Synthesis and molecular modeling of a lisinopril-tryptophan analogue inhibitor of angiotensin I-converting enzyme

Article abstract of DOI:10.1016/j.bmcl.2006.06.004

With a view to developing a more C-domain-selective angiotensin I-converting enzyme (ACE)-inhibitor, a novel analogue of lisinopril has been synthesized which incorporates a bulky P₂^′ tryptophan functionality. This inhibitor demonstr

Full text of DOI:10.1016/j.bmcl.2006.06.004

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4616–4619
Synthesis and molecular modeling of a lisinopril–tryptophan
analogue inhibitor of angiotensin I-converting enzyme
Aloysius T. Nchinda,a Kelly Chibale,a,b Pierre Redelinghuysa and Edward D. Sturrocka,*
aInstitute of Infectious Disease and Molecular Medicine, University of Cape Town, Observatory 7925, Cape Town, South Africa
bDepartment of Chemistry, University of Cape Town, Rondebosch 7701, Cape Town, South Africa
Received 17 April 2006; revised 2 June 2006; accepted 2 June 2006
Available online 19 June 2006
Abstract—With a view to developing a more C-domain-selective angiotensin I-converting enzyme (ACE)-inhibitor, a novel analogue
of lisinopril has been synthesized which incorporates a bulky P02 tryptophan functionality. This inhibitor demonstrated a significant-
ly increased specificity for the C-domain as compared with lisinopril. Molecular docking revealed hydrophobic and hydrogen-bond-
ing interactions with residues of the C-domain S02 subsite.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Lisinopril, a widely used commercially available ACE
inhibitor in the treatment of hypertension and myocar-
dial infarction, is a tight-binding competitive inhibitor
(Ki = 2.7 · 10ꢀ10).1–5 However, side effects associated
with ACE inhibitor use, such as angioedema and persis-
tent cough,6 have been attributed to a lack of adequate
C-domain-selectivity and an inappropriate suppression
of bradykinin hydrolysis by the ACE N-domain.
Recently, Georgiadis and co-workers reported the syn-
thesis of a highly C-domain-selective (ꢁ3000-fold) phos-
phinic acid ACE inhibitor, RXP A380.7 One key feature
of this C-domain-selectivity is a tryptophan moiety at
the P02 position. With a view to developing more C-do-
main-selective ACE inhibitors with similar pharmaco-
logical profiles to that of lisinopril, we determined the
effect of the introduction of a tryptophan moiety at
the P02 position of lisinopril.
thionyl chloride and methanol as described previous-
ly9,10) was effected using EDCÆHCl in the presence of
HOBt and diisopropylethylamine to afford the dia-
stereomeric pseudopeptide 4 in 74% yield.
Characterization of this diastereomeric mixture was
achieved from the mass (EI-MS) and H NMR spectro-
1
scopic data. The EI-MS data indicated a molecular ion
peak at 637 corresponding to M++H. The 1H NMR
spectrum showed two singlets at d 3.71 and 3.73 ppm
corresponding to the two methyl ester groups of the dia-
stereomers. When the diastereomeric mixture 4 was stir-
red in 4 N HCl in EtOAc at room temperature for 24 h,
followed by solvent removal and stirring in 0.5 N LiOH
for a further 5 h, the desired product 5 was obtained in
quantitative yield as a mixture of diastereoisomers
(Scheme 1). Purification and separation of the 60:40
mixture by HPLC delivered the two diastereomers (5a
and 5b) (Fig. 1).
A lisinopril analogue, incorporating a tryptophan moie-
ty at the P02 position, was synthesized as follows (Scheme
1): reductive amination of ethyl 2-oxo-4-phenyl butyrate
1 and N-e-(tert-butoxycarbonyl)-L-lysine 2 using an eth-
anolic solution of NaBH3CN8 gave key intermediate 3
as a 55:45 mixture of diastereoisomers. Coupling of
compound 3 with the L-tryptophan methyl ester (pre-
pared by methylation of tryptophan in the presence of
The two isomers (5a and 5b) of the lisinopril–tryptophan
analogues were tested for ACE inhibitory activity using
purified testis ACE (C-domain) and somatic ACE N-do-
main together with Z-Phe-His-Leu as substrate.11 The
ACE inhibitory activities were determined as previously
described11 with some modifications. Recombinant
truncated forms of the C- and N-domains12,13 were used
in the enzyme assays. Enzyme (4 lg/ml) was pre-incu-
bated with inhibitor at concentrations ranging from 0
to 500 lM. Residual enzyme activity was determined
from enzyme–inhibitor solution using Z-Phe-His-Leu
at two substrate concentrations. The enzyme assay was
Keywords: ACE inhibitor; Molecular docking; Hypertension;
C-domain.
*
Corresponding author. Tel.: +27 21 4066312; fax: +27 21
4066470; e-mail: sturrock@curie.uct.ac.za
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.004

