Substrate Model If neither X-ray data nor the amino acid sequence are available
for an enzyme – which is not uncommon for synthetically useful enzymes – one
can proceed as follows: A set of artificial substrates having a broad variety of
structures is subjected to an enzymatic reaction. The results, i.e., the reaction rates
and enantioselectivities, then allow to create a general structure of an imagined
‘ideal’ substrate, which an actual substrate structure should simulate as closely as
possible to ensure rapid acceptance by the enzyme and a high enantioselectivity.
This idealized substrate structure is then called a ‘substrate model’ (Fig. 2.11, left).
Such models have been developed for PLE [239] and Candida rugosa lipase
[369, 370]. Of course these crude models only yield reliable predictions if they
are based on a substantial number of test substrates.
To ensure optimal selectivity of PLE with methyl carboxylates, the α- and
β-substituents should be assigned according to their size (L ¼ large, M ¼ medium
and S ¼ small) with the preferably-accepted enantiomer being shown in Fig. 2.11 (left).
Active Site Model Instead of developing an ideal substrate structure one also can
delineate the structure of the (unknown) active site of the enzyme by the method
described above. Thus, substrates of varying size and polarity are used as probes to
measure the dimensions of the active site in an approach denoted as ‘substrate
mapping’ [371, 372]. Such active site models are frequently employed and they
usually resemble an arrangement of assumed ‘sites’ or ‘pockets’ which are usually
box- or cave-shaped. A relatively reliable active-site model for PLE [373] using
cubic-space descriptors was based on the evaluation of the results obtained from
over 100 substrates (Fig. 2.11, right).
Fig. 2.10 Tetrahedral oxyanion-intermediates during hydrolysis of (R)- and (S)-1-phenylethyl
acetate by Candida antarctica lipase B (Graphics prepared by PyMol v. 1.701, courtesy of Georg
Steinkellner)
82
2 Biocatalytic Applications
for an enzyme – which is not uncommon for synthetically useful enzymes – one
can proceed as follows: A set of artificial substrates having a broad variety of
structures is subjected to an enzymatic reaction. The results, i.e., the reaction rates
and enantioselectivities, then allow to create a general structure of an imagined
‘ideal’ substrate, which an actual substrate structure should simulate as closely as
possible to ensure rapid acceptance by the enzyme and a high enantioselectivity.
This idealized substrate structure is then called a ‘substrate model’ (Fig. 2.11, left).
Such models have been developed for PLE [239] and Candida rugosa lipase
[369, 370]. Of course these crude models only yield reliable predictions if they
are based on a substantial number of test substrates.
To ensure optimal selectivity of PLE with methyl carboxylates, the α- and
β-substituents should be assigned according to their size (L ¼ large, M ¼ medium
and S ¼ small) with the preferably-accepted enantiomer being shown in Fig. 2.11 (left).
Active Site Model Instead of developing an ideal substrate structure one also can
delineate the structure of the (unknown) active site of the enzyme by the method
described above. Thus, substrates of varying size and polarity are used as probes to
measure the dimensions of the active site in an approach denoted as ‘substrate
mapping’ [371, 372]. Such active site models are frequently employed and they
usually resemble an arrangement of assumed ‘sites’ or ‘pockets’ which are usually
box- or cave-shaped. A relatively reliable active-site model for PLE [373] using
cubic-space descriptors was based on the evaluation of the results obtained from
over 100 substrates (Fig. 2.11, right).
Fig. 2.10 Tetrahedral oxyanion-intermediates during hydrolysis of (R)- and (S)-1-phenylethyl
acetate by Candida antarctica lipase B (Graphics prepared by PyMol v. 1.701, courtesy of Georg
Steinkellner)
82
2 Biocatalytic Applications