Multiple reactive immunization towards the hydrolysis of organophosphorus nerve agents: Hapten design and synthesis

Article abstract of DOI:10.1016/S0968-0896(01)00231-0

We designed and synthesized a series of haptens to elicit catalytic antibodies with phosphatase activity against nerve agents. The design is based on the novel concept of multiple reactive immunization which aims to afford two or more catalytic residues w

Full text of DOI:10.1016/S0968-0896(01)00231-0

Bioorganic & Medicinal Chemistry 9 (2001) 3185–3195
Multiple Reactive Immunization Towards the Hydrolysis of
Organophosphorus Nerve Agents: Hapten Design and Synthesis
Haihong Huang, Wen-Ge Han, Louis Noodleman and Flavio Grynszpan*
The Scripps Research Institute, Department of Molecular Biology, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA
Received 21 February 2001; accepted 23 May 2001
Abstract—We designed and synthesized a series of haptens to elicit catalytic antibodies with phosphatase activity against nerve
agents. The design is based on the novel concept of multiple reactive immunization which aims to afford two or more catalytic
residues within the antibody’s binding cleft. The haptens showed the desired reactivity in vitro and were submitted for immuniza-
tion. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
atropine derivatives and oximes is well known their
beneficial effects are limited.2 Phosphotriesterase
enzymes and some of their mutants reportedly have
hydrolyzed a variety of organophosphate insecticides
and acetylcholinesterase inhibitors.3 Catalytic anti-
bodies4 with phosphatase-like activities elicited from
transition state analogues have also been proposed as
detoxifying and neutralization tools.5 In this paper we
describe a novel approach to the development of catalytic
antibodies with a wide phosphatase activity.
Protection against organophosphorus nerve agents is a
matter of international concern. ‘Nerve agent’, or ‘nerve
gas’, represents a generic name for a class of highly toxic
compounds with marked inhibitory activity on the
enzyme acetylcholinesterase. Some organophosphorus
compounds (e.g., sarin, paraoxon, tabun, soman, and
others) displaying extremely toxic activity are widely
used as insecticide toxins and are also present in warfare
arsenals as effective chemical weapons. The dangers and
lethality of the chemical arsenal render a sense of urgency
to the search for potentially effective detoxifying agents.
Antibody catalyzed reactions take place in an aqueous
environment, close to ambient temperature and at neu-
tral pH. These efficient catalysts are environmentally
friendly in their nature. Due to its protein character, this
class of biocatalysts is potentially biodegradable. Evi-
dently, biodegradation of the detoxifying agent after
decontamination, neutralization or destruction of the
toxic compound represents an optimal post-treatment
opportunity.
Many efforts have been invested towards the safe
destruction of these nerve agents.1 Nevertheless, efficient
environmentally compatible solutions are yet to be
achieved. The efficient neutralization of these toxin
molecules in vivo before they hit their target is also in
demand. Even though antidotal in vivo treatment with
*Corresponding author. Fax: +1-858-784-7559; e-mail: flavio@scripps.edu
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00231-0

3186
H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
Monoclonal antibodies can be fully humanized6 and in
principle can be used for therapeutic purposes.7 Hence,
antibodies catalyzing the hydrolysis of toxic organophos-
phorus compounds will be of pharmacological impor-
tance in the prevention of neurotoxin poisoning.
connected through an appropriate linker (square reac-
tive center attached to B in Scheme 1). The relative
positioning of the reactive centers in the hapten will
dictate the location of the catalytic groups in the raised
antibodies. Catalytic antibodies elicited by this method
will be easily identified by a rather simple Enzyme-
Linked Immunosorbent Assay (ELISA) test. Selected
antibodies binding to the challenging hapten will hold
the promise of being efficient catalysts for the first reac-
tion and in addition will present a second nucleophilic
group within the active site. Lack of the second reaction
and diffusion away from the antibody would result in
the neutralization of the second piece by a solvent
molecule and no evident binding of the hapten to the
antibody. The antibodies generated via double (or mul-
tiple) reactive immunization are expected to be effective
catalysts for the hydrolysis of a wide range of organo-
phosphate (phosphodi- and triesters) and organopho-
sphonate compounds in addition to the above
mentioned organophosphorus toxic agents and their
analogues.
In contrast to ordinary immunization, where the elicited
antibody arranges amino acids within its binding site
generating the required shape and charge com-
plementarity to the challenging non-reactive antigen,
reactive immunization is carried out with reactive com-
pounds in order to promote a chemical reaction during
the binding of the antigen to the antibody. Thus, fol-
lowing the initial binding event, the hapten might
encounter an occasional reactive group in the binding
site yielding a covalent bond between the antigen and
the antibody. The reactive immunization process is
dependent on the reactivity of the hapten as well as that
of the antibody. The binding energy achieved in this
step is greater than otherwise possible and the somatic
mutations will presumably stop at about this point.
Consequently, the resulting antibodies may present
relatively promiscuous binding pockets that may accept
broad ranges of substrates. Further mutations could
lead to antibodies that are more reactive than those
generated by the primary response and result in faster
reactions and tighter binding.
The mechanism of the antibody catalyzed reactions
cannot be anticipated with great reliability. On the one
hand, it is possible that the organophosphorus sub-
strates will phosphorylate the binding site yielding
covalent adducts. On the other hand, the particular
electrostatic environment and amino acid side chains
within the antibody’s binding site may activate molecules
of water and therefore hydrolyze the substrate via either
an associative or a dissociative mechanism10 resulting in a
catalytic process with multiple turnovers.11
The reactive immunization technique represents a major
step towards the generation of antibodies bearing
desired catalytic residues approaching enzymatic per-
formance.8 The power of this technique has been nicely
demonstrated by eliciting antibodies that catalyze the
aldol condensation reaction using the same mechanism
as the class I aldolase enzymes.8c Aldolase catalytic
antibodies approximate the rate acceleration of the nat-
ural enzymes used in glycolysis.8c Antibodies elicited via
reactive immunization are very broad on scope. More
than 100 aldehyde–aldehyde, aldehyde–ketone, and
ketone–ketone aldol addition reactions have been cata-
lyzed by aldolase antibodies. Some of these reactions
yielded the key pieces for the total synthesis of natural
products.9 While aldolase antibodies are very efficient,
phosphatase antibodies only achieved modest activities.
