Disturbed vesicular trafficking of membrane proteins in prion disease

Abstract

The pathogenic mechanism of prion diseases remains unknown. We recently reported that prion infection disturbs post-Golgi trafficking of certain types of membrane proteins to the cell surface, resulting in reduced surface expression of membrane proteins and abrogating the signal from the proteins. The surface expression of the membrane proteins was reduced in the brains of mice inoculated with prions, well before abnormal symptoms became evident. Prions or pathogenic prion proteins were mainly detected in endosomal compartments, being particularly abundant in recycling endosomes. Some newly synthesized membrane proteins are delivered to the surface from the Golgi apparatus through recycling endosomes, and some endocytosed membrane proteins are delivered back to the surface through recycling endosomes. These results suggest that prions might cause neuronal dysfunctions and cell loss by disturbing post-Golgi trafficking of membrane proteins via accumulation in recycling endosomes. Interestingly, it was recently shown that delivery of a calcium channel protein to the cell surface was impaired and its function was abrogated in a mouse model of hereditary prion disease. Taken together, these results suggest that impaired delivery of membrane proteins to the cell surface is a common pathogenic event in acquired and hereditary prion diseases.

Keywords: :

• Prion
• prion protein
• post-Golgi membrane trafficking
• insulin receptor
• prion disease
• neurodegeneration
• attractin
• Golgi apparatus

This article refers to:

Introduction

The normal cellular prion protein, designated PrP^C, is a glycosylphosphatidylinositol (GPI)-anchored cell surface glycoprotein most abundantly expressed by neurons and to a lesser extent in other tissues such as lymph nodes and spleen.¹ Conformational conversion of PrP^C into the abnormally folded, amyloidogenic isoform, PrP^Sc, is a central event in the pathogenesis of prion diseases, which include Creutzfeldt-Jacob disease (CJD) in humans and scrapie and bovine spongiform encephalopathy in animals.¹ Indeed, mice devoid of PrP^C (Prnp^0/0) are resistant to prions, neither producing PrP^Sc nor developing prion disease even after intracerebral inoculation with the prions.^2-5 However, the pathogenic mechanism underlying neuronal cell death in prion diseases remains largely unknown.

Cell surface expression levels of membrane proteins are tightly controlled via complicated vesicular transport systems. Newly synthesized membrane proteins are delivered from the endoplasmic reticulum (ER) to the Golgi apparatus, where sugar chains attach to the proteins, and then to the cell surface directly or indirectly via recycling endosome compartments.^6,7 The membrane proteins expressed on the surface are endocytosed to sorting endosome compartments, with some of them being delivered back to the cell surface directly or indirectly via recycling endosome compartments and others are transported to lysosomes for degradation.^6,7 Damage to the vesicular transport of membrane proteins would reduce surface expression levels of the membrane molecules, leading to functional deficiency of the molecules eventually causing damage to the cells.

We recently reported that prion infection disturbs post-Golgi vesicular trafficking and hinders membrane proteins from being delivered to the cell surface from the Golgi apparatus.⁸ These findings suggest that disturbance of post-Golgi vesicular trafficking is a pathogenic mechanism in prion diseases.

Prion Infection Impairs post-Golgi Vesicular Trafficking

A vesicular transport assay using a temperature-sensitive mutant of the vesicular stomatitis virus-G protein fused with green fluorescent protein, designated VSV-G(ts045)-GFP, is very useful to assess vesicular trafficking from the ER to the cell surface through the Golgi apparatus in cells.⁹ VSV-G(ts045)-GFP folds improperly at non-permissive temperature, thereby remaining in the ER. In contrast, at permissive temperature, this protein is properly folded and exits from the ER to the Golgi apparatus and then to the cell surface. VSV-G(ts045)-GFP was normally transported from the ER to the Golgi apparatus in 22L prion-infected N2aC24L1–3 cells at the permissive temperature, compared with uninfected N2aC24 cells (Fig. 1).⁸ However, the export of VSV-G(ts045)-GFP from the Golgi apparatus was significantly delayed in infected cells, and this delayed export of VSV-G(ts045)-GFP was recovered to a normal level in cured N2aC24L1–3 cells, in which prions had been completely eliminated by treatment with SAF32 anti-PrP antibody (Fig. 1).⁸ Similar delayed export of VSV-G(ts045)-GFP from the Golgi apparatus was observed in Chandler prion-infected N2aC24Chm cells.⁸ This delayed export was also recovered to a normal level in cured N2aC24Chm cells.⁸ These results indicate that prion infection disturbs post-Golgi vesicular trafficking, but does not affect vesicular trafficking from the ER to the Golgi apparatus.

