Biology Reference
In-Depth Information
occurs at polymorphic frequencies in the Japanese (28%) and African-American
(2.5%) populations (Fernandez-Salguero
et al
., 1995).
Presumably, the greater the number of homologous repetitive elements, the
greater the opportunity for gene conversion. Consistent with this assertion, multi-
gene families frequently exhibit extensive sequence homogeneity compatible
with the consequences of gene conversion. Thus, the human ribosomal RNA gene
family, which is composed of ~400 members arranged in tandem repeats on five
different chromosomes (
RNR1
, 13p12;
RNR2
, 14p12;
RNR3
, 15p12;
RNR4
,
21p12;
RNR5
, 22p12) are much more similar to one another than they are to
members of the rDNA family in other primates (Arnheim
et al
., 1980; Li, 1997).
This genomic organization also results in a greater degree of homogeneity within
rDNA clusters than between them (Gonzalez
et al
., 1992; Seperack
et al
., 1988), a
phenomenon also noted for the polyubiquitin genes (
UBA52
, 19p13.1;
UBB
,
17p11.1-p12;
UBC
, 12q24.3) where gene conversion appears to occur exclusively
within rather than between gene clusters (Sharp and Li, 1987). Other examples of
repetitive gene families being subject to the homogenizing effects of gene conver-
sion are those of the human immunoglobulin C
(
IGHA1
; 14q32-q33; Kawamura
et al
., 1992; McCormack
et al
., 1993), V
light chain (
IGKV
; 2p12; Huber
et al
.,
1993),
light chain
(
IGLC1
; 22q11.12; Udey and Blomberg, 1988) genes, U2 snRNA (
RNU2
; 17q21-
q22; Liao
et al
., 1997) genes and the T-lymphocyte antigen receptor
heavy chain (
IGHG1
; 14q32.33; Lefranc
et al
., 1986) and
(
TCRB
;
14q11.2; Funkhouser
et al
., 1997; Tunnacliffe
et al
., 1985) genes.
The pituitary-expressed growth hormone (
GH1
; 17q22-q24) gene promoter
region has been found to exhibit a very high level of sequence polymorphism with
8 variant nucleotides within a 134 bp stretch (Giordano
et al
., 1997). These eight
variable positions have been ascribed to 12 different haplotypes ranging in fre-
quency from 2% to 31% in the general population. Since they occur in the same
positions in which the
GH1
gene differs from the other placentally expressed
genes of the growth hormone cluster [two chorionic somatomammotropin (
CSH1
and
CSH2
) genes, a chorionic somatomammotropin 'pseudogene' (
CSHL1
) and a
second growth hormone gene (
GH2
)], the mechanism responsible is likely to be
gene conversion with the placentally expressed genes serving as donors of the
converted sequences. Various examples of gene conversion of functional genes
templated by pseudogenes have been documented as a cause of human pathology
(discussed in Chapter 6 section 6.1.6).
Studies of gene conversion in fungi have shown that gene conversion occurs not
only between alleles but also intrachromosomally between duplicated sequences
on either the same chromatid or the sister chromatid and even between sequences
on nonhomologous chromosomes. Does gene conversion in the human genome
occur predominantly within a single chromosome (i.e. through sister chromatid
exchanges) or instead between either homologous or nonhomologous chromo-
somes? Data from both the human ribosomal RNA (Seperack
et al
., 1988) and U2
snRNA gene families (Liao
et al
., 1997) have suggested that intra-chromosomal
events are more frequent than inter-chromosomal events. Inter-chromosomal gene
conversion events have been reported in human gene pathology (see Chapter 6,
section 6.1.6) but these are comparatively rare. In an evolutionary context,
possible examples of gene conversion operating between homologous loci on
