Alagille syndrome (AGS) is an autosomal dominant disorder characterised by the abnormal development of the liver, heart, skeleton, eye, and face. AGS is a highly variable disorder, and the severity varies from asymptomatic gene carriers through to lethality due to inoperable cardiac complications or end-stage liver disease. The AGS critical region was located on chromosome 20p between p12.1 and p11.23. It is thought that one gene located in this region causes this disorder.
At the time the project was commenced the gene mutated in AGS had not been identified so we sought to refine the location of the AGS locus by haplotype and linkage analysis using eight polymorphic CA repeat markers in nine AGS-affected families with 11 affected individuals (Chapter 3). Using MLINK (parametric, two-point) and GeneHunter (non-parametric, multipoint) analysis, the highest positive LOD scores were seen for D20S162 (Z = 0.72), D20S189 (Z = 0.75) and D20S186 (Z = 0.76) at θ= 0.00. We did not demonstrate significant linkage to any of the markers studied. However, haplotype analysis from Family 1 indicated that the disease gene was likely to be located close to the microsatellite D20S894.
The identification of Jagged1 as the gene mutated in AGS shifted the focus of this project to mutation analysis (Chapter 4). Using single stranded conformational polymorphism analysis and denaturing high performance liquid chromatography, we have screened 43 Australian AGS affected individuals from 35 families for mutations within Jagged1. Twenty four Jagged1 mutations (69%) were identified in the 35 affected families. The twenty two different types of mutations include six small deletions (27.3%), three small insertions (13.6%), six missense mutations (27.3%), three nonsense mutations (13.6%) and four splice site mutations (18.2%). These mutations are spread across the entire coding sequence of the gene and most are localised to highly conserved motifs of the protein predicted to be important for Jagged1 function. The deletions, insertions and nonsense mutations (55% of total) are predicted to result in truncation of the JAG1 protein. Approximately 14% of the mutations identified arose de novo and approximately 27% were inherited, however, the origin of inheritance was not defined for over half of the mutations identified in this study. While most mutations were unique, (63.6%), 8 had been identified previously.
In Chapter 5, five mutations occurring across splice sites were studied further to examine their effects upon Jagged1 mRNA processing. A splice donor site mutation in intron 11 was shown to cause the aberrant splicing of Jagged1 mRNA, leading to the premature termination of translation in exon 12. A deletion in this same splice donor site also led to the aberrant splicing of exon 11. Splice donor mutation in introns 3 and 6 resulted in the removal of exons 3 and 6 respectively, and consequent premature stop codons in exons 4 and 7. A splice acceptor site mutation in intron 13 led to two aberrantly spliced Jagged1 products, one with an in-frame deletion of exon 14, and the other missing both exons 14 and 15. These results showed that splice site mutations result in the aberrant splicing of Jagged1 mRNA and that RNA analysis may identify Jagged1 mutations previously missed by conventional screening methods.
In Chapters 6 and 7, the consequences of several Jagged1 mutations identified in the Australian AGS population were investigated. Transfection of COS-7 and HeLa cells with wild-type and mutant Jagged1 genes demonstrated that wild-type protein was localised to the cell surface, whereas both missense and truncated JAG1 proteins showed an intracellular localisation. However, further studies showed that the truncated molecules were also secreted from the cell. Co-localisation studies indicated that most of the missense mutant JAG1 proteins were being trapped within the ER, although some also reached the Golgi. Digestion of these missense mutant JAG1 proteins with endoglycosidases, showed that glycoslyation was different to wild-type in two of the missense mutant proteins. These results are consistent with the proposal that AGS is caused by haploinsufficiency for wild-type JAG1 at the cell surface in those individuals carrying missense Jaggedl mutations. Whether or not the truncated secreted mutant proteins bind to Notch and activate the Notch signalling pathway was investigated by establishing a co-culture system utilising JAG1 expressing COS-7 cells and Notchl expressing JurkatT-cells (Chapter 7). JurkatT-cell clumping was observed in the presence of wild-type JAG1 protein and to a lessor extent in the presence of both 896JAG1 and 1898JAG 1 proteins. Real-time RT-PCR showed that the 896JAG1 protein caused an increase in Hry and HERP2 gene expression similar to that caused by wild-type JAG1. Thus, a truncated, secreted JAG1 mutant was shown to activate the Notch signalling pathway in the Jurkat T-cells.
The results of these studies are consistent with the proposal that haploinsufficiency for wild-type Jagged1 is responsible for the AGS phenotype. However, an additional dosage-dependent mechanism may also be involved in those individuals carrying truncating Jagged1 mutations. A truncating mutant JAG1 protein was shown to activate Notch signalling in a similar manner to wild-type, but possibly in a diminished capacity. This, may ultimately cause a dominant negative effect on the Notch signalling pathway.