Article Text

Download PDFPDF

Screening British CADASIL families for mutations in theNOTCH3 gene
  1. R P J DE LANGE*,
  2. J BOLT,
  3. E REID§,
  4. R DA SILVA,
  5. D J SHAW,
  6. D M ST CLAIR*
  1. * University of Aberdeen, Medical School, Department of Mental Health, Foresterhill, Aberdeen AB25 2ZD, UK
  2. University of Aberdeen, Medical School, Department of Molecular and Cellular Biology, Foresterhill, Aberdeen AB25 2ZD, UK
  3. Hairmyres Hospital, Lanarkshire, UK
  4. § Department of Medical Genetics, Addenbrooke's Hospital, Cambridge CB2 2QQ, UK
  5. Department of Neurology, Institute of Neurosciences, Southern General Hospital, Glasgow G51 4TF, UK
  1. Dr St Clair, d.stclair{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Editor—CADASIL (Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leucoencephalopathy) is a hereditary form of multi-infarct vascular dementia.1Clinical symptoms often present in middle adult life (30-50 years of age) and include recurrent subcortical ischaemic strokes, migraine with or without aura, major psychiatric symptoms, and dementia. Magnetic resonance imaging (MRI) shows high intensity signal lesions, often confluent, and areas of cystic degeneration of the subcortical white matter and basal ganglia. Pathological examination shows multiple, small, deep cerebral infarcts, leucoencephalopathy, and a non-atherosclerotic, non-amyloid angiopathy involving mainly the small, deep, perforating cerebral arterioles. Severe alterations of vascular smooth muscle cells are evident on ultrastructural analysis.

The term CADASIL was adopted after linkage of French families with these symptoms to chromosome 19,2 3 but families with many of the features of CADASIL had been described by Worster-Droughtet al 4 5 in the 1930s as familial presenile dementia with spastic paralysis, by Sourander and Walinder6 as hereditary multi-infarct dementia, and by Stevens et al 7 as chronic familial vascular encephalopathy. In 1996, the responsible gene was identified as NOTCH3,8 a member of the Notch family of signalling proteins originally identified inDrosophila.9

Notch and Notch homologues control the ability of non-terminally differentiated cells to respond to differentiation/proliferation signals through local cell interactions.10 They are transmembrane proteins with distinct extracellular and intracellular domains. Notch is activated by binding of a ligand to the extracellular so-called epidermal growth factor (EGF) repeats.11 This is thought to release the intracellular domain which translocates to the nucleus to regulate the transcription of genes that ultimately determine cell fate.12 13

To date, 26 separate mutations have been found inNOTCH3, 24 as described by Joutelet al 14 in the French families and two additional mutations from American families, as reported by Meeks et al. 15 Twelve of these mutations are clustered in exon 4. All of these mutations predict the introduction or replacement of cysteine residues in the extracellular EGF repeat domain. CADASIL has also been reported in Dutch,16 German,17 Swiss,18Italian,19 20 American,21 22 and Japanese families.23 We report here the results of linkage analysis and screening for mutations in British families with a diagnosis of CADASIL.

DNA was available from four multiplex families from the central belt of Scotland, all of Scottish ancestry, a family from south west England of English ancestry, and one further isolated subject from Scotland with suspected CADASIL. We classified the diagnosis of CADASIL into definite, when there was neuropathological confirmation or evidence of genetic linkage to chromosome 19p12 or both, and probable, where there were clear clinical symptoms and MRI findings typical of CADASIL plus a positive family history. The clinical-demographic details and results described in the text below are summarised in table 1. In spite of extensive genealogical investigations we were unable to find a common ancestor for any of the families described.

Table 1

Diagnostic details for families studied

Families 1 and 2, both Scottish, had sufficient meioses available to perform analysis of genetic linkage to chromosome 19p12. We genotyped the families on a Perkin Elmer Applied Biosystems (PE ABI) 377 automated genotyper with 5′ FAM labelled primers for six polymorphic microsatellite markers, D19S226, D19S411, D19S885, D19S199, D19S923, and D19S841 flanking the NOTCH3gene (GDB accession numbers 188569, 199752, 608544, 182271, 611676, and 593357, respectively), and analysed the results using Perkin Elmer Genescan software (version 2.1).

Clear genetic linkage to chromosome 19p12 was found in both families. Family 2 was originally reported not to be linked to chromosome 19.24 However, further characterisation of the phenotypes in these families by MRI scanning, which had not been performed at the time of the initial reporting, required us to reclassify several cases showing key recombinants. We also determined haplotypes at NOTCH3 in the other multiplex families and found the same patterns of microsatellite allele sizes for families 3 and 4.

We then sequenced exon 4 of the NOTCH3 gene (GDB accession number AF058883) in all our families, by automated sequencing of PCR products. After checking size and yield on 1% agarose gel electrophoresis, PCR products were purified using Centrikon™ columns. Purified PCR products were sequenced by automated cycle sequencing using PE ABI BigDye™ chemistry. The products were run on polyacrylamide gels on a PE ABI 377 automated sequencer and analysed using PE ABI Factura (2.0.1) and Sequence Navigator (1.1) software.

The CADASIL disease mutation was found in exon 4 in all families in positions described earlier14 (table 1). The sequencing results corresponded with the genetic linkage analysis for all affected and unaffected members of families 1 and 2. The results also agreed with the predicted affected status based on clinical and MRI data for members of families 3-6. The C499T mutation coding for the R143C amino acid substitution is shared between families 3 and 4. Since they share the same haplotype it is likely that the mutation has the same ancestral origin.

In addition to the mutations, the sequencing also detected reported polymorphisms.14 For example, the G684A polymorphism in exon 4 was found with a frequency that matches the 0.17 described by Joutel et al. 14

In conclusion, our results confirm the broad geographical occurrence of CADASIL in Europe. It is conceivable that this rare Mendelian type stroke is still largely underdiagnosed and mutation analysis will lead to an increase in the number of diagnosed cases.


We thank Drs Durward and Bone, and Professor Behan (Southern General Hospital, Glasgow) and Dr J Gibson (Derriford Hospital, Plymouth) for access to DNA from families. This work was funded by the Stroke Association, Smith Charities, and Davidson Bequest.