.8 mL), 0.5 M (NH4)2CO3 (2.2 mL), and DMSO (0.6 mL). The
product was isolated as the TFA salt (1.6 mg, 1%).
The list of NOEs and J couplings observed for compounds 4, 5, 8,
9, 12, and 13 are presented together with JHH values in the
3
Supporting Information. In the NAMFIS analysis only NOEs
between hydrogens directly bound to the heavy atoms of the
macrocyclic ring were applied to eliminate possible errors intro-
duced by an incomplete set of conformations of the side chains.
The NAMFIS program deconvolutes the averaged NMR spec-
trumofa compound intoweightedcontributions fromtheoretical
conformers47 by performing a least-squares fit of NOE-derived
distances, and 3JHH values to the corresponding data were back-
calculated for each theoretical conformer. In other words it varies
the mole fraction of each optimized conformer until the best
possible fit to the available set of conformations is obtained.
Goodness of fit isexpressed as the “sumof the square differences”
(SSD) andthe root-mean-squaredeviation(rmsd) valuesbetween
the experimental and computed data.
NMR Spectroscopy Experiments and Computational Studies.
NMR studies were carried out using Varian Unity INOVA
spectrometers (1H at 900, 800, and 600 MHz) at 25 °C. Assign-
ment was made using NOESY, TOCSY, gCOSY, gHSQC, and
gHMBC experiments. Samples were dissolved in DMSO-d6
(Aldrich) for complete assignment including amide protons.
On the basis of previous experience, we assumed that using
DMSO instead of D2O would cause a slight redistribution of
conformation populations rather than result in the formation of
completely different orientations.50 Hence, the conformations
identified in DMSO-d6 are relevant in biological context. NOE
build-up studies were performed using the conventional “three-
π/2” NOESY pulse sequence74,75 without solvent suppression,
with mixing time of 40, 80, 120, 160, 200, 300, 400, or 500 ms.
The relaxation delay was set to 2.5 s; 16 scans were accumulated,
and 4096 and 512 points were used in the direct and indirect
dimensions, respectively. Interproton distances were calculated
using the initial rate approximation and the internal calibration
3
The JNH-HR couplings used in the NAMFIS analysis were
calculated for each computed conformation using the extended
Karplus equation developed for peptides.79,80 Corresponding
3JHH values for the vicinal protons attached to carbons only
were derived for the theoretically predicted conformers by the
equation presented by Haasnoot, De Leeuw, and Altona.81
Biochemical Evaluation. L-Leucine-p-nitroanilide (L-Leu-
pNA) was obtained from Sigma-Aldrich. [3H]AL-11 was ob-
ꢀ
tained from G. Toth, Biological Research Center (Szeged,
Hungary).63 All other reagents were of the highest grade com-
mercially available. CHO-K1 cells were kindly donated by the
Pasteur Institute (Brussels, Belgium).
Cell Culture, Transient Transfection, and Membrane Prepara-
tion. CHO-K1 and HEK293 cell lines were cultured in 75 and
500 cm2 culture flasks in Dulbecco’s modified essential medium
(DMEM) supplemented with L-glutamine (2 mM), 2% (v/v) of a
stock solution containing 5000 IU/mL penicillin and 5000 μg/
mL streptomycin (Invitrogen, Merelbeke, Belgium), 1% (v/v) of
a stock solution containing nonessential amino acids, 1 mM
sodium pyruvate, and 10% (v/v) fetal bovine serum (Invitrogen,
Merelbeke, Belgium). The cells were grown in 5% CO2 at 37 °C
until confluence was reached.
HEK293 cells were transiently transfected with plasmid
DNA, pCIneo containing the gene of human IRAP (kindly pro-
vided by Prof. M. Tsujimoto, Laboratory of Cellular Biochem-
istry, Saitama, Japan) or pTEJ482 carrying the complete human
AP-N cDNA.24 The transient transfection was performed as
described previously with 8 μL/mL Lipofectamine (Invitrogen,
Merelbeke, Belgium) and 1 μg/mL plasmid DNA.83 After trans-
fection, the cells were cultured for 2 more days. IRAP and AP-N
transfected HEK293 cells displayed 10 and 8 times higher enzyme
activity, respectively, than nontransfected cells.
CHO-K1 cell and transfected HEK293 cell membranes were
prepared as described previously.84 Briefly, the cells were har-
vested with 0.2% 2-[2-[bis(carboxymethyl)amino]ethyl(carboxy-
methyl)amino]acetic acid (EDTA) (w/v) (in phosphate-buffered
saline (PBS), pH 7.4) and centrifuged for 5 min at 500g at room
temperature. After resuspension in PBS, the number of cells were
counted and washed. The cells were then homogenized in 50 mM
Tris-HCl (at pH 7.4) using a Polytron homogenizer (10 s
at maximum speed) and a Potter homogenizer (30 strokes at
1000 rpm) and then centrifuged for 30 min (30000g at 4 °C).
