1.4,3.1 Haploinsufficiency.
DiGeorge syndrome usually occurs sporadically, but it may be inherited in an
autosomal dominant fashion (Rohn et al 1984). The general distinction of dominant mutations into loss of function, or gain of function categories (Wilkie A. 1994), suggests that all the CATCH 22 syndromes fall into the former, due to
haploinsufficiency of a gene or genes from 22ql 1. The study of chromosome deletion syndromes has isolated a number of haploinsufficient genes. Characterisation of some of these genes has implicated possible mechanisms by which abnormal gene dosage will result in an abnormal phenotype (reviewed in Fisher & Scambler 1994). One such example of a haploinsufficient gene is PAX-6, which maps to chromosome
1 Ip 13, and when deleted results in aniridia, a congenital disorder resulting in complete or partial absence of the iris (Ton et al 1991). PAX-6 plays an important role in development of the eye, by acting as a transcription factor. The authors involved in the isolation of PAX-6, suggest that inadequate gene dosage may impair developmental process that have stringent requirements for certain gene products at specific times and locations (such as is known for other homeobox containing genes), during the set up of morphogenetic fields. This is the simplest effect of
haploinsufficiency, and stems from insufficient protein production.
The protein products of many haploinsufficient genes form intermolecular protein complexes. If exact stoichiometry is required for these complexes then this will be disturbed by the loss of one copy of the gene product, a phenomenon that may be particularly likely when transcription factors are involved. This may be true for many other molecules, as seen with the structural protein ankyrin. An exact molar ratio with interacting proteins, in this case a 1:4 ratio with spectrin, seems to be critical for ankyrin to function (Davis & Lux. 1989).
Another example where levels of gene dosage are critical to the function of the encoded protein may occur in Greig syndrome (a disorder affecting limb and craniofacial development), resulting from the loss of a copy of the zinc-finger gene
GLI-3, which maps to chromosome 7pl3 (Vortkamp et al 1991). Fisher & Scambler argue that although there is not much data available concerning the biochemistry of
genes. In Drosophila, Kruppel, when present in high concentrations, forms
homodimers which repress further transcription. In low concentrations the monomer appears to act as a transcriptional activator, using the same DNA target sequences as are recognised by the repressor, suggesting that this regulatory system is exquisitely dosage sensitive.
These common modes of action may be the route by which a haploinsufficient gene results in an abnormal phenotype. Furthermore, it is not just transcriptional factors that require specific dosages to function, for example, haploinsufficiency for receptors and signalling molecules have been implicated in the aetiology of Hirschprung’s disease.
Most haploinsufficiency syndromes exhibit a wide variety of phenotypes between affected individuals, even within families. This may suggest that the non-deleted gene or genes, manifest a level of genetic control over the phenotype. Although this has yet to be documented for a haploinsufficient syndrome, examples of genetic control by the mutated loci in recessive disorders have been reported. Cystic fibrosis is caused by mutations in the CFTR gene, and results in a severe phenotype characterised by chronic sino-pulmonary disease, pancreatic exocrine insufficiency, and elevated electrolyte levels in sweat. However, the common missense mutation R117H is found to be associated with phenotypes of decreasing severity, resulting from the mutation occurring on two different chromosomal backgrounds. Alternative splicing of CFTR
exon 9, due to variants of the intron 8 splice acceptor, was attributed as a causal role of phenotype variability (Kiesewetter et al 1993).
Other background genetic effects may occur due to a second locus, sometimes known as “non-allelic non-complementation”. Hemizygosity might also increase the
sensitivity of the embryo to environmental insults, especially those that directly alter levels of gene expression, a mechanism likely to produce gross developmental abnormalities. One other possibility that cannot be ignored is that of the “second hit hypothesis” (Knudson et al 1985), a well known pathway to those who study
hereditary cancers. If a second hit occurred in hemizygous cells, after developmental lineage splits, a specific populations of cells may go on to form specific abnormalities, while those in another lineage group would be normal. An early hit could give rise to
The role of chance, although hard to scientifically demonstrate, must not be ruled out as an effector on the phenotype of aneusomic disorders.
1,4.3.2 A Contiguous Gene Syndrome v A Single Gene Disorder.
As previously mentioned, over 90% of CATCH 22 patients are deleted for the s c ll.l loci, indicating a common deletion of approximately 2Mb. This obviously suggests that these individuals will be hemizygous for any of the genes that lie within this interval. Due to the size of the common deletion and the variability of phenotype it was postulated that DGS is a contiguous gene syndrome.
