Novel amplification mechanism of prions through disrupting sortilin-mediated trafficking

ABSTRACT

Conformational conversion of the cellular prion protein, PrP^C, into the abnormally folded isoform of prion protein, PrP^Sc, which leads to marked accumulation of PrP^Sc in brains, is a key pathogenic event in prion diseases, a group of fatal neurodegenerative disorders caused by prions. However, the exact mechanism of PrP^Sc accumulation in prion-infected neurons remains unknown. We recently reported a novel cellular mechanism to support PrP^Sc accumulation in prion-infected neurons, in which PrP^Sc itself promotes its accumulation by evading the cellular inhibitory mechanism, which is newly identified in our recent study. We showed that the VPS10P sorting receptor sortilin negatively regulates PrP^Sc accumulation in prion-infected neurons, by interacting with PrP^C and PrP^Sc and trafficking them to lysosomes for degradation. However, PrP^Sc stimulated lysosomal degradation of sortilin, disrupting the sortilin-mediated degradation of PrP^C and PrP^Sc and eventually evoking further accumulation of PrP^Sc in prion-infected neurons. These findings suggest a positive feedback amplification mechanism for PrP^Sc accumulation in prion-infected neurons.

KEYWORDS:

• Prions
• prion protein
• protein degradation
• lysosome
• sortilin
• sorting
• VPS10P sorting receptor

INTRODUCTION

Prions are causative agents of prion diseases, a group of fatal neurodegenerative disorders including Creutzfeldt-Jakob disease in humans and bovine spongiform encephalopathy and scrapie in animals.¹ They are widely believed to consist of the abnormally folded, amyloidogenic isoform of prion protein, designated PrP^Sc.¹ PrP^Sc is produced through conformational conversion of the cellular prion protein, PrP^C, by unknown mechanisms.¹ PrP^C is a glycosylphosphatidylinositol (GPI)-anchored membrane glycoprotein expressed most abundantly in brains, particularly by neurons.² The constitutive conversion of PrP^C into PrP^Sc leads to accumulation of PrP^Sc in brains. We and others have shown that the conversion of PrP^C into PrP^Sc is a key pathogenic event in prion disease, by demonstrating that mice devoid of PrP^C neither developed the disease nor propagated prions or accumulated PrP^Sc in their brains after intracerebral inoculation with prions.^3–6 Most pathogens usually evade host defense mechanisms to propagate themselves in their hosts. However, the host defense mechanism against prions to suppress prion propagation, or PrP^Sc accumulation, remains unknown.

The vacuolar protein sorting-10 protein (VPS10P)-domain receptors, including sortilin, SorLA, SorCS1, SorCS2 and SorCS3, are multi-ligand type I transmembrane proteins abundantly expressed in brains and involved in neuronal function and viability.^7,8 They function as a cargo receptor to deliver a number of cargo proteins to their subcellular compartments through the VPS10P domain in the extracellular luminal N-terminus.^7,8 Accumulating lines of evidence indicate that altered VPS10P receptor-mediated trafficking could be involved in the pathogenesis of neurodegenerative disorders, including Alzheimer's disease^9–12 and frontotemporal lobar degeneration.¹³ Sortilin mediates intracellular trafficking of the amyloid precursor protein (APP)-cleaving enzyme BACE1¹⁴ and the neurotrophic factor receptors Trks.¹⁵ SorLA and SorCS1 are involved in APP transport.^9,11

We recently reported that sortilin negatively regulates PrP^Sc accumulation by sorting PrP^C and PrP^Sc to lysosomes for degradation, and that PrP^Sc accumulation itself impairs the sotrtilin-mediated degradation of PrP^C and PrP^Sc by stimulating lysosomal degradation of sortilin, thereby evoking further accumulation of PrP^Sc in prion-infected cells.¹⁶ These findings suggest that the sortilin-mediated lysosomal degradation of PrP^C and PrP^Sc could be a host defense mechanism against prions, and that prions, or PrP^Sc, could propagate in infected neurons by evading the sortilin-mediated defense mechanism by inducing lysosomal degradation of sortilin.

SORTILIN IS A NEGATIVE REGULATOR FOR PRP^Sc ACCUMULATION

We found that PrP^C directly interacts with sortilin, but not with other VPS10P molecules, on the plasma membrane in PrP^C-overexpressing neuroblastoma N2a cells, designated N2aC24 cells.¹⁶ The interaction of both molecules was also confirmed in mouse brain homogenates.¹⁶ SiRNA-mediated knockdown of sortilin increased PrP^Sc in prion-infected N2aC24L1-3 cells, which are N2aC24 cells persistently infected with 22L scrapie prions.¹⁶ In contrast, overexpression of sortilin in N2aC24L1-3 cells decreased PrP^Sc.¹⁶ We also showed that sortilin-knockout mice had accelerated prion disease caused by early accumulation of PrP^Sc in their brains after infection with RML scrapie prions.¹⁶ These results indicate that sortilin could negatively regulate PrP^Sc accumulation in prion-infected cells and mice.

