N-(Hydroxyaminocarbonyl)phenylalanine: A novel class of inhibitor for carboxypeptidase A

Article abstract of DOI:10.1016/S0968-0896(00)00235-2

N-(Hydroxyaminocarbonyl)phenylalanine (1) was designed rationally as a new type of inhibitor for carboxypeptidase A (CPA). The designed inhibitor was readily prepared from phenylalnine benzyl ester in two steps and evaluated to find that rac-1 inhibits CP

Full text of DOI:10.1016/S0968-0896(00)00235-2

Bioorganic & Medicinal Chemistry 9 (2001) 185±189
N-(Hydroxyaminocarbonyl)phenylalanine:
A Novel Class of Inhibitor for Carboxypeptidase A
Sang J. Chung and Dong H. Kim*
Center for Biofunctional Molecules and Department of Chemistry, Pohang University of Science and Technology,
San 31 Hyojadong, Namgu, Pohang 790-784, South Korea
Received 19 June 2000; accepted 19 August 2000
AbstractÐN-(Hydroxyaminocarbonyl)phenylalanine (1) was designed rationally as a new type of inhibitor for carboxypeptidase A
(CPA). The designed inhibitor was readily prepared from phenylalnine benzyl ester in two steps and evaluated to ®nd that rac-1
inhibits CPA in a competitive fashion with the Ki value of 2.09 mM. Surprisingly, inhibitor 1 having the d-con®guration is more
potent (Ki=1.54 mM) than its antipode by about 3-fold. A possible explanation for the stereochemistry observed in the inhibition of
CPA with 1 is presented. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
have been speculated to interact with the heme iron that
plays an important role in the catalytic pathway of 5-
lipoxygenase.12,13 We wish to describe in this paper
design and synthesis of a new type of potent CPA inhi-
bitors by exploiting the metal chelating property of
hydroxyurea.
The zinc peptidases constitute an important class of pro-
teolytic enzymes from both the mechanistic and biological
standpoint. Among these enzymes, carboxypeptidase A
(CPA) that cleaves the C-terminal amino acid residue
having a hydrophobic side chain has been most intensively
studied and serves as a prototypic zinc peptidase.1 CPA
has also been used as a model target enzyme for the
development of inhibitor design strategies that can be
useful for designing inhibitors of other zinc proteases of
medicinal interest such as angiotensin converting
enzyme and matrix metalloproteases.2 5 In principle,
inhibitors of zinc proteases are obtained by incorporat-
ing a zinc ligating functionality into the structural frame
of a substrate-like molecule that can be recognized and
accommodated by the target enzyme. Hydroxamate
moiety is one of the representative zinc ligating motifs
and has been extensively utilized for the design of inhi-
bitors for numerous zinc metalloproteases since Nishino
and Powers6 have reported its use as a zinc ligating
functionality in designing thermolysin inhibitors.7 9 N-
Hydroxyurea can be considered as hydroxamate of car-
bamic acid and thus is expected to chelate metal ions like
hydroxamate does. Hydroxyurea and 1-ethyl-3-hydro-
xyurea were reported to form complexes with Fe(III) and
Cu(II) ions.10 Some of hydroxyurea derivatives are potent
5-lipoxygenase inhibitors and under clinical evaluation as
anti-asthma agents.11 These hydroxyurea derivatives
Results and Discussion
CPA is a much studied and well characterized zinc con-
taining metalloprotease which catalyzes the cleavage of
the C-terminal amino acid residue having an aromatic
side chain from peptide substrate.1 Three binding and
one catalytic sites have been identi®ed at the active site
of CPA. The presence of a zinc ion that is coordinated
tightly to the backbone amino acid residues of His-69,
Glu-72 and His-196 is essential for the catalytic hydro-
lysis reaction. A molecule of water is bound loosely to
the zinc ion as the fourth ligand. The zinc ion activates
the water molecule in collaboration with the carboxylate
of Glu-270, generating a hydroxyl group that attacks at
the carbonyl carbon of the scissile peptide bond. The
other important role that the zinc ion plays in the cata-
lytic process is to stabilize the tetrahedral transition
state that is generated by the hydroxyl attack on the
peptide carbonyl in the reaction process. The other
principal binding sites are Arg-145 an0d a substrate
0
recognition pocket at the S1 subsite (S1 pocket). The
latter that is known as the primary substrate recognition
pocket is shaped complementary to an aromatic ring
present in the side chain of amino acids such as Phe, and
*Corresponding author. Tel.: +82-562-279-2101; fax: +82-562-279-
5877; e-mail: dhkim@vision.postech.ac.kr
0968-0896/01/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00235-2

186
S. J. Chung, D. H. Kim, / Bioorg. Med. Chem. 9 (2001) 185±189
Figure 2. Schematic representation showing the design rationale for 1
as an inhibitor of CPA.