Approximately ten years ago the phrase contiguous gene syndrome was coined to describe the general features of a group of syndromes which include the WAGR complex and Langer-Giedion syndrome (Schmickel R. 1986). The best way to describe these features is to discuss one of the syndromes within this group.
Individuals who show abnormalities associated with the WAGR complex, manifest Wilms’ tumour, aniridia, genitourinary abnormalities and mental retardation. The finding that patients with a Wilms’ tumour had a ten thousand fold increase risk of aniridia, compared to the general population, was first noted in 1964 (for review see Tay J. 1995). Molecular mapping, after the further association with genito-urinary anomalies, indicated that many patients had cytogenetically visible deletions on
chromosome 1 Ip 13. The observation that defines this as a contiguous gene syndrome was that more than one gene within this region contributes to the phenotype. Loss of heterozygosity for WTI, due to a “second hit” in the remaining copy, results in the formation of a Wilms’ tumour. A large deletion can also result in the loss of PAX-6,
which maps proximally to WTI, resulting in aniridia. However, chromosome
rearrangements, such as an interstitial deletion of 1 Ip 13, within the 3’ coding region of PAX-6 (Ton et al 1991), results in the less severe phenotype of aniridia alone. If DGS belonged to this family of disorders, a large deletion would result in the loss of all the contiguous genes, and would lead to the most severe phenotype. The phenotype resulting from a small interstitial deletion would correlate to the gene loss within this deletion, and unbalanced translocations may demonstrate a more severe phenotype, the effect of superimposing loss or gain of other autosomal genetic
material against loss of 22ql 1 genetic material (Emanuel 1988). It would then be expected that the less severe and specific phenotype of an individual with a balanced translocation would result from the interruption of sequences, a putative gene, within the breakpoint on 22ql 1.
Evidence against this theory came from two studies of families diagnosed to have CATCH22 symptoms. The proband of one family was diagnosed to have complete DGS, with severe defects including lAA, VSD and hypocalcaemic seizures (Wilson et al 1991). The female proband has three sibs; one unaffected sister, one brother with VSD, and another with coarctation of the aorta and patent ductus arteriosus.
Cytogenetic analysis and quantitative Southern analysis of the three affected sibs and their parents was performed using a probe isolated from the marker HP500. The results showed that each affected child had inherited the normal chromosome 22 from their father, and an interstitial deletion on the chromosome 22 inherited maternally. Retrospective diagnosis of the “normal” mother indicated mild facial dysmorphism which was also seen in the two brothers, however, that is the only real abnormality shared within the family even though they have an identical deletion. A similar finding was reported in a mother and daughter with VCFS. The mother had surgery as a child for a cleft palate and a congenital heart defect. She also had facial abnormalities consistent with VCFS and was noted to be a low achiever at school. The daughter was the second child, as a previous son had died in infancy as an unexplained cot death, although it had been noted that he had a heart murmur in the neonatal period. The second child had developmental delay, absent speech (although her occasional vocalisations were said to be nasal in sound), and facial features similar to her mother. No heart defects or clefting were observed. Molecular analysis showed again, an identical deletion within^2ql 1 in the commonly deleted region (Holder et al 1993).
The breakpoint of the balanced translocation in patient ADU maps to the DGCR, a translocation also present in the proband’s mother. If DGS is a single gene disorder then it would be a reasonable hypothesis to suggest that the gene involved is disrupted by the breakpoint, and this single gene is deleted or rearranged in all other affected
individuals. The phenotypic variability would then be explained by other genetic influences.
One initially exciting finding was that nearly 70% of all inherited cases of CATCH 22 were maternally derived (reviewed in Demczuks & Aurias 1995). However
uniparental disomy has been reported for both maternal and paternal chromosome 22s, and it is thus unlikely that this region is subjected to imprinting (Schinzel et al
1994, Peter et al 1995). One explanation for this significant departure from
randomness is an ascertainment bias due to the possibility that affected fathers have a lesser reproductive success, due to sociological factors or decreased fertility.
Mapping of genes to the DGCR will thus be the first step to the identification of those involved in the aetiology of DGS and the other CATCH 22 syndromes.
1.4,3.3 Positional Effects.
Although the ADU breakpoint may disrupt a gene of primary importance in DGS, another hypothesis is that the breakpoint exerts a positional effect on a gene or genes in the vicinity of the breakpoint.
