close

Вход

Забыли?

вход по аккаунту

?

Biosynthetic Inorganic Chemistry.

код для вставкиСкачать
Reviews
Y. Lu
DOI: 10.1002/anie.200600168
Chemical Biology
Biosynthetic Inorganic Chemistry
Yi Lu*
Keywords:
biocatalysis и bioinorganic chemistry и
biomimetic synthesis и protein
design
Angewandte
Chemie
5588
www.angewandte.org
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
Inorganic chemistry and biology can benefit greatly from each other.
Although synthetic and physical inorganic chemistry have been greatly
successful in clarifying the role of metal ions in biological systems, the
time may now be right to utilize biological systems to advance coordination chemistry. One such example is the use of small, stable, easyto-make, and well-characterized proteins as ligands to synthesize novel
inorganic compounds. This biosynthetic inorganic chemistry is
possible thanks to a number of developments in biology. This review
summarizes the progress in the synthesis of close models of complex
metalloproteins, followed by a description of recent advances in using
the approach for making novel compounds that are unprecedented in
either inorganic chemistry or biology. The focus is mainly on synthetic
?tricks? learned from biology, as well as novel structures and insights
obtained. The advantages and disadvantages of this biosynthetic
approach are discussed.
1. Introduction
Metal ions play important roles in catalyzing numerous
chemical and biological reactions, as the reactivity and
diversity can be modulated at high levels by not only different
metal ions, but also the different oxidation states of these
same metal ions.[1?6] The successful synthesis and application
of novel metal complexes or metalloenzymes can have a great
impact on all areas of chemistry and biology, whether it is
organic, medicinal, materials chemistry, or biochemistry and
cell biology. An important example of such an impact is the
development of metal catalysts for reactions such as nitrogen
fixation, methane hydroxylation, and CO oxidation or
insertion (Table 1).[7, 8] These catalysts are key components
for multibillion-dollar chemical industries that produce vital
raw chemicals, such as fertilizers.
Interestingly, many of the reactions in Table 1 can be
catalyzed by either synthetic catalysts or by biological
enzymes. A closer look at the two systems suggests that the
catalysts mediate the same reaction under different conditions: chemical systems generally operate at high temperatures and pressures in organic solvent, whereas biological
systems work at ambient temperatures and pressures in water.
Furthermore, biocatalysts often have high catalytic turnovers.
Therefore, the question is not whether one can develop
catalysts that can work under milder conditions with high
turnovers, but rather how such catalysts can be developed.
A cursory look at the examples in Table 1 may lead one to
conclude that one system may be better than the other;
however, each system has its own advantages and disadvantages (Table 2). Chemical catalysts are generally smaller,
easier to synthesize and characterize, cheaper to produce, and
more resistant to harsh conditions such as high temperatures
and pressures. These practical attributes make it difficult to
replace catalysts with enzymes in industrial-scale synthesis,
although enzymes possess a number of desirable features for
catalysts. First, they form more-rigid active sites to stabilize
the transition state and thus are more able to accelerate
catalysis at ambient temperature and pressure. Second, they
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
From the Contents
1. Introduction
5589
2. Application of Biosynthetic
Inorganic Chemistry in
Biomimetic Model Systems
5591
3. Biosynthetic Inorganic
Chemistry for the Synthesis of
New Inorganic Complexes
5595
4. Conclusions and Outlook
5598
direct reactions towards a specific
functional group of an organic substrate, thus eliminating or minimizing
the need for protecting groups, which
are often required in synthetic chemistry. Third, the catalytic
site in enzymes is protected, which results in higher catalyst
stability and higher turnover numbers. Fourth, enzymes
achieve high regio-, stereo-, and enantioselectivity. Fifth, the
complex protein framework is amenable to site-specific (as
opposed to solvent-induced) modulation of the secondary
coordination sphere of a metal-binding site. For example, the
pKa values for glutamic acid in the same protein vary between
4.0 and 8.2, depending on the local electrostatic environment
and hydrophobicity.[9] Finally, enzymes are environmentally
benign, as they are synthesized and applied under physiological conditions in aqueous solvents and with biocompatible
and biodegradable ligands.
Since each of the chemical and biological systems has its
own advantages and disadvantages, one may wonder if the
advantages of both systems can be combined. Towards this
end, a number of synthetic biomimetic model systems for
metalloproteins have been prepared from small organic
molecules and characterized.[10] Such a synthetic approach
has been very effective in the elucidation of structural and
functional properties of metalloproteins.
Despite tremendous progress in biomimetic modeling, it is
still relatively difficult to mimic some of the features of
metalloproteins, for example, site-specific modulation of the
secondary coordination sphere and the regio-, stereo-, and
enantioselectivity of a system. Models that reproduce both
the structure and the function of metalloproteins are rare.
These challenges, like perhaps all obstacles in synthetic
inorganic chemistry, can be overcome by ligand design.
Prudent choice of ligand has repeatedly resulted in better
models of metalloproteins,[10] especially when features in the
secondary coordination sphere such as hydrogen-bonding
[*] Prof. Y. Lu
Department of Chemistry
University of Illinois at Urbana-Champaign
Urbana, IL 61801 (USA)
Fax: (+ 1) 217-244-3186
E-mail: yi-lu@uiuc.edu
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5589
Reviews
Y. Lu
Table 1: Comparison of chemical and biological systems.
Reaction
Chemical System
Biological System
nitrogen fixation
N2(g) + 3 H2(g)!2 NH3(g)
(a-Fe catalyst, 400?550 8C, 100 atm)
CH4 + H2O!CO + 3 H2
(Ni catalyst, 700?900 8C, 1?25 bar)
CO + 2 H2 !CH3OH
(Cu/Zn Catalyst, 250?280 8C, 70?110 bar)
CO + H2O!CO2 + H2
(Fe/Cu catalyst, > 200 8C)
CH3OH + CO!CH3COOH
(Rh(I)I2(CO)2] , 120 8C, 30 atm)
N2 + 8 H+ + 8 e + 16 MgATP!2 NH3 + H2 + 16 MgADP + 16 Pi
(nitrogenase)
CH4 + O2 + NADH + H+!CH3OH + H2O + NAD+
(methane monooxygenase)
methane hydroxylation
CO oxidation
CO insertion
CO + H2O!CO2 + 2 H+ + 2e
(CO dehydrogenase)
CH3[M] + CO + HSCoA!CH3(CO)SCoA + H+ + [M][a]
(acetyl-CoA synthase)
[a] CH3[M] is a corrinoid?iron?sulfur protein that acts in the reaction as a methyl group donor; HS-CoA is coenzyme A.
Table 2: Advantages of chemical and biological systems.
Chemistry
Catalysts/Materials
Biology
Enzymes/Biomaterials
easy to synthesize
easy to characterize
more robust
mild conditions
rigidity/protection
site-specific modulation of the
2nd coordination sphere
regio- and enantioselectivity
environmentally friendly
approach: study native proteins and their variants
cheaper to produce
approach: create new structures and
functions by using a minimalist
approach
variants, which rendered the use of such proteins as ligands
impractical. However, advancements in biology, such as the
development of cloning, protein expression, site-directed
mutagenesis, and polymerase chain reaction, have made it
possible to reduce dramatically the time required to synthesize proteins and change amino acid residues. Whereas it used
to take many months or even many years to construct,
express, and purify a series of proteins, it now takes a matter
of weeks or even days; such progress is similar to that made in
the computer industry (Figure 1). Furthermore, hundreds of
milligrams or even grams of pure protein can now be
were introduced.[11, 12] The main difference between synthetic
models and metalloproteins is that the synthetic models often
contain water-insoluble organic ligands, whereas metalloproteins consist of polypeptides or proteins. Polypeptides and
proteins are unusual organic molecules in that they are
naturally soluble in water and can be synthesized under mild
conditions. Through specific folding of the polypeptides and
proteins, a rigid network can form that allows site-specific
modulation of the secondary coordination sphere. This
suggests that polypeptides or proteins were established in
the course of evolution as ligands for metal-based catalysts or
enzymes.
So why then should the ligands from biological systems
not be used to synthesize model compounds? It used to be
very difficult and expensive to make proteins and their
Yi Lu received his BS degree from Beijing
University (P.R. China) in 1986 and his PhD
from the University of California at Los
Angeles in 1992 under the guidance of Prof.
Joan Selverstone Valentine. After two years
of postdoctoral research at the California
Institute of Technology in the group of Prof.
Harry B. Gray, he began his independent
career as an Assistant Professor in the
Department of Chemistry at the University
of Illinois at Urbana?Champaign, where he
is now an Alumni Research Scholar Professor
of Chemistry. His interests lie in bioinorganic
chemistry.
5590
www.angewandte.org
Figure 1. Comparison of progresses made in the computer industry
and biology. A) According to Moore?s law, the number of transistors
and thus the speed of a computer doubles approximately every
18 months. B) Estimated average speeds of cloning, expression, and
purification of recombinant proteins (calculated as the inverse of
number of days it takes to complete the process). Techniques such as
recombinant DNA (Rec. DNA), site-directed mutagenesis (SDM), and
the polymerase chain reaction (PCR) have significantly accelerated the
biological progress.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
prepared. Thanks to the progress in biotechnology, certain
chemical synthesis steps in the production of pharmaceutical
products can be replaced by biosynthesis. As a result, the
number of synthesis steps is reduced, yields are increased, and
less chemical waste is produced,[13, 14] so that the difference in
efficiency between the synthesis of small organic molecules
and proteins is decreased. At the same time, progress in
biophysics has made possible the routine characterization of
metalloproteins by different spectroscopic techniques.[15?17]
Advances in structural biology, such as the availability of
high-energy synchrotron sources and high-field magnets and
their associated methodology developments, have also made
it possible to obtain three-dimensional crystal or NMR
structures of proteins in a short period of time. Thus, it is
now possible to synthesize and characterize novel inorganic
compounds using small, stable, easy-to-produce proteins with
characteristic scaffolds as ?ligands?.
The biosynthetic approach uses the same techniques as
the biochemical studies of native enzymes and their mutant
forms. However, the following examples below show that
there are distinct differences between the two approaches:
First, like all biomimetic approaches, the biosynthetic
approach is a minimalist approach; it focuses on determining
whether the necessary structural features identified in native
enzymes are sufficient to confer the structure and function of
the enzyme. Therefore, the two approaches are complementary. Second, stable, easy-to-produce, and well-characterized
proteins are used in the biosynthetic approach, and successful
models are often as good as, if not better than, the native
enzymes for structural and functional studies. Third, biosynthetic models may offer additional advantages, since different
metal-binding centers can be compared in the same protein
framework. Finally, the biosynthetic approach allows synthesis of new compounds, with structures and reactivities
unprecedented in either biology or inorganic chemistry.
2. Application of Biosynthetic Inorganic Chemistry
in Biomimetic Model Systems
Biomimetic model systems combine the benefits of both
chemistry and biology. In this process, one can learn the
minimal structural features necessary to form metal-binding
sites with desirable function and help to elucidate the
structure?function relationship of target proteins. At the
same time, new compounds with structures that are rare in
inorganic chemistry may be synthesized, which expands our
knowledge of coordination chemistry. The resulting compounds are often smaller and better defined structurally than
the native proteins, and can therefore be used in practical
applications.
Biosynthetic model systems with small, stable, and wellcharacterized proteins as ligands can achieve the same goals.
Since they use the same ligands as target proteins and operate
under almost the same physiological conditions, they can
often result in closer models and are more readily usable in
practical applications. A few examples will illustrate these
features.
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
2.1. Inorganic Biosynthesis with the Same Protein Scaffold and
Loop-Directed Mutagenesis
The CuA center is a mixed-valence dinuclear copper
center in which the copper ions are each coordinated to a
histidine residue (His 181 and His 224) and bridged by the
thiolate sulfur atoms of two cysteine ligands (Cys 216 and
Cys 220, Figure 2 A, top).[18?21] Weak ligands such as the
Figure 2. Crystal structures of A) the CuA center in cytochrome c
oxidase from P. denitrificans; and B) a biosynthetic CuA model in
azurin. The top figures show the CuA center with Cu2S2(Cys) in a plane
and the bottom figures show the CuA center viewed perpendicular to
the plane.
thioether sulfur atom of methionine (Met227) and the
backbone carbonyl oxygen atoms of isoleucine (Ile 180),
histidine (His 224), and glutamate (Glu 218) are also present
in axial positions (Figure 2 A, bottom). The CuA center is
unique in both inorganic chemistry and biology; it acts as an
electron-transfer center in cytochrome c oxidase (CcO; a
terminal oxidase in the respiratory chain of eukaryotic
mitochondria and some aerobic bacteria)[22?26] and nitrous
oxide reductase (N2OR; an enzyme responsible for the
reduction of N2O in denitrifying bacteria).[27, 28] It was also
the first biological system shown to contain a metal?metal
bond.[29] Many coordination compounds containing metal?
metal bonds are known with late transition metals; however
few contain first-row transition metals such as copper.[30]
Whereas dinuclear or multinuclear mixed-valence copper
complexes in which the unpaired electron is completely
localized (Class I) or partially delocalized (Class II) have
been known for many years,[31, 32] copper complexes, such as
the CuA center, with fully delocalized (Class III) mixedvalence states are rare.[33?40] In both cases, the synthesis of the
CuA center and elucidation of the structural features responsible for its functions are of great interest.
A number of synthetic models with organic ligands have
been reported for this purpose, most of which contain a
dinuclear mixed-valence copper center and a copper?copper
bond, but not bridging thiolate units.[33, 34, 38, 39, 41] A model
compound that contains bridging thiolate residues but lacks a
copper?copper bond also exists.[36] A dinuclear copper center
containing a copper?copper bond and two bridging nitrogen
atoms was also reported recently.[42] Stable, easy-to-make, and
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5591
Reviews
Y. Lu
well-characterized blue copper proteins such as amicyanin[43]
and azurin[44] were used as ligands to make a closer model of
the CuA center. Blue copper proteins and proteins containing
the CuA center have been shown to share close structural
homology.[45] Sequence alignment revealed that mononuclear
blue copper proteins and dinuclear CuA proteins differ mainly
in a single loop between two b-strands. By replacing the loop
in blue copper proteins with the corresponding loop in CuA
proteins (Figure 3), a close mimic of the CuA center was
Figure 3. Development of a biosynthetic CuA model in azurin through
loop-directed mutagenesis.
synthesized in blue copper proteins amicyanin[43] and
azurin.[44] A high-resolution crystal structure showed that
the CuA center in the azurin model is almost identical to that
of the native CuA center, including structural features in the
secondary coordination sphere, such as the nonbonding
interactions between the thioether group of methionine
residue and peptide carbonyl oxygen atoms (Figure 2 B).[46]
The synthesis of a close model of a native protein is only
one measure of the effectiveness of such an approach. How
much insight is gained into the structure and function of the
metal center is another important measure. The biosynthetic
CuA model also gives deeper insight in this regard. For
example, a stopped-flow study of copper incorporation into
metal-free CuA azurin revealed a tetragonal intermediate and
showed the importance of reductants in the formation of the
final CuA center.[47] Titration of different metal ions into the
model protein suggests that the CuA center as an MII M? I
center (M, M? = any metal) is strongly favored regardless of
the sequence of addition of metal ions.[48, 49] Mutagenesis
studies revealed that the conserved cysteine,[50] histidine,[51, 52]
and methinione[53] residues all play an important role in
maintaining the structure and in fine-tuning the function.
