On the nature of mutual inactivation between [Cp*Rh(bpy)(H2O)]2+ and enzymes - analysis and potential remedies

Article abstract of DOI:10.1016/j.molcatb.2010.01.006

Pentamethylcyclopentadienyl rhodium bipyridine ([Cp*Rh(bpy)(H₂O)]²⁺) is a versatile catalyst to promote biocatalytic redox reactions. However, its major drawback lies in the mutual inactivation of [Cp*Rh(bpy)(H₂O)]²⁺ and the biocatalyst. This interaction was investigated using the alcohol dehydrogenase from Thermus sp. ATN1 (TADH) as model enzyme. TADH binds 4equiv. of [Cp*Rh(bpy)(H₂O)]²⁺ without detectable decrease in catalytic activity and stability. Higher molar ratios lead to time-, temperature-, and concentration-dependent inactivation of the enzyme suggesting [Cp*Rh(bpy)(H₂O)]²⁺ to function as an 'unfolding catalyst'. This detrimental activity can be circumvented using strongly coordinating buffers (e.g. (NH₄)₂SO₄) while preserving its activity as NAD(P)H regeneration catalyst under electrochemical reaction conditions.

Full text of DOI:10.1016/j.molcatb.2010.01.006

Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
On the nature of mutual inactivation between [Cp*Rh(bpy)(H2O)]2+
and enzymes – analysis and potential remedies
Maël Poizat, Isabel W.C.E. Arends, Frank Hollmann∗
Delft University of Technology, Department of Biotechnology, Biocatalysis and Organic Chemistry Group, Julianalaan 136, 2628 BL Delft, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Pentamethylcyclopentadienyl rhodium bipyridine ([Cp*Rh(bpy)(H2O)]2+) is a versatile catalyst to pro-
mote biocatalytic redox reactions. However, its major drawback lies in the mutual inactivation of
[Cp*Rh(bpy)(H2O)]2+ and the biocatalyst. This interaction was investigated using the alcohol dehydroge-
nase from Thermus sp. ATN1 (TADH) as model enzyme. TADH binds 4 equiv. of [Cp*Rh(bpy)(H2O)]2+
without detectable decrease in catalytic activity and stability. Higher molar ratios lead to time-,
temperature-, and concentration-dependent inactivation of the enzyme suggesting [Cp*Rh(bpy)(H2O)]2+
to function as an ‘unfolding catalyst’. This detrimental activity can be circumvented using strongly coor-
dinating buffers (e.g. (NH4)2SO4) while preserving its activity as NAD(P)H regeneration catalyst under
electrochemical reaction conditions.
Article history:
Received 7 November 2009
Received in revised form 5 January 2010
Accepted 6 January 2010
Available online 11 January 2010
Keywords:
Asymmetric catalysis
Bioelectrochemistry
Cofactor regeneration
Enzyme inactivation
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
[20–22]. (4) Its catalytic activity is not confined to the regenera-
tion of reduced nicotinamide cofactors. Also in situ regeneration
NAD(P)-dependent oxidoreductases are valuable tools for the
synthesis of chiral compounds. Especially, the enantiospecific
reduction of prochiral ketones has attracted considerable aca-
demic [1,2] and industrial [3,4] interest. Due to the high cost of
the nicotinamide cofactors (NAD(P)), in situ cofactor regenera-
tion is required for preparative applications. Besides a wealth of
enzymatic cofactor regeneration approaches [5–10], a range of
non-enzymatic approaches have been proposed [11,12], with the
organometallic pentamethylcyclopentadienyl rhodium bipyridine
([Cp*Rh(bpy)(H2O)]2+) being by far the most successful non-
enzymatic regeneration catalyst. First reported in 1987 [13,14], it
has attracted great interest mainly for its versatility (Scheme 1):
(1) in contrast to most chemical reductants known, its catalyti-
cally active hydrido form ([Cp*Rh(bpy)H]+) specifically transfers
a hydride ion to the 4-position of NAD(P)+ (due to coordination
to the amide-carbonyl-O-atom) thereby exclusively forming the
enzymatically active, reduced 1,4-NAD(P)H [15,16]. (2) As a low
molecular weight compound, it does not – unlike most enzymatic
regeneration systems – distinguish between the phosphorylated
and non-phosphorylated nicotinamide cofactor [17,18]. (3) In situ
regeneration of its catalytically active form is feasible using chem-
ical [18], electrochemical [19], or even photochemical reduction
of oxidized nicotinamide cofactors [18,23], reductive regeneration
of flavins [18,24–27], and heme-groups and analogs [26,28,29] has
been reported. (5) Furthermore, [Cp*Rh(bpy)(H2O)]2+ shows excel-
lent activity and stability over a very broad temperature- and pH
range thereby outperforming any enzymatic regeneration system
reported [18]. Also, Steckhan and coworkers demonstrated that lig-
and engineering offers the opportunity of tailoring the catalytic
activity of [Cp*Rh(bpy)(H2O)]2+ and its derivates [19,30].