Putative position effects have been found in human genetic diseases for which the disease “causing” gene has been identified. These have included campomelic dysplasia (CD), a congenital skeletal malformation syndrome associated with sex- reversal, aniridia and Greig Syndrome. In each case a translocation breakpoint has been found that does not disrupt the genes S0X9, PAX6 and GLI-3 respectively, but maps between 10 and 185kb away.(Foster et al 1994, Fantes et al 1995, Vortkamp et al 1991).
Three types of position effect have been hypothesised (reviewed in Milot et al 1996). Position-effect variegation (PEV), first observed in Drosophila and yeast, occurs primarily through relocalisation of a gene into a heterochromatic environment which remains highly condensed throughout the cell cycle. This can lead to a shut down of gene expression. Although the exact mechanism is unknown, it is thought that condensed chromatin spreads from the transcriptionally inactive heterochromatin to silence the genes that would normally lie in a euchromatic environment (reviewed in Bedell et al 1996). The spreading can occur over a large distance and often results in a gradient effect, where genes closest to the rearrangement breakpoint are more
adversely affected. In mice, a PEV-like phenotype has been observed with some X- autosome translocations involving autosomal pigment (reviewed in Milot et al 1996). Spreading of X-chromosome inactivation from the X-inactivation centre, results in the loss of expression of the genes on the autosome, adjacent and near to the breakpoint, in a gradient type effect. PEV in humans has rarely been documented although one similarity has been seen recently with a subtype of acute myeloid leukaemia (AML), (reviewed in Roth S. 1996). A recurrent 8; 16 translocation associated with AML results in the formation of a fusion protein, between A/OZ (mapping to 8pl 1) and
CREBAÀnding protein, CRB (mapping to 16pl3) (Borrow et al 1996, Reifsynder et al
1996). MOZ is a homologue of SAS2, a protein important in transcriptional silencing, and both MOZ and SAS2 are thought to function as histone acetyltransferase enzymes. The highly condensed structural properties of heterchromatin rely on direct
interactions between multicomponent silencing factors and nucleosomes.
Nucleosomes are the basic subunits of chromatin, consisting of an octomer of histone proteins. Histones have been found to be hypoacetylated, and it is thought that acétylation may loosen their interactions resulting in a more relaxed DNA
configuration. However, transcriptional silencing also involves acétylation of other unknown proteins. It is thought that the in-frame fusion protein derived from MOZ and the multiple signalling protein CRB, leads to the creation of an abnormal
acétylation pattern of histones and other regulatory proteins that will interfere with gene activation and, in this case, result in cellular transformation.
Regulatory position effects (RPE) may occur when a gene and/or its regulatory regions become situated adjacent to another gene as a result of a genomic
rearrangement. The gene may come into close proximity with a novel regulatory region, resulting in up or down regulation of expression. Alternatively, the new location of the regulatory region may allow interaction with other genes, again resulting in abnormal transcription. It is known the RPE exist in mammals and can cause disease. Many examples or RPE result in activation or deregulation of cellular proto-oncogenes, and are associated with tumour induction, such as the coupling of the MYC oncogene with regulatory regions of the immunoglobulin genes. This results in Burkitt’s Lymphoma (Joos et al 1992).
The third type of position effect can be envisaged when a deletion or rearrangement removes part, or all, of the regulatory elements. This deletion position effect (DPE) may act even when these sequences are a large distance from the gene whose expression they regulate. Examples of DPE have been seen in thalassaemias and hereditary persistence of foetal haemoblobin, where deletions located far upstream or downstream have gene-specific effects on globin gene expression (reviewed in Milot
et al 1996).
Translocations as seen in campomelic dysplasia, aniridia and Greig syndrome may result in disease through any one of the described position effects. Evidence for the role of PEV, RPE and DPE is currently being reported through the use of transgenic animals (reviewed in Milot et al 1996), however, information is needed on the sequences disrupted by the breakpoints to allow the exact mechanisms involved in each case to be identified. Recently a testis specific cDNA was isolated that spanned a translocation breakpoint on human chromosome 17q24.3-q25.1 in a patient with CD and sex-reversal (Ninomiya et al 1996). Mutations have been identified in the gene
S0X9, the chromosome 17q24 gene responsible for CD. The testis-specific transcript was found to map proximally to S0X9. The authors suggested that this novel cDNA may function as a regulatory mRNA (as no long ORF was detected from the
sequence), and play a role in the sex-reversal seen in the CD patients. In this intriguing example, the translocation may exert a RPE on the proximal gene S0X9.
Further discussion of this cDNA can be found in 4.1.4.