Surprisingly, the dinuclear copper structure remains largely
intact, even when the histidine was replaced with residues
incapable of coordinating to the copper (e.g. Ala and Gly), as
evidenced by the similar characteristic purple color, as well as
UV/Vis, magnetic circular dichroism (MCD), resonance
Raman (RR), and electron nuclear double resonance
(ENDOR) spectra.[51, 52]
Until recently, there were not many examples of fully
delocalized (Class III) dinuclear mixed-valence copper compounds. Even rarer is the reversible conversion between
different classes, that is, between delocalized [Cu1.5иииCu1.5]
and trapped valence [CuIIиииCuI] states. The study of the CuA
5592
www.angewandte.org
center in azurin provided an interesting example of a
reversible transition between delocalized and trapped
mixed-valence states that is triggered solely by a change in
pH value.[54] The CuA center is known to be the electron entry
site in cytochrome c oxidase. However, its role in protoncoupled electron transfer has not been defined clearly. This
work showed that protonation of the C-terminal His residue
not only resulted in a trapped valance state of the CuA center,
but also caused a significant increase in the reduction
potential.[54] Since the corresponding C-terminal His residue
in cytochrome c oxidases is located along a major electrontransfer pathway from CuA center to heme a, and since
protonation can result in an increased reduction potential that
prevents electron transfer from the CuA center to heme a, the
results strongly indicate that the CuA center and the histidine
residue may play important roles in proton-coupled electron
transfer.[54]
These observations may be obtained from studies of
native CuA centers, in particular soluble fragments of CcO
containing the CuA center.[18?21] However, several studies have
demonstrated that biosynthetic models can provide unique
insights that are not readily obtainable from studies of either
native enzymes or synthetic model compounds. For example,
by using azurin as the ?ligand? and replacing the ligand loop
of the blue copper center with that of the CuA center, two
electron transfer (ET) centers may be placed in the same
protein framework. High-resolution crystal structures showed
that one of the copper atoms in the CuA center overlays
exactly with the blue copper center (Figure 4 A).[46] Therefore,
Figure 4. A) Overlay of the crystal structures of blue copper azurin (in
light blue) and the biosynthetic CuA model in azurin (in purple). The
copper site in the blue copper center (B) overlays almost exactly with
one of the copper sites in the purple CuA center (C).
this biosynthetic model provided a unique opportunity to
compare the two ET centers in the same protein framework
and exclude or at least minimize other factors that influence
its properties. For example, the same series of mutations at a
conserved axial methionine unit in the blue copper protein
(Met 121, Figure 4 B) and the CuA azurin (Met 123, Figure 4 C) indicate that the methionine residue has much less
influence on the reduction potential of the CuA center
(< 25 mV) than of the blue copper center (> 170 mV).[53] In
contrast, electron transfer from the same donor and at the
same distance to the CuA center is faster, despite its lower
reduction potential, than to the blue copper center.[55] Thus,
this work showed directly that the CuA center is a more
efficient ET center.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
These findings may be important in understanding the
different roles of the two copper centers. A much wider range
of reduction potentials (> 600 mV) is required for blue
copper proteins to transfer electrons to a variety of partners
in many different biological systems.[20, 21, 56, 57] In contrast, the
CuA center is part of cytochrome c oxidase, which is at the end
of the respiration chain with very small differences (< 50 mV)
in the potentials of the redox partners.[58] In this case, large
variations in the potential could hinder the flow of electrons
in the right direction. The diamond-shaped Cu2S2(Cys) core
structure of CuA provides a way to minimize changes in
reduction potential through variations of the axial ligands.
Despite the low driving force, however, the CuA center still
needs to transfer electrons to its partners at a desirable rate.
Therefore, it has to optimize its structure by lowering its
reorganization energy to be a more efficient ET center than
the blue copper center.
Loop-directed mutagenesis, in which ligand loops of a
template protein are swapped with those of a target protein
that shares similar structural homology, was employed to
make the biosynthetic models of the CuA center described
above. A preliminary survey of the protein structure data
bank (PDB) suggests that this approach is not limited to
specific cases but can be utilized generally to make biosynthetic models of many metalloproteins. For example, azurin,
amicyanin, and CuA proteins belong to a group with a
common scaffold, called the Greek Key b-barrel. It is the
second most abundant motif in the PDB, and a number of
proteins with diverse active-site structures and functions, such
as immunoglobin, beta-amylase, cytochrome c oxidase, nitrite
reductase, and superoxide dismutase, contain the same
scaffold (Figure 5).[59]
heme peroxidases contain a heme active site. However,
manganese peroxidase (MnP) contains an additional MnIIbinding site (Figure 6 A). MnII oxidation plays a critical role in
the function of MnP (Scheme 1) in the biodegradation of
lignin, the second most abundant biopolymer on earth (after
Figure 6. A) Comparison of manganese peroxidase (MnP) and cytochrome c peroxidase (CcP); B) Heme-manganese binding site in MnP
and CcP; C) Crystal structures of MnP (red) and a CcP variant that
mimics MnP (blue).
Figure 5. The Greek Key b-barrel motif appears in more than 600
proteins of different classes with diverse functions. As most of the
active sites are in the loop region, loop-directed mutagenesis can be
used to construct novel metalloproteins.
Most of the active sites in these proteins reside within or
between loops. Evidently, nature produced a thermodynamically stable scaffold upon which many different protein
structures and functions can be conferred by simply swapping
loops, which is much less disruptive to the protein structure
than changing scaffolds. This ?trick? is important for biosynthetic inorganic chemistry.
Another example of nature using the same protein
scaffold for different functions are heme peroxidases.[60] All
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Scheme 1. Catalytic cycles of CcP (top) and MnP (bottom). Trp = tryptophan; Por = heme porphyrin.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5593
Reviews
Y. Lu
cellulose), as well as being a key step in converting woody
plants to small chemicals such as those in petroleum.[61]
Interestingly, cytochrome c peroxidase (CcP) shares not
only the same overall structure, but also a similar structure
of the heme-binding site, the major differences being that CcP
lacks the MnII-binding site and contains Trp instead of Phe
around the heme-binding site (Figure 6 A,B).[62] Several
ligands to MnII in MnP are absent in CcP. To convert CcP,
an enzyme found in bakerIs yeast, into an MnP with similar
function, the residues in CcP were replaced with the residues
of MnP at the corresponding positions to give new CcP
variants with enhanced MnP function (Figure 6 C).[63?65]
Furthermore, replacing the Trp residues in CcP with the
Phe units from MnP showed that the Phe residues play
different roles in MnP function; whereas the Trp 191 Phe
mutation stabilized compound I (an intermediate), the
Trp 51 Phe mutation increased the activity significantly, as it
increased the reactivity of compound II, whose oxidation of
MnII is the rate-determining step in the reaction mechanism of
MnP.[66, 67]
2.2. Noncovalent Tuning of the Secondary Coordination Sphere
through Inorganic Biosynthesis
The noncovalent tuning of the secondary coordination
sphere of metal-binding sites is another ?trick? of biological
systems. For example, cytochrome P450 (Cyt P450) is one of
the most efficient enzymes for asymmetric oxygen-transfer
reactions, such as sulfoxidation, epoxidation, and hydroxylation, of a variety of substrates.[68] Its primary coordination
sphere consists of a heme unit with a thiolate Cys as its
proximal ligand (Figure 7 A). In view of the importance of
heme?thiolate ligation, the proximal ligand His in human
myoglobin was changed to a Cys.[69, 70] The spectrum of the
resulting protein was similar to that of Cyt P450. The protein
model also exhibited increased asymmetric sulfoxidation
activity. However, the same His-to-Cys mutation alone in
horse heart myoglobin (Mb) did not result in heme?thiolate
ligation.[71] Instead, an additional mutation of the distal
noncoordinating histidine to either a valine or isoleucine
resulted in a P450-like protein with an iron(III) resting state
(Figure 7 B).[71] This effect, termed the trans effect by the
authors, mimics the more hydrophobic environment of the
Cyt P450 (Figure 7 A) and contributes significantly to the
binding of the axial ligand and to the stability.
Similarly, mutation of the axial His ligand to Cys in CcP
resulted in a very unstable ligand that was rapidly oxidized to
cysteic acid.[72] It was then recognized that a nonpolar residue
next to the Cys unit is conserved in P450 proteins, while the
analogous amino acid in CcP is an aspartic acid (Figure 7 C).
Thus, further mutation of this Asp residue to Leu resulted in a
stable, pentacoordinate, high-spin heme with thiolate ligation.[73] The study also marked the first time a stable
cyanidoferric complex and the ferrous state of a model P450
was obtained.[73, 74]
Stable coordination of cysteine thiolate in the ferrous
carbonyl derivative of an engineered protein was also
obtained by mutations of the axial His and Met ligands in
cytochrome b562 to Cys and Gly, respectively (Figure 7 D).[75]
The key to success was the replacement of two glutamate
residues (Glu 4 and Glu 8), which are electrostatically repelled from the heme propionate groups, with Ser. These studies
demonstrated the importance of the secondary coordination
sphere around the primary coordination ligands in stabilizing
metal?ligand ligation.
2.3. Water and Hydrogen Bonding in Inorganic Biosynthesis
Figure 7. Key structural differences of Cyt P450 (A), Mb (B), CcP (C),
and Cyt b562 (D).
5594
www.angewandte.org
The third ?trick? emerging from recent investigations on a
number of metalloproteins is the precise positioning of
hydrogen-bonding networks in fine-tuning the function and
specificity of metalloproteins. For example, it has been shown
that most heme enzymes go through a similar heme?peroxo
intermediate and yet exhibit diverse enzymatic activities, such
as monooxygenase (Cyt P450), heme oxygenase (HO), and
oxidase (CcO). High-resolution crystal structures of both Cyt
P450 and heme oxygenase revealed the presence of water
molecules in the metal-binding sites (Scheme 2).[76, 77] The
number and exact location of the water molecules have been
shown to be critical in forming hydrogen-bonding networks
that can dramatically influence the reactivity. Although the
exact nature of the hydrogen-bonding networks in CcO has
not been revealed, functional studies of a biosynthetic model
in myoglobin support their importance in fine-tuning the
reactivity.[78] This work, complementary to work carried out
using synthetic models,[79, 80] is based on a structural comparison of myoglobin and heme?copper oxidase. Although both
proteins contain heme, CcO contains a copper center that is
approximately 4.7 J from the heme iron center (Figure 8 A).
To make a biosynthetic model of CcO, His ligands were
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
pathways of either reduced heme with O2 or oxidized heme
with H2O2, the biosynthetic model CuBMb is not simply an
HO model. Rather, CuBMb is at a branching point between
HO and CcO (Scheme 3); the extra proton may promote
either the CcO or the P 450 reaction, but not the HO reaction.
Therefore, such a biosynthetic model provides a unique
opportunity to investigate which properties direct the enzyme
activity.
Scheme 2. Proposed role of hydrogen bonds in HO and cyt P450.
Adapted from reference [76].
Scheme 3. Hydrogen bonds and protons influence the reaction path of
the common intermediate in heme enzymes. Adapted from reference [78].
Figure 8. A) Overlay of crystal structures of CcO (thick lines) with
sperm whale Mb (thin lines). Mb lacks the CuB-binding site since it
has one His, one Leu, and one Phe residue instead of three histidine
residues as in CcO; B) Overlay of the crystal structure of CcO (thick
lines) with a computer model of CuBMb (Leu 28 His/Phe 43 Phe, thin
lines).
introduced into the distal side of sperm whale Mb at the
corresponding locations (Figure 8 B).[81]
Spectroscopic studies indicate that the Mb model
(CuBMb) is very similar to CcO. Studies of this biosynthetic
model also indicated that the presence of a copper ion has a
critical influence on the redox potential of the heme Fe
center[82, 83] and can transform the oxygen-binding protein Mb
into an oxygen-activating enzyme.[78, 81] Interestingly, the
protein model in Mb generates verdoheme, a key intermediate in HO, rather than producing a ferryl?heme as in CcO.[78]
Control experiments indicated that protein instability or
altered dynamics of the binding site were not the cause for
this reaction, since replacement of CuII or CuI by redoxinactive ions such as ZnII and AgI, respectively, does not
promote the HO reaction.[78] Since all reactions were carried
out in the presence of catalase, the possibility of exogenous
peroxide as the cause for verdoheme formation was also
eliminated. More importantly, reaction of CuBMb with H2O2
(equivalent in oxidation state to 2e-reduced O2 but possessing two extra protons) produces a ferryl species similar to that
in CcO. Since HO results in verdoheme regardless of the
presence of metal ions in the distal site, or through reaction
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
3. Biosynthetic Inorganic Chemistry for the
Synthesis of New Inorganic Complexes
Biosynthetic inorganic chemistry has been quite effective
for making biomimetic models of complex metalloproteins
and for revealing new insights into the structural features
responsible for protein function. Its major advantage, however, is that it combines the benefits of inorganic chemistry
and biology to produce new inorganic complexes with
unprecedented structure and function.[84]
Natural metalloproteins normally consist of 20 amino
acids, less than half of which can bind as ligands to metal ions.
Furthermore, they employ only few metal ions and metalcontaining prosthetic groups (e.g., cobalamins and heme
groups). The use of nonnatural amino acids in metalloproteins
will dramatically increase the number of possible ligand
donor sets, and the introduction of inorganic compounds into
proteins will lead to new and complex active sites. The net
result will be a new, considerably expanded enzymology,
which can draw from a large reservoir of artificial metalloproteins for various reaction types and rates.
Moreover, such endeavors can also result in new inorganic
compounds with novel structure and activity. Especially
important in this respect is the adaptation of inorganic
catalysts for asymmetric reactions and for chemical transformations in water. Encapsulating these inorganic catalysts
with proteins can fulfill both purposes, as the natural chiral
environment of a protein can provide chiral discrimination.
The hydrophobic interior of proteins provides a suitable
surrounding for inorganic complexes that are active only in
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5595
Reviews
Y. Lu
organic solvent, while the hydrophilic exterior can increase
the water solubility. These requirements can all be accomplished with minimal modification of the ligands and thus
minimal effect on the catalyst reactivity.
3.1. Introducing Nonnatural Amino Acids into Metalloproteins
The properties of metalloproteins can be fine-tuned by
introducing nonnatural amino acids into the metal-binding
sites. For example, pyridine is a common ligand in synthetic
models of metal-binding sites containing imidazole histidines
in proteins.[10] The histidine ligands have been replaced by
pyridine ligands in a newly designed heme protein with four
a-helix units. The replacement of the two His ligands with two
pyridyl ligands decreased the protein-binding affinity of ferric
heme approximately 60 000-fold and increased the reduction
potential of the heme by 287 mV.[85] These differences should
be taken into consideration when using pyridines to model
histidines in metalloproteins.
Perhaps the biggest advantage of using nonnatural amino
acids in the metal-binding sites of proteins is the isostructural
replacement of ligands or residues around the secondary
coordination sphere, by which the electronic properties can
be changed through substitution of a single atom or group.
For example, both Cys and Met in blue copper protein
proteins are conserved (Figure 9). Replacing the cysteine with
Figure 9. Isostructural replacement of Cys and Met by nonnatural
amino acids.
selenocysteine in azurin resulted in marked changes in the
UV/Vis and EPR spectra (50-nm red shift of the visible
charge transfer band and twofold increase in the hyperfine
coupling constant),[86, 87] with little effect on the reduction
potential of the copper center.[86] In contrast, replacing the
methionine with selenomethionine or norleucine resulted in
little change in the UV/Vis and EPR spectra, but a dramatic
increase in the reduction potentials (25 and 140 mV, respectively, over the native protein).[88]
In addition to ligands in the primary coordination sphere,
residues in the secondary coordination sphere have also been
replaced with nonnatural amino acids. For example, replacement of the OH group at the para position of a conserved Tyr
unit near one of the Cys ligands of a rubredoxin with H, F,
NO2, and CN groups resulted in an increase in reduction
potentials of the iron?sulfur center, with electron-withdrawing groups leading to more positive potentials (Fig-
5596
www.angewandte.org
ure 10 A).[89] Furthermore, a conserved hydrogen bond
between the backbone amide groups and the cysteine ligand
of the Fe4S4 clusters was found in high-potential iron?sulfur
proteins (HiPIPs). Replacement of the backbone amide
groups with an ester linkage eliminated these hydrogen
bonds and lowered the reduction potential of the iron center
by approximately 100 mV (Figure 10 B).[90]
Figure 10. Using nonnatural amino acids to probe the role of Tyr in
the secondary coordination sphere in rubredoxin (A) and the backbone
carbonyl oxygen atom in HiPIP (B).