The pellet was resuspended in 50 mM Tris-HCl, centrifuged
(30 min at 30000g at 4 °C), and the supernatant was removed.
The resulting pellets were stored at -20 °C until used.
Enzyme Assay. Determination of the aminopeptidase cataly-
tic activity was based on the cleavage of the substrate L-leucine-
p-nitroanilide (L-Leu-pNA)84 to give L-leucine and p-nitroani-
line. This latter compound displays a characteristic light absorp-
tion maximum at 405 nm. Pellets, prepared as described above,
were thawed and resuspended, using a Polytron homogenizer, in
enzyme assay buffer containing 50 mM Tris-HCl (pH 7.4), 140
mMNaCl, 0.1%(w/v) bovine serum albumin (BSA), and 100μM
phenylmethylsulfonyl fluoride (PMSF). The incubation mixture
β
˚
distance between germinal methylene protons (1.78 A; Cys-H ,
Hcy-Hβ, and β3hTyr-Hγ), as calculated by comparison of the
relative build-up rates (rij = rref(σref/σij)(1/6), where rij denotes the
distance between protons i and j and σij is the normalized
intensity of their NOE). At least five mixing times yielding a
linear (r2 >0.95) initial NOE rate were used to estimate the σij
build-up rates. Peak intensities were calculated using normaliza-
tion of both cross-peaks with both diagonal peaks ([(xpeak1 ꢀ
xpeak2)2/(diagpeak1 ꢀ diagpeak2)2]0.5). Vicinal coupling con-
stants were derived from 1H NMR or E-COSY spectra76 using
4096 ꢀ 1024 data points followed by zero-filling up to 8192 ꢀ
8192 points.
Theoretical Conformational Analysis. Conformational energy
calculations were performed using the systematic search method
SPMC. To ensure that the entire potential energy surface is
covered, separate calculations using the OPLS-2005 and the
Amber* all atom force fields were employed as implemented in
the program MacroModel 9.02.77 The general Born solvent
accessible (GB/SA) surface area method developed by Still was
usedinallcalculations.78 The number oftorsionanglesallowedto
vary during each Monte Carlo step ranged from 1 to n - 1 where
n equals the total number of rotatable bonds. Amide bonds were
fixed in the trans configuration. For each force field, 10 000 Monte
Carlo steps were performed in separate calculations, followed by
a maximum of 5000 PR conjugate gradient minimization steps,
˚
and a cutoff of 1.0 A was applied; conformations within 42 kJ/mol
of the global minimum were retained. Still et al. have suggested
that locating the global minimum at least 7-12 times constitutes a
thorough search of a molecule’s potential energy surface.77 The
global minimum was found at least 20 times in all calculations.
The OPLS-2005 force field, used in the theoretical analysis, had
no low quality parameters and, apart from three medium torsions
used, only high quality stretching, bending, and torsional param-
eters. The Amber force field had no low quality parameters but
had 6, 18, and 55 medium quality stretch, bend, and torsion param-
eters in addition to the 332 high quality ones.
NAMFIS Analysis. Unrestrained conformational searches in
OPLS-2005 and Amber force fields yielded a total of 700-950
optimized theoretical conformations for each analyzed structure.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8069
References
comprised 50 μL of membrane homogenate, 200 μL of L-Leu-
pNA (1.5 mM), and 50 μL of enzyme assay buffer alone or with
the test compound. The amount of membrane homogenate
corresponded to 1.5 ꢀ 105 transfected HEK293 cells in each well.
Assays were carried out at 37 °C in 96-well plates (Medisch Labo
Service, Menen, Belgium), and the formation of p-nitroaniline
was followed by measuring the absorption at 405 nm every 5 min
between 10 and 50 min in a Tecan M200 96-well reader. The
enzymatic activities were calculated by linear regression analysis
of the timewise increase in absorption.
Stability Experiments. The stability of the compounds was
studied in the presence of CHO-K1 cell membranes. Membrane
pellets were thawed and resuspended using a Polytron homog-
enizer in 50 mM Tris-HCl (pH 7.4) enzyme assay buffer.
