The pacifastin inhibitor family web site

[Introduction] [Sequences] [Alignments] [Structures] [Phylogenetics] [References]

Created and maintained by Zoltán Gáspári
E-mail: szpari@para.chem.elte.hu

Last modified: 21.05.2003


Alignments

Nomenclature

Nomenclature of the sequences follows the rules described in Simonet et. al., (2002), i.e. 'Genus species pacifastin-related inhibitor n' or Genus species pacifastin-related domain n', according to whether the polypeptide is processed into individual monomers or not. Experimental evidence shows that the locust (Locusta migratoria and Schistocerca gregaria) inhibitors are processed, whereas the pacifastin light chain, containing 9 inhibitory domains, remains intact. In all other cases, we assumed that the polypeptide chains are not processed, based on the lack of dibasic cleavage sites characteristic of the locust sequences.

Sequence alignments

Multiple alignments of different types are available:
Protein sequences: [FASTA] [ClustalW aln]
Nucleotide sequences: [FASTA]
Dual alignment (protein + nucleotide sequences) [custom format]

The protein sequence alignment was constructed manually. The alignment of the corresponding coding nucleotide sequences was constructed based on the protein sequence alignment.


Alignment of selected members of the family. The disulfide pairing is indicated. Residues changing in accordance with the size of the binding loop are green, those involved in the core interaction are light blue.

Correlated mutation analysis (CMA)

Correlated mutation analysis was performed as described in Oliveira et al. (2002): Correlated mutation analyses on very large sequence families. ChemBioChem :1010-1017. and Oliveira et al. (2003): From sequence to structure and function with entropy-variability plots. Proteins, accepted.

Networks obtained consistently (i.e. with different parameters and grouping of the sequences) comprise the most conserved residues, namely the six cysteines, a glycine and the asparagine involved in binding loop stabilization (see in detail below). The other "network" includes the two residues of the proposed "core interaction", one in the first and another in the third β-strand.
An example to networks with a pairwise correlation cutoff of 0.75 and using groups of similar sequences (the corresponding residues of SGCI are shown in parentheses):
Network#   1:    7 (C)  24 (C)  25 (N)  27 (C)  29 (C)  33 (G)  38 (C)  44 (C)
Network#   2:   18 (F)  36 (A)
Network#   3:   19 (K)  22 (D)  35 (S)  45 (P)
Network #1 corresponds to the conserved residues, network#2 to the core interaction. The role of residues in network#3 is not clear when considering their positions in the 3D structure.

The protease binding loop

The protease binding loop is located between two disulfide bridges, the P3 and P3' residues are cysteines. The 9 P1 and 13 P1' residues form 28 pairs in the 59 sequences. The bining loop is stabilized by an almost perfectlty conserved asparagine. This residue is replaced with Asp in one sequence (PRPD1 from Pyrocoelia rufa, we have no information whether this is an effective inhibitor or not), and His in two cases (BMPD1 from Bombyx mori and MSPD1 from Manduca sexta). In the latter two closely related sequences the binding loop is one residue longer (5 residues between the flanking disulfide bridges) than usual, and the first residue of the binding loop is Leu instead of Ser/Thr found in the majority of the other inhibitors. This observation is consistent with the structural role of the conserved Asn.