Legume seed storage proteins are abundant with amyloid-forming sequences

Abstract: Fibrillation of food proteins is widely studied as a way to enhance protein functionality to form gels, films, coatings, and other materials. Legume proteins can be induced to self-assemble into amyloid-like fibrils (protein aggregates that are around 10 nm wide and up to a few μm long). Fibril assembly is commonly induced by heating at low pH that causes hydrolysis, and a subset (normally < 50%) of the protein fragments assemble into fibrils. The core region of the fibrils is composed of beta-sheets aligned perpendicular to the long fibril axis, with flanking surface-exposed regions of poorly-defined structure. Fibrillation of crude protein concentrates or isolates results in a heterogeneous mix of fibril morphologies, yet the exact identities of these amyloid-forming protein fragments is largely unknown. Here, we used LC-MS/MS to identify the fibril core regions of seed storage proteins from pea and soy, along with a bioinformatics analysis of lupin, pea, peanut and soy. The experimental and prediction results were generally in agreement, revealing dozens of fibril core peptides from each of the 7S and 11S proteins. Pea 7S globulins were found to be particularly rich in amyloid regions (108 unique peptides), while soy 11S had the least (43). The core regions were, on average, 11-13 aa in length. While each 7S and 11S variant had some unique amyloid regions, many were also mapped to conserved regions (across variants from the same plant and different plant sources). The abundance of amyloid-forming regions (average of 1 region/45 aa across 70 proteins), including in conserved domains, may indicate a biological role. Indeed, there is some evidence from other groups that seed storage proteins form amyloid in vivo. Insight into the amyloid-core forming regions will be key to the rational design of novel food/materials using nanofibrils derived from legume proteins, and perhaps to understanding biological roles.