Preincubation was carried out for 40 min at 37 °C in a finale
volume of 250 μL containing 150 μL of membrane homogenate
(corresponding to 4 ꢀ 105 CHO-K1 cells, 50 μL of enzyme assay
bufferwithoutorwith30 mM EDTA/600μM 1,10-phenantroline
(1,10-Phe), and 50 μL of enzyme assay buffer without or with the
different compounds or unlabeled Ang IV (60 μM for nonspecific
binding). The binding assay was then initiated by adding 50 μL
of enzyme assay buffer containing [3H]AL-11 (18 nM, without
(if chelators were already present in the preincubation medium)
or with 30 mM EDTA/600 μM 1,10-Phe), and the mixture was
further incubated for 30 min at 37 °C. The final chelator
concentrations were 5 mM EDTA and 100 μM 1,10-Phe. The
[3H]AL-11 concentration was 3 nM, and the final unlabeled
ligand concentrations ranged from 10-5 to 10-9 M. After incuba-
tion, the mixture was vacuum-filtered using an Inotech 24-well cell
harvester through GF/B glass fiber filters (Whatman) presoaked
in 1% (w/v) BSA. After drying, the radioactivity retained by the
filters was measured (after adding 3 mL of scintillation liquid
(Optiphase Hisafe)) using a β-counter (Perkin-Elmer). [3H]AL-11
was characterized as described in Demaegdt et al., 2009.
(1) Brookmeyer, R.; Johnson, E.; Ziegler-Graham, K.; Arrighi, H. M.
Forecasting the global burden of Alzheimer’s disease. Alzheimer’s
Dementia 2007, 3, 186–191.
(2) Small, G. W.; Rabins, P. V.; Barry, P. P.; Buckholtz, N. S.;
DeKosky, S. T.; Ferris, S. H.; Finkel, S. I.; Gwyther, L. P.;
Khachaturian, Z. S.; Lebowitz, B. D.; McRae, T. D.; Morris,
J. C.; Oakley, F.; Schneider, L. S.; Streim, J. E.; Sunderland, T.;
Teri, L. A.; Tune, L. E. Diagnosis and treatment of Alzheimer
disease and related disorders. Consensus statement of the American
Association for Geriatric Psychiatry, the Alzheimer’s Association,
and the American Geriatrics Society. JAMA, J. Am. Med. Assoc.
1997, 278, 1363–1371.
(3) Miguel-Hidalgo, J. J.; Alvarez, X. A.; Cacabelos, R.; Quack, G.
Neuroprotection by memantine against neurodegeneration induced
by beta-amyloid(1-40). Brain Res. 2002, 958, 210–221.
(4) Birks, J. Cholinesterase inhibitors for Alzheimer’s disease. Cochrane
Database Syst. Rev. 2006, CD005593.
(5) Cosman, K. M.; Boyle, L. L.; Porsteinsson, A. P. Memantine in the
treatment of mild-to-moderate Alzheimer’s disease. Expert Opin.
Pharmacother. 2007, 8, 203–214.
(6) Doraiswamy, P. M.; Xiong, G. L. Pharmacological strategies for
the prevention of Alzheimer’s disease. Expert Opin. Pharmacother.
2006, 7, 1–10.
(7) Raina, P.; Santaguida, P.; Ismaila, A.; Patterson, C.; Cowan, D.;
Levine, M.; Booker, L.; Oremus, M. Effectiveness of cholinesterase
inhibitors and memantine for treating dementia: evidence review
for a clinical practice guideline. Ann. Intern. Med. 2008, 148, 379–
397.
(8) Ruther, E.; Ritter, R.; Apecechea, M.; Freytag, S.; Gmeinbauer, R.;
Windisch, M. Sustained improvements in patients with dementia of
Alzheimer’s type (DAT) 6 months after termination of Cerebrolysin
therapy. J. Neural Transm. 2000, 107, 815–829.
(9) Nunomura, A.; Perry, G.; Smith, M. A. Prospects for Antioxidant
Therapy in Mild Cognitive Impairment and Alzheimer’S Disease.
In Oxidative Stress and Neurodegenerative Disorders; Qureshi,
G. A., Parvez, S. H., Eds.; Elsevier B.V.: Amsterdam, 2007; pp 451-
466.
(10) Sonnen, J. A.; Breitner, J. C.; Lovell, M. A.; Markesbery, W. R.;
Quinn, J. F.; Montine, T. J. Free radical-mediated damage to
brain in Alzheimer’s disease and its transgenic mouse models.
Free Radical Biol. Med. 2008, 45, 219–230.
Data Analysis. All experiments were performed at least twice
with duplicate determinations in each experiment. The calcula-
tion of IC50 values from competition binding (or enzyme inhibi-
tion) experiments was performed by nonlinear regression analysis
using GraphPad Prism 5.0. The equilibrium dissociation con-
stants (Ki values) of the tested compounds in the binding and
enzyme assays were calculated using the equationKi = [IC50/(1 þ
[L]/K)] in which [L] is the concentration of the free radioligand
(binding) or free substrate concentration (enzyme assay) and K is
the equilibrium dissociation constant (KD) of [3H]AL-11 (from
saturation binding experiments) or the Michaelis-Menten con-
stant (Km) for substrate cleavage.85
(11) Lonn, E.; Bosch, J.; Yusuf, S.; Sheridan, P.; Pogue, J.; Arnold,
J. M.; Ross, C.; Arnold, A.; Sleight, P.; Probstfield, J.; Dagenais,
G. R. Effects of long-term vitamin E supplementation on cardio-
vascular events and cancer: a randomized controlled trial. JAMA,
J. Am. Med. Assoc. 2005, 293, 1338–1347.
(12) Salloway, S.; Mintzer, J.; Weiner, M. F.; Cummings, J. L. Disease-
modifying therapies in Alzheimer’s disease. Alzheimer’s Dementia
2008, 4, 65–79.
(13) Upton, N.; Chuang, T. T.; Hunter, A. J.; Virley, D. J. 5-HT6
receptor antagonists as novel cognitive enhancing agents for
Alzheimer’s disease. Neurotherapeutics 2008, 5, 458–469.
(14) Wolfe, M. S. Therapeutic strategies for Alzheimer’s disease. Nat.
Rev. Drug Discovery 2002, 1, 859–866.
(15) Wright, J. W.; Harding, J. W. The angiotensin AT4 receptor subtype
as a target for the treatment of memory dysfunction associated with
Alzheimer’s disease. J. Renin-Angiotensin-Aldosterone Syst. 2008,
9, 226–237.
(16) Braszko, J. J.; Kupryszewski, G.; Witczuk, B.; Wisniewski, K.
Angiotensin II-(3-8)-hexapeptide affects motor activity, perfor-
mance of passive avoidance and a conditioned avoidance response
in rats. Neuroscience 1988, 27, 777–783.
(17) Braszko, J. J.; Wlasienko, J.; Koziolkiewicz, W.; Janecka, A.;
Wisniewski, K. The 3-7 fragment of angiotensin II is probably
responsible for its psychoactive properties. Brain Res. 1991, 542,
49–54.
(18) Chai, S. Y.; Yeatman, H. R.; Parker, M. W.; Ascher, D. B.;
Thompson, P. E.; Mulvey, H. T.; Albiston, A. L. Development
of cognitive enhancers based on inhibition of insulin-regulated
aminopeptidase. BMC Neurosci. 2008, 9 (Suppl. 2), S14.
(19) Hallberg, M. Targeting the insulin-regulated aminopeptidase/AT4
receptor for cognitive disorders. Drug News Perspect. 2009, 22,
133–139.
Acknowledgment. We gratefully acknowledge The Swedish
Research Council for economic support and the Swedish NMR
Centre (SNC) at the University of Gothenburg for a generous
allotment of spectrometer time. H.D. is a postdoctoral re-
searcher of the “Fund for Scientific Research;Flanders”
ꢀ
(FWO, Belgium). We are grateful to Prof. D. Tourwe and A.
Lukaszuk (Department of Organic Chemistry, Vrije Universi-
teit Brussel, Belgium) for the preparation of the AL-11 pre-
ꢀ
cursor and to G. Toth and E. Szemenyei (Biological Research
Center, Szeged, Hungary) for the tritiation of this precursor to
[3H]AL-11.63 We also thank Aleh Yahorau for conducting the
€
HRMS experiments and Henrik Schuller for the development
of a tool for computer-aided interproton distance calculations
from NOESY data.
(20) Albiston, A. L.; McDowall, S. G.; Matsacos, D.; Sim, P.; Clune, E.;
Mustafa, T.; Lee, J.; Mendelsohn, F. A.; Simpson, R. J.; Connolly,
L. M.; Chai, S. Y. Evidence that the angiotensin IV (AT(4))
receptor is the enzyme insulin-regulated aminopeptidase. J. Biol.
Chem. 2001, 276, 48623–48626.
Supporting Information Available: Additional synthetic and
analytical data, LC-UV and 1H NMR spectroscopy data used
in the identification of the isomers of LLL-configuration and the
corresponding epimers, and results from the configurational
analysis employing the NAMFIS methodology. This material is
available free of charge via the Internet at http://pubs.acs.org.
(21) Chai, S. Y.; Bastias, M. A.; Clune, E. F.; Matsacos, D. J.; Mustafa,
T.; Lee, J. H.; McDowall, S. G.; Paxinos, G.; Mendelsohn, F. A.;

8070 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Andersson et al.
Albiston, A. L. Distribution of angiotensin IV binding sites (AT4
receptor) in the human forebrain, midbrain and pons as visualised
by in vitro receptor autoradiography. J. Chem. Neuroanat. 2000,
20, 339–348.
with distance geometry calculations. J. Biomol. NMR 1997, 9,
25–34.
(44) Blackledge, M. J.; Bruschweiler, R.; Griesinger, C.; Schmidt, J. M.;
Xu, P.; Ernst, R. R. Conformational backbone dynamics of the
cyclic decapeptide antamanide. Application of a new multiconfor-
mational search algorithm based on NMR data. Biochemistry
1993, 32, 10960–10974.
(45) Bruschweiler, R.; Blackledge, M.; Ernst, R. R. Multi-conforma-
tional peptide dynamics derived from NMR data: a new search
algorithm and its application to antamanide. J. Biomol. NMR
1991, 1, 3–11.
(22) Fernando, R. N.; Larm, J.; Albiston, A. L.; Chai, S. Y. Distribu-
tion and cellular localization of insulin-regulated aminopeptidase
in the rat central nervous system. J. Comp. Neurol. 2005, 487, 372–
390.
(23) Moeller, I.;Paxinos, G.;Mendelsohn, F. A.;Aldred, G. P.;Casley, D.;
Chai, S. Y. Distribution of AT4 receptors in the Macaca fascicularis
brain. Brain Res. 1996, 712, 307–324.
(24) Olsen, J.; Cowell, G. M.; Konigshofer, E.; Danielsen, E. M.;
Moller, J.; Laustsen, L.; Hansen, O. C.; Welinder, K. G.; Engberg, J.;
Hunziker, W. Complete amino acid sequence of human intestinal
aminopeptidase N as deduced from cloned cDNA. FEBS Lett.
1988, 238, 307–314.
(25) Keller, S. R.; Scott, H. M.; Mastick, C. C.; Aebersold, R.;
Lienhard, G. E. Cloning and characterization of a novel insulin-
regulated membrane aminopeptidase from Glut4 vesicles. J. Biol.
Chem. 1995, 270, 23612–23618.
(26) Lukaszuk, A.; Demaegdt, H.; Feytens, D.; Vanderheyden, P.;
Vauquelin, G.; Tourwe, D. The replacement of His4 in angiotensin
IV by conformationally constrained residues provides highly po-
tent and selective analogues. J. Med. Chem. 2009, 52, 5612–5618.
(27) Lukaszuk, A.; Demaegdt, H.; Szemenyei, E.; Toth, G.; Tymecka,
D.; Misicka, A.; Karoyan, P.; Vanderheyden, P.; Vauquelin, G.;
Tourwe, D. Beta-Homo-amino acid scan of angiotensin IV. J. Med.
Chem. 2008, 51, 2291–2296.
(28) Albiston, A. L.; Morton, C. J.; Ng, H. L.; Pham, V.; Yeatman,
H. R.; Ye, S.; Fernando, R. N.; De Bundel, D.; Ascher, D. B.;
Mendelsohn, F. A.; Parker, M. W.; Chai, S. Y. Identification and
characterization of a new cognitive enhancer based on inhibition of
insulin-regulated aminopeptidase. FASEB J. 2008, 22, 4209–4217.
(29) Andersson, H.; Demaegdt, H.; Vauquelin, G.; Lindeberg, G.;
Karlen, A.; Hallberg, M. Ligands to the (IRAP)/AT4 receptor
encompassing a 4-hydroxydiphenylmethane scaffold replacing
Tyr2. Bioorg. Med. Chem. 2008, 16, 6924–6935.
(46) Landis, C.; Allured, V. S. Elucidation of solution structures by
conformer population analysis of NOE data. J. Am. Chem. Soc.
1991, 113, 9493–9499.
(47) Cicero, D. O.; Barbato, G.; Bazzo, R. NMR analysis of molecular
flexibility in solution: a new method for the study of complex
distributions of rapidly exchanging conformations. Application to
a 13-residue peptide with an 8-residue loop. J. Am. Chem. Soc.
1995, 117, 1027–1033.
(48) Pearlman, D. A. FINGAR, a new genetic algorithm-based method
for fitting NMR data. J. Biomol. NMR 1996, 8, 49–66.
(49) Wang, J.; Hodges, R. S.; Sykes, B. D. Generating multiple con-
formations of flexible peptides in solution on the basis of NMR
nuclear Overhauser effect data: application to desmopressin.