Thus, it does not astonish that
a
broad range of
[Cp*Rh(bpy)(H2O)]2+-catalyzed chemoenzymatic and elec-
troenzymatic reductions [23,31–38], oxidations [23] and
oxyfunctionalization [24,25,39,40] reactions have been reported.
Compared to enzymatic cofactor regeneration systems such as
alcohol dehydrogenases (ADHs), formate- and phosphite dehy-
drogenase, and NADH oxidases, [Cp*Rh(bpy)(H2O)]2+ exhibits a
somewhat lower catalytic activity in terms of kcat (0.5–10 min−1 vs.
ca. 100 min−1), which however is compensated by the significantly
lower molecular weight resulting in comparable specific activities
in terms of U mg−1. Furthermore, its catalytic promiscuity, stabil-
ity, and robustness to varying reaction conditions is unparalleled
by any enzymatic cofactor system known so far.
There is, however, one crucial aspect of [Cp*Rh(bpy)(H2O)]2+
that severely limits its general applicability for in situ regeneration
of oxidoreductases, which is the mutually inactivating interaction
between [Cp*Rh(bpy)(H2O)]2+ and proteins. While Steckhan and
coworkers explicitly ruled out such an interaction at least for the
alcohol dehydrogenases from horse liver and Thermoanaerobium
∗
Corresponding author. Tel.: +31 015 2781957; fax: +31 015 2781415.
E-mail address: f.hollmann@tudelft.nl (F. Hollmann).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.01.006

150
M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
Scheme 1. [Cp*Rh(bpy)(H2O)]2+ as a versatile catalyst to support chemoenzymatic redox reactions: the catalytically active hydrido species [Cp*Rh(bpy)H]+ (which can be
generated either electrochemically, chemically, or photochemically) can regenerate reduced nicotinamide and flavin cofactors allowing enzymatic redox reactions depending
on these cofactors. Also, indications exist that direct regeneration of reduced heme-groups is feasible.
brokii [32], we [41] and others [33,40] have observed this inter-
action with a range of oxidoreductases. A coordinative interaction
between nucleophilic protein residues and the free coordination
site at [Cp*Rh(bpy)(H2O)]2+ (obtained by water displacement)
appears to be a plausible mechanism. Tight coordination of pro-
tein residues to the ‘vacant’ Rh-binding site on the one hand would
explain the inactivity of [Cp*Rh(bpy)(H2O)]2+ as regeneration cat-
alyst under formate-driven conditions. On the other hand, this
coordination may be expected to impair enzyme activity if it occurs
with catalytically relevant amino acids, blocks the access to the
active site, or interferes with the native tertiary and quaternary
structure of the enzyme(s).
However, a systematic study investigating this interaction is
missing so far. Thus, the controversy about the inactivation has
remained rather descriptive and too much relying on assumptions
rather than pursuing an in-depth understanding of the underlying
inactivation mechanisms.
2. Materials and methods
Unless indicated otherwise, all chemicals were purchased in
analytical grade from Sigma–Aldrich and used without further
purification. Amino acids were purchased as pure compounds
(inner salts).
[Cp*Rh(bpy)(H2O)]2+ and TADH were synthesized according to
published procedures [18,42]. A detailed description can be taken
from the supporting information.
Activity assays for both [Cp*Rh(bpy)(H2O)]2+ and TADH
were performed spectrophotometrically based on the forma-
tion or depletion of the reduced nicotinamide cofactor (NADH,
ꢀmax = 340 nm, ε = 6.22 mM−1 cm−1). A detailed description of the
experimental details (pH, T, etc.) can be taken from the supporting
information.
For cyclic voltammetry, the electrochemical cell and conditions
reported by Hagen was used [43].
The aim of the present study therefore was to characterize
the inactivation mechanism. As a model enzyme, we chose the
recently described alcohol dehydrogenase from Thermus sp. ATN1
[42]. Based on these results we propose a range of potential solu-
tions.
3. Results
3.1. Interaction of [Cp*Rh(bpy)(H2O)]2+ with different enzymes
From previous investigations, a rather diverse picture about the
interaction of [Cp*Rh(bpy)(H2O)]2+ with proteins evolves. To clarify
whether this interaction is more likely the rule or rather the excep-
tion, we incubated [Cp*Rh(bpy)(H2O)]2+ with different enzymes
from various enzyme classes and organisms and determined
the residual [Cp*Rh(bpy)(H2O)]2+ activity in the formate-driven
reduction of NAD+ (Table 1). This reaction represents a simple
and sensitive probe for coordination strength of ligands to the
Rh-central atom as strongly coordinating ligands impede the coor-
dinative interaction of formate and the Rh-central atom [15,16].
With all enzyme preparations tested, a significant reduc-
tion of [Cp*Rh(bpy)(H2O)]2+ activity was observed, which we
attribute to the coordinative interaction between the enzymes
Table 1
Residual [Cp*Rh(bpy)(H2O)]2+-activity (0.02 mM) after 1 h incubation with various
enzymes (375 ␮g ml−1).
