Ruk

Figure 5.6 Genomic architecture at chromosome 17q21.31. Schematic representation of the 17q21.31 region showing two common haplotypes comprising low copy repeat regions (large block arrows) with shaded regions of the arrows indicating repeat subunits. (A) A deletion rearrangement spanning ~600-740 kb (includes the MAPT gene and five other genes not shown) is thought to occur through non-allelic homologous recombination on haplotype H2. (B) A common inversion spanning 900 kb, which includes the MAPT gene (and five other genes not shown), is present on haplotype H2. Adapted and reprinted by permission from Macmillan Publishers Ltd: Nature Genetics (Lupski 2006), copyright 2006.

Figure 5.6 Genomic architecture at chromosome 17q21.31. Schematic representation of the 17q21.31 region showing two common haplotypes comprising low copy repeat regions (large block arrows) with shaded regions of the arrows indicating repeat subunits. (A) A deletion rearrangement spanning ~600-740 kb (includes the MAPT gene and five other genes not shown) is thought to occur through non-allelic homologous recombination on haplotype H2. (B) A common inversion spanning 900 kb, which includes the MAPT gene (and five other genes not shown), is present on haplotype H2. Adapted and reprinted by permission from Macmillan Publishers Ltd: Nature Genetics (Lupski 2006), copyright 2006.

with a reasonably specific clinical phenotype; indeed Sharp and colleagues were able to identify a fifth individual with mental retardation based on their clinical features who also had the deletion at 17q21.31. In the same issue of Nature Genetics where their paper was published, two other papers reported the same deletion as being associated with idiopathic mental retardation and a similar clinical phenotype, which included characteristic facial features, hypotonia, friendly or amicable behaviour, and structural brain abnormalities on imaging. Koolen et al. used array CGH to screen 360 individuals with idiopathic mental retardation and found a person with the 600 bp deletion; screening an additional 480 affected individuals revealed one person with the same 600 bp deletion and one with a smaller 100 bp deletion but having one common breakpoint with the larger deletion (Koolen et al. 2006). Shaw-Smith and colleagues found three cases with the deletion and overall it is thought that this relatively large submicro-scopic deletion at 17q21.31 could account for 1% of cases of mental retardation (Lupski 2006). This deletion has since been extensively characterized among 22 individuals, establishing 'chromosome 17q21.31 microdeletion syndrome' as a clearly defined genomic disorder (OMIM 610443) with an estimated prevalence of one in 16 000 (Koolen et al. 2008). An inversion event in a parent is necessary for the deletion to occur and within the proximal breakpoint a rearrangement hotspot was noted to lie in a mobile DNA element, an L2 long interspersed element (LINE) motif (Koolen et al. 2008).

Box 5.11 Autism spectrum disorders (OMIM 209850)

Autism spectrum disorders are characterized by difficulties with language and social interactions together with restricted and repetitive behaviour with onset by 3 years of age. Such disorders are common, with an estimated prevalence of one in 166.

100K arrays, Friedman and colleagues were able to resolve deletions as small as 178 kb in size: in 100 children with idiopathic mental retardation, eight had de novo deletions and two had duplications (Friedman et al. 2006). More widespread application of array CGH technology in clinical genetic testing is proposed as a cost effective approach (Wordsworth et al. 2007).

5.4.3 De novo copy number mutations and autism

There has been great research interest in the genetic basis of autism (Box 5.11), with strong evidence to support a role for inherited variation but marked genetic heterogeneity and multiple genetic loci reported (Klauck 2006). Highly diverse cytogenetic abnormalities are found in more than 5% of autistic children, occurring across all chromosomes, with notable examples being duplication at 15q11-q13 and 16p13.1, and deletions at 2q37 and 22q13.3. However, the extent of submicroscopic structural variation remains unclear (Vorstman et al. 2006).

Sebat and colleagues sought to address this using high resolution CGH arrays, although the authors noted that they were probably failing to detect the vast majority of copy number variants because of their small size (Sebat et al. 2007). With this caveat in resolution imposed by available technology, the results were still striking. A significantly higher rate of de novo copy number mutation among patients with sporadic autism was found compared to healthy controls (10% versus 1%, P = 0.0005) (Sebat et al. 2007). The investigators were careful to exclude cases of syndromic autism (severe mental retardation or other congenital abnormalities) and patients with known cytogenetic abnormalities. Among

195 patients with autism spectrum disorders, 14 de novo copy number mutations were found. The authors estimated that once cytogenetically visible abnormalities were taken into account, the frequency of de novo copy number variation was 15% at current detection thresholds. The true figure may prove very much higher, with a diverse array of rare, highly penetrant sporadic copy number mutations postulated to underlie much of the sporadic nature of autism (Sebat et al. 2007).

