How do protein aggregates damage cells? Scientists believe they are toxic themselves, or that they sequester other proteins that have an essential cellular function. In the November 11 Science, researchers led by Frederic Rousseau and Joost Schymkowitz at VIB Switch Laboratory, Leuven, Belgium, add evidence for the latter. They designed a synthetic peptide consisting of an amyloidogenic fragment of vascular endothelial growth factor receptor 2 (VEGFR2). When added to endothelial cell cultures, which depend on VEGF signaling, the fragment triggered clumping of the receptor, and the cells died. When added to cells which express transgenic VEGFR2 but which do not normally make this receptor, the peptide seeded VEGFR2 aggregation, but the cells appeared normal. The findings suggest that at least in this case, toxicity occurs only through depriving cells of an essential protein.

“This is an artificial model, but I think it’s an interesting way to approach the question of how protein aggregation damages cells,” Lary Walker at Emory University, Atlanta, told Alzforum. “The paper demonstrates that amyloid can cause a toxic loss of function.” Walker was not involved in the research.

Synthetic Peptide Binds Native.

An amyloidogenic fragment (green) of VEGFR2 bound to the full protein at ribosomes (red) in endothelial cells (nuclei blue). [Courtesy of Science/AAAS.]

Previous studies of aggregating proteins have turned up evidence for both loss-of-function and gain-of-function mechanisms. For example, both phenomena occur in ataxia and amyotrophic lateral sclerosis (see Mar 2008 news; Oct 2015 conference newsOct 2016 news). Similarly, data conflict on how expanded repeats in C9ORF72 cause neurodegeneration in frontotemporal dementia and ALS (see Oct 2015 conference newsMar 2016 news). One difficulty in resolving this question is that researchers often do not know exactly what normal cellular function amyloidogenic proteins serve.

To overcome this limitation, the authors selected VEGFR2, a well-studied receptor with a known role in cell proliferation and migration. First author Rodrigo Gallardo identified short regions in mouse VEGFR2 that were predicted to be amyloidogenic. Many proteins have such sequences, but do not normally form amyloid, perhaps because the sequences are buried inside the folded protein or are surrounded by charged residues or unstructured sequences that hinder aggregation. The authors synthesized peptides consisting of segments of the VEGFR2 amyloidogenic region, and then tested them for their ability to bind full-length VEGFR2. They found that a peptide, dubbed vascin, with two tandem repeats joined by a proline-proline linker formed protofibrils and fibrils in solution.

To test for toxicity, the authors added vascin to human umbilical vein endothelial cell (HUVEC) cultures. They used both mouse vascin and human vascin, which vary by one amino acid and behave similarly. The cells took up either peptide, which bound to VEGFR2 in the cytoplasm. Vascin hung around ribosomes, suggesting it might glom onto VEGFR2 while the receptor is being translated (see image above). Less VEGFR2 made it to the cell surface, and more of it ended up in insoluble deposits, when vascin was present. Other receptors were unaffected. To test the functional consequences of this aggregation, the authors added VEGF to the cultures. Cells containing vascin had dampened signaling, suggesting a loss of function of VEGFR2.

This loss of receptor function correlated with death of endothelial cells. By contrast, HEK 293 cells transfected with VEGFR2 made aggregates when treated with vascin but stayed healthy. Because HEK 293 cells do not need VEGFR2 for survival, this suggested that HUVECs died from the loss of VEGF signaling and not from an inherent toxic property of the amyloid. Primary neurons also survived equally well when given vascin or a scrambled control peptide, further underlining that aggregates by themselves were not harmful.

Rousseau noted that vascin only triggered aggregation of the specific protein with which it shared homology. “Aggregation is really specific, not promiscuous,” he told Alzforum. In ongoing work, he is exploring whether he can exploit this specificity to knock down harmful proteins. Peptides targeting aggregation-prone regions would cause the unwanted protein to clump, sequestering it and silencing its normal biological activities. If aggregates by themselves are biologically inert, this should not harm the cell, the theory goes. The strategy has paid off in arabidopsis and in maize, in which Rousseau and colleagues knocked down a growth inhibitor to produce bigger plants (see Betti et al., 2016). Likewise, in mice infected with Staphylococcus, peptides directed against several bacterial proteins acted like antibiotics and killed the cells by aggregating the protein (see Bednarska et al., 2016). 

Rousseau and colleagues also want to know if the aggregates present in neurodegenerative diseases primarily damage cells through a loss-of-function mechanism. They are identifying proteins that interact with Aβ and might be sequestered by amyloid plaques. Then they will test whether aggregating just those proteins triggers toxicity. If Aβ takes out a protein needed by a particular cell type, that might explain some of the selective cell vulnerability to neurodegeneration seen in many diseases, Rousseau speculated.