Figure 1. Schematic representation of the active site of CPA.
the former is involved in forming bifurcated hydrogen
bonds with the terminal carboxylate of substrate (Fig. 1).1
Scheme 1.
The X-ray crystal structure of CPA complex formed with
Gly-l-Tyr, a slowly hydrolyzing substrate for CPA,
revealed that the terminal amino group and the carbox-
amide carbonyl group of the CPA bound Gly-l-Tyr form
a bidentate coordinative bond to the active site zinc
ion.14 The distance between one of carboxylate oxygens
of Glu-270 and a-amino nitrogen of the bound ligand is
found to be 2.99 A, suggesting that there might be
formed a hydrogen bond between them.
We envisioned that the replacement of Gly in Gly-l-Tyr
with hydroxyaminocarbonyl to give a N-substituted
hydroxyurea would convert the substrate into inhibitor of
CPA. Since hydroxyurea has been known to form chela-
tion complexes with metal ions such as Fe(III) and Cu(II),
compound 1 is thought to bind CPA with its hydroxy-
aminocarbonyl moiety forming a bidentate coordinative
bond to the active site zinc ion.10 The carboxylate and
phenyl ring in 1 would interact with the guanidinium
Figure 3. The Dixon plot.
Their inhibitory constants (Kis) obtained from Dixon's
plots (Fig. 3)15 are listed in Table 1. The CPA inhibitory
potency of 1 is comparable to that of 2-benzylsuccinic
acid, a well known CPA inhibitor but less than that of
2-benzyl-3-mercaptopropanoic acid by about two order
of magnitude (Table 1). Inhibitor 1 binds CPA with
higher binding anity than 3-hydroxyaminocarbonyl-2-
benzylpropanoic acid (Ki=24.2 mM)16 that bears a
strong zinc chelating functionality of hydroxamate by
about 12-fold. This observation supports the design
0
moiety of Arg-145 and the S1 pocket of CPA, respec-
tively. Hydrogen bonding between the Glu-270 carboxy-
late and the NH of the hydroxyurea moiety was
anticipated on the basis of the X-ray crystal structure of
14
.
CPA Gly-l-Tyr complex and such an interaction was
deemed to strengthen the binding of 1 to CPA. Figure 2
depicts schematically the possible binding mode of 1 to
CPA, leading to inhibition of the enzymic activity.
Table 1. Inhibitory constants for CPA inhibition
Inhibitor
Ki (mM)a
rac-1
l-1
2.09Æ0.16
4.56Æ0.38
1.54Æ0.10
58.50Æ2.94
18.98Æ1.1
Ðb
d-1
rac-2
l-2
Inhibitor 1 was readily prepared in racemic and optically
active forms by allowing phenylalanine benzyl ester to
react with O-benzylhydroxylamine in the presence of car-
bonyldiimidazole (CDI) followed by removal of the ben-
zyl groups by hydrogenolysis in the presence of Pd-C
catalyst (Scheme 1).
d-2
HBPAc
BSAe
BMPAg
24.2d
1.1Æ0.3f
0.011h
aKi values were calculated using GraFit1
.
bNo inhibition was observed up to 3.6 mM of d-2.
c3-Hydroxyaminocarbonyl-2-benzylpropanoic acid.
dref 16.
