Optical 2-benzyl-5-hydroxy-4-oxopentanoic acids against carboxypeptidase A: Synthesis, kinetic evaluation and X-ray crystallographic study

Article abstract of DOI:10.1016/j.cclet.2009.09.005

2-Benzyl-5-hydroxy-4-oxopentanoic acid 1 and its enantiomers were designed, synthesized and assayed for inhibitory activity against carboxypeptidase A (CPA, EC 3.4.17.1). To verify the role of the terminal hydroxyl group in 1 binding to CPA, 2-benzyl-5-benzyloxy-4-oxopentanoic acid 2 was also synthesized and evaluated. The inhibition constants show that both L-1 and D-1 were shown to have strong binding affinity with L-1 being more potent than its enantiomer by 165-fold. On the other hand, the inhibition constant of 2 increases 4-fold comparing with that of 1. In order to explore the exact binding mode of the hydroxyacteyl group of 1 to the active site zinc ion of CPA, we have solved the crystal structure of CPA complexed with L-1 up to 1.85 ? resolution. In CPA·L-1 complex, the phenyl ring is fitted in the substrate recognition pocket at the S^′₁ subsite, and the carboxylate forms bifurcated hydrogen bonds with the guanidinium moiety of Arg-145 and Arg-127 and a hydrogen bond with the phenolic hydroxyl of the down-positioned Tyr-248. The carbonyl oxygen of L-1 does coordinate to the active site zinc ion of CPA as expectedly. Unexpectedly, the terminal hydroxyl group of L-1 is engaged in hydrogen bonding with carbonyl oxygen of Ser-197 instead of coordinating to the active site zinc ion.

Full text of DOI:10.1016/j.cclet.2009.09.005

Available online at www.sciencedirect.com
Chinese Chemical Letters 21 (2010) 159–162
www.elsevier.com/locate/cclet
Optical 2-benzyl-5-hydroxy-4-oxopentanoic acids against
carboxypeptidase A: Synthesis, kinetic evaluation and X-ray
crystallographic study
Shou Feng Wang a, Jing-Yi Jin a, Zong Hao Zeng b, Guan Rong Tian a,
*
a Key Laboratory of Organism Functional Factors of Changbai Mountains of Ministry of Education,
Yanbian University, Yanji 133002, China
b Institute of Biophysics, Chinese Academy of Science, Beijing 100871, China
Received 17 June 2009
Abstract
2-Benzyl-5-hydroxy-4-oxopentanoic acid 1 and its enantiomers were designed, synthesized and assayed for inhibitory activity
against carboxypeptidase A (CPA, EC 3.4.17.1). To verify the role of the terminal hydroxyl group in 1 binding to CPA, 2-benzyl-5-
benzyloxy-4-oxopentanoic acid 2 was also synthesized and evaluated. The inhibition constants show that both L-1 and D-1 were
shown to have strong binding affinity with L-1 being more potent than its enantiomer by 165-fold. On the other hand, the inhibition
constant of 2 increases 4-fold comparing with that of 1. In order to explore the exact binding mode of the hydroxyacteyl group of 1 to
˚
the active site zinc ion of CPA, we have solved the crystal structure of CPA complexed with L-1 up to 1.85 A resolution. In CPAꢀL-1
complex, the phenyl ring is fitted in the substrate recognition pocket at the S01 subsite, and the carboxylate forms bifurcated hydrogen
bonds with the guanidinium moiety of Arg-145 and Arg-127 and a hydrogen bond with the phenolic hydroxyl of the down-
positioned Tyr-248. The carbonyl oxygen of L-1 does coordinate to the active site zinc ion of CPA as expectedly. Unexpectedly, the
terminal hydroxyl group of L-1 is engaged in hydrogen bonding with carbonyl oxygen of Ser-197 instead of coordinating to the
active site zinc ion.
# 2009 Guan Rong Tian. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
Keywords: 2-Benzyl-5-hydroxy-4-oxopentanoic acid; Carboxypeptidase A; X-ray crystallographic study
Zinc proteases, which are widely found in a variety of tissues, play important roles in numerous physiological and
pathological processes. The most studied drug design targets of zinc proteases are angiotensin-converting enzyme [1],
enkephalinase [2] and matrix metalloproteinases [3]. Inhibitors designed for most zinc proteases have focused on
small molecules capable of interacting with the primary recognition pocket, a peptidomimetic backbone and
incorporating a zinc-binding group (ZBG). The general strategy of incorporating a ZBG, such as carboxylic acid, thiol,
hydroxamic acid and phosphorus, for the designed inhibitors has been proven to be successful [4]. However, the
exploitation and discovery of more potent and selective ZBGs is still required [5].
* Corresponding author.
E-mail address: grtn@ybu.edu.cn (G.R. Tian).
1001-8417/$ – see front matter # 2009 Guan Rong Tian. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
doi:10.1016/j.cclet.2009.09.005

