Characterization of α-nitromethyl ketone as a new zinc-binding group based on structural analysis of its complex with carboxypeptidase A

Article abstract of DOI:10.1016/j.bmcl.2009.07.060

Zinc-binding groups (ZBGs) are exhaustively applied in the development of the new inhibitors against a wide variety of physiologically and pathologically important zinc proteases. Here the α-nitro ketone was presented as a new ZBG, which is a transition-s

Full text of DOI:10.1016/j.bmcl.2009.07.060

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5009–5011
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Characterization of
a-nitromethyl ketone as a new zinc-binding group based
on structural analysis of its complex with carboxypeptidase A
b
a,
*
Shou-Feng Wang a, Guan Rong Tian a, , Wen-Zheng Zhang , Jing-Yi Jin
*
a Key Laboratory of Biological Resources and Functional Molecules of the Changbai Mountains of Ministry of Education, Yanbian University, Yanji, Jilin 133002, China
b Institute of Biophysics, Chinese Academy of Science, Beijing 100871, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Zinc-binding groups (ZBGs) are exhaustively applied in the development of the new inhibitors against a
Received 10 April 2009
Revised 22 June 2009
Accepted 9 July 2009
Available online 12 July 2009
wide variety of physiologically and pathologically important zinc proteases. Here the a-nitro ketone was
presented as a new ZBG, which is a transition-state analog featured by the unique bifurcated hydrogen
bonds at the active site of carboxypeptidase A based on the structural analysis. Introduction of a nitro
group at the
a-position of the ketone could provide more non-covalent interactions without loss of the
abilities to form a tetrahedral transition-state analog.
Keywords:
Ó 2009 Elsevier Ltd. All rights reserved.
Zinc-binding group
Crystal structure
Transition-state analog
Zinc proteases are a family of enzymes having a catalytically
essential zinc ion at their active sites. They include a variety of
physiologically and pathologically important species, such as
angiotensin-converting enzyme (ACE) and matrix metalloproteas-
es, which are the most studied drug design targets.1,2 The inhibitor
design strategies widely applied to these zinc proteases generally
make use of a zinc-binding group (ZBG) that can form coordinative
bonds to the zinc ion at the active site of the enzymes, which are
usually lessons from the inhibition of carboxypeptidase A (CPA,
EC 3.4.17.1) as a leading representative.3–5 For example, captopril,6
an inhibitor of ACE, was discovered by application of the design
strategy that originated from 2-benzylsuccinic acid as a potent
inhibitor of CPA.7 The other important implication is that the inter-
actions of ketones with CPA were regarded as a model for the
inhibition of ACE by carbonyl compounds.8 Introduction of an
Nitro is a well-known electron-withdrawing group; besides, it
could form various hydrogen bonds with the residues at the active
site of CPA as shown in the previous report.11 As expected, intro-
duction of the nitro group at the
a-position of the ketone should
not only remain the abilities to reach a TS mimic but also provide
additional interactions with the residues at the active site of CPA.
Therefore, we designed 2-benzyl-5-nitro-4-oxopentanoic acid 5
as the CPA inhibitor. Based on the structural analysis of the CPA-
5 complex, we here present a-nitromethyl ketone as a new TS ana-
log ZBG.
All three optically active forms of 5 (RS, R, S) were synthesized
as shown in Scheme 1.12 Their inhibitory activities against CPA
were evaluated and the corresponding inhibitory constant (Ki) val-
ues are collected in Table 1.13 Firstly, it could be found that the Ki
values of the series of ketone-based inhibitors against CPA de-
creased with the increase of the pKa values of the same substituted
a-substituent with strong electron-withdrawing abilities could
enhance the electrophilicity of the carbonyl, which facilitated the
addition of a nucleophile followed by the formation of a hybridized
adduct that mimics the transition state (TS) occurring during nor-
mal substrate hydrolysis catalyzed by CPA. For example, a-bromo
ketone compound is a TS analog inhibitor of CPA but the non-
modified ketone is a substrate analog inhibitor.9,10 Because the
TS analogs utilize the binding energies more effectively, such ZBGs
derived inhibitors have been in extensive attentions as potential
candidate of drug.1
* Corresponding authors. Tel.: +86 4332732286; fax: +86 4332732456 (J.-Y.J.);
tel.: +86 4332762289; fax: +86 4332732456 (G.R.T.).