Isostructural replacement of ligands has been a common
practice in synthetic inorganic chemistry. Introducing nonnatural amino acids into metalloproteins finally allows
biosynthetic inorganic chemistry to achieve a similar level
of sophistication. Since proteins can provide a rigid network
for the metal-binding sites, isostructural replacement often
allows the introduction of ligands with little or no ability to
coordinate to metal ions (e.g., replacement of Met with Ile)
without affecting the overall characteristics of the metalbinding site. It also makes it easier to probe noncovalent
interactions in the secondary coordination sphere (e.g.,
through replacement of a peptide bond with an ester linkage[90]). This approach can reveal the role of specific residues
to an unprecedented level; it can be used to fine-tune
coordination complexes. For example, the use of isostructural
nonnatural amino acids at the Met position allowed deconvolution of different factors that influence the reduction
potential of the blue copper azurin through this axial
ligand.[88] A linear relationship between the reduction potential and the hydrophobicity of the side chains on the axial
ligand pointed to hydrophobicity as the dominant factor in
controlling the reduction potential[88] and could be helpful in
developing similar compounds with predicted reduction
potentials.
3.2. Introducing Nonnative Metal Cofactors into Proteins
The introduction of nonnative metal cofactors into
proteins exemplifies the combination of the benefits of
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
biology and inorganic chemistry in biosynthetic inorganic
chemistry. Three approaches have been applied towards this
goal: noncovalent, single-covalent, and dual-covalent attachment.
The majority of native metal cofactors or prosthetic
groups are incorporated into proteins through noncovalent
interactions such as hydrogen bonding, as well as electrostatic
and hydrophobic interactions. Because proteins have not
been evolved to bind nonnative metal cofactors specifically,
the introduction of these cofactors is a challenging task. The
most successful approach thus far involves the replacement of
the heme unit in heme proteins such as myoglobin with
modified metalloporphyrins.[91?99] Since the modified metalloporphyrins are similar in structure to native heme groups, it
is possible to confer novel reactivities to the protein through
the introduction of new functional groups without severely
changing the structure of the metal-binding site. For example,
modifications to the two propionate groups in heme resulted
in new functions such as protein?protein and protein?
molecule recognition as well as electron-transfer properties
and enhanced chemical reactivity (Figure 11 A),[99] which
manifests itself for example in an enhanced P450-like
dioxygen activation upon attachment of a flavin to one of
the propionate groups.[102] Furthermore, when the heme is
replaced with iron porphycene,[103] enhanced O2-binding
affinity is observed.
Many nonnative metal complexes do not resemble native
metal cofactors. Conjugates of these metal complexes with
biotin, however, can take advantage of the strong and specific
binding between biotin and the protein avidin. In this way,
biotinylated dirhodium(I) complexes have been introduced
into avidin or the closely related protein streptavidin to
produce novel catalysts for asymmetric hydrogenation with
up to 96 % enantiomeric excess (ee, Figure 11 B).[100, 104?111]
Another way to incorporate nonnative complexes is the
design of metalloproteins based on the crystal structures of
proteins and metal complexes. After careful inspection of the
structures of myoglobin and MIII(salophen) complexes (M =
Cr, Mn), modifications of both the protein and the metal
complex gave a new protein that exhibited asymmetric
sulfoxidation activity (Figure 11 C; salophen = N,N?-bis(salicylidene)-1,2-phenylenediamine dianion).[101, 112?114] Finally, a
mimetic protein with four helices has been designed de novo
that selectively binds a metalloporphyrin.[115] This example
marks the beginning of computationally developed, tailormade metalloproteins with nonnative metal complexes.
Noncovalent attachment is not the only way in which
native metal cofactors can be incorporated into proteins.
There are also covalent attachment strategies, in which
attachment is through either a single point (some protozoan
mitochondrial cytochrome c[116]) or two points (most cytochrome c[117]). These strategies have also been applied to
incorporate nonnative metal complexes.
For example, the introduction of 1,10-phenanthroline
copper[120] or iron?edta[121] into proteins by single-point
covalent attachment gave a sequence-specific nuclease
(edta = ethylenediaminetetraacetate). Covalent attachment
of a copper(II) 1,10-phenanthroline complex to a single
cysteine residue in an adipocyte lipid-binding protein gave a
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Figure 11. Introducing nonnative metal-containing cofactors into proteins through noncovalent bonds. A) Replacing heme in Mb with a
modified heme (adapted from reference [99]); B) Introducing organometallic complexes by taking advantage of strong noncovalent biotin
and avidin interactions (adapted from reference [100]); C) Replacing
heme in Mb with an Mn Schiff base complex (adapted from reference [113]).
catalyst that promotes highly enantioselective hydrolysis
(with up to 86 % ee) (Figure 12 A).[118, 122] This strategy was
also used to attach a ferrocene derivative covalently to azurin
(Figure 12 B). This novel organometalloprotein increased the
solubility of ferrocene and improved the stability of ferrocenium in water.[119] The secondary coordination sphere of the
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5597
Reviews
Y. Lu
Figure 13. Introducing nonnative metal-containing cofactors into proteins through dual-point covalent attachment. A computer model of
Mb(L72C/Y103C) with a MnIIIsalen complex covalently attached at two
points and overlaid with heme in myoglobin (adapted from reference [124]).
4. Conclusions and Outlook
Figure 12. Single-point covalent attachment of nonnative metal-containing cofactors to proteins. A) A computer model of an adipocyte
lipid-binding protein?phenanthroline complex (ALBP-Phen) (adapted
from reference [118]); B) A computer model of an azurin?ferrocene
organometalloprotein.[119]
protein has also been used to fine-tune the reduction potential
of ferrocene.[119]
Although single-point attachment allows the binding of
nonnative metal complexes to proteins with minimum
structural modifications, the conformational freedom of the
complexes inside the protein may not be restricted enough to
perform highly stereo- and enantioselective transformations.
For example, an Mn(salen) complex coupled by single-point
covalent attachment to a protein gave less than 10 % ee
(salen = N,N?-bis(salicylidene)ethylenediamine dianion).[123]
When the complex is attached in a specific location of Mb
by a dual-point covalent attachment, the enantioselectivity
increased to 51 % ee (Figure 13).[124] Clearly, such enantioselectivity is not yet useful for synthetic transformations;
however, it is understandable given the fact that myoglobin
has not evolved like other heme enzymes (e.g., Cyt P450) for
asymmetric catalysis, and that the substrate binding site has
not been modified to confer substrate selectivity. The
principle demonstrated here is nonetheless important in
guiding future research. For example, a combination of
these covalent attachment strategies[123, 124] with structurebased design or with directed evolution methods[125, 126] will
result in the next generation of asymmetric catalysts with
even higher efficiencies and selectivities.
5598
www.angewandte.org
Biosynthetic inorganic chemistry is a natural extension of
synthetic inorganic chemistry. Instead of using small organic
molecules, small, stable, easy-to-make, and well-characterized
proteins such as azurin and myoglobin are used as ?ligands?
for the synthesis of either biomimetic models or new
compounds. The chemical and biological approaches each
have advantages and disadvantages in terms of synthesis,
characterization, and properties (Table 2).
Synthesis. A synthetic approach is generally faster and
produces higher yields. However, recent developments in
molecular biology and protein enzymology have enabled the
routine production of hundreds of milligrams to grams of
protein in the laboratory and even higher yields with
industrial methods. By carrying out reactions with cells
containing the desirable protein models, the power of natural
amplification of genes and protein products can be exploited
for biosynthesis and biocatalysis. One advantage of biosynthesis over small molecule synthesis, in which modification of
the ligand might result in significant variations in yield and
have associated cost and time investments, is that a similar
modification of a protein ligand has a smaller influence on
these parameters. These features make the speed and yield of
biosynthesis closer to that of small molecule synthesis. In a
few selected cases, it is even easier to synthesize biosynthetic
models than some of the more complex synthetic models. For
example, synthetic models of heme?copper oxidases require
multistep syntheses that give relatively low yields.[79, 80] The
biosynthesis of heme?copper models in myoglobin[81] or
cytochrome c peroxidase[127] proceeds with similar yields and
preparation times as biosynthesis of other derivatives of the
same proteins. Regardless of the complexity of the modifications, approximately 100 mg L1 of pure biosynthetic models
can be obtained in less than a week. The routine production of
gram quantities of protein is thus possible. Although the
biosynthetic approach has been limited to a few cases in
industry,[13, 14] the gap between the two approaches is narrowing.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
Applications. Biosynthesis is often carried out in air at
ambient temperatures and pressures rather than in inert
atmospheres at extreme temperatures and pressures, as are
required for many syntheses of small molecules. Therefore,
biosynthesis could help to save energy and equipment costs.
Furthermore, biosynthesis is an environmentally benign
approach, since biodegradable materials are utilized and
water is the solvent. However, the products from biosyntheses
are not as robust as synthetic compounds, which will make the
long-term industrial use at high temperatures or pressures
difficult. This problem can be overcome by using protein
ligands that have been isolated from extremophilic organisms
(which live at high temperatures or in organic solvents such as
methanol).
A common criterion to evaluate the practical utility of
catalysts is the mass ratio between the product and the
catalyst. This is a good measure only if the catalysts are of
similar type and are synthesized by similar methods. In this
case, higher mass catalysts are generally more costly to
synthesize and to dispose of, and therefore smaller compounds would be preferable to a chemically synthesized
protein ligand of similar catalytic efficiency. However,
biologically synthesized protein ligands can almost always
be made at a much lower cost than their molecular weight
would suggest, and disposing a biocompatible and biodegradable protein ligand is also much less costly than disposing a
similarly sized organic ligand, which is often not biocompatible or biodegradable. In fact, the cost of synthesizing and
disposing of a small protein (e.g., MW = 10 000) by using
recombinant protein expression systems is very similar to that
of a much larger protein (e.g., MW = 100 000). Therefore,
multiple criteria are needed to evaluate different types of
catalysts,[128] for example, mass ratio, catalyst accessibility and
costs of synthesis and disposal, substrate scope, and, if
applicable, enantioselectivity.
Characterization. Most metal-based spectroscopic techniques, such as electronic spectroscopy (UV/Vis), electron
paramagnetic resonance (EPR), magnetic circular dichroism
(MCD), resonance Raman (RR), and X-ray absorption
spectroscopy (XAS), can be performed on both synthetic
and biosynthetic compounds without much difference in data
collection and interpretation. Even though recent developments in structural biology have made it possible to obtain
three-dimensional structures of biosynthetic compounds
routinely by using X-ray crystallography or NMR spectroscopy, it is still much easier to obtain structures of synthetic
compounds with higher resolution. To overcome these
limitations, one can use proteins with known 3D structures
as ligands. In this way, it is easier to grow diffraction-quality
crystals, and 3D structures can be obtained faster by comparing differences in the electron density map. The use of
synchrotron sources to obtain diffraction data may also help
to improve resolution.
Properties. Proteins provide a rigid network that helps to
stabilize metal-binding or substrate-binding sites as a result of
covalent bonds and noncovalent interactions (e.g., hydrogen
bonds, electrostatic and hydrophobic forces) in the primary
and secondary coordination spheres. They provide a better
environment for regio- and stereoselective binding and
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
catalysis, often without requiring protection of reactive
functional groups. Because they consist of chiral natural
amino acids, proteins are intrinsically enantioselective. Biosynthetic compounds with proteins can benefit from these
features. Indeed, a number of biophysical studies show that
biosynthetic models, such as a biosynthetic CuA center in
azurin (Figure 2 B),[46] are as rigid as the native template
proteins. Moreover, most synthetic modeling results in either
structural or functional models of target native proteins.
Because biosynthetic models use the same type of ligands and
are synthesized and characterized under the same conditions
as target proteins, they are often structural and functional
models at the same time. Recent advances in synthetic
chemistry have made it possible to design new organic ligands
that provide a rigid network similar to that in proteins,
including hydrogen-bonding interactions in the secondary
coordination sphere.[11, 12] Synthetic asymmetric catalysts have
a wider range of activities and a broader substrate scope.[129]
The wider range of activities arises mainly from the extensive
choice of ligands, metals, and metal-containing cofactors. The
broader substrate scope is attributable to a wide range of
accessible reaction conditions, such as low temperatures, high
pressures, and different organic solvents. Biosynthetic catalysts, however, are mostly restricted to only the 20 natural
amino acids, physiologically available metals or metal-containing cofactors, physiological conditions, and water as
solvent. These limitations make biocatalysts even more
remarkable, and make it even more imperative to learn
methods from nature to make better catalysts. The introduction of nonnatural amino acids and nonnative metal-containing cofactors into proteins is an important step in this
direction.
In summary, recent advances in a number of areas in
biology and chemistry have enabled the development of
biosynthetic inorganic methods with comparable preparation
times and yields to those of synthetic inorganic chemistry. In
many cases, biosynthetic inorganic chemistry can produce
close structural and functional models of more-complex
metalloproteins and novel inorganic compounds. These
developments, however, only demonstrate the potential that
the biosynthetic approach has to offer; this potential is far
from being realized. Although methodologies have been
developed to introduce nonnative metal cofactors into
proteins by noncovalent, single-, and double-covalent attachment, the use of these approaches in asymmetrically catalyzed
syntheses of chiral intermediates for fine chemicals or
pharmaceutical products is still lacking. In contrast to the
field of asymmetric catalysis, which has had many years of
development by a number of researchers, the field of
biosynthetic inorganic chemistry is still in its infancy. Much
more time and effort, including cross-fertilization of the
synthetic and biosynthetic approaches, will be required before
the full potential of biosynthetic inorganic chemistry can be
realized.
I wish to thank the members of my research group whose work
has been cited in the references for their dedication and hard
work, as well as Hee Jung Hwang, Thomas D. Pfister, Dewain
Garner, and Natasha Yeung for help with the preparation of
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5599
Reviews
Y. Lu
figures in this review. The work of my group described in this
review has been supported by the US National Science
Foundation and the National Institutes of Health.
Received: January 16, 2006
Published online: August 10, 2006
[1] S. J. Lippard, J. M. Berg, Principles of Bioinorganic Chemistry,
University Science Books, Mill Valley, CA, 1994.
[2] I. Bertini, H. B. Gray, S. J. Lippard, J. S. Valentine, Bioinorganic
Chemistry, University Science Books, Sausalito, CA, 1994.
[3] J. J. R. Fra?sto da Silva, R. J. P. Williams, The Biological
Chemistry of the Elements, the Inorganic Chemistry of Life,
Clarendon Press, Oxford, 1991.
[4] J. A. Cowan, Inorganic Biochemistry, an Introduction, VCH,
New York, 1993.
[5] W. Kaim, B. Schwederski, Bioinorganic Chemistry: Inorganic
Elements in the Chemistry of Life, Wiley, Chichester, 1994.
[6] R. H. Holm, E. I. Solomon, Chem. Rev. 1996, 96, 2237.
[7] Ullmann6s Encyclopedia of Industrial Chemistry, 6th ed., 1998
Electronic Release, 1998.
[8] Riegel6s Handbook of Industrial Chemistry, 10th ed., 2003.
[9] A. Fersht, Structure and Mechanism in Protein Science: A Guide
to Enzyme Catalysis and Protein Folding, Freeman, New York,
NY, 1999.
[10] R. H. Holm, E. I. Solomon, Chem. Rev. 2004, 104, 347.
[11] R. I. Cukier, D. G. Nocera, Annu. Rev. Phys. Chem. 1998, 49,
337.
[12] A. S. Borovik, Acc. Chem. Res. 2005, 38, 54.
[13] B. A. Anderson, M. M. Hansen, A. R. Harkness, C. L. Henry,
J. T. Vicenzi, M. J. Zmijewski, J. Am. Chem. Soc. 1995, 117,
12 358.
[14] B. A. Anderson, M. M. Hansen, J. T. Vicenzi, M. J. Zmijewski,
Process Chem. Pharm. Ind. 1999, 263.