Enzyme (source)
Residual activity [%]
None
100
10.8
17.6
0
Albumine (bovine serum)
Acylase I (Aspergillus mellus)
Aldehyde dehydrogenase (Saccharomyces
cerevisiae)
Esterase (Thermoanaerobium brokii)
Enoate reductase (Bacillus subtilis, cell crude
extract from Escherichia coli)
Lipase (Candida rugosa, Type VII)
Phytase (Aspergillus ficuum)
Protease (Aspergillus oryzae, Type XXIII)
0
15.3
and [Cp*Rh(bpy)(H2O)]2+
. It should, however, be mentioned
here that not in every case, detailed information about the
enzyme formulation was available. Thus, in some cases decreased
[Cp*Rh(bpy)(H2O)]2+-activity may also have been caused by inac-
tivating buffer components (vide infra).
13.1
19.4
15.3

M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
151
Fig. 1. Residual catalytic activity of [Cp*Rh(bpy)(H2O)]2+ after pre-incubation with amino acids (1 equiv.).
Given the heterogeneous nature of enzyme classes and protein
folds tested, we suspect that the interaction of [Cp*Rh(bpy)(H2O)]2+
with proteins is rather a general phenomenon than an exception.
In the absence of either amino acid KM (formate) was 93 mM
whereas this value dropped to 5 mM and 37 mM in the presence
of 1 equiv. of cysteine and histidine, respectively. In both cases
the maximum activity of [Cp*Rh(bpy)(H2O)]2+ (vmax, under the
given conditions 35 h−1) did not change. The axis interceptions
for the single Lineweaver–Burk plots were practically identical
suggesting a ‘competitive activation’ of [Cp*Rh(bpy)(H2O)]2+. The
Michaelis–Menten behavior in the presence of non-coordinating
amino acids such as alanine was super imposable to the curve
without any additional ligands (data not shown).
3.2. Interaction of [Cp*Rh(bpy)(H2O)]2+ with amino acids
Next, we investigated the influence of various natural amino
acids on the efficiency of [Cp*Rh(bpy)(H2O)]2+ for the formate-
driven reduction of NAD+. As shown in Fig. 1, a significant reduction
of activity (residual activity in parentheses) was observed only
in the presence of histidine (8.2%), methionine (71.2%), trypto-
phane (45.9%), and cysteine (0%). Thus, we conclude that under
physiological pH values the protolysis grade of ␣-amino function-
alities (pKa = 9.2–10.6) and of potentially coordinating OH-groups
(serine and threonine) as well as amines (lysine, pKa = 10.5) and
amides/guanidines (asparagine, glutamine, arginine, pKa ca. 12.5)
is not high enough to quantitatively coordinate to the Rh-central
atom. Also steric effects and/or differences in nucleophilicity may
play a role. The strong inactivating effect of histidine becomes clear
considering the relative high ratio (ca. 90%) of non-protonated imi-
dazolium over protonated at pH 7 (pKa = 6). The somewhat lower
effect of tryptophane might have been due to the lower basic-
ity on the one hand and greater steric constraints on the other
hand. Interestingly, cysteine which should be near-completely pro-
tonated at pH 7 showed the most significant interaction with
[Cp*Rh(bpy)(H2O)]2+, probably due to the high thiophilicity of the
soft heavy metal.
The
stoichiometry
of
the
interaction
between
[Cp*Rh(bpy)(H2O)]2+ and the inhibiting amino acids was examined
as well (Fig. 3). Thus, for cysteine, a clear 1:1 ratio was observed
while for histidine and tryptophane full inhibition was achieved
at 1:1.5 and 1:>2, respectively, reflecting the relative degree of
[Cp*Rh(bpy)(H2O)]2+-inhibition shown in Fig. 1. We attribute
the apparent deviation from equimolar stoichiometry in case of
histidine and tryptophane to lower coordination strength of these
ligands rather than to deviations from the assumed 1:1 stoichiom-
etry. Further studies however will be necessary to substantiate
this assumption.
3.3. Interaction of [Cp*Rh(bpy)(H2O)]2+ with TADH
The interaction of [Cp*Rh(bpy)(H2O)]2+ with isolated amino
acids is only a very restricted model for the mutual inactivation
The kinetics of this interaction was fast as exemplified with
cysteine. [Cp*Rh(bpy)(H2O)]2+ lost approx. 80% of its initial
NAD+-reduction activity after only 3 min and full inhibition was
observed within less than 1 h after addition of cysteine (supporting
information). These kinetics were corroborated by the spectral
changes, which [Cp*Rh(bpy)(H2O)]2+ underwent when incubated
with cysteine (supporting information). Characteristic absorp-
tion changes at 270 and 340 nm were observed and correlated
well with the time-course of decreasing catalytic activity of
[Cp*Rh(bpy)(H2O)]2+ (supportinginformation). Similar rapidkinet-
ics was also observed for histidine and tryptophane (data not
shown).
Interestingly, if the contact time between [Cp*Rh(bpy)(H2O)]2+
and cysteine or histidine was very short (i.e. when addition of these
amino acids to the assay mixture occurred immediately before
the addition of formate) a significant reduction of the apparent
Michaelis–Menten constant (KM) for formate was observed (Fig. 2).