This work focuses on meeting the challenge of generating
efficient phosphatase antibodies.
Results and Discussion
Design and synthesis of a phosphonate hapten for
double reactive immunization
Following the premises described above we designed the
first hapten in this series. Hapten 7 presents a soft-shell
labile phosphonate center that after hydrolytic nucleo-
philic attack triggers the latent quinone methide as the
second reactive species. Hapten 7 was prepared as
depicted in Scheme 2. Thus, the commercially available
4-hydroxymandelic acid (1) reacted with 5-aminohex-
anoic acid, which was previously protected as its corre-
sponding benzyl ester. The resulting phenol (2) reacted
with methyl ethyl phosphonic chloride, generated in situ
from ethyl phosphonic dichloride and one equivalent of
methanol. Subsequent substitution of the secondary OH
group with fluoride using the DAST reaction afforded
the precursor (4) for hapten 7. The carboxy group in 4
was deprotected by hydrogenation and then activated as
the corresponding N-succinimide derivative prior to
coupling to the carrier protein and immunization.
The haptens described in this work are designed to elicit
antibodies via ‘multiple’ reactive immunization, leading
to two or more catalytic residues strategically located in
the binding site. In principle, haptens with different
reactivities will afford different catalytic groups (e.g.,
Lys, Ser, Thr, Tyr, Cys, His) in the elicited biocatalysts.
In general, these haptens present only one perceptible
reactive center that upon reaction with the antibody
unleashes a second latent reactive center which is then
ready to capture a second nucleophilic residue within
the antibody’s binding site (Scheme 1).
When 4 was incubated with an excess of N-butylamine
in a solution of 95% dioxane in water, emulating the
conditions within the hydrophobic binding site of a
typical antibody,12 it afforded the expected derivative
(6) which was isolated by HPLC and characterized by
ESI-MS (calcd for MNa+: 346, found: 346). This
These haptens consist of two detachable parts: one car-
rying a reactive center (triangle reactive center attached
to A in Scheme 1) and another one bearing a second
tacit reactive center as well as the carrier protein

H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
3187
compound is the result of two nucleophilic attacks on 4,
the first (nucleophile=water or amine) on the phos-
phorus atom and the second (nucleophile=amine) on
the quinone methide generated in situ after the first
nucleophilic attack, in addition to the hydrolysis of the
benzyl ester moiety under the reaction conditions. We
expect hapten 7 to react similarly with nucleophilic
residues in the binding sites of the elicited antibodies.
The first nucleophilic group (i.e., a nucleophilic group
from an amino acid side chain or an activated water
molecule) in the antibody’s binding site will be phos-
phorylated, and the second one will react with the qui-
none methide after elimination of a fluoride ion
(Scheme 3). Since the linker-carrier protein construct is
attached to the quinone methide moiety, the success of
the double reactive immunization will be readily mon-
itored by an ELISA test.
Design and preparation of a phosphonate hapten for
multiple reactive immunization
Based on a similar general strategy to that used in the
design of hapten 7, the system could be pushed one step
further by the generation of antibodies with three active
site nucleophilic residues (Scheme 4).
For this purpose, a second hapten was designed includ-
ing a phosphonate center with two differentially labile
phenolates. One phenol moiety with an electron-with-
drawing group (EWG) and a second phenol carrying a
latent quinone methide species. In this case, the first
hydrolysis will result in the release of the phenol
including the EWG (the more labile substituent
attached to the phosphorus atom) and the monopho-
sphorylation of the antibody. This putative antibody-
phophonate will then undergo a second nucleophilic
attack triggering the quinone methide group. A third
nucleophile will then react with the last fragment bear-
ing the carrier protein and be readily identified by a
simple ELISA test. Alternatively, if the first hydrolysis
is caused by a water molecule, the resulting phospho-
nate could either diffuse away from the binding site or
react with a nucleophilic residue.
A suicide inhibitor, 4-(fluoromethyl)phenyl phosphate,
acting through a similar mechanism as described above
for 7, was successfully used to inactivate phosphotyr-
osine phosphatase.13 The relative stability of this inhi-
bitor under physiological conditions coupled with its
reactivity towards the target enzyme are features that gives
further support to the feasibility of our hapten design.
Scheme 1. Schematic representation of the multiple reactive immunization concept.

3188
H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
The key design feature in this hapten is the sequential
hydrolysis of the phenol substituents on the central phos-
phorus atom. The described course of events developed
during the immunization process is likely to bestow anti-
bodies including two residues participating in a phospho-
nate moiety as well as a third nucleophilic residue that
would react with the quinone methide formed in situ.
this method, the geometry optimization calculations
were carried out using the Amsterdam Density Func-
tional (ADF, Version 1999) package.18 For solvation
energy calculations, we used the MEAD (Macroscopic
Electrostatics with Atomic Detail) program suite
developed by Bashford.19
ADF/MEAD is a simpler and faster alternative to per-
To the best of our knowledge the experimental Khydrol
values for compounds structurally related to our hapten
design are not fully available. Assuming that the O–P
bond strength of given phenol phosphates correlates
with the O–H acidity of the corresponding phenols we
turned to the computational estimation of the pKa
values of the later species. We reasoned that the most
acidic phenol would constitute a more easily hydrolyz-
able group in the phosphonate hapten, accounting for
the required cleavage sequence.
forming
full
ADF/SCRF
calculations
for
pKa’s.16,17,22,26,27 We compared the two methods for 8.
ADF/SCRF gives a pKa=10.316a,17 using the constants
given in eq (2) (see computational methods in the
Experimental). This value is one pKa unit higher than
that obtained from ADF/MEAD calculation
(pKa=9.3). These two results bracket the experimental
value of pKa=10.0.
We calculated the pKa’s of a series of 29 phenols (Table
1 and Fig. 1) including 17 with reported acidity con-
stants in water at 25 ꢀC.20 The predicted pKa values
correlate very well with the reported experimental
results (Table 1) further endorsing the estimated
unknown pKa values. Table 1 shows that the ADF/
MEAD method predicts the same pKa serial order for
the different phenols as the experimental data and the
We used two methods for our pKa predictions. The first
one is the Schrodinger’s Jaguar pKa module,14,15 which
performs with a mean absolute error of 0.25 compared
to experimental results.14 For further assessing the
unknown pKa values, we also performed calculations
using another relatively straightforward method.16,17 In
Scheme 2. Synthesis of hapten 7.