E-mail addresses: grtn@ybu.edu.cn (G.R. Tian), jyjin-chem@ybu.edu.cn (J.-Y. Jin).
Scheme 1. Synthesis of 2-benzyl-5-nitro-4-oxopentanoic acid.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.060

5010
S.-F. Wang et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5009–5011
Table 1
14
Compound
Ki (l
M)
pKa
RS-7
RS-8
RS-99
RS-5
R-5
R1 = R2 = H
1 = Br, R2 = H
R1 = R2 = F
1 = NO2, R2 = H
207
15
0.2
0.43
0.16
31
4.76
2.90
0.50
1.68
R
R
S-5
acetic acids (Table 1),14 which suggest that the electron-withdraw-
ing abilities of the -substituents should be the key to the inhibi-
a
tory potency of the ketone-based inhibitors. Kinetic assay of
Figure 1. Difference electron density map for CPA-R-5 complex generated with
Fourier coefficient |Fo| ꢀ |Fc| phases calculated from the final model omitting the
bound inhibitor.
compound 5 as inhibitors against CPA also disclosed that R-5 is
the most potent inhibitor with inhibitory constant as 0.16
which is consistent to the -specificity of CPA.
We are mainly interested in the binding mode of the
lM,
L
a
-nitrom-
ethyl ketone at the active site of CPA. Crystals of the CPA-R-5 com-
plex for X-ray diffraction were then obtained by soaking CPA
crystals into the solution containing racemic 5 and determined at
a resolution of 1.85 Å (PDB code: 3FX6).15 The final model includ-
ing all residues of CPA and R-5 was refined in the range of 50.0–
1.85 Å (Table 2). Figure 1 depicts the stereoview of the difference
electron density in the region of the active site of CPA that is occu-
pied by R-5. Distances of important interactions between CPA and
R-5 in the complex are listed in Table 3. Structural analysis firstly
disclosed that R-5 occupies the S10 subsite of CPA, where the benzyl
moiety resides in the hydrophobic pocket and the carboxylate
makes a salt link with the guanidinium group of Arg-145. Tyr-
248 is found in the ‘so-called’ down position, which is responsible
to the formation of the hydrophobic pocket. Such binding modes
are commonly observed in X-ray crystal structures of CPA-inhibitor
complexes.16 It is also clearly observed that the ketonic hydrate of
the R-5 are bound at the active site of CPA. The so-called ‘gem-diol’
form with tetrahedron configuration exhibits the structurally
essential features of the transition state, that is, a bidentate chela-
tion to the zinc ion and two respective hydrogen bonds with Glu-
270 and Arg-127.17 Thus, the presented R-5 should be a TS analog
inhibitor of CPA. Besides the above structural features generally
Table 3
Selected CPA-R-5 interactions
0
Atom in R-5
Enzyme residue
Separation (ÅA)
O1
O2
O3
O4
O3
O4
O5
O5
Arg-145 guanidinium N1
Tyr-248 phenolic O
Zn2+
2.70
2.62
2.32
2.42
2.77
2.59
3.01
2.81
Zn2+
Arg-127 guanidinium N1
Glu-270 carboxylate O1
Arg-127 guanidinium N1
Arg-71 guanidinium N1
found in the complexes of CPA and its TS analog inhibitors, the
introduced nitro group also forms various hydrogen bonding with
the residues at the active site of CPA as expected. Both Arg-127 and
Arg-71 are involved in the hydrogen bond with the oxygen (O5) of
the nitro group. Thus, the bifurcated hydrogen bond is the other
salient structural feature of the CPA-R-5 complex, which should
be ascribed to the introduction of the nitro group (Fig. 2).