5.5 Inversions in health and disease

5.5.1 Inversions may cause severe disease

Inversions are reported to occur on all chromosomes and may involve two breaks on different arms (pericentric inversion) or the same arm (paracentric inversion) (Fig. 3.8A). Many inversions may not be associated with any phenotypic consequences, for example an inversion on chromosome 9 inv(9)(p11q13) is found in 1-3% of the population without any known clinical significance (Hsu et al. 1987; Yamada 1992). There are, however, some notable examples of inversions leading to severe disease, illustrated by inversions involving intron 22 and intron 1

of the F8 gene encoding factor VIII which are associated with severe haemophilia A (Box 5.12).

An inversion at the IDS gene in the same region of the X chromosome, Xq28, has been associated with another X-linked recessive disorder, Hunter syndrome (Bondeson et al. 1995). Here a deficiency of idunorate 2-sulphatase, encoded by IDS, can lead to severe disease with progressive damage to the brain and liver, often causing death by the age of 15 years.

Inversions of segmentally duplicated olfactory receptor genes at 4p16 and 8p23 were found to occur commonly among control subjects of European descent at 12.5% and 26%, respectively, with 2.5% of individuals heterozygous for both inversions (Giglio et al. 2001, 2002). This latter state, while carrying no phenotype for the affected individual, has been associated with having offspring carrying a recurrent translocation for the two loci denoted t(4;8) (p16;p23) (Section 3.4). Thus having two inversion polymorphisms in a heterozygous state on non-homologous chromosomes can lead to interchromosomal rearrangements. The translocation, in some individuals, is associated with Wolf-Hirschhorn syndrome (OMIM 194190). This syndrome is characterized by severe growth retardation, mental retardation, a characteristic facial appearance, and closure defects (for example cleft lip or palate, coloboma of the eye, and cardiac septal defects).

Box 5.12 Haemophilia A (OMIM 306700)

This is one of the commonest X-linked diseases affecting one in 5000 males. It is caused by a deficiency of coagulation factor VIII, which leads to easy bruising, haemorrhage into joints and muscles, and prolonged bleeding. Approximately 50% of people have a severe phenotype with less than 1% residual factor VIII, while 10% have moderate disease (2-5% residual factor VIII) and 40% have mild disease (5-30% residual factor VIII). The condition results from genetic variation at the F8 gene encoding factor VIII: in those with severe disease, half of cases arise from an inversion involving intron 22 and a homologous region telomeric to the F8 gene

(Lakich et al. 1993). A further 5% of severe cases were reported to arise from a second inversion, this time involving intron 1 (Bagnall et al. 2002). In both situations, inversions have arisen due to intrachromosomal recombination between homologous regions in introns 1 or 22, and regions telomeric to the F8 gene. A large gene deletion is thought to account for 5% of severe cases while the remainder, together with all moderate and mild cases, arise due to point mutations and small insertions/deletions of the F8 gene (Castaldo et al. 2007). These are thought to be newly occurring mutations in approximately one-third of cases.

5.5.2 Inversion and deletion at 17q21.31 with evidence of selection

Large inversions are thought to often arise from nonallelic homologous recombination involving highly homologous low copy repeat regions present in an inverted orientation (Shaw and Lupski 2004). Such a situation is seen in the complex genomic architecture on the long arm of chromosome 17 at 17q21.31 where both deletions and a large inversion have been identified (see Fig. 5.6) (Stefansson et al. 2005; Lupski 2006).

Stefansson and colleagues found a large inversion polymorphism that was common among Europeans (frequency of inversion 21%), but considerably rarer in Africans (6%) and almost absent among Asians (1%) (Stefansson et al. 2005). The inversion spans a region of particular interest in neurological diseases, with two highly divergent haplotypes previously identified involving MAPT (encoding microtubule-associated protein tau), designated H1 and H2, and shown to be associated with progressive supranuclear palsy (Baker et al. 1999) and Parkinson's disease (Skipper et al. 2004). Analysis of 60 microsatellite markers showed that the H2 haplotype was structurally distinct, and that a 900 kb segment was inverted compared to H1 and the reference assembly for the region. Analysis of human and chimpanzee (Pan troglodytes) clones suggested very great mutational differences and a very ancient divergence between the H1 and H2 lineages, such that the inversion polymorphism may be 3 million years old (Stefansson et al. 2005). Moreover, analysis in an Icelandic population showed that the H2 lineage was in fact undergoing positive selection, with carrier females of the inversion polymorphism having more children and higher recombination rates (Stefansson et al. 2005).

5.5.3 Finding inversions across the human genome

High throughput methods for detecting inversion polymorphisms are not currently available, however approaches using high density single nucleotide polymorphism (SNP) genotyping data have been proposed. For example, Bansal and colleagues developed a statistical method for the detection of large inversions based on unusual patterns of linkage disequilibrium indicative of inversions using the HapMap dataset (Bansal et al. 2007). When such patterns were present in a majority of chromosomes from the population, and were in an inverse orientation to the reference human genome sequence, this was taken as evidence of a candidate inversion. One hundred and seventy-six such events were reported, ranging in size from 200 kb to several megabases, however only a minority were likely to be real.

