The phospha-Michael addition of dimethyl- and diphenylphosphites to the η1-N-maleimidato ligand: Inhibition of serine hydrolases by half-sandwich metallocarbonyl azaphosphonates

Article abstract of DOI:10.1016/j.jorganchem.2008.10.057

Dialkyl- and diphenyl phosphites react with the (η⁵-C₅H₅)M(CO)_x(η ¹-N-maleimidato) (M = Fe, Mo; x = 2 or 3) complexes giving products of the phospha-Michael addition to the η¹-N-maleimidato

Full text of DOI:10.1016/j.jorganchem.2008.10.057

Journal of Organometallic Chemistry 694 (2009) 908–915
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
The phospha-Michael addition of dimethyl- and diphenylphosphites
to the
by half-sandwich metallocarbonyl azaphosphonates
g
1-N-maleimidato ligand: Inhibition of serine hydrolases
b
c
a
Bogna Rudolf a, , Michèle Salmain , Marcin Palusiak , Janusz Zakrzewski
*
a
´
´
Department of Organic Chemistry, University of Łódz 90-136 Łódz, Narutowicza 68, Poland
b Ecole Nationale Supérieure de Chimie de Paris, Laboratoire de Chimie et Biochimie des Complexes Moléculaires (UMR CNRS 7576),
11 rue Pierre et Marie Curie 75231 Paris cedex 05, France
c
´
´
Department of Crystallography, University of Łódz, 90-136 Łódz, Pomorska 149/153, Poland
a r t i c l e i n f o
a b s t r a c t
Dialkyl- and diphenyl phosphites react with the (g g
5-C5H5)M(CO)x(
or 3) complexes giving products of the phospha-Michael addition to the
these complexes (M = Fe, x = 2) was characterized by X-ray diffraction. The synthesized metallocarbonyl
azaphosphonates and the corresponding iron phosphonic acid act as inhibitors of certain serine hydro-
lases (AChE and BChE). The kinetic assays were performed and revealed that inhibition mechanism
depends strongly on the enzyme and the structure of the inhibitor.
1-N-maleimidato) (M = Fe, Mo; x = 2
Article history:
g
1-N-maleimidato ligand. One of
Received 9 September 2008
Received in revised form 24 October 2008
Accepted 27 October 2008
Available online 8 November 2008
Keywords:
Ó 2008 Elsevier B.V. All rights reserved.
Metallocarbonyl compounds
Bioorganometallic chemistry
Phosphonates
Serine hydrolases
Cholinesterases
1. Introduction
is also disclosed and discussed. The biochemical activity of some of
these new phosphonates towards selected serine hydrolases: ace-
Organometallic complexes carrying phosphonato groups (phos-
phonates or corresponding phosphonic acids) have attracted con-
siderable interest [1] within the last years due in part to their
ability to strongly bind to metal oxide solid supports. This feature
enables immobilization of organometallic moieties (e.g. redox-ac-
tive ferrocenyl groups) on robust inorganic backbones, which
may be of interest for many branches of materials science [2]. On
the other hand, phosphonates having azaheterocyclic substituents
may manifest interesting biological activities [3].
tylcholinesterase (AChE) and butyrylcholinesterase (BChE) was
investigated. It appeared that the diphenyl phosphonate ester de-
rived of 1a behaved as a competitive inhibitor of the serine ester-
ase BChE in a short timescale and inactivated this enzyme in a
time-dependent fashion by phosphonylation of its active site ser-
ine residue.
2. Results
We have been interested for a long time in the chemistry of the
2.1. Synthesis of phosphonates 2a–c and 3a–b
(g g
5-C5H5)M(CO)x( 1-N-maleimidato) complex 1a [4] and, more re-
cently, in that of its molybdenum and tungsten analogs 1b and 1c
[5]. The ethylenic bond of the maleimidato ligand in these com-
plexes is readily attacked by nucleophiles such as thiols, amines
or imidazoles. These reactions can be used to introduce IR-detect-
able moieties on peptides and proteins with potential applications
in immunoassays [6–10].
In this paper, we describe the addition of dialkyl phosphites to
the ethylenic bond in 1a–c leading to phosphonates bearing metal-
locarbonyl moieties. The X-ray structural study of phosphonate 2a
We found that complexes 1a–c react in dichloromethane at r.t.
with dimethyl- and diphenyl phosphite in the presence of a stoi-
chiometric amount of DBU to afford phosphonates 2a–c and 3a
as fairly air-stable yellow solids or oils (Scheme 1).
The 1H NMR spectrum of these compounds did not show ethyl-
enic protons signals characteristic of 1a–c, but instead complex
patterns, assignable to the CH–CH2 entity, coupled with 31P were
observed at 2.9–3.4 ppm. The 13C NMR spectrum of 2a showed
coupling of all carbons of the succinimide ring with 31P, indicating
creation of the C–P bond. The 31P NMR spectra of 2a–c and 3a dis-
played a singlet in the region between 10 and 25 ppm characteris-
tic for phosphonates. Their IR spectra apart from two
* Corresponding author. Tel.: +48 42 6355755.
E-mail address: brudolf@chemul.uni.lodz.pl (B. Rudolf).
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.10.057

B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
909
Table 2
O
O
Selected geometrical parameters of 2a (Å, °).
O
Fe–C6
Fe–C7
Fe–N8
O9–C9
O12–C12
P10–C10
P10–O20
P10–O21
P10–O22
1.786(3)
1.790(3)
1.975(2)
1.219(3)
1.210(3)
1.804(3)
1.458(3)
1.573(2)
1.568(2)
M
M
N
N
i
P
OR
OR
(CO)x
(CO)
x
O
O
1a-c a M = Fe, x = 2
b M = Mo, x = 3
c M = W, x = 3
2a-c R = Me
3a R = Ph
O2–P10–O21
O20–P10–C10
O12–C12–N8
O12–C12–C11
O7–C7–Fe
102.69(11)
113.76(13)
125.3(2)
124.6(2)
175.6(3)
Scheme 1. Conditions: (i) HP(@O)(OR)2, DBU in CH2Cl2.
metallocarbonyl stretching bands ꢀ2000 cmꢁ1 and an imide car-
bonyl stretching band ꢀ1650 cmꢁ1 displayed a strong absorption
band around 1200 cmꢁ1, assignable to the stretching vibration of
the P@O bond.