160
S.F. Wang et al. / Chinese Chemical Letters 21 (2010) 159–162
Carboxypeptidase A (CPA, EC 3.4.17.1), the first zinc protease whose crystal structure was solved by X-ray
crystallography, is one of the most extensively studied zinc-containing proteolytic enzymes and serves as a
prototypical enzyme to exploit ZBGs in the last 30 years [6]. Lee and Kim [7] designed and synthesized
hydroxyacetyl-based inactivators of CPA based on the crystal structure of CPA complexed with its substrate Gly-L-Tyr
(glycyl-L-tyrosine) in 1998 and the hydroxyacetyl group was thought to chelate the active site zinc ion. Recently
´
Slepokura and Lis [8] reported the crystal structures of dihydroxyacetone (DHA) monomer complexes with CaBr2 and
CdCl2. The crystal structures show that DHA molecules chelate the cations via both the hydroxyl and carbonyl groups.
However, the binding mode of hydroxyacetyl group to the active site zinc ion of zinc proteases is still not solved. It is
therefore thought to be necessary and interesting to explore the binding mode of the hydroxyacetyl group to the active
site zinc ion for further rational designing inhibitors of zinc proteases.
We reported herein the synthesis of 2-benzyl-5-hydroxy-4-oxopentanoic acid 1 and its enantiomers as inhibitors of
CPA. To verify the role of the terminal hydroxyl group in 1 binding to CPA, 2-benzyl-5-benzyloxy-4-oxopentanoic
acid 2, the analogue of 1, was also synthesized and evaluated. To further explore the exact binding mode of the
hydroxyacetyl group in 1 to the active site zinc ion of CPA, the complex of CPAꢀL-1 was performed and its crystal
structure was investigated using X-ray diffraction method.
Compound 1 was prepared utilizing 2-benzylsuccinic acid mono tert-butyl ester as starting material and its
enantiomers were prepared following the same procedure as for 1 starting with optical 2-benzylsuccinic acid mono
tert-butyl ester. Replacing the terminal bromo group of 4 with sodium formate dihydrate [9] followed by hydrolysis
[10] resulted in each of the forms of 1 required (Scheme 1). Compound 2 was synthesized by allowing benzyl bromide
to react with 5 in presence of freshly prepared Ag2O [11] followed by hydrolysis in acidic condition [10] (Scheme 1).
The prepared compounds were evaluated as CPA inhibitors by a standard procedure [15] to find that they inhibit the
enzyme in a competitive reversible fashion as can be seen from Fig. 1.
Their inhibitory constants (Kis) obtained from Dixon’s plots (Fig. 1) are listed in Table 1. All of these compounds
should be active-site-directed inhibitors because they satisfy the essential requirements for CPA recognition, i.e., a
terminal carboxylate, a hydrophobic side chain and a zinc-binding group [12]. The finding that the L-1
(Ki = 0.86 mmol Lꢁ1) shows 165-fold more potent inhibition against CPA than the corresponding D-1
(Ki = 143 mmol Lꢁ1) is consistent with the L-specificity of CPA [13]. The inhibition constant of rac-2
(Ki = 42 mmol Lꢁ1) increases 4-fold comparing with that of rac-1 (Ki = 8.0 mmol Lꢁ1), which suggests strongly
that the hydroxyl group of rac-1 must involve in additional binding interaction in the inhibitor binding to CPA.
˚
The X-ray crystal structure of the CPAꢀL-1 complex has been refined at 1.85 A resolution to a final R-factor of
0.238. The atomic coordinates and structure factors of the CPAꢀL-1 complex have been deposited with the Brookhaven
Protein Data Bank (PDB ID 3FVL).
The final model of the complex exhibits the details of L-1 binding to CPA (Fig. 2). The carboxylate of L-1 is
˚
engaged in bifurcated hydrogen bonding with one of nitrogen atoms in the guanidinium moiety of Arg-145 (3.15 A)
˚
and Arg-127 (2.76 A). The aromatic side chain of Tyr-248 is found in the ‘‘so-called’’ down position and its phenolic
˚
hydroxyl is engaged in hydrogen bonding with the terminal carboxylate of L-1 (2.54 A) and the zinc-bound water
molecule that is found in the native CPA is absent in the complex (Fig. 2, right). The phenyl ring of L-1 is
accommodated in the S01 hydrophobic pocket, the primary recognition site of CPA (Fig. 2, left). These binding
Scheme 1. Prepared 1 and 2: (a) (i) ClCOOBuꢁt, N-Me morphorine, Et2O, ꢁ10 8C, 15 min; (ii) CH2N2, Et2O, ꢁ10 8C, 15 min and then r.t.,
overnight; (iii) HBr (48%), CH2Cl2, r.t., 15 min; (b) HCO2Naꢀ2H2O, EtOH, reflux, 12 h; (c) TFA, CH2Cl2, r.t., 24 h; (d) BnBr, Ag2O, CH2Cl2, r.t.,
12 h.