Walker noted that several different types of toxicity are likely to occur in neurodegenerative disease. In some cases, aggregates may themselves cause harm, and in others, small soluble oligomers are believed to do the most damage (see Apr 2002 news). However, Walker agreed that protein sequestration can be toxic, pointing to the example of stress granules in ALS, which bind several proteins and RNA and slow down cellular metabolism (see Mar 2013 news; Nov 2013 conference newsDec 2013 news). “The mechanisms may vary from disease to disease,” he suggested.

Bart De Strooper, a co-author on the Science paper, agreed. “Further work will be needed to prove how general this [sequestering phenomenon] is, but it is an inspiring hypothesis,” he wrote to Alzforum.—Madolyn Bowman Rogers


  1. This paper reports on several important findings derived from a carefully designed study on the VEGFR2 protein. Unlike Aβ-protein or α-synuclein, VEGFR2 does not aggregate in vivo under normal or pathogenic conditions. However, VEGFR2 can be induced to aggregate via an interaction with a de novo designed peptide, referred to as a vascin. The sequence of vascin was derived from a specific VEGFR2 fragment, which was predicted to have amyloidogenic properties and is at the same time homologous to VEGFR2, which allows it to seed its aggregation in a selective way, i.e., without seeding the aggregation of other proteins.

    VEGFR2 was selected for this study because it has a well-defined physiological function associated with specific cells. As vascin is introduced into these specific cells, it causes VEGFR2 to aggregate, therefore inhibiting the ability of VEGFR2 to perform its normal function. This appears to be the source of VEGFR2 aggregation-induced toxicity in these specific cells, for which the normal function of VEGFR2 is essential. However, VEGFR2 aggregation in other types of cells does not cause toxicity as those other cells do not depend on the normal function of VEGFR2, which implies that VEGFR2 aggregates are not particularly toxic on their own.

    It is likely that other proteins that do not aggregate under normal or pathological conditions could be induced to aggregate through seeding by a de novo designed peptide that is homologous to the original protein. As in the case of VEGFR2, this induced aggregation could lead to a loss-of-function induced toxicity in those cells, for which the protein function is of essence. This is definitely a very important discovery, as it provides a way to control specific protein function by manipulating protein aggregation.

    Proteins associated with neurodegenerative diseases, such as Aβ-protein, α-synuclein, tau, and others, on the other hand, are aggregation-prone, at least under pathological conditions, but perhaps also under normal conditions, which makes them different from VEGFR2. Moreover, their physiological role is still not completely understood. For example, evidence is emerging that Aβ is an antimicrobial and antiviral protein (the associated Alzforum story was posted online several months ago). If the normal function of Aβ in vivo is to act mainly as an immune defense against microbes and viruses, then Aβ aggregation must occur under non-AD conditions. A loss of function would lead to an increase of microbial and viral infections, but would not necessarily lead to a massive neuronal loss observed in AD.

    Thus, there is to date no convincing evidence that AD is a consequence of a loss of function of Aβ. However, there is a correlation between neuronal loss and aberrant aggregation of tau into neurofibrillary tangles. Tau has an important physiological function, which is lost when tau aggregates into neurofibrillary tangles in AD. However, the molecular basis of AD is multifaceted and the exact molecular mechanisms underlying the pathology are not well understood, so it is too early to speculate as to which extent the findings of this study may be applicable to neurodegenerative diseases.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Research Brief: Gain or Loss of Function? Ataxin Mutation Cuts Both Ways
  2. C9ORF72 Mice Point to Gain of Toxic Function in ALS, FTD
  3. Grow that Axon: C9ORF72 Function Revealed?
  4. C9ORF72 Dipeptide Repeat Not Enough for Motor Neuron Disease
  5. Earliest Amyloid Aggregates Fingered As Culprits, Disrupt Synapse Function in Rats
  6. Disease Mutations Zip Lock Stress Granules in Proteinopathy, ALS
  7. Profilin-1 Links Cytoskeleton and RNA Aggregation in ALS
  8. Stress Relief: Anti-Stress Granule Therapy Saves ALS Models

Paper Citations

  1. . Sequence-Specific Protein Aggregation Generates Defined Protein Knockdowns in Plants. Plant Physiol. 2016 Jun;171(2):773-87. Epub 2016 May 4 PubMed.
  2. . Protein aggregation as an antibiotic design strategy. Mol Microbiol. 2016 Mar;99(5):849-65. Epub 2015 Dec 9 PubMed.

Further Reading

Primary Papers

  1. . De novo design of a biologically active amyloid. Science. 2016 Nov 11;354(6313) PubMed.