O6–C6–Fe
C6–Fe–C7
179.1(3)
93.62(15)
Fe–N8–C9–C10
–178.72(15)
ꢁ122.95(19)
52.9(2)
104.6(2)
ꢁ175.5(2)
N8–C9–C10–P10
O20–P10–C10–C9
C10–P10–O21–C21
C10–P10–O22–C22
2.2. X-ray structure of 2a
The crystallographic data and selected geometrical parameters
of 2a are collected in Tables 1 and 2 respectively. Compound 2a
crystallized in monoclinic P21/c space group with one molecule
per asymmetric unit cell. The central iron atom was four-coordi-
nated and the whole complex adopted the so-called ‘‘piano-stool”
structure which seems to be characteristic for this type of coordi-
nation environment (Fig. 1). The distance between the central
atom and the centroid of the Cp ring was 1.724(3) Å, i.e. very close
to the distance between the central atom and the best plane of this
ring for which a distance of 1.723(4) Å was measured. This arises
from the relatively similar distances between the iron and carbon
atoms forming Cp ring; the range is 2.066(5)–2.100(4) Å with a
maximum for C1, with a mean value of 2.082(15) Å. The best plane
of Cp ring formed an angle of 33.1(2)° with the best plane of the
ring of succinimidato ligand. It is worth mentioning that the succ-
inimidato ligand was partially twisted along the Fe–N bond. The
value of the torsion angle formed by Fe, N8, C10 and C11 atoms
is 177.8(1)°, which means that the succinimidato ligand was
twisted along the coordinating bond by more than 2°. As a result,
the angle between the Cp centroid, Fe and the succinimidato ring
centroid was equal to 121.3(1)° instead of value close to 146°
(180° ꢁ 33.1° gives 145.9°). Interestingly, the dimethyl phospho-
nate group was attached to the succinimidato ligand in the direc-
Table 1
Crystallographic data and structure refinement of 2a.
Fig. 1. Molecular structure of 2a. Thermal ellipsoids were drawn with 30%
probability.
Formula
C13H14FeNO7P
383.07
Monoclinic
P21/c
10.784(13)
12.551(10)
12.533(6)
111.43(5)
1579(2)
Formula weight
Crystal system
Space group
a (Å)
b (Å)
c (Å)
tion opposite to the Cp ring. As the result, the P10–O20 vector
formed a torsion angle of 163.7(5)° with the vector running be-
tween Fe center and the centroid of the Cp ring. There was a lack
of typical hydrogen bonds that could stabilize the crystal structure,
since there is a lack of typical proton donating groups. The only
intermolecular interaction, which can additionally stabilize the
crystal packing, seems to be a P@Oꢂ ꢂ ꢂC„O interaction [11] formed
by P10–O20 and C6–O6 bonds. The O20ꢂ ꢂ ꢂC6i (symmetry code i:
2 ꢁ x, ꢁ1/2 + y, 3/2 ꢁ z) distance equals 3.064(5) Å.
b (°)
V (Å3)
Z
4
Dcalc (g/cm3)
Crystal size (mm)
Temperature (K)
Radiation (Å)
h Minimum–maximum (°)
Dataset h; k; l
Total/unique data/Rint
1.611
0.11 ꢃ 0.14 ꢃ 0.16
293
Mo Ka, 0.71069
2.0–27.5
0:14; 0:16; ꢁ16:15
3815/3629/0.037
2557
2.3. Synthesis of phosphonic acid 4
Observed data [I > 2.0r(I)]
Number of reflections
Number of parameters
R, wR2, S
3629
210
Transesterification of 2a with TMSBr followed by reaction with
methanol afforded the metallocarbonyl phosphonic acid 4 (Scheme
2). The compound was soluble in water and methanol and insolu-
0.0326, 0.1104, 0.98
ꢁ0.35/0.43
Minimum/maximum residual density (e Åꢁ3
)

910
B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
compound was able to bind to both free enzymes and enzyme–
O
O
substrate complexes with different dissociation constants KI and
K0I. Finally, 3a behaved as a competitive inhibitor towards BChE
(this experiment was made possible because irreversible inhibition
occurs very slowly, see below), whereas it had no effect on AChE.
O
O
Fe
i then ii
N
Fe
P
N
OCH3
OCH3
P
(CO)2
OH
OH
(CO)2
O
O
2a
4
2.4.2. Irreversible inhibition experiments
Organophosphorus (OP) compounds of various structures
(phosphonates, phosphates, phosphofluoridates, phosphorami-
dates) have long been known as irreversible inhibitors of serine
hydrolases (esterases, lipases, proteases) [14–16]. The mechanism
by which these compounds inhibit these enzymes is well estab-
lished. It consists in the phosphylation of the active site serine res-
idue in a similar manner as the substrate temporarily acylates the
same site (Scheme 3).
Scheme 2. Conditions: (i) TMSBr, CH2Cl2 and (ii) MeOH.
ble in other organic solvents. The 1H NMR spectrum of 4 lacked the
two methyl protons doublets characteristic of 2a. The 31P signal
shifted downfield to 15.86 ppm. The same reaction attempted with
3a failed, leaving substrate 3a unchanged.
In this line, several OP chemicals have been employed as chem-
ical warfare agents (‘‘nerve agents”). Others are broadly used in
agriculture as insecticides. Only two compounds belonging to this
family are actually used as pharmaceuticals, namely echothiophate
for glaucoma treatment and metrifonate for the chemotherapy of
schisostomes [17]. In all these cases, the mechanism of action of
the OP compounds involved the irreversible inhibition of AChE.
