Metal ion catalysis in the hydrolysis of esters of 2-hydroxy-1, 10-phenanthroline: The effects of metal ions on intramolecular carboxyl group participation

Article abstract of DOI:10.1006/bioo.2000.1176

Rate constants have been determined for hydrolysis of the acetate, glutarate, and phthalate monoesters of 2-hydroxy-1,10-phenanthroline in water at 30°C and μ = 0.1 M with KCl. The hydrolysis reactions of the esters are hydroxide ion catalyzed at pH > 9. The phthalate and glutarate monoesters have in addition pH-independent reactions from pH 5.5 to 9 that involve intramolecular participation by the neighboring carboxylate anion. The pH-independent reaction of the glutarate monoester is ～5-fold faster than that of the phthalate monoester. The plots of log k_obsd vs pH for hydrolysis of the carboxyl substituted esters are bell shaped at pH < 5, which indicates a rapid reaction of the zwitterionic species (carboxyl anion and protonated phenanthroline nitrogen). The divalent metal ions, Cu²⁺, Ni²⁺, Zn²⁺, and Co²⁺, complex strongly with the esters; saturation occurs at metal ion concentrations less than 0.01 M. The 1:1 metal ion complexes have greatly enhanced rates of hydrolysis; the second-order rate constants for the OH^- reactions are increased by factors of 10⁵ to 10⁸ by the metal ion. The pH-rate constant profiles for the phthalate and glutarate ester metal ion complexes have a sigmoidal region below pH 6 that can be attributed to a metal ion-promoted carboxylate anion nucleophilic reaction. The carboxyl group reactions are enhanced 10²- to 10³-fold by the metal ions, which allows the neighboring group reaction to be competitive with the favorable metal ion-promoted OH^- reaction at pH < 6, but not at pH > 6. The half-lives of the pH-independent neighboring carboxyl group reactions of the Cu(II) complexes at 30°C are ～2 s. The other metal ion complexes are only slightly less reactive (half-lives vary from 2.5 to 40 s). These are the most rapid neighboring carboxyl group reactions that have been observed in ester hydrolysis.

Full text of DOI:10.1006/bioo.2000.1176

Bioorganic Chemistry 28, 357–373 (2000)
doi:10.1006/bioo.2000.1176, available online at http://www.idealibrary.com on
Metal Ion Catalysis in the Hydrolysis of Esters of 2-Hydroxy-1,
10-phenanthroline: The Effects of Metal Ions on Intramolecular
Carboxyl Group Participation
Thomas H. Fife and M. P. Pujari
Department of Biochemistry, University of Southern California, Los Angeles, California 90033
Received June 16, 2000
Rate constants have been determined for hydrolysis of the acetate, glutarate, and phthalate
monoesters of 2-hydroxy-1,10-phenanthroline in water at 30ЊC and ␮ ϭ 0.1 M with KCl. The
hydrolysis reactions of the esters are hydroxide ion catalyzed at pH Ͼ 9. The phthalate and
glutarate monoesters have in addition pH-independent reactions from pH 5.5 to 9 that involve
intramolecular participation by the neighboring carboxylate anion. The pH-independent reaction
of the glutarate monoester is ϳ5-fold faster than that of the phthalate monoester. The plots of
log kobsd vs pH for hydrolysis of the carboxyl substituted esters are bell shaped at pH Ͻ 5,
which indicates a rapid reaction of the zwitterionic species (carboxyl anion and protonated
phenanthroline nitrogen). The divalent metal ions, Cu2ϩ, Ni2ϩ, Zn2ϩ, and Co2ϩ, complex
strongly with the esters; saturation occurs at metal ion concentrations less than 0.01 M. The
1:1 metal ion complexes have greatly enhanced rates of hydrolysis; the second-order rate
constants for the OHϪ reactions are increased by factors of 105 to 108 by the metal ion. The
pH-rate constant profiles for the phthalate and glutarate ester metal ion complexes have a
sigmoidal region below pH 6 that can be attributed to a metal ion-promoted carboxylate anion
nucleophilic reaction. The carboxyl group reactions are enhanced 102- to 103-fold by the metal
ions, which allows the neighboring group reaction to be competitive with the favorable metal
ion-promoted OHϪ reaction at pH Ͻ 6, but not at pH Ͼ 6. The half-lives of the pH-independent
neighboring carboxyl group reactions of the Cu(II) complexes at 30ЊC are ϳ2 s. The other
metal ion complexes are only slightly less reactive (half-lives vary from 2.5 to 40 s). These
are the most rapid neighboring carboxyl group reactions that have been observed in ester
hydrolysis.
᭧ 2000 Academic Press
INTRODUCTION
A carboxyl group has been implicated in the catalytic action of the esterase and
peptidase Zn(II) metalloenzyme carboxypeptidase A (1–5). A nucleophilic mechanism
involving the ␥-carboxyl group of Glu-270 has been proposed for the enzyme-cata-
lyzed hydrolysis of ester substrates (3,6,7), although trapping experiments designed
to demonstrate the presence of an anhydride intermediate have not been successful
(8). In addition to the intrinsic chemical interest, knowledge of the mechanisms and
the magnitude of the rate constants for coordinated metal ion and carboxyl group
participation in the hydrolysis of acyl derivatives is necessary for the meaningful
assessment of proposed enzyme mechanisms.
357
0045-2068/00 $35.00
Copyright ᭧ 2000 by Academic Press
All rights of reproduction in any form reserved.

