| Abstract | The gelsolin homology (GH) domain has been found
to date exclusively in actin-binding proteins. In
humans, three copies of the domain are present in
CapG, five copies in supervillin, and six copies each in
adseverin, gelsolin, flightless I and the villins: villin,
advillin and villin-like protein. Caenorhabditis elegans
contains a four-GH-domain protein, GSNL-1. These
architectures are predicted to have arisen from gene
triplication followed by gene duplication to result in
the six-domain protein. The subsequent loss of one,
two or three domains produced the five-, four-, and
three-domain proteins, respectively. Here we conducted
BLAST and hidden Markov based searches of UniProt
and NCBI databases to identify novel gelsolin domain
containing proteins. The variety in architectures suggests
that the GH domain has been tested in many
molecular constructions during evolution. Of particular
note is flightless-like I protein (FLIIL1) from Entamoeba
histolytica, which combines a leucine rich
repeats (LRR) domain, seven GH domains, and a headpiece
domain, thus combining many of the features of
flightless I with those of villin or supervillin. As such,
the GH domain superfamily appears to have developed
along complex routes. The distribution of these proteins
was analyzed in the 343 completely sequenced
genomes, mapped onto the tree of life, and phylogenetic
trees of the proteins were constructed to gain
insight into their evolution.
|
