Hydrolysis of organophosphate esters: Phosphotriesterase activity of metallo-ss-lactamase and its functional mimics

Article abstract of DOI:10.1002/chem.201000282

The phosphotriesterase (PTE) activity of a series of binuclear and mononuclear zinc(II) complexes and metallo-β-lactamase (mβl) from Bacillus cereus was studied. The binuclear complex 1, which exhibits good mβl activity, shows poor PTE activity. In contrast, complex 2, a poor mimic of mβl, exhibits much higher activity than 1. The replacement of Cl ^- ligands by OH^- is important for the high PTE activity of complex 2 because this complex does not show any catalytic activi-ty in methanol. The natural enzyme mssl from B. cereus is also found to be an inefficient catalyst in the hydrolysis of phosphotriesters. These observations indicate that the binding of -lactam substrates at the binuclear zinc(II) center is different from that of phosphotriesters. Furthermore, phospho diesters, the products from the hydrolysis of triesters, significantly inhibit the PTE activity of mβl and its functional mimics. Although the mononuclear complexes 3 and 4 exhibited significant mβl activity, these complexes are found to be almost inactive in the hydrolysis of phosphotriesters. These observations indicate that the elimination of phosphodiesters from the reaction site is important for the PTE activity of zinc(II) complexes.

Full text of DOI:10.1002/chem.201000282

DOI: 10.1002/chem.201000282
Hydrolysis of Organophosphate Esters: Phosphotriesterase Activity of
Metallo-b-lactamase and Its Functional Mimics
A. Tamilselvi and Govindasamy Mugesh*[a]
Abstract:
The
phosphotriesterase
ty in methanol. The natural enzyme
mbl from B. cereus is also found to be
an inefficient catalyst in the hydrolysis
of phosphotriesters. These observations
indicate that the binding of b-lactam
substrates at the binuclear zinc(II)
center is different from that of phos-
photriesters. Furthermore, phospho-
diesters, the products from the hydroly-
sis of triesters, significantly inhibit the
PTE activity of mbl and its functional
mimics. Although the mononuclear
complexes 3 and 4 exhibited significant
mbl activity, these complexes are found
to be almost inactive in the hydrolysis
of phosphotriesters. These observations
indicate that the elimination of phos-
phodiesters from the reaction site is
important for the PTE activity of
zinc(II) complexes.
(PTE) activity of a series of binuclear
and mononuclear zinc(II) complexes
and metallo-b-lactamase (mbl) from
Bacillus cereus was studied. The binu-
clear complex 1, which exhibits good
mbl activity, shows poor PTE activity.
In contrast, complex 2, a poor mimic of
mbl, exhibits much higher activity than
1. The replacement of ClÀ ligands by
OHÀ is important for the high PTE ac-
tivity of complex 2 because this com-
plex does not show any catalytic activi-
Keywords: enzyme mimics
· lac-
tams · organophosphates · struc-
ture–activity relationships · zinc
Introduction
A comparison of the active site of PTE to that of other bi-
nuclear metal-containing enzymes reveals that PTE and the
nickel-containing urease have very similar coordination en-
vironments around the metal centers.[9] In both the cases,
each of the two metal ions is coordinated by two histidine
residues from the protein backbone. Whereas an aspartate
(Asp) residue is asymmetrically bound to one of the metal
ions, a carbamylated lysine and a hydroxide group act as
bridging ligands. The distances between two metal centers
(Zn···Zn in PTE: 3.4 ꢀ; Ni···Ni in urease: 3.5 ꢀ) are also
almost identical. The close structural resemblance of these
two proteins suggests that they may have a distant evolu-
tionary relationship.
Recently, Dong et al. have isolated a zinc-containing
methyl parathion hydrolase (MPH) from soil-dwelling bac-
teria that catalyzes the degradation of the organophospho-
rus pesticide methyl parathion.[10] Interestingly, the active
site of MPH has been shown to be very similar to that of
metallo-b-lactamase (mbl), which uses a binuclear zinc site
(Scheme 1) to hydrolyze the b-lactam ring in penicillins,
cephalosporins, and carbapenems. It is, however, not clear
whether mbl can hydrolyze organophosphorus triesters.
Herein, we show, for the first time, that synthetic zinc(II)
complexes that exhibit good mbl activity are poor mimics of
PTE and vice versa. We also show that the mbl from Bacil-
Synthetic organophosphorus compounds have been used ex-
tensively as pesticides, petroleum additives, and plasticiz-
ers.[1] In particular, organophosphorus triesters constitute a
major class of crop protectants and their widespread use in
agriculture leads to serious environmental problems.[2,3]
These compounds are highly toxic to mammals and they
affect the nerve system by inhibiting acetylcholine esterase
(AChE), a key enzyme involved in the breakdown of acetyl-
choline.[4] Therefore, degradation of organophosphorus
triesters and remediation of associated contaminated sites
are of worldwide concern.[5,6] In this regard, the bacterial
phosphotriesterase (PTE) enzyme plays an important role in
degrading a wide range of organophosphorus esters, includ-
ing agriculture pesticides, such as paraoxon and parathion,
and chemical warfare agents, such as sarin.[7,8]
[a] Dr. A. Tamilselvi, Prof. Dr. G. Mugesh
Department of Inorganic & Physical Chemistry
Indian Institute of Science, Bangalore 560 012 (India)
Fax : (+91)80-2360-0683/2360-1552
E-mail: mugesh@ipc.iisc.ernet.in
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201000282.
8878
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 8878 – 8886

FULL PAPER
Scheme 1. The chemical structures of the active sites of a) mbl from Ba-
cillus cereus (PDB code: 1BC2)[11] and b) PTE from Pseudomonas dimin-
uta (PDB code: 1HZY).[12]
Scheme 2. The hydrolysis of PNPDPP to DPP and p-nitrophenol
lus cereus is an inefficient catalyst in the degradation of par-
(PNP)[14] in the presence of zinc(II) complexes.
athion and related compounds.
PNPDPP (d=À18.8 ppm) and an increase in the peak for
DPP (d=À12.3 ppm) was observed (Figure S1 in the Sup-
porting Information). The rate of formation of DPP, howev-
er, decreased over a period of time and the reaction did not
proceed to completion.