358
FIFE AND PUJARI
A neighboring carboxyl group can participate in the hydrolysis of esters as an
intramolecular nucleophile without assistance from metal ions (9–15). However,
the rate constants in the nucleophilic reactions are considerably less than those in
carboxypeptidase A esterase reactions (16,17). The intramolecular carboxyl group
reactions of esters continue to be of great interest as examples of “steric fit” in
nucleophilic reactions (15).
The demonstration of bifunctional mechanisms in hydrolytic reactions is important
because many enzymatic reactions give bell-shaped pH-rate constant profiles that
may be due to such mechanisms. Esters with two substituent carboxyl groups can
give kobsd vs pH profiles that are bell shaped (13,14). The profile for the reaction of
salicyl succinate was interpreted in terms of the concerted bifunctional mechanisms
I and II where a carboxylate anion acts as a nucleophile and an un-ionized carboxyl
functions as a general acid with proton donation to oxygen (Scheme 1). However,
Maugh and Bruice (14) considered that the reactions of salicyl succinate involve
nucleophilic attack by a carboxylate anion and a changing inductive effect as the
second carboxyl group ionizes, or a “series nucleophilic” reaction in which there are
two stepwise nucleophilic processes. The same general conclusions were drawn in
regard to other proposed bifunctional mechanisms. A properly located chelated metal
ion might function with great effect analogous to a general acid in a bifunctional
reaction, and there would be no kinetic ambiguity (Eq. [1]).1
[1]
Determination of the effects of metal ions in intramolecular nucleophilic reactions
is difficult: first, because of the necessity of the metal ion binding to the substrate in
a position that is remote from the nucleophile, and secondly, because of the extremely
favorable metal ion-promoted OHϪ catalysis that is always observed in the hydrolysis
of esters having metal ion liganding groups (19–26) The facile OHϪ reaction restricts
competition from other mechanistic pathways (23). Rate enhancements in metal ion-
promoted OHϪ reactions of acyl derivatives are very large (19,20,23,25–31), ranging
up to 108-fold with Cu2ϩ (1010-fold with an acyl phosphate) (31). In contrast, a
significant effect (102- to 104-fold) has only been observed in metal ion-promoted
carboxyl nucleophilic reactions in cases where the leaving group is an aliphatic alcohol
and the metal ion can complex the leaving group oxygen in the transition state
(24,25). Complexed divalent metal ions were previously found to have little or no
1 For an analogy of general acid catalysis and metal ion catalysis in acetal hydrolysis see Ref. (18).

HYDROLYSIS OF ESTERS
359
SCHEME 1
effect in carboxyl group nucleophilic reactions of esters of 8-hydroxyquinoline and
2-carboxy-8-hydroxyquinoline (22,23). However, there have been no previous studies
of the effects of a metal ion strongly complexed to the carbonyl oxygen of an ester
with a carboxyl group substituent capable of functioning as a nucleophile.2,3 If a
carboxyl nucleophilic mechanism is to be considered chemically feasible for an
esterase metalloenzyme, then such a mechanism must be capable of generating much
larger rate constants and rate enhancements than have hitherto been observed in
chemical reactions. It is therefore important to investigate systems that are maximally
suited to provide large rate constants.
The 1,10-phenanthrolyl ligand has advantages for the study of metal ion effects,
among which is the very strong complexing ability of the phenanthroline nitrogens
(23,31–34). The position of the metal ion in a 1:1 complex can often be specified.
Furthermore, the neutral liganding groups will minimize the charge reduction on the
metal ion that results with negatively charged ligands. In esters of 2-hydroxy-1,10-
phenanthroline a metal ion will be complexed by the phenanthroline nitrogens and
can also complex with the ester carbonyl oxygen in a stable 6-membered chelate
ring. A phthalate monoester would allow intramolecular carboxyl group attack via a
kinetically favored 5-membered ring transition state in which the strength of metal
ion binding should increase as negative charge is generated on oxygen. Therefore,
we have determined the effects of the divalent metal ions Cu2ϩ, Ni2ϩ, Zn2ϩ, and Co2ϩ
in the reactions of 2-(1,10-phenanthrolyl) hydrogen phthalate (III) and for comparison,
the corresponding monoglutarate (IV) and acetate (V) esters (Scheme 2). The carboxyl
group substituent of III and IV will not be bound to the metal ion in a 1:1 complex
because of the unfavorable large ring size required for such chelation.
2 The metal ion complexation of the carbonyl oxygen of salicyl 1,10-phenanthroline-2-carboxylate
could be strong, but nucleophilic attack by the carboxylate anion would only result in a reversible
rearrangement and is therefore not detected (23).
3 Metal ion complexation of the carbonyl oxygen of amides results in rate retardation in the neighboring
carboxyl group reaction (32).