The prepared compounds were evaluated as CPA inhi-
bitors by a standard procedure to ®nd that they inhibit
the enzyme in a competitive reversible fashion as can be
seen from Figure 3.
erac-2-Benzylsuccinic acid.
fref 22.
grac-2-Benzyl-3-mercaptopropanoic acid.
href 23.

S. J. Chung, D. H. Kim, / Bioorg. Med. Chem. 9 (2001) 185±189
187
Scheme 2.
rationale that the a-NH in 1 would enhance the binding
anity through hydrogen bonding with the carboxylate
of Glu-270.
Figure 4. Schematic representation of binding modes of (a) l-1 and (b)
d-1 to the active site of CPA. In both cases, the hydroxyl group in the
form of anion forms a salt bridge with the active site zinc ion. In the case
of d-1, however, the hydroxyurea moiety experiences steric interactions
with the wall of the catalytic hole and as a result the hydroxyl is pushed
towards the zinc ion to have a stronger interaction with the zinc ion than
In order to ascertain the conjectural role of the N-
hydroxyaminocarbonyl moiety in 1 to form a chelate
coordinative bond to the zinc ion in binding of the inhi-
bitor to CPA, we have decided to evaluate the urea deri-
vative of Phe, i.e., 2. Compound 2 was prepared simply
by re¯uxing the mixture of phenylalanine benzyl ester
and sodium cyanate in THF followed by hydrogenolysis
to remove the benzyl group (Scheme 2).17
.
the corresponding interaction in the CPA l-1 complex.
to the imidazoles of His-69 and His-196, the carboxylate
of Glu-72, and a molecule of water is rested near the
core of CPA.21 It appears that the hydroxyurea moiety
of d-1 would project towards the core of the enzyme
when it binds CPA, but may hit the wall of the catalytic
hole. As a result, the moiety may be pushed towards the
zinc ion at the catalytic site and rest closer to the zinc ion
(Fig. 4b). The electrostatic interactions between the
deprotonated ureahydroxy and the zinc ion become for-
ti®ed. On the other hand, the hydroxyurea moiety of the
CPA bound l-1 orients near the entrance of the catalytic
hole again its hydroxyl being positioned towards the zinc
ion, but in this case it is not being pushed towards the
zinc ion because the moiety may be placed in open space
(Figure 4a). Accordingly, the two interacting species are
more distanced compared with that in the CPA complex
formed with d-1 and the binding force is reduced cor-
respondingly. Figure 4 illustrates schematically the
observed inhibitory stereochemistry. The rear atom in
the Newman projection in Figure 4 represents the a-
amino nitrogen.
The Ki value of rac-2 is increased by 28-fold compared
with that of rac-1 to be 58.5 mM, which suggests strongly
that the hydroxy group in the hydroxyurea moiety plays
a critical role in binding of 1 to CPA. Although the lit-
erature10 reports that hydroxyurea forms chelation
complexes with metal ions such as Fe(III) and Cu(II), it
was thought to be of interest to probe whether the
hydroxyurea moiety in 1 also coordinates to the active
site zinc ion in a bidentate fashion when the inhibitor
binds CPA. There are two acidic groups in 1 and we
found two pKa values of 2.22 and 6.16 that may be
attributed to the protonated carboxylate and the urea-
hydroxy groups, respectively. The pKa value of 6.16 for
the hydroxyurea suggests that under the assay conditions
the hydroxyurea moiety ligates the active site zinc ion in
a monodentate fashion rather than chelating the metal
ion as originally thought.
