Abstract
A question addressed in this study is: why similar enzymes are classified into different subclasses? As an example, PhaC synthases are classified according to four different classes (I, II, III and IV). To answer this question we proposed that besides the catalytic residues, the overall amino acids (AAs) present are responsible for the differences observed. The AAs' composition affects the structure/function/substrate specificity (SFS) of these enzymes. The differences between the classes in various PhaC synthases and proteases were analysed to support our argument. Homology and phylogenic tree of some selected PhaC synthases of different strains (representing the four classes) were demonstrated. The properties of a specific class of enzyme could not be changed into those of another by changing the catalytic residues. Moreover, these differences could not be detected from the proteins' 3D structures, despite clear differences at the AAs level. Another question was also addressed: could we benefit from the various existing protein databases in the field of biotechnology? To answer this, we introduced a model for an Experimental Design based on the information in the protein database (for strains available in our lab) regarding their ability to degrade castor oil. Two enzymes in the phenol degradation pathway, phenol 2-monooxygenase and catechol 1,2-dioxygenase, and a lipase enzyme were analysed. These enzymes were screened and analysed according to the BLAST-protein database and BRENDA. The comprehensive enzyme information system compared six strains against each other, including: Pseudomonas aeruginosa, Bacillus subtilis, Bacillus pumilus, Bacillus thuringiensis, Bacillus licheniformis, and Geobacillus stearothermophilus. Only P. aeruginosa proved to have the three required enzymes and was suitable for the production of lipases from castor oil (crude castor oil is usually contaminated with phenol) as indicated by the databases. In addition, in vivo castor oil degradation and in vitro lipase enzyme activity were analysed. The apparent lipase activity was 1070 Units/ml. Therefore, this new strategy is recommended to better understand the SFS as well as for using protein database in an Experimental Design.