Elsevier

The Lancet

Volume 350, Issue 9090, 22 November 1997, Pages 1511-1515
The Lancet

Early Report
Strong clustering and stereotyped nature of Notch3 mutations in CADASIL patients

https://doi.org/10.1016/S0140-6736(97)08083-5Get rights and content

Summary

Background

CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoencephalopathy) is commonly overlooked or misdiagnosed owing to its recent identification and its variable mode of presentation. The defective gene in CADASIL is Notch3, which encodes a large transmembrane receptor. To set up a diagnostic test and to delineate the NotchS domains involved in CADASIL, we undertook mutations analysis in this gene in a group of CADASIL patients.

Methods

50 unrelated patients with CADASIL and 100 healthy controls were screened for mutations along the entire Notch3 sequence, by means of single-strand conformation polymorphism, heteroduplex, and sequence analysis.

Findings

Strongly stereotyped mis-sense mutations, located within the epidermal-growth-factor-like (EGF-like) repeats, in the extracellular domain of Notch3, were detected in 45 patients. Clustering of mutations within the two exons encoding the first five EGF-like repeats was observed (32 patients). All these mutations lead to loss or gain of a cysteine residue and therefore to an unpaired number of cysteine residues within a given EGF domain. None of these mutations was found in the 100 controls.

Interpretation

Because of the strong clustering and highly stereotyped nature of the pathogenetic mutations detected in CADASIL patients, an easy and reliable diagnostic test for CADASIL is feasible. The findings suggest that aberrant dimerisation of NotchS, due to abnormal disulphide bridging with another NotchS molecule or with another protein, may be involved in the pathogenesis of this disorder.

Introduction

An autosomal dominant disorder that causes stroke has been reported in several families since 1977, under various names such as hereditary multi-infarct dementia.1, 2, 3, 4 In 1991, analysis of a very large pedigree allowed us to define the clinical, neuroimaging, and genetic characteristics of this disorder, now known by the acronym CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoencephalopathy).5, 6 The disorder most commonly manifests as transient or permanent ischaemic strokes, which occur at a mean age of 45 years (range 27–65). Most of these strokes are classic lacunar infarcts that arise in the absence of hypertension or any other recognised vascular risk factor. Subcortical dementia, in all cases associated with pseudobulbar palsy, is the second commonest clinical manifestation of CADASIL. All individuals with symptoms have prominent signal abnormalities on brain magnetic resonance imaging (MRI): on T2-weighted images there are symmetrical and extensive hypersignals within the cerebral white matter; and on Tl-weighted images there are well-delineated hypointense lesions highly suggestive of small infarcts within the deep white matter and basal ganglia (figure 1). MRI abnormalities can also be detected before the appearance of symptoms. The arteriopathy underlying the disorder is neither atherosclerotic nor amyloid and involves the media of small cerebral arteries. However, lesions may be observed to a lesser extent in extracerebral arteries, including skin arterioles. Ultrastructural examination reveals abnormal patches of a granular osmiophilic material within the basal membranes of vascular smooth-muscle cells7, 8 (figure 1). Mapping of the mutated gene on chromosome 19 provided an indirect genotypic diagnostic test for the disorder, based on genetic linkage analysis.6

Since 1991, an increasing number of families with clinical and neuroimaging features similar to those previously described have been reported from all over the world; more than 200 such families have been referred to our department. Detailed clinical and genotyping analysis of these families showed that stroke was not the sole phenotype of CADASIL. Other commonly observed symptoms are attacks of migraine with aura (about 30% of patients) and mood disorders (about 20%).9, 10

The possibility of a diagnosis of CADASIL may readily occur to a clinician aware of this disorder when investigating a large family with several affected members; the diagnosis can then be confirmed by means of linkage analysis. However, the hereditary nature of the disorder present in a CADASIL patient is often not apparent, since the penetrance of this disorder is complete only at around 50–60 years of age. Indirect genotypic diagnostic testing is precluded when family members are not available for linkage analysis. CADASIL diagnosis may also be overlooked owing to its variable mode of presentation, often leading to erroneous diagnoses (such as multiple sclerosis, Alzheimer's dementia, or Binswanger's disease) and therefore to inappropriate care for the patient.

In 1996, we identified Notch3 as the defective gene in CADASIL.11 Members of the Notch gene family encode evolutionarily conserved transmembrane receptors, and are involved in cell fate specification during embryonic development.12 The Notch3 gene includes 33 exons encoding a protein of 2321 aminoacids. Its extracellular domain contains 34 epidermal-growth-factor-like (EGF-like) repeats (unpublished results). From data on Drosophila, this extracellular domain seems to be involved in ligand binding, whereas the intracellular domain carries the intrinsic signal-transducing activity.13, 14 CADASIL gene identification provided the basic information needed to set up a direct genotypic diagnostic test. However, Notch3 is a large gene, and direct genotypic screening has proved difficult in disorders due to mutations in large genes except when mutations are clustered.15 As a first step to set up a reliable and feasible diagnostic test, we undertook Notch3 mutations analysis in a group of 50 unrelated patients with CADASIL, searching for potential clustering; we also tested 100 control individuals so that we could characterise Notch3 polymorphisms to be differentiated from pathogenetic mutations.

Section snippets

Methods

We screened 50 unrelated caucasian patients referred because of a strong suspicion of CADASIL, including 42 patients from Europe, seven from the USA, and one from North Africa. The patients were selected on the basis of the following criteria: a history of recurrent subcortical ischaemic strokes, migraine with aura, vascular dementia of unknown aetiology, or a combination of these features; the presence of hypersignals on T2W1 cerebral MRI and images suggestive of small, deep infarcts on brain

Results

Screening of the entire Notch3 coding sequence detected heterozygous mutations in 45 of the 50 patients tested (table 1). All these mutations are mis-sense mutations located within the EGF-like repeats. They are strongly stereotyped; all of them are predicted to replace the wild-type aminoacid with a cysteine residue (37 patients) or to replace one of the six highly conserved cysteine residues with another aminoacid (eight patients). None of the mutations involving cysteine residues was

Discussion

We identified pathogenetic mutations in Notch3 in 45 of 50 unrelated patients with CADASIL. All these mutations are located within the extracellular domain, with strong clustering within exons 3 and 4, which encode the first five EGF repeats. These mutations led in all cases to an odd number of cysteine residues in the EGF-like repeats of the Notch3 protein.

These results have strong implications for diagnostic testing, since screening of only two of the 33 exons could identify the pathogenetic

References (24)

  • P Sourander et al.

    Hereditary multi-infarct dementia

    Acta Neuropathol

    (1977)
  • P Davous et al.

    Démence sous corticale familiale avec leucoencéphalopathie artériopathique: observation clinicopathologique.

    Rev Neurol (Paris)

    (1991)
  • Cited by (0)

    View full text