Comparison of a second genome sequence based on fosmid paired-end sequencing data with the reference human assembly has also been informative, with 56 likely inversion breakpoints greater than 8 kb in size identified (Tuzun et al. 2005). This approach utilized a human fosmid DNA genomic library generated from an anonymous North American female (donor GM15510) to identify orientation discrepancies, and the authors validated a large proportion of those identified.

Comparison of human and chimpanzee genome sequences has also allowed definition of potential sites of inversions between the two species (Feuk et al. 2005; Szamalek et al. 2006). Feuk and colleagues identified 1576 putative inversions in such regions, ranging in size from 23 bp to 62 Mb, with 33 inversions greater than 100 kb and the highest count seen on the X chromosome (Feuk et al. 2005). Of 27 putative inversions tested, 23 were experimentally validated including a novel 4.3 Mb inversion at 7p14. Three of these inversions were found to be polymorphic among a panel of ten individuals of European descent: a 730 kb inversion at 7p22 (5% minor allele frequency), a 13 kb inversion at 7q11 (30%), and a 1 kb inversion at 16q24 (48%).

Among apparently healthy human subjects, 182 inversions greater than 1 kb in size had been reported by December 2007 (Database of Genomic Variants, http://projects.tcag.ca/variation/) (Zhang et al. 2006). As described earlier, pericentric inversions on chromosome 9, inv(9)(p11q12)/inv(9)(p11q13), are found in about 2% of the normal population and are regarded as polymorphic variants without significant consequences: homologous regions in the long and short arms of chromosome 9 around the breakpoints may facilitate recombination and inversion (Park et al. 1998). Other large inversions have been found. For example, on chromosome Xq28, mis-pairing between two large inverted repeats was found to lead to a 48 kb inversion of a region containing the

FLN1/EMD genes in 18 out of 108 human X chromosomes analysed (Small et al. 1997).

5.6 Summary

This chapter has highlighted the importance of pathogenic copy number variation, in particular how this relates to the growing list of genomic disorders. Here genomic architecture, most notably highly homologous stretches of DNA, is highly significant in modulating chromosomal rearrangements and genomic disease (Inoue and Lupski 2002). Examples of hotspots of recurrent rearrangements and disease include chromosome 22q11 (DiGeorge syndrome and velocardiofacial syndrome; Box 5.2) and the proximal short arm of chromosome 17 (CMT type 1A; Box 5.4). A number of different genomic disorders have been described to illustrate the basis and nature of recurrent and non-recurrent disease. Sites involving reciprocal genomic disorders were noted such as duplication, deletion, and inversions at 7q11.23 (Williams-Beuren syndrome); or at 17p11.2 involving the dosage-sensitive gene peripheral myelin protein-22 (PMP22) leading to demyelinating peripheral neuropathies (CMT type 1A associated with duplications, and deletions with HNPP) (Lupski et al. 1991; Chance et al. 1993). Other genomic disorders have been described showing parent of origin effects such as Prader-Willi and Angelman syndromes.

The importance of recent advances in the analysis of submicroscopic structural variation have also been emphasized by progress that has been achieved in the diagnosis and understanding of mental retardation. The analysis of copy number variation has led, for example, to the identification of previously unrecognized genomic disorders such as 17q21.31 microdeletion syndrome, which may account for 1% of cases of idiopathic mental retardation (Lupski 2006). Many other new syndromes and underlying structural causes have been defined by high resolution mapping techniques to define pathogenic copy number variation, representing major advances of significant clinical importance. Array CGH techniques developed for a research setting are now being implemented for clinical diagnostic use, highlighting the utility of such advances to direct patient care.

Terminal deletions and subtelomeric events have also been reviewed, together with structural variation i nvolving a change in orientation in the form of inversions. Inversions are recognized to occur in all chromosomes and can lead to severe disease, as seen with inversions involving the F8 gene and haemophilia A, or reach high frequencies in particular populations associated with positive selection (Lakich et al. 1993; Stefansson et al. 2005). Systematic surveys to detect inversions within human populations are now becoming feasible, as they are for other classes of submicroscopic variation.

The role of segmental duplications and in particular low copy repeats in genomic disorders has been described. In the next chapter the nature and importance of segmental duplications are described taking a broad view of the field, which allows analysis of their extent and evolutionary significance. Further examples involving copy number variation and other genetic variation are also described in Chapter 6 through review of the genetics of colour vision and rhesus blood groups.

5.7 Reviews

Reviews of subjects in this chapter can be found in the following publications:

Topic

Mechanisms of chromosomal rearrangements and genomic disorders

Clinical and genetic aspects of Parkinson's disease Charcot-Marie-Tooth disease Williams-Beuren syndrome Cri du chat syndrome Chromosomal rearrangements at chromosome 22q11 Genetics of autism spectrum disorder Molecular genetics of haemophilia A

References

Lupski 1998; Shaffer and Lupski 2000; Inoue and Lupski 2002; Shaw and Lupski 2004; Lupski and Stankiewicz 2005; Gu et al.

0 0

Post a comment