Kinetic and thermodynamic studies on the cyanation reactions and base-on/base-off equilibria of alkyl-13-epicobalamins

Article abstract of DOI:10.1039/b304069k

Ligand substitution equilibria of two different 13-epicobalamins (X-13-epiCbl, X = NCCH₂ and CN^-) with cyanide have been studied. It was found that CN^- substitutes the 5,6- dimethylbenzimidazole (DMBz) moiety in the α-position to form X(CN)Cbl-13epi, which for X = NCCH₂ in the presence of CN^- subsequently gives (CN)₂Cbl-13epi. The kinetics of the displacement of DMBz by CN^- showed saturation behaviour at high cyanide concentration and the limiting rate constants are characterized by the activation parameters: X = NCCH₂, ΔH^≠ = 83 ± 1 kJ mol^-1, ΔS^≠ = +77 ± 4 J K^-1 mol^-1, ΔV^≠ = +13.3 ± 1.0 cm³ mol^-1; X = CN^-, ΔH^≠ = 106 ± 1 kJ mol^-1, ΔS^≠ = +82 ± 4 J K^-1 mol^-1 and ΔV^≠ = +14.8 ± 0.5 cm³ mol^-1. These parameters are interpreted in terms of a limiting D mechanism. The rate constants for the displacement of DMBz in the case of the 13-epicobalamins were found to be slower than those obtained in the case of the analogous alkylcobalamins, and consequently, the thermodynamic equilibrium constants for the 13-epicobalamins were found to be smaller than those obtained in the case of the alkylcobalmins. This clearly shows the effect of the epimerization of the e-side chain attached to the C-13 of the corrin ring on the rate and equilibrium constants for these ligand displacement reactions. The Royal Society of Chemistry 2003.

Full text of DOI:10.1039/b304069k

View Article Online / Journal Homepage / Table of Contents for this issue
Kinetic and thermodynamic studies on the cyanation reactions and
base-on/base-off equilibria of alkyl-13-epicobalamins†
Mohamed S. A. Hamza,a,b Xiang Zou,c Kenneth L. Brownc and Rudi van Eldik*a
a Institute for Inorganic Chemistry, University of Erlangen-Nürnberg, Egerlandstr.1,
91058 Erlangen, Germany. E-mail: vaneldik@chimie.uni-erlangen.de
b Department of Chemistry, Faculty of Science, Ain Shams University, Cairo, Egypt
c Department of Chemistry and Biochemistry, Ohio University, Athens, Ohio 45701, USA
Received 14th April 2003, Accepted 29th May 2003
First published as an Advance Article on the web 24th June 2003
Ligand substitution equilibria of two different 13-epicobalamins (X-13-epiCbl, X = NCCH2 and CNϪ) with
cyanide have been studied. It was found that CNϪ substitutes the 5,6-dimethylbenzimidazole (DMBz) moiety in
the α-position to form X(CN)Cbl-13epi, which for X = NCCH2 in the presence of CNϪ subsequently gives
(CN)2Cbl-13epi. The kinetics of the displacement of DMBz by CNϪ showed saturation behaviour at high cyanide
concentration and the limiting rate constants are characterized by the activation parameters: X = NCCH2, ∆H≠ =
83 1 kJ molϪ1, ∆S≠ = ϩ77 4 J KϪ1 molϪ1, ∆V ≠ = ϩ13.3 1.0 cm3 molϪ1; X = CNϪ, ∆H≠ = 106 1 kJ molϪ1, ∆S≠ =
ϩ82 4 J KϪ1 molϪ1 and ∆V ≠ = ϩ14.8 0.5 cm3 molϪ1. These parameters are interpreted in terms of a limiting D
mechanism. The rate constants for the displacement of DMBz in the case of the 13-epicobalamins were found to be
slower than those obtained in the case of the analogous alkylcobalamins, and consequently, the thermodynamic
equilibrium constants for the 13-epicobalamins were found to be smaller than those obtained in the case of the
alkylcobalmins. This clearly shows the effect of the epimerization of the e-side chain attached to the C-13 of the
corrin ring on the rate and equilibrium constants for these ligand displacement reactions.
ation. X-Ray crystal structures of CN-13-epiCbl21,22 and
Introduction
molecular mechanics calculations23 showed that epimerization
There is significant interest in studying the axial ligand substi-
tution reactions of cobalt–corrinoid complexes since it is
known that the two active cobalamins (methylcobalamin,
MeCbl, and coenzyme B12, AdoCbl) undergo substitution of
at C13 leads to twisting of the C pyrrol ring and an “upwardly”
axial disposition of the e side chain. Epimerization at C13
causes a decrease in pKbase-off of about 0.83 0.14 across the
series of XCbl and a corresponding increase in KCo, for a given
their axial benzimidazole ligand by a protein histidine residue
X. KCo was found to be 6.9
1.8 fold higher for the X-13-
during complexation to the CH3Cbl-dependent methionine
synthase,1,2 the class I3 AdoCbl-dependent mutases (methyl-
malonyl-coenzyme A mutase and glutamate mutase),4,5 and the
class III AdoCbl-dependent -lysine-5,6-aminomutase.6 As a
result, comparison of the ligand substitution reactions trans
to the axial alkyl ligand in coenzyme B12 and its derivatives to
that trans to a non-alkyl ligand is of interest to further our
understanding of the mechanisms of these reactions.