H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
3189
Jaguar results with a few exceptions. The only devia-
tions were found for the phenols with ÀNO2 (9, 10 and
11) and ÀCF3 groups (20 and 21). It is highly significant
for our purposes that both theoretical methods predict
the largest pKa value for 36, and the smallest pKa values
for phenols 10 and 11.21 The relative predicted ordering
of the pKa’s of the above mentioned phenols as well as
the predicted pKa shift (ꢀpKa=1.5 for Jaguar and 1.8
for ADF/MEAD) are in good agreement between the
two methods. The ideal combination of phenol sub-
stituents to the phosphonate core of the designed hapten
would include the most and the least acidic phenols to
guarantee the programmed sequential cleavage. Dis-
regarding stability considerations, a hapten including
the combination of phenols 36 and 10 or 11 would be
the most suitable for our purposes.
Scheme 3. Double reactive immunization with hapten 7.
Scheme 4. Multiple reactive immunization leading to three different nucleophile catalytic residues in the antibody binding site.

3190
H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
Figure 1. Structures of phenols used for pKa calculations.
Table 1. Experimental and calculated pKa values for the selected phenols
Phenol
Exp. data
Jaguar
ADF/MEAD
Phenol
Exp. data
Jaguar
ADF/MEAD
8
9.98
8.38
7.16
5.22
9.42
9.08
9.89
9.29
10.25
10.08
9.8
8.6
7.1
5.7
9.6
9.5
9.8
9.1
10.3
9.8
9.3
7.3
3.5
4.1
8.2
7.5
9.4
7.7
23 (mSOCH3)
24 (pSO2CH3)
25 (mSO2CH3)
26 (pCOCH3)
27 (mCOCH3)
28 (pCO2CH3)
29 (mCO2CH3)
30 (pCF2H)
31 (mCF2H)
32 (pCFH2)
33 (mCFH2)
34
8.75
7.83
9.33
8.05
9.19
9.4
8.2
9.3
8.2
9.3
8.8
9.6
9.3
9.5
9.5
9.5
10.2
10.4
10.8
8.3
6.2
8.3
8.5
9.7
7.5
11.3
8.6
9 (mNO2)
10 (pNO2)
11 (m,pNO2)
12 (pCl)
13 (mCl)
14 (pF)
15 (mF)
16 (pCH3)
17 (mCH3)
18 (pCN)
19 (mCN)
20 (pCF3)
21 (mCF3)
22 (pSOCH3)
11.2
10.3
9.5
11.0
11.1
9.3
11.3
11.5
7.808.0 5.4
8.6
9.3
9.1
7.2
7.8
9.3
8.68
8.28
35
36
9.06.5
The relative stability of the hapten under physiological
conditions combined with its sequential reactivity is pivo-
tal in our approach. Therefore, the delicate balance
between reactivity and stability of the antigen directed our
synthetic efforts. After several synthetic attempts experi-
menting with different phenols (e.g., 24, 27, 28, and 29),
we successfully completed the preparation of the designed
hapten using the combination of phenols 36 and 27
(Scheme 5). All other attempts resulted in unstable com-
pounds that would not be appropriate for immunization.
elimination side products. Hapten 44 is relatively stable
with an estimated half-life longer than 48 h.
Cleavage of the 3-Hydroxyacetophenone moiety of 44
under mild aqueous acidic conditions was monitored by
HPLC. This reaction represents the first step in the
reactivity of this hapten as hitherto described in Scheme 4
and is in agreement with the predicted relative reactivity
of 27 and 36. This result which can then be followed by
the hydrolysis of the resulting phosphonate confirms the
soundness of our hapten design. The chemical proper-
ties of 44 indicate that it is suitable for the planned
immunization.
Thus, 3-hydroxy-3-(4-hydroxy-phenyl)-propionic acid
(39) was obtained after deprotection of the aldol pro-
duct between 4-(trimethyl-silanyloxy)-benzaldehyde (37)
and acetic acid benzyl ester. The key phosphonate
intermediate (41) was prepared after the installation of a
five-carbon chain linker. Selective cleavage of the benzyl
ester followed by succinimide activation of the resulting
acid and DAST reaction at À78 ꢀC (carefully followed
by TLC) afforded 44 in reasonable yield. The latter
reaction sequence efficiently reduced the appearance of
In the event of a successful immunization with hapten
44, the elicited antibodies will include two residues par-
ticipating in a phosphonate moiety. This potential ‘che-
lating site’, determined by the two nucleophilic residues,
could potentially result in catalytic metalloantibodies.
The chemistry of these intriguing bioorganometallic
complexes will be explored.

H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
3191
Scheme 5. Synthesis of hapten 44.
Conclusions
compounds were visualized by irradiation with UV light
and/or by treatment with a solution of phosphomolyb-
dic acid followed by heating or by immersing in an
iodine bath. Flash chromatography: silica gel Merk 60
(particle size 230–400 mesh) eluent given in parentheses.
NMR: Bruker AMX 300, Bruker AMX 400, The che-
mical shifts (d) are given in ppm, and the coupling con-
stants J are given in Hz. The NMR spectra were
recorded in CDCl3 unless stated otherwise.
We accomplished the synthesis of two haptens for mul-
tiple reactive immunization. The haptens were designed
to react, and therefore elicit, with more than one
nucleophile in the antibody binding cleft. Haptens 7 and
44 were recently conjugated with ovalbumin and bovine
serum albumin (BSA) and submitted for immunization.
The activities of the elicited catalytic antibodies will be
reported in due course.
In our program, we seek to induce antibodies with multi-
ple functionalities arranged in optimal positions to facil-
itate catalysis. The generation of two or more catalytic
residues with the ability to communicate through hydro-
gen bonding or transfer is of great importance in enzyme
mimetics design since such a feature is key to many enzy-
matic processes. Thus, the synthesis of a second set of
haptens with a latent o-quinone methide is under way.