S.F. Wang et al. / Chinese Chemical Letters 21 (2010) 159–162
161
Fig. 1. The Dixon plot.
Table 1
Inhibitory constants for CPA inhibition.
Compounds
Ki (mmol Lꢁ1
)
rac-1
L-1
D-1
8.0
0.86
143
42
rac-2
˚
Fig. 2. The stereo-view of the CPAꢀL-1 complex (left) and the interactions of L-1 with the residue of CPA [distance (A)] (right).
interactions are reminiscent of those reported the complexes of CPAwith its inhibitors [14]. It is interestingly to notice
˚
that the carbonyl oxygen of L-1 is separated from the active site zinc ion by 2.25 A, which suggests that it is ligated to
˚
the zinc ion. However, the terminal hydroxyl group of L-1 is separated from the active site zinc ion by 4.04 A,
suggesting that it is not ligated to the zinc ion. In addition, its carbonyl oxygen is involved in hydrogen bonding with
˚
one of carboxylate oxygen atoms of Glu-270 with bonding distance of 2.63 A. Most interestingly, a hydrogen bond is
˚
formed between the terminal hydroxyl group and the carbonyl oxygen of Ser-197 with bonding distance of 2.71 A

162
S.F. Wang et al. / Chinese Chemical Letters 21 (2010) 159–162
(Fig. 2, right). It should attribute to the additional hydroxyl-carbonyl-type hydrogen bond that makes the inhibitory
activity of rac-1 against CPA more potent than that of rac-2. The binding mode of hydroxyacetyl group of L-1 with
CPA presented is not consistent with the binding mode proposed by Lee and Kim [7] or the exact binding mode of
´
DHA to metal ion such as Cd (II) and Ca (II) reported by Slepokura and Lis [8].
In conclusion, compound 1 and its enantiomers capable of hydroxyacetyl group were synthesized and evaluated.
The inhibitory constants show that L-1 has more potent binding affinity than its corresponding D-1 by 165-fold. The
crystal structure of CPAꢀL-1 complex reveals that L-1 does with its carbonyl oxygen coordinating to the active site zinc
ion of CPA. Unexpectedly, the terminal hydroxyl group of L-1 is involved in hydrogen bonding with the carbonyl
oxygen of Ser-197 instead of coordinating to the active site zinc ion, which is not consistent with the binding mode
proposed in previous papers. The design strategy demonstrated and the exact binding mode of hydroxyacetyl group of
L-1 to the active site zinc ion of CPA in this work may be of considerable interest to those who are engaged in the
rational design of inhibitors for zinc-containing proteases.
References