A. T. Nchinda et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4616–4619
4617
COOEt
O
NHBoc
COOH
1
i
EtOOC
*
NHBoc
N
H
3
H2N
COOH
ii
2
NHBoc
NH2
EtOOC
*
H
N
COOH
COOMe
NH
H
N
H
N
COOH
iii
N
*
O
H
O
NH
4
5
Scheme 1. Reagents and conditions: (i) ketone 1 (4.0 equiv), amino acid 2 (1.0 equiv), NaBH3CN (2.0 equiv), 50% EtOH/H2O, rt, 12 h;
(ii) L-tryptophan methyl ester, EDCÆHCl, HOBt, i-Pr2NEt (1.0 equiv), dry DMF, rt, 72 h; (iii) a—4 N HCl, EtOAc, rt, 24 h and b—0.5 N LiOH,
THF–MeOH, rt, 5 h.
5a
Lisinopril
5b
H2N
H2N
H2N
HOOC
H
N
HOOC
H
N
HOOC
COOH
NH
COOH
NH
N
N
N
N
H
H
H
COOH
O
O
O
Ph
Ph
Ph
Figure 1. Structures of lisinopril and diastereomers 5a and 5b.
adapted so that it could be carried out in a 96-well plate
and the fluorescence was measured at Ex = 360 nm;
Em = 486 nm on a Varian Cary Eclipse plate reader.
All assays were conducted in triplicate. Inhibition curves
were plotted using GraphPad Prism 4.01 and Ki values
were determined from Dixon plots of 1/V versus [I].
The ACE inhibitory activities of 5a and 5b are shown
in Table 1.
To support these observations, the synthesized com-
pounds 5a and 5b were subsequently modeled into the
binding sites of the testis ACE (C-domain) crystal struc-
ture14 and a homology model of sACE N-domain.
Molecular modeling calculations were performed using
the DISCOVER module of INSIGHT II (Accelrys
Inc., Version 98.0) on a Silicon Graphics Octane 1 work-
station. The starting structure was the X-ray crystal
structure of testis ACE complexed with the inhibitor lis-
inopril14 and the homology modeled structure of the
N-domain. After removing the crystallographic water
molecules and adding hydrogens, the CVFF and the
ESFF (metal adapted) force-fields were used in all ener-
gy minimizations and dynamic runs. The conjugate gra-
dient minimization algorithm was used after running
1000 initial steps, and then 3000 cycles of molecular
dynamics followed by 3000 cycles of energy minimiza-
tion in an NTV ensemble, at a temperature of 300 K.
The compounds all showed Ki values in the micromolar
range as compared with the nanomolar Ki value of lisin-
opril. Incorporation of a tryptophan moiety at the P02 po-
sition of lisinopril (compounds 5a and 5b) resulted in a
marked increase in C-domain-selectivity (25- to 100-fold)
as compared with lisinopril (2.6-fold) (Table 1). This
observation is in agreement with previously reported
work by Georgiadis et al.7 suggesting that a large P02 res-
idue is an important feature for C-domain-selectivity.