Computational methods
We used two methods for our pKa predictions. The first
one is the Schrodinger’s Jaguar pKa module.14,15 This
module utilizes ab initio quantum chemical methods to
predict pKa’s in aqueous media. It employs a combina-
tion of correlated ab initio quantum chemistry, self-
consistent reaction field (SCRF), continuum treatment
of solvation,22 and systematic corrections to account for
approximations in both the ab initio and continuum
solvation methods. In this method, the geometry opti-
mization calculations are performed at the B3LYP/6-
31G** level of theory. Then a larger basis set cc-
pVTZ++ is used to get the single point energies. Zero
point energies are estimated and the solvation free
energies are obtained through the SCRF procedure.22
Experimental
General remarks
Reagents were purchased from Aldrich. Thin layer
chromatography (TLC): silica gel plates Merk 60F
,
254

3192
H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
Correction factors are derived from a simple linear
form.14 After all the systematic corrections are added,
the calculation yields good agreement with experimental
results as described in the Jaguar manual. In the parti-
cular calculations for the test set of phenols, Jaguar
pKa’s module performs with a mean absolute error of
0.25 compared to experimental results.14
AH
!
AÀ þ Hþ;
ð1Þ
ð2Þ
1:37pKa ¼ fEðAÀÞ þ EðHþÞ À EðAHÞ þ Ecorr
g
þ fGsolðAÀÞ À GsolðAHÞg À 268:26
¼ PA þ ꢀGðdeprotÞ À 268:26
For further assessing the unknown pKa values, we also
performed calculations using another relatively
straightforward method.16,17 In this method, the geo-
metry optimization calculations were carried out using
the Amsterdam Density Functional (ADF, Version
1999) package.18 The parametrization of Vosko, Wilk
and Nusair23 was used for the local density approxima-
tion term, and the Becke24 and Perdew25 nonlocal cor-
rections were used for the nonlocal exchange and
correlation terms. The molecular orbitals were expan-
ded in an uncontracted triple-z Slater-type orbital basis
set, along with a single set of polarization functions,
which constitutes a basis set IV in the ADF code. The
inner core shells were treated by the frozen core
approximation.
In eq (2) E(AÀ), E(H+) and E(AH) represent the gas-
phase energies of the deprotonated phenol, proton and
neutral phenol, respectively. Ecorr is a correction term to
the proton affinity PA, including zero point energy
(ꢀZPE) and 5/2RT work. We only calculated the
ꢀZPE=À8.8 kcal/mol for phenol (8). For simplicity,
we then used ꢀZPE=À8.8 kcal/mol through all the pKa
calculations. Gsol(AÀ) and Gsol(AH) are the solvation
free energies for the deprotonated phenol and phenol,
respectively À268.26 comes from the sum of the solva-
tion free energy of a proton (À260.5 kcal/mol)19,28À30
(using the estimated value of Noyes),29 and the transla-
tion entropy contribution to the gas-phase free energy
of a proton (ÀTꢀSgas(H+)=À7.76 kcal/mol at 298K
and 1 atm pressure).28
After the geometry optimization, we used Chargefit, a
modified version of the CHELPG code26,27 to fit the
point charges from the molecular electrostatic potentials
(ESP) calculated by ADF. The total net charge of the
molecule and the three Cartesian dipole moment com-
ponents from density functional calculations were
adopted as constraint conditions for the fit. The fitted
points lay on a cubic grid between the van der Waals
radius and the outer atomic radius with a grid spacing
of 0.2 A. The outer atomic radius for all atoms used was
5.0A , and the van der Waals radii for C, O, N, H, F,
Cl, and S were 1.67, 1.4, 1.55, 1.2, 1.7, 1.9 and 1.8 A,
respectively. In order to minimize the uncertainties in
the fitting procedure, the singular value decomposition
(SVD) method27 was introduced into the code to obtain
a model with stable atomic charges and an accurate
molecular dipole moment. For solvation energy calcu-
lations, we used the MEAD (Macroscopic Electrostatics
with Atomic Detail) program suite developed by Bash-
ford.19 In this approach, the solute–solvent interaction
is treated within the framework of classical electro-
statics. The solute is represented by a set of atomic
charges and Born radii, and solvent as a continuous
dielectric medium (with e=80.0). The dielectric bound-
ary between the interior and the exterior is defined by
the surface of contact of a 1.4 A sphere rolling over the
superposition of spheres defined by the Born radii of the
atoms. The free energy difference for charging the solute
in vacuum and in solution is calculated by solving the
macroscopic Poisson equation using a numerical finite-
difference method. We used the same radii for the various
atoms as in the charge-fit calculations, except for the C, N
and O atoms with 2.0A in the –NO2 and –CN groups.
These radii were made larger due to the multiple bond
character of the –NO2 and –CN groups (in agreement with
the Jaguar radii22).
5-Aminovaleric benzyl ester. 5-Aminovaleric acid
hydrochloride (7.68 g, 50mmol) and 4-toluenesulfonic
acid monohydrate (10.46 g, 55 mmol) were added to
freshly distilled benzyl alcohol (25 mL) in a 250mL
round bottle flask equipped with a Dean–Stark trap.
The mixture was refuxed until all the water was
removed from the reaction mixture. The mixture was
then cooled to room temperature. Ethyl ether (500 mL)
was added to the mixture and cooled down to 0 ꢀC for 2
h. Crystalline 5-aminovaleric benzyl ester salt was col-
lected by filtration. After recrystallization from ether/
ethyl acetate. The white crystalline salt (11.2 g) was
collected in 91% yield. Spectroscopic data was in
agreement with that reported in the literature.31
5-[2-Hydroxy-2-(4-hydroxy-phenyl)-acetylamino]-penta-
noic acid benzyl ester (2). dl-4-hydroxymandelic acid (2
g, 11.9 mmol), 1-hydroxybenzotriazole monohydrate
(1.82 g, 11.9 mmol), Diisopropylethylamine (3.85 g, 29.8
mmol) and 5-aminovaleric benzyl ester (5.68 g, 13.1
mmol) were dissolved in DMF (50mL). (3-dimethy-
aminopropyl)-3-ethylcarbodiimide hydrochloride anhy-
drous (EDC, 2.281 g, 11.9 mmol) was added and the
solution was stirred at ambient temperature for 24 h.
The DMF was then removed under reduced pressure
and the residue dissolved in ethyl acetate (50mL). After
washing with water, citric acid and brine, the organic
phase was dried over anhydrous Na2SO4. Flash column
chromatography (ethyl ether:ethyl acetate) afforded the
purified product (3.1 g, 72% yield). ESI-MS: calcd for
C20H23NO5 MH+: 358.2, found: 358. 1H NMR
(400 MHz) d 1.53 (m, 2H), 1.62 (m, 2H), 2.35 (t, 2H),
3.23 (m, 2H), 4.88 (s, 1H), 5.09 (s, 2H) 6.58 (d, 2H,
J=8.5 Hz), 7.04 (d, 2H, J=8.5 Hz), 7.34 (m, 5H) ppm.