360
FIFE AND PUJARI
SCHEME 2
EXPERIMENTAL
Materials. 2-Hydroxy-1,10-phenanthroline was prepared by the synthetic scheme
previously outlined (35) and was recrystallized from a benzene-ethanol mixture, mp
159–160 ЊC; lit. (35), mp 159–160 ЊC. The synthesis of the esters III and IV
was carried out by dissolving freshly prepared 2-hydroxy-1,10-phenanthroline in
tetrahydrofuran that had been distilled and dried. An equivalent amount of the appro-
priate anhydride was added and the solution was refluxed for at least 48 h. The solvent
was then removed by rotary evaporation. The oil residue was then extracted several
times with dry benzene. The solid that was obtained was recrystallized from dry
chloroform or benzene. The acetate ester V was prepared similarly except that excess
acetic anhydride was employed. 2-(1,10-Phenanthrolyl) hydrogen phthalate (III) had
mp 142ЊC after recrystallization. Anal. Calcd for C20H12N2O4: C, 69.76; H, 3.51; N,
8.12. Found: C, 69.94; H, 3.49; N, 8.12. 2-(1,10-Phenanthrolyl) hydrogen glutarate
(IV) had mp 148 ЊC after recrystallization. Anal. Calcd for C17H14N2O4: C, 65.80;
H, 4.55; N, 9.03. Found: C, 65.69; H, 4.53; N, 9.13. 2-(1,10-Phenanthrolyl) acetate
(V) had mp 152 ЊC after recrystallization. Anal. Calcd for C14H10N2O2: C, 70.58; H,
4.23; N, 11.76. Found: C, 70.55; H, 4.30; N, 11.78. Spectra were consistent with the
expected structures. The esters are sensitive to air, but can be stored in a desiccator
in the dark at low temperature.
Acetonitrile was Eastman Kodak Spectrograde. Amine buffer components were
recrystallized or distilled prior to use.
Kinetic measurements. The rates of hydrolysis of III–V were measured with a
Beckman DU-7500 or a Pye Unicam SP8-100 spectrophotometer by following the
absorbance increase at 300–320 nm. The buffers employed were formate, acetate,
cacodylate, 2,6-lutidine, N-ethylmorpholine, Tris, morpholine, and carbonate. All
buffer solutions were maintained at a constant ionic strength of 0.1 M with KCl. The
buffer concentration was 0.02 M; buffer catalysis is negligible at that concentration.
A typical kinetic run was initiated by injecting 20–30 ␮L of a 0.004 M stock solution

HYDROLYSIS OF ESTERS
361
of the ester in acetonitrile into 3 mL of buffer maintained at 30 ЊC. Product spectra
at the conclusion of the reactions matched those of identical concentrations of 2-
hydroxy-1,10-phenanthroline and the appropriate carboxylic acid. The solutions em-
ployed for kinetic measurements in the absence of metal ions included 2 ϫ 10Ϫ5
M EDTA.
Reactions that were too fast to be measured with a conventional spectrophotometer
were monitored with a Durrum D-110 stopped-flow spectrophotometer equipped with
a Hewlitt-Packard Model 1207B storage oscilloscope. The substrate was dissolved
at the desired concentration in 10Ϫ2–10Ϫ3 M HCl, ␮ ϭ 0.1 M with KCl. This solution
was introduced into one of two identical drive syringes. The second syringe contained
the appropriate buffer maintained at the same ionic strength and the chosen metal
ion (Cu2ϩ, Ni2ϩ, Zn2ϩ, and Co2ϩ). With each buffer three to four reactions were
tabulated. The observed reactions followed good first-order kinetics, and infinity
points were stable. The hydrolysis of phthalic anhydride or glutaric anhydride has
no effect on the observed kinetics in the reactions of III and IV monitored at 320
nm. Pseudo-first-order rate constants (kobsd ) were computer-calculated.
Reaction mixture pH values were measured with a Radiometer-type PHM 22A pH
meter equipped with a combination electrode. In calculating second-order rate con-
stants for hydroxide ion catalysis (kOH ) the ion product of water, Kw, was taken to
be 1.47 ϫ 10Ϫ14 at 30ЊC.
RESULTS
In Fig. 1 a plot is shown of log kobsd vs pH for hydrolysis of 2-(1,10-phenanthrolyl)-
hydrogen phthalate (III). The plot has a slope of 1.0 at high pH, which indicates a
FIG. 1. Plot of log kobsd vs pH for the hydrolysis of 2-(1,10-phenanthrolyl) hydrogen phthalate in
H2O at 30ЊC, ␮ ϭ 0.1 M with KCl, in the presence of 0.002 M Cu2ϩ
in the absence of metal ion (᭪).
(
) and 0.004 M Zn2ϩ (ⅷ), and

