Functional amyloids are widely distributed in bacteria and play important roles in biofilm formation and microbial physiology. However, most currently known bacterial amyloids have been identified through sequence homology to a limited number of prototype proteins, such as the curli subunit CsgA of Escherichia coli. This approach may overlook amyloidogenic sequences that lack recognizable similarity to these canonical systems. In this study, a cross-species, motif-based computational strategy was used to explore whether conserved sequence features derived from mammalian serum amyloid A (SAA) proteins could provide clues for identifying potential amyloidogenic motifs in bacterial proteomes. Comparative analysis of mammalian SAA isoforms identified a conserved sequence segment with predicted aggregation propensity, within which the hydrophobic motif SIAIILCILIL was observed in murine SAA3. Database searches revealed that similar sequence motifs occur in several proteins encoded by Gram-positive bacteria, including multiple proteins in Clostridioides difficile. To further explore whether C. difficile produces extracellular structures capable of interacting with amyloid-binding dyes, Congo Red-supplemented agar assays were performed. After 48 h of growth, both clinical isolates and a laboratory reference strain exhibited Congo Red-binding colony phenotypes. Because Congo Red binding can arise from several extracellular components and cannot be attributed to a specific protein or sequence motif, these observations should be interpreted cautiously. Taken together, this study presents a motif-based computational framework for identifying candidate amyloidogenic motifs across species and highlights sequence features in bacterial proteomes that may warrant further biochemical and structural investigation. The results should be regarded as hypothesis-generating and provide a basis for future experimental validation of potential amyloid-forming proteins in bacteria.
Weichen Gong (Thu,) studied this question.