CHEMCATCHEM
FULL PAPERS
From the results presented in Figure 2, the following conclu-
sions could be drawn:
Variation of the substrate led to the following observations
and trends:
1) The best results in terms of enantioselectivity for any
ligand are obtained with Sav-WT.
1) Conversion of bulky substrate 1a generally led to the high-
est observed enantioselectivities in favor of the S product.
2) Reducing the size of the substituent at position 1 of the di-
hydroisoquinoline moiety from phenyl to methyl resulted
in a considerable decrease in enantioselectivity.
3) Introduction of methoxy substituents at positions 6 and 7
of the dihydroisoquinoline moiety only moderately affected
the selectivity.
2) Mutating single residues at position S112 or K121, or mu-
tating both residues simultaneously, leads to a decrease in
enantioselectivity for all ligands tested. The S112Y mutant led,
in combination with l-LeuNH2, to an inversion of enantioselec-
tivity compared to Sav-WT (63% ee of (S)-2a in Sav-WT and
20% ee of (R)-2a in the presence of S112Y).
3) All mutants at position K121 typically lead to high con-
versions with all amino amide ligands tested. In the presence
of either l-ProNH2 or l-ValNH2, significantly improved conver-
sions compared to Sav-WT were observed with K121N, K121A,
K121H, and K121Y mutants.
4) With 2-substituted 1-pyrrolines, conversions were generally
good to excellent.
5) The highest ee value for 1-pyrrolines was observed for the
bulky substrate carrying a cyclohexyl substituent in combi-
nation with l-ThrNH2. The ee values for the benzyl- and
phenyl-substituted substrates 1e and 1 f, respectively, were
generally low.
4) In the presence of the L124K, L124Y, or L124F mutant,
a marked decrease in enantioselectivity compared with that in
the presence of Sav-WT is observed for all ligands except l-
ProNH2. Conversions were generally good to excellent with the
L124Y or L124F mutant.
5) Good to excellent conversions were observed for all
double mutant–ligand combinations, which were accompanied
by a decrease in enantioselectivity (except l-ProNH2).
Finally, the substrate scope of the artificial imine reductases
was evaluated (Figure 3). For this purpose, six prochiral cyclic
Conclusions
To readily access large chemical diversity, a new artificial metal-
loenzyme design based on the biotin–streptavidin technology
has been presented. Relying on three-legged piano stool com-
plexes and tethering the biotin anchor on the Cp* moiety al-
lowed us to screen various bidentate ligands for the asymmet-
ric reduction of cyclic imines. An initial screening led to the
identification of amino amides as versatile bidentate ligands in
conjunction with the {IrCp*biotin} moiety. Genetic diversity was
introduced by site-directed mutagenesis. Both chemical diver-
sity and genetic diversity were shown to have a significant
effect on the activity and selectivity of the resulting artificial
metalloenzyme.
By taking into account the versatility of the {Cp*MLn} moiety
in catalysis, we reasoned that the strategy disclosed herein will
find wide application for the chemical optimization of artificial
metalloenzymes. Current efforts are aimed at the structural
and kinetic characterization of such hybrid catalysts.
Figure 3. Selection of substrates for the asymmetric transfer hydrogenation
of cyclic imines.
imines were tested in the presence of artificial metalloenzymes
[IrCp*biotin(amino amide)Cl]ꢀSav-WT (amino amide=l-ProNH2,
l-ValNH2, l-LeuNH2, l-IleNH2, l-ThrNH2, GlyNH2). The results of
this screening are summarized in Table 3.
Experimental Section
General method for the asymmetric transfer hydrogenation
Buffer A (100 mL, 0.6m in MOPS in Milli-Q H2O, pH 7.8) was placed
in a polypropylene (PP) tube, followed by the addition of the bio-
tinylated metal complex [MCp*biotinCl2]2 stock solution (3.75 mL,
5.0 mm in DMSO) and the ligand stock solution (3.75 mL, 11 mm in
Milli-Q H2O or DMSO, depending on the ligand). The mixture was
agitated for 30 min at 308C and 600 rpm in a thermo mixer for
precomplexation. The corresponding lyophilized Sav mutant
(1.6 mg) was dissolved in buffer B (100 mL, 0.6m in MOPS, 3.0m in
HCOONa in Milli-Q H2O, pH 7.8). Then, Sav-mixture (100 mL) was
added to the PP tube containing the metal complex and agitation
was continued for 15 min at 308C and 600 rpm to ensure binding
of the biotinylated complex to Sav. Finally, the substrate stock solu-
tion was added (7.5 mL, 1m in DMSO) and the mixture was agitated
at 508C for 18 h. Subsequently, NaOH(aq) (60 mL, 5m solution) was
added to the reaction mixture, followed by the addition of CH2Cl2
Table 3. Screening of different substrates for the [IrCp*biotin(amino ami-
de)Cl]ꢀSav-WT-mediated transfer hydrogenation.
Entry Ligand
Yield (ee)[a] [%]
1a
1b
1c
1d
1e
1 f
1
2
3
4
5
6
l-ProNH2 47 (67) 99 (5) 100 (6) 69 (À32) 60 (À11) 100 (18)
l-ValNH2 70 (63) 100 (17) 100 (17) 84 (À44) 80 (À22) 86 (16)
l-LeuNH2 96 (63) 100 (3) 100 (18) 95 (À45) 96 (À31) 100 (20)
l-IleNH2 89 (65) 100 (20) 100 (43) 83 (À32) 79 (À22) 10 (18)
l-ThrNH2 96 (63) 100 (25) 100 (41) 98 (À57) 95 (À35) 91 (20)
GlyNH2
94 (43) 100 (4) 100 (18) 96 (À46) 94 (À9)
79 (2)
[a] Positive and negative ee values correspond to S and R enantiomers, re-
spectively. The absolute configuration of the amines 2d–f was not deter-
mined. The best results are highlighted in boldface.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1010 – 1014 1013