What triggers the vascular Aβ deposits known as cerebral amyloid angiopathy (CAA)? In the November 16 Nature, researchers led by Jonas Neher at the German Center for Neurodegenerative Diseases in Tübingen, Germany, nominate medin, a small aggregating fragment of a vascular cell adhesion protein called lactadherin. In vitro, medin bound Aβ, promoting fibrilization. In a mouse model of amyloidosis, injecting aggregated medin accelerated Aβ deposition, while knocking it out nearly abolished CAA. Medin permeates CAA in human brains as well, and higher expression of lactadherin correlates with cognitive decline and Alzheimer’s risk.

  • Medin, a fragment of a cell adhesion protein, forms amyloid in blood vessels.
  • This amyloid promotes vascular amyloid-β deposits by fibrilizing with Aβ.
  • Lowering medin protects against CAA, suggesting therapeutic potential.

“Medin plays an important role in neurovascular disease that has been overlooked,” Neher told Alzforum. He noted that few previous studies have investigated this protein, and he suggested that targeting medin could offer a new therapeutic angle for improving vascular function in the aging brain.

Martin Citron at UCB Pharma in Brussels believes the data are compelling. “It’s an exciting story, and I think it will make some waves,” he told Alzforum. Berislav Zlokovic and Kassandra Kisler at the University of Southern California, Los Angeles, called the paper a tour de force, but noted more work is needed to elucidate exactly how medin affects amyloid pathology and cognition (full comment below).

Medin Malfeasance. In the cortices of 2-year-old amyloidosis mice (top), Aβ deposits (red) in both parenchyma and blood vessels are packed with medin (green). Abolishing medin (bottom) lessens plaques and nearly eliminates cerebral amyloid angiopathy. [Courtesy of Wagner et al., 2022.]

Medin is cleaved from lactadherin, also known as milk fat globule-epidermal growth factor 8 (MFGE8), a membrane glycoprotein that connects smooth muscle cells in the vasculature to the surrounding basement membrane (Larsson et al., 2006). Medin is the major component of the amyloid found in the medial layer of the aorta in nearly all Caucasians over the age of 50 (Mucchiano et al., 1992; Häggqvist et al., 1999; Peng et al., 2005). Medin deposits have long been associated with aortic aneurysms, but were more recently linked to cerebrovascular disease and Alzheimer’s (Peng et al., 2007; Karamanova et al., 2020; Migrino et al., 2020). However, whether medin is a cause or consequence of AD is unclear.

Neher and colleagues turned to mice to investigate. They had previously identified medin aggregates in the blood vessels of year-old wild-type mice, and tied these deposits to loss of vascular elasticity, suggesting mice could be a good model organism for medin amyloid (Degenhardt et al., 2020). To dovetail that with Aβ amyloidosis, they chose to study the APP23 model, which develops parenchymal plaques at 6 months and CAA at 12 (Calhoun et al., 1999). 

Joint first authors Jessica Wagner and Karoline Degenhardt immunostained cortical brain sections of 2-year-old mice for medin, and found that Aβ plaques and vascular deposits were packed with it. When they eliminated medin by crossing APP23 mice with MFGE8 knockouts, plaque area dropped by half, and CAA was slashed by 85 percent (see image above). Conversely, injecting medin aggregates into the hippocampi of 3-month-old APP23 mice boosted plaque density six months later. Depleting medin from the samples before injection prevented plaques. Together, these data suggested that medin promotes Aβ deposition.

Plaque Promotion. Nine-month-old APP23 mice (left) have few amyloid plaques (blue) in the hippocampus, unless they were previously injected with seeds of aggregated medin (middle) or Aβ (right). [Courtesy of Wagner et al., 2022.]