362
FIFE AND PUJARI
hydroxide ion-catalyzed reaction. At pH 6–9 there is a pH-independent reaction, and
at pH Ͻ 6 the plot of Fig. 1 is bell shaped. The reaction scheme is that of Eq. [2].
[2]
The equation for kobsd is then
k1K1aH ϩ k2K1K2 ϩ kOHK1K2 [OHϪ]
kobsd
ϭ
,
[3]
2
aH ϩ K1aH ϩ K1K2
where K1 and K2 are the first and second dissociation constants. Thus, the pH-
independent region of Fig. 1 represents a reaction of the anionic species, while the
bell shaped portion of the profile is due to a reaction of the neutral species, or more
likely a kinetic equivalent zwitterion (ionized carboxyl group and protonated nitrogen).
Values of the rate constants are given in Table 1. The apparent pKa values of 4.0 and
5.0 provide a good fit to the experimental data employing Eq. [3]. There is, however,
some positive deviation at low pH (Ͻ 3) which indicates the incursion of a protonated
species reaction. The solid line in Fig. 1 is theoretical for Eq. [3] and the constants
in Table 1.
A similar plot of log kobsd vs pH is also produced in the hydrolysis of 2-(1,10-
phenanthrolyl) hydrogen glutarate (IV), as seen in Fig. 2, and again Eq. [3] provides
a good fit for the experimental data with pKapp 3.5 and 4.5. The rate constants for
IV in Table 1 are several-fold larger than those of III.
The divalent metal ions Cu2ϩ, Ni2ϩ, Co2ϩ, and Zn2ϩ bind strongly to III–V. Figure
3 is a plot of kobsd vs the concentration of Zn2ϩ in the hydrolysis of IV at pH 3.2.
Saturation occurs at Zn2ϩ concentrations less than 0.005 M. The equation for kobsd at
constant pH is
kMKM [M2ϩ]
(1 ϩ KM [M2ϩ])
kobsd
ϭ
,
[4]
where KM is the metal ion association constant and kM is the rate constant at the

HYDROLYSIS OF ESTERS
TABLE 1
363
Values of the Rate Constants for Hydrolysis of 2-(1,10-Phenanthrolyl) hydrogen Phthalate (III),
Glutarate (IV), and Acetate (V) in Water at 30ЊC and ␮ ϭ 0.1 M with KCl
Compound
Metal ion
k1 (sϪ1
2 ϫ 10Ϫ3
)
k2 (sϪ1
)
kOH (MϪ1 sϪ1
)
pKapp
III
None
Cu2ϩ
Ni2ϩ
Zn2ϩ
Co2ϩ
None
Cu2ϩ
Ni2ϩ
Zn2ϩ
Co2ϩ
None
Cu2+
Ni2ϩ
Zn2ϩ
Co2ϩ
1.3 ϫ 10Ϫ4
4.2 ϫ 10Ϫ1
1.8 ϫ 10Ϫ1
4.5 ϫ 10Ϫ2
1.6 ϫ 10Ϫ2
7.0 ϫ 10Ϫ4
2.5 ϫ 10Ϫ1
1.3 ϫ 10Ϫ1
1.3 ϫ 10Ϫ1
6.7 ϫ 10Ϫ2
2.5
4.0, 5.0
4.3
4.5
5.0
5.1
3.5, 4.5
4.4
4.15
4.15
4.2
1.1 ϫ 108
9.0 ϫ 107
2.1 ϫ 106
1.0 ϫ 106
12.2
IV
V
1 ϫ 10Ϫ2
6.8 ϫ 107
2.7 ϫ 107
4.3 ϫ 107
4.0 ϫ 106
10.2
1.2 ϫ 106
2.7 ϫ 105
9.0 ϫ 104
8.8 ϫ 104
specified pH for hydrolysis at a saturating concentration of the metal ion. Values of
the association constants are given in Table 2.
Plots of log kobsd vs pH are shown in Figs. 1 and 2 at saturating concentrations of
Cu2ϩ, Zn2ϩ, and Co2ϩ. Similar plots were also obtained with Ni2ϩ, but were omitted
FIG. 2. Plot of log kobsd vs pH for the hydrolysis of 2-(1,10-phenanthrolyl) hydrogen glutarate in
H2O at 30ЊC, ␮ ϭ 0.1 M with KCl, in the presence of 0.003 M Zn2ϩ (ⅷ) and 0.003 M Co2ϩ (᭾), and
in the absence of metal ion (᭪).

364
FIFE AND PUJARI
FIG. 3. Plot of kobsd vs the concentration of Zn2ϩ at pH 3.20 in the hydrolysis of 2-(1,10-phenanthrolyl)
hydrogen glutarate in H2O at 30ЊC, ␮ ϭ 0.1 M with KCl.
from the figures for clarity. The plots have a slope of 1.0 at pH Ͼ 6, and a pH-
independent region from pH 4 to 6, which is followed by a downward bend to again
give a line of slope 1.0 at lower pH. The reaction scheme for hydrolysis of the 1:1
metal ion complexes VI–VIII is given in Eq. [5].
TABLE 2
Values of the Association Constants for Binding of Divalent Metal Ions with 2-(1,10-Phenanthrolyl)
hydrogen Phthalate (III), Glutarate (IV), and Acetate (V) in
Water at 30ЊC and ␮ ϭ 0.1 M with KCl
Compound
Metal ion
KM MϪ1
III
Cu2ϩ
Ni2ϩ
Zn2ϩ
Co2ϩ
Cu2ϩ
Ni2ϩ
Zn2ϩ
Co2ϩ
Cu2ϩ
Ni2ϩ
Zn2ϩ
Co2ϩ
2 ϫ 104a
7 ϫ 103a
2.5 ϫ 104b
7 ϫ 103b
1 ϫ 103c
3.5 ϫ 103a
5 ϫ 103a
4 ϫ 103a
105d
IV
V
3 ϫ 104d
3 ϫ 104d
1 ϫ 104d
a At pH 3.20.
b At pH 4.28.
c At pH 2.80.
d At pH 6.20.

