There is no treatment for CADASIL, the heritable small vessel disease that causes migraines, stroke, and even dementia. Its full name is a mouthful: cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Taking a step toward therapy, a paper in the July 11 Journal of Experimental Medicine reports that propping up Notch 3 signaling in mice protects cells that ensheathe retinal blood vessels. The authors, led by Joseph Arboleda-Velasquez and Patricia D’Amore at the Schepens Eye Research Institute of the Massachusetts Eye and Ear Infirmary, Boston, also claim that an antibody that activates Notch 3 receptors rescues at least some retinal blood vessels affected by a CADASIL-causing Notch mutation.

“The results are promising. They have certainly moved the field forward,” said Raj Kalaria at Newcastle University in the United Kingdom.

Damage to small blood vessels in the brain can impair cognition and may increase risk for dementia (see Feb 2012 newsApr 2017 news). Aging and cardiovascular disease can damage small vessels, but in rare cases, mutations are to blame, such as those causing CADASIL. “CADASIL is quite rare, but it’s useful for looking at general principles of this kind of disease,” noted Roy Weller, University of Southampton, U.K.

CADASIL mutations occur in extracellular domains of the Notch 3 receptor (Chabriat et al., 2009). Expressed predominantly in smooth muscle cells that form the outer layers of blood vessels, Notch 3 helps these cells form and maintain their identity, and protects them from apoptosis (Domenga et al., 2004; Baeten and Lilly, 2015). CADASIL mutations interfere with Notch 3 activation and cause aggregates of the protein’s extracellular domain to form in blood vessel walls. But researchers still debate whether the mutations cause a loss of normal function or a gain of toxic function, or both.

To get a handle on this, co-first authors Arturo Israel Machuca-Parra and Alexander Bigger-Allen asked how smooth muscle cells would fare without Notch 3 or if they expressed a mutant form of the receptor. The authors created three strains of mice: a Notch 3 knockout (N3KO); the same knockout conditionally expressing human wild-type Notch 3; or the knockout expressing human Notch 3 with the C455R CADASIL mutation (Arboleda-Velasquez et al., 2011; Arboleda-Velasquez et al., 2008; Mitchell et al., 2001). Both transgenes were under the control of a smooth muscle cell promoter.

Machuca-Parra and colleagues examined the retina because this extension of the central nervous system is vulnerable to CADASIL, is surrounded by a boundary similar to the blood-brain barrier, and has a stereotypical pattern of branching blood vessels for easy monitoring. The authors tested the integrity of the mural cells that coat retinal blood vessels, using an antibody to the mural cell protein α-smooth muscle actin (SMA). In wild-type mice and in N3KOs expressing human wild-type Notch 3, SMA outlined the arteries and arterioles of the retinas. In contrast, there was little SMA in retinal vessels of the N3KOs, or the knockouts that expressed the C455R transgene (see image below). That the C455R mutant and the full knockout had similar pathology suggested to the authors that C455R is a loss-of-function mutation.

Skimpy Attire: Mice expressing the C445R Notch 3 mutant in smooth muscle cells (bottom) have less α-smooth muscle actin in retinal blood vessels (right) and the whole retina (red in left) than do control animals (top). [Image courtesy of Alexander Bigger-Allen.]

Can the C445R mutant muster any Notch 3 activity? To test this, Machuca-Parra and Bigger-Allen used A13, an antibody that binds the Notch 3 extracellular domain and acts as an agonist (Li et al., 2008). The authors suspected A13 would activate C445R mutants also, because the antibody binds at a site that is distant from CADASIL mutations. Indeed, A13 activated a Notch 3 signaling reporter in cells expressing either wild-type or mutant Notch 3. This key experiment suggested it might be possible to buoy Notch 3 signaling in C445R CADASIL. 

The authors next injected A13 once a week for five weeks into the peritoneal cavities of week-old C455R or N3KO mice. As expected, the antibody had no effect on retinal arterioles in the N3KO mice, but doubled the SMA coverage of C445R Notch 3 arterioles (see image below). A13-treated C445R transgenics had more Notch 3 intracellular domain fragments in the brain, as well as higher plasma levels of the Notch 3 extracellular domain, indicating elevated processing of Notch by γ-secretase and ADAM proteases. Levels of endostatin, a fragment of collagen 18α1, a potential CADASIL biomarker, crept up as well (Primo et al., 2016). Arboleda-Velasquez thought this boded well for the clinic. “A big barrier in the field has been the lack of biomarkers,” he said.

Antibody to the Rescue: A13 acts as a Notch 3 agonist (right two columns). It restores smooth muscle actin (white) in retinal arterioles (green) expressing the C455R CADASIL mutation. Left two columns show control antibody. [Image courtesy of Alexander Bigger-Allen.]

“I am encouraged by this study. It’s an approach that would be useful to explore further,” said Kalaria. But he was disappointed that it was limited to young mice, when CADASIL’s hallmark symptoms usually appear in middle age. “I really would have liked to see what happens to older animals. Whether the antibody will work in that case is a big question,” he said. Arboleda-Velasquez agreed this is important to resolve.

