.
Angewandte
Communications
ceutical intermediate for the preparation of amphetamine,[18a]
(R)-selegiline,[18b] and cathepsin K inhibitors.[18c] No enzyme
could catalyze these hydroxylation reactions with high regio-
and enantioselectivity.[6a,15a,c] Herein, we report the evolution
of P450pyr hydroxylase to turn the terminal selectivity fully
into subterminal selectivity and also gain excellent enantio-
selectivity for the hydroxylation of alkanes 1 and 3.
The evolution studies started with the identification of the
key amino acid residues by docking n-octane (1) onto the X-
ray crystal structure of P450pyr hydroxylase; 22 residues
located either within 6 ꢀ of the substrate or in the substrate-
accessing channel and “big loop” were selected (see Figure S1
in the Supporting Information). Iterative saturation muta-
genesis (ISM) was applied for the evolution.[5c] P450pyr
variants at each selected site were generated by PCR using
NNK degenerate codon and screened for the subterminal
hydroxylation of 1 on a microtiter plate by high-throughput
screening (HTS). Variants with enhanced regio- and/or
enantioselectivity were selected for individual biotransfor-
mation in a shake flask to confirm the selectivity, and the best
mutant was used as the template for the subsequent round of
evolution.
In the first three rounds of evolution, a HTS assay based
on the hydroxylation of 4-nitrophenetole was used.[13,15c]
Whereas wild-type P450pyr showed only terminal hydroxyl-
ation of n-octane (1), the screening of 4136 clones in the first
round, 3948 clones in the second round, and 3760 clones in the
third round led to the identification of the P450pyr triple
mutant N100S/F403I/T186I with 40% subterminal selectivity
(i.e., regioselectivity in favor of 2-hydroxylation) and 44% of
the activity of wild-type P450pyr for terminal hydroxylation.
The limited improvement of the subterminal selectivity in
the first three rounds of evolution is possibly due to the use of
an artificial substrate in the HTS assay. To solve this problem
and also to monitor enantioselectivity during evolution,[11a,19]
a novel colorimetric HTS assay was developed to determine
both the regio- and enantioselectivity of the subterminal
hydroxylation of n-octane (1). Figure 1A illustrates the
principle of the HTS assay. Three NAD+-dependent alcohol
dehydrogenases (ADHs) that are highly specific for the
oxidation of 5, (S)-2, and (R)-2, respectively, are used to
separately oxidize the product mixture from a P450pyr-
variant-catalyzed hydroxylation of 1 in three parallel experi-
ments. In the presence of NAD+, NBT, and PMS, the
concentrations of 5, (S)-2, and (R)-2 can be obtained on the
basis of the UV absorption of formazan at 580 nm. The
subterminal selectivity and the enantioselectivity can be
calculated by using Equation (1) and (2), respectively. To
prove the concept, we screened a number of ADHs and found
CpSADH from Candida parapsilosis,[20] PfODH from Pichia
finlandica,[21] and YAD from Saccharomyces cerevisiae[7a] to
be highly specific for the oxidation of (S)-2, (R)-2, and 5,
respectively. Whereas YAD is commercially available,
CpSADH and PfODH were prepared by engineering Escher-
ichia coli expressing the corresponding His-tagged enzyme,
growing the cells, and purifying the enzymes with affinity
chromatography. To validate the assay, 81 samples containing
5, (S)-2, and (R)-2 in different ratios with a total concen-
tration of 1.0 mm were separately oxidized with YAD,
CpSADH, and PfODH for 30 min. The UV absorption at
580 nm was readily determined by the use of a microtiter-
plate reader and found to be proportional to the concen-
tration of 5, (S)-2, and (R)-2, respectively (see Figure S7A).
The determined content of 2 and the ee value of (S)-2 were
very close to the true values (Figure 1B). Further experiments
showed that the assay was independent of the total concen-
tration of 5, (S)-2, and (R)-2 (see Figure S7B).
The colorimetric HTS assay was used for the 4th–6th
rounds of evolution of P450pyr. After biotransformation of
the variants on a 96-well plate, the supernatants were divided
into three plates for the HTS assay. The developed colori-
metric HTS assay was proven to be accurate and applicable
for the evolution. Figure 1C shows a representative example
of the determination of the subterminal selectivity and
enantioselectivity of a P450pyr mutant for the hydroxylation
of n-octane (1) during evolution. The values determined from
the HTS assay are nearly the same as those obtained by GC
analysis of the product of biohydroxylation with the same
mutant in a shake flask.
The best mutant from the 4th round, N100S/F403I/T186I/
L302V, increased the subterminal selectivity to 92% and gave
(S)-2 with 72% ee (Table 1). The best mutant from the 5th
round, I83F/N100S/F403I/T186I/L302V, displayed higher
than 99% subterminal selectivity to provide (S)-2 with an
Table 1: Directed evolution of P450pyr hydroxylase for the regio- and enantioselective subterminal hydroxylation of n-octane (1) to (S)-2-octanol (2).
Round No. of sites No. of
No. of positive Best mutant
clones
Subterminal ee of
Activity
Relative
activity
[%][d]
saturated
clones
selectivity
(S)-2 [U(gcdw)À1 [c]
]
screened identified
[%][a]
[%][b]
WT[e]
nil
22
21
20
19
18
17
nil
nil
10
20
40
40
nil
N100S
N100S/F403I
N100S/T186I/F403I
N100S/T186I/L302V/F403I
I83F/N100S/T186I/L302V/F403I
A77Q/I83F/N100S/T186I/L302V/F403I
0
5
33
40
92
nil
1.8
100
49
110
44
94
97
1
2
3
4
5
6
4136
3948
3760
3572
3384
3196
ND[e] 0.9
ND
56
72
95
98
2.0
0.8
1.7
1.8
1.7
100
120
>99
>99
90
[a] The subterminal selectivity of the best mutant in each round was determined by GC analysis of the products of the biotransformation of n-octane (1;
5 mm) with E. coli cells (2 gcdwLÀ1) expressing the P450pyr mutant in potassium phosphate buffer (100 mm, pH 8.0; 10 mL) containing glucose (2%,
w/v) at 308C and 250 rpm for 4 h. [b] The ee value of (S)-2 was determined by GC analysis on a chiral stationary phase. [c] The activity is the specific
activity determined for the first 30 min of the biotransformation. [d] The activity of the mutant is expressed as a percentage relative to the activity of
E. coli cells expressing P450pyr for the terminal hydroxylation of n-octane (1) under the same conditions. [e] ND: not determined, WT: wild type.
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!