As can be seen from Table 1, d-1 is more potent than l-
1 by 3-fold. The reversal of stereochemistry in the inhi-
bition in comparison with stereochemistry of substrate
is of interest, although some precedences of such rever-
sal have been reported. Holmquist and Vallee reported
that N-mercaptoacetylphenylalanine obtained from
d-Phe is more potent than its l-enantiomer by about
23-fold.18 More recently, Jin and Kim19 reported that
N-formyl-N-hydroxy a-amino acid methyl esters
derived from d-Leu is about 70-fold more potent than
that prepared from l-Leu in inhibition of thermolysin,20
a zinc containing protease whose active site structure
and catalytic mechanism resemble those of CPA. How-
ever, in the case of compound 2 where the N-hydroxy is
lacking, the l-isomer was found to be much potent than
its enantiomer (Table 1), which suggests that the bind-
ing of 2 is resulted primarily from the interactions of the
carboxylate and the phenyl ring with the guanidinium
moiety of Arg-145 and the substrate recognition pocket,
respectively, as does substrate.
In summary, racemic and optically active N-hydroxy-
urea derivatives of phenylalanine, 1 were designed and
synthesized as inhibitors of CPA. The inhibitory kinetics
showed that they are indeed potent competitive inhibi-
tors for CPA. Surprisingly, d-1 is more potent than l-1.
Hydorxyurea derivatives of an amino acid or peptide
that can be readily prepared may thus be potentially
useful in the design of inhibitors that are e€ective for
zinc proteases of medicinal interest.
Experimental
Melting points were taken on a Thomas±Hoover capil-
1
lary melting point apparatus and were uncorrected. H
NMR and 13C NMR spectra were recorded on a Bruker
AM 300 (300 MHz) NMR spectrometer using tetra-
methylsilane as the internal standard. IR spectra were
obtained with a Bruker Equinox 55 FT-IR spectro-
photometer. Silica gel 60 (230±400 mesh) was used for
¯ash chromatography and thin layer chromatography
(TLC) was carried out on silica coated glass sheets (Merck
silica gel 60 F-254). Elemental analyses were performed at
The observed reversal of stereochemistry in the inhibi-
tion of CPA by the hydroxyurea inhibitors may be envi-
sioned on the basis of the unique position that the zinc ion
takes in the active site. The zinc ion that is coordinated

188
S. J. Chung, D. H. Kim, / Bioorg. Med. Chem. 9 (2001) 185±189
the Center for Biofunctional Molecules, Pohang Uni-
versity of Science and Technology, Pohang, Korea.
+ 16.99ꢀ (c 0.36, DMF). (R)-2: mp 191±192 ꢀC; [a]2D0
16.95ꢀ (c 0.59, DMF).
N-(Hydroxyaminocarbonyl)phenylalanine (1). A solution
of carbonyldiimidazole (810 mg, 5 mmol), O-benzyl-
hydroxylamine hydrochloride (800mg, 5 mmol), and
triethylamine (1.4 mL, 10mmol) in dichloromethane (40
mL) was stirred at room temperature for 1 h, and then
phenylalanine benzyl ester p-toluenesulfonate (2.15 g,
5 mmole) was added. After stirring for additional 4 h, the
reaction mixture was washed with 1 N HCl, saturated
aqueous solution of sodium bicarbonate, brine, and dried
over anhydrous sodium sulfate. After evaporation of the
solvent, the residue was chromatographed on a silica gel
column (30% ethyl acetate in hexane) to a€ord 1 (1.7 g,
84%) as benzyl ester. A solution of the benzyl ester
(1.5 g, 3.7 mmole) in methanol was stirred under hydro-
gen atmosphere in the presence of 10% Pd-C for 2 h at
room temperature and ®ltered. The ®ltrate was con-
centrated under the reduced pressure and ether was
added slowly to a€ord a crystalline product, which was
recrystallized from methanol-ether to give analytically
pure 1 (700mg, 84%). Mp 132±134 ꢀC; IR (KBr) 3290,
Determination of pKa
The pKa value of 1 was determined using a AT-400
Potentiometric Titrator (Kyoto Electronics Manufactur-
ing Co.). A solution of 1 (22.4 mg, 0.1 mmole) in water
(60 mL) was adjusted to pH 1.85 with 1 N HCl, and
titrated with 0.05 N aqueous NaOH solution. The pKa
values were obtained directly from the titrator.