Because a smooth muscle cell promoter drives Notch 3 expression in the test animals, the authors don’t know how A13 might affect other cells that may normally make Notch 3. They will test this with a new mouse line that expresses a humanized Notch 3 gene under the control of the mouse’s endogenous promoter.

Axel Montagne, Zhen Zhao, and Berislav Zlokovic at the University of Southern California in Los Angeles suggested the analysis should be extended beyond the retina to white-matter regions of the brain, such as the corpus callosum and internal capsule, which are particularly damaged in CADASIL. They also suggested monitoring mural cells that lack SMA, since at least in some forms of CADASIL, SMA-negative mural cells may be particularly damaged (Ghosh et al., 2015).

Arboleda-Velasquez wants to test a Notch 3 antibody in clinical trials eventually. He thinks the approach should help the 5-10 percent of CADASIL patients whose mutations map to the Notch 3 ligand-binding domain. He also suggested that people with Notch 3 variations that interfere with Notch 3 signaling, but are not classified as CADASIL mutations, may benefit as well. “The therapy is absolutely relevant to both groups of patients,” he said. Even CADASIL patients with mutations in regions outside the Notch 3 ligand-binding domain might be helped by an antibody like A13, he added.

The new findings may have implications beyond CADASIL. “Anything that could increase mural cell survival or maintain their function is likely to be important for vessel diseases,” said Weller. He claimed studies by Roxana Aldea in Roxana Carare's lab, also at the University of Southampton, suggest the importance of smooth muscle cell contraction for clearance of fluid and soluble metabolites from the brain. A breakdown in such clearance might cause buildup of toxic proteins, including Aβ (Aug 2012 news).—Marina Chicurel 


  1. Machuca-Parra and colleagues present an interesting study showing that Notch3 signaling is linked to vascular smooth muscle cells (VSMCs) coverage in retinal vessels and demonstrate that restoring Notch3 signaling via genetic rescue and using a Notch3 agonist antibody (A13) prevents small vessel disease (SVD) phenotype (e.g., VSMC loss) in both mouse models of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) and in Notch3 knockout (KO) mice.

    Pathology of brain vascular mural cells, including both VSMCs and pericytes, is central to both CADASIL patients and transgenic CADASIL models. The authors examined the α-smooth muscle actin (SMA)-positive VSMCs coverage in the retina and found reduced VSMCs coverage in Notch3 KO mice that can be rescued genetically by re-expressing wild-type Notch3. In contrast, re-expression of CADASIL C455R mutant in Notch3 KO mice neither rescued VSMCs nor contributed to VSMCs loss. These observations raise a couple of questions. What is the impact of Notch3 on the pericyte cell population covering brain capillaries? Are the effects restricted to the SMA-positive VSMCs pool? A previous report found no changes associated with VSMCs, but reported a significant decrease in pericytes attributed to Notch aggregation in a mouse CADASIL model carrying the R169C Notch3 mutation (Ghosh et al., 2015). Although it is known that Notch3 may regulate the expression of standard pericyte markers (i.e., PDGFRβ, NG2, and Desmin), investigation of the pericyte population lacking SMA in the brain by using a conditional Notch3 KO mouse model be the next step toward new insights into the role of mural cells and toward understanding the significance of Notch3 signaling in CADASIL.

    Furthermore, the authors provided evidence that the A13 antibody is efficient in HEK cells in vitro and can cross the blood-retinal barrier to reach the perivascular space. However, the only in vivo evidence demonstrating that mural cells can be activated by A13 administration is the presence of active cleaved Notch3 in brain vessels, which suggests no antibody uptake by mural cells but a direct binding. Further analyses have to be completed in regard to the distribution of Notch3 in vessels to understand how restoration of Notch3 signaling contributes to VSMC versus pericyte function in CADASIL mice.

    Finally, although the authors studied the retina for obvious reasons, further studies need to be performed on the brain parenchyma, especially in white matter (WM) regions such as the corpus callosum and the internal capsule because of the association between CADASIL and WM disease. Is it possible to restore Notch3 signaling and prevent mural cell loss in those specific WM structures using A13 antibody? And importantly, is the WM integrity preserved in A13-treated as compared to control IgG-treated CADASIL mice?


    . Pericytes are involved in the pathogenesis of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Ann Neurol. 2015 Dec;78(6):887-900. Epub 2015 Oct 7 PubMed.

  2. This well-designed study takes two big steps forward for the CADASIL field by (1) showing a therapeutic effect of a Notch 3-activating antibody in a mouse model, and (2) providing evidence for the potential utility of the Notch 3 extracellular domain as a translatable disease biomarker.

    I think the CADASIL field has argued long enough about whether the disease results from gain versus loss of Notch 3 function. When arguments like this go on for decades, the answer is usually “both.” It’s nice to see this paper move past that tangle and begin translating known biology to therapy.

    The potential clinical significance of these findings is even greater given recent evidence that Notch 3 mutations may be a major contributor to cerebral small vessel disease in the general population (see, for example, Rutten et al., 2016). In addition, given that small vessel disease often coexists with and likely exacerbates Alzheimer’s pathology (Kisler et al., 2017) and vice versa, it would be of interest to look for beneficial effects of the Notch 3-activating antibody in Alzheimer’s models. 


