You are here

Article Text

Article menu

Download PDFPDF

Research paper
CADASIL with cord involvement associated with a novel and atypical NOTCH3 mutation
  1. Paul Bentley1,2,
  2. Tao Wang3,
  3. Omar Malik2,
  4. Richard Nicholas2,
  5. Maria Ban4,
  6. Stephen Sawcer4,
  7. Pankaj Sharma1,2
  1. 1Imperial College Cerebrovascular Research Unit, Imperial College London, Charing Cross Campus, London, UK
  2. 2Department of Neurology, Imperial College Healthcare NHS Trust, London, UK
  3. 3Medical Genetics Research Group, University of Manchester, Manchester, UK
  4. 4Molecular Biology Department, Cambridge University, Cambridge, UK
  1. Correspondence to Dr P Bentley, Imperial College Cerebrovascular Research Unit (ICCRU), Imperial College Hospitals, Charing Cross Campus, Fulham Palace Rd, London W6 8RF, UK; p.bentley{at}imperial.ac.uk

Abstract

Background Cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoencephalopathy (CADASIL) is a hereditary cause of cerebral small-vessel disease associated with one of many recognised mutations of the NOTCH3 gene. Spinal cord involvement is not a recognised feature. The authors describe a unique CADASIL pedigree that manifested a stereotypical pattern of cord lesions, in association with a novel and atypical NOTCH3 mutation.

Methods Clinical, radiological, laboratory and genetic characterisation of three affected family members. The associated NOTCH3 mutation was further evaluated by site-directed mutagenesis, immunohistochemistry, CBF1-transcription reporter assay, and screened for in 100 unrelated pathologically confirmed multiple sclerosis (MS) patients.

Results Three members of a family presented with CADASIL caused by a novel NOTCH3 missense mutation, C212Y. Two daughters of the proband also manifested a distinctive pattern of cord lesions confined to the posterocentral zone, cerebral lesions showing both a demyelinating and a typical CADASIL topography, positive antinuclear antibodies and intrathecally derived oligoclonal bands. The mutation occurred in exon 4—that is, outside the Notch3 ligand-binding domain—yet unusually for this location impaired Notch function as assessed by Jagged1 signal transduction. The C212Y mutation did not occur in 100 separate MS cases.

Conclusions This is the first description of an inherited pattern of cord lesions in association with CADASIL. The fact that certain features of dysregulated immunity also occurred, in association with a novel and atypical loss-of-function NOTCH3 mutation, supports evidence for functional interactions of Notch3 with the immune system, in addition to its vascular support role.

  • CADASIL
  • NOTCH3
  • genetic
  • stroke
  • cord lesions
  • mutations
  • cerebrovascular disease
  • genetics
  • multiple sclerosis
  • neuroimmunology

https://doi.org/10.1136/jnnp.2010.223297

Statistics from Altmetric.com

    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.

    Introduction

    Cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoencephalopathy (CADASIL) is a well-recognised monogeneic disease associated with recurrent small-vessel ischaemic strokes, migraine and dementia.1 The underlying molecular defect of CADASIL resides in Notch3—a transmembrane protein expressed in vascular smooth muscle cells—resulting in impaired cerebrovascular autoregulation, hypoperfusion and ischaemia.2 3 More than 100 deleterious mutations of the NOTCH3 gene have been described, with variable consequences on cell function,2 4 and, in a minority of cases, accounting for variability in disease phenotype.5–8

    Since CADASIL often presents with a relapsing–remitting course, and widespread, discrete cerebral and brainstem white-matter lesions, diagnostic confusion can occur with multiple sclerosis (MS).9 10 However, the pathological and genetic bases of these two diseases are distinct, with one large, well-characterised cohort of MS patients failing to demonstrate any of the common NOTCH3 mutations.11 Moreover, detailed radiological and cerebrospinal fluid (CSF) characteristics of the two diseases are generally distinct, with, for example, cord lesions and CSF antibodies often being present in MS but not in CADASIL.12 13 We describe here a family with CADASIL that also displayed certain features of MS—including cord lesions and CSF oligoclonal bands—in association with a novel NOTCH3 mutation. As well as characterising the affected individuals' clinical, radiological and laboratory features, we assess the NOTCH3 mutation in terms of its effects on cell function and determine whether it is found in a separate group of pathologically confirmed MS.