The corrin ring (macrocyclic tetrapyrrole ring) does not pos-
sess a plane of symmetry, and consequently, there is a possibil-
ity of axial ligand diastereomerism when the two axial ligands
are different.7–13 The “upper”, or β-face, of the corrinoid, with
its upward projecting a, c and g acetamide side chain, is less
sterically hindered than the “lower”, or α-face, which is brack-
eted by the downward projecting b, d and e propionamides and
the secondary amide f side chain. The electronic structure of
the corrin ring has a significant effect on the properties of the
Co() ion. It has been shown recently that displacement of H
at C10 by an electron-withdrawing NO group increases the pKa
for coordinated H2O, renders the Co–O bond more ionic
and significantly deactivates the metal ion towards ligand
substitution.14
It is known that many structurally modified analogs of
AdoCbl, including those modified in the side chains,15 in the
corrin ring itself,16 in the nucleotide loop15a,17 and in the adenine
moiety itself,18 are active coenzymes with some AdoCbl-
dependent enzymes. These active analogs are very important in
understanding the mechanism by which the B12-requiring
enzymes activate AdoCbl. An interesting example of these
active coenzyme analogs is the C13 epimer,19,20 in which the e
propionamide side chain adopts an “upwardly” axial configur-
epiCbl’s than for the XCbl’s.24
We have been interested in the application of different kinetic
and thermodynamic high pressure techniques to study the
ligand substitution reactions of a series of alkylcobalamins
(XCbl) in efforts to understand the mechanism of these reac-
tions and to investigate how the alkyl group (X) in the trans
position controls the kinetics and thermodynamics of these
reactions. In this work, the fascinating influence of a single
cobalt–carbon bond in tuning the reactivity of the Co()
center was clearly demonstrated.25–29 We found that the reaction
of XCbl with CNϪ, where X = NCCH2, CF3 and CNϪ,
proceeded via the formation of an intermediate X(CN)Cbl,
which subsequently reacts, in the presence of CNϪ, to give
(CN)2CblϪ.26 We also found that the mechanism of substituting
the axial DMBz trans to the X group in a series of XCbl by
CNϪ occurred via a dissociative mechanism.25,26
In order to further improve our understanding of the reac-
tion mechanism of such cobalamins and their derivatives, we
have now investigated the effect of epimerization at C13 on the
kinetics and thermodynamics of ligand substitution reactions
of these complexes. We have studied the cyanation reactions of
two alkyl-13-epicobalamins, in which the alkyl group was
NCCH2 and CNϪ. For this purpose the kinetics of the substi-
tution reactions with cyanide was studied as a function of
nucleophile concentration, temperature and pressure.
Experimental
Materials
All chemicals were of analytical grade and used as received.
CAPS buffer was purchased from Sigma. HClO4, NaClO4 and
NaCN were purchased from Merck. Ultra pure water was
used in the kinetic and thermodynamic measurements. The
sample preparations and all measurements were carried out in
† Electronic supplementary information (ESI) available: Figs. S1–3:
additional spectroscopic and kinetic data. See http://www.rsc.org/
suppdata/dt/b3/b304069k/
D a l t o n T r a n s . , 2 0 0 3 , 2 9 8 6 – 2 9 9 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
2986

View Article Online
diffuse light since alkyl-13-epicobalamins, like all other alkyl-
cobalamins, are known to be very light sensitive.30
and is stable in the dark. We previously reported that the reac-
tions of XCbl with CNϪ to form X(CN)CblϪ, where X = Me,
CH2Br, Pr, Et and Ado, are also too fast to be monitored by
stopped-flow techniques.25,26
Cyanocobalamin was supplied by Sigma. H2OCbl was from
Roussel. The CN-13-epiCbl was prepared as described in the
literature19,20,24 by acid-catalyzed epimerization of CNCbl in
anhydrous trifluoroacetic acid. CN-13-epiCbl was converted to
the alkyl-13-epiCbl’s as described previously.24 These complexes
were purified and characterized by UV-visible, 1H, 13C and
amide 15N NMR spectroscopy as described previously.22,31
Kinetics of the reaction of ꢀ-NCCH2-13-epiCbl with cyanide
Characteristic spectral changes are observed for the reaction of
β-NCCH2-13epi-Cbl with cyanide (see ESI,† Fig. S1). The
product spectrum suggests the formation of the intermediate
complex, NCCH2(CN)-13-epiCbl, since new bands appeared at
383, 568 and 606 nm (sh), and this complex is light sensitive, but
indefinitely stable in the dark. Fig. 1 shows a plot of kobs vs.
[CNϪ] for the reaction of 4 × 10Ϫ5 M β-NCCH2-13-epiCbl with
excess CNϪ ([CNϪ] = 0.025–1.0 M) at pH 11.0, I = 1.0 M
(NaClO4) and 5.0 ЊC. This plot shows that the observed rate
constant decreases on increasing the [CNϪ] and then reaches a
limiting value of kobs at high [CNϪ], which equals k1 in reaction
(2) according to the rate law (derived under steady state condi-
tions for the intermediate species) given in (3), indicating that
Instrumentation and measurements
The pH of the solutions was measured using a Mettler Delta
350 pH meter with a combined glass electrode. It was calibrated
with standard buffer solutions at pH 7.0 and 10.0. UV-Vis
spectra were recorded on Shimadzu UV-2101 and Hewlett
Packard 8452A spectrophotometers.