The authors tested this idea in vitro, where they found that medin and Aβ40 bound each other to form fibrils. This makes sense, because medin has a very similar amino acid sequence to Aβ, the authors noted. By itself, medin forms an amorphous, disordered amyloid. When it co-aggregated with Aβ, the resulting deposits were looser and less fibrillar compared to pure amyloid β aggregates, as seen by electron microscopy. Not only did the two proteins fibrilize together, but seeds of aggregated medin could accelerate aggregation of Aβ40 in solution. Again, this suggested that medin could kick off Aβ amyloidosis.

Are the findings relevant to people? The authors stained for medin in postmortem frontal lobe sections from 16 AD patients. The fragment was abundant in vascular amyloid (see image below), but unlike in mice, did not occur in parenchymal plaques. To glean more clues to human expression of MFGE8, the authors analyzed RNA-Seq data from 566 participants in the ROSMAP longitudinal cohort. These data showed that vascular smooth muscle cells were likely the primary source of lactadherin, with endothelial cells secondary, and very little expression from other brain cells. This contrasts with mice, where astrocytes also express the protein, perhaps explaining why medin occurs in parenchymal plaques in their brains.

Medin Packs CAA. In a cortical blood vessel from an AD patient (left), medin (green), Aβ (blue), and smooth muscle cell protein (red) co-localize almost perfectly (right). [Courtesy of Wagner et al., 2022.]

Moreover, in the ROSMAP cohort, AD brains expressed more MFGE8 mRNA than did control brains, and this correlated with faster cognitive decline on the MMSE. This association was independent of plaque and tangle burden, and unaffected by age, sex, level of education, APOE genotype, or postmortem interval. The data hint that medin independently accelerates cognitive decline. Other genetic data supports this relationship. The FinnGen genetic study recently identified an MFGE8 mutation that truncates the protein, eliminating the medin sequence, and this protects against coronary disease and dementia (Ruotsalainen et al., 2022; Jukarainen et al., 2022). 

Neher believes medin is a promising therapeutic target. Because the fragment has no known physiological role, eliminating it should not cause side effects. Also, being in the vasculature, medin might be easily accessible to drugs. His group has made mice with humanized medin and is exploring whether immunotherapy could mop up these deposits. They are also investigating whether there are different isoforms of medin, and whether medin amyloid contributes to other cerebrovascular diseases.

David Holtzman at Washington University in St. Louis noted that medin knockouts abolish CAA less thoroughly than do ApoE knockouts, suggesting ApoE may be a more effective target. “However, there may be advantages in targeting medin, and further studies to assess this are warranted,” he wrote to Alzforum (full comment below).

Citron sees several avenues for future research. He wondered if medin might interact more with Aβ peptides incorporating the Dutch mutation, which form mainly vascular amyloid, than it does with other Aβ species, perhaps explaining why the Dutch variant preferentially causes vascular pathology. He also suggested that a ligand that binds medin might make a good PET tracer for specifically imaging CAA in living people. CAA is currently assessed by looking for signs of microhemorrhages on MRI, or by amyloid PET scans, where the signal can be swamped by parenchymal plaques. A specific PET tracer could be quite useful, Citron suggested.

Steven Greenberg at Massachusetts General Hospital, Boston, noted that disentangling causes from effects of CAA and AD is challenging, and more work will be needed to prove a causal role for medin. “The experiments in this paper nonetheless form a strong foundation for further study of this potentially important contributor to the pathogenesis of the two diseases,” he wrote (comment below).—Madolyn Bowman Rogers


  1. This interesting study, suggesting that medin promotes vascular amyloid accumulation in AD, expands on the authors’ and others’ previous work. Using two mouse models of Aβ amyloidosis, and bolstered by human data and structural data, the authors have achieved a tour de force. However, this study brings up several questions that remain to be answered.

    The authors’ data suggests that medin generated by vascular smooth-muscle cells, and possibly other cells of the neurovascular unit, could play a role in accelerating Aβ pathology and CAA, based on their immunostaining and RNA-Seq analysis. But, how medin interacts with Aβ to promote fibrilization remains still to be investigated. It will be also important to determine which cells are responsible for medin expression and its interaction with Aβ, as this information will be helpful in identifying targets for therapeutic interventions. And, importantly, what would be the effect of medin load on cognitive function?

