Optimizing grid preparation methods for TEM imaging of amyloid-forming proteins

Transmission electron microscopy (TEM), together with Thioflavin T (ThT) fluorescence assays, is widely used to visualize amyloid fibrils and to characterize the kinetics of amyloid formation. However, discrepancies between ThT fluorescence data and TEM observations are sometimes reported, which may arise from limitations in fibril visualization by TEM. In particular, TEM imaging can be strongly influenced by the sample loading procedure on the grid, which governs fibril deposition and retention. In this work, five different grid preparation methods were compared to evaluate their efficiency in detecting and visualizing human islet amyloid polypeptide (hIAPP) fibrils, which are present in 95% of patients with type 2 diabetes mellitus. The methods were assessed based on detection speed, morphological representation, fibril abundance and grid contamination. The two best-performing methods were further evaluated for detecting early hIAPP aggregates and subsequently applied to another amyloid forming protein, namely amyloid-β 42 (Aβ₄₂), which is involved in Alzheimer's disease. Among the tested approaches, method 2 (a droplet-deposition protocol) and method 3 (a centrifugation-based loading protocol) provided the most efficient fibril detection and morphological representation. Method 2 was identified as the best compromise between rapid detection, experimental simplicity, and low grid contamination, and was further tested under different buffer conditions. Overall, this comparative study demonstrated that variations in grid preparation protocols can significantly influence TEM observations and provide practical guidance for selecting optimal conditions for amyloid fibril imaging depending on experimental objectives.