    Materials and methods

    Subjects

    The female proband (A) in her fifties and her three daughters (B, C, D), aged between mid-thirty and late twenties, were evaluated. All of them consented for NOTCH3 mutation analysis and for their anonymous test results to be published. Both parents of the proband were deceased. All grandchildren were well, except for the son of D who at 3 years old displayed mildly impaired cognitive–behavioural, but not physical, milestones. Subjects A, C and D, but not B, tested positive for the same NOTCH3 mutation, and were further evaluated with laboratory and imaging tests. B had no relevant medical history.

    NOTCH3 genotyping

    Genomic DNA was extracted from a 10 ml blood sample from each individual. Exons 3 and 4 of the NOTCH3 gene were amplified by PCRs using Abgene Reddymix Taq (Abgene) in a volume of 25 μl. PCR primers and reaction conditions have been described previously.14 PCR products were purified from agarose gel using the GENECLEAN II Kit (Bio 101, Inc). Standard cycle sequencing reactions using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Warrington, UK) contained 3–10 ng of purified PCR product in 10 μl and were performed using the forward and reverse primers used for initial amplification. The sequencing reactions were then ethanol precipitated, dried and analysed on an ABI PRISM 3700 DNA Analyzer.

    Site-directed mutagenesis

    Construction of full-length human NOTCH3 complementary DNA (cDNA) was made from an imaging clone (#2090922) and DNA fragments were generated by reverse transcriptase-PCR.15 NOTCH3 cDNA was subcloned into expression vector pcDNA3.1. The following mutations were introduced by site-directed mutagenesis using the QuickChange Site-Directed Mutagenesis Kit (Stratagene): C212Y, C212S, C440G. All the cDNA sequences in expression vectors were confirmed by DNA sequencing. HEK293 cells, a human embryonic kidney cell line (ATCC, CRL-1573), were cultured in Dulbecco's modified Eagle's medium, supplemented with 10% fetal bovine serum, 2 mM l-glutamine and 100 U/ml streptomycin/penicillin. Cells were transiently transfected with 2 μg of endotoxin-free plasmid DNA using lipofectAMINE transfection reagents (Invitrogen).15

    Western blotting, immunocytochemistry and functional assay

    Western blotting was carried out with whole-cell lysates 48 h post-transfection using a rabbit polyclonal Notch3 antibody (M-134, Santa Cruz).15 For immunocytochemistry, HEK293 cells were seeded on cover slips in a 6-well plate. Twenty-four hours after transfection an antibody specific for the extracellular domain of Notch3 (R&D System) and an AlexaFlour488-conjugated secondary antibody (Molecular Probes) were used to visualise the cell-surface Notch3 proteins. Notch3 function was assessed by the ability of the Notch3 ligand Jagged1 to activate transcription via the CBF1/RBP-Jk promoter. A firefly luciferase reporter construct containing 4×CBF1 responsive elements (0.2 μg DNA/well) and renilla luciferase control (0.05 μg DNA/well) were co-transfected with wild-type Notch3 or mutant constructs C212Y, C212S, C440G (0.2 μg DNA/well) into HEK293 cells, in a 96-well plate. After 24 h, 2×104/well HEK293 cells that stably express the Notch3 ligand Jagged1, or the same number of control HEK293 cells, were overlaid on transfected cells and co-cultured for a further 24 h. Luciferase activity was measured with a luminometer using the Dual-Glo luciferase kit (Promega), and presented as relative luciferase units.15 Experiments were performed in triplicate, and run twice for confirmation.

    Screening of NOTCH3 mutation in MS brain bank

    One hundred subjects with MS confirmed at postmortem from the UK MS Tissue bank (ethical approval 08/MRE09/31) were typed for the NOTCH3 mutation described here, C212Y (methods as above). Within this cohort, nine subjects had a family history and their mean age of MS onset was 33±10 years (mean±SD, range 10–63 years) with a mean age of death of 62±13 years (range 34–68 years).