In contrast to complex 1, the binuclear zinc(II) complex 2,
containing chloride ligands, showed excellent PTE activity.
When the reaction was carried out in CHES buffer at
pH 9.0, complex 2 efficiently catalyzed the hydrolysis of
PNPDPP to produce DPP (Figure 1). After 14 h, the com-
plete disappearance of the peak corresponding to PNPDPP
was observed (Figure S2 in the Supporting Information).
Results and Discussion
Phosphotriesterase activity of zinc(II) complexes: We have
recently reported the metallo-b-lactamase activity of the bi-
nuclear complexes 1 and 2 and the mononuclear complexes
3 and 4.[13] A detailed study of the hydrolysis of b-lactam
ring in penicillin and oxacillin mediated by these complexes
(1–4) has revealed the structural requirements for good mbl
activity. The binuclear complex 1, containing a water mole-
cule, exhibited much higher mbl activity than complex 2,
which does not have any water molecules but has labile
chloride ligands. We have also shown that the mononuclear
complexes 3 and 4 with coordinated water molecules exhibit
significant mbl activity. The higher activity of 3 and 4 com-
pared with ZnACHTUNGTRENNUNG(OAc)2·6H2O indicated that the activation of
water molecules in complexes 3 and 4 by the N,N-dimethyl
amino group is responsible for their higher catalytic activity.
In the present work, we studied the PTE activity of com-
plexes 1–5 to understand the correlation between the mbl
and PTE activities.
The PTE activity of complexes 1–5 was studied by
31P NMR spectroscopy and reverse-phase HPLC methods.
In the 31P NMR spectra, the disappearance of signals for the
phosphotriester substrate p-nitrophenyl diphenylphosphate
(PNPDPP) and the appearance of signals for diphenylphos-
phate (DPP) were followed in CHES buffer at pH 9.0
(Scheme 2). Although complex 1 exhibited excellent mbl ac-
tivity,[13] this complex showed very poor activity in the hy-
drolysis of PNPDPP. When complex 1 (20 mm) was added to
PNPDPP, a considerable decrease in the peak due to
Figure 1. The 31P NMR spectra obtained for the reaction of PNPDPP
(20 mm) with complex 2 (20 mm) in CHES buffer (200 mm) at pH 9.0.
Chem. Eur. J. 2010, 16, 8878 – 8886
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8879

G. Mugesh and A. Tamilselvi
The binuclear zinc complex 5, with two independent zinc(II)
centers,[15] exhibited much lower catalytic activity than 2,
however, this complex was found to be a better catalyst
than 1 in the hydrolysis of PNPDPP (Figure S3 in the Sup-
porting Information). The mononuclear complexes 3 and 4
did not show any appreciable activity in CHES buffer. It
should be noted that complexes 3 and 4 exhibited significant
mbl activity in the hydrolysis of b-lactam substrates such as
penicillin and oxacillin.[13]
observed at d=À10.9 ppm is not due to the hydrolyzed
product (DPP), but due to the formation of another phos-
photriester, diphenyl methylphosphate (DPMP), by metha-
nolysis as shown in Scheme 3. Brown and co-workers have
The replacement of the labile chloride ligands by water
(hydroxide) ligands appears to be responsible for the PTE
activity of complexes 2 and 5. Although complexes 2 and 5
are quite stable in acetonitrile, they are slowly converted
into complexes 6 and 7, respectively, in aqueous solution, as
evidenced by ESI–MS analysis. The protonation state of
zinc-bound water/hydroxide, however, depends on the pH of
the solution. The formation of binuclear metal complexes
with terminal hydroxide ligands has been well documented
in the literature.[16]
Scheme 3. The methanolysis of PNPDPP to DPMP and p-nitrophenol
(PNP) in the presence of complex 1.
shown in a series of studies that the methanolysis of phos-
photriesters by transition-metal complexes are generally
much faster than the hydrolysis reactions.[17] They have also
shown that simple zinc(II) salts, such as Zn
ACHTUNGRTENUN(NG OTf)2 and Zn-
AHCTUNTGRENNG(UN ClO4)2, can efficiently degrade paraoxon and fenitrothion
by a methanolysis pathway.[18] Furthermore, alcoholysis reac-
tions have been shown to proceed with different selectivities
than hydrolysis. For example, the reactions of some nerve
agents with methoxide proceed largely to displace the thio-
late group, leading to the formation of the corresponding
methoxyesters.[19–20] The selection of substrates for such stud-
ies is important because the alcoholysis reactions produce a
new phosphotriester instead of a phosphodiester.
In contrast to 1, complex 2 did not show any noticeable
activity when the reaction was carried out in methanol (Fig-
ure S4 in the Supporting Information). This is presumably
due to the high stability of complex 2 in methanol. Since the
zinc-coordinated chloride ligands are intact in methanol,
there is no nucleophilic hydroxyl or methoxy group in the
À
Interestingly, complex 1 very effectively catalyzed the
complex to cleave the P O bonds. Complex 2 was also
À
cleavage of the P O bond in PNPDPP in methanol. At
found to be inactive in hydrolyzing PNPDPP when DMSO
was used as the solvent. The HPLC and mass spectral analy-
ses indicated that the chloride ligands in 2 remain coordinat-
ed to zinc in methanol. The binuclear zinc complex 5, with
two independent zinc(II) centers, was also found to be a rel-
atively poor catalyst in the methanolysis of PNPDPP. Al-
though the chloride ligands in complex 5 are slowly replaced
by methoxide (or methanol), only a small amount of
PNPDPP was converted to DPMP after 150 min (Figure S5
in the Supporting Information).