J. Am. Chem. Soc. 1995, 117, 8627–8634.
(50) Thepchatri, P.; Cicero, D. O.; Monteagudo, E.; Ghosh, A. K.;
Cornett, B.; Weeks, E. R.; Snyder, J. P. Conformations of lauli-
malide in DMSO-d6. J. Am. Chem. Soc. 2005, 127, 12838–12846.
(51) Bifulco, G.; Dambruoso, P.; Gomez-Paloma, L.; Riccio, R. De-
termination of relative configuration in organic compounds by
NMR spectroscopy and computational methods. Chem. Rev. 2007,
107, 3744–3779.
(52) Matsumori, N.;Kaneno, D.;Murata, M.;Nakamura, H.;Tachibana,
K. Stereochemical determination of acyclic structures based on
carbon-proton spin-coupling constants. A method of configuration
analysis for natural products. J. Org. Chem. 1999, 64, 866–876.
ꢁ
(53) Fares, C.; Hassfeld, J.; Menche, D.; Carlomagno, T. Simultaneous
ꢀ
(30) Axen, A.; Andersson, H.; Lindeberg, G.; Ronnholm, H.; Kortesmaa,
determination of the conformation and relative configuration of
archazolide A by using nuclear Overhauser effects, J couplings, and
residual dipolar couplings. Angew. Chem., Int. Ed. 2008, 47, 3722–
3726.
J.; Demaegdt, H.; Vauquelin, G.; Karlen, A.; Hallberg, M. Small
potent ligands to the insulin-regulated aminopeptidase (IRAP)/AT4
receptor. J. Pept. Sci. 2007, 13, 434–444.
ꢀ
ꢀ
(31) Axen, A.; Lindeberg, G.; Demaegdt, H.; Vauquelin, G.; Karlen, A.;
Hallberg, M. Cyclic insulin-regulated aminopeptidase (IRAP)/
AT4 receptor ligands. J. Pept. Sci. 2006, 12, 705–713.
(32) Carpino, L. A.; Han, G. Y. 9-Fluorenylmethoxycarbonyl amino-
protecting group. J. Org. Chem. 1972, 37, 3404–3409.
(33) Angell, Y. M.; Alsina, J.; Albericio, F.; Barany, G. Practical
protocols for stepwise solid-phase synthesis of cysteine-containing
peptides. J. Pept. Res. 2002, 60, 292–299.
(54) Sanchez-Pedregal, V. M.; Santamaria-Fernandez, R.; Navarro-
Vazquez, A. Residual dipolar couplings of freely rotating groups in
small molecules. Stereochemical assignment and side-chain con-
formation of 8-phenylmenthol. Org. Lett. 2009, 11, 1471–1474.
(55) Thiele, C. M. Simultaneous assignment of all diastereotopic pro-
tons in strychnine using RDCs: PELG as alignment medium for
organic molecules. J. Org. Chem. 2004, 69, 7403–7413.
(56) Thiele, C. M. Residual dipolar couplings (RDCs) in organic
structure determination. Eur. J. Org. Chem. 2008, 5673–5685.
(57) Demaegdt, H.; Laeremans, H.; De Backer, J.-P.; Mosselmans, S.;
Le, M. T.; Kersemans, V.; Michotte, Y.; Vauquelin, G.; Vander-
heyden, P. M. L. Synergistic modulation of cystinyl aminopepti-
dase by divalent cation chelators. Biochem. Parmacol. 2004, 68,
893–900.
(34) Wallace, T. J.; Mahon, J. J. Reactions of thiols with sulfoxides. II.
Kinetics and mechanistic implications. J. Am. Chem. Soc. 2002, 86,
4099–4103.
(35) Otaka, A.; Koide, T.; Shide, A.; Fujii, N. Application of
dimethylsulphoxide(DMSO)/trifluoroacetic acid(TFA) oxidation
to the synthesis of cystine-containing peptide. Tetrahedron Lett.
1991, 32, 1223–1226.
(36) Tam, J. P.; Wu, C. R.; Liu, W.; Zhang, J. W. Disulfide bond
formation in peptides by dimethyl sulfoxide. Scope and applica-
tions. J. Am. Chem. Soc. 2002, 113, 6657–6662.
(37) Tang, Z. L.; Pelletier, J. C. Preparation of amides from acids and
resin bound azides: suppression of intramolecular lactam forma-
tion. Tetrahedron Lett. 1998, 39, 4773–4776.
(58) Demaegdt, H.; Lenaerts, P. J.; Swales, J.; De Backer, J. P.; Laere-
mans, H.; Le, M. T.; Kersemans, K.; Vogel, L. K.; Michotte, Y.;
Vanderheyden, P.; Vauquelin, G. Angiotensin AT4 receptor ligand
interaction with cystinyl aminopeptidase and aminopeptidase N:
[
125I]angiotensin IV only binds to the cystinyl aminopeptidase apo-
enzyme. Eur. J. Pharmacol. 2006, 546, 19–27.