SORTILIN TRAFFICS PRP^C TO NON-RAFT DOMAINS AND TO LATE ENDOSOMES/LYSOSOMES

PrP^C is synthesized in the endoplasmic reticulum (ER) and trafficked to the plasma membrane through the Golgi apparatus.¹⁷ PrP^C undergoes several posttranslational modifications during its biosynthesis, including cleavage of the N-terminal signal peptide, removal of the C-terminal peptide for attachment of a GPI anchor at the C-terminus and formation of a disulfide bond at the C-terminal domain in the ER, and addition of two core N-linked oligosaccharides at the C-terminal domain in the ER that are further modified in the ER and then in the Golgi apparatus.¹⁷ Like other GPI-anchored proteins, PrP^C is predominantly located at raft domains and, to a lesser extent, at non-raft domains.^16,17 After internalization, some PrP^C molecules are delivered back to the plasma membrane directly or indirectly via the recycling endosome compartments and others are transported to lysosomes for degradation.¹⁸ Copper and zinc stimulate endocytosis of PrP^C by binding to histidine residues in the octapeptide repeat (OR) region located in the N-terminal domain.^19–21 It has been postulated that PrP^C interacts with an as yet unidentified raft molecule via the N-terminal domain including the OR region, thereby being retained at raft domains.²⁰ The binding of copper or zinc to the OR region causes structural changes in the N-terminal interacting region of PrP^C, thereby PrP^C leaves raft domains to non-raft domains to be endocytosed via the clathrin-dependent pathway.²⁰ Low-density lipoprotein receptor-related protein 1 has been reported to be involved in the clathrin-dependent endocytosis of PrP^C.²² The clathrin-independent pathways including caveolae, which is considered to be formed by clustering raft domains, or caveolae-like domains have been also reported to mediate the endocytosis of PrP^C.¹⁸

We found that sortilin was predominantly located at non-raft domains in prion-uninfected N2aC24 cells.¹⁶ Sortilin knockout caused marked shift in localization of PrP^C from non-raft domains to raft domains in N2aC24 cells and mouse brains.¹⁶ These findings suggest that sortilin could function to recruit PrP^C from raft domains to non-raft domains. We also found that, after internalization, PrP^C was transported to both late endosomes and recycling endosomes in N2aC24 cells.¹⁶ However, PrP^C was preferentially transported to recycling endosomes with reduced localization at late endosomes/lysosomes in sortilin-knockdown and -knockout N2aC24 cells,¹⁶ indicating that sortilin also could function as an endocytic receptor for PrP^C at non-raft domains to be sent to lysosomes for degradation (Fig. 1A). Consistent with this, sortilin-deficient N2aC24 cells showed higher PrP^C on their plasma membranes than control N2aC24 cells.¹⁶ Sortilin-knockout mice also showed higher PrP^C in their brains compared to WT mice.¹⁶ Moreover, inhibition of lysosomal enzymes by NH₄Cl increased PrP^C markedly in N2aC24 cells, but only slightly in sortilin-knockout N2aC24 cells.¹⁶

FIGURE 1. A model of the sortilin-mediated intracellular trafficking of PrP^C and PrP^Sc in prion-uninfected and infected neurons. (A) Sortilin-dependent and -independent endocytosis of PrP^C in uninfected neurons. Sortilin mediates endocytosis of PrP^C on the plasma membrane (PM), particularly at non-raft domains, via the clathrin-dependent pathway to early endosomes (EE) and then traffics it to late endosomes/lysosomes (LE/Lys) for degradation. Other PrP^C molecules are trafficked either to LE/Lys for degradation or to the recycling endosome (RE) pathway in a sortilin-independent way. There also might be sortilin-dependent and -independent trafficking pathways from the Golgi apparatus (GA) to LE/Lys for degradation. (B) Intracellular trafficking of PrP^C and PrP^Sc in prion-infected neurons. Prion infection stimulates lysosomal degradation of sortilin via an unknown mechanism, thereby impairing the sortilin-mediated trafficking of PrP^C and PrP^Sc to LE/Lys for degradation. As a result, PrP^C and PrP^Sc are increased at raft domains and endocytosed via the sortilin-independent pathway to RE, causing accumulation of PrP^Sc and increasing conversion of PrP^C into PrP^Sc in prion-infected neurons. PrP^Sc could undergo retrograde transport to the GA. However, sortilin might also be functionally impaired in the GA, thereby being unable to traffic PrP^Sc in the GA to LE/Lys for degradation. The decreased degradation of PrP^Sc in LE/Lys and the increased conversion of PrP^C into PrP^Sc in raft domains or RE could both contribute to the constitutive production of PrP^Sc in prion-infected neurons. Dashed arrows indicate restricted trafficking.

The plasma membrane or raft domains are considered to be major sites for the conversion of PrP^C into PrP^Sc,²³ although the exact site of PrP^Sc production remains controversial. It is thus likely that sortilin could negatively regulate PrP^Sc accumulation by reducing PrP^C on the plasma membrane, particularly at raft domains through recruiting PrP^C to non-raft domains from raft domains and sorting it to the late endosome/lysosome protein degradation pathway.