[15] Inorganic Electronic Structure and Spectroscopy, Vol. I (Eds.:
E. I. Solomon, A. B. P. Lever), Wiley-Interscience, New York,
1999.
[16] Inorganic Electronic Structure and Spectroscopy, Vol. II (Eds.:
E. I. Solomon, A. B. P. Lever), Wiley-Interscience, New York,
1999.
[17] Physical Methods in Bioinorganic Chemistry: Spectroscopy and
Magnetism (Ed.: L. Que, Jr.), University Science Books, Mill
Valley, CA, 2000.
[18] C. Dennison, G. W. Canters, Recl. Trav. Chim. Pays-Bas 1996,
115, 345.
[19] H. Beinert, Eur. J. Biochem. 1997, 245, 521.
[20] A. J. Vila, C. O. FernRndez in Handbook on Metalloproteins
(Eds.: I. Bertini, A. Sigel, H. Sigel), Dekker, New York, NY,
2001, p. 813.
[21] Y. Lu in Biocoordination Chemistry, Vol. 8 (Eds.: L. Que, Jr.,
W. B. Tolman), Elsevier, Oxford, UK, 2004, pp. 91.
[22] M. Saraste, Q. Rev. Biophys. 1990, 23, 331.
[23] B. G. MalmstrSm, Chem. Rev. 1990, 90, 1247.
[24] G. T. Babcock, M. Wikstrom, Nature 1992, 356, 301.
[25] S. M. Musser, M. H. B. Stowell, S. I. Chan, FEBS Lett. 1993,
327, 131.
[26] J. A. GarcTa-Horsman, B. Barquera, J. Rumbley, J. Ma, R. B.
Gennis, J. Bacteriol. 1994, 176, 5587.
[27] P. M. H. Kroneck, J. Riester, W. G. Zumft, W. E. Antholine,
Biol. Met. 1990, 3, 103.
[28] W. G. Zumft, P. M. H. Kroneck, Adv. Inorg. Biochem. 1996, 11,
193.
[29] N. J. Blackburn, M. E. Barr, W. H. Woodruff, J. van der Oost, S.
de Vries, Biochemistry 1994, 33, 10 401.
[30] F. A. Cotton, R. A. Walton, Multiple Bonds Between Metal
Atoms, Wiley, New York, 1982.
5600
www.angewandte.org
[31] B. J. Hathaway in Comprehensive Coordination Chemistry, the
Synthesis, Reactions, Properties & Application of Coordination
Compounds, Vol. 5, 1st ed. (Eds.: G. Wilkinson, R. D. Gillard,
J. A. McCleverty), Pergamon, Oxford, England, 1987, p. 533.
[32] M. Dunaj-Jurco, G. Ondrejovic, M. Melnik, J. Garaj, Coord.
Chem. Rev. 1988, 83, 1.
[33] C. Harding, V. McKee, J. Nelson, J. Am. Chem. Soc. 1991, 113,
9684.
[34] M. E. Barr, P. H. Smith, W. E. Antholine, B. Spencer, J. Chem.
Soc. Chem. Commun. 1993, 1649.
[35] C. Harding, J. Nelson, M. C. R. Symons, J. Wyatt, J. Chem. Soc.
Chem. Commun. 1994, 2499.
[36] R. P. Houser, V. G. Young, Jr., W. B. Tolman, J. Am. Chem. Soc.
1996, 118, 2101.
[37] D. D. LeCloux, R. Davydov, S. J. Lippard, Inorg. Chem. 1998,
37, 6814.
[38] C. He, S. J. Lippard, Inorg. Chem. 2000, 39, 5225.
[39] R. Gupta, Z. H. Zhang, D. Powell, M. P. Hendrich, A. S.
Borovik, Inorg. Chem. 2002, 41, 5100.
[40] X.-M. Zhang, M.-L. Tong, X.-M. Chen, Angew. Chem. 2002,
114, 1071; Angew. Chem. Int. Ed. 2002, 41, 1029.
[41] D. D. LeCloux, R. Davydov, S. J. Lippard, J. Am. Chem. Soc.
1998, 120, 6810.
[42] S. B. Harkins, J. C. Peters, J. Am. Chem. Soc. 2004, 126, 2885.
[43] C. Dennison, E. Vijgenboom, S. de Vries, J. van der Oost, G. W.
Canters, FEBS Lett. 1995, 365, 92.
[44] M. Hay, J. H. Richards, Y. Lu, Proc. Natl. Acad. Sci. USA 1996,
93, 461.
[45] G. J. Steffens, G. Buse, Hoppe-Seyler6s Z. Physiol. Chem. 1979,
360, 613.
[46] H. Robinson, M. C. Ang, Y.-G. Gao, M. T. Hay, Y. Lu, A. H. J.
Wang, Biochemistry 1999, 38, 5677.
[47] X. Wang, M. C. Ang, Y. Lu, J. Am. Chem. Soc. 1999, 121, 2947.
[48] M. T. Hay, R. M. Milberg, Y. Lu, J. Am. Chem. Soc. 1996, 118,
11 976.
[49] M. T. Hay, Y. Lu, J. Biol. Inorg. Chem. 2000, 5, 699.
[50] H. J. Hwang, N. Nagraj, Y. Lu, Inorg. Chem. 2006, 45, 102 ? 107.
[51] X. Wang, S. M. Berry, Y. Xia, Y. Lu, J. Am. Chem. Soc. 1999,
121, 7449.
[52] S. M. Berry, X. Wang, Y. Lu, J. Inorg. Biochem. 2000, 78, 89.
[53] H. J. Hwang, S. M. Berry, M. J. Nilges, Y. Lu, J. Am. Chem. Soc.
2005, 127, 7274.
[54] H. J. Hwang, Y. Lu, Proc. Natl. Acad. Sci. USA 2004, 101,
12 842.
[55] O. Farver, Y. Lu, M. C. Ang, I. Pecht, Proc. Natl. Acad. Sci.
USA 1999, 96, 899.
[56] H. B. Gray, B. G. MalmstrSm, R. J. P. Williams, J. Biol. Inorg.
Chem. 2000, 5, 551.
[57] E. I. Solomon, D. W. Randall, T. Glaser, Coord. Chem. Rev.
2000, 200?202, 595.
[58] B. E. Ramirez, B. G. MalmstrSm, J. R. Winkler, H. B. Gray,
Proc. Natl. Acad. Sci. USA 1995, 92, 11 949.
[59] Y. Lu, S. M. Berry, T. D. Pfister, Chem. Rev. 2001, 101, 3047.
[60] A. Gengenbach, X. Wang, Y. Lu in Fundamentals and Catalysis
of Oxidative Delignification Processes, Vol. 785 (Ed.: D. S.
Argyropoulos), American Chemical Society, Washington, DC,
2001, p. 487.
[61] ?Manganese and Its Role in Biological Processes?: M. H. Gold,
H. L. Youngs, M. D. S. Gelpke, Met. Ions Biol. Syst. 2000, 37,
559.
[62] M. Sundaramoorthy, K. Kishi, M. H. Gold, T. L. Poulos, J. Biol.
Chem. 1994, 269, 32 759.
[63] B. K. Yeung, X. Wang, J. A. Sigman, P. A. Petillo, Y. Lu, Chem.
Biol. 1997, 4, 215.
[64] S. K. Wilcox, C. D. Putnam, M. Sastry, J. Blankenship, W. J.
Chazin, D. E. McRee, D. B. Goodin, Biochemistry 1998, 37,
16 853.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
[65] X. Wang, Y. Lu, Biochemistry 1999, 38, 9146.
[66] A. Gengenbach, S. Syn, X. Wang, Y. Lu, Biochemistry 1999, 38,
11 425.
[67] T. D. Pfister, A. J. Gengenbach, S. Syn, Y. Lu, Biochemistry
2001, 40, 14 942.
[68] P. R. O. de Montellano, Cytochrome P450 Structure, Mechanism, and Biochemistry, 2nd ed., Plenum, New York, 1995.
[69] S. Adachi, S. Nagano, Y. Watanabe, K. Ishimori, I. Morishima,
Biochem. Biophys. Res. Commun. 1991, 180, 138.
[70] S. Adachi, S. Nagano, K. Ishimori, Y. Watanabe, I. Morishima,
T. Egawa, T. Kitagawa, R. Makino, Biochemistry 1993, 32, 241.
[71] D. P. Hildebrand, J. C. Ferrer, H.-L. Tang, M. Smith, A. G.
Mauk, Biochemistry 1995, 34, 11 598.
[72] K. Choudhury, M. Sundaramoorthy, A. Hickman, T. Yonetani,
E. Woehl, M. F. Dunn, T. L. Poulos, J. Biol. Chem. 1994, 269,
20 239.
[73] J. A. Sigman, A. E. Pond, J. H. Dawson, Y. Lu, Biochemistry
1999, 38, 11 122.
[74] R. Perera, M. Sono, J. A. Sigman, T. D. Pfister, Y. Lu, J. H.
Dawson, Proc. Natl. Acad. Sci. USA 2003, 100, 3641.
[75] T. Uno, A. Yukinari, Y. Tomisugi, Y. Ishikawa, R. Makino, J. A.
Brannigan, A. J. Wilkinson, J. Am. Chem. Soc. 2001, 123, 2458.
[76] H. Fujii, X. Zhang, T. Tomita, M. Ikeda-Saito, T. Yoshida, J.
Am. Chem. Soc. 2001, 123, 6475.
[77] T. Matsui, M. Furukawa, M. Unno, T. Tomita, M. Ikeda-Saito, J.
Biol. Chem. 2005, 280, 2981.
[78] J. A. Sigman, H. K. Kim, X. Zhao, J. R. Carey, Y. Lu, Proc. Natl.
Acad. Sci. USA 2003, 100, 3629.
[79] J. P. Collman, R. Boulatov, C. J. Sunderland, L. Fu, Chem. Rev.
2004, 104, 561.
[80] E. Kim, E. E. Chufan, K. Kamaraj, K. D. Karlin, Chem. Rev.
2004, 104, 1077.
[81] J. A. Sigman, B. C. Kwok, Y. Lu, J. Am. Chem. Soc. 2000, 122,
8192.
[82] X. Zhao, N. Yeung, Z. Wang, Z. Guo, Y. Lu, Biochemistry 2005,
44, 1210.
[83] X. Zhao, M. J. Nilges, Y. Lu, Biochemistry 2005, 44, 6559.
[84] Y. Lu, Curr. Opin. Chem. Biol. 2005, 9, 118.
[85] H. K. Privett, C. J. Reedy, M. L. Kennedy, B. R. Gibney, J. Am.
Chem. Soc. 2002, 124, 6828.
[86] S. M. Berry, M. D. Gieselman, M. J. Nilges, W. A. van der
Donk, Y. Lu, J. Am. Chem. Soc. 2002, 124, 2084.
[87] M. Ralle, S. M. Berry, M. J. Nilges, M. D. Gieselman, W. A.
van der Donk, Y. Lu, N. J. Blackburn, J. Am. Chem. Soc. 2004,
126, 7244.
[88] S. M. Berry, M. Ralle, D. W. Low, N. J. Blackburn, Y. Lu, J. Am.
Chem. Soc. 2003, 125, 8760.
[89] D. W. Low, M. G. Hill, J. Am. Chem. Soc. 1998, 120, 11 536.
[90] D. W. Low, M. G. Hill, J. Am. Chem. Soc. 2000, 122, 11 039.
[91] T. Hayashi, A. Tomokuni, T. Mizutani, Y. Hisaeda, H. Ogoshi,
Chem. Lett. 1998, 1229.
[92] T. Hayashi, Y. Hitomi, T. Ando, T. Mizutani, Y. Hisaeda, S.
Kitagawa, H. Ogoshi, J. Am. Chem. Soc. 1999, 121, 7747.
[93] T. Hayashi, T. Matsuda, Y. Hisaeda, Chem. Lett. 2003, 32, 496.
[94] H. Sato, T. Hayashi, T. Ando, Y. Hisaeda, T. Ueno, Y.
Watanabe, J. Am. Chem. Soc. 2004, 126, 436.
[95] T. Hayashi, T. Nakagawa, K. Harada, T. Matsuo, Y. Hitomi, Y.
Hisaeda, Chem. Lett. 2004, 33, 1512.
[96] H. Sato, M. Watanabe, Y. Hisaeda, T. Hayashi, J. Am. Chem.
Soc. 2005, 127, 56.
[97] T. Hayashi, Y. Hitomi, T. Takimura, A. Tomokuni, T. Mizutani,
Y. Hisaeda, H. Ogoshi, Coord. Chem. Rev. 1999, 190?192, 961.
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
[98] I. Hamachi, S. Shinkai, Eur. J. Org. Chem. 1999, 539.
[99] T. Hayashi, Y. Hisaeda, Acc. Chem. Res. 2002, 35, 35.
[100] M. Skander, N. Humbert, J. Collot, J. Gradinaru, G. Klein, A.
Loosli, J. Sauser, A. Zocchi, F. Gilardoni, T. R. Ward, J. Am.
Chem. Soc. 2004, 126, 14 411.
[101] T. Ueno, M. Ohashi, M. Kono, K. Kondo, A. Suzuki, T. Yamane,
Y. Watanabe, Inorg. Chem. 2004, 43, 2852.
[102] T. Matsuo, T. Hayashi, Y. Hisaeda, J. Am. Chem. Soc. 2002, 124,
11 234.
[103] T. Hayashi, H. Dejima, T. Matsuo, H. Sato, D. Murata, Y.
Hisaeda, J. Am. Chem. Soc. 2002, 124, 11 226.
[104] M. E. Wilson, G. M. Whitesides, J. Am. Chem. Soc. 1978, 100,
306.
[105] C.-C. Lin, C.-W. Lin, A. S. C. Chan, Tetrahedron: Asymmetry
1999, 10, 1887.
[106] J. Collot, J. Gradinaru, N. Humbert, M. Skander, A. Zocchi,
T. R. Ward, J. Am. Chem. Soc. 2003, 125, 9030.
[107] C. M. Thomas, T. R. Ward, Chem. Soc. Rev. 2005, 34, 337.
[108] G. Klein, N. Humbert, J. Gradinaru, A. Ivanova, F. Gilardoni,
U. E. Rusbandi, T. R. Ward, Angew. Chem. Int. Ed. 2005, 44,
7764.
[109] M. Skander, C. Malan, A. Ivanova, T. R. Ward, Chem.
Commun. 2005, 4815.
[110] T. R. Ward, Chem. Eur. J. 2005, 11, 3798.
[111] C. Letondor, N. Humbert, T. R. Ward, Proc. Natl. Acad. Sci.
USA 2005, 102, 4683.
[112] M. Ohashi, T. Koshiyama, T. Ueno, M. Yanase, H. Fujii, Y.
Watanabe, Angew. Chem. 2003, 115, 1035; Angew. Chem. Int.
Ed. 2003, 42, 1005.
[113] T. Ueno, T. Koshiyama, M. Ohashi, K. Kondo, M. Kono, A.
Suzuki, T. Yamane, Y. Watanabe, J. Am. Chem. Soc. 2005, 127,
6556.
[114] Y. Watanabe, T. Hayashi, Prog. Inorg. Chem. 2005, 54, 449.
[115] F. V. Cochran, S. P. Wu, W. Wang, V. Nanda, J. G. Saven, M. J.
Therien, W. F. DeGrado, J. Am. Chem. Soc. 2005, 127, 1346.
[116] G. W. Pettigrew, J. L. Leaver, T. E. Meyer, A. P. Ryle, Biochem.
J. 1975, 147, 291.
[117] R. A. Scott, A. G. Mauk, Cytochrome c: A Multidisciplinary
Approach, University Science Books, Sausalito, CA, 1996.
[118] R. R. Davies, M. D. Distefano, J. Am. Chem. Soc. 1997, 119,
11 643.
[119] H. J. Hwang, J. R. Carey, E. T. Brower, A. J. Gengenbach, J. A.
Abramite, Y. Lu, J. Am. Chem. Soc. 2005, 127, 15 356.