room temperature,
a complete disappearance of the
PNPDPP peak at d=À18.2 ppm was observed within ap-
proximately 300 min (Figure 2). The analysis of the products
obtained from the reaction mixture indicated that the signal
In agreement with the 31P NMR spectroscopic experi-
ments, the HPLC data also indicate that complex 2 is more
active than complex 1. The plots of the percentage of
PNPDPP versus time for the hydrolysis in the presence of
zinc(II) complexes are shown in Figure 3 and the corre-
sponding t50 values (the time required for the 50% hydroly-
sis of the substrate) for PNPDPP are summarized in
Table 1. These data indicate that complexes 2 and 5 hydro-
lyze PNPDPP with t50 values of 204.9 and 1282.1 min, re-
spectively, in CHES buffer, and these values are much lower
than that of the control reaction (t50 >4400 min). Further-
more, the complete hydrolysis of the substrate (100% hy-
drolysis) was observed with complexes 2 and 5. The free
Figure 2. The 31P NMR spectra obtained for the methanolysis of
PNPDPP (20 mm) by complex 1 (20 mm).
ligand
(2,6-bis[(dimethylamino)methyl]-4-methylphenol)
8880
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 8878 – 8886

The Hydrolysis of Organophosphate Esters
FULL PAPER
hancement in the initial rate was observed when the solvent
was changed from methanol to buffer (Table 1 and
Figure 3). Furthermore, complex 2 was found to be active at
concentrations as low as 10 mol% (Figure S6 in the Sup-
porting Information), indicating the catalytic nature of the
hydrolysis. The other binuclear zinc(II) complex 5 exhibits
good activity both in CHES buffer and methanol, although
the rate of hydrolysis in buffer is about two times higher
than that of the methanolysis.
Effect of the phosphodiesters on PTE activity: The relative-
ly poor catalytic activity of complex 1 in buffer is probably
due to the inhibition of the reaction by the product (DPP).
It is possible that DPP can block the zinc(II) coordination
sites in complex 1 by forming a bridge between the two
metal ions. In such a case, complex 1 cannot bring the hy-
drolysis reaction to completion once a considerable amount
of DPP has been generated in the reaction. It should be
noted that Anslyn and co-workers have recently proposed
the formation of a phosphodiester-bridged binuclear zinc(II)
complex during the transesterification reaction.[21] On the
other hand, Lippard and co-workers reported that DPP can
bind strongly to binuclear zinc(II) complexes, both as a ter-
minal and as a bridging ligand.[22] To understand the binding
of DPP to complex 1, we carried out the hydrolysis of
PNPDPP by complex 1 in the presence of DPP. The reac-
tions were followed by 31P NMR spectroscopy and ESI–MS.
The MS analysis of the reaction mixture containing complex
1, PNPDPP, and DPP indicated the replacement of the
bridging acetate ligand in complex 1 by DPP anion leading
to the formation of complex 8 (Scheme 4). The 31P NMR
Figure 3. The hydrolysis of PNPDPP (0.20ꢂ10À3 m) by a) complex
2
(0.21ꢂ10À3 m), b) complex 5 (0.21ꢂ10À3 m), c) complex 1 (0.21ꢂ10À3 m),
and d) the control A) ethanol (35%) or B) methanol was used in 20 mm
CHES buffer, pH 9.0. To increase the solubility of the PNPDPP in the re-
action mixture, The reaction was monitored by reverse-phase HPLC and
the amount of p-nitrophenol at any given time was calculated from the
calibration plot.
Table 1. The t50 values and initial rates (v0) of hydrolysis/methanolysis of
PNPDPP by complexes 1, 2, and 5 at RT.
Complex
In CHES buffer, pH 9.0
Methanol
v0 [mmminÀ1
]
[c]
[c]
t50 [min][a]
v0 [mmminÀ1
]
t50 [min][a]
1
2
5
3103.0
204.9
4.50ꢂ10À2
55.89ꢂ10À2
5.91ꢂ10À2
1.89ꢂ10À2
202.8
> 3026[d]
1705.6
52.37ꢂ10À2
0.34ꢂ10À2
3.03ꢂ10À2
0.15ꢂ10À2
1282.1
control
> 4400[b]
> 3766[e]
Scheme 4. The formation of the phosphate-bridged binuclear zinc(II)
complex 8 by a ligand exchange reaction.
[a] The time required for 50% hydrolysis of the substrate (t50) was deter-
mined by monitoring the increase in peak area due to p-nitrophenol.
[PNPDPP]: 0.20ꢂ10À3 m; [catalyst]: 0.21ꢂ10À3 m; in EtOH/CHES buffer,
pH 9.0. [b] After 4400 min, only 24% hydrolysis was observed. [c] Calcu-
lated from the initial 5–10% of the reaction by monitoring the increase
in the peak area of p-nitrophenol at 238C. [d] After 3026 min, only 7%
hydrolysis was observed. [e] After 3766 min, only 3.6% hydrolysis was
observed.
spectrum also indicated the formation of complex 8. The
peak observed for this complex (d=À4.36 ppm) is shifted
downfield compared with that of DPP (d=À12.32 ppm), in-
dicating that the DPP ligand binds strongly to the binuclear
metal center. As the coordination of the DPP ligand in a
bridging mode may block the substrate binding, a considera-
ble decrease in the activity was observed upon accumulation
of DPP. The spectrum recorded in methanol after 5 h of re-
action time, showed only DPMP and small amount of com-
plex 8 (Figure S7 in the Supporting Information). These ob-
servations clearly suggest that the low catalytic activity of
complex 1 in buffer compared with that in methanol is due
to the inhibition by the products formed in the reaction.