In the same line, several diesters of simple phosphonic acids of
the general formula alkyl-P(@O)(OEt)(OC6H4-pNO2) are potent
2.4. Biochemical studies
2.4.1. Reversible inhibition experiments
It has been shown recently that simple dialkyl phenyl phos-
phates were able to inhibit AChE and/or BChE is a reversible man-
ner by a competitive or partially competitive mechanism [12].
Earlier, Casida and coll. showed that several phosphonic acids de-
rived from the agrochemical ethephon were competitive inhibitors
of BChE [13]. We thought it interesting to examine whether com-
plexes 2a and 3a together with phosphonic acid 4 were also able
to inhibit AChE and BChE and what the inhibition mechanism
may be. In a typical experiment, enzyme was mixed with complex
at a fixed concentration and substrate at various concentrations
and the enzyme activity of the samples was measured accordingly.
Initial hydrolysis rates were plotted versus substrate concentration
and non linear regression analysis of the data yielded the dissocia-
tion constants gathered in Table 3.
irreversible inhibitors of AChE with IC50 values of ca. 0.1
compounds inhibit to a lesser extent the serine protease chymo-
trypsin (CT) with IC50 ranging from 1 M to 1 mM depending of
lM. These
l
the alkyl chain length [18]. Other phosphonates containing a
p-nitrophenyl group were also shown to irreversibly inhibit tryp-
sin, CT and BChE [19,20]. More recently, van Koten and coll. re-
ported the synthesis of a phosphonate derivative carrying an
organometallic pincer complex that site-selectively labeled the ac-
tive site serine of the lipase cutinase [21]. Diphenyl phosphonate
esters with a peptidic substituent were shown as irreversible
inhibitors of certain serine proteases [22].
It was immediately observed that 2a (at a concentration up to
1.8 mM) had no effect on the activity of AChE and BChE, indicating
that this compound was not able to bind to any of these enzymes.
The phosphonic acid derivative 4 was shown to behave as a mixed-
type inhibitor for both AChE and BChE, indicating that this
Therefore, we thought it interesting to investigate whether the
new phosphonate diester complexes 2a and 3a were also able to
inactivate the serine hydrolases AChE, BChE and CT in a time-
dependent manner. In a typical experiment, enzyme was incubated
with complex in large excess. Aliquots were removed from time to
time and the enzyme activity of the mixtures measured with
appropriate substrates. No loss of activity was observed for the
mixtures of AChE and either 2a or 3a compared to control, indicat-
ing that none of these phosphonate diesters were able to inactivate
the enzyme. Conversely, gradual loss of enzymatic activity of the
mixtures of BChE and 2a, BChE and 3a and CT and 3a was noticed.
For the two mixtures containing 3a, the % remaining activity plot-
ted as a function of time followed as expected a first order law
from which was deduced the kobs/[I] (kobs is the pseudo-first order
Table 3
Dissociation constants KI and K0I for the inhibition of AChE and BChE by complexes 2a,
3a and 4.
Complex
Enzyme
Type of inhibition
Inhibition constants
2a
AChE
BChE
No inhibition
No inhibition
–
–
3a
4
AChE
BChE
No inhibition
Competitive
–
KI = 0.14 mM
AChE
Mixed
KI = 0.6 mM
K0I = 2.4 mM
KI = 0.9 mM
K0I = 1.7 mM
BChE
Mixed
rate constant;
I
is 3a) of 8.6 0.4 Mꢁ1 minꢁ1 for BChE and
0.66 0.04 Mꢁ1 minꢁ1 for CT (Fig. 2).
A
-
O
P
O
P
O
P
O
1
1
2
1
2
2
H O
- R OH
OR
R
OR
OR
2
OR
OR
OR
R
3
+
+
Enz OH HO
R
Enz
O
Enz OH
Enz
O
P
R
3
2
3
3
"spontaneous regeneration"
O
P
H O
2
OH
R
"aging"
2
Enz
O
+
R OH
3
B
R
R
O
H O
2
- R'XH
R
+
OH
+
Enz
O
X
Enz
O
Enz OH
Enz OH
R
X
R'
-
R'
O
O
O
X = NH or O
Scheme 3. (A) Mechanism of inhibition of serine hydrolases by OP compounds. (B) Simplified mechanism of hydrolysis catalyzed by serine hydrolases.

B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
911
0.5
0
-0.5
-1
y = -0,0006x + 0,095
R2 = 0,9892
-1.5
-2
y = -0,0086x - 0,2579
R2 = 0,9766
CT
BChE
-2.5
-3
0
500
1000
1500
2000
2500
t (min)
Fig. 2. Time-dependent inactivation of BChE and CT by 3a at 1 mM.
In the case of the BChE and 2a, the gradual decrease of enzy-
matic activity did not obey a first order kinetics law (results not
shown). In fact, it was observed the gradual appearance of a precip-
itate that may have been responsible for the loss of activity. Indeed
resuspension of the precipitate completely restored the enzymatic
activity of the mixture as compared to control experiment.
Further experiments were then carried out with 3a and BChE.
First, the ability to inhibit BChE was also evidenced by measuring
the IC50 after incubation for 60 min, i.e. the concentration of com-
plex that decreases by 50% the enzyme activity. It was equal to
16%
14%
12%
10%
8%
- PRAL
+ PRAL 2 mM
220 lM. Second, experiments were also carried out to shed a light
on the mechanism of inhibition of BChE. If 3a inactivates BChE via
the mechanism depicted in Scheme 3, it is likely that the enzyme
can be reactivated via dephosphonylation, either spontaneously
[23] or by addition of a reactivator such as the oxime pralidoxime
[24]. Therefore, an enzyme sample incubated overnight with 3a
was first submitted to gel filtration so as to remove excess com-
plex. The enzymatic activity of the purified protein was low (1.8%
with respect to control sample), indicating that inhibition by 3a
was indeed irreversible in the short term. Then one half of the pro-
tein solution was allowed to stand at room temperature while pra-
lidoxime was added to the other half (final concentration 2 mM).
The % reactivation measured after 15 and 24 h are reported as a
bar chart (Fig. 3).
