Biology Reference
In-Depth Information
permutations, thereby giving eukaryotic promoters their modular character
(Mushegian and Koonin, 1996; reviewed by Nussinov, 1990).
Transcriptional activators bind to promoter sequences in different combina-
tions and permutations, some factors being coordinated by
cis
-acting DNA
sequence motifs, others by protein-protein interactions on the promoter. The
effect of any given activator is therefore likely to depend on the other activator
and repressor proteins present which may bind either competitively or coopera-
tively. Such a process of 'combinatorial control' has, by maximizing flexibility,
potentiated the rapid evolution of quite elaborate gene regulatory networks
(Ptashnel, 1997). Those seeking a reference source to
cis
-acting DNA sequence
motifs may consult the Transcription Regulatory Regions Database (
http:
//www.bionet.nsc.ru/trrd
) or the Eukaryotic Promoter Database (
http:
//www.epd.isb-sib.ch
) both of which contain information on regulatory regions
of eukaryotic genes and include data on transcription factor binding sites, pro-
moters, enhancers and silencers.
If we take the simple case of two genes recently duplicated at the genomic DNA
level, then we might reasonably expect that their promoter regions would initially
be very similar. As a consequence, the expression profiles of these genes might
also be expected to be very similar if not identical. However, gene duplication
generates redundancy allowing the promoters to be freed from the constraints of
selection thereby enabling them to acquire mutations. Such mutations can lead to
the inactivation of the gene by abolishing its expression (see Chapter 6) or alter-
natively can, in rather more subtle ways, serve to change its expression pattern.
In principle, promoters can change either by the slow steady acquisition of sin-
gle base-pair substitutions in pre-existing sequence motifs or, more dramatically
and abruptly, by
promoter shuffling
, the gain or loss of individual regulatory ele-
ments exchanged between genes in cassette fashion (Surguchov, 1991). Promoter
shuffling may have occurred in cases of homologous genes that contain dissimilar
upstream regulatory elements but also in cases of nonhomologous genes contain-
ing similar upstream regulatory elements. For promoter shuffling to be viable, the
transposed sequence element must be capable of exerting an influence on the
transcription of the gene that has just acquired it. That this proposition is a rea-
sonable one is evidenced by the results of studies reported by Kermekchiev
et al
.
(1991). These authors tested 27 combinations of different promoters and
enhancers and found that the relative
in vitro
efficiency of the enhancers was
roughly the same irrespective of the promoter used. Another way in which pro-
moter sequences can change abruptly is through gene conversion as described for
the human growth hormone (
GH1
; 17q22-q24; Giordano
et al
., 1997) and
-globin
(
HBG1
,
HBG2
; 11p15.5; Chiu
et al
., 1997) gene promoters.
In subsequent sections, the similarities and differences between the promoters
of extant mammalian genes will be explored. In addition, some of the mechanisms
by which evolution has recruited specific sequences to a promoter function and
fine-tuned their interactions with transcription factors will be described.