In conclusion, we successfully introduced the nitro group to the
ketone skeleton and then developed
new ZBG. The presented ketone-based ZBG not only is a TS analog
a-nitromethyl ketone as a
but also exhibits various hydrogen bonds with the residues at the
Table 2
Data collection and refinement statistics for the CPA-R-5 complex
Space group
Unit cell0
P21
a, b, c (AÅ)
73.36, 59.75, 99.51
a
, b,
c
(°)
90.00, 104.04, 90.00
0
Resolution range (ÅA)
Number of unique reflections
Overall completeness (%)
Rmerge (%)a
50.00–1.85
67,762
89.6
4.7
23.6
R factorb
Rms deviationsc
0
Bonds (AÅ)
Angles (°)
Dihedrals (°)
0.005
1.3
23.2
P
P
a
Rmerge for data sets for replicate reflections, R = ||Fhi| ꢀ h|Fh|i|/ h|Fh|i,
|Fhi| = scaled structure factor for reflection h in data set i, h|Fh|i = average structure
factor for reflection h calculated from replicate data.
P
P
b
R factor, R = ||Fo| ꢀ |Fc||/ |Fo|; |Fo| and |Fc| are the observed and calculated
structure factors, respectively.
Figure 2. Schematic representation of the important interactions of R-5 with the
active site of CPA.
c
Rms: root mean square.

S.-F. Wang et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5009–5011
5011
solutions were filtered (GHP Acrodic syringe filter, pore size 0.2
use. Perkin–Elmer HP 8453 UV–vis spectrometer was used for UV absorbance
measurements. The enzyme stock solution was added to a solution containing
l
m) before
active site of CPA, which may be helpful to further development of
new zinc proteases inhibitors.
Cl-CPL (final concentrations: 50 and 100 lM) and inhibitor (five different final
concentrations in the range of 0.5ꢀ2.0 Ki) in 0.05 M Tris/0.5 M NaCl, pH 7.5
buffer (1 mL cuvette), and the change in absorbance at 320 nm was measured
immediately. The final concentration of CPA was 20 nM. Initial velocities were
then calculated from the linear initial slopes of the change in absorbance
where the amount of substrate consumed was less than 10%. 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.18
14. Dean, J. A. In Lange’s Handbook of Chemistry, 15th ed.; McGraw-Hill: New York,
1999.
15. CPA was washed three times with water to remove toluene used in
packaging and then dissolved in 1.2 M LiCl. After centrifugation, the enzyme
solution was then diluted to 12–15 mg/mL with a buffer solution of 1.2 M
LiCl, 25 mM Tris, pH 7.5. An aliquot of this solution was then pipetted into a
Acknowledgement
The work was financially supported by the Science and Technol-
ogy Committee of Jilin Province, China (No. 20060568).
References and notes
1. Babine, R. E.; Bender, S. L. Chem. Rev. 1997, 97, 1359.
2. Leung, D.; Abbenante, G.; Fairlie, D. P. J. Med. Chem. 2000, 43, 305.
3. Jacobsen, F. E.; Lewis, J. A.; Cohen, S. M. ChemMedChem 2007, 2, 152.
4. Lipscomb, W. N.; Sträter, N. Chem. Rev. 1996, 96, 2375.
5. Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. H. Chem. Rev. 1999, 99,
2735.
6. Ondetti, M. A.; Rubin, B.; Cushman, D. W. Science 1977, 196, 441.
7. Byers, L. D.; Wolfenden, R. Biochemistry 1973, 12, 2070.
8. Sugimoto, T.; Kaiser, E. T. J. Am. Chem. Soc. 1978, 100, 7750.
9. Christianson, D. W.; Lipscomb, W. N. J. Am. Chem. Soc. 1986, 108, 4998.
10. Jin, J.-Y.; Wang, S.-F.; Xuan, W.; Sheng, J.-W.; Wang, S.-H.; Tian, G. R. Chin. J.
Chem. 2008, 26, 153.