[120] R. Chen, Trends Biotechnol. 2001, 19, 13.
[121] S. A. Datwyler, C. F. Meares, Met. Ions Biol. Syst. 2001, 38, 213.
[122] D. Qi, C.-M. Tann, D. Haring, M. D. Distefano, Chem. Rev.
2001, 101, 3081.
[123] M. T. Reetz, M. Rentzsch, A. Pletsch, M. Maywald, Chimia
2002, 56, 721.
[124] J. R. Carey, S. K. Ma, T. D. Pfister, D. K. Garner, H. K. Kim,
J. A. Abramite, Z. Wang, Z. Guo, Y. Lu, J. Am. Chem. Soc.
2004, 126, 10 812.
[125] M. T. Reetz, A. Zonta, K. Schimossek, K. Liebeton, K.-E.
Jaeger, Angew. Chem. 1997, 109, 2961; Angew. Chem. Int. Ed.
1997, 36, 2830.
[126] M. T. Reetz, Proc. Natl. Acad. Sci. USA 2004, 101, 5716.
[127] J. A. Sigman, B. C. Kwok, A. Gengenbach, Y. Lu, J. Am. Chem.
Soc. 1999, 121, 8949.
[128] E. N. Jacobsen, N. S. Finney, Chem. Biol. 1994, 1, 85.
[129] Comprehensive Asymmetric Catalysis I?III, Vol. 1, (Eds.: E. N.
Jacobsen, A. Pfaltz, H. Yamamoto), Springer, New York, 1999.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5601
roach are discussed.
1. Introduction
Metal ions play important roles in catalyzing numerous
chemical and biological reactions, as the reactivity and
diversity can be modulated at high levels by not only different
metal ions, but also the different oxidation states of these
same metal ions.[1?6] The successful synthesis and application
of novel metal complexes or metalloenzymes can have a great
impact on all areas of chemistry and biology, whether it is
organic, medicinal, materials chemistry, or biochemistry and
cell biology. An important example of such an impact is the
development of metal catalysts for reactions such as nitrogen
fixation, methane hydroxylation, and CO oxidation or
insertion (Table 1).[7, 8] These catalysts are key components
for multibillion-dollar chemical industries that produce vital
raw chemicals, such as fertilizers.
Interestingly, many of the reactions in Table 1 can be
catalyzed by either synthetic catalysts or by biological
enzymes. A closer look at the two systems suggests that the
catalysts mediate the same reaction under different conditions: chemical systems generally operate at high temperatures and pressures in organic solvent, whereas biological
systems work at ambient temperatures and pressures in water.
Furthermore, biocatalysts often have high catalytic turnovers.
Therefore, the question is not whether one can develop
catalysts that can work under milder conditions with high
turnovers, but rather how such catalysts can be developed.
A cursory look at the examples in Table 1 may lead one to
conclude that one system may be better than the other;
however, each system has its own advantages and disadvantages (Table 2). Chemical catalysts are generally smaller,
easier to synthesize and characterize, cheaper to produce, and
more resistant to harsh conditions such as high temperatures
and pressures. These practical attributes make it difficult to
replace catalysts with enzymes in industrial-scale synthesis,
although enzymes possess a number of desirable features for
catalysts. First, they form more-rigid active sites to stabilize
the transition state and thus are more able to accelerate
catalysis at ambient temperature and pressure. Second, they
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
From the Contents
1. Introduction
5589
2. Application of Biosynthetic
Inorganic Chemistry in
Biomimetic Model Systems
5591
3. Biosynthetic Inorganic
Chemistry for the Synthesis of
New Inorganic Complexes
5595
4. Conclusions and Outlook
5598
direct reactions towards a specific
functional group of an organic substrate, thus eliminating or minimizing
the need for protecting groups, which
are often required in synthetic chemistry. Third, the catalytic
site in enzymes is protected, which results in higher catalyst
stability and higher turnover numbers. Fourth, enzymes
achieve high regio-, stereo-, and enantioselectivity. Fifth, the
complex protein framework is amenable to site-specific (as
opposed to solvent-induced) modulation of the secondary
coordination sphere of a metal-binding site. For example, the
pKa values for glutamic acid in the same protein vary between
4.0 and 8.2, depending on the local electrostatic environment
and hydrophobicity.[9] Finally, enzymes are environmentally
benign, as they are synthesized and applied under physiological conditions in aqueous solvents and with biocompatible
and biodegradable ligands.
Since each of the chemical and biological systems has its
own advantages and disadvantages, one may wonder if the
advantages of both systems can be combined. Towards this
end, a number of synthetic biomimetic model systems for
metalloproteins have been prepared from small organic
molecules and characterized.[10] Such a synthetic approach
has been very effective in the elucidation of structural and
functional properties of metalloproteins.
Despite tremendous progress in biomimetic modeling, it is
still relatively difficult to mimic some of the features of
metalloproteins, for example, site-specific modulation of the
secondary coordination sphere and the regio-, stereo-, and
enantioselectivity of a system. Models that reproduce both
the structure and the function of metalloproteins are rare.
These challenges, like perhaps all obstacles in synthetic
inorganic chemistry, can be overcome by ligand design.
Prudent choice of ligand has repeatedly resulted in better
models of metalloproteins,[10] especially when features in the
secondary coordination sphere such as hydrogen-bonding
[*] Prof. Y. Lu
Department of Chemistry
University of Illinois at Urbana-Champaign
Urbana, IL 61801 (USA)
Fax: (+ 1) 217-244-3186
E-mail: yi-lu@uiuc.edu
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5589
Reviews
Y. Lu
Table 1: Comparison of chemical and biological systems.
Reaction
Chemical System
Biological System
nitrogen fixation
N2(g) + 3 H2(g)!2 NH3(g)
(a-Fe catalyst, 400?550 8C, 100 atm)
CH4 + H2O!CO + 3 H2
(Ni catalyst, 700?900 8C, 1?25 bar)
CO + 2 H2 !CH3OH
(Cu/Zn Catalyst, 250?280 8C, 70?110 bar)
CO + H2O!CO2 + H2
(Fe/Cu catalyst, > 200 8C)
CH3OH + CO!CH3COOH
(Rh(I)I2(CO)2] , 120 8C, 30 atm)
N2 + 8 H+ + 8 e + 16 MgATP!2 NH3 + H2 + 16 MgADP + 16 Pi
(nitrogenase)
CH4 + O2 + NADH + H+!CH3OH + H2O + NAD+
(methane monooxygenase)
methane hydroxylation
CO oxidation
CO insertion
CO + H2O!CO2 + 2 H+ + 2e
(CO dehydrogenase)
CH3[M] + CO + HSCoA!CH3(CO)SCoA + H+ + [M][a]
(acetyl-CoA synthase)
[a] CH3[M] is a corrinoid?iron?sulfur protein that acts in the reaction as a methyl group donor; HS-CoA is coenzyme A.
Table 2: Advantages of chemical and biological systems.
Chemistry
Catalysts/Materials
Biology
Enzymes/Biomaterials
easy to synthesize
easy to characterize
more robust
mild conditions
rigidity/protection
site-specific modulation of the
2nd coordination sphere
regio- and enantioselectivity
environmentally friendly
approach: study native proteins and their variants
cheaper to produce
approach: create new structures and
functions by using a minimalist
approach
variants, which rendered the use of such proteins as ligands
impractical. However, advancements in biology, such as the
development of cloning, protein expression, site-directed
mutagenesis, and polymerase chain reaction, have made it
possible to reduce dramatically the time required to synthesize proteins and change amino acid residues. Whereas it used
to take many months or even many years to construct,
express, and purify a series of proteins, it now takes a matter
of weeks or even days; such progress is similar to that made in
the computer industry (Figure 1). Furthermore, hundreds of
milligrams or even grams of pure protein can now be
were introduced.[11, 12] The main difference between synthetic
models and metalloproteins is that the synthetic models often
contain water-insoluble organic ligands, whereas metalloproteins consist of polypeptides or proteins. Polypeptides and
proteins are unusual organic molecules in that they are
naturally soluble in water and can be synthesized under mild
conditions. Through specific folding of the polypeptides and
proteins, a rigid network can form that allows site-specific
modulation of the secondary coordination sphere. This
suggests that polypeptides or proteins were established in
the course of evolution as ligands for metal-based catalysts or
enzymes.
So why then should the ligands from biological systems
not be used to synthesize model compounds? It used to be
very difficult and expensive to make proteins and their
Yi Lu received his BS degree from Beijing
University (P.R. China) in 1986 and his PhD
from the University of California at Los
Angeles in 1992 under the guidance of Prof.
Joan Selverstone Valentine. After two years
of postdoctoral research at the California
Institute of Technology in the group of Prof.
Harry B. Gray, he began his independent
career as an Assistant Professor in the
Department of Chemistry at the University
of Illinois at Urbana?Champaign, where he
is now an Alumni Research Scholar Professor
of Chemistry. His interests lie in bioinorganic
chemistry.
5590
www.angewandte.org
Figure 1. Comparison of progresses made in the computer industry
and biology. A) According to Moore?s law, the number of transistors
and thus the speed of a computer doubles approximately every
18 months. B) Estimated average speeds of cloning, expression, and
purification of recombinant proteins (calculated as the inverse of
number of days it takes to complete the process). Techniques such as
recombinant DNA (Rec. DNA), site-directed mutagenesis (SDM), and
the polymerase chain reaction (PCR) have significantly accelerated the
biological progress.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
prepared. Thanks to the progress in biotechnology, certain
chemical synthesis steps in the production of pharmaceutical
products can be replaced by biosynthesis. As a result, the
number of synthesis steps is reduced, yields are increased, and
less chemical waste is produced,[13, 14] so that the difference in
efficiency between the synthesis of small organic molecules
and proteins is decreased. At the same time, progress in
biophysics has made possible the routine characterization of
metalloproteins by different spectroscopic techniques.[15?17]
Advances in structural biology, such as the availability of
high-energy synchrotron sources and high-field magnets and
their associated methodology developments, have also made
it possible to obtain three-dimensional crystal or NMR
structures of proteins in a short period of time. Thus, it is
now possible to synthesize and characterize novel inorganic
compounds using small, stable, easy-to-produce proteins with
characteristic scaffolds as ?ligands?.
The biosynthetic approach uses the same techniques as
the biochemical studies of native enzymes and their mutant
forms. However, the following examples below show that
there are distinct differences between the two approaches:
First, like all biomimetic approaches, the biosynthetic
approach is a minimalist approach; it focuses on determining
whether the necessary structural features identified in native
enzymes are sufficient to confer the structure and function of
the enzyme. Therefore, the two approaches are complementary. Second, stable, easy-to-produce, and well-characterized
proteins are used in the biosynthetic approach, and successful
models are often as good as, if not better than, the native
enzymes for structural and functional studies. Third, biosynthetic models may offer additional advantages, since different
metal-binding centers can be compared in the same protein
framework. Finally, the biosynthetic approach allows synthesis of new compounds, with structures and reactivities
unprecedented in either biology or inorganic chemistry.
2. Application of Biosynthetic Inorganic Chemistry
in Biomimetic Model Systems
Biomimetic model systems combine the benefits of both
chemistry and biology. In this process, one can learn the
minimal structural features necessary to form metal-binding
sites with desirable function and help to elucidate the
structure?function relationship of target proteins. At the
same time, new compounds with structures that are rare in
inorganic chemistry may be synthesized, which expands our
knowledge of coordination chemistry. The resulting compounds are often smaller and better defined structurally than
the native proteins, and can therefore be used in practical
applications.
Biosynthetic model systems with small, stable, and wellcharacterized proteins as ligands can achieve the same goals.
Since they use the same ligands as target proteins and operate
under almost the same physiological conditions, they can
often result in closer models and are more readily usable in
practical applications. A few examples will illustrate these
features.
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
2.1. Inorganic Biosynthesis with the Same Protein Scaffold and
Loop-Directed Mutagenesis
The CuA center is a mixed-valence dinuclear copper
center in which the copper ions are each coordinated to a
histidine residue (His 181 and His 224) and bridged by the
thiolate sulfur atoms of two cysteine ligands (Cys 216 and
Cys 220, Figure 2 A, top).[18?21] Weak ligands such as the
Figure 2. Crystal structures of A) the CuA center in cytochrome c
oxidase from P. denitrificans; and B) a biosynthetic CuA model in
azurin. The top figures show the CuA center with Cu2S2(Cys) in a plane
and the bottom figures show the CuA center viewed perpendicular to
the plane.
thioether sulfur atom of methionine (Met227) and the
backbone carbonyl oxygen atoms of isoleucine (Ile 180),
histidine (His 224), and glutamate (Glu 218) are also present
in axial positions (Figure 2 A, bottom). The CuA center is
unique in both inorganic chemistry and biology; it acts as an
electron-transfer center in cytochrome c oxidase (CcO; a
terminal oxidase in the respiratory chain of eukaryotic
mitochondria and some aerobic bacteria)[22?26] and nitrous
oxide reductase (N2OR; an enzyme responsible for the
reduction of N2O in denitrifying bacteria).[27, 28] It was also
the first biological system shown to contain a metal?metal
bond.[29] Many coordination compounds containing metal?
metal bonds are known with late transition metals; however
few contain first-row transition metals such as copper.[30]
Whereas dinuclear or multinuclear mixed-valence copper
complexes in which the unpaired electron is completely
localized (Class I) or partially delocalized (Class II) have
been known for many years,[31, 32] copper complexes, such as
the CuA center, with fully delocalized (Class III) mixedvalence states are rare.[33?40] In both cases, the synthesis of the
CuA center and elucidation of the structural features responsible for its functions are of great interest.
A number of synthetic models with organic ligands have
been reported for this purpose, most of which contain a
dinuclear mixed-valence copper center and a copper?copper
bond, but not bridging thiolate units.[33, 34, 38, 39, 41] A model
compound that contains bridging thiolate residues but lacks a
copper?copper bond also exists.[36] A dinuclear copper center
containing a copper?copper bond and two bridging nitrogen
atoms was also reported recently.[42] Stable, easy-to-make, and
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5591
Reviews
Y. Lu
well-characterized blue copper proteins such as amicyanin[43]
and azurin[44] were used as ligands to make a closer model of
the CuA center. Blue copper proteins and proteins containing
the CuA center have been shown to share close structural
homology.[45] Sequence alignment revealed that mononuclear
blue copper proteins and dinuclear CuA proteins differ mainly
in a single loop between two b-strands. By replacing the loop
in blue copper proteins with the corresponding loop in CuA
proteins (Figure 3), a close mimic of the CuA center was
Figure 3. Development of a biosynthetic CuA model in azurin through
loop-directed mutagenesis.
synthesized in blue copper proteins amicyanin[43] and
azurin.[44] A high-resolution crystal structure showed that
the CuA center in the azurin model is almost identical to that
of the native CuA center, including structural features in the
secondary coordination sphere, such as the nonbonding
interactions between the thioether group of methionine
residue and peptide carbonyl oxygen atoms (Figure 2 B).[46]
The synthesis of a close model of a native protein is only
one measure of the effectiveness of such an approach. How
much insight is gained into the structure and function of the
metal center is another important measure. The biosynthetic
CuA model also gives deeper insight in this regard. For
example, a stopped-flow study of copper incorporation into
metal-free CuA azurin revealed a tetragonal intermediate and
showed the importance of reductants in the formation of the
final CuA center.[47] Titration of different metal ions into the
model protein suggests that the CuA center as an MII M? I
center (M, M? = any metal) is strongly favored regardless of
the sequence of addition of metal ions.[48, 49] Mutagenesis
studies revealed that the conserved cysteine,[50] histidine,[51, 52]
and methinione[53] residues all play an important role in
maintaining the structure and in fine-tuning the function.