(59) Kobori, T.; Goda, K.; Sugimoto, K.; Ota, T.; Tomisawa, K.
Preparation of Peptide Derivatives as Angiotensin IV Receptor
Agonists. WO 97/03093 A1, 1997.
(60) Kobori, T.; Goda, K.; Sugimoto, K.; Ota, T.; Tomisawa, K.
Preparation of Amino Acid Derivatives as Angiotensin IV Recep-
tor Agonists. WO 98/05624 A1, 1998.
(38) Brandt, M.; Gammeltoft, S.; Jensen, K. Microwave heating for
solid-phase peptide synthesis: general evaluation and application
to 15-mer phosphopeptides. Int. J. Pept. Res. Ther. 2006, 12, 349–
357.
ꢀ
(39) Vazquez, J.; Albericio, F. A useful and sensitive color test to
(61) Sardinia, M. F.; Hanesworth, J. M.; Krebs, L. T.; Harding, J. W.
AT4 receptor binding characteristics: D-amino acid- and glycine-
substituted peptides. Peptides 1993, 14, 949–954.
monitor aldehydes on solid-phase. Tetrahedron Lett. 2001, 42,
6691–6693.
(40) Kay, C.; Lorthioir, O. E.; Parr, N. J.; Congreve, M.; McKeown,
S. C.; Scicinski, J. J.; Ley, S. V. Solid-phase reaction monitoring;
chemical derivatization and off-bead analysis. Biotechnol. Bioeng.
2000, 71, 110–118.
(62) Sardinia, M. F.; Hanesworth, J. M.; Krishnan, F.; Harding,
J. W. AT4 receptor structure-binding relationship: N-terminal-
modified angiotensin IV analogues. Peptides 1994, 15, 1399–1406.
(63) Demaegdt, H.; Lukaszuk, A.; De Buyser, E.; De Backer, J. P.;
Szemenyei, E.; Toth, G.; Chakravarthy, S.; Panicker, M.; Michotte,
Y.; Tourwe, D.; Vauquelin, G. Selective labeling of IRAP by the
tritiated AT4 receptor ligand [3H]angiotensin IV and its stable
analog [3H]AL-11. Mol. Cell. Endocrinol. 2009, 311, 77–86.
(64) Demaegdt, H.; Laeremans, H.; De Backer, J. P.; Mosselmans, S.;
Le, M. T.; Kersemans, V.; Michotte, Y.; Vauquelin, G.; Vander-
heyden, P. M. Synergistic modulation of cystinyl aminopeptidase by
divalent cation chelators. Biochem. Pharmacol. 2004, 68, 893–900.
(41) Vojkovsky, T. Detection of secondary amines on solid phase. Pept.
Res. 1995, 8, 236–237.
(42) Nevins, N.; Cicero, D.; Snyder, J. P. A test of the single-conforma-
tion hypothesis in the analysis of NMR data for small polar
molecules: a force field comparison. J. Org. Chem. 1999, 64,
3979–3986.
(43) Beckers, M. L.; Buydens, L. M.; Pikkemaat, J. A.; Altona, C.
Application of a genetic algorithm in the conformational analysis
of methylene-acetal-linked thymine dimers in DNA: comparison

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8071
(65) Kenny, A. J. Regulatory peptide metabolism at cell surfaces: the
key role of endopeptidase-24.11. Biomed. Biochim. Acta 1986, 45,
1503–1513.
(66) Lalu, K.; Lampelo, S.; Vanha-Perttula, T. Characterization of
three aminopeptidases purified from maternal serum. Biochim.
Biophys. Acta 1986, 873, 190–197.
(67) Laustsen, P. G.; Vang, S.; Kristensen, T. Mutational analysis of the
active site of human insulin-regulated aminopeptidase. Eur. J.
Biochem. 2001, 268, 98–104.
(68) Schmidt, B.; Lindman, S.; Tong, W.; Lindeberg, G.; Gogoll, A.;
Lai, Z.; Thornwall, M.; Synnergren, B.; Nilsson, A.; Welch, C. J.;
Sohtell, M.; Westerlund, C.; Nyberg, F.; Karlen, A.; Hallberg, A.
Design, synthesis, and biological activities of four angiotensin II
receptor ligands with gamma-turn mimetics replacing amino acid
residues 3-5. J. Med. Chem. 1997, 40, 903–919.
(69) Sigel, H.; Martin, R. B. Coordinating properties of the amide bond.
Stability and structure of metal ion complexes of peptides and
related ligands. Chem. Rev. 1982, 82, 385–426.