    Results

    Clinical

    A Caucasian care-assistant in her fifties (A; figure 1A) presented with a 2-year history of recurrent ischaemic strokes and transient ischaemic attacks, including dysphasia, anosmia and ageusia, progressive depression, mental slowing and impaired concentration and memory. Past medical history consisted of migraine without aura, treated dysthyroidism and borderline hypertension. She smoked 30 cigarettes per day from age 15, but did not abuse alcohol. Her mother had a history of lifelong migraine and died in her fifties of breast carcinoma. Neurological examination revealed abducens and oculomotor nerve palsies, dysarthria, spasticity and generalised hyperreflexia. Cardiovascular examination was unremarkable other than a blood pressure of 150/92 mm Hg.

    Figure 1
    Figure 1

    (A) Family lineage indicating members with Notch3 mutations and CADASIL (black circles). First and fourth generations were not tested being deceased and under 18 years old, respectively, although the subject indicated by hatched shading had migraine. (B) Genotype analysis of affected individuals showing heterozygous missense mutation C212Y. CADASIL, cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoencephalopathy.

    Her daughter C, in her early thirties, presented with a 10-year history of migraine, which on occasion was associated with transient (<1 h) expressive aphasia. She also had a 5-year history of a relapsing–remitting brainstem and cord syndrome that had been diagnosed clinically as MS. Clinical symptoms included vertigo, vomiting, diplopia, painful eye movements, paraesthesia and weakness of the legs, gait imbalance, Lhermitte's phenomenon and urinary retention. Relapses developed gradually over days or weeks. Three of these were treated with 5-day courses of intravenous methylprednisolone, which typically resulted in significant albeit incomplete improvement over the following few weeks. Neurological examination demonstrated complex ophthalmoparesis, skew deviation, spasticity, cerebellar ataxia, generalised hyperreflexia, absent abdominal reflexes and bilateral sensory disturbance in the legs. Another daughter D, in her late twenties, presented with migraine without aura. She was on the oral contraceptive pill. Her neurological examination was unremarkable, and this remained so for 2 years (ie, until present day). Both daughters had normal cardiovascular examinations including blood pressure, and neither smoked or abused alcohol, or recreational drugs.

    Imaging

    T2-weighted MRI brain scans in patients A, C and D showed appearances consistent with CADASIL: extensive, largely confluent, white-matter hyperintensities concentrated around the juxtacortical temporal poles and external capsules. In addition, C showed multiple ovoid lesions in the corona radiata, peritrigone, lateral and medial pons and lower medulla, some of which were enlarged and some of which were constricted over 5 years, suggestive of an inflammatory demyelinating disorder such as MS. (figure 2A). The MRI cord of daughters C and D showed multiple areas of high intensity confined to the posterocentral cord that tapered cranio-caudally on sagittal sections. Lesions did not enhance (figure 2B). The brain and cord MRI of patient B and cord MRI of patient A were normal.

    Figure 2
    Figure 2

    MRI brain (A) and cord (B) in two affected daughters. Both show cerebral features of CADASIL including temporal lobe hyperintensities seen here, but patient C also shows multiple pons, medulla and corona radiata lesions, some of which increased and some of which decreased in size and intensity over time, consistent with MS. Both daughters also showed stereotypical lesions confined to the posterocentral cord. CADASIL, cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoencephalopathy.

    Other investigations

    CSF examination of patient D revealed five white cells (60% mononuclear cells, 40% lymphocytes), protein 43 mg/dl, normal glucose and oligoclonal bands on protein electrophoresis unmatched on serum electrophoresis. Patient A showed the following mildly abnormal blood results: serum cholesterol 5.9 mmol/l, alanine transaminase 60 iu/l, alkaline phosphatase 132 iu/L, thyroid stimulating hormone 7.2 mu/l. Patient C showed positive antinuclear antibodies (1:80), and anti-dsDNA (Crithidia) level 187 units (range 0–30). Her HLA-typing was DR13,15, DRB 52,52, DQ6. However, the following tests were all normal or negative: glucose, full blood count, renal function, electrolytes, erythrocyte sedimentation rate (ESR), thrombophilia screen, serum electrophoresis and immunoglobulins, vitamin B12, folate, homocysteine, anticardiolipin and lupus anticoagulant, angiotensin converting enzyme (ACE), extractable nuclear antigen (ENA), anti-neutrophil cytoplasmic antibody (ANCA), serology for syphilis, HIV-1 and 2, very-long chain fatty acid levels and anti-aquaporin antibodies. Patient D tested normal for all of the above. Patients C and D also showed normal visual, brainstem, somatosensory-evoked cortical potentials, and electroretinograms.

    Genotyping and functional characterisation

    All three patients showed a previously undescribed heterozygous missense mutation C212Y in exon 4 of NOTCH3, which occurs within epidermal growth factor-like repeat (EGFR) 5 (figure 1B). This G-to-A mutation at nucleotide 713 (c713G→A) results in the substitution of tyrosine for cysteine at codon 212. We subsequently evaluated the functional properties of C212Y, as well as an alternative, previously described4 pathogenic mutation at the same codon, C212S, in porcine aortic smooth muscle and HEK293 cells. Both C212Y and C212S NOTCH3 mutants were expressed in HEK293 cells as indicated by western blotting (figure 3A), and appropriately targeted the cell surface as indicated by extracellularly confined labelled anti-Notch3 antibodies (figure 3B). However, the C212Y mutant failed to activate the Notch-specific downstream transcriptional factor CBF1/RBP-Jκ following exposure to the Notch3 ligand Jagged1, in contrast to the C212S mutation2 whose function was similar to that of wild-type NOTCH3. Another previously identified mutation, C440G, which is located within the putative ligand-binding domain (EGFR 10/11), impaired Jagged1-CBF1 signal transduction, validating the assay.

    Figure 3
    Figure 3

    Functional characterisation of C212Y Notch3 mutation. (A) Western blotting performed 48 h after transfection shows successful expression of Notch3 proteins using Notch3 antibody. Control cells were transfected with empty pcDNA 3.1 vectors. (B) Wild-type (WT) Notch3 and Notch3 mutants (C212Y and C212S) were transfected in HEK293 cells. Immunofluorescent staining performed 24 h post-transfection using an antibody against the extracellular domain of Notch3 protein (green; left column) shows cell surface localisation of both WT Notch3 and the mutants C212Y and C212S. Nuclei of the same cells were DAPI stained (red; right column). (C) Notch3 activity was determined by Jagged ligand-induced CBF1/RBP-Jκ activation using luciferase assay. Significant Jagged-induced increases in luciferase activity were found for WT Notch3 and C212S, but not for C212Y, C440R or empty pcDNA 3.1 vectors. Results are presented as mean relative luciferase unit±SE. *p<0.05, **p<0.01. DAPI, 4′,6-diamidino-2-phenylindole.

    In view of the presence of features of both MS and CADASIL in daughters C and D, and a novel NOTCH3 mutation, we tested for the same mutation, C212Y, in our independent sample of 100 pathologically confirmed cases of MS. None of these MS cases demonstrated this mutation.

    Discussion

    We report a unique pedigree in which CADASIL co-existed with cord lesions, associated with a novel and anomalous loss-of-function NOTCH3 mutation. All three affected family members displayed characteristic radiological appearances of CADASIL—namely, anterior temporal lobe hyperintensities on T2-MRI. However, two affected daughters demonstrated additional features suggestive of central inflammation: namely, gradual (rather than sudden) onset deteriorations and a steroid-responsive, relapsing–remitting course (patient C); multiple ovoid or flame-shaped, periventricular, pontine and medullary lesions that increased or decreased in size over time (patient C); cord lesions (patients C and D) and CSF oligoclonal bands unmatched in the serum (patient D). The clinical and genetic features are discussed.