SORTILIN IS INVOLVED IN DEGRADATION OF PRP^Sc

We also found that sortilin could function to direct PrP^Sc for degradation.¹⁶ Sortilin interacted with PrP^Sc in prion-infected N2aC24L1-3 cells.¹⁶ Sortilin-knockout significantly slowed down the degradation of PrP^Sc in N2aC24 cells infected with RML or 22L prions.¹⁶ PrP^Sc is found at various intracellular compartments, including the plasma membrane, various endosomal compartments such as early and late endosomes, recycling endosomes, and lysosomes, and the Golgi apparatus.¹⁸ Enzymatic release of PrP^C from the plasma membrane by phosphoinositide-specific phospholipase C was shown to reduce PrP^Sc in infected cells,²⁴ and formation of PrP^Sc was inhibited by lowered temperature,²⁵ which blocks the endocytosis and internalization of PrP^C. These suggest that the conversion of PrP^C into PrP^Sc might occur at the plasma membrane, where exogenous PrP^Sc is likely to first contact endogenous PrP^C, or after its internalization in the endosomal compartment. Internalized PrP^Sc could also undergo retrograde transport to the Golgi apparatus and/or to the ER,^26,27 where the transported PrP^Sc might trigger the conversion of PrP^C into PrP^Sc. PrP^Sc molecules on the plasma membrane are trafficked to lysosomes for degradation via the endolysosomal pathway.^28,29 The PrP^Sc retrogradely transported to the Golgi apparatus are subjected to Golgi quality control and trafficked to lysosomes for degradation.²⁷ Sortilin localizes in para-nuclear vesicles, in the trans-Golgi network, and on the plasma membrane.^30,31 It is thus possible that sortilin could be involved in both degradation trafficking pathways of PrP^Sc. However, sortilin and PrP^Sc molecules differed in their microdomain localization on the plasma membrane in N2aC24L1-3 cells. Sortilin was predominantly detected in non-raft fractions while PrP^Sc was exclusively located in raft fractions (Fig. 1B).¹⁶ Therefore, the sortilin-mediated lysosomal degradation of PrP^Sc located on the plasma membrane might be a minor event.

PRP^Sc STIMULATES DEGRADATION OF SORTILIN IN LYSOSOMES

Interestingly, we found that sortilin was markedly reduced in both prion-infected cells and mouse brains, and that the reduced sortilin levels in prion-infected cells were recovered by treatment with lysosomal inhibitors but not with proteasomal inhibitor.¹⁶ These findings suggest that sortilin is increasingly degraded in lysosomes in prion-infected cells. We also found that PrP^Sc accumulation preceded the reduction of sortilin in N2aC24 cells freshly infected with RML prions.¹⁶ Immunofluorescent staining showed that sortilin was barely detectable in PrP^Sc-positive cells but still abundantly observed in PrP^Sc-negative cells.¹⁶ It is thus likely that PrP^Sc produced after prion infection could stimulate sortilin degradation in lysosomes in a cell-autonomous fashion, and that the negative role of sortilin in PrP^Sc accumulation could be impaired in prion-infected cells, therefore PrP^Sc progressively accumulates in prion-infected neurons.

CONCLUSIONS

We presented a novel accumulation mechanism of PrP^Sc through degradation of sortilin. Sortilin could form the host defense mechanism against prions, by functioning to sort PrP^C and PrP^Sc to the late endosomal/lysosomal compartments for degradation (Fig. 1A, B). Conversely, PrP^Sc itself stimulates degradation of sortilin in lysosomes, reducing sortilin levels and impairing its defense function against prions. As a result, PrP^C is increasingly converted to PrP^Sc, and PrP^Sc degradation is delayed, and eventually PrP^Sc progressively accumulates in prion-infected cells (Fig. 1B). Accelerating the sortilin-mediated lysosomal degradation of PrP^C and PrP^Sc might be beneficial for treatment of prion diseases.

DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

The authors declare no competing interests.

ACKNOWLEDGMENTS

We would like to thank Prof. Horiuchi (Hokkaido University) for anti-PrP antibody clone 132 and N2a cells, and Prof. Doh-ura (Tohoku University) for ScN2a cells. We also would like to thank Mitsuru Tomita, Masashi Yano, Junji Chida, Hideyuki Hara and Nandita Rani Das at Tokushima University and Anders Nykjaer at Aarhus University for their contributions.

FUNDING

This work was partly supported by Pilot Research Support Program in Tokushima University, Naito Foundation, JSPS KAKENHI 26460557 and MEXT KAKENHI 17H05702 to KU, and JSPS KAKENHI 26293212, MEXT KAKENHI 15H01560 and 17H05701, and Practical Research Project for Rare/Intractable Diseases of the Japan Agency for Medical Research and Development (AMED) to SS.