General remarks for kinetic experiments
All solutions were prepared by dissolving in 0.5 M
NaCl, pH 7.5, 0.05 M Tris bu€er solution. Stock assay
solutions were ®ltered before use. Carboxypeptidase A
was purchased from Sigma Chemical Co. (Allan form,
twice recrystallized from bovine pancreas, aqueous sus-
pension in toluene) and used without further puri®cation.
CPA stock solutions were prepared by dissolving the
enzyme in the bu€er. Hippuryl-l-phenylalnine (Hipp-l-
Phe) purchased from Sigma Chemical Co. was used as
substrate for CPA. The change in the absorbance at
254 nM that is due to the generation of hippuric acid
form the substrate was followed at 25 ꢀC on a Perkin±
Elmer HP 8453 UV±vis spectrometer.
1
1654, 1635, 1541cm 1; H NMR (300MHz) d 8.46 (s,
1H), 7.27±7.16 (m, 5H), 6.52 (d, J=8.26 Hz, 1H), 4.37
(ddd, J=13.27, 7.85, 5.53Hz, 1H), 3.08±2.94 (m, 2H); 13C
NMR 174.28, 161.61, 138.46, 130.03, 129.06, 127.27,
54.08, 37.70. Anal. Calcd. for C10 H12N2O4: C, 53.57; H,
5.39; N, 12.49. Found: C, 53.93; H, 5.52; N, 12.40.
Determination of Ki
Compound 1 was prepared in the same fashion as the
above, starting with optically active phenylalanine ben-
zyl ester p-toluenesulfonate. (S)-1: mp 136±138 ꢀC; [a]D20
+42.17ꢀ (c 0.54, EtOH). (R)-1: mp 136±138 ꢀC;
[a]2D0 42.78ꢀ (c 1.2, EtOH);
The initial velocities were obtained from the linear plot
of the substrate hydrolysis monitored by following the
increase of absorbance at 254 nM. The Ki values were
then estimated from the semireciprocal plot of the initial
velocity versus the concentration of the inhibitors
according to the method of Dixon. Two or three con-
centrations of the substrate were used. Typically, the
enzyme stock solution was added to the solution of the
inhibitor and the substrate in the bu€er (1 mL cuvette),
and the rate of absorption increase at 254 nM was
recorded immediately.
N-(Aminocarbonyl)phenylalanine (2).
A mixture of
phenylalanine benzyl ester p-toluenesulfonate (4.27 g,
10 mmole) and sodium cyanate (2.6 g, 40 mmole) in
THF was re¯uxed for 4 h, and the solvent was evapo-
rated under reduced pressure. The residue was dissolved in
hot ethyl acetate and the insoluble material was removed
by ®ltration. The ®ltrate was chilled in an ice±water bath
to a€ord a crystalline solid, which was recrystallized
from ethyl acetate to give 2 as its benzyl ester (2.1 g,
70.4%). A solution of the benzyl ester (1.1 g, 3.69mmole)
in MeOH was stirred under hydrogen atmosphere in the
presence of 10% Pd-C (200 mg) for 2 h and ®ltered. The
®ltrate was concentrated to give a solid residue, which
was recrystallized from water to give an analytically
pure 2 (740 mg, 96%). Mp 182±183 ꢀC; IR (KBr) 3455,
Acknowledgements
The authors gratefully acknowledge the Korea Science
and Engineering Foundation for the generous support
of this work.
1
3313, 1697, 1636, 1558cm 1; H NMR (DMSO-d6) d
12.76 (s, 1H), 7.41±7.27 (m, 5H), 6.29 (d, J=8.22 Hz, 1H),
5.75 (s, 2H), 4.44 (ddd, J=13.14, 7.88, 5.34Hz, 1H), 3.12
(dd, J=13.73, 5.12Hz, 1H), 2.96 (dd, J=13.70, 7.88Hz,
1H); 13C NMR 174.02, 158.21, 137.56, 129.28, 128.22,
126.44, 53.75, 37.63. Anal. calcd for C10H12N2O3: C,
57.68; H, 5.81; N, 13.45. Found: C, 57.32; H, 6.16; N,
13.46.