    One of the most striking findings is that the two affected daughters each showed multiple, highly stereotyped lesions restricted to the posterocentral cord on T2-weighted MRI. A previous radiological survey of 25 randomly selected CADASIL patients found no abnormalities on conventional MRI sequences.12 No other reports have indicated cord disease in CADASIL, except for one case where a clinical syndrome suggestive of anterior spinal infarction was described without radiological confirmation.16 We suggest either of two possible explanations for the cord lesions observed here. First, in the context of a relapsing–remitting, steroid-responsive cord syndrome with a subset of brain lesions obeying an MS-type distribution, positive oligoclonal bands and a high antinuclear antibody titre, these cord lesions may also reflect an inflammatory demyelinating process. However, the shape and topography of the cord lesions, as well as the fact that they did not enhance, and the similarity in their appearance between family members make them atypical for MS, in which cord lesions are typically lateral and asymmetric. Second, the fact that the lesions taper longitudinally, as seen with spinal cord infarction,17–19 and occur in the posterocentral cord—where small-calibre perforating arteries of the cord are longest in length20—leads us to speculate that the lesions reflect small-vessel ischaemia (and so parallel the aetiology of cerebral lesions in CADASIL). Previously observed changes in the magnetisation transfer ratio of the cord in CADASIL have been interpreted as indicating microcirculatory dysfunction.12

    A further unusual finding for CADASIL is that CSF from one daughter revealed an unmatched pattern of oligoclonal bands, suggesting primary central nervous system inflammation. The CSF characteristics in CADASIL are largely normal, with only a mildly elevated protein level (ie, 45–75 mg/dl) found in up to one-third of patients, and only 1 out of 87 patients being found to have positive CSF oligoclonal bands.13 While two other CADASIL cases have been reported as showing CSF oligoclonal bands,13 21 it is unclear from any of these reports whether such bands were unmatched on serum electrophoresis—that is, suggesting intrathecal rather than systemic synthesis—and whether other atypical clinical or genetic features co-existed. If the cord lesions seen in our patients on MRI were due to ischaemia (discussed above), the presence of positive CSF oligoclonal bands leads us to speculate a possible small-vessel vasculitic aetiology.

    As well as describing the unique co-occurrence of CADASIL with a distinctive pattern of cord lesions, we describe here a novel and atypical NOTCH3 mutation. The vast majority of the 170 or so pathogenic NOTCH3 mutations currently described show very similar CADASIL phenotypes,22 with only a few mutations—mainly unusual deletion or frame-shift mutations—resulting in distinct clinical syndromes.5–7 Therefore, the reason why the missense mutation described here, C212Y, should result in an unusual clinical phenotype is not at first clear, especially as it represents a cysteine substitution within an extracellular EGFR domain—in keeping with the vast majority of described NOTCH3 mutations. However, we also found that C212Y resulted in impaired Notch3 nuclear signalling—which is unexpected, because more than 95% of NOTCH3 mutations retain cellular function and because, when they are loss-of-function mutations, they are limited almost exclusively to mutations in domain EGFR 10/11, that is within the putative ligand-binding domain.2 8 23 In support of this, a previously described, alternative pathogenic mutation at the same codon (C212S) as the mutation described here4 was shown by our assay to be associated with retained nuclear signalling. Interestingly, the only previously published2 example of a loss-of-function NOTCH3 mutation located outside the ligand-binding domain also involved a cysteine-to-tyrosine mutation (C543Y) as found here, suggesting that tyrosine may influence protein function through remote allosteric effects. However, whether any unusual clinical features were associated with this latter mutation was not commented upon.

    A critical question is whether the syndrome presented here arose solely from the single NOTCH3 mutation or from two different genetic (or genetic/environmental) events. The fact that the mother, who also possessed the NOTCH3 mutation C212Y, did not manifest cord lesions, in contrast to the daughters, favours the possibility that a further germ-line event (either genetic or epigenetic) may have supervened. Alternatively, this fact might suggest that the C212Y mutation is only partially penetrant with respect to cord lesions. In any case, in support of the claim that the C212Y mutation was responsible for the entire neurological syndrome, either by itself or in conjunction with a secondary factor, are the findings that: (1) the cord lesions are strikingly similar in topography between siblings who possessed the C212Y mutation, but critically are absent in the third sister with the wild-type NOTCH genotype; (2) the cord lesions are atypical for common acquired demyelinating disorders, for example, MS and (3) the underlying mutation is exceptional among NOTCH3 mutations in being a loss-of-function mutation located outside the putative ligand-binding domain.