Fig. 2. Michaelis–Menten behavior of formate in the formate-driven reduction of
NAD+ in the presence (ꢀ) and absence (ꢀ) of cysteine. Inset: Lineweaver–Burk plot.

152
M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
Fig. 4. Residual [Cp*Rh(bpy)(H2O)]2+-activity after incubation with TADH at differ-
ent molar ratios and incubation times: 1 h (black), 4 h (grey), 24 h (white).
Fig. 3. Stoichiometry of the cysteine- (♦), histidine- (ꢀ), and tryptophane (ꢀ)-
related inhibition of [Cp*Rh(bpy)(H2O)]2+
.
tion temperature slightly influenced the degree of inhibition. For
example performing the incubation experiments at 40 ◦C instead
of ambient temperature resulted in an increased inactivation by
approximately 10% (data not shown).
observed between [Cp*Rh(bpy)(H2O)]2+ and proteins. For example,
pKa values can be significantly modulated by the local electrostatic
potential within an enzyme. We therefore further investigated
the interaction between [Cp*Rh(bpy)(H2O)]2+ and the recently
described alcohol dehydrogenase from Thermus sp. ATN1 (TADH)
[42]. This enzyme appeared a suitable model system for vari-
ous reasons. On the one hand fermentative production at scale is
straightforward in Escherichia coli as host system expressing TADH
encoded on a pET vector [42] and using the autoinduction system
reported by Studier [44]. Furthermore, one-step purification via
heat-treatment of harvested E. coli yielded sufficiently pure protein
(>95% as judged by SDS gel analysis) to exclude perturbing inter-
actions with host proteins and other components such as protein
stabilizers.
The effect of varying [Cp*Rh(bpy)(H2O)]2+ concentrations on the
activity of the biocatalyst is shown in Fig. 5.
To our surprise, no decrease in TADH activity was observed until
a [Cp*Rh(bpy)(H2O)]2+ to TADH ratio of up to 4. Furthermore, the
thermal stability of TADH (in terms of half life time at 60 ◦C) was not
impaired by the presence of 4 equiv. of [Cp*Rh(bpy)(H2O)]2+ (data
not shown). Thus, we assume that these four, probably surface-
exposed, metal binding sites are not crucial for TADH activity or
-stability. Increasing the relative amount of [Cp*Rh(bpy)(H2O)]2+ to
TADH resulted in decreased residual enzyme activity. Interestingly,
the temperature-dependency of the inactivation was not uniform.
Raising the incubation temperature from 25 ◦C to 40 ◦C did not
significantly influence the degree of TADH inactivation. At 60 ◦C,
however, TADH inactivation after 1 h was significantly higher, at
least for [Cp*Rh(bpy)(H2O)]2+ to TADH ratios higher than 10 (Fig. 5).
This inactivation is accompanied by significant changes in the cir-
cular dichroism spectrum of TADH at 270 and 300 nm indicating
loss of secondary structural elements (supporting information). It
should be mentioned here that the experiments shown in Fig. 5
were performed at pH 6 representing the optimal pH range for
TADH-catalyzed reduction reactions. Therefore, direct compari-
In a first set of experiments we determined the residual activity
of [Cp*Rh(bpy)(H2O)]2+ in the presence of different molar ratios
with TADH and different incubation times (Fig. 4).
The degree of [Cp*Rh(bpy)(H2O)]2+-inhibition correlated both
with the molar ratio to TADH, and incubation time. Up to 4 equiv. of
[Cp*Rh(bpy)(H2O)]2+ were quickly (within minutes) and quantita-
tively inactivated suggesting the presence of four easily accessible
‘[Cp*Rh(bpy)(H2O)]2+-binding sites’ per subunit interacting with
the metal center. Upon prolonged incubation at least 20 equiv.
of [Cp*Rh(bpy)(H2O)]2+ were inactivated by TADH. The incuba-
Fig. 5. Residual TADH activity after 1 h incubation with [Cp*Rh(bpy)(H2O)]2+ at 25 ◦C (black), 40 ◦C (grey), 60 ◦C (white).

M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
153
Fig. 6. Effect of various coordinating components on the residual [Cp*Rh(bpy)(H2O)]2+ (black) and TADH activity (grey).
son with the previous experiments (Fig. 4) may not be possible
3.5. Cyclic voltammetry
due to differences in proteolysis degree of crucial amino acids
involved. However, we observed a significant stabilization of TADH
towards [Cp*Rh(bpy)(H2O)]2+ when performing the incubation
experiments at pH 8. For example, TADH retained more than 80%
of its initial activity at pH 8 while at pH 6 only 25% were recov-
ered (in both cases 1 h incubation at ambient temperature, 100-fold
molar excess of [Cp*Rh(bpy)(H2O)]2+). We hypothesize that the
deprotonated hydroxyl complex ([Cp*Rh(bpy)(OH)]+, pKa = 8.5)
containing the poor leaving group OH− interacts to a much lesser
extent with nucleophilic protein residues than the aquo complex
([Cp*Rh(bpy)(H2O)]2+). This led us to further investigate the influ-
ence of some potentially strongly coordinating additives on the
On the one hand strongly coordinating buffers stabilize TADH
towards [Cp*Rh(bpy)L]n+ (with L being the ligand) but, on the
other hand inactivate the metal complex towards formate-
driven regeneration of NADH. However, the catalytic hydrido
species ([Cp*Rh(bpy)H]+) can also be formed electrochemically.