100 lL glass dialysis button. A 7-kDa molecular weight cutoff membrane,
which was pre-washed by doubly distilled and deionized water followed by
equilibrated in 0.15 M LiCl, was placed over the button and secured with a
rubber O-ring. The button was then dialyzed against 0.2 M LiCl, 25 mM Tris,
pH 7.5, and at 4 °C. Crystals appeared after about 4 days on the bottom and
edges of the well. The crystals were then crosslinked using a buffer solution
containing 0.15 M LiCl and 0.02% (v/v) glutaraldehyde for 90 min. The
crystals were then transferred to
a soaking solution of the same buffer
11. Wang, S.-H.; Wang, S.-F.; Xuan, W.; Zeng, Z.-H.; Jin, J.-Y.; Ma, J.; Tian, G. R.
Bioorg. Med. Chem. 2008, 16, 3596.
containing 50 M racemic 5 and stored for 7 days in a cold room (4 °C)
l
before data collection. Single crystals grew in the P21 space group with three
molecules in one asymmetric unit. Crystallographic data were collected
using a Rigaku RA-Micro 7 Desktop Rotating Anode X-ray Generator with a
Cu target operating at 40 kV 0ꢂ 20 mA and R-Axis IV++ imaging-plate detector
at a wavelength of 1.5418 ÅA. A 0.5-mm collimator was used to keep the
whole crystal bathed in the X-ray beam. A total of 180 images with 1.0°
oscillation were collected. The collected intensities were indexed, integrated,
corrected for absorption, scaled and merged using HKL 2000. Diffraction
images were processed with the CCP4 program suite. The structure of native
12. Compound RS-5: oil; IR (film): 3446, 1735, 1705, 1560 cmꢀ1
;
1H NMR
(300 MHz, CDCl3) d 7.30–7.25 (m, 5H), 5.70 (s, 2H), 3.09 (dd, J = 5.4, 13.5 Hz,
1H), 2.93 (dd, J = 9.3, 17.4 Hz, 1H), 2.82 (dd, J = 8.3, 13.5 Hz, 1H), 2.65 (dd,
J = 3.6, 17.4 Hz, 1H); 13C NMR (75 MHz, CDCl3) d 195.5, 179.3, 137.4, 129.0,
128.9, 127.1, 83.2, 41.8, 39.9, 37.6. FT-ICR MS: Anal. Calcd for C12H13NO5:
251.0794. Found: 250.0740 [MꢀH].Compound R-5: ½a D20
ꢁ
= + 29.5 (c, 0.5, CHCl3);
FT-ICR MS: Anal. Calcd for C12H13NO5: 251.0794. Found: 250.0743
[MꢀH].Compound S-5: ½a D20
= ꢀ28.8 (c, 0.6, CHCl3); FT-ICR MS: Anal. Calcd
ꢁ
19
for C12H13NO5: 251.0794. Found: 250.0744 [MꢀH].
CPA (PDB code 1M4L) was used as the starting model. The programs COOT
20
13. Determination of Ki value. CPA was purchased from Sigma Chemical Co. (Allan
form, twice crystallized from bovine pancreas, aqueous suspension in toluene)
and used without further purification for kinetic assays. Lithium chloride and
and CNS were used in the model building and refinement, respectively. The
randomly selected 5% of the data were set aside for the Rfree calculation.
Water molecules were gradually added to the model with CNS.
16. Christianson, D. W.; Lipscomb, W. M. Acc. Chem. Res. 1989, 22, 62.
17. Kim, H.; Lipscomb, W. N. Biochemistry 1990, 29, 5546.
18. Dixon, M. Biochem. J. 1953, 55, 170.
19. Emsley, P.; Cowtan, K. Acta Crystallogr., Sect. D 2004, 60, 2126.
20. Brunger, A. T.; Adams, P. D.; Clore, G. M. Acta Crystallogr., Sect. D 1998, 54, 905.
Tris were obtained from Sigma. O-(trans-p-Chlorocinnamoyl)-L-b-
phenyllactate (Cl-CPL) was used as the substrate in the kinetic study. All
solutions for kinetic study were prepared by dissolving in doubly distilled and
deionized water. CPA stock solutions were prepared by dissolving CPA in
0.05 M Tris/0.5 M NaCl, pH 7.5 buffer solution and their concentrations were
estimated from the absorbance at 278 nm (e278 = 64,200). The stock assay