and Zn
the initial rates observed in the presence of the ligand and
Zn(OAc)2·2H2O were almost identical to that of the control
ACHTUNGTRENNUNG(OAc)2·2H2O did not show any noticeable activity;
ACHTUNGTRENNUNG
experiment. The initial rate (v0) for the hydrolysis of
PNPDPP by complex 1 (4.50ꢂ10À2 mmminÀ1) was found to
be almost ten times lower than that observed for the metha-
nolysis (52.37ꢂ10À2 mmminÀ1). For complex 2, a 164-fold en-
Chem. Eur. J. 2010, 16, 8878 – 8886
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8881

G. Mugesh and A. Tamilselvi
The efficient and complete hydrolysis of PNPDPP by
complex 2 is unexpected because the hydrolytic reactions of
phosphotriesters promoted by transition-metal complexes
are not generally catalytic due to the binding of anionic
phosphodiesters that inhibit further turnover.[17–20] Although
a substantial rate acceleration has been obtained by bridging
the binuclear metal centers with two phosphoryl oxygen
atoms,[23] a significant product inhibition was observed
during the hydrolysis of phosphodiesters by several binu-
clear zinc(II) complexes. Meyer and co-workers have dem-
onstrated that the hydrolysis of phosphodiesters, such as
bis(4-nitrophenyl) phosphate (BNPP), by pyrazolate-based
zinc(II) complexes is strongly inhibited by the products.[24]
They have also shown that a larger Zn···Zn separation is
generally suited for an efficient bridging of phophodies-
ters.[24b,25] The higher catalytic activity of complex 2 com-
pared with that of 1 in buffer is due to the absence of any
significant inhibition by DPP in the reaction catalyzed by
complex 2. The relatively shorter Zn···Zn distance (3.223 ꢀ)
and the sterically bulky amino substituents in complex 2
may disfavor a strong binding of DPP at the binuclear
zinc(II) center.
nificant inhibition observed by DPP up to a concentration
of 200 mm (Figure 4). These observations indicate that the
difference in the activities of complexes 1 and 2 can be at-
tributed to the various degrees of inhibition by the products.
To understand the nature of species produced in the reac-
tions, we have determined the initial rates at different pH
values. At low pH values, complex 2 was found to be inac-
tive due to the protonation of ZnÀOH species (Figure 5).
To support the conclusions drawn from the NMR spectro-
scopic experiments, we have studied the effect of DPP on
the hydrolysis of PNPDPP by HPLC.[26] When the hydrolysis
of PNPDPP by complexes 1 and 5 was carried out in the
presence of various concentrations of DPP, a decrease in the
activity was observed (Figure 4). In particular, DPP almost
Figure 5. The effect of pH on the reaction rate for the hydrolysis of
PNPDPP by complex 2.
This is in agreement with the report of Samples et al. that
the catalytic activity of the binuclear metal site is generally
lost upon protonation of the hydroxide bridge at lower pH
values.[27] The activity of complex 2 increases drastically with
an increase in the pH of the reaction medium. This indicates
that the replacement of chloride ligands by hydroxide is re-
sponsible for the high catalytic activity. At a pH above 8.0,
the conversion of complex 2 to the corresponding hydroxy
species 6 takes place rapidly with an increase in the pH. At
a pH below 8.0, a considerable amount of protonation takes
place, leading to a decrease in the catalytic activity. This is
in agreement with the report of Lippard et al. that the rate-
limiting nucleophilic attack of the zinc-bound hydroxy spe-
cies at the substrate, followed by fast protonation of the in-
termediates leads to the formation of the hydrolyzed prod-
ucts.[28]
Figure 4. The inhibition of the PTE activity of the binuclear complexes 1
(c), 2 (a), and 5 (b) by DPP in CHES buffer, pH 9.0.
Mechanism for the hydrolysis of phosphotriesters by com-
plex 2: Based on the above observations, we proposed the
mechanism as shown in Scheme 5 for the hydrolysis of
PNPDPP by complex 2. According to this mechanism, the
initial replacement of the chloride ligands in complex 2 by
OHÀ ligands leads to the formation of complex 6, which acts
as the catalyst in the reaction. Although the binuclear com-
plexes may become mononuclear in solution, the mass spec-
tral data indicate that complexes 2 and 6 do not undergo
any dissociation in solution to give the corresponding mono-
nuclear species. An attack of the phosphate group at one of
the zinc ions generates the catalyst–substrate complex 9. As
shown by Lippard et al. for the hydrolysis of b-lactams, this
binding orients the substrate and the hydroxy group in a fa-
completely inhibited the activity of complex 1 with an IC50
value of 73.4 mm. Furthermore, the methanolysis of
PNPDPP by complex 1 was also strongly inhibited by DPP
(Figure S8 in the Supporting Information) and the inhibition
of activity in methanol was found to be similar to that in
buffer. A relatively weak inhibition was observed when
complex 5 was used as the catalyst. This confirms that the
low catalytic activity of complex 1 is mainly due to the
strong binding of DPP at the zinc(II) center, as shown in
Scheme 4. Interestingly, the hydrolytic activity of complex 2
was unaffected upon treatment with DPP. There was no sig-
8882
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 8878 – 8886

The Hydrolysis of Organophosphate Esters
FULL PAPER
Table 2. A comparison of b-lactamase and PTE activities of complexes
1, 2, 5, and BcII.
Complex
b-lactamase activity
PTE activity
t50 [min][d] v0 [mmminÀ1
]
[c]
[h]
t50 [min][a]
v0 [mmminÀ1
]
1
2
5
125.0
2653.0
31.4
9.64
0.89
29.82
3.69
5145.0
220.0
0.01
1.73
0.19
0.11
0.02
>2900[e]
>2800[f]
>2800[g]
BcII
control
364.5
>1800[b]
0.09
[a] Time required for 50% hydrolysis of oxacillin (t50) was determined by
monitoring the decrease in peak area due to oxacillin. [Oxacillin]: 1.0ꢂ
10À3 m; [catalyst]: 0.2ꢂ10À3 m; in CHES buffer, pH 9.0. [b] After
1800 min, only 29% hydrolysis was observed. [c] Calculated from the ini-
tial 5–10% of the reaction by monitoring the decrease in the peak area
of oxacillin at 238C. [d] Time required for 50% hydrolysis of substrate
(t50) was determined by monitoring the increase in peak area due to p-ni-
trophenol. [PNPDPP]: 1.0ꢂ10À3 m; [catalyst]: 0.2ꢂ10À3 m; in CHES
buffer, pH 9.0. [e] After 2900 min, only 45% hydrolysis was observed.
