INSIGHTS | PERSPECTIVES
By Hendrik F. T. Klare and
Silicon is found in nature in many in- organic forms, some of which are con- structed by living organisms. Yet, no known biological molecule contains a carbon–silicon (C–Si) bond, and no bi- ological processes to form C–Si bonds
have been identified. On page 1048 of this issue, Kan et al. (1) show that natural as well
as reengineered enzymes can promote C–Si
bond formation. The resulting chiral compounds mostly consist of one of the two possible stereoisomers (enantiomers).
Methods for C–Si bond formation by
chemical means have matured over the
past decades. A typical reaction for creating asymmetric (chiral) C–Si compounds is
the transition-metal–catalyzed insertion of
diazo compounds 1, which are bench-stable
carbenoid precursors, into the silicon–
hydrogen bond of triorganosilanes 2 (see the
figure). State-of-the-art catalysts are based
on copper (2), iridium (3), and rhodium (4)
coordinated by elaborate chiral ligands or
small peptides (5). Enantioselection is generally high—that is, the reaction products
are almost pure enantiomers.
Bearing this model reaction in mind,
Kan et al. investigated the enzyme cytochrome c from Rhodothermus marinus
(Rma cyt c) as a catalyst for this synthetically useful transformation. This small
protein contains a heme c moiety, which is
an iron(II) porphyrin complex that is covalently attached via sulfide linkages to two
cysteine residues. The central iron(II) atom
is axially coordinated by the nitrogen atom
of a histidine residue and a sulfur atom of
methionine side chain M100, resulting in
octahedral complexation (see the figure).
This structural motif plays a key role in
electron-transfer processes, but no catalytic
functions in living systems have been found
to date. In contrast, other heme-containing
proteins such as cytochrome P450 and myo-
globin variants have been shown to form
metallocarbenoids with selected diazo-based
carbene precursors, thereby functioning as
biocatalysts for several unnatural reactions
(6–10). Building on this precedent, Kan et al.
evaluated these enzymes as promoters for
formation of the non-native C–Si bond. The
desired bond did indeed form, but control
of the stereochemical course of the reaction
The authors next investigated Rma cyt
c. This choice was beneficial in two re-
spects. First, the wild-type protein Rma
cyt c catalyzed the silicon–hydrogen inser-
tion reaction with high enantioselection
[>97% enantiomeric excess (ee)]. Second,
an existing x-ray structure of the enzyme
(11) provided direct insight into its active
site, serving as an excellent starting point
for further improvement by directed evo-
lution. Following this approach, Kan et al.
selected methionine residue M100, which
needs to dissociate to make room for iron–
carbenoid formation, as well as valine V75
and methionine M103, which are in close
proximity to the active iron heme center
(see the figure). They prepared mutants of
the wild-type protein with sequential site-
The enzyme Rma cyt c V75T M100D
M103E (T, threonine; D, aspartic acid; E,
glutamic acid) emerged as the superior mutant, catalyzing the C–Si bond formation
with unprecedented efficiency (>1500 turnovers and >99% ee), even higher than that
of traditional transition-metal catalysts. In
view of the fact that natural enzymes have
evolved to perform with well-defined functions on specific substrates, it is impressive
to see that only a few mutations are needed
to reengineer a natural enzyme to catalyze
a reaction that was previously reserved for
synthetic chemists. Moreover, the novel enzyme accepts a broad range of differently
substituted hydrosilanes 2 and tolerates
various functional groups, including free
alcohols and amines. In all cases, Kan et al.
observed preferential C–Si bond formation,
producing almost all new tetraorganosi-lanes 3 as single enantiomers.
Kan et al. performed their experiments
in vitro with non-natural, tailor-made substrates that are not likely to survive in a
real biological system. Triorganosilanes 2
are nonpolar, hydrocarbon-like compounds
that are likely to be prone to hydrolysis or
oxidation before they can pass through the
cell membrane. Yet, the authors show that
the same reaction also proceeds in vivo—
that is, in whole cells of Escherichia coli expressing the heme protein Rma cyt c V75T
The beauty and value of Kan et al.’s accomplishment lie in the enzyme-promoted
formation of an unnatural bond. This
closes a crucial gap between biological and
chemical catalysis. The impact is unforeseeable, but it seems that we are a big step
closer to potentially facilitating industrially
relevant reactions such as alkene hydrosilylation with biomolecules. j
1. S. B. J. Kan, R. D. Lewis, K. Chen, F. H. Arnold, Science 354,
2. Y.-Z.Zhang,S.-F.Zhu,L.-X. Wang,Q.-L.Zhou, Angew. Chem.
Int. Ed. 47, 8496 (2008).
3. Y. Yasutomi,H.Suematsu, T.Katsuki, J.Am.Chem.Soc.
132, 4510 (2010).
4. D.Chen,D.-X.Zhu,M.-H.Xu, J.Am.Chem.Soc.138,1498
5. R.Sambasivan,Z. T.Ball, J.Am.Chem.Soc.132,9289
6. H.Renata,Z.J. Wang, F.H.Arnold, Angew.Chem.Int.Ed.
54, 3351 (2015).
7. T.K.Hyster, T.R. Ward, Angew.Chem.Int.Ed.55,7344
8. P. S. Coelho, E. M. Brustad, A. Kannan, F. H. Arnold, Science
339, 307 (2013).
9. Z. J. Wang, N. E. Peck, H. Renata, F. H. Arnold, Chem.Sci. 5,
10. V. Tyagi, R. B. Bonn, R. Fasan, Chem. Sci. 6, 2488 (2015).
11. M. Stelter et al., Biochemistry 47, 11953 (2008).
Institut für Chemie, Technische Universität Berlin, 10623
Berlin, Germany. Email: email@example.com
BIOCHEMIS TR Y
Teaching nature the unnatural
A reengineered enzyme catalyzes C−Si bond formation
RO R' N N
N2 R3Si H
1 Diazo compound
3 Carbon–silicon product
bound to heme
Enzyme catalysis of C–Si bonds
Kan et al. show that the enzyme Rma cyt c can
catalyze the formation of C–Si bonds. Altering just a
few amino acids in the enzyme improves its activity
beyond that of transition-metal catalysts.