Surprisingly, the dinuclear copper structure remains largely
intact, even when the histidine was replaced with residues
incapable of coordinating to the copper (e.g. Ala and Gly), as
evidenced by the similar characteristic purple color, as well as
UV/Vis, magnetic circular dichroism (MCD), resonance
Raman (RR), and electron nuclear double resonance
(ENDOR) spectra.[51, 52]
Until recently, there were not many examples of fully
delocalized (Class III) dinuclear mixed-valence copper compounds. Even rarer is the reversible conversion between
different classes, that is, between delocalized [Cu1.5иииCu1.5]
and trapped valence [CuIIиииCuI] states. The study of the CuA
5592
www.angewandte.org
center in azurin provided an interesting example of a
reversible transition between delocalized and trapped
mixed-valence states that is triggered solely by a change in
pH value.[54] The CuA center is known to be the electron entry
site in cytochrome c oxidase. However, its role in protoncoupled electron transfer has not been defined clearly. This
work showed that protonation of the C-terminal His residue
not only resulted in a trapped valance state of the CuA center,
but also caused a significant increase in the reduction
potential.[54] Since the corresponding C-terminal His residue
in cytochrome c oxidases is located along a major electrontransfer pathway from CuA center to heme a, and since
protonation can result in an increased reduction potential that
prevents electron transfer from the CuA center to heme a, the
results strongly indicate that the CuA center and the histidine
residue may play important roles in proton-coupled electron
transfer.[54]
These observations may be obtained from studies of
native CuA centers, in particular soluble fragments of CcO
containing the CuA center.[18?21] However, several studies have
demonstrated that biosynthetic models can provide unique
insights that are not readily obtainable from studies of either
native enzymes or synthetic model compounds. For example,
by using azurin as the ?ligand? and replacing the ligand loop
of the blue copper center with that of the CuA center, two
electron transfer (ET) centers may be placed in the same
protein framework. High-resolution crystal structures showed
that one of the copper atoms in the CuA center overlays
exactly with the blue copper center (Figure 4 A).[46] Therefore,
Figure 4. A) Overlay of the crystal structures of blue copper azurin (in
light blue) and the biosynthetic CuA model in azurin (in purple). The
copper site in the blue copper center (B) overlays almost exactly with
one of the copper sites in the purple CuA center (C).
this biosynthetic model provided a unique opportunity to
compare the two ET centers in the same protein framework
and exclude or at least minimize other factors that influence
its properties. For example, the same series of mutations at a
conserved axial methionine unit in the blue copper protein
(Met 121, Figure 4 B) and the CuA azurin (Met 123, Figure 4 C) indicate that the methionine residue has much less
influence on the reduction potential of the CuA center
(< 25 mV) than of the blue copper center (> 170 mV).[53] In
contrast, electron transfer from the same donor and at the
same distance to the CuA center is faster, despite its lower
reduction potential, than to the blue copper center.[55] Thus,
this work showed directly that the CuA center is a more
efficient ET center.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
These findings may be important in understanding the
different roles of the two copper centers. A much wider range
of reduction potentials (> 600 mV) is required for blue
copper proteins to transfer electrons to a variety of partners
in many different biological systems.[20, 21, 56, 57] In contrast, the
CuA center is part of cytochrome c oxidase, which is at the end
of the respiration chain with very small differences (< 50 mV)
in the potentials of the redox partners.[58] In this case, large
variations in the potential could hinder the flow of electrons
in the right direction. The diamond-shaped Cu2S2(Cys) core
structure of CuA provides a way to minimize changes in
reduction potential through variations of the axial ligands.
Despite the low driving force, however, the CuA center still
needs to transfer electrons to its partners at a desirable rate.
Therefore, it has to optimize its structure by lowering its
reorganization energy to be a more efficient ET center than
the blue copper center.
Loop-directed mutagenesis, in which ligand loops of a
template protein are swapped with those of a target protein
that shares similar structural homology, was employed to
make the biosynthetic models of the CuA center described
above. A preliminary survey of the protein structure data
bank (PDB) suggests that this approach is not limited to
specific cases but can be utilized generally to make biosynthetic models of many metalloproteins. For example, azurin,
amicyanin, and CuA proteins belong to a group with a
common scaffold, called the Greek Key b-barrel. It is the
second most abundant motif in the PDB, and a number of
proteins with diverse active-site structures and functions, such
as immunoglobin, beta-amylase, cytochrome c oxidase, nitrite
reductase, and superoxide dismutase, contain the same
scaffold (Figure 5).[59]
heme peroxidases contain a heme active site. However,
manganese peroxidase (MnP) contains an additional MnIIbinding site (Figure 6 A). MnII oxidation plays a critical role in
the function of MnP (Scheme 1) in the biodegradation of
lignin, the second most abundant biopolymer on earth (after
Figure 6. A) Comparison of manganese peroxidase (MnP) and cytochrome c peroxidase (CcP); B) Heme-manganese binding site in MnP
and CcP; C) Crystal structures of MnP (red) and a CcP variant that
mimics MnP (blue).
Figure 5. The Greek Key b-barrel motif appears in more than 600
proteins of different classes with diverse functions. As most of the
active sites are in the loop region, loop-directed mutagenesis can be
used to construct novel metalloproteins.
Most of the active sites in these proteins reside within or
between loops. Evidently, nature produced a thermodynamically stable scaffold upon which many different protein
structures and functions can be conferred by simply swapping
loops, which is much less disruptive to the protein structure
than changing scaffolds. This ?trick? is important for biosynthetic inorganic chemistry.
Another example of nature using the same protein
scaffold for different functions are heme peroxidases.[60] All
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Scheme 1. Catalytic cycles of CcP (top) and MnP (bottom). Trp = tryptophan; Por = heme porphyrin.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5593
Reviews
Y. Lu
cellulose), as well as being a key step in converting woody
plants to small chemicals such as those in petroleum.[61]
Interestingly, cytochrome c peroxidase (CcP) shares not
only the same overall structure, but also a similar structure
of the heme-binding site, the major differences being that CcP
lacks the MnII-binding site and contains Trp instead of Phe
around the heme-binding site (Figure 6 A,B).[62] Several
ligands to MnII in MnP are absent in CcP. To convert CcP,
an enzyme found in bakerIs yeast, into an MnP with similar
function, the residues in CcP were replaced with the residues
of MnP at the corresponding positions to give new CcP
variants with enhanced MnP function (Figure 6 C).[63?65]
Furthermore, replacing the Trp residues in CcP with the
Phe units from MnP showed that the Phe residues play
different roles in MnP function; whereas the Trp 191 Phe
mutation stabilized compound I (an intermediate), the
Trp 51 Phe mutation increased the activity significantly, as it
increased the reactivity of compound II, whose oxidation of
MnII is the rate-determining step in the reaction mechanism of
MnP.[66, 67]
2.2. Noncovalent Tuning of the Secondary Coordination Sphere
through Inorganic Biosynthesis
The noncovalent tuning of the secondary coordination
sphere of metal-binding sites is another ?trick? of biological
systems. For example, cytochrome P450 (Cyt P450) is one of
the most efficient enzymes for asymmetric oxygen-transfer
reactions, such as sulfoxidation, epoxidation, and hydroxylation, of a variety of substrates.[68] Its primary coordination
sphere consists of a heme unit with a thiolate Cys as its
proximal ligand (Figure 7 A). In view of the importance of
heme?thiolate ligation, the proximal ligand His in human
myoglobin was changed to a Cys.[69, 70] The spectrum of the
resulting protein was similar to that of Cyt P450. The protein
model also exhibited increased asymmetric sulfoxidation
activity. However, the same His-to-Cys mutation alone in
horse heart myoglobin (Mb) did not result in heme?thiolate
ligation.[71] Instead, an additional mutation of the distal
noncoordinating histidine to either a valine or isoleucine
resulted in a P450-like protein with an iron(III) resting state
(Figure 7 B).[71] This effect, termed the trans effect by the
authors, mimics the more hydrophobic environment of the
Cyt P450 (Figure 7 A) and contributes significantly to the
binding of the axial ligand and to the stability.
Similarly, mutation of the axial His ligand to Cys in CcP
resulted in a very unstable ligand that was rapidly oxidized to
cysteic acid.[72] It was then recognized that a nonpolar residue
next to the Cys unit is conserved in P450 proteins, while the
analogous amino acid in CcP is an aspartic acid (Figure 7 C).
Thus, further mutation of this Asp residue to Leu resulted in a
stable, pentacoordinate, high-spin heme with thiolate ligation.[73] The study also marked the first time a stable
cyanidoferric complex and the ferrous state of a model P450
was obtained.[73, 74]
Stable coordination of cysteine thiolate in the ferrous
carbonyl derivative of an engineered protein was also
obtained by mutations of the axial His and Met ligands in
cytochrome b562 to Cys and Gly, respectively (Figure 7 D).[75]
The key to success was the replacement of two glutamate
residues (Glu 4 and Glu 8), which are electrostatically repelled from the heme propionate groups, with Ser. These studies
demonstrated the importance of the secondary coordination
sphere around the primary coordination ligands in stabilizing
metal?ligand ligation.
2.3. Water and Hydrogen Bonding in Inorganic Biosynthesis
Figure 7. Key structural differences of Cyt P450 (A), Mb (B), CcP (C),
and Cyt b562 (D).
5594
www.angewandte.org
The third ?trick? emerging from recent investigations on a
number of metalloproteins is the precise positioning of
hydrogen-bonding networks in fine-tuning the function and
specificity of metalloproteins. For example, it has been shown
that most heme enzymes go through a similar heme?peroxo
intermediate and yet exhibit diverse enzymatic activities, such
as monooxygenase (Cyt P450), heme oxygenase (HO), and
oxidase (CcO). High-resolution crystal structures of both Cyt
P450 and heme oxygenase revealed the presence of water
molecules in the metal-binding sites (Scheme 2).[76, 77] The
number and exact location of the water molecules have been
shown to be critical in forming hydrogen-bonding networks
that can dramatically influence the reactivity. Although the
exact nature of the hydrogen-bonding networks in CcO has
not been revealed, functional studies of a biosynthetic model
in myoglobin support their importance in fine-tuning the
reactivity.[78] This work, complementary to work carried out
using synthetic models,[79, 80] is based on a structural comparison of myoglobin and heme?copper oxidase. Although both
proteins contain heme, CcO contains a copper center that is
approximately 4.7 J from the heme iron center (Figure 8 A).
To make a biosynthetic model of CcO, His ligands were
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
pathways of either reduced heme with O2 or oxidized heme
with H2O2, the biosynthetic model CuBMb is not simply an
HO model. Rather, CuBMb is at a branching point between
HO and CcO (Scheme 3); the extra proton may promote
either the CcO or the P 450 reaction, but not the HO reaction.
Therefore, such a biosynthetic model provides a unique
opportunity to investigate which properties direct the enzyme
activity.
Scheme 2. Proposed role of hydrogen bonds in HO and cyt P450.
Adapted from reference [76].
Scheme 3. Hydrogen bonds and protons influence the reaction path of
the common intermediate in heme enzymes. Adapted from reference [78].
Figure 8. A) Overlay of crystal structures of CcO (thick lines) with
sperm whale Mb (thin lines). Mb lacks the CuB-binding site since it
has one His, one Leu, and one Phe residue instead of three histidine
residues as in CcO; B) Overlay of the crystal structure of CcO (thick
lines) with a computer model of CuBMb (Leu 28 His/Phe 43 Phe, thin
lines).
introduced into the distal side of sperm whale Mb at the
corresponding locations (Figure 8 B).[81]
Spectroscopic studies indicate that the Mb model
(CuBMb) is very similar to CcO. Studies of this biosynthetic
model also indicated that the presence of a copper ion has a
critical influence on the redox potential of the heme Fe
center[82, 83] and can transform the oxygen-binding protein Mb
into an oxygen-activating enzyme.[78, 81] Interestingly, the
protein model in Mb generates verdoheme, a key intermediate in HO, rather than producing a ferryl?heme as in CcO.[78]
Control experiments indicated that protein instability or
altered dynamics of the binding site were not the cause for
this reaction, since replacement of CuII or CuI by redoxinactive ions such as ZnII and AgI, respectively, does not
promote the HO reaction.[78] Since all reactions were carried
out in the presence of catalase, the possibility of exogenous
peroxide as the cause for verdoheme formation was also
eliminated. More importantly, reaction of CuBMb with H2O2
(equivalent in oxidation state to 2e-reduced O2 but possessing two extra protons) produces a ferryl species similar to that
in CcO. Since HO results in verdoheme regardless of the
presence of metal ions in the distal site, or through reaction
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
3. Biosynthetic Inorganic Chemistry for the
Synthesis of New Inorganic Complexes
Biosynthetic inorganic chemistry has been quite effective
for making biomimetic models of complex metalloproteins
and for revealing new insights into the structural features
responsible for protein function. Its major advantage, however, is that it combines the benefits of inorganic chemistry
and biology to produce new inorganic complexes with
unprecedented structure and function.[84]
Natural metalloproteins normally consist of 20 amino
acids, less than half of which can bind as ligands to metal ions.
Furthermore, they employ only few metal ions and metalcontaining prosthetic groups (e.g., cobalamins and heme
groups). The use of nonnatural amino acids in metalloproteins
will dramatically increase the number of possible ligand
donor sets, and the introduction of inorganic compounds into
proteins will lead to new and complex active sites. The net
result will be a new, considerably expanded enzymology,
which can draw from a large reservoir of artificial metalloproteins for various reaction types and rates.
Moreover, such endeavors can also result in new inorganic
compounds with novel structure and activity. Especially
important in this respect is the adaptation of inorganic
catalysts for asymmetric reactions and for chemical transformations in water. Encapsulating these inorganic catalysts
with proteins can fulfill both purposes, as the natural chiral
environment of a protein can provide chiral discrimination.
The hydrophobic interior of proteins provides a suitable
surrounding for inorganic complexes that are active only in
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5595
Reviews
Y. Lu
organic solvent, while the hydrophilic exterior can increase
the water solubility. These requirements can all be accomplished with minimal modification of the ligands and thus
minimal effect on the catalyst reactivity.
3.1. Introducing Nonnatural Amino Acids into Metalloproteins
The properties of metalloproteins can be fine-tuned by
introducing nonnatural amino acids into the metal-binding
sites. For example, pyridine is a common ligand in synthetic
models of metal-binding sites containing imidazole histidines
in proteins.[10] The histidine ligands have been replaced by
pyridine ligands in a newly designed heme protein with four
a-helix units. The replacement of the two His ligands with two
pyridyl ligands decreased the protein-binding affinity of ferric
heme approximately 60 000-fold and increased the reduction
potential of the heme by 287 mV.[85] These differences should
be taken into consideration when using pyridines to model
histidines in metalloproteins.
Perhaps the biggest advantage of using nonnatural amino
acids in the metal-binding sites of proteins is the isostructural
replacement of ligands or residues around the secondary
coordination sphere, by which the electronic properties can
be changed through substitution of a single atom or group.
For example, both Cys and Met in blue copper protein
proteins are conserved (Figure 9). Replacing the cysteine with
Figure 9. Isostructural replacement of Cys and Met by nonnatural
amino acids.
selenocysteine in azurin resulted in marked changes in the
UV/Vis and EPR spectra (50-nm red shift of the visible
charge transfer band and twofold increase in the hyperfine
coupling constant),[86, 87] with little effect on the reduction
potential of the copper center.[86] In contrast, replacing the
methionine with selenomethionine or norleucine resulted in
little change in the UV/Vis and EPR spectra, but a dramatic
increase in the reduction potentials (25 and 140 mV, respectively, over the native protein).[88]
In addition to ligands in the primary coordination sphere,
residues in the secondary coordination sphere have also been
replaced with nonnatural amino acids. For example, replacement of the OH group at the para position of a conserved Tyr
unit near one of the Cys ligands of a rubredoxin with H, F,
NO2, and CN groups resulted in an increase in reduction
potentials of the iron?sulfur center, with electron-withdrawing groups leading to more positive potentials (Fig-
5596
www.angewandte.org
ure 10 A).[89] Furthermore, a conserved hydrogen bond
between the backbone amide groups and the cysteine ligand
of the Fe4S4 clusters was found in high-potential iron?sulfur
proteins (HiPIPs). Replacement of the backbone amide
groups with an ester linkage eliminated these hydrogen
bonds and lowered the reduction potential of the iron center
by approximately 100 mV (Figure 10 B).[90]
Figure 10. Using nonnatural amino acids to probe the role of Tyr in
the secondary coordination sphere in rubredoxin (A) and the backbone
carbonyl oxygen atom in HiPIP (B).