HYDROLYSIS OF ESTERS
365
VI
[5]
In these reactions the ratio of metal ion to reactant is Ͼ 100:1 to ensure a 1:1 complex.
The equation for kobsd as a function of pH at saturating metal ion concentration is then
k2ЈKaaH ϩ kOHЈKaKaЈ
kobsd
ϭ
.
[6]
2
aH ϩ KaaH ϩ KaKaЈ
The slopes of 1.0 at pH Ͼ 6 show that pKaЈ has not been approached at the highest
pH employed, that is, aH Ͼ KaЈ at all experimental pH values.4 Equation [6] then
reduces to Eq. [7] at aH Ͼ KaЈ. At Ka Ͼ aH Eq. [8] is then obtained.
k2ЈKa ϩ kOH ЈKaKaЈ/aH
kobsd
ϭ
[7]
[8]
aH ϩ Ka
k
OH ЈKaЈ
aH
kobsd ϭ k2Ј ϩ
.
The lower arm of the profile will have kobsd given by
k2ЈKa
aH
kobsd
ϭ
.
[9]
4 The pKa values for acid ionization of the aquo complexes of the metal ions at 25ЊC are Cu2ϩ, 6.8;
Zn2ϩ, 8.8; Co2ϩ, 8.9; and Ni2ϩ, 10.6 (36).

366
FIFE AND PUJARI
These equations assume a mechanism involving intramolecular attack of metal ion
bound OHϪ on the carbonyl carbon at high pH. However, at pH values below the
pKa of metal ion bound water, a kinetically equivalent reaction would be attack of
external hydroxide ion on the metal ion complex VII. Equations [10] and [11] would
then follow, so that at high pH Eq. [12] results.
k2ЈKaaH ϩ kOH KaKw
kobsd
kobsd
ϭ
ϭ
[10]
2
aH ϩ KaaH
k2ЈKa ϩ kOH Ka[OHϪ]
aH ϩ Ka
[11]
[12]
kobsd ϭ kOH [OHϪ].
The values of kOH, the second-order rate constant for hydroxide ion catalysis of the
hydrolysis of the metal ion complex, are reported in Table 1. The lines in Figs. 1 and
2 for the metal ion-promoted reactions are theoretical for Eq. [11] and the values of
the constants in Table 1.
Hydroxide ion catalysis also occurs in the hydrolysis of 2-(1,10-phenanthrolyl)
acetate (V) as seen in Fig. 4. The value of kOH at 30ЊC is 10.2 MϪ1 secϪ1, which is
only slightly less than the kOH found for IV. However, in contrast with IV, the reaction
of V becomes slow at low pH. The divalent metal ions again bind strongly to the
phenanthroline nitrogens of V. In Fig. 4 the plot of log kobsd vs pH for hydrolysis of
FIG. 4. Plot of log kobsd vs pH for the hydrolysis of 2-(1,10-phenanthrolyl) acetate in H2O at 30ЊC,
␮ ϭ 0.1 M with KCl, in the presence of 0.003 M Cu2ϩ
of metal ion (᭪).
(
) and 0.003 M Co2ϩ (᭾) and in the absence

HYDROLYSIS OF ESTERS
367
the Cu(II) and Co(II) complexes of V is shown. Only metal ion-promoted OHϪ
reactions are observed. The plots are linear with slopes of 1.0. The plot for hydrolysis
of the Cu(II) complex is linear at pH values from 3 to 6.5 with kOH ϭ 1.2 ϫ 106
MϪ1 sϪ1.
DISCUSSION
Intramolecular nucleophilic carboxyl group participation occurs in the hydrolysis
of several types of esters, including glutarate and phthalate monoesters (9–15,37).
The “steric fit” of the carboxyl group to the carbonyl carbon is highly important in
the nucleophilic reactions (10,15); nucleophilic attack via a 5-membered ring transition
state is 230-fold more favorable than through a 6-membered ring (10). This difference
is most likely due to the increased restriction of unfavorable rotational states in the
ester reacting via a 5-membered ring (10,15).
The neighboring carboxylate anions of III and IV are sterically able to participate
in the hydrolysis reactions as efficient nucleophiles. The large rate constants in the
pH-independent reactions (pH 6–9) show convincingly that carboxyl participation is
occurring. The anhydrides produced in these reactions hydrolyze rapidly (10,12).5
Phthalic anhydride has been shown to be an intermediate in the hydrolysis of phthalate
monoesters (12). The ease with which nucleophilic attack can take place might be
expected to be greater with III than IV since a 5-membered ring transition state can
be formed in the former case. The carboxyl group of III is rigidly held adjacent to
the carbonyl group. Nevertheless, the rate constant k2 for the nucleophilic reaction
of the anionic species of III (Eq. [3]) is 5-fold less than with IV. This is most likely
a reflection of the more deactivated carbonyl group of the phthalate ester. Note that
the kOH values for III and IV are in the same ratio as the k2 values, while kOH for
IV and the acetate ester V are nearly the same.
The bell-shaped kobsd vs pH profiles at low pH indicate that the neutral species of
III and IV also undergo a rapid reaction. It is likely that the kinetically equivalent
zwitterion is the reactive species (ionized carboxyl group and protonated phenanthro-
line nitrogen) since the carboxyl anion would be a better nucleophile than the neutral
species and the protonated phenanthroline would provide a better leaving group (IX)
(Scheme 3). Again, k1 is larger in the case of the glutarate ester than with III.
Metal ion catalysis. Metal ion effects in the intramolecular nucleophilic reactions
of esters with good leaving groups have previously been undetectable; the metal ion-
promoted OHϪ reaction is the favored pathway even at low pH in the hydrolysis of
phenolic esters (23). When the leaving group is an aliphatic alcohol, a complexed
metal ion can provide assistance for breakdown of a tetrahedral intermediate to
products (24,25). With a poor leaving group the OHϪ pathway is relatively less
favorable, which then allows other mechanisms to be observable. In none of the
previously studied nucleophilic reactions has it been sterically possible for the metal
ion to complex favorably with the carbonyl oxygen. Such complexation should further
increase the efficiency of the metal ion-promoted OHϪ pathway but might also greatly
5 The rate constant for pH-independent hydrolysis of phthalic anhydride in H2O at 30ЊC has been
reported as 1.2 ϫ 10Ϫ2 sϪ1 (12).