Analytical HPLC was performed on a 4.6 × 250 mm
Beckman C8 ultrasphere column while semipreparative HPLC
was performed on a 10 × 250 mm Beckman C8 ultrasphere
column.24
Kinetic measurements were carried out on an Applied
Photophysics SX 18MV stopped-flow instrument coupled to an
online data acquisition system. At least eight kinetic runs were
recorded under all conditions, and the reported rate constants
represent the mean values. All kinetic measurements were
carried out under pseudo-first order conditions, i.e. the nucleo-
phile concentration was at least in ten-fold excess. Measure-
ments under high pressure were carried out using a home-made
high pressure stopped-flow instrument.32 Kinetic traces were
analysed with the OLIS KINFIT program (Bogart, GA, USA).
The UV-Vis spectrophotometers and stopped-flow instru-
ments were thermostated to the desired temperature 0.1 ЊC.
Values of ∆H≠ and ∆S≠ were calculated from the slopes and
intercepts, respectively, of plots of ln(k/T ) vs. 1/T , and values
of ∆V ≠ were calculated from the slope of plots of ln(k) vs.
pressure.
(3)
k
Ϫ2 > k1 under such conditions. Interesting to note is that in the
case of NCCH2Cbl, the reverse was found, viz. k1 ӷ kϪ2
.
26 The
intercept of the plot can be assigned to a contribution of the
back reaction, whereas the observed curvature can be con-
sidered as evidence in favor of a limiting D mechanism as
outlined in eqn. (2). The kinetic data are for the substitution
reaction that involves displacement of α-DMBz by CNϪ as
shown in reaction (1).
Results and discussion
Preliminary experiments at pH 11.0, in which the UV-Vis spec-
trum was scanned in the range of 300 to 700 nm, showed that
CNϪ reacts rapidly with the X-13-epiCbl complexes and that
the equilibria are established within the mixing and measure-
ment time. The reactions investigated in the present study can
be summarized by reaction (1), in which the displacement of
DMBz by cyanide (KCN1) to give X(CN)-13-epiCbl occurs.
(1)
Fig. 1 Plot of kobs vs. [CNϪ] for the reaction between β-NCCH2-
13-epiCbl and CNϪ at pH 11.0, 5.0 ЊC and I = 1.0 M (NaClO4); the solid
line is a fit to eqn. (3) in the text.
On the basis of all the available data, the suggested mechan-
ism for the reaction between β-NCCH2-13-epiCbl and CNϪ
can be represented by reaction (2), which involves dechelation
of DMBz to form a six-coordinate aqua intermediate. The data
in Fig. 1 were fitted to eqn. (3) and resulted in k1 = 14.0 0.6
In a preliminary study we first attempted to follow the
kinetics of the reaction of Me-13-epiCbl with CNϪ at pH 11.0
(Caps buffer) and 5.0 ЊC. This reaction was found to be too fast
to be followed on the stopped-flow instrument (deadtime of ca.
2–4 ms). The product of the reaction, Me(CN)-13-epiCbl,
shows a slight increase in absorbance in the range 580–600 nm
s
Ϫ1, kϪ2 = 32 2 sϪ1 and k2/kϪ1 = 25 7 MϪ1 at 5.0 ЊC, from
which an overall equilibrium constant, KCN = k1k2/kϪ1kϪ2 = 11
(2)
D a l t o n T r a n s . , 2 0 0 3 , 2 9 8 6 – 2 9 9 1
2987

View Article Online
Table 1 Kinetic data for the reaction of (X)-13-epiCbl with CNϪ as a
function of temperaturea
kobs b/sϪ1
T /ЊC
X = CN
X = NCCH2
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
14.9 0.4
30
56
100
1
4
6
0.00203 0.0002
0.00404 0.0002
0.0090 0.0005
0.01845 0.0001
0.0368 0.0007
0.0725 0.0007
180 13
∆H≠/kJ molϪ1
106
ϩ82
1
4
83
ϩ77
1
4
∆S≠/J KϪ1 molϪ1
a Experimental conditions: for X = NCCH2: [NCCH2-13-epiCbl] =
4 × 10Ϫ5 M, [CNϪ] = 1.0 M, pH 11.0, I = 1.0 M (NaClO4). For
X = CN: [CN-13-epiCbl] = 5 × 10Ϫ5 M, [CNϪ] = 0.25 M, pH 11.0,
I = 0.5 M (NaClO4). b Under the selected experimental conditions,
kobs = k1
2 MϪ1 can be calculated. This value is in a very good agreement
with the spectrophotometrically determined value of 13.3
1.5 MϪ1 reported below, but is significantly smaller than that
reported for NCCH2Cbl, viz. 63.8 M .