[f] After 2800 min, only 31% hydrolysis was observed. [g] After
2800 min, only 28% hydrolysis was observed. [h] Calculated from the ini-
tial 5–10% of the reaction by monitoring the increase in the peak area of
p-nitrophenol at 238C.
of complexes 1 and 5 were, however, extremely low, indicat-
ing that the products from the PTE-type reactions show
stronger inhibition than those of the mbl reactions. On the
other hand, the PTE activity of complex 2 was found to be
much higher than the mbl activity of this complex. The t50
Scheme 5. The proposed mechanism for the hydrolysis of PNPDPP by
complex 2.
value obtained for the hydrolysis of PNPDPP by
2
(220 min) was almost ten times lower than that observed for
the hydrolysis of oxacillin (2653 min). In contrast, the t50
vorable position for nucleophilic attack.[29] A signal at d=
À13.09 ppm in the 31P NMR spectrum can be assigned to
the catalyst–substrate complex (Figure 1). The formation of
complex 9 was also observed in the ESI–MS spectrum. In
the second step, the nucleophilic attack of zinc-bound hy-
droxide at the phosphorus center led to the elimination of
p-nitrophenol. This results in the formation of complex 10 in
which the DPP anion is bridged between two zinc ions. As
previously mentioned, this complex is expected to be quite
unstable due to the unfavorable binding of the phosphate
group. A rapid reaction of complex 10 with water regener-
ates the catalytically active complex 6.
value for the hydrolysis of PNPDPP by complex
5
(>2900 min) was about 100 times higher than that obtained
for the hydrolysis of oxacillin (31 min) (Table 2).
As shown in Figure 6A, complex 1 was able to hydrolyze
oxacillin completely (100% hydrolysis) within 1500 min,
however, this complex hydrolyzed only 20% of PNPDPP
during the same period. On the other hand, complex 2 hy-
drolyzed more than 95% of PNPDPP within 800 min, but
the hydrolysis of oxacillin was not complete even after
3000 min (Figure 6B). In contrast to 2, complex 5 was found
to be almost inactive in the hydrolysis of PNPDPP and only
approximately 5% conversion was observed after 250 mi-
nutes (Figure S9 in the Supporting Information). These ob-
servations suggest that the two zinc(II) ions in complex 2
are favorably positioned for the efficient hydrolysis of
PNPDPP.
A comparison of b-lactamase and PTE activities: To under-
stand the difference in the structural requirement in the
zinc(II) complexes for the hydrolysis of phosphotriester and
b-lactams, we studied both mbl and PTE activities under
identical experimental conditions. A 5:1 mixture of sub-
strate/catalyst was used to keep the reactions under catalytic
conditions. Although the mbl activities of complexes 1 and 2
were reported in our previous studies,[13] we have repeated
the experiments in CHES buffer at pH 9.0 to ensure that
the mbl and PTE activities are obtained under identical ex-
perimental conditions (Table 2). A comparison of the t50
values obtained for the hydrolysis of oxacillin (b-lactam sub-
strate) and PNPDPP indicates that complexes 1 and 5 exhib-
it high mbl activity. Interestingly, complex 5, with two inde-
pendent zinc(II) centers, was found to be a remarkably effi-
cient catalyst with t50 value of 31.0 min. The PTE activities
Hydrolysis of PNPDPP by mbl (BcII): In addition to syn-
thetic zinc(II) complexes, we have also studied the PTE ac-
tivity of mbl from B. cereus. Since both mbl and PTE have
binuclear zinc active sites, it was thought worthwhile to test
the hydrolysis of PNPDPP by mbl. Interestingly, the
31P NMR spectrum indicated that some PNPDPP was rapid-
ly hydrolyzed by mbl to produce DPP (Figure S10 in the
Supporting Information), however, after some amount of
PNPDPP (ca. 10–15%) had been hydrolyzed, there was no
further hydrolysis observed even after 10 h. It was also ob-
served that the amount of DPP produced in the reaction de-
Chem. Eur. J. 2010, 16, 8878 – 8886
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8883

G. Mugesh and A. Tamilselvi
the b-lactam substrate, the t50 value could not be obtained
for the hydrolysis of PNPDPP even after 1500 min (Fig-
ure S11 in the Supporting Information). After this period,
only about 20% of the substrate was converted to products.
These observations suggest that the presence of the binu-
clear zinc(II) center is not sufficient to perform efficient hy-
drolysis of phosphotriesters. The PTE enzyme may follow
some interesting mechanisms to eliminate the products, par-
ticularly the phosphodiesters, from the active site.
Conclusion
We have shown, for the first time, that the binuclear zinc(II)
complexes that exhibit good metallo-b-lactamase (mbl) ac-
tivity are poor mimics of PTE. A comparison of the PTE ac-
tivities of bi- and mononuclear zinc complexes indicates that
a binuclear metal site is required for the PTE activity. The
replacement of ClÀ ligands by OHÀ is important for the high
PTE activity of complex 2 because this complex is found to
be a poor catalyst in methanol. Metallo-b-lactamase from B.
cereus is found to be an inefficient catalyst in the hydrolysis
of phosphotriesters. These observations indicate that the
binding of b-lactam substrates at the binuclear zinc(II)
center is different from that of phosphotriesters. Further-
more, phosphodiesters, the products from the hydrolysis of
phosphotriesters, significantly inhibit the PTE activity of
mbl and its functional mimics. This study suggests that binu-
clear zinc(II) complexes capable of eliminating the products
from the catalyst–product complexes would serve as good
PTE mimics.
*
*
Figure 6. The comparison of PTE ( ) and mbl ( ) activities of A) com-
plex 1 and B) complex 2. The hydrolysis experiments were carried out in
20 mm CHES buffer, pH 9.0. For the PTE activity, ethanol (35%) was
added to the reaction mixture to maintain the solubility of PNPDPP. The
concentrations of oxacillin or PNPDPP were fixed at 1ꢂ10À3 m and the
zinc(II) complexes at 0.2ꢂ10À3 m.
pends on the concentration of mbl employed for the reac-
tions. These observations suggest that the PTE activity of
mbl is significantly inhibited by DPP, which is similar to the
inhibition of the PTE activity of complex 1. We presume
that the formation of an enzyme–product complex, as shown
in Scheme 6, does not allow the complete hydrolysis of the
substrate.