According to Steckhan et al. [19] this should also be possi-
ble in the presence of strongly coordinating buffers (supporting
information). Therefore, we investigated the electrocatalytic activ-
ity of [Cp*Rh(bpy)(H2O)]2+ equilibrated in (NH4)2SO4-containing
buffer (Fig. 7).
The increase of cathodic peak current in the presence of NAD+
(catalytic effect) indicates that indeed the electrochemical forma-
tion of [Cp*Rh(bpy)H]+ and its reaction with NAD+ is not inhibited
by the presence of NH3. Similarly, catalytic effects were observed
also in the presence of Cl− and TRIS (supporting information).
Overall, we conclude, that strongly coordinating components
can prevent [Cp*Rh(bpy)(H2O)]2+-related enzyme inactivation
while preserving its activity for electrochemical NADH regenera-
tion.
stability of TADH in the presence of [Cp*Rh(bpy)(H2O)]2+
.
3.4. On the stability of TADH in the presence of
[Cp*Rh(bpy)(H2O)]2+ and various coordinating buffers
It is well known that various N- or S-containing compounds can
strongly coordinate to Rh. We hypothesized that such coordina-
tive saturation of the Rh-central atom might stabilize TADH (and
other proteins in general) towards [Cp*Rh(bpy)(H2O)]2+. Therefore,
we screened a range of putative good coordinating compounds
such as nitrogen-containing buffers and halogenides (supporting
information) on their effect on TADH-stability in the presence
of [Cp*Rh(bpy)(H2O)]2+ (Fig. 6).
their inhibiting effect on formate-driven, [Cp*Rh(bpy)(H2O)]2+
catalyzed NAD+ reduction and residual TADH activity was observed.
To probe the reversibility/irreversibility of
A clear correlation between
-
[Cp*Rh(bpy)(H2O)]2+-related inhibition of TADH we performed a
series of competition experiments. Thus, TADH-samples treated
with [Cp*Rh(bpy)(H2O)]2+ were supplemented with 100 mM
(NH4)2SO4 after 2 and 24 h, respectively. After 2 h, almost full
initial TADH activity could be restored upon (NH4)2SO4 addi-
tion. If however, this treatment was initiated after 24 h TADH
activity could not be restored anymore. Furthermore, after 24 h
incubation in the presence of [Cp*Rh(bpy)(H2O)]2+ a yellow,
protein-containing precipitate was observed. Obviously, at rela-
tively short incubation times TADH is reversibly inhibited while
after prolonged incubation times it is irreversibly inactivated.
Fig. 7. Cyclic voltammogram of [Cp*Rh(bpy)(H2O)]2+ (overnight equilibrated in
100 mM (NH4)2SO4), without NAD+ (light grey), in the presence of 1 equiv. NAD+
(dark grey), and in the presence of 2 equiv. NAD+ (black).

154
M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
Scheme 2. Proposed coordination/deprotonation mechanism explaining the activating effect of cysteine and histidine at short incubation times and the inhibiting effect at
prolonged incubation times.
4. Discussion
the central atom. Compared to the water ligand, these protonated
amino acid species are better leaving groups explaining that vmax
is reached at lower formate concentrations. It should be empha-
sized here that this activating effect can only be observed at short
contact times between [Cp*Rh(bpy)(H2O)]2+ and the amino acid(s).
At prolonged incubation times deprotonation of His and Cys occurs
resulting in thermodynamically and kinetically stable adducts with
which substitution of the ligand by formate or protein residues
does not occur. In other words, the KM value for other ligands has
increased dramatically.
[Cp*Rh(bpy)(H2O)]2+ is a useful catalyst for the in situ regen-
eration of various oxidoreductase cofactors and prosthetic groups
(Scheme 1) which is well-documented by its various applica-
tions. Less well-documented however is that it has sometimes a
detrimental effect on enzyme stability let alone the underlying
inactivation mechanism. Aim of this study was to shed some light
on this inactivation effect.
4.1. The interaction between [Cp*Rh(bpy)(H2O)]2+ and amino
acids
4.2. The interaction between [Cp*Rh(bpy)(H2O)]2+ and TADH
Inspired by the study by Hildebrand and Lütz [33] we validated
their observation that among the isolated amino acids primarily
cysteine, histidine, and tryptophane (and methionine albeit to a
lower extend) exceed an inhibiting effect on [Cp*Rh(bpy)(H2O)]2+
suggesting these amino acids to be the coordinating residues within
a protein (Fig. 1). Especially cysteine and histidine had a very
interesting effect on [Cp*Rh(bpy)(H2O)]2+ activity. After prolonged
incubation, these ‘ligands’ severely impaired the formate-driven
reduction of NAD+ which we attribute to the formation of a tight
(thermodynamically and kinetically stable) bond to the Rh-central
atom thereby preventing formate coordination. But at very short
incubation times, these amino acids exhibited an activating effect
in the sense that the affinity of formate to Rh (expressed as KM
value) was drastically increased (Fig. 2). These observations may be
rationalized by the hypothetical coordination equilibrium outlined
in Scheme 2.