Figure 1. VSV-G(ts045)-GFP transport assay in uninfected N2aC24, infected N2aC24L1–3, and cured N2aC24L1–3 cells. Cells were transfected with the vector encoding VSV-G(ts045)-GFP and incubated at non-permissive temperature. Fluorescent intensities for VSV-G(ts045)-GFP at the Golgi region against those in the whole cell were determined in the randomly selected transfected cells (n = 14–16) at various times after the cells were transferred to permissive temperature. The Golgi area was immunohistochemically delineated using the Golgi marker GM130. Trafficking of VSV-G(ts045)-GFP in the cells was assessed as the accumulation kinetics of the protein in the Golgi area. Post-Golgi trafficking of VSV-G(ts045) was significantly delayed in infected N2aC24L1–3 cells, compared with that in uninfected N2aC24 cells. No significant difference in the VSV-G(ts045)-GFP trafficking was observed between uninfected N2aC24 and cured N2aC24L1–3 cells. Data were analyzed using the Student t test. * P < 0.05, ** P < 0.01. These were modified from Uchiyama et al.⁸

Prion Infection Selectively Impairs Trafficking of Membrane Proteins to the Cell Surface

To investigate the consequences of the disturbed post-Golgi vesicular trafficking, surface membrane proteins in uninfected N2aC24, infected N2aC24L1–3, and cured N2aC24L1–3 cells were labeled with biotins and subjected to western blotting after purification of the biotinylated proteins. Three membrane proteins we examined, including PrP, the insulin receptor α subunit (IRα) and attractin, whose mutation causes prion disease-like spongiform neurodegeneration in animals,¹⁰ were significantly less biotinylated in infected cells than in uninfected and cured cells,⁸ indicating that surface expression of these membrane proteins is lower in infected cells than in uninfected and cured cells. Less than 1% of the biotinylated PrP from infected cells were PK-resistant,⁸ indicating that most of the biotinylated PrP in infected cells are PrP^C. Total proteins (unbiotinylated and biotinylated) composed of IRα and attractin were the same in these cells although, due to accumulation of PrP^Sc, total PrP was higher in infected cells than in uninfected and cured cells.⁸ Consistent with these results, the cell surface staining of PrP^C and IRα was reduced in infected cells.⁸ Instead, these molecules were aberrantly accumulated at a paranuclear region, most parts of which were co-stained with the Golgi markers TGN38 and giantin.⁸ Uninfected and cured cells exhibited staining for PrP^C and IRα predominantly on the cell surface. PrP^C, IRα and another GPI-anchored protein, Thy-1, were also less biotinylated in brain slices from RML prion- and 22L prion-infected, terminally ill mice than in control uninfected brain slices (Fig. 2).⁸ However, the glutamate receptor subunits GluR3 and NR1 were similarly biotinylated in infected and uninfected brain slices (Fig. 2).⁸ PrP^C and Thy-1 are GPI-anchored molecules, IR and attractin are single-pass transmembrane molecules, and GluR3 and NR1 are multipass transmembrane molecules. These results indicate that prion infection might impair post-Golgi trafficking of certain types of membrane proteins.

