Promoters and transcription factors - Human Gene Evolution - page 232

retention of the functional redundancy could allow diversification of promoter
elements leading to independent regulation.
There are examples of divergently transcribed gene pairs that are not obviously
related evolutionarily and which are unlikely to have arisen by gene duplication
and inversion. Thus the human phosphoribosylaminoimidazole carboxylase
( PAICS ) and phosphoribosylpyrophosphate amidotransferase ( PPAT ) genes,
which encode enzymes of the de novo purine biosynthesis pathway, are only 229 bp
apart on chromosome 4q12 but encode products that are not structurally related
(Brayton et al ., 1994). Similarly, the human isocitrate dehydrogenase 3 ( IDH3G )
and signal sequence receptor 8 ( SSR4 ) genes on chromosome Xq28 are separated
by a 133 bp sequence of CpG island character, share overlapping promoter ele-
ments but encode proteins which have no obvious common function or evolu-
tionary history (Brenner et al ., 1997). Finally, some genes may be transcribed in
opposite directions but do not necessarily share promoter elements nor encode
proteins which share any obvious function, for example the human minichromo-
some maintenance 4 ( MCM4 ) and DNA-activated protein kinase ( PRKDC ) genes
which are separated by ~700 bp on chromosome 8q11 (Connelly et al ., 1998).

5.1.6 5 and 3 untranslated regions of genes

The 5 untranslated regions (UTRs) of vertebrate genes are often evolu-

tionarily conserved (Duret et al ., 1993; Lipman 1997) but to a lesser extent than
their corresponding coding sequences. Thus, whereas the average degree of
nucleotide sequence identity shared between the coding regions of human and
murine genes is ~85%, the 5

and 3

UTRs exhibit sequence identities of 67% and
69% respectively (Makalowski et al ., 1996). In practical terms, this degree of
sequence divergence is sufficient to allow the discrimination of different mam-
mals simply by reference to their UTR sequences (e.g. Soteriou et al ., 1995).
Within the 5

and 3

UTRs, highly conserved regions (HCRs) can be found. These
were defined by Duret et al . (1993) as sequences of at least 100 bp that exhibit >70%
homology between species, and which diverged more than 300 Myrs ago. Since
such sequences would be expected, in the absence of selective pressure, to share only
~30% similarity, evolutionary conservation implies function. Comparison of mam-
malian and avian genes reveals that ~17% and ~30% of genes contain HCRs in
their 5

and 3

UTRs respectively (Duret et al ., 1993). When mammalian and
fish genes are compared, the proportion of genes whose 3

UTRs and 3

UTRs possess an HCR
falls to 5% (Duret et al ., 1993). Since HCRs occur relatively infrequently within
introns, the evolutionary constraints would appear to operate at the level of the
mature mRNA (Duret et al ., 1993). The HCRs in the 5 UTRs of the creatine kinase

B, c- jun and actin genes span the CCAAT and TATA boxes and are therefore likely
to play a role in transcriptional regulation (Duret et al ., 1993). Other HCRs in the 5

UTRs of the transforming growth factor 3 and ferritin heavy chain genes span ele-

ments known to be involved in translation (Duret et al ., 1993). Some HCRs in 3

UTRs are thought to play a role in mRNA degradation (e.g. c- fos and transferrin
receptor genes) whereas others may be important for mRNA transport and transla-
tion (Duret et al ., 1993). HCRs in 3

UTRs appear to be preferentially associated
with widely expressed genes especially those encoding DNA-binding proteins and
cytoskeletal proteins (Duret et al ., 1993).