Ϫ1 33 This difference shows
that the affinity of CNϪ for cobalamin is ca. 6 times higher than
for 13-epicobalamin. The value of k1 for the reaction of
26
NCCH2Cbl with CNϪ (64.7 sϪ1
)
is significantly higher than
that obtained here for NCCH2-13-epiCbl (viz. 14.0 sϪ1). The
value of k2/kϪ1, which represents the efficiency of cyanide com-
pared to DMBz of the nucleotide loop to scavenge the six-co-
ordinate intermediate, was found to be 25 MϪ1 for of NCCH2-
13-epiCbl, which is very similar to that of NCCH2Cbl (viz. 27.9
Fig. 2 (a) Plot of ln kobs vs. pressure for the reaction between
β-NCCH2-13-epiCbl and CNϪ measured at 1.0 M CNϪ; the best fit of
the data (solid line) gives ∆V ≠ = ϩ13.3 1.0 at 0.0 ЊC. (b) Plot of ln kobs
vs. pressure for the reaction between β-CN-13-epiCbl and CNϪ
M
Ϫ1).26 The difference between k1 for NCCH2Cbl and NCCH2-
13-epiCbl can indeed be ascribed to the presence or absence of
the e-side chain on the α face, which decreases k1 in the case of
13-epicobalamin and consequently also decreases the overall
equilibrium constant. This suggests that in the normal Cbl, the
presence of a “downward” projecting propionamide side chain
at C13 sterically destabilizes the base-on species.
measured at 0.25 M CNϪ; the best fit of the data (solid line) gives ∆V ≠
ϩ14.8 1.2 cm3 molϪ1 at 30.0 ЊC.
=
The reaction between β-NCCH2-13-epiCbl and CNϪ was
studied as a function of temperature and pressure at a high
cyanide concentration (1.0 M), i.e. where kobs = k1, and the
results are reported in Table 1 and Fig. 2(a), respectively.
Fig. 2(a) demonstrates a good linear correlation between ln(k)
and pressure. The activation parameters ∆H≠ and ∆S≠ were
found to be 83 1 kJ molϪ1 and ϩ77 4 J KϪ1 molϪ1, respect-
ively, and the activation volume, ∆V ≠ = ϩ13.3 1.0 cm3 molϪ1
at 0.0 ЊC. These data, along with the observed rate law
and saturation kinetics for cyanide substitution, suggest that
the first step of the reaction of β-NCCH2-13-epiCbl with CNϪ
(k1) indeed follows a limiting D mechanism, i.e. a dissociative
dechelation of DMBz followed by coordination of water to the
vacant coordination site. It was found previously that ∆H≠, ∆S≠
and ∆V ≠ for the reaction of NCCH2Cbl with CNϪ are 85 2 kJ
molϪ1, ϩ97 6 J KϪ1 molϪ1 and ϩ12.7 0.5 cm3 molϪ1, respect-
ively.26 It seems clear that these data are indeed comparable,
suggesting that a limiting D mechanism is operating in both
cases. We have also recently reported values for ∆V ≠ of ϩ13.1
and ϩ14.8 cm3 molϪ1 for the reactions of CNCbl and β-CF3Cbl
with CNϪ.25,26 Furthermore, the volume of activation for the
reaction of β-(N-methylimidazolyl)cobalamin with N-methyl-
imidazole was also reported to be significantly positive, viz.
ϩ15.0 0.7 and ϩ16.8 1.1 cm3 molϪ1 at 5 × 10Ϫ3 and 1 M
N-methylimidazole, respectively, which correspond to the
aquation of (N–MeIm)2Cblϩ and the dechelation reaction of
the α-DMBz of (N–MeIm)Cblϩ, respectively.34 These ∆V ≠
values are typical of those expected for a limiting D substi-
tution mechanism for an octahedral complex.35,36
Fig. 3 Plot of kobs vs. [CNϪ] for the reaction between CN-13-epiCbl
and CNϪ at pH 11.0, 30.0 ЊC and I = 0.5 M (NaClO4); the solid line is a
fit to eqn. (3) in the text.
Kinetics of the reaction of CN-13-epiCbl with CN؊
Fig. 3 shows a plot of kobs vs. [CNϪ] for the reaction of 5 × 10Ϫ5
M CN-13-epiCbl with excess CNϪ ([CNϪ] = 0.01 to 0.25 M) at
pH 11.0, I = 0.5 M (NaClO4) and 30.0 ЊC. This plot also shows
saturation kinetics at high cyanide concentrations but in the
opposite direction from the NCCH2-13-epiCbl case reported
above. A limiting value of kobs is reached at high [CNϪ], and the
intercept and observed curvature can be interpreted in the same
way as before. The reported kinetic data are for the substitution
D a l t o n T r a n s . , 2 0 0 3 , 2 9 8 6 – 2 9 9 1
2988

View Article Online
of α-DMBz by CNϪ, since new bands were observed at 367, 540
(sh) and 580 nm, characteristic for (CN)2-13-epiCbl.
The data in Fig. 3 were fitted to eqn. (3) and resulted in k1 =
0.0207 0.0002 sϪ1, kϪ2 = (2.0 0.3) × 10Ϫ3 sϪ1 and k2/kϪ1
=
31 2 MϪ1 at 25.0 ЊC, from which an overall equilibrium con-
stant KCN = k1k2/kϪ1kϪ2 = 320 60 MϪ1 was calculated. This
value is in a good agreement with the spectrophotometrically
determined value of 350 21 MϪ1 reported below. Interestingly,
in this case, k1 ӷ kϪ2, which accounts for the opposite trend in
the cyanide concentration dependence of kobs (compare Figs. 1
and 3) and is similar to that reported previously in the case of
CNCbl.26 The limiting rate constant value k1 (0.0207 sϪ1) is
smaller than that obtained in the case of CNCbl (0.042 sϪ1).26,37
However, the value of k2/kϪ1 (31 2 MϪ1), which represents the
efficiency of cyanide compared to DMBz of the nucleotide loop
to scavenge the six-coordinate intermediate, was found to
be similar to that obtained previously in the case of CNCbl
(36.9 MϪ1).26
Fig. 4 Plot of kobs vs. [HClO4] for the reaction between β-NCCH2-
13epiCbl and HClO4 at 5.0 ЊC and I = 2 M (NaClO4); the solid line is a
fit to eqn. (4) in the text.