6%
4%
2%
0%
15 h
24 h
Fig. 3. Time-dependent reactivation of BChE inhibited by 3a. Inhibited BChE was
incubated in Tris–HCl pH 7.6 as is or in the presence of 2 mM pralidoxime.
Very slow reactivation of enzyme towards BTCh hydrolysis was
observed with time. Pralidoxime noticeably accelerated enzyme
reactivation. It is interesting to note that only 15% of enzymatic
activity was spontaneously recovered for the modified BChE after
10 days, so that the simultaneous dealkylation process, i.e. the
so-called aging process, yielding a permanently inhibited enzyme
(see Scheme 3) may have occurred concurrently.
Eventually, the sample of BChE treated overnight with 3a and
submitted to gel filtration was analyzed by IR spectroscopy. In
the 1800–2200 cmꢁ1 spectra range, the spectrum displayed two
weak but clearly detectable bands at 2054 and 1983 cmꢁ1, the lat-
ter one being weaker that the former (Fig. 4, upper trace). For com-
parison, the IR spectrum of 3a was recorded in the same conditions
CO ligands. Why it only changed the asymmetric vibration mode
remains unclear.
This set of experiments gave a good indication that 3a inhibited
BChE by phosphonylation of its active site serine residue (Ser 196
in horse serum BChE). The inhibited enzyme undergoes a very slow
dephosphonylation process leading to partial activity recovery.
This process is accelerated by addition of pralidoxime.
3. Discussion
Addition of dialkyl phosphates to electron-poor ethylenic bonds
(phospha-Michael reaction) has been widely studied in the field of
organic compounds [25]. For example, Tan and coll. have studied
this reaction using maleimides as Michael acceptors [26]. As part
of our earlier work, we showed that the reaction between thiols
and the maleimidato ligand in complexes 1a–c did occur but its
rate was significantly slower compared to that with N-ethylmalei-
mide. In this work, we show that addition of dialkylphosphites to a
(Fig. 4, lower trace). As expected, compound 3a gave two
mCO
bands at 2054 and 2006 cmꢁ1 characteristic of the symmetric
and asymmetric stretching vibration modes of the two CO ligands.
It indicated that the metallocarbonyl complex was indeed cova-
lently bound to the protein. Furthermore, the protein environment
in which the complex was embedded affected the vibration of the

912
B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
0.003
0.0025
0.002
0.0015
0.001
0.0005
0
-0.0005
-0.001
2200
2150
2100
2050
2000
wavenumber (cm )
1950
1900
1850
1800
-1
Fig. 4. IR spectrum of 3a (upper trace) and BChE-3a (lower trace) in the mCO spectral range. See Experimental part for details.
metal coordinated maleimidato ligand is also feasible as in the or-
ganic series. The phospho-derivatives obtained from compound 1a
were sufficiently stable to perform some biological tests on the two
cholinesterases AChE and BChE.
What may be deduced from the kinetics experiments is that the
phosphonic acid complex probably binds to the cholinesterases in
a manner that differs from that of the diphenyl ester complex.
Complex 3a operates selectivity between AChE and BChE while
complex 4 behaves in a very similar manner towards both en-
zymes. We may hypothesize that the latter compound does not
penetrate very deeply within the enzymes’ active site as the main
difference between both enzymes from a steric point of view lies in
the wideness of the gorge and the substrate binding site. Con-
versely, complex 3a likely occupies BChE’s active site. Furthermore,
the phenoxy groups seem to play a role in the binding of 3a to
BChE as 2a bound neither to AChE nor to BChE. The inability of
the diphenyl ester complex to bind to AChE may also be related
to steric reasons, as the size of this complex may be too large with
respect to the diameter of the gorge that is much narrower in the
case of AChE.
Given the fact that 3a likely binds to the esterasic subsite of
BChE’s active site and possesses two phenoxy leaving groups, it
was not unexpected that this compound should be able to phos-
phonylate the catalytically active serine residue. Indeed, 3a was
shown to slowly and almost fully inactivate BChE. Furthermore
as this inactivation was partly reversible in the long term and reac-
tivation was accelerated by addition of the strong nucleophile pra-
lidoxime, it is highly probable that inhibition occurred through
enzyme phosphonylation. Conversely, the inability of the complex
to bind to AChE probably forbade inactivation to take place.
There is an on-going interest in the design of new inhibitors of
cholinesterases because these molecules represent the only class of
drugs administered to patients with Alzheimer’s disease [27].
Among them, phosphonate derivatives may be of interest because
of their stability in water. Because cholinesterases are also the tar-
get for certain agrochemicals (insecticides) and nerve agents, they
have stimulated numerous studies. Therefore, the active site of
AChE and BChE is now well known thanks to a wealth of X-ray
crystallographic data. It consists of a 20 Å length gorge at the bot-
tom of which are located the esterasic subsite including the cata-
lytically active residues Ser and His and the anionic subsite that
accommodates the cationic part of the substrate. Aminoacids com-
position of this latter site differs between AChE and BChE [28],
allowing BChE to catalyze the hydrolysis of larger substrates such
as butyrylcholine. The gorge of AChE is lined with 14 aromatic res-
idues while 6 of these residues are aliphatic in the case of BChE,
making the volume of the gorge 200 Å3 larger for the latter protein
[29]. Both enzymes also display a peripheral anionic site at the en-
trance of the gorge where ligands such as propidium bind.
On the other hand, several research groups have recently re-
ported very interesting results regarding organometallic enzyme
inhibitors for which the transition metal either suitably organizes
the ligand in the three-dimensional space [30] or brings an addi-
tional reactivity [31]. In this area, several polypyridyl Ru(II) com-
plexes were shown to behave as potent inhibitors of AChE with
IC50 in the submicromolar range [32].
