Biology Reference
In-Depth Information
in proteins and are known as domains . Evolutionarily mobile domains are termed
modules . Motifs can be as simple as in the case of the hexamer repeat unit that forms
a left-handed parallel
-helix found in UDP-N-acetylglucosamine acyltransferase
or more complex as in the 21-26 amino acid zinc finger motif (Henikoff et al .,
1997). For practical purposes, a protein domain may be regarded as a structurally
and/or functionally discrete portion of a protein that can fold independently into a
stable tertiary structure and may be composed of single or multiple motifs. Since
the zinc finger motif can fold independently, it may actually be regarded as a small
domain in its own right. Most proteins contain two or more domains although
some may just contain a single domain (Doolittle, 1995).
During evolutionary time, the amino acid sequences of proteins are eroded at
a much faster rate than are the corresponding 3D structures. In other words,
although it may be difficult to discern sequence conservation, structural conser-
vation may still be evident (Creighton, 1993). One example of this is provided by
the
-defensin
( DEFB1 , DEFB2 ; 8p23) genes which, despite a complete lack of DNA sequence
homology, and major differences between their encoded proteins in terms of
their cysteine spacing and disulfide pairing, have nevertheless evolved from a
common ancestor (Lei et al ., 1997).
It is often quite difficult to ascertain whether structural similarities are homolo-
gous (i.e. based upon divergent evolution from a common ancestor) or analogous (i.e.
based upon convergent evolution to a physically favoured secondary or tertiary
structure). The
-defensin ( DEFA1 , DEFA4 , DEFA5 , DEFA6 ; 8p23) and
barrel domain is common among enzymes and examples of this
domain are thought to share a common ancestor rather than to have evolved con-
vergently to a stable fold (Reardon and Farber, 1995). Such examples are a reflection
of the evolution of extant protein structures from a small number of basic folds
encoded by primordial exons.
Primordial exons encoding functional domains are now widely dispersed
among many diverse proteins (Doolittle, 1995). Examples include the ankyrin
repeat (Bork, 1993), the spectrin repeat (Pascual et al ., 1997), the EGF-like domain
(Campbell and Bork 1993), the WD-repeat (Neer et al ., 1994), the ATPase domain
(Bork et al ., 1992) and kringles (Patthy et al ., 1984). The EGF-like domain is pre-
sent in ~1% of human proteins whilst the immunoglobulin domain may be even
more prevalent (Henikoff et al ., 1997). Other domains such as the Kunitz domain
which is less prevalent, may have a more recent origin (Ikeo et al ., 1992).
How many primordial exons were required to construct the huge array of extant
proteins? Dorit et al . (1990) used the frequency with which exons in a 1200 exon
database had been 'reused' between genes encoding different proteins to assess the
size of the underlying exon pool; they estimated that the number of primordial
exons was between 1000 and 7000. A similar study succeeded in distilling 1410
polypeptide chains down to 112 analogous fold families whose members exhibited
an average of <18% sequence identity (Orengo et al ., 1993). Patthy (1991b) has
pointed out, however, that this type of approach may tend to underestimate the true
number of exons since it may not adequately have considered the pool of exons that
participate in exon shuffling only rarely. Databases of protein folds and their fami-
lies are available online at http://www.embl-ebi.ac.uk/dali/ (Homology-derived
Structures of Proteins), http://www.mips.biochem.mpg.de/ (Munich Information
/
 
Search WWH ::




Custom Search