4618
A. T. Nchinda et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4616–4619
Table 1. ACE inhibitory activity profiles of lisinopril and its derivatives 5a and 5b
Compound
Total PE (kcal/mol)
Ki
Selectivity Ki N/C
C-dom
N-dom
C-dom
N-dom
Lisinopril
ꢀ890.0
ꢀ869.6
ꢀ830.6
ꢀ743.7
ꢀ630.1
ꢀ704.3
51.0 nM
7.0 lM
26.3 lM
131.5 nM
NI
NI
2.6
>100
>25
5a
5b
NI: No inhibition up to a concentration of 500 lM.
Total potential energy (PE) is the sum of the energies calculated for all the bonded-atom interactions and for all the nonbonded-atom pairs.
All calculations were carried out in a dielectric constant
˚
of 1.00 and a cut-off distance of 9.50 A. The structures
Tyr523, and Phe527. The binding of compounds 5a
and 5b in the C-domain active site resembled that of lis-
inopril, where the phenyl group of the P1 phenylalanine
interacted with the hydrophobic Val518 and the polar
Ser355 residues of the S1 subsite and the lysyl group
was accommodated in the S02 subsite. The distance be-
tween the phenylalanine-carbonyl and the Zn-atom in
the active site of the C-domain was observed to be
of compounds 5a and 5b were generated with standard
bond lengths and angles using the builder tool of IN-
SIGHT II software (Accelrys Inc.) and then minimized.
The initial position of the compounds in the C- and N-
domain active sites of ACE was obtained by superim-
posing their important pharmacophoric groups on the
corresponding atoms of lisinopril in the tACE-lisinopril
complex. After removal of the reference inhibitor (lisin-
opril), the structure of the inhibitor-complex was refined
by running energy minimizations and molecular dynam-
ics. These compounds were found to occupy the S1, S2,
and S02 subsites, and an example of compound 5a in
the active site of testis ACE is shown in Figure 2. From
the molecular docking experiments, the energy-mini-
mized bound conformers of the inhibitors in the active
site of the C- and N-domains were obtained. The inhib-
itors all exhibited negative total potential energies for
the protein–ligand interaction. These were consistent
with the determined biological binding affinities for the
relevant domains and it was an indication of the relative
stability of the enzyme–ligand complex (Table 1). The
docking of compounds 5a and 5b into both domains
showed consistency between the interactions of key
ACE active site residues and the inhibitors, and also re-
vealed that the synthesized compounds bound to the C-
and N-domains in a similar fashion.
˚
˚
2.73 A for compound 5a and 2.74 A for compound 5b.
These values are within the range of a reasonable
H-bonding interaction.
The S02 subsite of the enzyme readily accommodated the
tryptophan moiety of compounds 5a and 5b. This obser-
vation is in agreement with previous studies which have
highlighted the importance of a deep S02 pocket as a
determinant of C-domain-selectivity.7,15 The P20 trypto-
phan moieties of compounds 5a and 5b demonstrated
additional hydrophobic and hydrogen-bonding interac-
tions with S02 residues Thr282, Val379, Val380, and
Asp453 of the C-domain active site (Fig. 2), and may
thus explain the increased C-domain-selectivity of these
compounds as compared with lisinopril.
Here, the aromatic ring of the P02 tryptophan increases
the overall hydrophobicity of the molecule and conse-
quently the interaction with the hydrophobic Val379
and Val380 residues in the C-domain. These residues
are replaced by the polar residues Ser357 and Thr358
in the N-domain. The other striking feature is the ability
to establish a strong H-bonding interaction between the
indole-NH and Asp453 in the C-domain which is re-
placed by a less polar Glu431 in the N-domain. This ami-
no acid’s side chain is longer than that of Asp, decreasing
the size of the binding pocket. Interactions with the S20
residues of the C- and N-domains are absent in the case
of lisinopril which lacks a bulky P02 functionality. From
The residues forming the subsites of the C- and N-do-
main active site were defined as those located at a dis-
˚
tance less than 6.00 A from the important
functionalities of the inhibitors. The active site residues
for the C-domain were Gln281, Thr282, His353, Ala354,
Ser355, Ala356, Val379, Val380, His383, Glu384,
His387, Phe391, Glu411, Asp453, Phe457, Phe460,
Lys+511, Phe512, His513, Ser516, Val518, Tyr520,
Figure 2. Interactions of Compound 5a with S1, S01 and S02 active site residues (A) and binding pockets (B) of tACE.