4. Conclusions
The phospha-Michael addition of anions of dialkylphosphites to
the double bond of the maleimidato ligand coordinated to several
transition metals successfully yielded phosphonate diester deriva-
tives that were characterized by classical spectroscopic means
including X-ray crystallography. The dimethyl phosphonate ester
of the iron compound was converted into the phosphonic acid
via transesterification with TMSBr. Biochemical activity studies of
the iron derivatives revealed an interesting behavior of these com-
plexes vis-à-vis certain serine hydrolases. While the dimethyl
derivative had no effect on AChE and BChE, the diphenyl adduct
The kinetics assays performed with the new phosphonate dies-
ters 2a and 3a and the phosphonic acid 4 gave rather unexpected
results. The phosphonic acid behaved as a mixed-type inhibitor to-
wards both AChE and BChE with moderate dissociation constants.
The complex bound to the free enzymes with more affinity than to
the enzyme–substrate complexes. On the other hand, the phospho-
nate diphenyl ester was a competitive inhibitor of BChE alone. The
question now arises about the mode of binding of these complexes
to the cholinesterases.

B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
913
selectively inhibited BChE in a competitive fashion and the phos-
phonic acid complex inhibited both AChE and BChE in a mixed-
type fashion. Furthermore, the diphenyl ester complex inactivated
both BChE and chymotrypsin (and not AChE) in a time-dependent
fashion with rates depending on the enzyme. Further studies car-
ried out with BChE revealed that inactivation occurred via phos-
phonylation of the catalytically active serine residue, thus
providing an original way to site-selectively label these enzymes
with a metallocarbonyl moiety.
2c: Yield 72 mg (75%), 1H NMR (CDCl3, d in ppm): 2.7–3.5 (m,
3H, succinimide), 3.79 (d, J = 12 Hz, 3H, OCH3), 3.88 (d, J = 12 Hz,
3H, OCH3), 5.68 (s, 5H, Cp) 31P NMR (CDCl3, d in ppm): 23.46. IR
(KBr) (cmꢁ1): 2046, 1959, (C„O), 1659 (C@O imide), 1319 (P@O).
Anal. Calc. for C14H14NO8PW C, 31.19; H, 2.62; N, 2.6; found: C,
31.08; H, 2.91; N, 2.87.
5.3.2. Metallocarbonyl diphenyl phosphonate complex 3a
An argon-saturated solution of metallocarbonyl complex 1a
(49 mg, 0.18 mmol) in dichloromethane was treated with diph-
enylphosphite (55.7 ll, 0.36 mmol) containing DBU (27 mg,
5. Experimental
0.18 mmol). The reaction mixture was stirred 48 h at room tem-
perature. The mixture was purified by column chromatography.
A yellow band containing a small amount of substrate 1a was
eluted with chloroform, followed by a second yellow band of
product eluted with ethyl acetate/chloroform 1:3. Crystallization
from diethyl ether/heptane (3:1) afforded analytically pure sam-
ple 3a. Yield 49.5 mg (54%). 1H NMR (CDCl3, d in ppm): 2.95–
3.2 (m, 2H, CH2), 3.5–3.7 (m, H, CH), 5.03 (s, 5H, Cp), 7.15–7.35
5.1. General remarks
NMR spectra were recorded in CDCl3, CD3OH and D2O on VAR-
IAN GEMINI 200BB (200 MHz for 1H, 31P and 13C) and AVANCE
TM DRX500 (500 MHz for 1H) spectrometers and referenced to
internal TMS. IR spectra were recorded as KBr pellets on a FT-IR
NEXUS (Thermo Nicolet) spectrometer. All solvents were purified
according to standard procedures. Chromatographic separations
were performed on Silica gel Merck 60 (230–400 mesh ASTM).
All reactions were carried out under argon.
(m, 10H). 31P NMR (CDCl3, d in ppm): 15.87. IR (
m
in cmꢁ1):
2039, 1994 (C„O), 1648 (C@O imide), 1274 (P@O). Anal. Calc.
for C23H18FeNO7P C, 54.46; H, 3.58; N, 2.76; found: C, 54.19; H,
3.32; N, 2.49.
5.2. Materials
5.3.3. Phosphonic acid 4
TMSBr (25 ll, 0.188 mmol) was added to a solution of 2a
Complexes 1a–c were synthesized according to literature pro-
cedures [4,5]. 1,8-diazabicyclo- (5,4,0)undec-7-ene (DBU) was
purchased from Aldrich. Acetylthiocholine (ATCh), acetylcholines-
(36 mg, 0.094 mmol) in dry dichloromethane (2 ml). After 1 h at
r.t., excess TMSBr was removed by rotary evaporation and MeOH
was added and stirred for 1 h. After evaporation of solvent, com-
pound 4 was obtained as an orange oil. Yield 15 mg (45.5%). 1H
NMR (CD3OH, d in ppm): 2.8–3.5 (m, 3H, succinimide), 5.22 (s,
terase (AChE) from electrical eel, type VI-S (EC 3.1.1.7) and a-chy-
motrypsin (CT) from bovine pancreas type II (EC 3.4.21.1) were
obtained from Sigma. Tetramethylsilyl bromide (TMSBr), dimeth-
ylphosphite diphenylphosphite butyrylcholinesterase (BChE) from
horse serum (EC 3.1.1.8) and butyrylthiocholine iodide (BTCh)
were obtained from Fluka. Suc-Ala-Ala-Ala-Pro-Phe-pNA (Suc-
AAPFpNa) was obtained from Bachem. 5,50-Dithiobis(2-nitroben-
zoic acid) (DTNB) was obtained from Lancaster. Pralidoxime was
obtained from Acros.
5H, Cp). 31P NMR (D2O, d in ppm): 15.86. IR ( in cmꢁ1): 2051,
m
2002 (C„O), 1589 (C@O imide), 1186 (P@O). FAB MS m/e = 356
(M+H)+.