[1] R. Behnia, A. Molteni, Curr. Pharm. Des. 9 (2000) 763;
H.P. Lefebvre, S.A. Brown, V. Chetboul, Curr. Pharm. Des. 13 (2007) 1347.
[2] V. Thanawala, V.J. Kadam, R. Ghosh, Curr. Drug Targets 9 (2008) 887.
[3] J.W. Skiles, N.C. Gonnella, A.Y. Jeng, Curr. Med. Chem. 8 (2001) 425;
B.G. Rao, Curr. Pharm. Des. 11 (2005) 295.
[4] S.P. Gupta, Chem. Rev. 107 (2007) 3042.
[5] F. Mannello, G. Tonti, S. Papa, Curr. Cancer Drug Targets 5 (2005) 285.
[6] W.N. Lipscomb, N. Stra¨ter, Chem. Rev. 96 (1996) 2375;
D.H. Kim, Curr. Top. Med. Chem. 4 (2004) 1217.
[7] D.H. Kim, K.J. Lee, Bioorg. Med. Chem. Lett. 7 (1997) 2607;
K.J. Lee, D.H. Kim, Bioorg. Med. Chem. 6 (1998) 1613.
´
[8] K. Slepokura, T. Lis, Acta Cryst. C 64 (2008) m127.
[9] Z.G. Tang, H.Q. Tian, D.W. Ma, Tetrahedron 60 (2004) 10547.
[10] C.B. Stephen, A.I. Khadiga, Synthesis 2000 (2000) 1369.
[11] B. Abderrahim, S. Gilles, Tetrahedron Lett. 38 (1997) 5945.
[12] D.W. Christianson, W.N. Lipscomb, Acc. Chem. Res. 22 (1989) 62.
[13] D.H. Kim, S. Chung, Tetrahedron: Asymmetry 10 (1999) 3769.
[14] S. Mangani, Coordin. Chem. Rev. 120 (1992) 309.
[15] Determination of Ki value. CPA was purchased from Sigma Chemical Co. (Allan form, twice crystallized from bovine pancrease, aqueous
suspension in toluene) and used without further purification. CPA stock solutions were prepared by dissolving the enzyme in 0.05 mol Lꢁ1 Tris/
0.5 mol Lꢁ1 NaCl, pH 7.5 buffer solution. O-(trans-p-chlorocinnamoyl)-L-b-phenyllactate (Cl-CPL), which was synthesized as described in
the literature [16], was used as substrate. Typically, the enzyme stock solution was added to a solution containing Cl-CPL (final concentrations:
50 mmol Lꢁ1 and 100 mmol Lꢁ1) and inhibitor (five different final concentrations in the range of 0.5Ki to 2Ki mmol Lꢁ1) in 0.05 mol Lꢁ1 Tris/
0.5 mol Lꢁ1 NaCl, pH 7.5 buffer solution (1 mL cuvette), and the change in absorbance at 320 nm was measured immediately using a Perkin–
Elmer HP 8453 UV/Vis spectrometer at 25 8C. Initial velocities were then calculated from the linear initial slopes of the change in absorbance.
The Ki values were then estimated from the semireciprocal plot of the initial velocity versus the concentration of the inhibitor according to the
method of Dixon [17]. The correlation coefficients for the Dixon plots were above 0.990.
[16] J. Suh, E.T. Kaiser, J. Am. Chem. Soc. 98 (1976) 1940.
[17] M. Dixon, Biochem. J. 55 (1953) 170.