Isostructural replacement of ligands has been a common
practice in synthetic inorganic chemistry. Introducing nonnatural amino acids into metalloproteins finally allows
biosynthetic inorganic chemistry to achieve a similar level
of sophistication. Since proteins can provide a rigid network
for the metal-binding sites, isostructural replacement often
allows the introduction of ligands with little or no ability to
coordinate to metal ions (e.g., replacement of Met with Ile)
without affecting the overall characteristics of the metalbinding site. It also makes it easier to probe noncovalent
interactions in the secondary coordination sphere (e.g.,
through replacement of a peptide bond with an ester linkage[90]). This approach can reveal the role of specific residues
to an unprecedented level; it can be used to fine-tune
coordination complexes. For example, the use of isostructural
nonnatural amino acids at the Met position allowed deconvolution of different factors that influence the reduction
potential of the blue copper azurin through this axial
ligand.[88] A linear relationship between the reduction potential and the hydrophobicity of the side chains on the axial
ligand pointed to hydrophobicity as the dominant factor in
controlling the reduction potential[88] and could be helpful in
developing similar compounds with predicted reduction
potentials.
3.2. Introducing Nonnative Metal Cofactors into Proteins
The introduction of nonnative metal cofactors into
proteins exemplifies the combination of the benefits of
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
biology and inorganic chemistry in biosynthetic inorganic
chemistry. Three approaches have been applied towards this
goal: noncovalent, single-covalent, and dual-covalent attachment.
The majority of native metal cofactors or prosthetic
groups are incorporated into proteins through noncovalent
interactions such as hydrogen bonding, as well as electrostatic
and hydrophobic interactions. Because proteins have not
been evolved to bind nonnative metal cofactors specifically,
the introduction of these cofactors is a challenging task. The
most successful approach thus far involves the replacement of
the heme unit in heme proteins such as myoglobin with
modified metalloporphyrins.[91?99] Since the modified metalloporphyrins are similar in structure to native heme groups, it
is possible to confer novel reactivities to the protein through
the introduction of new functional groups without severely
changing the structure of the metal-binding site. For example,
modifications to the two propionate groups in heme resulted
in new functions such as protein?protein and protein?
molecule recognition as well as electron-transfer properties
and enhanced chemical reactivity (Figure 11 A),[99] which
manifests itself for example in an enhanced P450-like
dioxygen activation upon attachment of a flavin to one of
the propionate groups.[102] Furthermore, when the heme is
replaced with iron porphycene,[103] enhanced O2-binding
affinity is observed.
Many nonnative metal complexes do not resemble native
metal cofactors. Conjugates of these metal complexes with
biotin, however, can take advantage of the strong and specific
binding between biotin and the protein avidin. In this way,
biotinylated dirhodium(I) complexes have been introduced
into avidin or the closely related protein streptavidin to
produce novel catalysts for asymmetric hydrogenation with
up to 96 % enantiomeric excess (ee, Figure 11 B).[100, 104?111]
Another way to incorporate nonnative complexes is the
design of metalloproteins based on the crystal structures of
proteins and metal complexes. After careful inspection of the
structures of myoglobin and MIII(salophen) complexes (M =
Cr, Mn), modifications of both the protein and the metal
complex gave a new protein that exhibited asymmetric
sulfoxidation activity (Figure 11 C; salophen = N,N?-bis(salicylidene)-1,2-phenylenediamine dianion).[101, 112?114] Finally, a
mimetic protein with four helices has been designed de novo
that selectively binds a metalloporphyrin.[115] This example
marks the beginning of computationally developed, tailormade metalloproteins with nonnative metal complexes.
Noncovalent attachment is not the only way in which
native metal cofactors can be incorporated into proteins.
There are also covalent attachment strategies, in which
attachment is through either a single point (some protozoan
mitochondrial cytochrome c[116]) or two points (most cytochrome c[117]). These strategies have also been applied to
incorporate nonnative metal complexes.
For example, the introduction of 1,10-phenanthroline
copper[120] or iron?edta[121] into proteins by single-point
covalent attachment gave a sequence-specific nuclease
(edta = ethylenediaminetetraacetate). Covalent attachment
of a copper(II) 1,10-phenanthroline complex to a single
cysteine residue in an adipocyte lipid-binding protein gave a
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Figure 11. Introducing nonnative metal-containing cofactors into proteins through noncovalent bonds. A) Replacing heme in Mb with a
modified heme (adapted from reference [99]); B) Introducing organometallic complexes by taking advantage of strong noncovalent biotin
and avidin interactions (adapted from reference [100]); C) Replacing
heme in Mb with an Mn Schiff base complex (adapted from reference [113]).
catalyst that promotes highly enantioselective hydrolysis
(with up to 86 % ee) (Figure 12 A).[118, 122] This strategy was
also used to attach a ferrocene derivative covalently to azurin
(Figure 12 B). This novel organometalloprotein increased the
solubility of ferrocene and improved the stability of ferrocenium in water.[119] The secondary coordination sphere of the
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5597
Reviews
Y. Lu
Figure 13. Introducing nonnative metal-containing cofactors into proteins through dual-point covalent attachment. A computer model of
Mb(L72C/Y103C) with a MnIIIsalen complex covalently attached at two
points and overlaid with heme in myoglobin (adapted from reference [124]).
4. Conclusions and Outlook
Figure 12. Single-point covalent attachment of nonnative metal-containing cofactors to proteins. A) A computer model of an adipocyte
lipid-binding protein?phenanthroline complex (ALBP-Phen) (adapted
from reference [118]); B) A computer model of an azurin?ferrocene
organometalloprotein.[119]
protein has also been used to fine-tune the reduction potential
of ferrocene.[119]
Although single-point attachment allows the binding of
nonnative metal complexes to proteins with minimum
structural modifications, the conformational freedom of the
complexes inside the protein may not be restricted enough to
perform highly stereo- and enantioselective transformations.
For example, an Mn(salen) complex coupled by single-point
covalent attachment to a protein gave less than 10 % ee
(salen = N,N?-bis(salicylidene)ethylenediamine dianion).[123]
When the complex is attached in a specific location of Mb
by a dual-point covalent attachment, the enantioselectivity
increased to 51 % ee (Figure 13).[124] Clearly, such enantioselectivity is not yet useful for synthetic transformations;
however, it is understandable given the fact that myoglobin
has not evolved like other heme enzymes (e.g., Cyt P450) for
asymmetric catalysis, and that the substrate binding site has
not been modified to confer substrate selectivity. The
principle demonstrated here is nonetheless important in
guiding future research. For example, a combination of
these covalent attachment strategies[123, 124] with structurebased design or with directed evolution methods[125, 126] will
result in the next generation of asymmetric catalysts with
even higher efficiencies and selectivities.
5598
www.angewandte.org
Biosynthetic inorganic chemistry is a natural extension of
synthetic inorganic chemistry. Instead of using small organic
molecules, small, stable, easy-to-make, and well-characterized
proteins such as azurin and myoglobin are used as ?ligands?
for the synthesis of either biomimetic models or new
compounds. The chemical and biological approaches each
have advantages and disadvantages in terms of synthesis,
characterization, and properties (Table 2).
Synthesis. A synthetic approach is generally faster and
produces higher yields. However, recent developments in
molecular biology and protein enzymology have enabled the
routine production of hundreds of milligrams to grams of
protein in the laboratory and even higher yields with
industrial methods. By carrying out reactions with cells
containing the desirable protein models, the power of natural
amplification of genes and protein products can be exploited
for biosynthesis and biocatalysis. One advantage of biosynthesis over small molecule synthesis, in which modification of
the ligand might result in significant variations in yield and
have associated cost and time investments, is that a similar
modification of a protein ligand has a smaller influence on
these parameters. These features make the speed and yield of
biosynthesis closer to that of small molecule synthesis. In a
few selected cases, it is even easier to synthesize biosynthetic
models than some of the more complex synthetic models. For
example, synthetic models of heme?copper oxidases require
multistep syntheses that give relatively low yields.[79, 80] The
biosynthesis of heme?copper models in myoglobin[81] or
cytochrome c peroxidase[127] proceeds with similar yields and
preparation times as biosynthesis of other derivatives of the
same proteins. Regardless of the complexity of the modifications, approximately 100 mg L1 of pure biosynthetic models
can be obtained in less than a week. The routine production of
gram quantities of protein is thus possible. Although the
biosynthetic approach has been limited to a few cases in
industry,[13, 14] the gap between the two approaches is narrowing.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
Applications. Biosynthesis is often carried out in air at
ambient temperatures and pressures rather than in inert
atmospheres at extreme temperatures and pressures, as are
required for many syntheses of small molecules. Therefore,
biosynthesis could help to save energy and equipment costs.
Furthermore, biosynthesis is an environmentally benign
approach, since biodegradable materials are utilized and
water is the solvent. However, the products from biosyntheses
are not as robust as synthetic compounds, which will make the
long-term industrial use at high temperatures or pressures
difficult. This problem can be overcome by using protein
ligands that have been isolated from extremophilic organisms
(which live at high temperatures or in organic solvents such as
methanol).
A common criterion to evaluate the practical utility of
catalysts is the mass ratio between the product and the
catalyst. This is a good measure only if the catalysts are of
similar type and are synthesized by similar methods. In this
case, higher mass catalysts are generally more costly to
synthesize and to dispose of, and therefore smaller compounds would be preferable to a chemically synthesized
protein ligand of similar catalytic efficiency. However,
biologically synthesized protein ligands can almost always
be made at a much lower cost than their molecular weight
would suggest, and disposing a biocompatible and biodegradable protein ligand is also much less costly than disposing a
similarly sized organic ligand, which is often not biocompatible or biodegradable. In fact, the cost of synthesizing and
disposing of a small protein (e.g., MW = 10 000) by using
recombinant protein expression systems is very similar to that
of a much larger protein (e.g., MW = 100 000). Therefore,
multiple criteria are needed to evaluate different types of
catalysts,[128] for example, mass ratio, catalyst accessibility and
costs of synthesis and disposal, substrate scope, and, if
applicable, enantioselectivity.
Characterization. Most metal-based spectroscopic techniques, such as electronic spectroscopy (UV/Vis), electron
paramagnetic resonance (EPR), magnetic circular dichroism
(MCD), resonance Raman (RR), and X-ray absorption
spectroscopy (XAS), can be performed on both synthetic
and biosynthetic compounds without much difference in data
collection and interpretation. Even though recent developments in structural biology have made it possible to obtain
three-dimensional structures of biosynthetic compounds
routinely by using X-ray crystallography or NMR spectroscopy, it is still much easier to obtain structures of synthetic
compounds with higher resolution. To overcome these
limitations, one can use proteins with known 3D structures
as ligands. In this way, it is easier to grow diffraction-quality
crystals, and 3D structures can be obtained faster by comparing differences in the electron density map. The use of
synchrotron sources to obtain diffraction data may also help
to improve resolution.
Properties. Proteins provide a rigid network that helps to
stabilize metal-binding or substrate-binding sites as a result of
covalent bonds and noncovalent interactions (e.g., hydrogen
bonds, electrostatic and hydrophobic forces) in the primary
and secondary coordination spheres. They provide a better
environment for regio- and stereoselective binding and
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
catalysis, often without requiring protection of reactive
functional groups. Because they consist of chiral natural
amino acids, proteins are intrinsically enantioselective. Biosynthetic compounds with proteins can benefit from these
features. Indeed, a number of biophysical studies show that
biosynthetic models, such as a biosynthetic CuA center in
azurin (Figure 2 B),[46] are as rigid as the native template
proteins. Moreover, most synthetic modeling results in either
structural or functional models of target native proteins.
Because biosynthetic models use the same type of ligands and
are synthesized and characterized under the same conditions
as target proteins, they are often structural and functional
models at the same time. Recent advances in synthetic
chemistry have made it possible to design new organic ligands
that provide a rigid network similar to that in proteins,
including hydrogen-bonding interactions in the secondary
coordination sphere.[11, 12] Synthetic asymmetric catalysts have
a wider range of activities and a broader substrate scope.[129]
The wider range of activities arises mainly from the extensive
choice of ligands, metals, and metal-containing cofactors. The
broader substrate scope is attributable to a wide range of
accessible reaction conditions, such as low temperatures, high
pressures, and different organic solvents. Biosynthetic catalysts, however, are mostly restricted to only the 20 natural
amino acids, physiologically available metals or metal-containing cofactors, physiological conditions, and water as
solvent. These limitations make biocatalysts even more
remarkable, and make it even more imperative to learn
methods from nature to make better catalysts. The introduction of nonnatural amino acids and nonnative metal-containing cofactors into proteins is an important step in this
direction.
In summary, recent advances in a number of areas in
biology and chemistry have enabled the development of
biosynthetic inorganic methods with comparable preparation
times and yields to those of synthetic inorganic chemistry. In
many cases, biosynthetic inorganic chemistry can produce
close structural and functional models of more-complex
metalloproteins and novel inorganic compounds. These
developments, however, only demonstrate the potential that
the biosynthetic approach has to offer; this potential is far
from being realized. Although methodologies have been
developed to introduce nonnative metal cofactors into
proteins by noncovalent, single-, and double-covalent attachment, the use of these approaches in asymmetrically catalyzed
syntheses of chiral intermediates for fine chemicals or
pharmaceutical products is still lacking. In contrast to the
field of asymmetric catalysis, which has had many years of
development by a number of researchers, the field of
biosynthetic inorganic chemistry is still in its infancy. Much
more time and effort, including cross-fertilization of the
synthetic and biosynthetic approaches, will be required before
the full potential of biosynthetic inorganic chemistry can be
realized.
I wish to thank the members of my research group whose work
has been cited in the references for their dedication and hard
work, as well as Hee Jung Hwang, Thomas D. Pfister, Dewain
Garner, and Natasha Yeung for help with the preparation of
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5599
Reviews
Y. Lu
figures in this review. The work of my group described in this
review has been supported by the US National Science
Foundation and the National Institutes of Health.
Received: January 16, 2006
Published online: August 10, 2006
[1] S. J. Lippard, J. M. Berg, Principles of Bioinorganic Chemistry,
University Science Books, Mill Valley, CA, 1994.
[2] I. Bertini, H. B. Gray, S. J. Lippard, J. S. Valentine, Bioinorganic
Chemistry, University Science Books, Sausalito, CA, 1994.
[3] J. J. R. Fra?sto da Silva, R. J. P. Williams, The Biological
Chemistry of the Elements, the Inorganic Chemistry of Life,
Clarendon Press, Oxford, 1991.
[4] J. A. Cowan, Inorganic Biochemistry, an Introduction, VCH,
New York, 1993.
[5] W. Kaim, B. Schwederski, Bioinorganic Chemistry: Inorganic
Elements in the Chemistry of Life, Wiley, Chichester, 1994.
[6] R. H. Holm, E. I. Solomon, Chem. Rev. 1996, 96, 2237.
[7] Ullmann6s Encyclopedia of Industrial Chemistry, 6th ed., 1998
Electronic Release, 1998.
[8] Riegel6s Handbook of Industrial Chemistry, 10th ed., 2003.
[9] A. Fersht, Structure and Mechanism in Protein Science: A Guide
to Enzyme Catalysis and Protein Folding, Freeman, New York,
NY, 1999.
[10] R. H. Holm, E. I. Solomon, Chem. Rev. 2004, 104, 347.
[11] R. I. Cukier, D. G. Nocera, Annu. Rev. Phys. Chem. 1998, 49,
337.
[12] A. S. Borovik, Acc. Chem. Res. 2005, 38, 54.
[13] B. A. Anderson, M. M. Hansen, A. R. Harkness, C. L. Henry,
J. T. Vicenzi, M. J. Zmijewski, J. Am. Chem. Soc. 1995, 117,
12 358.