13C NMR (100.6 MHz) d 14.20, 21.70, 28.80, 33.60,
38.90, 60.70, 66.40, 73.70, 115.7, 128.3, 128.6, 130.6,
135.8, 156.4, 173.8 ppm.
Finally, we calculated the pKa associated with the process
[eqs (1) and (2)]:

H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
3193
5-{2-[4-(Ethyl-methoxy-phosphinoyloxy)-phenyl]-2-hydr-
oxy-acetylamino}-pentanoic acid benzyl ester (3). Diiso-
propylethylamine (520mg, 4 mmol) was added to a
solution of ethyl phosphonate dichloride (293.8 mg, 2
mmol) in CH2Cl2 (10mL). Methanol (64 mg, 2 mmol)
was added and the solution was stirred at ambient tem-
perature for 1 h. dl-4-Hydroxymandelic-5-amide vale-
ric benzyl ester (714 mg, 2 mmol) was added followed
by diisopropylethylamine (390mg, 1.50mmol). After
stirring for 3 h at ambient temperature, the reaction
mixture was washed with brine and the solvents
removed under reduced pressure. The residue was pur-
ified by flash column chromatography using a stepwise
gradient strarting with ethyl ether/ethyl acetate and
ending with ethyl acetate/acetoniltrile. Purification
afforded 509 mg of product (55% yield). ESI-MS: calcd
for C23H30NO7P MNa+: 486, found: 486. 1H NMR
(400 MHz) d 1.14 (3H, m), 1.47 (2H, m), 1.60(2H, m),
1.70(2H, m), 2.31 (2H, m), 3.17 (2H, m), 3.70(3H, m),
4.75 (1H, s), 5.06 (2H, s), 6.97 (2H, d, J=8.5 Hz), 7.06
(2H, d, J=8.5 Hz), 7.30(5H, m) ppm. 13C NMR
(100.6 MHz) d 6.28, 17.50, 18.93, 21.89, 28.70, 33.52,
38.59, 52.75, 66.08, 72.98, 120.24, 121.3, 127.6 128.09,
128.43, 135.78, 136.70, 149.90, 172.29, 173.12 ppm.
42, 0.04 mmol) and N-hydroxysuccinimide (9.2 mg, 0.08
mmol) in dioxane (5 mL). The solution was stirred at
ambient temperature for 16 h. The urea was then
removed by filtration and the solvent was evaporated.
The crude compound was purified by flash chromato-
graphy using a gradient of ethyl acetate to acetonitrile.
The products were isolated with 90% yield.
Compound 7. ESI-MS: calcd for C20H26FN2O8P MNa+
1
495, found 495. H NMR (400 MHz) d 1.22 (3H, tt),
1.70(2H, m), 1.79 (2H, m), 1.85 (2H, m), 1.91 (2H, m),
2.64 (2H, t), 2.85 (4H, s), 3.37 (2H, m), 5.72 (1H, d,
J=48 Hz), 6.68 (2H, d), 7.23 (2H, d, J=8.5 Hz), 7.42
(2H, d, J=8.5 Hz) ppm. 13C NMR (100.6 MHz): d 6.43,
13.0, 17.78, 19.19, 21.78, 25.31, 25.31, 25.56, 28.40,
30.51, 38.42, 90.22, 92.09, 120.60, 128.30, 168.31, 169.18
ppm.
Compound 43. 1H NMR (400 MHz) d 1.23 (3H, m), 1.39
(2H, m), 1.54 (2H, m), 1.70(2H, m), 2.30(2H,m), 2.58
(3H, s), 2.64 (2H, m), 2.84 (4H, s), 3.28 (2H, m), 5.90
(2H, d, J=48 Hz), 7.18 (1H, d, J=8.5 Hz), 7.28 (2H,d,
J=8.5 Hz) 7.32 (1H, d, J1=8.8 Hz, J2=7.5 Hz), 7.33
(1H, s), 7.67 (1H, d, J=7.5 Hz), 7.72 (1H, d, J=8.8 Hz)
ppm.
5-{2-[4-(Ethyl-methoxy-phosphinoyloxy)-phenyl]-2-fluoro-
acetylamino}-pentanoic acid benzyl ester (4). Diethyl-
amino sulfur trifluoride (DAST, 184.7 mL, 1.4 mmol)
was added to a solution of 1 (464 mg, 1 mmol) in
CH2Cl2 (20mL) at 0 ꢀC. After stirring at 0 ꢀC for 1 h,
the solution was washed three times with brine and the
solvent was removed under reduced pressure. The crude
product was purified by flash chromatography using a
gradient of ethyl acetate to acetonitrile to afford 325.5
mg of 2 (70% yield). The silica gel was neutralized with
triethylamine, washed with methanol and equilibrated
with the eluent prior to chromatography. ESI-MS: calcd
3-Hydroxy-3-(4-hydroxy-phenyl)-propionic acid (39).32
n-Butyl lithium (35.6 mmol, 23.7 mL 1.5 M in hexane)
was added to a solution of diisopropylamine (5 mL,
35.6 mmol) in THF (45 mL). The mixture was stirred
for 15 min at À78 ꢀC. Benzyl acetate (5.94 mL, 39.1
mmol) was then added to the reaction mixture. After
stirring for one
h
at À78 ꢀC, 4-trimethylsiloxy-
benzaldehyde (37, 6.9 g, 35.6 mmol) was also added.
After 3 h at À78 ꢀC, the reaction was quenched with
NH4Cl (saturated solution, 50mL) and the pH adjusted
between 5 to 6. The reaction mixture was extracted with
ethyl acetate (3Â50mL) and the combined organic lay-
ers were dried over Na2SO4 and evaporated. The solid
residue containing 38 was dissolved in 0.05% TFA in
acetonitrile/methanol (1:1, 50mL). After 12 h stirring at
ambient temperature, the solvent was removed and the
residue was dissolved in ethanol (25 mL). Pd/C (0.6 g,
10%) was carefully added to the reaction vessel. The
mixture was stirred under hydrogen atmosphere (1 atm)
for 3 h and filtered through Celite. The solvent was
evaporated to afford 5.06g of 39 (yield 78%). The pro-
duct (ESI-MS: calcd for C9H10O4 MH+ 183, found 183)
was immediately used in the following reaction with no
further purification.