Experimental Section
General procedure: The buffers used for this investigation (MES,
pH 5.5–6.4; PIPES, pH 6.5–7.0; HEPES, pH 7.1–8.0; TAPS, pH 8.1–8.9;
and CHES, pH 9.0–9.5), diphenyl phosphoryl chloride, diethyl phosphoryl
chloride, diethyl thiophosphoryl chloride, and deuterated solvents were
purchased from Sigma–Aldrich Chemical Company. HPLC grade sol-
vents were purchased from Merck. Other chemicals, including 2-methyl
imidazole, were of the highest purity purchased from local suppliers. All
reactions were carried out at room temperature unless otherwise stated.
Solvents were purified by standard procedures and were freshly distilled
prior to use. 1H (400 MHz), 13C (100.56 MHz), and 31P (161.9 MHz)
NMR spectra were obtained on a Bruker 400 MHz NMR spectrometer.
Chemical shifts are cited in ppm with respect to SiMe4 as an internal
standard (1H and 13C) and phosphoric acid (31P) as an external standard.
Mass spectral studies were carried out on a Q-TOF micro mass spectrom-
eter with ESIMS mode analysis as well as in the Bruker ESIMS ion-trap
mode. In the case of isotopic patterns, the value given is for the most in-
tense peak. The HPLC experiments were performed on an analytical
HPLC system fitted with a PDA detector controlled by EMPOWER soft-
ware (Waters Corporation, Milford MA). The organophosphates were
isolated and purified by using a reverse-phase flash chromatography (Bi-
otage) system.
The observations made by 31P NMR spectroscopy are fur-
ther supported by the HPLC experiments. Whereas the t50
value (364.5 min) obtained for the hydrolysis of oxacillin by
the enzyme BcII (at 44.4 nm) indicated a facile hydrolysis of
Synthesis of ligand HL1: This ligand was synthesized by following the
method reported in the literature.[15] Ethanolamine (3.05 g, 50 mmol) was
added to a solution of 2-methylimidazole (4.11 g, 50 mmol) in water
(70 mL). Formaldehyde (8.12 g, 100 mmol, 37% in water) was slowly
added and the solution was adjusted to pH 11 with KOH. The reaction
Scheme 6. The possible formation of a stable enzyme–product complex
during the hydrolysis of PNPDPP by metallo-b-lactamase from Bacillus
cereus.
8884
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 8878 – 8886

The Hydrolysis of Organophosphate Esters
FULL PAPER
mixture was heated to reflux for 24 h, during which time a suspension
was formed. This suspension was filtered and washed with water to yield
HL1 as a white solid (7.80 g, 47%). 1H NMR (D2O/HCl): d=2.43 (s,
3H), 3.47 (t, 2H), 3.83 (t, 2H), 4.39ppm (s, 2H).
in methanol and the progress of the reaction was followed with respect
to time.
Determination of b-lactamase and phosphotriesterase activities: The ki-
netics of the hydrolysis of PNPDPP was followed by HPLC with a re-
verse-phase column. All reactions were carried in methanol or in CHES
buffer (0.02m pH 9.0). The stock solutions of test complexes, PNPDPP,
and oxacillin were prepared in the appropriate solvent and used immedi-
ately for the experiments to avoid any possible decomposition. The con-
centrations of stock solutions of PNPDPP, oxacillin, and the complexes
were fixed at 10ꢂ10À3 m. In a typical kinetic experiment, a sample vial
containing the test complexes (1.5 mL) with the appropriate concentra-
tion of PNPDPP or oxacillin was incubated. At various time intervals, ali-
quots (10 mL) were removed from the reaction mixture and injected di-
rectly onto the HPLC column. In the case of b-lactamase activity, the
compounds were eluted in a linear gradient mode with a mixture of 30–
Synthesis of 5: Anhydrous ZnCl2 (81.9 mg, 0.6 mmol) was added to a sus-
pension of HL1 (100 mg, 0.3 mmol) in acetonitrile. The reaction was con-
tinued for 1 h and the desired product was obtained from the clear solu-
tion by crystallization. Crystals suitable for the single-crystal X-ray dif-
fraction studies were obtained by the slow evaporation of the filtrate
(73.9 mg, 41%). 1H NMR (CD3OD): d=2.29 (brs, 24H), 3.50 (s, 8H),
7.14 ppm (s, 4H); 13C NMR (CD3OD): d=165.3, 155.1, 130.4, 124.8,
(OH2)4]+.
109.4, 101.8, 59.2, 44.2 ppm; MS (ESI): m/z: 534.9 [Zn2ACTHNUTRGENNUG(HL1)ACHTUTGNRENNUGN
Synthesis of PNPDPP: A mixture of diphenyl chlorophosphate (1 mL,
5 mmol), p-nitrophenol (0.696 g, 5 mmol), and triethylamine (0.7 mL) in
diethyl ether (20 mL) was stirred at room temperature for 12 h. After the
completion of reaction, the reaction mixture was poured into water and
the compound was extracted from the aqueous layer with diethyl ether.
The combined organic fractions were evaporated to dryness to give
yellow oil, which was subjected to reverse-phase flash chromatography to
give PNPDPP in a pure form (0.156 g, 84%). 1H NMR (CDCl3): d=
7.23–7.26 (m, 6H), 7.33–7.40 (m, 6H), 8.24 ppm (d, J=9.2 Hz, 2H);
13C NMR (CDCl3): d=120.5, 121.4, 126.3, 126.6, 130.6, 145.6, 150.5,
155.5 ppm; 31P NMR (CDCl3): d=À18.3 ppm.
80% acetonitrile and 0.1% TFA over 8.5 min at
a flow rate of
1.0 mLminÀ1. The reaction product (oxacillin turnover product, retention
time: 3.66 min) and oxacillin (retention time: 4.24 min) were separable
by column chromatography. The chromatograms were extracted at
254 nm and the concentration of oxacillin or the hydrolyzed product was
determined for each injection from the peak area by using a calibration
plot. In the case of phosphotriesterase activity, the compounds were
eluted in an isocratic mode with a mixture of 80% methanol and 0.1%
TFA in water over 6.5 min at a flow rate of 1.0 mLminÀ1. The reaction
product p-nitrophenol (retention time: 3.78 min) and PNPDPP (retention
time: 5.27 min) were separable on a column. The chromatograms were
extracted at 305 nm and the concentration of p-nitrophenol was deter-
mined for each injection from the peak area by using a calibration plot.
