Genes and gene families - Human Gene Evolution - page 181

nontranscribed spacer region ( Figure 4.25 ; Gonzalez et al ., 1992). The 13 kb (45S)
transcript is then processed to generate the 28S, 18S and 5S mature rRNA mole-
cules with respective lengths of 1.8 kb, 0.15 kb, and 5.8 kb.
The 28S rRNA contains both conserved and variable regions. The former are
constant in size and sequence whereas the latter are variable (Gorski et al ., 1987)
and exhibit species-specific differences between primates (Gonzalez et al ., 1988;
1990). The variable regions potentiate hairpin loop formation and are essential for
rRNA secondary structure (Gorski et al ., 1987).
The number of rRNA genes on a given chromosome varies in the population,
while in any one individual, rRNA gene number varies between clusters
(Arnheim et al ., 1980). The rRNA clusters appear to have evolved in concerted
fashion. Thus rRNA genes on non-homologous chromosomes are far more simi-
lar to one another than would be expected if these genes had evolved indepen-
dently. Several mechanisms have been proposed to account for this relative
homogeneity of rDNA including unequal crossing over between rDNA sequences
on nonhomologous chromosomes (Worton et al ., 1988). Gene conversion may
then act so as to homogenize rDNA repeats within each cluster. On the basis of
linkage disequilibrium data, Seperack et al . (1988) have suggested that gene con-
version and unequal crossing over may also occur within a chromosome (i.e. by
sister chromatid exchanges).

5S ribosomal RNA genes. The human 5S rRNA ( RN5S1 ) genes occur in clus-
ters of 2.3 kb and 1.6 kb tandem repeat units which differ from each other by
virtue of a deletion in the 3

flanking region (Sorensen and Frederiksen, 1991).
There appear to be 300-400 copies of 5S rRNA genes per haploid human genome.
Between 100 and 150 genes are derived from the 2.3 kb cluster (probably located
at chromosome 1q42), 5-10 from the 1.6 kb cluster whereas the remaining
200-300 genes/gene variants are not found in repeat structures and may be dis-
persed around the genome. In addition to these genes, a large number of 5S rRNA
pseudogenes exist which brings the total number of 5S rRNA homologous
sequences per haploid genome to around 2000.

Small nuclear RNA genes. The human U2 snRNA ( RNU2 ) genes are clustered
in tandem arrays of between 6 and 30 copies at 17q21-q22 (Van Arsdell and
Weiner, 1984; Westin et al ., 1984). The primate U2 snRNA arrays have evolved in
concerted fashion with each repeat being essentially homogeneous within a
species although somewhat different between species, an observation consistent
with the action of gene conversion (Liao and Weiner, 1995; Liao et al ., 1997;
Matera et al ., 1990). By contrast to the situation in higher primates, the U2 snRNA
genes of both the mouse and the prosimian, Galago crassicaudatus , are dispersed
rather than clustered suggesting that the arrays characteristic of higher primates
may have resulted from amplification of a common ancestral gene (Matera et al .,
1990). Once established in the simian lineage, however, the U2 tandem repeat
array has remained at the same chromosomal locus through multiple speciation
events over a period of >35 Myrs (Pavelitz et al ., 1995).
Some 30 copies of the human U1 snRNA ( RNU1 ) genes are also clustered at
1p36.3. This site is distinct from the cluster of U1 snRNA pseudogenes at 1q12-q22