1
for C23H29FNO6P MNa+: 488, found: 488. H NMR
(400 MHz) d 1.21 (3H, m), 1.50(2H, m), 1.60(2H, m),
1.90(2H, m), 2.40(2H, m), 3.32 (2H, m), 3.80(3 H, m),
5.12 (2H, s), 5.70(2H, d, J=48 Hz), 7.20(2H, d, J=8.5
Hz), 7.30(5H, m), 7.40(2H, d, J=8.5 Hz).
5-{2-[4-(Ethyl-methoxy-phosphinoyloxy)-phenyl]-2-fluoro-
acetylamino}-pentanoic acid (5). Pd/C (10%, 3 mg) was
added to a solution of 2 (30 mg, 0.065 mmol) in ethanol
(10mL). The solution was stirred under the hydrogen
atmosphere for 1 h. The reaction mixture was then fil-
tered through Celite and the solvent was removed under
reduced pressure to afford 24 mg of pure 3 (99% yield).
ESI-MS: calcd for C16H23FNO6P MH+: 376.12, found:
1
376. H NMR (400 MHz) d 1.22 (3H, m), 1.50(2H, m),
5-[3-Hydroxy-3-(4-hydroxy-phenyl)-propionylamino]-pen-
tanoic acid benzyl ester (40). 3-Dimethyamino propyl)-
3-ethylcarbodiimide hydrochloride anhydrous (EDC,
1.18 g, 6.2 mmol) was added to a solution of 39 (1.04 g,
6.2 mmol), 1-hydroxybenzotriozle monohydrate (0.94 g,
6.2 mmol), diisopropylethylamine (1.99 g, 15.5 mmol)
and 5-aminovaleric benzyl ester (2.95 g, 6.8 mmol) in
DMF (50mL). The solution was stirred at ambient
temperature for 24 h. The DMF was then removed
under reduced pressure and the residue was dissolved in
ethyl acetate (50mL). After washing with water, citric
acid, and brine, the solution was dried (Na2SO4), the
1.60(2H, m), 1.89 (2H, m), 2.30(2H, m), 3.20(2H, m),
3.80(3H, m), 5.70(1H, d, J=48 Hz), 7.20(2H, J=8.5 Hz),
7.40(2H, J=8.5Hz) ppm. 13C NMR (100.6 MHz) d 6.29,
17.64, 19.07, 20.80, 21.72, 28.62, 33.30, 38.58, 53.05, 90.07,
91.95, 120.61, 128.20, and 168.42 ppm.
General procedure for the preparation of the N-
hydroxysuccinimide esters 7 and 43
1,3-Dicyclohexylcarbodiimide (16.5 g, 0.08 mmol) was
added to a solution of the corresponding acids (5 and

3194
H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
crude compound was purified by flash column chromato-
graphy using a 1:1 mixture of ethyl ether/ethyl acetate
as eluent to afford 1.67 g of 40 (55% yield). ESI-MS:
43 (10 mg, 0.017 mmol) in CH2Cl2 (20mL) at À78 ꢀC.
After stirring at that temperature for 3 h the solution
was washed three times with brine and the solvent was
removed under reduced pressure. The crude product
was purified by flash chromatography using a gradient
of ethyl acetate to acetonitrile to afford 3 mg of 44 (30%
yield). The silica gel was neutralized with triethylamine,
washed with methanol and equilibrated with the eluent
prior to chromatography. ESI-MS: calcd for
C28H32FN2O9P MH+ 591, found 591. 1H NMR
(400 MHz) d 1.22 (3H, m), 1.38 (2H, m), 1.54 (2H, m),
1.70(2H, m), 2.15 (2H, m), 2.58 (3H, s), 2.64 (2H, m),
2.84 (4H, s), 3.28 (2H, m), 5.90(1H, d, J=48.0Hz),
6.78 (2H, d, J=8.5 Hz), 7.18 (1H, dd, J1=7.5 Hz,
J2=8.8 Hz), 7.20(1H, d, J=8.5Hz), 7.32 (1H, m,
J1=8.8 Hz, J2=7.5Hz), 7.33 (1H, s), 7.67 (1H, d, J=8.5
Hz), 7.72 (1H, d, J=8.8 Hz) ppm.
1
calcd for C21H25NO5 MH+ 372, found 372. H NMR
(400 MHz) d 1.47 (2H, m), 1.60(2H, m), 2.36 (2H, s),
2.48 (2H, m), 3.23 (2H, m), 4.99 (1H, dd, J1=8.8 Hz,
J2=3.2 Hz), 5.11 (2H, s), 6.76 (2H, d, J=8.5 Hz), 7.15
(2H, d, J=8.5 Hz), 7.33 (5 H, m) ppm. 13C NMR
(100.6 MHz) d 21.96, 28.82, 33.66, 38.90, 39.11, 44.74,
66.37, 70.63, 115.45, 126.94, 128.26, 128.58,134.54,
135.77, 155.83, 172.06, 173.58 ppm.
5-(2-{4-[(3-Acetyl-phenoxy)-ethyl-phosphinoylmethyl]-
phenyl}-2-hydroxy-acetylamino)-pentanoic acid benzyl
ester (41). Diisopropylethylamine (260mg, 2 mmol)
was added to a solution of ethyl phosphonate dichloride
(146.9 mg, 1 mmol) in CH2Cl2 (5 mL). 2-Hydro-
xyacetophenone (152.15 mg, 1 mmol) was added and
the solution was stirred at ambient temperature for 1 h.