To avoid any significant changes in the concentrations of the starting ma-
terials only the first 5–10% of conversion was followed in most of the
cases. To calculate the percentage conversion in the presence of the en-
zymes and complexes, the test samples were injected onto the column at
regular time intervals.
Synthesis of parathion:
A mixture of diethyl chlorothiophosphate
(0.943 mL, 5 mmol), p-nitrophenol (0.696 g, 5 mmol), and triethylamine
(0.7 mL) in diethyl ether (20 mL) was stirred at room temperature for
12 h. After this period, the reaction mixture was poured into water and
the compound was extracted from the aqueous layer with diethyl ether.
The combined organic fractions were evaporated to dryness to give
yellow oil, which was subjected to reverse-phase flash chromatography to
1
obtain parathion (0.995 g, 68%) in pure form. H NMR (CDCl3): d=1.45
(t, 6H, J=6.8 Hz), 4.33 (q, J=6.8 Hz, 4H), 7.41 (d, J=7.2 Hz, 2H),
8.30 ppm (d, J=7.6 Hz, 2H); 13C NMR (CDCl3): d=16.4, 66.0, 122.1,
125.9, 145.3, 156.0, 188.8 ppm; 31P NMR (CDCl3): d=62.1 ppm.
Synthesis of paraoxon: A mixture of diethyl chlorophosphate (0.860 mL,
5 mmol), p-nitrophenol (0.696 g, 5 mmol), and triethylamine (0.7 mL) in
diethyl ether (20 mL). After stirring the reaction mixture for 12 h at
room temperature, the reaction mixture was poured into water and the
compound was extracted from the aqueous layer with diethyl ether. The
combined organic fractions were evaporated to dryness to give a yellow
oil, which was subjected to reverse-phase flash chromatography to obtain
paraoxon in a pure form (1.018 g, 74%). 1H NMR (CDCl3): d=1.38 (t,
J=6.8 Hz, 6H), 4.26 (q, J=7.2 Hz, 4H), 7.39 (d, J=9.2 Hz, 2H),
8.25 ppm (d, J=6.8 Hz, 2H); 13C NMR (CDCl3): d=16.1, 65.2, 115.6,
120.6, 125.7, 144.7, 155.6 ppm; 31P NMR (CDCl3): d=À7.0 ppm.
X-ray crystallography: X-ray crystallographic studies were carried out on
a Bruker CCD diffractometer with graphite-monochromatized MoKa ra-
diation (l=0.71073 ꢀ) controlled by a Pentium-based PC running the
SMART software package.[30] Single crystals were mounted at room tem-
perature on the ends of glass fibers and data were collected at room tem-
perature. The structures were solved by direct methods and refined by
using the SHELXTL software package.[31] All non-hydrogen atoms were
refined anisotropically, and hydrogen atoms were assigned idealized loca-
tions. Empirical absorption corrections were applied to all structures by
using SADABS.[32]
Acknowledgements
This study was supported by the Department of Science and Technology
(DST), New Delhi, India. G.M. acknowledges the DST for the award of
Ramanna and Swarnajayanti fellowships and A.T. thanks the University
Grants Commission (UGC) for a research fellowship.
[1] J. Dragun, A. C. Kuffner, R. W. Schneiter, Chem. Eng. 1984, 91, 65–
70.
[2] D. B. Barr, R. Bravo, G. Weerasekera, L. M. Caltabiano, R. D.
Whitehead, A. O. Olsson, S. P. Caudill, S. E. Schober, J. L. Pirkle,
E. J. Sampson, R. J. Jackson, L. L. Needham, Environ. Health Per-
spect. 2004, 112, 186–200.
[3] a) B. E. Mileson, J. E. Chambers, W. L. Chen, W. Dettbarn, M.
Ehrich, A. T. Eldefrawi, D. W. Gaylor, K. Hamernik, E. Hodgson,
A. G. Karczmar, S. Padilla, C. N. Pope, R. J. Richardson, D. R. Saun-
ders, L. P. Sheets, L. G. Sultatos, K. B. Wallace, Toxicol. Sci. 1998,
41, 8; b) J. E. Casida, G. B. Quistad, Chem. Res. Toxicol. 2004, 17,
983–998.
Crystal data for 5: C16H30Cl4N6O4Zn2; Mr =643.00; monoclinic; space
group P21/n; a=7.7441(13), b=13.369(2), c=12.092(2) nm; a=
102.841(3)8; V=1220.6(4) ꢀ3; Z=2; 1calcd =1.750 gcmÀ3; MoKa radiation
(l=0.71073 ꢀ); T=293(2) K; R1 =0.0441 (I>2s(I)); wR2 =0.1277 (all
data). Further details about the crystal structures can be obtained from
the CCDC.[33]
[4] M. Lotti, Toxicology 2002, 181, 245–248.
[5] a) Food and Agriculture Organization. The Ticking Time Bomb:
Toxic Pesticide Waste Dumps; News Highlights, 2001, http://
www.fao.org/News/2001/010502-e.htm; b) Food and Agriculture Or-
ganization. Prevention and Disposal of Obsolete Pesticides, 2001,
http://www.fao.org/WAICENT/FAOINFO/AGRICULT/AGP/
AGPP/Pesticid/Disposal/index_en.htm.
[6] S. Bird, T. D. Sutherland, C. Gresham, J. Oakeshott, C. Scott, M. Ed-
dleston, Toxicology 2008, 247, 88–92.
[7] a) G. Amitai, R. Adani, G. Sod-Moriah, I. Rabinovitz, A. Vincze, H.