    Recently, loss-of-function NOTCH3 mutations have been associated with a CADASIL phenotype different to that of the commoner gain-of-function variety, with the former group showing a significantly higher rate of asymptomatic cerebral white-matter lesions than the latter.8 The implication was that pathological processes other than granular osmiophilic material deposition may co-occur in CADASIL following loss-of-function NOTCH3 mutations. By extension to our report and noting that the loss-of-function mutation described here, in occurring distant to the putative ligand-binding domain, was even more anomalous than the subgroup of loss-of-function NOTCH3 mutations reported by Monet-Leprêtre et al8 lead us to speculate that central nervous inflammation may be part of the wider CADASIL phenotypic spectrum, in the presence of certain atypical NOTCH3 mutations. In support of this, CADASIL is associated with blood–brain barrier breakdown,13 raised CSF complement levels24 and vasculitis-like pathology,25 while impairment of Notch3 function dysregulates pathogenic T cells in an animal model of MS.26 Whether markers of central nervous inflammation are more likely in other CADASIL patients with loss-of-function NOTCH3 mutations would be an interesting future line of enquiry.

    Finally, given the possibility that the novel NOTCH3 mutation described here may have predisposed to an MS-like phenotype, we questioned whether this mutation might also be found in 100 cases of pathologically confirmed MS (including nine cases where there a family history of MS existed). This analysis failed to identify the C212Y mutation in any of these separate cases, which concurs with an earlier study from our group in which common NOTCH3 mutations or microsatellite linkers were also not found among a well-characterised MS cohort.11 However, since both of these cohorts are intentionally made up of ‘typical’ MS patients, it is conceivable that NOTCH3 mutations may be found among unselected MS patients, especially those showing atypical features.

    In conclusion, our report is the first clear description of a NOTCH3 mutation associated with spinal cord disease, in addition to cerebral ischaemia. As well as extending the spectrum of CADASIL phenotypic variants, the presence of demyelinating inflammatory features in our patients supports evidence for the Notch3 receptor in an immunological, as well as vascular support, role.