The reaction between CN-13-epiCbl and CNϪ was studied as
a function of temperature and pressure at a high cyanide con-
centration (0.25 M), i.e. where kobs = k1, and the results are
reported in Table 1 and Fig. 2(b), respectively. Fig. 2(b) demon-
strates a good linear correlation between ln(k) and pressure.
The activation parameters ∆H≠ and ∆S≠ were found to be 106
with increasing acid concentration. This linear increase in kobs
is ascribed to an acid-catalyzed dechelation reaction, the slope
of which represents the rate constant for the acid-catalyzed
process (kH) and the intercept of which represents the rate
constant for the spontaneous dechelation of DMBz (viz. k1).
The increase in kobs with increasing acid concentration was
in this case not as prominent as in the case of XCbl,27 because
the value of pKbase-off in the present case is only 0.85,24 which
means that more acidic conditions are required to reach
significant concentrations of the protonated base-off cobal-
amin. It is reasonable to conclude that dechelation of DMBz to
give β-NCCH2-13-epiCbl (base-off ) is catalyzed by acid, in
agreement with recently reported data for the dechelation of
DMBz in a series of XCbl’s.27
1 kJ molϪ1 and ϩ82
4 J KϪ1 molϪ1, respectively, and the
activation volume, ∆V ≠ = ϩ14.8 1.2 cm3 molϪ1 at 30.0 ЊC.
These data along with the observed rate law suggest that the
reaction of CN-13-epiCbl with CNϪ also follows a limiting D
mechanism involving dissociative dechelation of DMBz to
form a six-coordinate intermediate aqua complex. This mech-
anism is in agreement with our suggestion for the reactions of
NCCH2-13-epiCbl and XCbl (X = CNϪ, CF3, NCCH2)25,26 with
CNϪ and that suggested by Reenstra and Jencks,37 where they
also found that the rate of CNϪ addition to CNCbl reaches a
limiting value at high [CNϪ].
In terms of the overall base-on/base-off equilibration for the
spontaneous and acid-catalyzed pathways discussed in detail
before,27 the rate law for the acidification reaction of β-NCCH2-
13-epiCbl is given in eqn. (4), in which k1 and kϪ1 represent the
rate constants for the spontaneous dechelation and chelation
of X-13-epiCbl in reaction (2), respectively, kH represents the
acid-catalysed reaction path, and KBz represents the acid-
dissociation constant for the dechelated DMBzHϩ species.
Kinetics of the base-on/base-off equilibration
We 27 have recently shown that evidence for the suggested dis-
sociative mechanism for the reaction of XCbl with CNϪ could
be obtained from the direct measurement of the dechelation
rate constant of DMBz (k1) through acidification of XCbl to
produce the protonated base-off species. However, in the pres-
ent study (X-13-epiCbl), pKbase-off values are less than those
reported for XCbl, viz. pKbase-off for CN-13-epiCbl and NCCH2-
13-epiCbl were found to be Ϫ0.9 and 0.85, respectively.24 These
low values require a very high acid concentration to dechelate
DMBz in the α-position. On the other hand, the DMBz dis-
sociation rate of NCCH2-13-epiCbl and CN-13-epiCbl can be
followed with the stopped flow technique, since, from the elec-
tron inductive effect of these substituents, the rate of DMBz
dissociation is expected to be slower for these complexes than
for those with strongly donating X substituents. The DMBz
dissociation rate of other X-13-epiCbl’s, such as Me-13-epiCbl
and Et-13-epiCbl, are much too fast to be monitored by
stopped-flow techniques.
(4)
A non-linear least squares fit of the data shown in Fig. 4 to
the rate law in (4) using pKBz = 5.56,38 results in k1 = 39 3 sϪ1
k
,
Ϫ1 = (1.5 0.1) × 105 sϪ1 and kH = 121 7 MϪ1 sϪ1. It follows
that KCo (= kϪ1/k1) equals 3700, which is less than the value
previously reported from thermodynamic considerations (viz.
5.1 × 104).24 The value of k2/kϪ1 calculated from the linear fit of
the experimental data obtained from the cyanide concentration
dependence (see above) to eqn. (3) was found to be 25 MϪ1
.
Kinetic results for the stopped-flow “pH-jump” reaction of
6 × 10Ϫ5 M β-NCCH2-13-epiCbl (pH 5, unbuffered solution)
with different concentrations of HClO4 (0.01–1.0 M) at I = 2 M
(NaClO4) and 5.0 ЊC are shown in Fig. 4. The values of kobs first
decrease with increasing acid concentration in the lower con-
centration range (0.005–0.05 M), and then gradually increase
linearly with increasing acid concentration (0.05–0.3 M). A
combination of these concentration dependencies results in the
curved dependence observed in Fig. 4. The first part of the plot
at low acid concentration is incomplete since it was not possible
to follow the reaction at lower acid concentrations where it
becomes too fast to be followed by stopped-flow and the change
in absorbance becomes too small. The second part of the plot is
expected to level off to reach k1, but shows an increase in kobs
Combined with the value of kϪ1 (1.5 × 105 sϪ1), the value of k2
can be calculated to be 3.75 × 106 MϪ1 sϪ1. The value of k1 is not
in good agreement with the acid-independent limiting rate con-
stant (14.0 sϪ1) obtained when the nucleophile concentration
dependence was studied for the reaction of β-NCCH2-13-
epiCbl and CNϪ (see above). The likely reason for the inconsis-
tency of the results is the low value of pKbase-off which requires a
very high concentration of HClO4. In addition, the Hϩ-
catalyzed contribution to the observed rate constant is already
greater than the uncatalyzed rate at [HClO4] = 0.5 M, such that
the limiting value of k1 cannot be reached.