40 (357.16 mg, 1 mmol) was added followed by diiso-
propylethylamine (195 mg, 1.50mmol). After stirring
for 3 h at ambient temperature, the reaction mixture
was washed with brine and the solvents removed under
reduced pressure. The residue was purified by flash col-
umn chromatography using a stepwise gradient strart-
ing with ethyl ether/ethyl acetate and ending with ethyl
acetate/acetonitrile. Purification afforded 290.6 mg of
product (50% yield). ESI-MS: calcd for C31H36NO8P
MNa+ 604, found 604. 1H NMR (400 MHz) d 1.24 (3H,
m), 1.41 (2H, m), 1.54 (2H, m), 1.98 (2H, m), 2.28 (2H,
t), 2.40(2H, m), 2.46 (3H, s), 3.14 (2H, m), 4.93 (1H,
m), 5.02 (2H, s), 6.28 (1H, d, J=5.7 Hz), 7.05 (2H, d,
J=8.5 Hz), 7.20(2H, d, J=8.5 Hz), 7.30(5H, m), 7.34
(1H, dd, J1=8.8 Hz, J2=7.5 Hz), 7.36 (1H, s), 7.64 (1H,
d, J=7.5 Hz), 7.51 (1H, d, J=8.8 Hz) ppm. 13C NMR
(100.6 MHz) d 6.48, 18.54, 19.95, 21.90, 26.68, 28.68,
33.57, 38.80, 44.44, 66.22, 70.16, 120.07, 120.48, 125.1,
125.30, 127.10, 128.21, 128.52, 130.0, 135.70, 138.70,
140.49, 149.30, 150.05, 171.72, 173.28 ppm.
Acknowledgements
We are grateful to Schrodinger, Inc. for a temporary
license of Jaguar. F.G. thanks the J. S. Guggenheim
Memorial Foundation for a fellowship. We thank Pro-
fessor R. A. Lerner for financial support and insightful
discussion. We thank Professor E. J. Baerends and the
Theoretical Chemistry group at the Free University of
Amsterdam for use of the ADF code.
References and Notes
1. (a) Holmes, F. W.; Placitas, N. M. USA NATO Sci. Ser.
1998, 22, 159. (b) Yang, Y.-Ch. Acc. Chem. Res. 1999, 32, 109.
2. Dunn, M. A.; Sidell, F. R. JAMA 1989, 262, 649.
3. (a) Watkins, L. J.; Mahoney, H. J.; Mc Culloch, J. K.;
Raushel, F. M. L. Biol. Chem. 1997, 272, 25596. (b) Scanlan,
T. S.; Reid, R. C. Chem. Biol. 1995, 2, 71.
4. For a review on catalytic antibodies see: Blackburn, G. M.;
Datta, A.; Denham, H.; Wentworth, P. Adv. Phys. Org. Chem.
1998, 31, 249.
5-(3-{4-[(3-Acetyl-phenoxy)-ethyl-phosphinoyloxy]-phenyl}-
3-hydroxy-propionylamino)-pentanoic acid (42). Pd/C
(10%, 3 mg) was added to a solution of 41 (25 mg, 0.043
mmol) in ethyl acetate (5 mL). The solution was stirred
under hydrogen atmosphere for 40min and monitored
by TLC. The reaction mixture was then filtered through
Celite and the solvent was removed under reduced
pressure to afford 21 mg of pure 42 (99% yield). ESI-
5. (a) Amital, H.; Tur-Kaspa, I.; Tashma, Z.; Hendler, I.;
Shoenfeld, Y. Mil. Med. 1996, 161, 7. (b) Rosenblum, J. S.;
Lo, L.Ch.; Li, T.; Janda, K. D.; Lerner, R. A. Angew. Chem.,
Int. Ed. Engl 1995, 34, 2275. (c) Lavey, B. J.; Janda, K. D. J.
Org. Chem. 1996, 61, 7633. (d) Lavey, B. J.; Janda, K. D.
Bioorg. Med. Chem. Lett. 1996, 6, 1523. (e) Spivak, D. A.;
Hoffman, T. Z.; Moore, A. H.; Taylor, M. T.; Janda, K. D.
Bioorg. Med. Chem. 1999, 7, 1145. (f) Renard, P.-Y.; Vayron,
P.; Taran, F.; Mioskovski, Ch. Tet. Lett. 1999, 40, 281. (g)
Vayron, P.; Renard, P.-Y.; Valleix, A.; Mioskowski, Ch.
Chem. Eur. J. 2000, 6, 1050. (h) Vayron, P.; Renard, P.-Y.;
Taran, F.; Creminon, Ch.; Frobert, Y.; Grassi, J.; Mioskovski,
Ch. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 7058. (i) Yli-Kau-
haluoma, J.; Humppi, T.; Yliniemela, A. Acta Chem. Scan.
1999, 53, 473. (j) Brimfield, A. A.; Lenz, D. E.; Maxwell,
D. M.; Broomfield, C. A. Chem.-Biol. Interac. 1993, 87, 95.
6. (a) Berkower, I. Curr. Opin Biotechnol. 1996, 7, 622. (b)
Rybak, S. M.; Hoogenboom, H. R.; Meade, H. M. Proc. Natl.
Acad. Sci. U.S.A. 1992, 89, 3165.
1
MS: calcd for C24H30NO8P MH+ 492, found 492. H
NMR (400 MHz) d 1.28 (3H, m), 1.42 (2H, m), 1.51
(2H, m), 2.09 (2H, m), 2.25 (2H, m), 2.49 (2H, m), 2.60
(3H, s), 5.01 (1H, m), 6.76 (1H, d, J=8.5 Hz), 7.08 (2H,
d, J=8.5 Hz), 7.27 (2H, d, J=8.4 Hz), 7.40(2H, dd,
J1=8.40Hz, J2=7.0Hz), 7.60(2H, s), 7.73 (2H, d,
J=7.0Hz) ppm. 13C NMR (100.6 MHz) d 6.45, 18.51,
19.92, 21.90, 26.7, 28.60, 33.40, 38.85, 44.70, 47.78,
70.17, 92.00, 120.14, 120.50, 125.30, 127.16, 130.15,
138.70, 171.96, 176.66, 191.40, 197.34 ppm.
5-(3-{4-[(3-Acetyl-phenoxy)-ethyl-phosphinoyloxy]-phenyl}-
3-fluoro-propionylamino)-pentanoic acid 2,5-dioxo-pyrro-
lidin-1-yl ester (44). Diethylamino sulfur trifluoride
(DAST, 6.6 mL, 0.05 mmol) was added to a solution of
7. Mets, B.; Winger, G.; Cabrera, C.; Seo, S.; Jamdar, S.;
Yang, G.; Zhao, K.; Briscoe, R. J.; Almonte, R.; Woods,
J. H.; Landry, D. W. Proc. Natl. Acad. Sci. U.S.A. 1998, 95,
10176.