    . Archetypal NOTCH3 mutations frequent in public exome: implications for CADASIL. Ann Clin Transl Neurol. 2016 Nov;3(11):844-853. Epub 2016 Sep 28 PubMed.

    . Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer disease. Nat Rev Neurosci. 2017 Jul;18(7):419-434. Epub 2017 May 18 PubMed.

  3. I appreciate the comment by Drs. Zlokovic, Zhao, and Montagne on our work addressing the role of Notch3 signaling in mural cell loss.

    They raised important questions:

    • What is the impact of Notch3 on the pericyte cell population covering brain capillaries? Are the effects restricted to the SMA-positive VSMCs pool?

    I agree this is a relevant point that was not addressed in the current publication, and concur that the issue will be properly resolved using a conditional Notch3 KO mouse model allowing for abrogation of Notch3 signaling in specific mural cell subpopulations. We have recently generated such a model and have begun preliminary analyses.

    • Is it possible to restore Notch3 signaling and prevent mural cell loss in those specific WM structures using A13 antibody? And importantly, is WM integrity preserved in A13-treated as compared to control IgG-treated CADASIL mice?

    Unambiguous analyses of the brain vasculatures will require characterization of complete vasculatures similar to what we have accomplished in the retina. This may be done by implementing tissue-clearing methodologies such as CLARITY. 


    . CLARITY for mapping the nervous system. Nat Methods. 2013 Jun;10(6):508-13. PubMed.

  4. This study is an important advance for CADASIL. CADASIL patients have no available therapies. They desperately need increased awareness of this terrible disease and an interest from researchers in developing disease-modifying drugs.

    Dr. Arboleda-Velazquez's works suggests that antibodies that activate the genetic target of CADASIL, Notch3, could play an important therapeutic benefit in CADASIL models. Hopefully these studies continue to advance and translate into benefit for CADASIL patients.

  5. This is a very exciting study. These results add an interesting perspective on the still-controversial question of the mutational spectrum of CADASIL (see for examples Rutten et al., 2013, and Moccia et al., 2015). Perhaps study of additional mutations, using the same experimental design, will help further address this question. This would also help delineate the therapeutic potential of the A13 antibody for treatment and prevention of CADASIL. 

    Well done!


    . Hypomorphic NOTCH3 alleles do not cause CADASIL in humans. Hum Mutat. 2013 Nov;34(11):1486-9. Epub 2013 Oct 7 PubMed.

    . Hypomorphic NOTCH3 mutation in an Italian family with CADASIL features. Neurobiol Aging. 2015 Jan;36(1):547.e5-11. Epub 2014 Aug 27 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Silent Vascular Disease May Hasten Dementia Progression
  2. Vascular Disease in 50s Begets Brain Amyloid in 70s
  3. Brain Drain—“Glymphatic” Pathway Clears Aβ, Requires Water Channel

Paper Citations

  1. . Cadasil. Lancet Neurol. 2009 Jul;8(7):643-53. PubMed.
  2. . Notch3 is required for arterial identity and maturation of vascular smooth muscle cells. Genes Dev. 2004 Nov 15;18(22):2730-5. PubMed.
  3. . Differential Regulation of NOTCH2 and NOTCH3 Contribute to Their Unique Functions in Vascular Smooth Muscle Cells. J Biol Chem. 2015 Jun 26;290(26):16226-37. Epub 2015 May 8 PubMed.
  4. . Hypomorphic Notch 3 alleles link Notch signaling to ischemic cerebral small-vessel disease. Proc Natl Acad Sci U S A. 2011 May 24;108(21):E128-35. Epub 2011 May 9 PubMed.
  5. . Linking Notch signaling to ischemic stroke. Proc Natl Acad Sci U S A. 2008 Mar 25;105(12):4856-61. Epub 2008 Mar 17 PubMed.
  6. . Functional analysis of secreted and transmembrane proteins critical to mouse development. Nat Genet. 2001 Jul;28(3):241-9. PubMed.
  7. . Modulation of Notch signaling by antibodies specific for the extracellular negative regulatory region of NOTCH3. J Biol Chem. 2008 Mar 21;283(12):8046-54. Epub 2008 Jan 8 PubMed.
  8. . Blood biomarkers in a mouse model of CADASIL. Brain Res. 2016 Aug 1;1644:118-26. Epub 2016 May 10 PubMed.
  9. . Pericytes are involved in the pathogenesis of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Ann Neurol. 2015 Dec;78(6):887-900. Epub 2015 Oct 7 PubMed.

Further Reading


  1. . Therapeutic NOTCH3 cysteine correction in CADASIL using exon skipping: in vitro proof of concept. Brain. 2016 Apr;139(Pt 4):1123-35. Epub 2016 Feb 19 PubMed.

Primary Papers

  1. . Therapeutic antibody targeting of Notch3 signaling prevents mural cell loss in CADASIL. J Exp Med. 2017 Aug 7;214(8):2271-2282. Epub 2017 Jul 11 PubMed.