Microdeletions (Genetics)

In the 1980s, the development of in situ hybridization techniques, particularly those utilizing fluorescent markers (FISH) (see Article 22, FISH, Volume 1) (Pinkel et al., 1988), led to the discovery of the first cryptic deletion syndromes, that is, those in which the missing material was not visible using conventional microscopy. Discovery of most microdeletion syndromes resulted from collaborations between clinicians, cytogeneticists, and molecular geneticists. Chromosome markers such as translocation breakpoints found in patients with abnormal phenotypes were often pivotal in identifying the chromosomal region of interest following which positional cloning methods resulted in the identification and characterization of the gene(s) of interest (Tommerup, 1993). For example, the first direct evidence that hemizygosity at the ELN locus contributes to the Williams syndrome phenotype followed a report that a t(6;7)(p21.1;q11.23) translocation was segregating in a family with dominant supra valvular aortic stenosis (SVAS) (Ashkenas, 1996). Subsequent studies have identified a number of genes implicated in the William’s contiguous gene phenotype, and in common with many other human microdeletion syndromes, hemizygosity at one or more loci leading to the disruption of expression of dosage-sensitive genes appears to be the principal mutational mechanism underlying the clinical phenotypes.

There have been over 60 reported microdeletion syndromes (see Article 87, The microdeletion syndromes, Volume 2) involving virtually all chromosomes (Table 1). Although haploinsufficiency of dosage-sensitive genes is thought to be the main mutational mechanism(s) underlying the associated clinical phenotype(s), there are rare examples where the deletion’s breakpoints per se appear to be the mutational mechanism mediated by the breakpoint disrupting a regulatory element that controls a nearby gene (e.g, 3′ interstitial deletions in cases with aniridia; Lauderdale et al., 2000; Crolla and van Heyningen, 2002).


Several terminal, or more correctly, subtelomeric deletions have been described following analysis of patients with normal karyotypes using chromosome-specific sub-telomere FISH probes (Knight et al., 1997). These studies were instrumental in defining the so-called terminal deletion syndromes that are distributed throughout the genome and were found in ~5% patients with idiopathic developmental delay with or without associated congenital abnormalities (de Vries et al., 2003). A proportion of the deletions observed in these studies were the unbalanced products of balanced parental rearrangements (usually translocations), and reciprocal duplications have also been reported.

The molecular mechanisms underlying the formation of at least some of the cryptic rearrangements are being elucidated, some of which are thought to arise following unequal meiotic recombination mediated by the proximity of nonallelic segmental duplications particularly in the pericentromeric and telomeric regions (Emanuel and Shaikh, 2001). However, alternative mechanisms may operate. For example, the presence in some mothers with children with Angelman’s syndrome and del(15)(q11-q13) of a submicroscopic heterozygous inversion at the regions defined by flanking segmental duplications has been proposed to represent “an intermediate estate” that facilitates the formation of a deletion in an offspring (Gimelli et al., 2003). Clearly, genomic organization has a pivotal role in determining the position and frequency of specific microdeletion syndromes and the several models proposed to date may not be exhaustive (Emanuel and Shaikh, 2001).

The discussion so far has focused on techniques used to identify imbalances at specific genomic locations. Leading up to the publication of the advanced draft assembly of the Human Genome Sequence in 2001, attention switched to the innovative use of fully sequenced clones (principally Bacterial or P1 Artificial Chromosomes, BAC/PACs). A method called “Array-CGH” was developed (see Article 55, Polymorphic inversions, deletions, and duplications in gene mapping, Volume 1) in which selected BAC/PAC clones (with an average insert size of ~150 Kb) were selected sequentially at contiguous genomic locations ~ 1 Mb apart and then spotted onto glass slides to act as targets for a comparative genomic hybridization (CGH) experiment using a probe mixture comprising differentially labeled DNA derived from both normal and clinically affected individuals (Fiegler et al., 2003). This approach opened up the possibility of detecting copy number changes either throughout the genome with a resolution only constrained by the density of the clones selected (i.e., at 1 Mb or complete coverage using tiling path clones (Ishkanian et al., 2004)) or at high density but focused on specific genomic regions.

A number of studies utilizing Array-CGH have illustrated the potential of this approach but has also provided unexpected insights into the number and distribution of large genomic imbalances distributed throughout the genome that may turn out to be nonpathogenic polymorphisms (see below). Array-CGH focused on specific chromosomal regions has been used to define critical deletion regions for specific phenotypes such as Congenital Aural Atresia in patients with 18q22.3-18q23 deletions (Veltman et al., 2003), 1p36 deletions (Yu et al., 2003), and imbalances involving 17p11-p12 (Shaw et al., 2004). Targeted Array-CGH led to the refinement of a critical deletion interval in 8q12 in some patients with CHARGE syndrome (a developmental disorder with a nonrandom pattern of congenital abnormalities). Sequence analysis of genes found in the deleted segment found mutations in CHD7 in 10 out of 17 patients with CHARGE syndrome (Vissers et al., 2004), thereby illustrating the utility of Array-CGH for the identification of some novel disease-related gene loci.

Array-CGH has also been used in whole-genome scans using BACs at 1-Mb resolution on chromosomally normal patients with idiopathic developmental delay or mental retardation with and without associated congenital abnormalities. The first studies reported similar results (Table 2). Shaw-Smith et al. (2004) and Vis-sers et al. (2003) used 1-Mb Array-CGH on similar patient groups and showed that ~25% carried genomic imbalances, that is, deletions or duplications. One-third of these imbalances were subsequently found to be carried by a clinically normal parent and were considered to be polymorphic copy number changes. However, the remaining two-thirds were novel de novo microdeletions/duplications (see Article 52, Algorithmic improvements in gene mapping, Volume 1), which further studies will determine whether they are directly responsible for the clinical phenotypes observed.