Leader, B. Chefetz, L. Leibovitz-Persky, D. Friesem, Y. Hadar,
FEBS Lett. 1998, 438, 195–200; b) N. Sethunathan, T. J. Yoshida,
Hydrolysis of PNPDPP, parathion, and paraoxon: The hydrolysis reac-
tions of PNPDPP, paraoxon, and parathion were followed by 31P NMR
spectroscopy. All NMR spectroscopic experiments were carried out at
293 K. The chemical shifts are given in ppm relative to phosphoric acid
as an external reference. In each experiment, the test solution contained
20 mm of complex and 20 mm of PNPDPP, paraoxon, or parathion. The
experiments were carried out in either CHES buffer (200 mm, pH 9.0) or
Chem. Eur. J. 2010, 16, 8878 – 8886
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8885

G. Mugesh and A. Tamilselvi
Microbiology 1973, 19, 873–875; c) C. M. Serdar, D. T. Gibson,
D. M. Munnecke, J. H. Lancaster, Appl. Environ. Microbiol. 1982,
44, 246–249; d) H. Shim, F. M. Raushel, Biochemistry 2000, 39,
7357–7364.
[22] T. Tanase, J. W. Yun, S. J. Lippard, Inorg. Chem. 1996, 35, 3585–
3594.
[23] J. Chin, Curr. Opin. Chem. Biol. 1997, 1, 514–521.
[24] a) B. Bauer-Siebenlist, F. Meyer, E. Farkas, D. Vidovic, J. A. Cuesta-
Seiji, R. Herbst-Irmer, H. Pritzkow, Inorg. Chem. 2004, 43, 4189–
4202; b) B. Bauer-Siebenlist, F. Meyer, E. Farkas, D. Vidovic, S. De-
chert, Chem. Eur. J. 2005, 11, 4349–4360.
[8] B. K. Singh, Nat. Rev. Microbiol. 2009, 7, 756–764.
[9] E. Jabri, M. B. Carr, R. P. Hausinger, P. A. Karplus, Science 1995,
268, 998–1004.
[10] Y.-J. Dong, M. Bartlam, L. Sun, Y.-F. Zhou, Z.-P. Zhang, C.-G.
Zhang, Z. Rao, X.-E. Zhang, J. Mol. Biol. 2005, 353, 655–663.
[11] S. M. Fabiane, M. K. Sohi, T. Wan, D. J. Payne, J. H. Bateson, T.
Mitchell, B. Sutton, Biochemistry 1998, 37, 12404–12411.
[12] M. M. Benning, H. Shim, F. M. Raushel, H. M. Holden, Biochem-
istry 2001, 40, 2712–2722.
[13] A. Tamilselvi, M. Nethaji, G. Mugesh, Chem. Eur. J. 2006, 12, 7597–
7606.
[14] The bright yellow color developed during the reaction indicates that
PNP is largely deprotonated at pH 9.0.
[25] F. Meyer, Eur. J. Inorg. Chem. 2006, 3789–3800, and references
therein.
[26] Diphenyl phosphate (DPP) required for the inhibition study was
generated in situ by adding diphenyl phosphoryl chloride (DPPC) to
CHES buffer at pH 9.0. When DPPC was added to the buffer, rapid
hydrolysis took place and the quantitative formation of DPP was
observed. This conversion can be easily followed by 31P NMR spec-
troscopy. When DPPC was added to the buffer, the peak at d=
À5.1 ppm for DPPC disappeared completely to produce a new
signal at d=À11.9 ppm for DPP.
[15] The crystal structure of complex 5 is almost identical to that of the
related CuII complex reported: W. L. Driessen, D. Rehorst, J. Reed-
ijk, I. Mutikainen, U. Turpeinen, Inorg. Chim. Acta 2005, 358, 2167–
2173.
[27] C. R. Samples, T. Howard, F. M. Raushel, V. J. DeRose, Biochem-
istry 2005, 44, 11005–11013.
[28] N. V. Kaminskaia, C. He, S. J. Lippard, Inorg. Chem. 2000, 39, 3365–
3373.
[16] a) S. Albedyhl, D. Schnieders, A. Jancsꢃ, T. Gajda, B. Krebs, Eur. J.
Inorg. Chem. 2002, 1400–1409; b) H. Arii, Y. Funahashi, K. Jitsuka-
wa, H. Masuda, Dalton Trans. 2003, 2115–2116; c) G. Yang, R.
Miao, Y. Li, J. Hong, C. Zhao, Z. Guo, L. Zhu, Dalton Trans. 2005,
1613–1619.
[17] a) T. Liu, A. A. Neverov, J. S. W. Tsang, R. S. Brown, Org. Biomol.
Chem. 2005, 3, 1525–1533; b) S. A. Melnychuk, A. A. Neverov, R. S.
Brown, Angew. Chem. 2006, 118, 1799–1802; Angew. Chem. Int. Ed.
2006, 45, 1767–1770; c) D. R. Edwards, G. E. Garrett, A. A. Never-
ov, R. S. Brown, J. Am. Chem. Soc. 2009, 131, 13738–13748.
[18] W. Desloges, A. A. Neverov, R. S. Brown, Inorg. Chem. 2004, 43,
6752–6761.
[19] a) Y.-C. Yang, J. A. Baker, J. R. Ward, Chem. Rev. 1992, 92, 1729–
1743; b) Y.-C. Yang, Acc. Chem. Res. 1999, 32, 109–115.
[20] Y.-C. Yang, F. J. Berg, L. L. Szafraniec, W. T. Beaudry, C. A.
Bunton, A. Kumar, J. Chem. Soc. Perkin Trans. 2 1997, 607–613.
[21] K. Worm, F. Chu, K. Matsumoto, M. D. Best, V. Lynch, E. V.
Anslyn, Chem. Eur. J. 2003, 9, 741–747.
[29] N. V. Kaminskaia, B. Spingler, S. J. Lippard, J. Am. Chem. Soc. 2000,
122, 6411–6422.
[30] SMART, version 5.05, Bruker AXS: Madison, 1998.
[31] a) SHELXTL, V97–2, G. M. Sheldrick, Siemens Industrial Automa-
tion Inc., Madison, 1997; b) SHELX-97, Program for crystal struc-
ture solution and refinement, G. M. Sheldrick, University of Gçttin-
gen, Gçttingen (Germany), 1997.
[32] SADABS, Version 2, Multi-Scan Absorption Correction Program,
G. M. Sheldrick, University of Gçttingen, Gçttingen (Germany),
2001.
[33] CCDC-764156 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/
data_request/cif.
Received: February 2, 2010
Published online: June 23, 2010
8886
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 8878 – 8886