[14] B. A. Anderson, M. M. Hansen, J. T. Vicenzi, M. J. Zmijewski,
Process Chem. Pharm. Ind. 1999, 263.
[15] Inorganic Electronic Structure and Spectroscopy, Vol. I (Eds.:
E. I. Solomon, A. B. P. Lever), Wiley-Interscience, New York,
1999.
[16] Inorganic Electronic Structure and Spectroscopy, Vol. II (Eds.:
E. I. Solomon, A. B. P. Lever), Wiley-Interscience, New York,
1999.
[17] Physical Methods in Bioinorganic Chemistry: Spectroscopy and
Magnetism (Ed.: L. Que, Jr.), University Science Books, Mill
Valley, CA, 2000.
[18] C. Dennison, G. W. Canters, Recl. Trav. Chim. Pays-Bas 1996,
115, 345.
[19] H. Beinert, Eur. J. Biochem. 1997, 245, 521.
[20] A. J. Vila, C. O. FernRndez in Handbook on Metalloproteins
(Eds.: I. Bertini, A. Sigel, H. Sigel), Dekker, New York, NY,
2001, p. 813.
[21] Y. Lu in Biocoordination Chemistry, Vol. 8 (Eds.: L. Que, Jr.,
W. B. Tolman), Elsevier, Oxford, UK, 2004, pp. 91.
[22] M. Saraste, Q. Rev. Biophys. 1990, 23, 331.
[23] B. G. MalmstrSm, Chem. Rev. 1990, 90, 1247.
[24] G. T. Babcock, M. Wikstrom, Nature 1992, 356, 301.
[25] S. M. Musser, M. H. B. Stowell, S. I. Chan, FEBS Lett. 1993,
327, 131.
[26] J. A. GarcTa-Horsman, B. Barquera, J. Rumbley, J. Ma, R. B.
Gennis, J. Bacteriol. 1994, 176, 5587.
[27] P. M. H. Kroneck, J. Riester, W. G. Zumft, W. E. Antholine,
Biol. Met. 1990, 3, 103.
[28] W. G. Zumft, P. M. H. Kroneck, Adv. Inorg. Biochem. 1996, 11,
193.
[29] N. J. Blackburn, M. E. Barr, W. H. Woodruff, J. van der Oost, S.
de Vries, Biochemistry 1994, 33, 10 401.
[30] F. A. Cotton, R. A. Walton, Multiple Bonds Between Metal
Atoms, Wiley, New York, 1982.
5600
www.angewandte.org
[31] B. J. Hathaway in Comprehensive Coordination Chemistry, the
Synthesis, Reactions, Properties & Application of Coordination
Compounds, Vol. 5, 1st ed. (Eds.: G. Wilkinson, R. D. Gillard,
J. A. McCleverty), Pergamon, Oxford, England, 1987, p. 533.
[32] M. Dunaj-Jurco, G. Ondrejovic, M. Melnik, J. Garaj, Coord.
Chem. Rev. 1988, 83, 1.
[33] C. Harding, V. McKee, J. Nelson, J. Am. Chem. Soc. 1991, 113,
9684.
[34] M. E. Barr, P. H. Smith, W. E. Antholine, B. Spencer, J. Chem.
Soc. Chem. Commun. 1993, 1649.
[35] C. Harding, J. Nelson, M. C. R. Symons, J. Wyatt, J. Chem. Soc.
Chem. Commun. 1994, 2499.
[36] R. P. Houser, V. G. Young, Jr., W. B. Tolman, J. Am. Chem. Soc.
1996, 118, 2101.
[37] D. D. LeCloux, R. Davydov, S. J. Lippard, Inorg. Chem. 1998,
37, 6814.
[38] C. He, S. J. Lippard, Inorg. Chem. 2000, 39, 5225.
[39] R. Gupta, Z. H. Zhang, D. Powell, M. P. Hendrich, A. S.
Borovik, Inorg. Chem. 2002, 41, 5100.
[40] X.-M. Zhang, M.-L. Tong, X.-M. Chen, Angew. Chem. 2002,
114, 1071; Angew. Chem. Int. Ed. 2002, 41, 1029.
[41] D. D. LeCloux, R. Davydov, S. J. Lippard, J. Am. Chem. Soc.
1998, 120, 6810.
[42] S. B. Harkins, J. C. Peters, J. Am. Chem. Soc. 2004, 126, 2885.
[43] C. Dennison, E. Vijgenboom, S. de Vries, J. van der Oost, G. W.
Canters, FEBS Lett. 1995, 365, 92.
[44] M. Hay, J. H. Richards, Y. Lu, Proc. Natl. Acad. Sci. USA 1996,
93, 461.
[45] G. J. Steffens, G. Buse, Hoppe-Seyler6s Z. Physiol. Chem. 1979,
360, 613.
[46] H. Robinson, M. C. Ang, Y.-G. Gao, M. T. Hay, Y. Lu, A. H. J.
Wang, Biochemistry 1999, 38, 5677.
[47] X. Wang, M. C. Ang, Y. Lu, J. Am. Chem. Soc. 1999, 121, 2947.
[48] M. T. Hay, R. M. Milberg, Y. Lu, J. Am. Chem. Soc. 1996, 118,
11 976.
[49] M. T. Hay, Y. Lu, J. Biol. Inorg. Chem. 2000, 5, 699.
[50] H. J. Hwang, N. Nagraj, Y. Lu, Inorg. Chem. 2006, 45, 102 ? 107.
[51] X. Wang, S. M. Berry, Y. Xia, Y. Lu, J. Am. Chem. Soc. 1999,
121, 7449.
[52] S. M. Berry, X. Wang, Y. Lu, J. Inorg. Biochem. 2000, 78, 89.
[53] H. J. Hwang, S. M. Berry, M. J. Nilges, Y. Lu, J. Am. Chem. Soc.
2005, 127, 7274.
[54] H. J. Hwang, Y. Lu, Proc. Natl. Acad. Sci. USA 2004, 101,
12 842.
[55] O. Farver, Y. Lu, M. C. Ang, I. Pecht, Proc. Natl. Acad. Sci.
USA 1999, 96, 899.
[56] H. B. Gray, B. G. MalmstrSm, R. J. P. Williams, J. Biol. Inorg.
Chem. 2000, 5, 551.
[57] E. I. Solomon, D. W. Randall, T. Glaser, Coord. Chem. Rev.
2000, 200?202, 595.
[58] B. E. Ramirez, B. G. MalmstrSm, J. R. Winkler, H. B. Gray,
Proc. Natl. Acad. Sci. USA 1995, 92, 11 949.
[59] Y. Lu, S. M. Berry, T. D. Pfister, Chem. Rev. 2001, 101, 3047.
[60] A. Gengenbach, X. Wang, Y. Lu in Fundamentals and Catalysis
of Oxidative Delignification Processes, Vol. 785 (Ed.: D. S.
Argyropoulos), American Chemical Society, Washington, DC,
2001, p. 487.
[61] ?Manganese and Its Role in Biological Processes?: M. H. Gold,
H. L. Youngs, M. D. S. Gelpke, Met. Ions Biol. Syst. 2000, 37,
559.
[62] M. Sundaramoorthy, K. Kishi, M. H. Gold, T. L. Poulos, J. Biol.
Chem. 1994, 269, 32 759.
[63] B. K. Yeung, X. Wang, J. A. Sigman, P. A. Petillo, Y. Lu, Chem.
Biol. 1997, 4, 215.
[64] S. K. Wilcox, C. D. Putnam, M. Sastry, J. Blankenship, W. J.
Chazin, D. E. McRee, D. B. Goodin, Biochemistry 1998, 37,
16 853.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
Angewandte
Chemie
Biomimetic Synthesis
[65] X. Wang, Y. Lu, Biochemistry 1999, 38, 9146.
[66] A. Gengenbach, S. Syn, X. Wang, Y. Lu, Biochemistry 1999, 38,
11 425.
[67] T. D. Pfister, A. J. Gengenbach, S. Syn, Y. Lu, Biochemistry
2001, 40, 14 942.
[68] P. R. O. de Montellano, Cytochrome P450 Structure, Mechanism, and Biochemistry, 2nd ed., Plenum, New York, 1995.
[69] S. Adachi, S. Nagano, Y. Watanabe, K. Ishimori, I. Morishima,
Biochem. Biophys. Res. Commun. 1991, 180, 138.
[70] S. Adachi, S. Nagano, K. Ishimori, Y. Watanabe, I. Morishima,
T. Egawa, T. Kitagawa, R. Makino, Biochemistry 1993, 32, 241.
[71] D. P. Hildebrand, J. C. Ferrer, H.-L. Tang, M. Smith, A. G.
Mauk, Biochemistry 1995, 34, 11 598.
[72] K. Choudhury, M. Sundaramoorthy, A. Hickman, T. Yonetani,
E. Woehl, M. F. Dunn, T. L. Poulos, J. Biol. Chem. 1994, 269,
20 239.
[73] J. A. Sigman, A. E. Pond, J. H. Dawson, Y. Lu, Biochemistry
1999, 38, 11 122.
[74] R. Perera, M. Sono, J. A. Sigman, T. D. Pfister, Y. Lu, J. H.
Dawson, Proc. Natl. Acad. Sci. USA 2003, 100, 3641.
[75] T. Uno, A. Yukinari, Y. Tomisugi, Y. Ishikawa, R. Makino, J. A.
Brannigan, A. J. Wilkinson, J. Am. Chem. Soc. 2001, 123, 2458.
[76] H. Fujii, X. Zhang, T. Tomita, M. Ikeda-Saito, T. Yoshida, J.
Am. Chem. Soc. 2001, 123, 6475.
[77] T. Matsui, M. Furukawa, M. Unno, T. Tomita, M. Ikeda-Saito, J.
Biol. Chem. 2005, 280, 2981.
[78] J. A. Sigman, H. K. Kim, X. Zhao, J. R. Carey, Y. Lu, Proc. Natl.
Acad. Sci. USA 2003, 100, 3629.
[79] J. P. Collman, R. Boulatov, C. J. Sunderland, L. Fu, Chem. Rev.
2004, 104, 561.
[80] E. Kim, E. E. Chufan, K. Kamaraj, K. D. Karlin, Chem. Rev.
2004, 104, 1077.
[81] J. A. Sigman, B. C. Kwok, Y. Lu, J. Am. Chem. Soc. 2000, 122,
8192.
[82] X. Zhao, N. Yeung, Z. Wang, Z. Guo, Y. Lu, Biochemistry 2005,
44, 1210.
[83] X. Zhao, M. J. Nilges, Y. Lu, Biochemistry 2005, 44, 6559.
[84] Y. Lu, Curr. Opin. Chem. Biol. 2005, 9, 118.
[85] H. K. Privett, C. J. Reedy, M. L. Kennedy, B. R. Gibney, J. Am.
Chem. Soc. 2002, 124, 6828.
[86] S. M. Berry, M. D. Gieselman, M. J. Nilges, W. A. van der
Donk, Y. Lu, J. Am. Chem. Soc. 2002, 124, 2084.
[87] M. Ralle, S. M. Berry, M. J. Nilges, M. D. Gieselman, W. A.
van der Donk, Y. Lu, N. J. Blackburn, J. Am. Chem. Soc. 2004,
126, 7244.
[88] S. M. Berry, M. Ralle, D. W. Low, N. J. Blackburn, Y. Lu, J. Am.
Chem. Soc. 2003, 125, 8760.
[89] D. W. Low, M. G. Hill, J. Am. Chem. Soc. 1998, 120, 11 536.
[90] D. W. Low, M. G. Hill, J. Am. Chem. Soc. 2000, 122, 11 039.
[91] T. Hayashi, A. Tomokuni, T. Mizutani, Y. Hisaeda, H. Ogoshi,
Chem. Lett. 1998, 1229.
[92] T. Hayashi, Y. Hitomi, T. Ando, T. Mizutani, Y. Hisaeda, S.
Kitagawa, H. Ogoshi, J. Am. Chem. Soc. 1999, 121, 7747.
[93] T. Hayashi, T. Matsuda, Y. Hisaeda, Chem. Lett. 2003, 32, 496.
[94] H. Sato, T. Hayashi, T. Ando, Y. Hisaeda, T. Ueno, Y.
Watanabe, J. Am. Chem. Soc. 2004, 126, 436.
[95] T. Hayashi, T. Nakagawa, K. Harada, T. Matsuo, Y. Hitomi, Y.
Hisaeda, Chem. Lett. 2004, 33, 1512.
[96] H. Sato, M. Watanabe, Y. Hisaeda, T. Hayashi, J. Am. Chem.
Soc. 2005, 127, 56.
[97] T. Hayashi, Y. Hitomi, T. Takimura, A. Tomokuni, T. Mizutani,
Y. Hisaeda, H. Ogoshi, Coord. Chem. Rev. 1999, 190?192, 961.
Angew. Chem. Int. Ed. 2006, 45, 5588 ? 5601
[98] I. Hamachi, S. Shinkai, Eur. J. Org. Chem. 1999, 539.
[99] T. Hayashi, Y. Hisaeda, Acc. Chem. Res. 2002, 35, 35.
[100] M. Skander, N. Humbert, J. Collot, J. Gradinaru, G. Klein, A.
Loosli, J. Sauser, A. Zocchi, F. Gilardoni, T. R. Ward, J. Am.
Chem. Soc. 2004, 126, 14 411.
[101] T. Ueno, M. Ohashi, M. Kono, K. Kondo, A. Suzuki, T. Yamane,
Y. Watanabe, Inorg. Chem. 2004, 43, 2852.
[102] T. Matsuo, T. Hayashi, Y. Hisaeda, J. Am. Chem. Soc. 2002, 124,
11 234.
[103] T. Hayashi, H. Dejima, T. Matsuo, H. Sato, D. Murata, Y.
Hisaeda, J. Am. Chem. Soc. 2002, 124, 11 226.
[104] M. E. Wilson, G. M. Whitesides, J. Am. Chem. Soc. 1978, 100,
306.
[105] C.-C. Lin, C.-W. Lin, A. S. C. Chan, Tetrahedron: Asymmetry
1999, 10, 1887.
[106] J. Collot, J. Gradinaru, N. Humbert, M. Skander, A. Zocchi,
T. R. Ward, J. Am. Chem. Soc. 2003, 125, 9030.
[107] C. M. Thomas, T. R. Ward, Chem. Soc. Rev. 2005, 34, 337.
[108] G. Klein, N. Humbert, J. Gradinaru, A. Ivanova, F. Gilardoni,
U. E. Rusbandi, T. R. Ward, Angew. Chem. Int. Ed. 2005, 44,
7764.
[109] M. Skander, C. Malan, A. Ivanova, T. R. Ward, Chem.
Commun. 2005, 4815.
[110] T. R. Ward, Chem. Eur. J. 2005, 11, 3798.
[111] C. Letondor, N. Humbert, T. R. Ward, Proc. Natl. Acad. Sci.
USA 2005, 102, 4683.
[112] M. Ohashi, T. Koshiyama, T. Ueno, M. Yanase, H. Fujii, Y.
Watanabe, Angew. Chem. 2003, 115, 1035; Angew. Chem. Int.
Ed. 2003, 42, 1005.
[113] T. Ueno, T. Koshiyama, M. Ohashi, K. Kondo, M. Kono, A.
Suzuki, T. Yamane, Y. Watanabe, J. Am. Chem. Soc. 2005, 127,
6556.
[114] Y. Watanabe, T. Hayashi, Prog. Inorg. Chem. 2005, 54, 449.
[115] F. V. Cochran, S. P. Wu, W. Wang, V. Nanda, J. G. Saven, M. J.
Therien, W. F. DeGrado, J. Am. Chem. Soc. 2005, 127, 1346.
[116] G. W. Pettigrew, J. L. Leaver, T. E. 
Документ
Категория
Без категории
Просмотров
1
Размер файла
1 274 Кб
Теги
chemistry, inorganic, biosynthetical
1/--страниц
Пожаловаться на содержимое документа