Similar results were obtained for the reaction of CN-13-
epiCbl with HClO4. The data for the reaction of 8 × 10Ϫ5
M
CN-13-epiCbl with HClO4 (0.01–1.0 M) at 30.0 ЊC are shown in
D a l t o n T r a n s . , 2 0 0 3 , 2 9 8 6 – 2 9 9 1
2989

View Article Online
Table 2 Kinetic, thermodynamic and activation parameter data for the reaction of (X)Cbl and (X)-13-epiCbl with CNϪ
Cbla
13-epiCblb
NCCH2
X =
NCCH2
CN
CN
k1 c/sϪ1
64.7 0.7
27.4 0.8
27.9 2.7
63.8 1.4e
0.042 0.001
14.0 0.6
0.0207 0.0002
kϪ1 c/sϪ1
(7 3) × 10Ϫ5
32
25
2
7
(2.0 0.3) × 10Ϫ3
(k2/kϪ1)c/MϪ1
KCN d/MϪ1
36.9 1.5
1 × 104 f
105
ϩ81
31
2
13.3 1.5
350 21
∆H≠/kJ molϪ1
∆S≠/J KϪ1 molϪ1
∆V ≠/cm3 molϪ1
85
ϩ97
2
6
2
6
83
ϩ77
1
4
106
ϩ82
1
4
ϩ12.7 0.5
ϩ13.1 0.3
ϩ13.3 1.0
ϩ14.8 1.2
a Data from ref. 26 unless otherwise stated. b This work. c Calculated from a non-linear fit of the experimental data to eqn. (3) in the text.
d Determined by spectrophotometric method. e Ref. 33. f Ref. 37.
the ESI†(Fig. S2). The data again show a decrease in kobs and
then level off at high concentration of HClO4, and no sign for
an acid catalyzed reaction was observed in this case. This is
expected since the pKbase-off value in this case is Ϫ0.9, which will
require a much higher acid concentration for the acid-catalyzed
pathway to play a meaningful role.24 The non-linear least
squares fit of the data to the rate-law in eqn. (4) without the
kH[Hϩ] term, resulted in k1 = 0.18 0.03 sϪ1 and kϪ1 = (1.79
0.06) × 104 sϪ1, respectively. It follows that KCo = 9.9 × 104,
which is less than that reported before, viz. 2.9 × 106.24 The
value of k2/kϪ1 calculated from the linear fit of the experimental
data obtained from the cyanide concentration dependence
(see above) to eqn. (3) was found to be 31 MϪ1. Combined with
the value of kϪ1 (1.79 × 104 sϪ1), it follows that k2 = 5.6 ×
tively, and Ax is the absorbance at any cyanide concentration.
The values of KCN1 and A∞ were calculated from eqn. (5), and
KCN1 was found to be 13.3 1.5 MϪ1. This value is in a good
agreement with that determined kinetically (11 2 MϪ1). The
analysis of these data by plotting log (Ax Ϫ A0)/(A∞ Ϫ Ax) vs.
log[CNϪ] gave a good linear plot with a slope of 1.09 0.05,
which indicates one CNϪ ligand is coordinated to the cobalt
atom.
A similar spectrophotometric titration was carried out for
CN-13-epiCbl with CNϪ and the spectral changes that accom-
pany this titration are shown in Fig. 6. The spectrophotometric
titration was monitored by following the increase in absorbance
at 580 nm, where the largest change in absorbance occurred,
and selected data are shown in the ESI†(Fig. S3). The
value of KCN1 in this case was found to be 350 21 MϪ1. This
value is in a good agreement with that determined kinetically
(viz. 320 MϪ1).
105 MϪ1 sϪ1
.
By way of comparison, the value of k1 for the dechelation
and kϪ1 for the chelation reaction of NCCH2-13-epiCbl were
found to be 39 and 1.5 × 105 sϪ1, respectively, whereas those for
CN-13-epiCbl were found to be 0.18 and 1.79 × 104 sϪ1, respect-
ively. These rate constants nicely show the effect of the axial
group (X) in the upper β-position on the substitution rate con-
stant of DMBz located in the α-position, trans to X, by CNϪ.
Equilibrium measurements
The spectrophotometric titration of β-NCCH2-13-epiCbl with
CNϪ was monitored by following the increase in absorbance at
600 nm, where the largest change in absorbance occurred.
Selected data are shown in Fig. 5, where the solid line represents
a fit of the data to eqn. (5).
