Biology Reference
In-Depth Information
FIGURE 6.2
De novo enzyme design workflow. A theozyme for the reaction of interest is placed into a scaffold and the resulting structure
designed.
114
modeling as described by Zhang et al.
42
Other, more ad hoc, approaches relying on
chemical intuition or biological precedent will be described below.
Once a theozyme has been devised, the next step in the process is to graft it into a protein
scaffold. In this step, called matching, a library of scaffold protein structures is searched for
attachment sites for the theozyme. A theozyme can be attached in a scaffold if the theozyme
ligand can be placed in a scaffold cavity, and if simultaneously every theozyme side-chain can
be grafted onto a scaffold backbone position, such that the desired relative geometric
orientation between theozyme side-chains and theozyme ligand is maintained, with no clashes
between the theozyme components and the scaffold backbone. Thus, the rigid-body orientation
between the substrate and the to-be-designed enzyme is determined at this stage, and hence
matching algorithms fall into category 3 of computational protein design algorithms.
Several matching algorithms have been proposed, such as a simple enumerative algorithm
that places each side-chain sequentially,
43
an algorithm that places the ligand for each
theozyme side-chain in parallel and then employs six-dimensional geometric hashing to
determine ligand positions that can make all desired theozyme contacts,
44
or an algorithm
that scans through all pairwise combinations of scaffold residues to determine geometric
overlap with theozyme side-chains.
45
Independent of the matching algorithm used, the
result of the matching stage is a set of so-called matches, which are models featuring the
desired minimal active site in a protein cavity. However, the theozyme usually only contains
a handful of side-chains (2
4), while in the matches usually several dozen side-chains are
within the shell contacting the placed ligand. This means that only a subset of the side-
chains making up the new active site have been assigned their ideal identities at this stage.
To determine sequence identities for the remaining active site residues, standard side-chain
placement algorithms are run in the next stage of the designing process. Usually several