368
FIFE AND PUJARI
SCHEME 3
increase the efficiency of the carboxyl nucleophilic reaction. The present study shows
that this is indeed the case.
The divalent metal ions, Cu2ϩ, Ni2ϩ, Zn2ϩ, Co2ϩ, bind very strongly to III–V. A
stable 6-membered chelate ring (VI–VIII) can be formed with complexation of the
carbonyl oxygen. Enhancements of 105- to 108-fold are obtained for kOH in the apparent
OHϪ reactions of the 1:1 complexes of III and IV as compared with the reactions
in the absence of metal ions. The reactions could involve either an internal attack of
metal ion bound OHϪ (X) or the attack of external OHϪ (XI) on the metal ion
complex (Scheme 4). The hydrolysis of the metal ion complexes of N-acylimidazoles
must involve a reaction of metal ion bound OHϪ because the alternative external
attack analogous to XI demands a second-order rate constant greater than that of a
diffusion-controlled reaction (30). Cupric ion is the most effective promoter of the
hydrolysis reactions of III–V, while Co2ϩ is least effective in the series of metal ions,
although Co2ϩ is still capable of producing a sizable rate enhancement.
At pH Ͻ 7 the metal ion-promoted reactions of III and IV become pH independent.
Subsequently, the profiles of Figs. 1 and 2 break downward to give slopes of 1.0.
These reactions are clearly dependent on the neighboring carboxyl group; at lower
pH (Ͻ 6) the only proton that can dissociate from the metal ion complex is that of
SCHEME 4

HYDROLYSIS OF ESTERS
369
the carboxyl group. Note that the Cu2ϩ-promoted hydrolysis of the acetate ester V
has a log kobsd vs pH profile that is linear with a slope of 1.0 at pH values as low as
3.5. The pKapp values in the reactions of III and IV are well below the pKa of metal
ion-bound water (36). At higher pH the plots of log kobsd vs pH are linear to pH
values above 7; the water pKa values must then be still higher. Since the metal ion
saturation is maintained at low pH values, as shown by the plots of kobsd vs [M2ϩ],
the pH-independent reaction and subsequent downward bend of the log kobsd vs pH
profile are due to the ionization of the neighboring carboxyl group.
The observed metal ion-promoted reactions in Figs. 1 and 2 could be influenced
by the changing inductive effect of the carboxyl group produced by ionization. An
un-ionized carboxyl group is more electron withdrawing than a carboxylate anion,
which would lead to greater reactivity for the un-ionized species in a OHϪ reaction.
Therefore, the reactions might be considered as OHϪ reactions of the ionized and
un-ionized species with the near pH-independent regions of the profiles representing
a transition between these reactions. However, with the glutarate monoester IV the
carboxyl inductive effect must be transmitted through three saturated carbon atoms
and would therefore be small. On that basis it would be predicted that there would
be a wider separation of the rate constants kOH Ј for the reactions of the two species
with III than IV; but the opposite is the case. It can be seen in Fig. 2 that the near
pH-independent reaction of the Co(II) complex of IV extends over 2 pH units, which
is clearly more than can be accounted for by a small change in the inductive effect.
Thus, the metal ion-promoted reactions at pH Ͻ 6 involve participation by the
carboxylate anion.
A nucleophilic reaction (XII) is analogous to the reactions in the absence of metal
ions (Scheme 5). The enhancements in k2Ј in the carboxyl nucleophilic reaction
provided by the complexed metal ions range from 102- to 103-fold. A metal ion-
promoted carboxyl group reaction can therefore be competitive with the metal ion-
promoted OHϪ reaction at pH Ͻ 6. The rate constants k2Ј for reactions of the Cu(II)
and Ni(II) complexes of III are sufficiently large that they are comparable to the rate
constant for the pH-independent hydrolysis of phthalic anhydride (12).5
The apparent pKa values of 4.2–4.4 in the hydrolysis of the metal ion complexes
SCHEME 5