The protonated amino acids represent the primary products in
the coordination reaction to the Rh-central atom. After deprotona-
tion, tight Rh–S and Rh–N bonds are formed showing no exchange
tendency with, e.g. formate (or protein residues) and thereby yield-
ing catalytically inactive complexes; i.e. formate cannot substitute
the tightly binding ligands. If however the time-scale for deproto-
nation is too short (less than 1–2 min), formate competes with the
less strongly binding, protonated amino acids for coordination to
The results presented here reveal that a simple stoichiome-
try for the interaction between TADH and [Cp*Rh(bpy)(H2O)]2+
cannot be made. At least 20 [Cp*Rh(bpy)(H2O)]2+-binding events
were observed in TADH/[Cp*Rh(bpy)(H2O)]2+ titration experi-
ments (Fig. 4) but more may be assumed. TADH contains in total
9 histidines and 5 cysteines (supporting information), which do
not explain the observed ‘stoichiometry’ of at least 20. Quantita-
tive analysis is complicated by the only partial inhibition by amino
acids such as methionine, tryptophane (Fig. 1). Therefore, we made
a rough estimation taking into account all amino acids present in
TADH. Thecontributionofeach aminoacidto[Cp*Rh(bpy)(H2O)]2+
-
inhibition was weighted by its influence on [Cp*Rh(bpy)(H2O)]2+ as
isolated amino acid (Fig. 1). Thus, an overall numerical stoichiom-
etry of 1:24.2 (TADH/[Cp*Rh(bpy)(H2O)]2+) was estimated which
is in good accordance with the results obtained in Figs. 4 and 5. Of
course this suggests an interaction of [Cp*Rh(bpy)(H2O)]2+ with
every (or most) amino acid residue(s) within TADH, which can
only occur if the protein is fully unfolded. Two to four of these
binding sites seem to be surface-exposed as judged by the fast
inhibition of up to 4 equiv. of [Cp*Rh(bpy)(H2O)]2+ within 1–2 h.
Interestingly, these interactions did not impair TADH activity.
Unfortunately, at present no crystal structure or homology model of
TADH is available to further rationalize this stoichiometry. Further
[Cp*Rh(bpy)(H2O)]2+-binding occurred at a lower rate and showed
Scheme 3. Proposed mechanism for the [Cp*Rh(bpy)(H2O)]2+-catalyzed inactivation of TADH. Blue: hydrophilic amino acid residues, yellow: hydrophobic amino acid
residues, red: [Cp*Rh(bpy)(H2O)]2+. (For interpretation of the references to colour in this scheme, the reader is referred to the web version of the article.)

M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
155
a detrimental effect on TADH activity. It is interesting to note that
raising the incubation temperature from ambient temperature to
40 ◦C had only a minor effect on TADH inhibition. If however, the
incubation temperature was raised to 60 ◦C, TADH inhibition was
much more pronounced (Fig. 5). 60 ◦C is the optimal growth tem-
perature of Thermus sp. ATN1 [42], the original host organism.
Thus, it may be expected that TADH exhibits high conformational
flexibility at this temperature to function optimally in its native
environment while it is more rigid at lower temperatures. There-
fore we hypothesize a ‘[Cp*Rh(bpy)(H2O)]2+-catalyzed’ unfolding
of TADH (Scheme 3).
This mechanism is based on the generally accepted mechanism
for thermal inactivation of proteins comprising: (1) the catalyti-
cally active enzyme is in a reversible equilibrium with an inactive,
partially unfolded structure. Thus, hydrophobic amino acids, nor-
mally located in the protein core are temporarily surface-exposed.
(2) These hydrophobic residues can agglomerate to reduce their
thermodynamically unfavorable solvent exposure yielding the irre-
versibly denaturated enzyme [45]. According to this mechanism,
we propose that the binding of two to four [Cp*Rh(bpy)(H2O)]2+
influenced neither the catalytic activity of TADH nor the equi-
librium between active and partially unfolded protein (neither
TADH’s catalytic activity nor its thermal stability was impaired
in the presence of up to 4 equiv. of [Cp*Rh(bpy)(H2O)]2+). How-
ever, further binding of [Cp*Rh(bpy)(H2O)]2+ to the partially
unfolded enzyme stabilized this form, possibly by blocking essen-
tial inter-protein interaction and/or by electrostatic repulsion.
Thus, [Cp*Rh(bpy)(H2O)]2+ binding shifts the equilibrium between
active and partially unfolded protein and thereby accelerates
TADH denaturation. This assumption is further supported by the
time-dependent switch from reversible inhibition to irreversible
inactivation of TADH ((NH4)2SO4 treatment could restore TADH
activity after 2 h but not after 24 h).
here might eventually lead to robust and efficient electroenzymatic
reduction reactions.