References
1. Christianson, D. W.; Libscomb, W. N. Acc. Chem. Res.
1989, 22, 62.
2. Ondetti, M. A.; Rubin, B.; Cushman, D. W. Science 1977,
196, 441.
3. Kim, D. H.; Guinosso, C. J.; Buzby, G. C.; Herbst, D. R.;
McCaully, R. J. J. Med. Chem. 1983, 26, 394.
4. Ghosh, S. S.; Said-Nejad, O.; Roestamadji, J.; Mobashery,
S. J. Med. Chem. 1992, 35, 4175.
5. Browner, M. F.; Smith, W. W.; Castelhano, A. L. Bio-
chemistry 1995, 34, 6602.
Optically active 2 was prepared in the same fashion as the
above, starting with optically active phenylalanine benzyl
ester p-toluenesulfonate. (S)-2: mp 191±192 ꢀC; [a]D20

S. J. Chung, D. H. Kim, / Bioorg. Med. Chem. 9 (2001) 185±189
189
6. Nishino, N.; Powers, J. C. Biochemistry 1979, 18, 4340.
7. Gowravaram, M. R.; Tomczuk, B. E.; Johnson, J. S.;
Delecki, D.; Cook, E. R.; Ghose, A. K.; Mathiowetz, A. M.;
Spurlino, J. C.; Rubin, B.; Smith, D. L.; Pulvino, T.; Wahl, R.
C. J. Med. Chem. 1995, 38, 2570.
8. Turbanti, L.; Cerbai, G.; DiBugno, C.; Giorgi, R.; Garzelli,
G.; Criscuoli, M.; Renzetti, A. R.; Subissi, A.; Bramanti, G. J.
Med. Chem. 1993, 36, 699.
14. Christianson, D. W.; Lipscomb, W. N. Proc. Natl. Acad.
Sci. USA 1986, 83, 7568.
15. Dixon, M. Biochem., J. 1953, 55, 170.
16. Jin, Y. H.; PhD Thesis, Pohang Uiversity of Science and
Technology, Pohang, Korea, 1997, p 76.
17. Kehm, B. B.; Whitehead, C. W. Org. Synth. 1963, Coll. 4,
515.
18. Holmquist, B.; Valee, B. L. Proc. Natl. Acad. Sci. USA.
1979, 76, 6216.
19. Jin, Y.; Kim, D. H. Tetrahedron: Asymmetry 1997, 8,
3699.
9. Penning, T. D.; Askonas, L. J.; Djuric, S. W.; Haack, R.
A.; Yu, S. S.; Mickener, M. L.; Krivi, G. G.; Pylas, E. Y.
Bioorg. Med. Chem. Lett. 1995, 5, 2517.
10. Harmon, R. E.; Dabrowiak, J. C.; Brown, D. J.; Gupta, S.
K.; Herbert, M.; Chitharanjan, D. J. Med. Chem. 1970, 13, 577.
11. Brooks, C. D. W.; Summers, J. B. J. Med. Chem. 1996, 39,
2629.
12. Greco, M. N.; Hageman, W. E.; Powell, E. T.; Tighe, J. J.;
Persico, F. J. J. Med. Chem. 1992, 35, 3180.
20. Matthews, B. W. Acc. Chem. Res. 1988, 21, 333.
21. Rees, D. C.; Lewis, M.; Libscomb, W. N. J. Mol. Biol.
1983, 168, 367.
22. Byers, L. D.; Wolfenden, R. Biochemistry 1973, 12,
2070.
23. Ondetti, M. A.; Condon, M. E.; Reid, J.; Sato, E. F.;
Cheung, H. S.; Cushman, D. W. Biochemistry 1979, 18,
1427.
13. Summers, J. B.; Mazdiyasni, H.; Holms, J. H.; Ratajczyk,
J. D.; Dyer, R. D.; Carter, G. W. J. Med. Chem. 1987, 30, 574.