370
FIFE AND PUJARI
of the glutarate ester are in accord with the expected pKa of the ␥-carboxyl group.
However, the apparent pKa values found in the hydrolysis of the metal ion complexes
of the phthalate ester (4.3–5.1) are higher than expected and show some dependence
on the identity of the metal ion. The measured pKa of monomethyl phthalate (30 ЊC)
is 3.32 (12), and that of phthalic acid monoamide is 3.75 (38). Any direct interaction
of the carboxyl group with the metal ion should stabilize the carboxylate anion and
therefore reduce the pKa. Likewise, a very favorable equilibrium prior to the rate-
limiting step would reduce the apparent pKa (39). The high pKapp values for III could
reflect either a possible hydrogen bonding interaction of the carboxyl group with
metal ion-bound water that stabilizes the proton on the carboxyl oxygen, or proximity
of the carboxyl to the negative charge of the partially ionized metal ion bound water.
On the other hand, in the stable extended conformation of the glutarate monoester,
the carboxyl group can point out into the solvent (15).
A change in rate-determining step can also give rise to a downward bend in a pH
rate constant profile but is unlikely in the reactions of III and IV. Breakdown of a
tetrahedral intermediate should be rate determining in the scheme of Eq. [13] at all
pH in view of the greater ease of the reverse step than expulsion of the leaving group
to give products. The steady-state equation for kobsd in the neighboring carboxyl group
reaction at metal ion saturation is Eq. [14],
[13]
kfkp
Ka
kobsd
ϭ
,
[14]
ͫ ͬ
kr ϩ kp Ka ϩ aH
where kf, kr, and kp are the rate constants for the formation of a tetrahedral intermediate
and the decomposition of the intermediate to reactants and products, respectively.
Release of a carboxylate anion from the tetrahedral intermediate to regenerate the
reactant should be relatively favorable. Therefore, if as expected, the rate constant
kr Ͼ kp, then kp is the rate constant of the rate-determining step. Breakdown to products
would be more difficult than is usually the case because it would involve breaking
a stable 6-membered chelate ring. A high pKapp is then best explained by an environ-
mentally perturbed pKa for the carboxyl group of III in the metal ion complex.
A metal ion complexed to the carbonyl oxygen will greatly facilitate the nucleophilic
attack step but will inhibit breakdown of a tetrahedral intermediate to products by
restricting electron release from oxygen (Eq. [13]). Large rate enhancements are

HYDROLYSIS OF ESTERS
371
achieved by metal ion complexation despite this necessary compromise of effects.
The opposition of metal ion effects on the two steps of the reaction explain in part
the large difference in catalytic efficiency for a chelated metal ion in the OHϪ and
carboxyl group-promoted reactions; the OHϪ reaction should proceed with rate-
determining nucleophilic attack, whereas the carboxyl group reaction should have
rate-determining breakdown of the tetrahedral intermediate to products.
The metal ion-promoted carboxyl nucleophilic reactions of III and IV are the
most favorable intramolecular carboxyl group reactions that have been observed. The
reactions are general examples of metal ion promotion in intramolecular reactions
where the metal ion can favorably complex the carbonyl group oxygen. The log rate
constant pH profiles of Figs. 1 and 2 show that both of the possible mechanisms,
metal ion-promoted OHϪ catalysis and metal ion-promoted carboxyl group attack,
are represented on a single profile, and this is because of the increased favorability
of the carboxyl group reaction. The former mechanism is predominant at pH Ͼ 6,
and the latter at pH Ͻ 6.
Carboxypeptidase A. The pH-rate constant profiles for the hydrolysis of the metal
ion complexes of III and IV are similar in shape to the plots of kcat vs pH obtained
in the carboxypeptidase A (CPA)-catalyzed hydrolysis of cinnamoyl-L, ␤-phenyllactate
(16,17). The ascending arm of those plots occurs at pH Ͼ 8.5, i.e., ϳ2 pH units
higher than with III and IV. The sigmoidal portion of the kcat vs pH plot at lower
pH has a pKapp of 6.2, which may be the pKa of the ␥-carboxyl group of Glu-270.
The pH-independent value of kcat in the enzymatic reaction (195 sϪ1 at 30 ЊC) (16)
is only ϳ500-fold larger than k2Ј with III and IV. However, in contrast with the
excellent leaving group of III and IV, the leaving group alcohol of the ␤-phenyllactate
ester is of much higher pKa; that pKa was calculated to be 14.7 (40). Therefore, the
enzymatic reaction is considerably more favorable than the metal ion-promoted car-
boxyl nucleophilic reactions of III and IV.6 The steric fit of the carboxyl group to
the reaction center of the phthalate ester III would be difficult to improve upon. The
rate constants k2Ј for III and IV are then close to maximal for metal ion promoted
carboxyl group participation in the reactions of esters having a phenanthrolyl liganding
group. A simple Glu-270 nucleophilic mechanism in the CPA-catalyzed hydrolysis
of esters can then be considered doubtful on chemical grounds. Metal ion-promoted
OHϪ catalysis, as with III and IV at pH Ͼ 6, is the only mechanism found to date
that is capable of generating rate constants of the magnitude obtained in the CPA
reactions. Thus, a reasonable mechanism for the enzyme reaction would be XIII
(8,41) (Scheme 6). Mechanism XIII is kinetically equivalent to a mechanism in which
the Glu-270 anion partially abstracts a proton from a metal ion bound water molecule in
the transition state, but XIII employs the better nucleophile. The kinetically equivalent
carboxylate anion general base mechanism is not likely since there should be a
significant concentration of Zn(II)–OHϪ at pH values from 6 to 9. Even if the
concentration was 1% at pH 7 (assuming a pKa of metal ion bound water of 9), the
reaction should go 100% via the ionized species because of the great nucleophilicity.
6 When the leaving group is a phenol with a carbonate diester substrate for CPA, the values of kcat are
reduced (40).

372
FIFE AND PUJARI
SCHEME 6
ACKNOWLEDGMENT
This work was supported by a research grant from the National Science Foundation.
REFERENCES
1. Lipscomb, W. N. (1970) Acc. Chem. Res. 3, 81.
2. Ludwig, M. A., and Lipscomb, W. N. (1973) in “Inorganic Biochemistry” (Eichhorn, G., Ed.), Vol.
1, p. 438, Elsevier, New York.
3. Kaiser, E. T., and Kaiser, B. L. (1972) Acc. Chem. Res. 5, 219.
4. Christianson, D. W., and Lipscomb, W. N. (1989) Acc. Chem. Res. 22, 62.
5. Auld, D. S. and Vallee, B. L. (1987) in “Hydrolytic Enzymes” (Neuberger, A., and Broclehurst, K.,
Eds.), Chap 4, p. 201, Elsevier, New York.
6. Suh, J., Cho, W., and Chung, S. (1985) J. Am. Chem. Soc. 107, 4530.
7. Britt, B. M., and Peticolas, W. L. (1992) J. Am. Chem. Soc. 114, 5295.
8. Osumi, A., Rahmo, A., King, S., Przystas, T. J., and Fife, T. H. (1994) Biochemistry 33, 14750.
9. Bruice, T. C., and Benkovic, S. J. (1966) “Bioorganic Mechanisms,” Benjamin, New York, and
references therein.
10. Bruice, T. C., and Pandit, U. K. (1960) J. Am. Chem. Soc. 82, 5858.
11. Gaetjens, E., and Morawetz, H. (1960) J. Am. Chem. Soc. 82, 5328.
12. Thanassi, J. W., and Bruice, T. C. (1966) J. Am. Chem. Soc. 88, 747.
13. Morawetz, H., and Oreskes, I. (1958) J. Am. Chem. Soc. 80, 2591.
14. Maugh II, T., and Bruice, T. C. (1971) J. Am. Chem. Soc. 93, 3237.
15. Lightstone, F. C., and Bruice, T. C. (1996) J. Am. Chem. Soc. 118, 2595.
16. King, S. W., and Fife, T. H. (1983) Biochemistry 22, 3603.
17. Hall, P. L., Kaiser, B. L., and Kaiser, E. T. (1969) J. Am. Chem. Soc. 91, 485.
18. Fife, T. H., and Przystas, T. J. (1987) J. Chem. Soc. Perkin Trans. 2, 143.
19. Hay, R. W. (1987) in “Comprehensive Coordination Chemistry” (Wilkinson, G., Ed.), pp. 411–485,
Pergamon, Oxford, and references therein.
20. Hay, R. W., and Clark, C. R. (1977) J. Chem. Soc., Dalton Trans., 1866, 1993.
21. Wells, M. A., and Bruice, T. C. (1977) J. Am. Chem. Soc. 99, 5341.
22. Fife, T. H., and Squillacote, V. L. (1978) J. Am. Chem. Soc. 100, 4787.
23. Fife, T. H., Przystas, T. J., and Squillacote, V. L. (1979) J. Am. Chem. Soc. 101, 3017.
24. Fife, T. H., and Przystas, T. J. (1980) J. Am. Chem. Soc. 102, 7297.
25. Fife, T. H., and Przystas, T. J. (1982) J. Am. Chem. Soc. 104, 2251.
26. Fife, T. H., and Przystas, T. J. (1985) J. Am. Chem. Soc. 107, 1041.
27. Groves, J. T., and Dias, R. M. (1979) J. Am. Chem. Soc. 101, 1033.
28. Groves, J. T., and Chambers, R. R. (1984) J. Am. Chem. Soc. 106, 630.
29. Chin, J. (1991) Acc. Chem. Res. 24, 145.
30. Fife, T. H., and Przystas, T. J. (1986) J. Am. Chem. Soc. 108, 4631.
31. Fife, T. H., and Pujari, M. P. (1990) J. Am. Chem. Soc. 112, 5551.
32. Fife, T. H., and Squillacote, V. L. (1977) J. Am. Chem. Soc. 99, 3762.

HYDROLYSIS OF ESTERS
373
33. Holyer, R. H., Hubbard, C. D., Kettle, S. F. A., and Wilkins, R. G. (1965) Inorg. Chem. 4, 929.
34. Bell, R. S., and Sutin, N. (1962) Inorg. Chem. 1, 359.
35. Fife, T. H., and Pujari, M. P. (1988) J. Am. Chem. Soc. 110, 7790.
36. Basolo, F., and Pearson, R. G. (1967) “Mechanisms of Inorganic Reactions,” 2nd ed., p. 32, Wiley,
New York.
37. Fersht, A. R., and Kirby, A. J. (1968) J. Am. Chem. Soc. 90, 5833.
38. Kortum, G., Vogel, W., and Andrussow, K. (1961) “Dissociation Constants of Organic Acids in
Aqueous Solution,” p. 392, 513, Butterworths, London.
39. Bruice, T. C., and Schmir, G. L. (1959) J. Am. Chem. Soc. 81, 4552.
40. King, S. W., Lum, V. R., and Fife, T. H. (1987) Biochemistry 26, 2294.
41. Fife, T. H. (1991) “Perspectives in Bioinorganic Chemistry” (Hay, R. W., Ed.), p. 43, JAI Press, London.