H. Huang et al. / Bioorg. Med. Chem. 9 (2001) 3185–3195
3195
8. (a) Wirshing, P.; Ashley, J. A.; Lo, L.Ch.; Janda, K. D.;
Lerner, R. A. Science 1995, 270, 1775. (b) Wagner, J.; Lerner,
R. A.; Barbas, C. F., III. Science 1995, 270, 1797. (c) Barbas,
C. F., III; Heine, A.; Zhong, G.; Hoffmann, T.; Gramati-
kova, S.; Bjornestedt, R.; List, B.; Anderson, J.; Stura, E.;
Wilson, I. A.; Lerner, R. A. Science 1997, 278, 2085. (d)
Lo, L.Ch.; Wentworth, P.; Jung, K. W.; Yoon, J.; Ashley,
J. A.; Janda, K. D. J. Am. Chem. Soc. 1997, 119, 10251. (e)
Lin, C.-H.; Hoffman, T. Z.; Xie, Y.; Wirsching, P.; Janda,
K. D. Chem. Commun. 1998, 10, 1075. (f) Zhong, G.; Lerner,
R. A.; Barbas, C. F., III. Angew. Chem., Int. Ed. Engl. 1999,
38, 3738.
9. For a recent review on organic synthesis supported by cat-
alytic antibodies see: Hasserodt, J. Synlett 1999, 12, 2007.
10. Florian, J.; Warshel, A. J. Phys. Chem. B 1998, 102, 719.
11. Lockridge, O.; Blong, R. E.; Masson, P.; Froment, M.-T.;
Millard, C. B.; Broomfield, C. A. Biochem. 1977, 36, 786.
12. Shabat, D.; Itzhaky, H.; Reymond, J.-L.; Keinan, E.
Nature 1995, 374, 143.
Theoretical Chemistry, Vrije Universiteit, Amsterdam, The
Netherlands, 1999.
19. (a) Bashford, D.; Karplus, M. Biochemistry 1990, 29,
10219. (b) Lim, C.; Bashford, D.; Karplus, M. J. Phys. Chem.
1991, 95, 5610.
20. (a) Bordwell, F. G.; Cooper, G. D. J. Am. Chem. Soc.
1952, 74, 1058. (b) Schultz, T. W.; Bearden, A. P.; Jaworska,
J. S. SAR and QSAR in Envirom. Res. 1996, 5, 99. (c) Abra-
ham, M. A.; Grellier, P. L.; Prior, D. V.; Duce, P. P. J. Chem.
Soc. Perkin Trans 2 1989, 699.
21. Quantitatively, the Jaguar predictions are in better agree-
ment with the available experimental pKa data than ADF/
MEAD. However, part of this better agreement with Jaguar is
due to the ‘systematic corrections’ (i.e., least square fitting). In
contrast, no systematic corrections were made to the ADF/
MEAD results.
22. Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy,
R. B.; Ringnalda, M. N.; Sitkoff, D.; Honig, B. J. Phys. Chem.
1996, 100, 11775.
13. Myers, J. K.; Widlanski, T. S. Science 1993, 262, 1451.
14. Jaguar, Schrodinger Inc., 1500 SW First Ave. Suite 1180,
Portland, OR 97201, USA.
15. Klicic, J. J.; Friesner, R. A.; Liu, S.-Y.; Ringnalda, M. N.
In preparation.
23. Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58,
1200.
24. Becke, A. D. J. Chem. Phys. 1986, 84, 4524.
25. Perdew, J. P. Phys. Rev. B 1986, 33, 8822. Perdew, J. P.
Phys. Rev. B 1986, 34, 7406 erratum.
16. (a) Richardson, W. H.; Peng, C.; Bashford, D.; Noodle-
man, L.; Case, D. A. Int. J. Quantum Chem. 1997, 61, 207. (b)
Chen, J. L.; Noodleman, L.; Case, D. A.; Bashford, D. J.
Phys. Chem. 1994, 98, 11059.
17. Li, J.; Fisher, C. L.; Konecny, R.; Bashford, D.; Noodle-
man, L. Inorg. Chem. 1999, 38, 929.
18. ADF 1999.02, Baerends, E. J.; Berces, A.; Bo, C.; Boer-
rigter, P. M.; Cavallo, L.; Deng, L.; Dickson, R. M.; Ellis, D.
E.; Fan, L.; Fischer, T. H.; Fonseca Guerra, C.; Gisbergen, S.
J. A.; Groeneveld, J. A.; Gritsenko, O. V.; Harris, F. E.; van
den Hoek, P.; Jacobsen, H.; van Kessel, G.; Kootstra, F.; van
Lenthe, E.; Osinga, V. P.; Philipsen, P. H. T.; Post, D.; Pye,
C.; Ravenek, W.; Ros, P.; Schipper, P. R. T.; Schreckenbach,
G.; Snijders, J. G.; Sola, M.; Swerhone, D.; te Velde, G.; Ver-
nooijs, P.; Versluis, L.; Visser, O.; van Wezenbeek, E.; Wiese-
neker, G.; Wolff, S. K.; Woo, T. K.; Ziegler. T. Department of
26. Breneman, C. M.; Wiberg, K. B. J. Comput. Chem. 1990,
11, 361.
27. (a) Mouesca, J.-M.; Chen, J. L.; Noodleman, L.; Bash-
ford, D.; Case, D. A. J. Am. Chem. Soc. 1994, 116, 11898. (b)
Chargefit2000: Li, J.; Ullmann, M.; Chen, J.L.; Mouesca, J.-
M.; Peng, C. Y.; Case, D. A.; Bashford, D.; Noodleman. L.
Department of Molecular Biology, The Scripps Research
Institute, 2000.
28. Tawa, G. J.; Topol, I. A.; Burt, S. K.; Caldwell, R. A.;
Rashin, A. A. J. Chem. Phys 1998, 109, 4852.
29. Noyes, R. M. J. Am. Chem. Soc. 1962, 84, 512.
30. Reiss, H.; Heller, A. J. Phys. Chem. 1985, 89, 4207.
31. Davis, A. P.; Menzer, S.; Walsh, J. J.; Williams, D. J.
Chem. Commun. 1996, 3, 453.
32. La Manna, A.; Pagani, G.; Pratesi, P.; Ricciardi, M. L. Il
Farmaco Ed. Sci. 1964, 19, 506.