Ax = A0 ϩ (A∞ Ϫ A0)KCN1[CNϪ]/(1 ϩ KCN1[CNϪ]) (5)
The values of A0 and A∞ represent the absorbances of
β-NCCH2-13-epiCbl and (β-NCCH2)(α-CN)-13-epiCbl, respec-
Fig. 6 UV-Vis spectra of CN-13-epiCbl recorded in the presence
of various concentrations of CNϪ (5 × 10Ϫ4–0.011 M) at pH 11.0,
I = 0.5 and at 30.0 ЊC.
Overall discussion
Table 2 summarizes the kinetic, thermodynamic and activation
parameter data obtained for the reaction of XCbl and X-13-
epiCbl with CNϪ. It is clear from this table that the values of
KCN for the reactions of NCCH2-13-epiCbl and CN-13-epiCbl
with CNϪ (13.3 and 350 MϪ1, respectively) are smaller than
those obtained for the same reactions of NCCH2Cbl and
CNCbl with cyanide (63.8 and 104 MϪ1, respectively).33,37 Table
2 also shows the significant decrease in the rate constant, k1, for
13-epicobalamins compared to the cobalamins. These differ-
ences are ascribed to the epimerization of the e side chain from
a “downward” axial position to an “upward” axial position,
which makes the displacement of DMBz (α-position) by CNϪ
slower in the case of the 13-epicobalamins and the overall equi-
librium constant smaller. It is already known that epimerization
Fig. 5 Change in absorbance at 600 nm on addition of CNϪ to
NCCH2-13-epiCbl; the solid line is a fit to eqn. (5) in the text and results
in KCN1 = 13.3 1.5 MϪ1
.
D a l t o n T r a n s . , 2 0 0 3 , 2 9 8 6 – 2 9 9 1
2990

View Article Online
8 Y. W. Alelyunas, P. E. Fleming, R. G. Finke, T. G. Pagano and
at C13 causes a decrease in the pKbase-off value of about 0.83
0.14 across the series of XCbl’s and a corresponding increase in
KCo for a given X. KCo was found to be 6.9 1.8 fold higher for
the X-13-epiCbl’s than for the XCbl’s.24
L. G. Marzilli, J. Am. Chem. Soc., 1991, 113, 3781.
9 K. L. Brown and X. Zou, J. Am. Chem. Soc., 1992, 114, 9643.
10 X. Zou, K. L. Brown and C. Vaughn, Inorg. Chem., 1992, 31, 1552.
11 K. L. Brown and X. Zou, Inorg. Chem., 1992, 31, 2541.
12 K. L. Brown, L. Salmon and J. A. Kirby, Organometallics, 1992, 11,
422.
13 X. Zou and K. L. Brown, J. Am. Chem. Soc., 1993, 115, 6689.
14 H. M. Marques, L. Knapton, X. Zou and K. L. Brown, J. Chem.
Soc., Dalton Trans., 2002, 3195.
15 (a) M. I. Yakusheva, A. A. Pozanskaya, T. A. Pospeiova, I. P.
Rudakova, A. M. Yurkevich and V. A. Yakovlev, Biochim. Biophys.
Acta., 1977, 484, 216; (b) T. Toraya, E. Krodel, A. S. Mildvan and
R. H. Abeles, Biochemistry, 1979, 189, 417; (c) C. G. D. Morley,
R. L. Bakley and H. P. C. Hogenkamp, Biochemistry, 1968, 7, 1231;
(d ) D. L. Anton, P. K. Tsai and H. P. C. Hogenkamp, J. Biol. Chem.,
1980, 255, 4507; (e) T. Toraya and S. Fukui, Adv. Chem. Ser., 1980,
191, 139.
In terms of the substitution lability of the X-13-epiCbl’s, the
results of this study have shown that epimerization at C13 has a
sensitive influence on the relative magnitudes of the rate con-
stants in the reaction sequence (2), which through rate law (3)
can have a remarkable influence on the observed dependence of
the pseudo first order substitution rate constant on the entering
nucleophile concentration as demonstrated by the results in
Figs. 1 and 3, and summarized in Table 2. This is caused by the
relative size of k1 and kϪ2, which controls the rate of the ligand
dissociation steps in the forward and reverse directions, and is
very sensitive to epimerization at C13.
Structural studies on the ground state trans effect suggest
that epimerization at C13 does not affect the Co–C bond
enthalpies, and would therefore be unlikely to affect the
enthalpies of activation for Co–C bond cleavage but tend to
increase the activation entropies.39 The values of ∆H≠, ∆S≠ and
∆V ≠ for the reaction of X-13-epiCbl with CNϪ are included
in Table 2. The values for ∆S≠ and ∆V ≠ show the same
mechanistic trends and support the suggested dissociative
mechanism. The substitution reactions of β-NCCH2-13-epiCbl
and β-CN-13-epiCbl with CNϪ involve rate-determining dis-
placement of α-DMBz that proceeds through a limiting D
mechanism.
This study has revealed how the alkyl ligands in XCbl and
X-13-epiCbl control the kinetics and thermodynamics of the
reaction with CNϪ, and how epimerization at C-13 decreases
the rate constant for the axial ligand substitution reaction, but
does not affect the mechanism of the substitution process.
16 (a) T. Tamao, Y. Morikawa, S. Shimizu and S. Fukui, Biochim.
Biophys. Acta., 1968, 151, 260; (b) T. Toraya, T. Shirakashi, S. Fukui
and H. P. C. Hogenkamp, Biochemistry, 1975, 14, 3949.