Table 1 Microdeletion syndromes

Chromosome Locus/gene Syndrome
1p36.3 DVL1 Monosomy 1p36
2p21 SIX3 Holoprosencephaly 2
2q13 NPHP1 Nephronophthisis 6
2q22.3 ZFHX1B Mowat-Wilson
2q37 HPE6 Holoprosencephaly 6
2q37 AHO Albright hereditary osteodystrophy
2q31-32 HOXD9; HOXD13 Synpolydactyly
EVX2
3p23 VHL Von-Hippel Lindau
3q23 BPES Blepharophimosis-ptosis-epicanthus-inversus
4p16 WHS Wolf-Hirschorn
4q25 PIX2 Rieger
5p15 CDC Cri-du-Chat
5q13.2 NIPBL Cornelia de Lange
5q22.2 APC Familial adenomatous polyposis
5q35.2 MSX2 Parietal Foramina
5q35 NSD1 Sotos
7p21 TWIST1 Saethre-Chotzen
7p13 GLI3 Greig cephalopolysyndactyly
7q11.23 ELN Williams
7q36.3 SHH Holoprosencaphaly 3
7q36.3 HLXB9 Sacral agenesis
8q12 CHD7 CHARGE
8q24.1 EXT1 Langer Giedion
8q24.1 TRPS1 Trichorhinophalangeal syndrome
9q22.3 PTCH Holoprosencaphaly 7
10p15 GATA3 Hypoparathryoidism, deafness, renal dysplasia
10p13 DGSII DiGeorge
11p15.5 IGF2 Beckwith-Wiedemann
11p13 PAX6 Aniridia
11p13 WT1 Wilms tumor
11p13 WAGR Wilms, Aniridia, genital anomalies, retardation
11p11.2 EXT2 Multiple exostoses
11p11.2 ALX4 Parietal Foramina contiguous gene deletion
11q23-24.1 MLL/ETS1 Jacobsen
12q24.1 PTPN11 Noonan
13q14 RB1 Retinoblastoma
13q32 ZIC2 Holoprosencaphaly 5
14q12-13 PAX9 Autosomal dominant hypodontia
15q13 SNRPN/UBE3A Prader-Willi/Angelman
16p13.3 CREBBP Rubinstein-Taybi
16p13.3 TSC2 Tuberous sclerosis 2
16p13.3 PKD1 Polycystic kidney disease
17p13.3 LIS1 Miller-Dieker
17p11.2 RAI1 Smith Magenis
17p11.2 PMP22 Hereditary neuropathy with liability to pressure
palsy
17q11.2 NF1 Neurofibromatosis 1
17q12-21 SOST Van Buchem disease
18p11.3 TGIF Holoprosencaphaly 4
20p11.23 JAG1 Alagille
21q22.3 TMEM1 Holoprosencaphaly 1

Table 1 (continued)

Chromosome Locus/gene Syndrome
22q11.2 TUPLE1;TBX1 DiGeorge I
22q13.3 ProSAP2 Monosomy 22q13.3
Xp22.33Yp11.32 SHOX Leri-Weill dyschondrosteosis, Madelung
deformity, short stature
Xp22.33 STS Steroid sulfatase
Xp22.33 KAL1 Kallmann
Xp22.33 MLS Micopthalmia and linear skin defects
Xp21.2 DAX1 Adrenal hypoplasia congenita
Xp21 GK Glycerol kinase deficiency
Xp21 DMD Duchenne muscular dystrophy
Xq22 PLP Pelizaeus-Merzbacher disease
Xq25 XLP X-linked lymphoproliferative disease
Xq26.2 ZIC3 X-linked hererotaxy
Yp11.32 SRY Ambiguous genitalia, sex reversal

Two subsequent papers reported large-scale copy number variations (deletions and duplications) ranging in size from 100Kb to 2Mb in two ethnically and geographically distinct populations of normal control individuals (Table 2). Iafrate et al. (2004), using Array-CGH, reported 255 variable loci in 55 individuals compared to Sebat et al. (2004) who used an oligonucleotide array and demonstrated 76 genomic changes in 20 individuals. Interpretation of what constitutes a pathogenic genomic imbalance compared to a polymorphism, particularly those involving microdeletions, will only be possible once the technical parameters, including the choice and density of clones used in the array, as well as an understanding of the sequences involved in “common” imbalances are fully understood (Carter, 2004).

Table 2 Number and distribution of large genomic imbalances in affected and control populations

A. Clinically affected populations Study Method No. Deln No. Dups Total % imbalances
Vissers et al. (2003) A-CGH 3 2 20 25
Shaw-Smith et al. (2004) A-CGH 7 5 50 24
Total (de novo) 10* (14%) (8) 7* (10%) (1) 70 24
B. Clinically normal populations
Study Method No studied Resolution (Kb) Total No (%) imbalances
Sebat etal .(2004) ROMA 20 >100 221 CNPs (ave: 11.5) 70 (32%)
Iafrate et al. (2004) Array-CGH 55 >150 255 LCVs (ave: 6.5) 142 (55%)

In conclusion, our understanding of the clinical effects and mutational mechanisms associated with genomic imbalances, particularly deletions, has come a long way since the first cytogenetic karyotype/phenotype correlations in the 1960s and the descriptions of the first microdeletion syndromes in the 1980s. The future development of high-density genomic arrays promises further insights into the complex structure of the “normal” human genome, from which novel microdeletion syndromes will emerge providing new diagnostic strategies and insights into disease-associated human gene(s).

Next post:

Previous post: