Zinc enhances liquid-liquid phase separation of Tau protein and aggravates mitochondrial damages in cells.

Intraneuronal neurofibrillary tangles composed of Tau aggregates have been widely accepted as an important pathological hallmark of Alzheimer's disease. Liquid-liquid phase separation (LLPS) of Tau can lead to its aggregation, and Tau aggregation can then be enhanced by zinc. However, it is unclear whether zinc modulates the formation of Tau stress granules in cells. We herein report that zinc promotes the formation of stress granules containing a pathological mutant ΔK280 of full-length human Tau. Furthermore, zinc promotes LLPS of ΔK280 of full-length Tau, shifting the equilibrium phase boundary to a lower protein concentration, and modulates the liquid nature of droplets formed by this pathological mutation. Zinc also promotes pathological phosphorylation of ΔK280 in neuronal cells, and aggravates mitochondrial damage and elevates reactive oxygen species production induced by Tau aggregation. Importantly, we show that treatment of cells with zinc increases the interaction between full-length Tau and G3BP1 inside stress granules to promote the formation of Tau filaments and increase Tau toxicity in neuronal cells. Collectively, these results demonstrate how Tau condensation and mitochondrial damages induced by Tau aggregation are enhanced by zinc to deteriorate the pathogenesis of Alzheimer's disease, bridging the gap between Tau LLPS and aggregation in neuronal cells.

Pathological concentration of zinc dramatically accelerates abnormal aggregation of full-length human Tau and thereby significantly increases Tau toxicity in neuronal cells.

A pathological hallmark of Alzheimer disease and other tauopathies is the formation of neurofibrillary tangles mainly composed of bundles of fibrils formed by microtubule-associated protein Tau. Here we study the effects of Zn2+ on abnormal aggregation and cytotoxicity of a pathological mutant ΔK280 of full-length human Tau. As revealed by Congo red binding assays, transmission electron microscopy, immunofluorescence, Western blot, and immunogold electron microscopy, pathological concentration of Zn2+ dramatically accelerates the fibrillization of ΔK280 both in vitro and in SH-SY5Y neuroblastoma cells. As evidenced by annexin V-FITC apoptosis detection assay and MTT reduction assay, pathological concentration of Zn2+ remarkably enhances ΔK280 fibrillization-induced apoptosis and toxicity in SH-SY5Y cells. Substitution of Cys-291 and Cys-322 with Ala, however, essentially eliminates such enhancing effects of Zn2+ on the fibrillization and the consequent cytotoxicity of ΔK280. Furthermore, Zn2+ is co-localized with and highly enriched in amyloid fibrils formed by ΔK280 in SH-SY5Y cells. The results from isothermal titration calorimetry show that Zn2+ binds to full-length human Tau by interacting with Cys-291 and Cys-322, forming a 1:1 Zn2+-Tau complex. Our data demonstrate that zinc dramatically accelerates abnormal aggregation of human Tau and significantly increases Tau toxicity in neuronal cells mainly via bridging Cys-291 and Cys-322. Our findings could explain how pathological zinc regulates Tau aggregation and toxicity associated with Alzheimer disease.

Sequence-dependent abnormal aggregation of human Tau fragment in an inducible cell model.

A pathological hallmark of Alzheimer disease (AD) is the accumulation of misfolded hyperphosphorylated microtubule-associated protein Tau within neurons, forming neurofibrillary tangles and leading to synaptic dysfunction and neuronal death. Here we study sequence-dependent abnormal aggregation of human fragment Tau244-372 in an inducible cell model. As evidenced by confocal laser scanning microscopy, Western blot, and immunogold electron microscopy, fibril-forming motifs are essential and sufficient for abnormal aggregation of Tau244-372 in SH-SY5Y neuroblastoma cells induced by Congo red: when its two fibril-forming segments PHF6 and PHF6* are deleted, Tau244-372 does lose its ability to form fibrils in SH-SY5Y cells, and the replacement of PHF6 and PHF6* with an unrelated amyloidogenic sequence IFQINS from human lysozyme does rescue the fibril-forming ability of Tau244-372 in SH-SY5Y cells. By contrast, insertion of a non-fibril forming peptide GGGGGG does not drive the disabled Tau244-372 to misfold in SH-SY5Y cells. Furthermore, as revealed by quantum dots based probes combined with annexin V staining, annexin V-FITC apoptosis detection assay, and immunofluorescence, fibril-forming motifs are essential and sufficient for early apoptosis of living SH-SY5Y cells induced by abnormal aggregation of Tau244-372. Our results suggest that fibril-forming motifs could be the determinants of Tau protein tending to misfold in living cells, thereby inducing neuronal apoptosis and causing the initiation and development of AD.

The role of crowded physiological environments in prion and prion-like protein aggregation.

Prion diseases and prion-like protein misfolding diseases are related to the accumulation of abnormal aggregates of the normal host proteins including prion proteins and Tau protein. These proteins possess self-templating and transmissible characteristics. The crowded physiological environments where the aggregation of these amyloidogenic proteins takes place can be imitated in vitro by the addition of macromolecular crowding agents such as inert polysaccharides. In this review, we summarize the aggregation of prion proteins in crowded physiological environments and discuss the role of macromolecular crowding in prion protein aggregation. We also summarize the aggregation of prion-like proteins including human Tau protein, human α-synuclein, and human copper, zinc superoxide dismutase under macromolecular crowding environments and discuss the role of macromolecular crowding in prion-like protein aggregation. The excluded-volume effects caused by macromolecular crowding could accelerate the aggregation of neurodegenerative disease-associated proteins while inhibiting the aggregation of the proteins that are not neurodegenerative disease-associated.

The contrasting effect of macromolecular crowding on amyloid fibril formation.

BACKGROUND: Amyloid fibrils associated with neurodegenerative diseases can be considered biologically relevant failures of cellular quality control mechanisms. It is known that in vivo human Tau protein, human prion protein, and human copper, zinc superoxide dismutase (SOD1) have the tendency to form fibril deposits in a variety of tissues and they are associated with different neurodegenerative diseases, while rabbit prion protein and hen egg white lysozyme do not readily form fibrils and are unlikely to cause neurodegenerative diseases. In this study, we have investigated the contrasting effect of macromolecular crowding on fibril formation of different proteins. METHODOLOGY/PRINCIPAL FINDINGS: As revealed by assays based on thioflavin T binding and turbidity, human Tau fragments, when phosphorylated by glycogen synthase kinase-3ß, do not form filaments in the absence of a crowding agent but do form fibrils in the presence of a crowding agent, and the presence of a strong crowding agent dramatically promotes amyloid fibril formation of human prion protein and its two pathogenic mutants E196K and D178N. Such an enhancing effect of macromolecular crowding on fibril formation is also observed for a pathological human SOD1 mutant A4V. On the other hand, rabbit prion protein and hen lysozyme do not form amyloid fibrils when a crowding agent at 300 g/l is used but do form fibrils in the absence of a crowding agent. Furthermore, aggregation of these two proteins is remarkably inhibited by Ficoll 70 and dextran 70 at 200 g/l. CONCLUSIONS/SIGNIFICANCE: We suggest that proteins associated with neurodegenerative diseases are more likely to form amyloid fibrils under crowded conditions than in dilute solutions. By contrast, some of the proteins that are not neurodegenerative disease-associated are unlikely to misfold in crowded physiological environments. A possible explanation for the contrasting effect of macromolecular crowding on these two sets of proteins (amyloidogenic proteins and non-amyloidogenic proteins) has been proposed.