Further studies will be necessary to establish a general mech-
anism explaining the role of [Cp*Rh(bpy)(H2O)]2+ in enzyme
inactivation. However, the data presented here give a strong indi-
cation that its major role lies in the stabilization of unfolded,
catalytically inactive states. If so, stabilization of these structural
elements, e.g. via multiple-point crosslinking and/or immobiliza-
tion would represent another simple solution to the inactivation
problem. Then, coordinative saturation of [Cp*Rh(bpy)(H2O)]2+
would not be necessary to stabilize the enzyme (as it would be
geometrically stabilized) and en masse inactivation of the metal
complex would not occur as the majority of binding amino acids
would remain inaccessible in the protein core. Again, the contra-
dictory results reported [33,41] on this require further investigation
to clarify this issue.
5. Conclusions
In the present contribution we have, for the first time, performed
an in-depth analysis of the mutually inactivating interaction
between [Cp*Rh(bpy)(H2O)]2+ and enzymes. Based on the stoi-
chiometric and kinetic analysis, we hypothesize an inactivation
mechanism and propose several remedies.
These studies will put the basis for various future developments
using [Cp*Rh(bpy)(H2O)]2+
.
Acknowledgements
Dr. Peter-Leon Hagedoorn (Department of Biotechnology, Delft
University of Technology) is acknowledged for assistance with the
cyclovoltammetric measurements. We also thank Dr. Katja Bühler
and Prof. Dr. Andreas Schmid (Dortmund University of Technology)
for kind provision with pASZ encoding TADH.
This mechanism would also explain the stabilization observed
for 2-hydroxy biphenyl 3-monooxygenase (HbpA) immobilized on
Eupergit [45]. Here, multiple-point attachment of the enzyme to
the carrier might have stabilized the catalytically active tertiary and
M. Poizat acknowledges financial support by the Conseil
régional d’Ile de France and the Ecole Normale Supérieure.
quaternary structure of the enzyme against [Cp*Rh(bpy)(H2O)]2+
.
Appendix A. Supplementary data
4.3. Potential remedies
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcatb.2010.01.006.
Based on the observations reported in this contribution we
discuss a range of potential solutions to the mutual inactivation
of [Cp*Rh(bpy)(H2O)]2+ with enzymes. One obvious solution was
reported by Lütz and coworker [33]. Physical separation of the
enzyme(s) and [Cp*Rh(bpy)(H2O)]2+ was shown to greatly stabilize
both. However, diffusion limitation severely limited the catalytic
performance of the single components in terms of turnover fre-
quency and total turnover number of the catalysts.
Enzyme engineering to generate inert variants does not seem
viable considering the multiplicity of [Cp*Rh(bpy)(H2O)]2+ bind-
ing sites within TADH. Exchanging all of these residues for sure is
tedious and most probably results in catalytically inactive/impaired
TADH variants. Furthermore, this would represent just a case-
specific solution lacking general applicability.
Therefore, we propose to further explore strongly coordinating
ligands to protect the enzyme(s) from [Cp*Rh(bpy)(H2O)]2+-related
inactivation. For example ammonium containing buffers proved to
be very efficient to stabilize TADH (Fig. 6) while electrocatalytic
NAD(P)H regeneration activity of [Cp*Rh(bpy)(NH3)]2+ (at present
we can only speculate on the exact nature of the ammonium adduct,
possibly NH3 is deprotonated in the Rh-ligand sphere yielding
[Cp*Rh(bpy)(NH2)]+) was not impaired (Fig. 7). The contradictory
indications from the literature reporting stabilizing effects [31,41]
and no effect [33] necessitate further in-depth studies to clarify
the possibility of a stabilizing effect. However, the results reported
References
[1] K. Drauz, H. Waldmann, Enzyme Catalysis in Organic Synthesis, Wiley-VCH,
Weinheim, 2002.
[2] K. Faber, Biotransformations in Organic Chemistry, Springer, Berlin, 2004.
[3] A. Liese, K. Seelbach, C. Wandrey, Industrial Biotransformations, Wiley-VCH,
Weinheim, 2006.
[4] A. Schmid, F. Hollmann, J.B. Park, B. Bühler, Curr. Opin. Biotechnol. 13 (2002)
359–366.
[5] H. Chenault, G. Whitesides, Appl. Biochem. Biotechnol. 14 (1987) 147–197.
[6] H. Zhao, W.A. van der Donk, Curr. Opin. Biotechnol. 14 (2003) 583–589.
[7] W.A. van der Donk, H. Zhao, Curr. Opin. Biotechnol. 14 (2003) 421–426.
[8] R. Wichmann, D. Vasic-Racki, Technology Transfer in Biotechnology: from Lab
to Industry to Production, vol. 92, Springer-Verlag, Berlin, 2005, pp. 225–260.
[9] W. Liu, P. Wang, Biotechnol. Adv. 25 (2007) 369–384.