    Additional work is warranted to tease out the details of medin function and mechanism in relation to amyloid pathology and cognitive decline.

  2. I think this is a very nice paper and shows good evidence that medin does influence CAA formation and would be a potential novel target for CAA treatment.

    The effect of knocking out medin on CAA is not nearly as strong as knocking out ApoE, which, in similar models, blocks CAA formation completely.  However, there may be advantages in targeting medin and further studies to assess this are warranted.

  3. This very interesting article points to the 50-amino-acid peptide, medin, as a candidate cofactor in promoting Aβ deposition, particularly in cerebral blood vessels, i.e., cerebral amyloid angiopathy (CAA). The suggestion, based on ROSMAP data, that the precursor protein of medin, MFGE8, is upregulated in cerebral smooth muscle and endothelial cells from Alzheimer’s disease brains is particularly intriguing, because it suggests medin expression is localized to the cerebral blood vessels.

    Disentangling causes from effects of CAA and of Alzheimer’s disease can be very challenging and it is not yet clear that medin will fall into the cause rather than the effect category. Nonetheless, the experiments in this paper form a strong foundation for further study of this potentially important contributor to the pathogenesis of these two diseases.

Make a Comment

To make a comment you must login or register.


Research Models Citations

  1. APP23

Paper Citations

  1. . Lactadherin binds to elastin--a starting point for medin amyloid formation?. Amyloid. 2006 Jun;13(2):78-85. PubMed.
  2. . Senile aortic amyloid. Evidence for two distinct forms of localized deposits. Am J Pathol. 1992 Apr;140(4):871-7. PubMed.
  3. . Medin: an integral fragment of aortic smooth muscle cell-produced lactadherin forms the most common human amyloid. Proc Natl Acad Sci U S A. 1999 Jul 20;96(15):8669-74. PubMed.
  4. . Medin-amyloid: a recently characterized age-associated arterial amyloid form affects mainly arteries in the upper part of the body. Amyloid. 2005 Jun;12(2):96-102. PubMed.
  5. . Role of aggregated medin in the pathogenesis of thoracic aortic aneurysm and dissection. Lab Invest. 2007 Dec;87(12):1195-205. Epub 2007 Oct 1 PubMed.
  6. . Endothelial Immune Activation by Medin: Potential Role in Cerebrovascular Disease and Reversal by Monosialoganglioside-Containing Nanoliposomes. J Am Heart Assoc. 2020 Jan 21;9(2):e014810. Epub 2020 Jan 13 PubMed.
  7. . Cerebrovascular medin is associated with Alzheimer's disease and vascular dementia. Alzheimers Dement (Amst). 2020;12(1):e12072. Epub 2020 Aug 25 PubMed.
  8. . Medin aggregation causes cerebrovascular dysfunction in aging wild-type mice. Proc Natl Acad Sci U S A. 2020 Sep 22;117(38):23925-23931. Epub 2020 Sep 8 PubMed.
  9. . Neuronal overexpression of mutant amyloid precursor protein results in prominent deposition of cerebrovascular amyloid. Proc Natl Acad Sci U S A. 1999 Nov 23;96(24):14088-93. PubMed.
  10. . Inframe insertion and splice site variants in MFGE8 associate with protection against coronary atherosclerosis. Commun Biol. 2022 Aug 17;5(1):802. PubMed.
  11. . Genetic risk factors have a substantial impact on healthy life years. Nat Med. 2022 Sep;28(9):1893-1901. Epub 2022 Sep 12 PubMed.

Further Reading

Primary Papers

  1. . Medin co-aggregates with vascular amyloid-β in Alzheimer's disease. Nature. 2022 Dec;612(7938):123-131. Epub 2022 Nov 16 PubMed.