Title: Homozygous <scp>NOTCH</scp> 3 null mutation and impaired <scp>NOTCH</scp> 3 signaling in recessive early‐onset arteriopathy and cavitating leukoencephalopathy
Abstract: Report13 April 2015Open Access Homozygous NOTCH3 null mutation and impaired NOTCH3 signaling in recessive early-onset arteriopathy and cavitating leukoencephalopathy Tommaso Pippucci Tommaso Pippucci U.O. Genetica Medica, Policlinico Sant'Orsola-Malpighi, Bologna, Italy Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Alessandra Maresca Alessandra Maresca IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Pamela Magini Pamela Magini Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Giovanna Cenacchi Giovanna Cenacchi Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Vincenzo Donadio Vincenzo Donadio IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Search for more papers by this author Flavia Palombo Flavia Palombo Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Valentina Papa Valentina Papa Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Alex Incensi Alex Incensi IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Search for more papers by this author Giuseppe Gasparre Giuseppe Gasparre Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Maria Lucia Valentino Maria Lucia Valentino IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Carmela Preziuso Carmela Preziuso Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy Search for more papers by this author Annalinda Pisano Annalinda Pisano Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy Search for more papers by this author Michele Ragno Michele Ragno Divisione di Neurologia, Ospedale Mazzoni, Azienda Sanitaria Unica Regionale, Ascoli Piceno, Italy Search for more papers by this author Rocco Liguori Rocco Liguori IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Carla Giordano Carla Giordano Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy Search for more papers by this author Caterina Tonon Caterina Tonon Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Unità Risonanza Magnetica Funzionale, Policlinico S.Orsola-Malpighi, Bologna, Italy Search for more papers by this author Raffaele Lodi Raffaele Lodi Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Unità Risonanza Magnetica Funzionale, Policlinico S.Orsola-Malpighi, Bologna, Italy Search for more papers by this author Antonia Parmeggiani Antonia Parmeggiani Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy U.O. Neuropsichiatria Infantile, Policlinico S.Orsola-Malpighi, Bologna, Italy Search for more papers by this author Valerio Carelli Corresponding Author Valerio Carelli IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Marco Seri Corresponding Author Marco Seri U.O. Genetica Medica, Policlinico Sant'Orsola-Malpighi, Bologna, Italy Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Tommaso Pippucci Tommaso Pippucci U.O. Genetica Medica, Policlinico Sant'Orsola-Malpighi, Bologna, Italy Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Alessandra Maresca Alessandra Maresca IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Pamela Magini Pamela Magini Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Giovanna Cenacchi Giovanna Cenacchi Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Vincenzo Donadio Vincenzo Donadio IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Search for more papers by this author Flavia Palombo Flavia Palombo Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Valentina Papa Valentina Papa Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Alex Incensi Alex Incensi IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Search for more papers by this author Giuseppe Gasparre Giuseppe Gasparre Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Maria Lucia Valentino Maria Lucia Valentino IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Carmela Preziuso Carmela Preziuso Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy Search for more papers by this author Annalinda Pisano Annalinda Pisano Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy Search for more papers by this author Michele Ragno Michele Ragno Divisione di Neurologia, Ospedale Mazzoni, Azienda Sanitaria Unica Regionale, Ascoli Piceno, Italy Search for more papers by this author Rocco Liguori Rocco Liguori IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Carla Giordano Carla