A note on clustering the functionally-related paralogues and orthologues of proteins: a case of the FK506-binding proteins (FKBPs)

Abstract

The expression patterns of 18 FK506-binding proteins (FKBPs) encoded in the human genome have been established whereas the functional significance of the numerous ORFs coding for FKBP-like sequences remains unknown. Nominal masses of the human FKBPs vary from 12 to 135 kDa. Some large FKBPs consist up to four repeats of the 12 kDa FK506-like binding domain (FKBD) whereas other large FKBPs contain one FKBD linked to different functional domains such as TPRs, leucine-zipper, calmodulin-binding domain etc. The genomes of other eukaryotic organisms, namely D. melanogaster, C. elegans, A. thaliana, S. pombe and S. cerevisiae encode different numbers of the FKBPs' paralogues some of which are orthologues to the human FKBPs. A library of novel algorithms was developed and used for computation of the level of conservation of the hydrophobicity and bulkiness profiles, and the amino acid compositions (AACs) of 247 aligned sequences of FKBPs. The pairwisely-compared hydrophobicity and bulkiness profiles for some combinations of the aligned sequences of the FKBDs yielded high values of the correlation coefficients (CCF). The AACs of some combinations of the aligned sequences of the FKBDs also differed to a low degree. The functionally-related orthologues and paralogues of the FKBPs were clustered by using the following criteria: 1 degrees apparent conservation of the crucial amino acid (AA) residues for peptidylprolyl cis/trans isomerase (PPIase) acitity and binding of some immunosuppressive drugs; 2 degrees convergence of the three mentioned above properties of the polypeptide chain; 3 degrees similarity in the sequence attributes pI and total hydrophobicity index (HI). The clustering method was used for setting up several hypotheses on the emergence of certain classes of the FKBPs in the eukaryotic kingdom.