17 (a) T. Toraya and A. Ishida, J. Biol. Chem., 1991, 266, 5430;
(b) T. Toraya, S. Miyoshi, M. Mori and K. Wada, Biochim. Biophys.
Acta., 1994, 1204, 169.
18 (a) G. N. Sando, R. L. Blakley, H. P. C. Hogenkamp and P. J.
Hoffman, J. Biol. Chem., 1975, 250, 8774; (b) K. Ushio, S. Fukui and
T. Toraya, Biochim. Biophys. Acta, 1984, 788, 318; (c) T. Toraya,
T. Matsumoto, M. Ichikawa, T. Itoh, T. Sugawara and Y. Mizuno,
J. Biol. Chem., 1986, 261, 9289.
19 R. Bonnett, J. M. Godfrey, V. B. Math, E. Edmond, H. Evans and
O. J. R. Hodder, Nature, 1971, 229, 473.
20 R. Bonnett, J. M. Godfrey and V. B. Math, J. Chem. Soc. C, 1971,
3736.
21 H. Stoeckli-Evans, E. Edmond and D. C. Hodgkin, J. Chem. Soc.,
Perkin Trans. 2, 1972, 605.
22 K. L. Brown, D. R. Evans, J. D. Zubkoswski and E. J. Valente, Inorg.
Chem., 1996, 35, 415.
23 H. M. Marques and K. L. Brown, J. Mol. Struct. (THEOCHEM),
1995, 340, 97.
24 K. L. Brown and G.-Z. Wu, Inorg. Chem., 1994, 33, 4122.
25 M. S. A. Hamza, X. Zou, K. L. Brown and R. van Eldik,
Inorg. Chem., 2001, 40, 5440.
26 M. S. A. Hamza, X. Zou, K. L. Brown and R. van Eldik, J. Chem.
Soc., Dalton Trans., 2002, 3832.
27 M. S. A. Hamza, X. Zou, K. L. Brown and R. van Eldik, Eur. J.
Inorg. Chem., 2003, 268.
28 N. E. Brasch, M. S. A. Hamza and R. van Eldik, Inorg. Chem., 1997,
36, 3216.
Acknowledgements
The authors gratefully acknowledge financial support from the
Deutsche Forschungsgemeinschaft (to R. v. E.), the Alexander
von Humboldt Foundation (fellowship to M. S. A. H.), and the
National Institute of General Medical Sciences, USA (Grant
GM 48858 to K. L. B.). M. S. A. H. thanks the Ain Shams
University for sabbatical leave.
29 M. S. A. Hamza, A. G. Cregan, N. E. Brasch and R. van Eldik,
Dalton Trans., 2003, 596.
30 Chemistry and Biochemistry of B12, ed. R. Banerjee, Wiley & Sons
Inc.New York, 1999.
31 K. L. Brown, H. B. Brooks, B. D. Gupta, M. Victor, H. M. Marques,
D. C. Scooby, W. J. Goux and R. Timkovich, Inorg. Chem., 1991, 30,
3430.
References
1 C. L. Drennan, S. Huang, J. T. Drummond, R. G. Matthews and
M. L. Ludwig, Science, 1994, 266, 1669.
2 C. L. Drennan, R. G. Matthews and M. L. Ludwig, Curr. Opin.
Struct. Biol., 1994, 4, 919.
3 B. T. Golding and W. Buckel, in Comprehensive Biological Catalysis,
ed. M. L., Sinnott, Academic Press, London, 1997, vol. III, pp. 239–
259.
32 R. van Eldik, W. Gaede, S. Wieland, T. Kraft, M. Spitzer and
D. M. Palmer, Rev. Sci. Instrum., 1993, 64, 1355.
33 K. L. Brown, J. Am. Chem. Soc., 1987, 109, 2277.
34 A. G. Cregan, N. E. Brasch and R. van Eldik, Inorg. Chem., 2001,
40, 1430.
4 F. Mancia, N. H. Keep, A. Nakagawa, P. F. Leadlay, S. McSweeney,
B. Ramussen, P. Boscke, O. Diat and P. R. Evans, Structure, 1996, 4,
229.
5 R. Reiter, G. Gruber, G. Jogl, V. G. Wagner, H. Bothe, V. Buckel and
C. Kratky, Structure, 1999, 7, 891.
6 C. H. Chang and P. A. Frey, J. Biol. Chem., 2000, 275, 106.
7 K. L. Brown, X. Zou and L. Salmon, Inorg. Chem., 1991, 30,
1949.
35 T. W. Swaddle, Rev. Phys. Chem. Jpn., 1980, 50, 230.
36 (a) R. van Eldik, T. Asano and W. J. Le Noble, Chem. Rev., 1989,
89, 549; (b) A. Drljaca, C. D. Hubbard, R. van Eldik, T. Asano,
M. V. Basilevsky and W. J. Le Noble, Chem. Rev., 1998, 98, 2167.
37 W. W. Reenstra and W. P. Jencks, J. Am. Chem. Soc., 1979, 101, 5780.
38 K. L. Brown, J. M. Hakimi, D. M. Nuss, Y. D. Montejano and
D. W. Jacobsen, Inorg. Chem., 1984, 23, 1463.
39 K. L. Brown, X. Zou and D. R. Evans, Inorg. Chem., 1994, 33, 5713.
D a l t o n T r a n s . , 2 0 0 3 , 2 9 8 6 – 2 9 9 1
2991