Giordano Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy Search for more papers by this author Caterina Tonon Caterina Tonon Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Unità Risonanza Magnetica Funzionale, Policlinico S.Orsola-Malpighi, Bologna, Italy Search for more papers by this author Raffaele Lodi Raffaele Lodi Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Unità Risonanza Magnetica Funzionale, Policlinico S.Orsola-Malpighi, Bologna, Italy Search for more papers by this author Antonia Parmeggiani Antonia Parmeggiani Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy U.O. Neuropsichiatria Infantile, Policlinico S.Orsola-Malpighi, Bologna, Italy Search for more papers by this author Valerio Carelli Corresponding Author Valerio Carelli IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy Search for more papers by this author Marco Seri Corresponding Author Marco Seri U.O. Genetica Medica, Policlinico Sant'Orsola-Malpighi, Bologna, Italy Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy Search for more papers by this author Author Information Tommaso Pippucci1,2,‡, Alessandra Maresca3,4,‡, Pamela Magini2, Giovanna Cenacchi4, Vincenzo Donadio3, Flavia Palombo2, Valentina Papa4, Alex Incensi3, Giuseppe Gasparre2, Maria Lucia Valentino3,4, Carmela Preziuso5, Annalinda Pisano5, Michele Ragno6, Rocco Liguori3,4, Carla Giordano5, Caterina Tonon4,7, Raffaele Lodi4,7, Antonia Parmeggiani2,8, Valerio Carelli 3,4 and Marco Seri 1,2 1U.O. Genetica Medica, Policlinico Sant'Orsola-Malpighi, Bologna, Italy 2Dipartimento di Scienze Mediche Chirurgiche (DIMEC), University of Bologna, Bologna, Italy 3IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy 4Unita' di Neurologia, Dipartimento di Scienze Biomediche e Neuromotorie (DIBINEM), University of Bologna, Bologna, Italy 5Dipartimento di Scienze Radiologiche, Oncologiche ed Anatomopatologiche, Sapienza, University of Rome, Rome, Italy 6Divisione di Neurologia, Ospedale Mazzoni, Azienda Sanitaria Unica Regionale, Ascoli Piceno, Italy 7Unità Risonanza Magnetica Funzionale, Policlinico S.Orsola-Malpighi, Bologna, Italy 8U.O. Neuropsichiatria Infantile, Policlinico S.Orsola-Malpighi, Bologna, Italy ‡These authors contributed equally to this work *Corresponding author. Tel: +39 051 4966747; Fax: +39 051 2092751; E-mail: [email protected] *Corresponding author. Tel: +39 051 2088421/051 6363694; Fax: +39 051 2088416; E-mail: [email protected] EMBO Mol Med (2015)7:848-858https://doi.org/10.15252/emmm.201404399 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Abstract Notch signaling is essential for vascular physiology. Neomorphic heterozygous mutations in NOTCH3, one of the four human NOTCH receptors, cause cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL). Hypomorphic heterozygous alleles have been occasionally described in association with a spectrum of cerebrovascular phenotypes overlapping CADASIL, but their pathogenic potential is unclear. We describe a patient with childhood-onset arteriopathy, cavitating leukoencephalopathy with cerebral white matter abnormalities presented as diffuse cavitations, multiple lacunar infarctions and disseminated microbleeds. We identified a novel homozygous c.C2898A (p.C966*) null mutation in NOTCH3 abolishing NOTCH3 expression and causing NOTCH3 signaling impairment. NOTCH3 targets acting in the regulation of arterial tone (KCNA5) or expressed in the vasculature (CDH6) were downregulated. Patient's vessels were characterized by smooth muscle degeneration as in CADASIL, but without deposition of granular osmiophilic material (GOM), the CADASIL hallmark. The heterozygous parents displayed similar but less dramatic trends in decrease in the expression of NOTCH3 and its targets, as well as in vessel degeneration. This study suggests a functional link between NOTCH3 deficiency and pathogenesis of vascular leukoencephalopathies. Synopsis A recessive homozygous protein-truncating NOTCH3 mutation is found in a patient with devastating childhood-onset, vascular leukoencephalopathy, suggesting that this could be the underlying mechanism in other patients with similar severe pathology. For the first time, a recessive homozygous protein-truncating mutation in NOTCH3 has been found in a patient with a devastating childhood-onset, vascular leukoencephalopathy. This mutation is shown to abolish NOTCH3 expression and signaling, with downregulation of NOTCH3 target genes. In the patient, vessels were characterized by SMC degeneration, similar to that observed in CADASIL. However, GOM deposits, the CADASIL hallmark, were absent, mirroring the Notch3−/− mouse model. Both the heterozygous consanguineous parents displayed similar but less dramatic trends in NOTCH3 signaling impairment and vascular damage. Introduction The Notch signaling pathway is an ancient inter-cellular signaling mechanism playing central roles in vascular physiology (Gridley, 2007). Notch3, one of the four mammalian Notch family receptors, is a heterodimeric, single-pass transmembrane protein functioning as transcriptional activator. It is composed of a 34 epidermal growth factor-like repeats (EGFRs) extracellular domain (Notch3ECD) non-covalently attached to the transmembrane/intracellular domain (Notch3TM/IC) (Kopan & Ilagan, 2009). Notch3 is predominantly expressed in the smooth muscle cells (SMCs) surrounding small arteries and in pericytes around capillaries (Joutel et al, 2010b; Lewandowska et al, 2010). Notch3 knockout mice (Notch3−/−) show marked alteration of arterial SMCs, pointing to a critical role of Notch3 in the maturation and maintenance of arteries (Joutel, 2010a). Heterozygous NOTCH3 mutations underlie cerebral arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL, MIM 125310), a disorder of the small arterial vessels of the brain that represents the most common heritable cause of stroke and progressive ischemic dementia in the adults. CADASIL is inherited dominantly, with > 500 families reported worldwide and de novo events observed sporadically (Coto et al, 2006; Chabriat et al, 2009). Virtually all mutations (> 95%) are highly stereotyped missense mutations that abolish an existent cysteine residue of the 34 EGFRs of NOTCH3ECD (mainly EGFRs 2–5) or insert a new one, with the final effect of introducing an odd number of cysteines (Chabriat et al, 2009). In CADASIL, aberrant accumulation of mutant NOTCH3ECD is observed and specific granular osmiophilic material (GOM) deposits appear around the degenerated vascular SMCs (Chabriat et al, 2009), but it remains debated whether NOTCH3ECD is or is not a principal GOM component (Joutel, 2010a). CADASIL-associated mutations confer NOTCH3ECD propensity to self-aggregate, sequestering wild-type NOTCH3 and other extracellular molecules (Duering et al, 2011). Anomalous accumulation of such aggregates within vessel walls, possibly leading to GOM deposition, is considered a likely pathogenic mechanism. Nonetheless, CADASIL mutations have been shown to reflect hypomorphic receptor activity in mouse models that remarkably parallel the human condition (Arboleda-Velasquez et al, 2011). Thus, a chronic reduction of Notch3 signaling may plausibly lead to vascular SMC degeneration and ultimately to ischemic disease (Arboleda-Velasquez et al, 2011). Few hypomorphic NOTCH3 mutations (two distinct small out-of-frame deletions and a nonsynonymous nonsense substitution) have been observed in three different patients having a clinical and/or familial history compatible with CADASIL or CADASIL-like conditions (Dotti et al, 2004; Weiming et al, 2013; Erro et al, 2014). Interestingly, the nonsense mutation, a p.R103X substitution, has been described in an independent family in association with a phenotype of ischemic strokes but with incomplete penetrance (Rutten et al, 2013). Taken as a whole, these findings support the hypothesis that heterozygous hypomorphic NOTCH3 alleles may predispose to a spectrum of cerebrovascular phenotypes overlapping CADASIL. These alleles act with highly variable penetrance, in agreement with the observation that hypomorphic alleles have been reported occasionally also in normal subjects (Rutten et al, 2013). To date, null homozygous NOTCH3 alleles have never been reported in humans. Here we describe a patient, previously diagnosed as having Sneddon syndrome (Parmeggiani et al, 2000), displaying arteriopathy and cavitating, early-onset leukoencephalopathy. In this patient we identified a homozygous NOTCH3 nonsense mutation, which abolishes NOTCH3 expression and causes deregulation of NOTCH3 downstream target genes. Results Brain MRI The last MRI scan in the proband, performed at 23 years of age, showed an enlargement of the lateral ventricles (left > right), thinning of the corpus callosum, atrophy of the basal ganglia, reduced volume of brainstem and cerebellum, and diffuse cerebral white matter hyperintensity on T2-weighted images, with relative U fibers sparing (Fig 1A, B and E). The hyperintense cerebral white matter showed severe, diffuse cavitations in association with chronic multiple lacunar infarctions in the basal ganglia, thalamus, pons and bulb (Fig 1B and E) and one acute ischemic lesion in the pons (1D). Brain 3D TOF (time of flight) (Fig 1C) showed two small saccular aneurisms in the right M1 (ø 4 mm) and left M2 (ø 2.5 mm) segments of middle cerebral arteries. Disseminated microbleeds were present in both infra- and supra-tentorial structures (Fig 1F) on susceptibility-weighted imaging (SWI). Figure 1. Brain MRI study of the proband and parents A–F. Brain MRI of the proband at 23 years. (A) Sagittal fast spin echo (FSE) T1-weighted image shows thinning of the corpus callosum, dilation of the IV ventricle and of the cisterna magna and reduced volume of vermis and brainstem; two lacunar lesions are evident in the dorsal pons. (B) T2-weighted axial fluid-attenuated inversion recovery (FLAIR) images show hyperintense periventricular white matter with several lacunar lesions also in the basal ganglia and thalami. (C) MR angiography 3D TOF (time of flight) reconstruction shows two saccular aneurisms in the right M1 (ø 4 mm) and left M2 (ø 2.5 mm) segments of middle cerebral arteries (arrows). (D) A recent ischemic hyperintense lesion is detected on diffusion tensor imaging (DTI) in the right side of the dorsal pons. Severe cavitations and lateral ventricles dilation (left > right) on FLAIR images (B, E) and diffuse microbleeds as small hypointense foci on SWI are visible (F) in the same slices of (E). R = right, L = left. G, H. Brain MRI scans of the asymptomatic parents of the proband show multiple focal hyperintensities on T2-weighted images in the periventricular and subcortical cerebral white matter, expression of gliosis secondary to chronic small vessel ischemic changes, more evident in the father (G), 56 years, than in the mother (H), 54 years. Download figure Download PowerPoint Brain MRI and MR angiography showed no significant changes in the asymptomatic parents, respectively, at 54 and 56 years, except for multiple focal hyperintensities on T2-weighted fluid-attenuated inversion recovery (FLAIR) images in the periventricular and subcortical cerebral white matter expression of gliosis secondary to chronic small vessel ischemic changes, more evident in the father (Fig 1G) than in the mother (Fig 1H). Genetic study Based on the assumption that the causative mutation was inherited in the homozygous state, whole exome sequencing (WES) detected 23 rare (minor allele frequency < 1%) homozygous variants. Only 8 of these were within large homozygous genomic regions (> 5 Mb), which are known to have higher probability to harbor the pathogenic mutation (McQuillan et al, 2008). Only three were predicted as pathogenic by at least 2 out of the 4 in silico pathogenicity predictors used and 1 was a truncating mutation. Of these four variants, 2 were discarded since within genes already reported to be responsible for phenotypically divergent recessive diseases: KANK2, implicated in palmoplantar keratoderma and woolly hair (MIM 616099) and CHRNG, implicated in multiple pterygium syndrome (MIM 253290). This filtering procedure (Supplementary Fig S1A) left 2 final candidate variants: a nonsense p.C966* variant (NM_000435:c.C2898A) in NOTCH3 and a missense p.R65H variant (NM_001040664) in PPAN/PPAN-P2RY11 (Supplementary Fig S1B). Between these 2 final candidates, NOTCH3 mutation emerged as the most likely explanation for the disease pathogenesis, as supported by mutation type (nonsense versus missense), alternate allele frequency (novel versus 0.002% in the EXAC database, http://exac.broadinstitute.org/), deeper intolerance to genic variation (5.0 versus 10.8/17.1 RVIS percentile) and consistency of the pathology observed in the proband with protein function, tissue pattern expression and existent association with the disorder (Supplementary Fig S1B). In the proband, NOTCH3 lay in one of the long autozygous regions, a 7.9-Mb-long region on chromosome 19 (Supplementary Fig S1C, left panel). NOTCH3 c.C2898A was confirmed in the patient in the homozygous state and detected in the consanguineous patient's parents in the heterozygous state (Supplementary Fig S1C, right panel). In addition to public databases, the mutation was absent from > 200 in-house control exomes and in 500 regional control chromosomes analyzed by direct sequencing. By quantitative mRNA analysis in skeletal muscle biopsies (Fig 2A), we observed dramatic reduction of NOTCH3 expression in the proband and demonstrated a more moderate reduction of NOTCH3 expression in his father, whereas there was no relevant deregulation of NOTCH3 in his mother. In 3 CADASIL patients, reduction of NOTCH3 expression was comparable to that observed in the father. In the proband, direct sequencing of the NOTCH3 cDNA obtained by retrotranscription of the residual mRNA detected only the 2898A allele (mutant), while in the heterozygous parents, the C2898 allele (wild-type) appeared to be predominant (Fig 2B). cDNA of the two CADASIL patients carrying the c.C3016T (p.R1006C) mutation revealed balanced composition of C3016 and 3016T alleles. These findings suggest that the c.C2898A protein-truncating substitution induces the decay of the mutant mRNA molecule, while classical CADASIL-causing c.C3016T change does not. Figure 2. NOTCH3 and NOTCH3 targets expression profile Gene expression of NOTCH3, HES1, HEY1, HEYL and the 17 recently identified targets were established by a real-time PCR panel assay in skeletal muscle of controls (n = 6), proband (II:1), parents (mother I:1, father I:2) and CADASIL patients (n = 3). Graph shows gene expression fold changes relative to controls and normalized on GAPDH (reference gene), expressed as mean of two experiments ± SEM. Among the 17 novel assayed target genes, RCAN2, ANGPT4, HP and SORBS were not detected in any sample and were therefore not reported. Green and red lines indicate 2.0 and 0.5 fold changes, respectively. Statistical significances and P-values are reported in Supplementary Table S1. Direct sequencing of NOTCH3 cDNA from skeletal muscle of controls, proband, parents and two CADASIL patients. In the proband, residual mutant cDNA is amplified and the sequence shows only the mutant allele (A allele, arrow). Predominance of the wild-type allele (C allele, arrow) in the parents documents mRNA-mediated decay of the mutant allele (A allele), in contrast to what was observed for a canonical CADASIL mutation where there is balanced composition of mutant and wild-type alleles (arrows). Download figure Download PowerPoint We examined expression levels of canonical NOTCH3 target genes (HES1, HEY1, HEYL) and of 17 potential target genes (PERP, PLN, TBX2, SUSD5, TIMP4, PTP4A3, GRIP2, KCNA5, NRIP2, S1PR3, PGAM2, CDH6, XIRP1, RCAN2, ANGPT4, HP and SORBS2) homologous to murine genes found to be robustly downregulated in caudal distal arteries of Notch3−/− mice (Fouillade et al, 2013). Among 17 potential NOTCH3 downstream target genes, 4 (RCAN2, ANGPT4, HP and SORBS2) were not detectable in none of the samples (Supplementary Table S1). Genes of the HES and HES-related families were not downregulated, consistent with what was reported in Notch3−/− mice (Fouillade et al, 2013). Global alteration in the expression profiles of the 13 detectable potential target genes was consistent with the levels of reduced expression of NOTCH3 itself. The proband displayed global deregulation of target genes, with four downregulated genes (TBX2, KCNA5, NRIP2, CDH6) and one upregulated gene (TIMP4) (Supplementary Table S1; Fig 2A). Gene expression was almost unaltered in the mother (Supplementary Table S1; Fig 2A). Altered expression was observed in the father and in CADASIL patients, where three genes (GRIP2, NRIP2, XIRP1) and four genes (PERP, TBX2, S1PR1, XIRP1), respectively, resulted significantly downregulated (Supplementary Table S1; Fig 2A). Magnitude in fold changes of deregulated genes was generally greater in the proband than in his father and in CADASIL patients. Muscle histology Standard staining of proband's muscle biopsy showed mild variation of the fiber size. Pathological changes were evident in small vessels and capillaries, which presented a generalized thickening of the walls (Supplementary Fig S2A and C). Muscle biopsies of the proband's parents showed similar changes (Supplementary Fig S2E and G). In the mother, inflammatory infiltration around a blood vessel was also evident (Supplementary Fig S2H). More details about muscle histology are included in the Supplementary Information (Supplementary Fig S3 and Supplementary Methods and Results). Characterization of skin and skeletal muscle vessels Analysis of vessel wall structure was performed by immunofluorescence, immunohistochemistry and transmission electron microscopy (TEM) both on skin and on skeletal muscle. By immunofluorescence, increased deposition and altered distribution of collagen IV, with a clear derangement of collagen fibers, were prominent in the proband, in both tissues examined (Fig 3A and B). These features were paralleled by attenuation and disorganization of SMCs of the tunica media, as evaluated by immunohistochemistry on skeletal muscle (Fig 3C). Similar changes were observed in a CADASIL patient (Fig 3A–C) that showed a near complete loss of SMCs of the tunica media of skeletal muscle vessels, and in the parents (Supplementary Fig S4A–C) where the alterations were less pronounced. Changes in vessels' structure were confirmed by transmission electron microscopy (TEM) analysis of skin biopsy, both in the proband (Fig 3D) and, to a lesser extent, in his parents (Supplementary Fig S4D). Proband's skin vessels were characterized by multilayering and shedding of the basal membrane from plasmalemma into the stroma: Parallel rows of banded collagen fibrils were oriented perpendicular to and intimately associated with the plasma membrane. In the interstitial stroma, collagen fibrils appeared quantitatively more represented, while elastic fibers were rarefied. Most importantly, no deposits of granular osmiophilic material (GOM), a hallmark of CADASIL vascular injury, were ever observed, in contrast to CADASIL patients (Fig 3E). Figure 3. Morphologic analysis of vessels in skin and skeletal muscle of the proband as compared to control and CADASIL patient A–C. Histopathological examination of skin (A) and skeletal muscle biopsies (B, C) of control (left), proband (middle) and CADASIL patient (right). (A) Collagen IV staining (in green) in skin vessels: Collagen wall appears compact in the control, while it is disorganized in the skin vessels of the proband and of the CADASIL patient. Derangement of the collagen wall in single collagen fibers is more evident in the proband than in the CADASIL patient, where collagen wall is quite compact. Vessel's endothelium is delineated by ULEX staining (in red). Collagen IV staining in skeletal muscle vessels (B) recapitulates the skin picture. Smooth muscle actin immunostaining of skeletal muscle biopsies (C) shows attenuated SMCs in the tunica media of both the proband and the CADASIL patient, with foci characterized by a complete SMCs loss (particularly in the CADASIL patient) and thinning of the vessel wall (arrows). D, E. Ultrastructural analysis of skin biopsies. (D) Skin biopsy of the proband: (left panel) At low magnification, the tunica media of a vessel shows irregular SMCs surrounded by a markedly thickened basal membrane with a multilayered aspect (arrows); SMCs are identified by the presence of "focal adhesions" (square boxes); (middle panel) collagen fibrils appear organized in a parallel pathway along SMC (arrow), note the "focal adhesions" (square box); (right panel) collagen fibrils ("CF") are more represented in dense bundles than elastic ones ("E"). No GOM is detected, in contrast to CADASIL patient (E), showing deposits of granular osmiophilic material (arrows) located between SMC plasmalemma and basal lamina. F. KCNA5 immunostaining of skeletal muscle biopsies of control (left), proband (middle) and CADASIL patient (right). A global decrease in reactivity is evident in the vessels of both the proband and the CADASIL patient. Note multiple areas in which KCNA5 is barely detectable (arrows). Data information: Scale bars: (A) 50 μm; (B) 30 μm; (C, F) 10 μm; (D, left panel) 7 μm; (D, middle and right panel) 1 μm. Download figure Download PowerPoint Finally, on the basis of the results of gene expression analysis (Fig 2A), we examined the expression of KCNA5 protein in skeletal muscle biopsy of the proband, his parents and a CADASIL patient by immunohistochemistry with specific antibodies. Consistent with KCNA5 mRNA levels, reduced immunoreactivity was observed in the proband, his father and the CADASIL patient, while the proband's mother was similar to the control (Fig 3F and Supplementary Fig S4E). Discussion In this study, we identified a null homozygous NOTCH3 mutation