aspect lies in the selection problem [337]: In order to identify the one (or the
few) mutant protein(s) with improved properties amongst the vast number of
variants (typically 10
4 –10
6 ), which are (more or less) randomly generated, an
efficient screening method is required to find the tiny needle in the very big
haystack. Adequate screening methods usually rely on spectral changes during
catalysis. The drawback of this first-generation screening method is the requirement for a chromogenic or fluorogenic ‘reporter group’ in the substrate, which
usually consists of a large (hetero)aromatic moiety which needs (at least) 10–14
π-electrons to be ‘visible’ by UV/VIS or fluorescence spectroscopy. Classic
reporter groups are (colorless) p-nitrophenyl derivatives, such as esters or
glycosides, which liberate the (yellow) p-nitrophenolate anion upon enzyme
catalysis. The latter can be spectrophotometrically monitored at 410 nm
(Scheme 2.43). Unfortunately, by introduction of the chromogenic reporter
group, the original substrate (e.g., a methyl ester) is modified to a structurally
very different p-nitrophenyl substrate ester analog. Since the mutants are
screened for optimal activity/selectivity on the surrogate substrate, their performance with the ‘real’ (methyl ester) substrate will be less efficient. In order to
create ‘real’ mutant enzymes for ‘real’ substrates, more sophisticated screening
methods are recommended based either on a multienzyme assay for acetate
(produced during ester hydrolysis, Scheme 2.43) [338], MS analysis of (deuterated) ‘pseudo-enantiomeric’ products, or time-resolved IR thermogravimetry
[339–343]. After all, 34% of random single amino acid replacements yield an
inactive protein [344] and you always get what you screen for [345].
Scheme 2.43 shows the use of a surrogate ester substrate bearing a chromogenic
( p-nitrophenyl) reporter group for the screening of Pseudomonas aeruginosa lipase
mutants possessing improved enantioselectivities for the kinetic resolution of an
α-chiral long-chain fatty acid [346]. Pure substrate enantiomers were separately
tested in 96-well microtiter plates using a plate reader for the readout of
enantioselectivity. After four rounds of epPCR at a low mutation rate, mutant A
(E ¼ 11) was obtained. Sequence analysis of mutant A (and several other positive
hits) revealed that position #155 was a ‘hot spot’ for beneficial variations. Hence,
mutant A was improved via saturation mutagenesis through variation of all remaining
19 amino acids at position #155 yielding mutant B (E ¼ 20). Another round of epPCR
on B gave mutant C (E ¼ 25), which could not be further improved. At high
mutations rates, epPRC of the wildtype enzyme gave only slightly improved variants
D and E (E ¼ 3.0 and 6.5, respectively) indicating further ‘hot regions’. Combinatorial multiple cassette mutagenesis (CMCM) of the wild-type enzyme in the ‘hot
region’ of amino acids 160–163 gave mutant G (E ¼ 30). The latter could be further
improved by DNA-shuffling with mutant genes D and E to finally yield mutant J,
which exhibited a top value of E > 51 among ~40,000 mutants screened.
An example for the successful generation of highly enantioselective esterase
mutants capable of hydrolyzing acetate esters of tert-alcohols is shown in Scheme
2.44 [347, 348]. In order to improve the modest enantioselectivity of wild-type Bacillus
subtilis esterase (E ¼ 5 and 43), a library of ca. 5000 mutants was constructed, which
78
2 Biocatalytic Applications
few) mutant protein(s) with improved properties amongst the vast number of
variants (typically 10
4 –10
6 ), which are (more or less) randomly generated, an
efficient screening method is required to find the tiny needle in the very big
haystack. Adequate screening methods usually rely on spectral changes during
catalysis. The drawback of this first-generation screening method is the requirement for a chromogenic or fluorogenic ‘reporter group’ in the substrate, which
usually consists of a large (hetero)aromatic moiety which needs (at least) 10–14
π-electrons to be ‘visible’ by UV/VIS or fluorescence spectroscopy. Classic
reporter groups are (colorless) p-nitrophenyl derivatives, such as esters or
glycosides, which liberate the (yellow) p-nitrophenolate anion upon enzyme
catalysis. The latter can be spectrophotometrically monitored at 410 nm
(Scheme 2.43). Unfortunately, by introduction of the chromogenic reporter
group, the original substrate (e.g., a methyl ester) is modified to a structurally
very different p-nitrophenyl substrate ester analog. Since the mutants are
screened for optimal activity/selectivity on the surrogate substrate, their performance with the ‘real’ (methyl ester) substrate will be less efficient. In order to
create ‘real’ mutant enzymes for ‘real’ substrates, more sophisticated screening
methods are recommended based either on a multienzyme assay for acetate
(produced during ester hydrolysis, Scheme 2.43) [338], MS analysis of (deuterated) ‘pseudo-enantiomeric’ products, or time-resolved IR thermogravimetry
[339–343]. After all, 34% of random single amino acid replacements yield an
inactive protein [344] and you always get what you screen for [345].
Scheme 2.43 shows the use of a surrogate ester substrate bearing a chromogenic
( p-nitrophenyl) reporter group for the screening of Pseudomonas aeruginosa lipase
mutants possessing improved enantioselectivities for the kinetic resolution of an
α-chiral long-chain fatty acid [346]. Pure substrate enantiomers were separately
tested in 96-well microtiter plates using a plate reader for the readout of
enantioselectivity. After four rounds of epPCR at a low mutation rate, mutant A
(E ¼ 11) was obtained. Sequence analysis of mutant A (and several other positive
hits) revealed that position #155 was a ‘hot spot’ for beneficial variations. Hence,
mutant A was improved via saturation mutagenesis through variation of all remaining
19 amino acids at position #155 yielding mutant B (E ¼ 20). Another round of epPCR
on B gave mutant C (E ¼ 25), which could not be further improved. At high
mutations rates, epPRC of the wildtype enzyme gave only slightly improved variants
D and E (E ¼ 3.0 and 6.5, respectively) indicating further ‘hot regions’. Combinatorial multiple cassette mutagenesis (CMCM) of the wild-type enzyme in the ‘hot
region’ of amino acids 160–163 gave mutant G (E ¼ 30). The latter could be further
improved by DNA-shuffling with mutant genes D and E to finally yield mutant J,
which exhibited a top value of E > 51 among ~40,000 mutants screened.
An example for the successful generation of highly enantioselective esterase
mutants capable of hydrolyzing acetate esters of tert-alcohols is shown in Scheme
2.44 [347, 348]. In order to improve the modest enantioselectivity of wild-type Bacillus
subtilis esterase (E ¼ 5 and 43), a library of ca. 5000 mutants was constructed, which
78
2 Biocatalytic Applications