A. T. Nchinda et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4616–4619
4619
the molecular docking experiments and the observed
References and notes
ACE inhibitory activities, together with previous studies
by Sieburth and co-workers,16 the most potent com-
pound 5a possesses an S-configuration at the stereogenic
center bearing the zinc-binding carboxylate group.
1. Borghi, C.; Ambrosioni, E. Am. Heart J. 2003, 145, 80.
2. Buikema, H. Evidence-based Cardiovasc. Med. 2003, 7,
131.
3. Cushman, D. W.; Cheung, H. S.; Sbo, E. F.; Ondetti, M.
A. Biochemistry 1977, 16, 5484.
Docking experiments revealed that the total potential
energies for the complex of compound 5a with the C-do-
main and the N-domain were ꢀ869.6 and ꢀ630.1 kcal
molꢀ1, respectively. Whereas, for compound 5b they
were ꢀ830.6 and ꢀ704.3 kcal molꢀ1 for the C-domain
and N-domain, respectively (Table 1). This is in agree-
ment with the observed biological inhibiting binding
affinity for compound 5a, which was found to be
>100-fold more C-domain-selective as compared with
lisinopril (2.6-fold) and more C-domain-selective as
compared with compound 5b.
4. Ondetti, M. A.; Rubin, B.; Cushman, D. W. Science 1977,
196, 441.
5. Brown, N. J.; Vaughan, D. E. Circulation 1998, 97, 1411.
6. Morimoto, T.; Gandhi, T. K.; Fiskio, J. M.; Seger, A. C.;
So, J. W.; Cook, E. F.; Fukui, T.; Bates, D. W. J. Eval.
Clin. Pract. 2004, 10, 499.
7. Georgiadis, D.; Cuniasse, P.; Cotton, J.; Yiotakis, A.;
Dive, V. Biochemistry 2004, 43, 8048.
8. Schnorrenberg, G.; Roos, O.; Losel, W.; Weidemann, I.;
Gaida, W.; Hoefke, W.; Arndts, D.; Streller, I. In USPTO
Patent Full text and Image Database, 1989, Boehringer
Ingelheim GmbH: US, 1–18.
9. Hvidt, T.; Szarek, W. A. Can. J. Chem. 1988, 66, 779.
10. Peng, S.; Winterfeldt, E. Liebigs Ann. Chem. 1990, 313.
11. Almquist, R. G.; Chao, W. R.; Ellis, M. E.; Johnson, H.
L. J. Med. Chem. 1980, 23, 1392.
12. Yu, X. C.; Sturrock, E. D.; Wu, Z.; Biemann, K.; Ehlers,
M. R. W.; Riordan, J. F. J. Biol. Chem. 1997, 272, 3511.
13. Corradi, H. R.; Schwager, S. L. U.; Nchinda, A. T.;
Sturrock, E. D.; Acharya, K. R. J. Mol. Biol. 2006, 357,
964.
14. Natesh, R.; Schwager, S. L. U.; Sturrock, E. D.; Acharya,
K. R. Nature 2003, 421, 551.
15. Acharya, K. R.; Sturrock, E. D.; Riordan, J. F.; Ehlers,
M. R. Nat. Rev. Drug Disc. 2003, 2, 891.
In summary, we have described the synthesis of lisinopril
analogues incorporating a bulky tryptophan moiety at
the P02 position. ACE inhibitory activities and molecular
docking experiments have shown that this incorporation
leads to a marked increase in C-domain-selectivity.
Acknowledgments
The authors wish to thank the South African National
Research Foundation, the University of Cape Town
and the Wellcome Trust (U.K.) for their generous finan-
cial support.
16. Kim, J.; Hewitt, G.; Carroll, P.; Sieburth, S. M. J. Org.
Chem. 2005, 70, 5781.