Biology Reference
In-Depth Information
1990a) and minisatellite sequences (Chandley and Mitchell, 1988; Wahls
et al
.,
1990b). Recombinational breakpoints have also been found to be associated with
topoisomerase I cleavage sites in the rat genome (Bullock
et al
., 1985). The major-
ity of these cleavage sites contain the sequences CTT and GTT. It may therefore be
that the process of nonhomologous recombination is mediated by topoisomerase I.
Gene distribution and density.
Several thousand genes have now been mapped
to within single chromosome bands. Some 80% map to the G-C rich R bands
whilst 20% map to G bands (Bickmore and Sumner, 1989; Craig and Bickmore,
1993). A similar distribution is apparent for CpG islands (
see
Section 1.1.1,
CpG
islands
): 86% are located in R bands (Craig and Bickmore, 1994; Larsen
et al
.,
1992). 'Housekeeping' genes are strictly confined to the R bands together with
about half of the tissue-specific genes (Holmquist, 1992) whereas the remainder of
the tissue-specific genes are present in the G bands. One of the four recognized
types or 'flavors' of R band, known as T bands, are often found at telomeres,
exhibit the highest G+C content and contain between 58% and 68% of R band
genes as well as the majority of CpG islands (Collins
et al
., 1996; Holmquist,
1992).
Chromosomes 13 and 18 appear to possess a relatively low gene density and
chromosome 19 a relatively high density as evidenced by the chromosomal
assignment of some 320 cDNAs derived from a human brain cDNA library
(Polymeropoulos
et al
., 1993). Interestingly, DNA excision repair may be prefer-
entially directed toward regions of high gene density (Surrales
et al
., 1997), a
reflection perhaps of the preferential repair of actively transcribed gene
sequences.
Isochores.
The human genome is a mosaic of large (>300 kb) DNA segments or
isochores that are compositionally homogeneous and which can be subdivided
into a small number of families characterized by different degrees of GC-richness
(30-60%) (Bernardi
et al
., 1993a,b). Five families have been identified: L1 and L2
which are GC-poor and comprise 62% of the genome and H1, H2, and H3 which
are GC-rich and represent 22%, 9%, and 3% of the genome respectively. Gene con-
centration varies between isochores: 34% of human genes are located in L1 and
L2 isochores, 38% in the H1 and H2 isochores, and 28% in the H3 isochores
(Mouchiroud
et al
., 1991; Saccone
et al
., 1996). The banding pattern of chromo-
somes reflects the isochore organization: thus, the G bands are formed by L1 and
L2 isochores whilst the T bands are formed by the H2 and H3 families. A recent
study of two primate globin pseudogenes which reside in different isochore com-
partments has provided evidence that isochores have arisen as a result of muta-
tional bias rather than from the action of selection (Francino and Ochman, 1999).
Matrix attachment regions.
Chromatin is attached to the nuclear matrix or scaf-
fold at specific sites known as matrix or scaffold attachment regions
(MARs/SARs). The organization of chromatin with respect to the nuclear scaffold
is thought to determine chromosome architecture in terms of its functional
domains; this in turn influences gene activity (reviewed by Dillon and Grosveld,
1994 and Walter
et al
., 1998). Indeed, MARs may function so as to place genes at
