The N-terminal domain of the prion protein is required and sufficient for liquid–liquid phase separation: A crucial role of the Aβ-binding domain

Abstract

Formation of biomolecular condensates through liquid–liquid phase separation (LLPS) has been described for several pathogenic proteins linked to neurodegenerative diseases and is discussed as an early step in the formation of protein aggregates with neurotoxic properties. In prion diseases, neurodegeneration and formation of infectious prions is caused by aberrant folding of the cellular prion protein (PrP C ). PrP C is characterized by a large intrinsically disordered N-terminal domain and a structured C-terminal globular domain. A significant fraction of mature PrP C is proteolytically processed in vivo into an entirely unstructured fragment, designated N1, and the corresponding C-terminal fragment C1 harboring the globular domain. Notably, N1 contains a polybasic motif that serves as a binding site for neurotoxic Aβ oligomers. PrP can undergo LLPS; however, nothing is known how phase separation of PrP is triggered on a molecular scale. Here, we show that the intrinsically disordered N1 domain is necessary and sufficient for LLPS of PrP. Similar to full-length PrP, the N1 fragment formed highly dynamic liquid-like droplets. Remarkably, a slightly shorter unstructured fragment, designated N2, which lacks the Aβ-binding domain and is generated under stress conditions, failed to form liquid-like droplets and instead formed amorphous assemblies of irregular structures. Through a mutational analysis, we identified three positively charged lysines in the postoctarepeat region as essential drivers of condensate formation, presumably largely via cation–π interactions. These findings provide insights into the molecular basis of LLPS of the mammalian prion protein and reveal a crucial role of the Aβ-binding domain in this process.

Biological Functions of the Intrinsically Disordered N-Terminal Domain of the Prion Protein: A Possible Role of Liquid–Liquid Phase Separation

“…Consistent with the concept that biomolecular condensates can be precursors of pathogenic protein aggregates, it was shown that after LLPS of recombinant full-length PrP, a rapid liquid-solid phase transition occurred, leading to the formation of β-sheet-rich and PK-resistant amyloid [94]. Notably, the unstructured N-terminal domain of PrP was required for the initial LLPS, supporting our finding that LLPS of PrP-C1 and -C2 is impaired [99]. Interestingly, it was already described some time ago that large globular protein assemblies preceded the formation of PrP145X amyloid fibrils in vitro.…”

“…Our study revealed that the intrinsically disordered N1 fragment of PrP is necessary and sufficient for the formation of biomolecular condensates, emphasizing the concept that intrinsically disordered and low-complexity regions are important drivers of phase separation [100]. Furthermore, a mutational analysis revealed that LLPS of N1 is governed at the molecular level mainly by intermolecular cation-π interactions of the positively charged residues in PB1 and PB2 with aromatic side chains [99].…”

“…In a recent study, we provided insight into the mechanism underlying the propensity of the mammalian prion protein to undergo LLPS [99]. Our study revealed that the intrinsically disordered N1 fragment of PrP is necessary and sufficient for the formation of biomolecular condensates, emphasizing the concept that intrinsically disordered and low-complexity regions are important drivers of phase separation [100].…”

“…Although there are only few publications to date that have examined LLPS of PrP in detail [94][95][96][97][98][99], they provided experimental evidence that LLPS could play a role in regulating (patho)physiological activities of PrP and formation of infectious prions. Consistent with the concept that biomolecular condensates can be precursors of pathogenic protein aggregates, it was shown that after LLPS of recombinant full-length PrP, a rapid liquid-solid phase transition occurred, leading to the formation of β-sheet-rich and PK-resistant amyloid [94].…”

“…Another intriguing observation is that binding of neurotoxic Aβ oligomers to the polybasic motifs converts liquid-like droplets of full length PrP into hydrogel and induces a conformation change of PrP [97,102], suggesting that aberrant phase transition of PrP C may be associated with the activity of the PrP/Aβ complex to induce neurotoxic signaling [97]. Thus far, a possible role of LLPS in the neuroprotective activities of full length PrP or the liberated N-terminal fragments has only been indirectly demonstrated: PrP variants that lack a stress protective activity in cell culture and animal models [58,67] failed to undergo LLPS in vitro [94,99]. Vice versa, N1 formed biomolecular condensates in vitro [99] and showed stress-protective activity in cell culture models [69,71].…”