5.4. X-ray structural determination of 2a
Crystals of 2a were obtained from a diethyl ether and heptane
solution of complex. Single crystals suitable for X-ray measure-
ment were mounted on glass-fibres. X-ray diffraction data were
collected at 293 K on Rigaku AFC5S four-circle diffractometer using
5.3. Syntheses
5.3.1. Metallocarbonyl dimethyl phosphonate complexes 2a–c
An argon-saturated solution of metallocarbonyl complex 1a–c
(0.18 mmol) in dichloromethane was treated with dimethylphos-
Mo K
a radiation. The unit cells were determined from 25 reflec-
tions. The structure was solved by direct methods using SHELXS-97
[33] and refined by full-matrix least square method on F2 using
SHELXL-97 [34]. After the refinement with isotropic displacement
parameters, refinement was continued with anisotropic displace-
ment parameters for all non-hydrogen atoms. All hydrogen atoms
were placed on geometrically idealized positions and constrained
to ride on their parent atoms, with a C–H distance of 0.950 Å and
Uiso(H) = 1.2Ueq(C). The molecular geometry was calculated by
PARST97 [35] and PLATON [36]. Fig. 1 has been prepared using PLATON
[36]. Table 2 contains selected geometrical parameters.
phite (33 ll, 0.36 mmol) containing DBU (27 mg, 0.18 mmol). The
reaction mixture was stirred overnight at room temperature. The
mixture was purified by column chromatography. A yellow band
containing a small amount of the substrates 1a–c was eluted with
chloroform, followed by second band (yellow for 2a and orange for
2b–c) of product eluted with chloroform-methanol 4:1 for 2a and
CH2Cl2/ethyl acetate 3:1 for 2b–c. Crystallization from diethyl
ether/heptane (3:1) afforded analytically pure samples 2a–b.
2a:Yield 48.6 mg (69%) 1H NMR(CDCl3, din ppm): 2.7–3.35 (m, 3H,
succinimide), 3.84 (d, J = 11 Hz, 3H, OCH3), 3.78 (d, J = 12.8 Hz, 3H,
OCH3), 5.04 (s, 5H, Cp). 13C NMR (CDCl3, d in ppm): 33.6 (d,
JCP = 3.7 Hz), 42.0 (d, JCP = 140.5 Hz), 52.9 (d, JCP = 6.9 Hz), 53.8 (d,
JCP = 6.7 Hz), 85.09 (s, Cp), 184.1 (d, JCP = 6.3 Hz), 187.9 (d, JCP = 6.1 Hz),
5.5. Enzyme assays
5.5.1. Butyrylcholinesterase
Activity measurements were performed on the basis of the Ell-
man’s thiol assay according to Doctor [37] with slight modification.
A stock solution of BChE (10 mg/ml) was prepared in 50 mM phos-
212.0 (s).31P NMR (CDCl3, d in ppm): 12.57. IR ( in cmꢁ1): 2056, 2009
m
(C„O), 1641 (C„O imide), 1228 (P„O). Anal. Calc. for C13H14FeNO7P
C, 40.76; H, 3.68; N, 3.66; found: C, 40,41; H, 3.82; N, 3.85.
phate buffer pH 8.0. The solution was diluted to 20
same buffer and 10 l were dispensed in wells of a microplate
(Greiner). The substrate mixture was prepared by mixing BTCh
(50 mM in H2O, 50 l) and DTNB (1 mM in 100 mM Tris–HCl,
lg/ml in the
2b: Yield 65 mg (80%), 1H NMR (CDCl3, d in ppm): 2.9–3.4 (m,
3H, succinimide), 3.87 (d, J = 12 Hz, 3H, OCH3), 3.80 (d, J = 12 Hz,
3H, OCH3), 5.57 (s, 5H, Cp), 31P NMR (CDCl3, d in ppm): 21.89. IR
l
l
(m
in cmꢁ1): 2054, 1977 (C„O), 1651 (C@O imide), 1288 (P@O).
0.01% gelatine pH 7.6, 5 ml) and incubated at 30 °C. The enzymatic
Anal. Calc. for C14H14MoNO8P C, 37.27; H, 3.13; N, 3.1; found: C,
reaction was initiated by dispensing the substrate mixture into the
37.41; H, 3.58; N, 3.49.
wells (200 ll). The optical density at 415 nm was monitored dur-

914
B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
ing 4 min with a microplate reader (model 680, Biorad) thermo-
stated at 30 °C. In these conditions, BChE has a specific activity of
6 u/mg of solid).
transferred into adjoining wells. A mixture of substrate (0.5 mM)
and DTNB (1 mM) in 100 mM Tris–HCl pH 7.6 (200 l) was added
l
to the wells and the optical density was immediately monitored at
415 nm during 4 min at 30 °C. The IC50 was calculated from linear
regression of plot of logit(% residual activity) versus log([3a]) (lo-
git(% residual activity) = log(% residual activity/(100 ꢁ % residual
activity)).
5.5.2. Acetylcholinesterase
Activity measurements were performed as above except that
BTCh was replaced by ATCh and the enzyme solution was diluted
to 0.8 lg/ml. In these conditions, AChE has a specific activity of
166 u/mg of protein.
5.8. Time-dependent inactivation of AChE and BChE
5.5.3. Chymotrypsin
BChE (10 mg/ml in 50 mM phosphate buffer pH 8, 5
mixed with 50 mM Tris–HCl buffer pH 7.6 (175 l) and inhibitor
(10 mM in MeOH, 20 l). A control experiment was performed
where the complex solution was replaced by MeOH. Aliquots
(2 l) were withdrawn at recorded times, diluted to the tenth
and the enzymatic activity was measured under standard condi-
tions (see above). The logarithm of the % residual activity was plot-
ted as a function of time to deduce the pseudo-first order rate
ll) was
Activity measurements were performed according to the assay
described by DelMar et al. [38] transposed to a microplate format.
A 1 mM stock solution of SucAAPFpNa was prepared in DMSO. The
working solution was prepared by diluting 1 part of stock solution
with 8 parts of 50 mM Tris–HCl, 20 mM CaCl2 pH 7.6. A stock solu-
tion of CT was prepared in 1 mM HCl (10 mg/ml, kept at ꢁ20 °C for
no more than one week). Immediately before the assay, the en-
l
l
l
zyme was diluted to 20
pensed into wells of a microplate. The enzymatic reaction was
initiated by addition of substrate (180 l). The optical density at
l
g/ml in 1 mM HCl and 20
l
l were dis-
constant kobs
(0.178 mg/ml).
