The liquid–liquid phase separation (LLPS) of proteins has been found ubiquitously in eukaryotic cells, and is critical in the control of many biological processes by forming a temporary condensed phase with different bimolecular components. TDP-43 is recruited to stress granules in cells and is the main component of TDP-43 granules and proteinaceous amyloid inclusions in patients with amyotrophic lateral sclerosis (ALS). TDP-43 low complexity domain (LCD) is able to de-mix in solution, forming the protein condensed droplets, and amyloid aggregates would form from the droplets after incubation. The molecular interactions regulating TDP-43 LCD LLPS were investigated at the protein fusion equilibrium stage, when the droplets stopped growing after incubation. We found the molecules in the droplet were still liquid-like, but with enhanced intermolecular helix–helix interactions. The protein would only start to aggregate after a lag time and aggregate slower than at the condition when the protein does not phase separately into the droplets, or the molecules have a reduced intermolecular helix–helix interaction. In the protein condensed droplets, a structural transition intermediate toward protein aggregation was discovered involving a decrease in the intermolecular helix–helix interaction and a reduction in the helicity. Our results therefore indicate that different intermolecular interactions drive LLPS and fibril formation. The discovery that TDP-43 LCD aggregation was faster through the pathway without the first protein phase separation supports that LLPS and the intermolecular helical interaction could help maintain the stability of TDP-43 LCD.

No abstract

Liquid-liquid phase separation (LLPS) of proteins has been found ubiquitously in eukaryotic cells, critical in the controlling of many biological processes through forming a temporary condensed phase with different bimolecular components. TDP-43 is recruited to stress granules in cells and is the main component of TDP-43 granules and proteinaceous amyloid inclusions in patients with amyotrophic lateral sclerosis (ALS). TDP-43 low complexity domain (LCD) is able to demix in solution forming the protein condensed droplets. The molecular interactions regulating its LLPS were investigated at the protein fusion equilibrium stage, where the droplets stopped growing. We found the molecules in the droplet were still liquid-like but with enhanced intermolecular helix-helix interaction in the LCD. The protein would start to aggregate after about 200 minutes of lag time and aggregate slower than at the condition when the protein does not phase separate or the molecules have a reduced intermolecular helical interaction. A structural transition intermediate towards protein aggregation was also discovered involving a decrease of the intermolecular helix-helix interaction and a reduction in the helicity. Therefore, LLPS and the intermolecular helical interaction could help maintain the stability of TDP-43 LCD.

Solid-state nuclear magnetic resonance (SSNMR), a technique capable of studying solid or semisolid biological samples, was first applied to study the cell differentiation and mineralization using the whole-cell sample. Mesenchymal stromal cells (MSCs) with multipotent differentiation capacity were induced to differentiate into osteoblasts. The whole differentiation process, osteoblast mineralization and the mineral maturation, was investigated using SSNMR, providing intact, atomic level information on the cellular mineral structural transformation. Our research indicated the extent of osteoblast mineralization could vary significantly for different cell populations whereas the difference was not easily shown by other means of characterization. The SSNMR spectra revealed hydroxylapatite (or hydroxyapatite [HAP]) formation around 2 to 4 weeks after osteogenic induction for MSCs with a high differentiation potency. The early mineral phase deposit before HAP formation contained a high amount of HPO 4 2À . The structures of minerals in the extracellular matrix (ECM) of osteoblasts could evolve for a period of time, even after the incubation of cells has been stopped. This observation was only possible by studying the sample in an intact state, where ECM was not disturbed. These findings improved our understanding of MSCs, which had wide applications in bone regeneration and tissue engineering. Meanwhile, this work demonstrated the advantage of studying these cellular systems as a whole without any mineral extraction, which had been largely overlooked.