“This paper is among the first to show just how important amyloid formation is in normal cellular processes in mammalian cells,” said David Teplow at the University of California, Los Angeles. He praised the thoroughness of the work, noting, “In every possible way you could imagine from a physical biochemistry point of view, [the authors] showed that these things are classic amyloid structures. It is beautiful and impressive work.”

In the last few years, accumulating evidence indicates that amyloids can serve multiple physiologic functions, particularly in yeast and bacteria. To date, known roles mainly include structural tasks such as protein storage, scaffolding, cell adhesion, and biofilm formation (see ARF related news story; ARF news story; and Ramsook et al., 2010). Amyloids have not previously been connected to regulated cell death pathways.

Programmed cell death can be either apoptotic, in which cells shrivel neatly away, or necrotic, in which dying cells rupture messily (see Vandenabeele et al., 2010). Receptor interacting protein kinases 1 and 3 (RIP1 and RIP3) act as master switches that regulate cell fate. Depending on physiological conditions, these proteins can mediate cell survival, apoptosis, or necrosis. To initiate necrosis, RIP1 and RIP3 form a protein complex known as the necrosome. The kinases then phosphorylate and activate downstream proteins in the necrotic pathway (see Sun et al., 2012; Zhao et al., 2012). RIP1 and RIP3 are known to bind each other through regions called RIP homotypic interaction motifs (RHIMs), but the nature of the necrosome had never been elucidated.

To examine this structure, first author Jixi Li coexpressed tagged human RIP1 and RIP3 in cell cultures. Surprisingly, the proteins formed very large complexes. Li and colleagues used a wide variety of biochemical techniques, including electron microscopy, circular dichroism spectroscopy, Fourier transform infrared spectroscopy, and solid-state nuclear magnetic resonance spectroscopy, to show that these complexes are amyloid fibrils. The fibrils display classic amyloidogenic properties. They bind amyloid dyes such as thioflavin T and Congo red, and are highly stable, dissolving only under harsh chemical conditions, the authors report. In addition, RIP1/RIP3 complexes can seed rapid polymerization in monomeric solutions of RIP kinases, another property of amyloid.

Structural analysis revealed that the fibrils consist of a buried amyloid core surrounded by long, flexible extensions. The RIP kinase domains are found in these extensions and likely maintain their globular shape and activity, the authors note. To define the amyloid-forming region, the researchers made short peptides from the RHIM domain of each protein. They found that fragments as short as 19 amino acids (RIP3) and 31 amino acids (RIP1) still bound each other. By mutating residues in these core regions one by one, the authors showed that only mutations in the central six or seven amino acids of each RHIM disrupted amyloid formation. This confirmed that the amyloid core is very small.

The authors then looked at function. Human cell cultures transfected with wild-type RIP kinases underwent necrosis when stimulated with the appropriate cell death signals. By contrast, cells transfected with amyloid-defective RIP mutants thrived, ignoring death signals. Likewise, treating cell cultures with amyloid dye inhibited the formation of the amyloid complex and prevented necrosis. These experiments indicate that the amyloid structure is essential for proper necrotic signaling. Intriguingly, the mutants that could not form amyloid also had little kinase activation, suggesting that amyloid formation helps turn on the kinases.

Based on these findings, the authors propose a model in which the long, disordered flanking sequences of RIP monomers hide the core RHIM sequences under normal cellular conditions, preventing the proteins from interacting with each other to form amyloid. Cell death signals switch on the kinases, and they autophosphorylate. The charged phosphate groups repel each other, opening up the proteins and exposing the RHIM regions. RIP1 and RIP3 can then bind to each other, which spurs further auto- and cross-phosphorylation, leading to even more aggregation in a feed-forward loop, the authors suggest. This model makes sense, Teplow agreed, because cells would need to control the formation of amyloid. Otherwise, the amyloid fibrils might choke the cell, or initiate unwanted cell death under normal conditions. Phosphorylation may act as a sensitive regulatory mechanism over this process.

What happens to necrotic amyloid fibrils after their original cells die? Wu does not know, but said she is investigating this question. She wonders whether these necrotic amyloid structures could cross-talk with β amyloid fibrils, perhaps helping to worsen Alzheimer’s disease. In future work, she and Chan will test this idea in vitro and in AD mouse models.

Cell death expert Peter Vandenabeele at VIB and Ghent University, Belgium, agreed, writing to ARF that there might be a link to AD: “[Because] amyloid structures may induce co-aggregation of other amyloidogenic proteins, it is conceivable that necrosis processes may contribute to AD pathogenesis.” He added that if so, it might be possible to delay AD onset with RIP kinase inhibitors (see full comment below).—Madolyn Bowman Rogers.

Reference:
Li J, McQuade T, Siemer AB, Napetschnig J, Moriwaki K, Hsiao YS, Damko E, Moquin D, Walz T, McDermott A, Chan FK, Wu H. The RIP1/RIP3 necrosome forms a functional amyloid signaling complex required for programmed necrosis. Cell. 2012 Jul 20;150(2):339-50. Abstract