. The same procedure was applied with AChE
l
415 nm was monitored during 2 min with a microplate reader. In
5.9. Time-dependent inactivation of CT
these conditions, CT has a specific activity of 68 2 u/mg of solid.
CT (10 mg/ml in 1 mM HCl, 100
Tris–HCl buffer pH 7.6 (350 ll) and 3a (10 mM in MeOH, 50 l
l
l) was mixed with 50 mM
l).
A control experiment was performed for which 3a was replaced
by MeOH. Aliquots (2 l) were withdrawn at recorded times, di-
luted to 20 g/ml and the enzymatic activity was measured under
5.6. Reversible inhibition assays
AChE (0.78
mixed with MeOH (20
a mixture of ATCh (0.1–0.5 mM) and 1 mM DTNB (200
l
g/ml in 50 mM phosphate buffer pH 8.0, 10
l
l
l) was
l) and
l). The
l
ll) or inhibitor (10 mM in MeOH, 20
l
l
standard conditions (see above). The logarithm of the % residual
optical density was monitored at 415 nm during 4 min at 30 °C.
activity was plotted as a function of time to deduce the pseudo-
The same procedure was applied with BChE using an enzyme solu-
first order rate constant kobs.
tion of 20 lg/ml and a substrate concentration range of 0.05–
0.3 mM. The interaction between the enzymes and the inhibitors
can be described by Scheme 4 where E is the enzyme, S in the sub-
strate, I is the inhibitor, ES is the enzyme–substrate complex and P1
and P2 are the products. KI and K0I are the inhibition constants
reflecting the interaction of inhibitor with free enzyme and en-
zyme–substrate complex, respectively.
5.10. Reactivation studies
BChE (10 mg/ml in 50 mM phosphate buffer pH 8, 5
mixed with 50 mM Tris–HCl buffer pH 7.6 (185 l) and 3a
(20 mM in MeOH, 10 l). The mixture was incubated overnight at
ll) was
l
l
room temperature then submitted to gel filtration to eliminate ex-
cess complex (5 ml bed volume D-salt column, Pierce chemicals)
using the same Tris buffer as eluent. One half of the protein solu-
tion collected (0.25 ml) was allowed to stand at room temperature
The Michaelis–Menten equation reflecting both interactions is:
mmax ꢃ ½Sꢄ
ð1=a0Þ ꢃ mmax ꢃ ½Sꢄ
m
¼
¼
;
a
ꢃ KM
þ
a0 ꢃ ½Sꢄ
ð
a=
a0Þ ꢃ KM þ ½Sꢄ
and aliquots (10 ll) were removed at specified times and the enzy-
a
and a0 are:
matic activity determined as above. Pralidoxime was added to the
second half of the solution of blocked BChE (final concentration
2 mM). After 15 h, the residual enzymatic activity was measured
on a 10-ll aliquot and corrected by spontaneous hydrolysis of
BTCh in the presence of pralidoxime. The reactivation rate R was
where the modifying factors
½Iꢄ
½Iꢄ
:
a
¼ 1 þ
and a0 ¼ 1 þ
KI0
a0)KM were estimated by non linear
KI
KM,
v
max, (1/a0
)
vmax and (
a
/
calculated according to [39] from the following equation:
regression analysis of rate versus substrate concentration plots in
the absence or presence of inhibitor with Kaleidagraph.
ꢀ
ꢁ
m0
m0mI
ꢁ
mR
R ¼ 1 ꢁ
ꢃ 100;
5.7. Measurement of IC50
where v0 is the substrate hydrolysis rate by intact BChE, vR is the
hydrolysis rate by reactivated BChE and vI is the hydrolysis rate
by inhibited BChE.
BChE (100
l
g/ml, 90
l
l) was mixed with 3a (0–20 mM, 10
ll) in
wells of a microtiter plate. After 1 h, the mixtures (10
ll) were
5.11. IR analysis of BChE 3a
K
m
ES
+
E + P + P
1 2
E + S
+
BChE (10 mg/ml in 50 mM phosphate buffer pH 8, 50
mixed with 50 mM Tris–HCl buffer pH 7.6 (130 l) and 3a (10 mM
in MeOH, 20 l). The mixture was incubated in the dark overnight
ll) was
l
I
I
l
and submitted to gel filtration to eliminate excess complex (5 ml
bed volume Econopac P6 column, Biorad) using 0.15 M NaCl as elu-
ent. The protein solution was concentrated with a centrifugal filter
(Ultrafree 0.5, 10 K NMWL, Millipore). The resulting solution
K'
K
I
I
EI + S
ESI
Scheme 4. Reversible inhibition. Competitive type: K0I = 0; Mixed type: KI – K0I;
Non-competitive type: KI = K0I.
(10 ll) was deposited on a 6 mm diameter nitrocellulose membrane

B. Rudolf et al. / Journal of Organometallic Chemistry 694 (2009) 908–915
915
[13] N. Zhang, J.E. Casida, Bioorg. Med. Chem. 10 (2002) 1281–1290.
[14] J.C. Powers, J.L. Asgian, O.D. Ekici, K.E. James, Chem. Rev. 102 (2002) 4639–
4750.
(Transfer blot, Biorad). The IR spectrum of the dry membrane was re-
corded with an FT- spectrometer (Tensor 27, Bruker) and the absor-
bance of a blank membrane was manually subtracted.
[15] J.-F. Cavalier, G. Buono, R. Verger, Acc. Chem. Res. 33 (2000) 579–589.
ˇ
ˇ
[16] J. Patocka, K. Kuca, D. Jun, Acta Med. (Hradec Králové) 47 (2004) 215–228.
[17] J.E. Casida, G.B. Quistad, Chem. Biol. Interact. 157 (2005) 277–283.
[18] E.L. Becker, T.R. Fukuto, D.C. Canham, E. Boger, Biochemistry 2 (1963) 72–76.
[19] Q. Zhao, I.M. Kovach, A. Bencsura, A. Papathanassiu, Biochemistry 33 (1994)
8128–8138.