    References

    1. ↵
      1. Chabriat H,
      2. Joutel A,
      3. Dichgans M,
      4. et al
      . Cadasil. Lancet Neurol 2009;8:643–53.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Joutel A,
      2. Monet M,
      3. Domenga V,
      4. et al
      . Pathogenic mutations associated with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy differently affect Jagged1 binding and Notch3 activity via the RBP/JK signaling Pathway. Am J Hum Genet 2004;74:338–47.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Joutel A,
      2. Monet-Leprêtre M,
      3. Gosele C,
      4. et al
      . Cerebrovascular dysfunction and microcirculation rarefaction precede white matter lesions in a mouse genetic model of cerebral ischemic small vessel disease. J Clin Invest 2010;120:433–45.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Joutel A,
      2. Vahedi K,
      3. Corpechot C,
      4. et al
      . Strong clustering and stereotyped nature of Notch3 mutations in CADASIL patients. Lancet 1997;350:1511–15.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Joutel A,
      2. Chabriat H,
      3. Vahedi K,
      4. et al
      . Splice site mutation causing a seven amino acid Notch3 in-frame deletion in CADASIL. Neurology 2000;54:1874–5.
      OpenUrlFREE Full Text
    6. ↵
      1. Dotti MT,
      2. De Stefano N,
      3. Bianchi S,
      4. et al
      . A novel Notch3 frameshift deletion and mitochondrial abnormalities in a patient with CADASIL. Arch Neurol 2004;61:942–5.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Arboleda-Velasquez JF,
      2. Lopera F,
      3. Lopez E,
      4. et al
      . C455R notch3 mutation in a Colombian CADASIL kindred with early onset of stroke. Neurology 2002;59:277–9.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Monet-Leprêtre M,
      2. Bardot B,
      3. Lemaire B,
      4. et al
      . Distinct phenotypic and functional features of CADASIL mutations in the Notch3 ligand binding domain. Brain 2009;132(Pt 6):1601–12.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Desmond DW,
      2. Moroney JT,
      3. Lynch T,
      4. et al
      . The natural history of CADASIL: a pooled analysis of previously published cases. Stroke 1999;30:1230–3.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. O'Riordan S,
      2. Nor AM,
      3. Hutchinson M
      . CADASIL imitating multiple sclerosis: the importance of MRI markers. Mult Scler 2002;8:430–2.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Broadley SA,
      2. Sawcer SJ,
      3. Chataway SJ,
      4. et al
      . No association between multiple sclerosis and the Notch3 gene responsible for cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL). J Neurol Neurosurg Psychiatry 2001;71:97–9.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Rocca MA,
      2. Filippi M,
      3. Herzog J,
      4. et al
      . A magnetic resonance imaging study of the cervical cord of patients with CADASIL. Neurology 2001;56:1392–4.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Dichgans M,
      2. Wick M,
      3. Gasser T
      . Cerebrospinal fluid findings in CADASIL. Neurology 1999;53:233.
      OpenUrlFREE Full Text
    14. ↵
      1. Wang T,
      2. Sharma SD,
      3. Fox N,
      4. et al
      . Description of a simple test for CADASIL disease and determination of mutation frequencies in sporadic ischaemic stroke and dementia patients. J Neurol Neurosurg Psychiatry 2000;69:652–4.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Wang T,
      2. Holt CM,
      3. Xu C,
      4. et al
      . Notch3 activation modulates cell growth behaviour and cross-talk to Wnt/TCF signalling pathway. Cell Signal 2007;19:2458–67.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Hutchinson M,
      2. O'Riordan J,
      3. Javed M,
      4. et al
      . Familial hemiplegic migraine and autosomal dominant arteriopathy with leukoencephalopathy (CADASIL). Ann Neurol 1995;38:817–24.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Faig J,
      2. Busse O,
      3. Salbeck R
      . Vertebral body infarction as a confirmatory sign of spinal cord ischemic stroke: Report of three cases and review of the literature. Stroke 1998;29:239–43.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Bergqvist CAG,
      2. Goldberg HI,
      3. Thorarensen O,
      4. et al
      . Posterior cervical spinal cord infarction following vertebral artery dissection. Neurology 1997;48:1112–15.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Mascalchi M,
      2. Cosottini M,
      3. Ferrito G,
      4. et al
      . Posterior spinal artery infarct. Am J Neuroradiol 1998;19:361–3.
      OpenUrlAbstract
    20. ↵
      1. Mawad ME,
      2. Rivera V,
      3. Crawford S,
      4. et al
      . Spinal cord ischemia after resection of thoracoabdominal aortic aneurysms: MR findings in 24 patients. Am J Neurorad 1990;11:987–91.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Chabriat H,
      2. Vahedi K,
      3. Iba-Zizen MT,
      4. et al
      . Clinical spectrum of CADASIL: a study of 7 families. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Lancet 1995;346:934–9.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Singhal S,
      2. Bevan S,
      3. Barrick T,
      4. et al
      . The influence of genetic and cardiovascular risk factors on the CADASIL phenotype. Brain 2004;127:2031–8.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Peters N,
      2. Opherk C,
      3. Zacherle S,
      4. et al
      . CADASIL-associated Notch3 mutations have differential effects both on ligand binding and ligand-induced Notch3 receptor signaling through RBP-Jk. Exp Cell Res 2004;299:454–64.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Unlü M,
      2. de Lange RP,
      3. de Silva R,
      4. et al
      . Detection of complement factor B in the cerebrospinal fluid of patients with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy disease using two-dimensional gel electrophoresis and mass spectrometry. Neurosci Lett 2000;282:149–52.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Rafalowska J,
      2. Dziewulska D,
      3. Fidzianska A
      . CADASIL: what component of the vessel wall is really a target for Notch 3 gene mutations? Neurol Res 2004;26:558–62.
      OpenUrlCrossRefPubMed
    26. ↵
      1. Jurynczyk M,
      2. Jurewicz A,
      3. Raine CS,
      4. et al
      . Notch3 inhibition in myelin-reactive T cells down-regulates protein kinase C theta and attenuates experimental autoimmune encephalomyelitis. J Immunol 2008;180:2634–40.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Funding PS and PB hold Department of Health (UK) Senior Fellowships. All tissue samples were supplied by the UK Multiple Sclerosis Tissue Bank (http://www.ukmstissuebank.imperial.ac.uk), which is funded by the MS Society of Great Britain and Northern Ireland (registered charity 207495).

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.