(76) Griesinger, C.; Sørensen, O. W.; Ernst, R. R. Practical aspects of
the E.COSY technique. Measurement of scalar spin-spin coupling
constants in peptides. J. Magn. Reson. 1987, 75, 474–492.
(77) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C.
Macromodel;an integrated software system for modeling organic
and bioorganic molecules using molecular mechanics. J. Comput.
Chem. 1990, 11, 440–467.
(78) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T.
Semianalytical treatment of solvation for molecular mechanics
and dynamics. J. Am. Chem. Soc. 1990, 112, 6127–6129.
(79) Kessler, H.; Griesinger, C.; Lautz, J.; Muller, A.; Vangunsteren,
W. F.; Berendsen, H. J. C. Conformational dynamics detected by
nuclear magnetic resonance NOE values and J coupling constants.
J. Am. Chem. Soc. 1988, 110, 3393–3396.
(80) Schmidt, J. M. A versatile component-coupling model to account
for substituent effects: Application to polypeptide φ and χ1 torsion
related 3J data. J. Magn. Reson. 2007, 186, 34–50.
(70) Ye, S.; Chai, S. Y.; Lew, R. A.; Ascher, D. B.; Morton, C. J.;
Parker, M. W.; Albiston, A. L. Identification of modulating
residues defining the catalytic cleft of insulin-regulated amino-
peptidase. Biochem. Cell Biol. 2008, 86, 251–261.
(71) Tholander, F.; Muroya, A.; Roques, B. P.; Fournie-Zaluski, M. C.;
Thunnissen, M. M.; Haeggstrom, J. Z. Structure-based dissection
of the active site chemistry of leukotriene A4 hydrolase: implica-
tions for M1 aminopeptidases and inhibitor design. Chem. Biol.
2008, 15, 920–929.
(72) Ito, K.; Nakajima, Y.; Onohara, Y.; Takeo, M.; Nakashima,
K.; Matsubara, F.; Ito, T.; Yoshimoto, T. Crystal structure of
aminopeptidase N (proteobacteria alanyl aminopeptidase) from
Escherichia coli and conformational change of methionine 260
involved in substrate recognition. J. Biol. Chem. 2006, 281, 33664–
33676.
(73) Ye, S.; Chai, S. Y.; Lew, R. A.; Albiston, A. L. Insulin-regulated
aminopeptidase: analysis of peptide substrate and inhibitor bind-
ing to the catalytic domain. Biol. Chem. 2007, 388, 399–403.
(74) Bodenhausen, G.; Wagner, G.; Rance, M.; Sorensen, O. W.; Wuthrich,
K.;Ernst,R.R.Longitudinal2-spin order in 2D exchange spectroscopy
(NOESY). J. Magn. Reson. 1984, 59, 542–550.
(81) Haasnoot, C. A. G.; Deleeuw, F. A. A. M.; Altona, C. The
relationship between proton-proton NMR coupling-constants
and substituent electronegativities;I. An empirical generaliza-
tion of the Karplus equation. Tetrahedron 1980, 36, 2783–2792.
(82) Johansen, T. E.; Scholler, M. S.; Tolstoy, S.; Schwartz, T. W.
Biosynthesis of peptide precursors and protease inhibitors using
new constitutive and inducible eukaryotic expression vectors.
FEBS Lett. 1990, 267, 289–294.
(83) Le, M. T.; De Backer, J. P.; Hunyady, L.; Vanderheyden, P. M.;
Vauquelin, G. Ligand binding and functional properties of human
angiotensin AT1 receptors in transiently and stably expressed
CHO-K1 cells. Eur. J. Pharmacol. 2005, 513, 35–45.
(84) Demaegdt, H.; Vanderheyden, P.; De Backer, J. P.; Mosselmans,
S.; Laeremans, H.; Le, M. T.; Kersemans, V.; Michotte, Y.;
Vauquelin, G. Endogenous cystinyl aminopeptidase in Chinese
hamster ovary cells: characterization by [125I]Ang IV binding and
catalytic activity. Biochem. Pharmacol. 2004, 68, 885–892.
(85) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibitor which causes 50%
inhibition (IC50) of an enzymatic reaction. Biochem. Pharmacol.
1973, 22, 3099–3108.
(75) Meier, B. H.; Ernst, R. R. Elucidation of chemical exchange
networks by two-dimensional NMR spectroscopy: the hepta-
methylbenzenonium ion. J. Am. Chem. Soc. 1979, 101, 6441–6442.
(86) The difference in the magnitude of the SSD values for 12 and 13
originates from the nature of the SSD. Its absolute value depends
on the number of used constraints.42