[20] A. Tramontano, B. Ivanov, G. Gololobov, S. Paul, Appl. Biochem. Biotechnol. 83
(2000) 233–243.
[21] C.A. Kruithof, M.A. Casado, G. Guillena, M.R. Egmond, A. van der Kerk-van Hoof,
A.J.R. Heck, R.J.M. Klein Gebbink, G. Van Koten, Chem. Eur. J. 11 (2005) 6869–
6877.
Acknowledgements
The French Ministry of Foreign Affairs and the Polish Ministry of
Science and Education are gratefully acknowledged for scientific
exchanges through the ‘‘POLONIUM” programme.
Appendix A. Supplementary material
[22] J. Oleksyszyn, J.C. Powers, Biochemistry 30 (1991) 485–493.
[23] A.N. Davison, Biochem. J. 60 (1955) 339–346.
[24] I.B. Wilson, S. Ginsburg, Arch. Biochem. Biophys. 54 (1955) 569–571.
[25] D. Enders, A. Saint-Dizier, M.-I. Lannou, A. Lenzen, Eur. J. Inorg. Chem. (2006)
29–49.
[26] Z. Jiang, Y. Zhang, W. Ye, C.-H. Tan, Tetrahedron Lett. 48 (2007) 51–54.
[27] A. Martinez, A. Castro, Expert Opin. Investig. Drug. 15 (2006) 1–12.
[28] P. Taylor, Z. Radic, N.A. Hosea, S. Camp, P. Marchot, H.A. Berman, Toxicol. Lett.
82/83 (1995) 453–458.
CCDC 701469 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif. Supplementary data associated with this article
can be found, in the online version, at doi:10.1016/
j.jorganchem.2008.10.057.
[29] A. Saxena, A.M.G. Redman, X. Jiang, O. Lockridge, B.P. Doctor, Biochemistry 36
(1997) 14642–14651.
[30] (a) E. Meggers, G.E. Atilla-Gokcumen, H. Bregman, J. Maksimoska, S.P.
Mulcahy, N. Pagano, D.S. Williams, Synlett (2007) 1177–1189;
(b) H. Bregman, D.S. Williams, G.E. Atilla, P.J. Carroll, E. Meggers, J. Am. Chem.
Soc. 126 (2004) 13594–13595;
(c) G.E. Atilla-Gokcumen, D.S. Williams, H. Bregman, N. Pagano, E. Meggers,
ChemBioChem 7 (2006) 1443–1450;
(d) E. Meggers, Curr. Opin. Chem. Biol. 11 (2007) 287–292.
[31] W.H. Ang, A. De Luca, C. Chapuis-Bernasconi, L. Juillerat-Jeanneret, M. Lo Bello,
P.J. Dyson, ChemMedChem 2 (2007) 1799–1806.
[32] S.P. Mulcahy, R. Korn, X.L. Xie, E. Meggers, Inorg. Chem. 47 (2008) 5030–5032.
[33] G.M. Sheldrick, SHELXS, Program for crystal structure solution, University of
Göttingen, Germany, 1997.
[34] G.M. Sheldrick, SHELXL, Program for refinement of crystal structures, University
of Göttingen, Germany, 1997.
[35] M. Nardelli, J. Appl. Crystallogr. 29 (1996) 296.
[36] A.L. Spek, PLATON – Molecular geometry program, University of Utrecht, The
Netherlands, 1998.
References
[1] T.L. Schull, D.A. Knight, Coord. Chem. Rev. 249 (2005) 1269–1282.
[2] J. Wu, Y. Song, E. Hang, H. Hou, Y. Fan, Y. Zhou, Chem. Eur. J. 12 (2006) 5823–
5831.
[3] K. Moonen, I. Laureyn, C.V. Stevens, Chem. Rev. 104 (2004) 6177–6215.
[4] B. Rudolf, J. Zakrzewski, Tetrahedron Lett. 35 (1994) 9611–9612.
[5] B. Rudolf, M. Palusiak, J. Zakrzewski, M. Salmain, G. Jaouen, Bioconjugate Chem.
16 (2005) 1218–1224.
[6] B. Rudolf, J. Zakrzewski, M. Salmain, G. Jaouen, New J. Chem. (1998) 813–818.
[7] B. Rudolf, J. Zakrzewski, G. Celichowski, M. Salmain, A. Vessières, G. Jaouen,
Appl. Organomet. Chem. 18 (2004) 105–110.
[8] N. Fischer-Durand, M. Salmain, B. Rudolf, A. Vessières, J. Zakrzewski, G. Jaouen,
ChemBioChem 5 (2004) 519–525.
[9] N. Fischer-Durand, M. Salmain, B. Rudolf, L. Jugé, V. Guérineau, O. Laprévote, A.
Vessières, G. Jaouen, Macromolecules 40 (2007) 8568–8575.
[10] P. Haquette, M. Salmain, K. Svedlung, A. Martel, B. Rudolf, J. Zakrzewski, S.
Cordier, T. Roisnel, C. Fosse, G. Jaouen, ChemBioChem 8 (2007) 224–231.
[11] F.H. Allen, C.A. Baalham, J.P.M. Lommerse, P.R. Raithby, Acta Crystallogr. B 54
(1998) 320–329.
[12] K.-S. Law, R.A. Acey, C.R. Smith, D.A. Benton, S. Soroushian, B. Eckenrod, R.
Stedman, K.A. Kantardjieff, K. Nakayama, Biochem. Biophys. Res. Commun. 355
(2007) 371–378.
[37] B.P. Doctor, L. Toker, E. Roth, I. Silman, Anal. Biochem. 166 (1987) 399–403.
[38] E.G. DelMar, C. Largman, J.W. Brodrick, M.C. Geokas, Anal. Biochem. 99 (1979)
316–320.
[39] D. Jun, L. Musilovac, K. Kuca, J. Kassa, J. Bajgar, Chem. Biol. Interact. (2008),
doi:10.1016/j.cbi.2008.05.004.