[10] Y.H. Kim, Y.J. Yoo, Enzyme Microb. Technol. 44 (2009) 129–134.
[11] F. Hollmann, A. Schmid, Biocatal. Biotransf. 22 (2004) 63–88.
[12] F. Hollmann, K. Hofstetter, A. Schmid, Trends Biotechnol. 24 (2006) 163–171.
[13] U. Kölle, M. Grätzel, Angew. Chem. Int. Ed. 26 (1987) 567–570.
[14] R. Ruppert, S. Herrmann, E. Steckhan, Tetraherdon Lett. 28 (1987) 6583–6586.
[15] H.C. Lo, O. Buriez, J.B. Kerr, R.H. Fish, Angew. Chem. Int. Ed. 38 (1999)
1429–1432.
[16] H.C. Lo, C. Leiva, O. Buriez, J.B. Kerr, M.M. Olmstead, R.H. Fish, Inorg. Chem. 40
(2001) 6705–6716.
[17] R. Ruppert, S. Herrmann, E. Steckhan, J. Chem. Soc. -Chem. Commun. (1988)
1150–1151.
[18] F. Hollmann, B. Witholt, A. Schmid, J. Mol. Catal. B: Enzym. 19–20 (2002)
167–176.
[19] E. Steckhan, S. Herrmann, R. Ruppert, E. Dietz, M. Frede, E. Spika,
Organometallics 10 (1991) 1568–1577.

156
M. Poizat et al. / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 149–156
[20] Z.Y. Jiang, C.Q. Lu, H. Wu, Ind. Eng. Chem. Res. 44 (2005) 4165–4170.
[21] S.H. Lee, D.H. Nam, C.B. Park, Adv. Synth. Catal. 351 (2009) 2589–2594.
[22] S.H. Lee, D.H. Nam, J.H. Kim, J.-O. Baeg, C.B. Park, ChemBioChem 10 (2009)
1621–1624.
[33] F. Hildebrand, S. Lütz, Chem. Eur. J. 15 (2009) 4998–5001.
[34] F. Hildebrand, S. Lütz, Tetrahedron Asymm. 18 (2007) 1187–1193.
[35] K. Delecouls-Servat, A. Bergel, R. Basseguy, Bioprocess. Biosyst. Eng. 26 (2004)
205–215.
[23] F. Hollmann, A. Kleeb, K. Otto, A. Schmid, Tetrahedron Asymm. 16 (2005)
3512–3519.
[36] K. Delecouls-Servat, R. Basseguy, A. Bergel, Chem. Eng. Sci. 57 (2002)
4633–4642.
[24] S. Unversucht, F. Hollmann, A. Schmid, K.-H. van Pée, Adv. Synth. Catal. 347
(2005) 1163–1167.
[25] F. Hollmann, P.C. Lin, B. Witholt, A. Schmid, J. Am. Chem. Soc. 125 (2003)
8209–8217.
[37] J. Cantet, A. Bergel, M. Comtat, Enzyme Microb. Technol. 18 (1996) 72–79.
[38] H.C. Lo, R.H. Fish, Angew. Chem. Int. Ed. 41 (2002) 478–481.
[39] F. Hollmann, A. Schmid, E. Steckhan, Angew. Chem. Int. Ed. 40 (2001) 169–
171.
[26] J.H. vanEsch, M.A.M. Hoffmann, R.J.M. Nolte, J. Org. Chem. 60 (1995) 1599–1610.
[27] G. de Gonzalo, G. Ottolina, G. Carrea, M.W. Fraaije, Chem. Commun. (2005)
3724–3726.
[40] J.D. Ryan, R.H. Fish, D.S. Clark, ChemBioChem 9 (2008) 2579–2582.
[41] J. Lutz, F. Hollmann, T.V. Ho, A. Schnyder, R.H. Fish, A. Schmid, J. Organometal.
Chem. 689 (2004) 4783–4790.
[28] P.A. Gosling, R.J.M. Nolte, J. Mol. Catal. A: Chem. 113 (1996) 257–267.
[29] F. Hollmann, A. Schmid, J. Inorg. Biochem. 103 (2009) 313–315.
[30] F. Hildebrand, C. Kohlmann, A. Franz, S. Lütz, Adv. Synth. Catal. 350 (2008)
909–918.
[31] V. Höllrigl, K. Otto, A. Schmid, Adv. Synth. Catal. 349 (2007) 1337–1340.
[32] D. Westerhausen, S. Herrmann, W. Hummel, E. Steckhan, Angew. Chem. Int. Ed.
31 (1992) 1529–1531.
[42] V. Höllrigl, F. Hollmann, A. Kleeb, K. Bühler, A. Schmid, Appl. Microbiol. Biotech-
nol. 81 (2008) 263–273.
[43] W.R. Hagen, Eur. J. Biochem. 182 (1989) 523–530.
[44] F.W. Studier, Protein Expr. Purif. 41 (2005) 207–234.
[45] V.V. Mozhaev, Trends Biotechnol. 11 (1993) 88–95.