Figure 2. Western blotting of surface and total membrane proteins in the brains of uninfected and infected mice. Brains were freshly removed from uninfected control and terminally ill mice and coronally sliced. The brain slices were then subjected to biotin labeling of membrane proteins. The biotin-labeled surface proteins were purified using avidin-beads and subjected to western blotting for detection of IRα, PrP, Thy-1, GluR3, and NR1. Total expression levels of each protein were also determined. Biotinylated levels of IRα, PrP, and Thy-1, but not GluR3 and NR1, were lower in infected brains than in uninfected brains. Total expression levels of each protein were the same except for PrP. Total PrP was increased in infected brains due to accumulation of PrP^Sc. β-actin is an internal control. These were modified from Uchiyama et al.⁸

Trafficking of Membrane Proteins Is Commonly Impaired in Prion Diseases

Prion diseases in humans manifest as sporadic, hereditary, and acquired types. Hereditary prion diseases including familial CJD and fatal familial insomnia (FFI) are causatively associated with specific mutations of the PrP gene. Interestingly, reduced surface expression was also shown for GPI-anchored GFP in N2a cells expressing the FFI-linked PrP mutant, D177N/Met-128 PrP,¹¹ and for the voltage-gated calcium channel α₂δ-1 subunit in the primary cerebellar granule neurons from transgenic mice expressing familial CJD-associated PrP mutants, PG14 PrP and D177N/Val-128.¹² Taken together with our results obtained from prion-infected cells and mice,⁸ trafficking of membrane proteins to the cell surface could be commonly impaired in acquired and hereditary prion diseases.

Disturbed Trafficking of Membrane Proteins Is an Early Event in Prion Diseases

To investigate the role of the disturbed post-Golgi vesicular trafficking in the pathogenesis of prion disease, PrP^C and IRα were biotin-labeled in brain slices from mice sacrificed at 2, 62 to 68, and 80 to 87 days post-inoculation (dpi) with RML prions and at the terminal stage. The infected mice became ill at around 112 dpi. PrP^C and IRα were biotinylated at the same level in uninfected and infected brains at 2 dpi.⁸ However, at 80–87 dpi, they were significantly less biotinylated in infected brains than in uninfected brains.⁸ They were also less biotinylated in infected brains at 62–68 dpi.⁸ PrP^Sc was detectable in the brains from 68 dpi and in an increased amount at 87 dpi, reaching its maximal level at the terminal stage.⁸ These results suggest that post-Golgi trafficking is disturbed in brains from the early stage of the infection soon after accumulation of PrP^Sc.

Early impairment of post-Golgi trafficking was also suggested in infected mice transgenic for the GPI-anchored membrane protein PrP-EGFP, a fusion protein of PrP with EGFP.¹³ Tg mice expressing PrP-EGFP together with wild-type PrP^C developed the disease after inoculation with RML prions, with accumulation of PrP-EGFP in the Golgi apparatus in neurons from the early stages after infection.¹³ It was also reported that impaired delivery of the voltage-gated calcium channel α₂δ-1 subunit to the cell surface in Tg(PG14) mice is an early pathogenic event, by demonstrating that glutamate release and calcium influx evoked by K⁺-induced depolarization in the cerebellar synaptosomes was impaired well before clinical symptoms became evident in Tg(PG14) mice.¹² These results suggest that impairment of trafficking of membrane proteins to the surface is a common early pathogenic event in acquired and hereditary prion diseases.

Possible Mechanisms for Impaired Trafficking of Membrane Proteins in Prion Diseases

Subcellular localization of PrP^Sc was determined in infected N2aC24L1–3 and N2aC24Chm cells using anti-PrP monoclonal Ab mAb132, which is able to specifically visualize PrP^Sc under partially denatured conditions by binding to residues 119–127 of mouse PrP. PrP^Sc was abundantly detected at a paranuclear region, which was preferentially costained with the recycling endosome marker transferrin receptor and to a lesser extent with other endosome markers, including another recycling endosome marker Rab11, the early endosome marker EEA1, and the late endosome marker Rab9.⁸ Only a small proportion of PrP^Sc were detected in the Golgi apparatus.⁸ Aberrant accumulation of PrP^Sc in recycling endosomes has also been reported in other infected culture cells and in neurons of RML prion-infected mice.^14,15 These results indicate that PrP^Sc accumulates throughout endosomal compartments, particularly abundantly in the recycling endosome compartment, but little in the Golgi apparatus. Newly synthesized membrane proteins are delivered from the ER to the Golgi apparatus and then to the cell surface directly or indirectly via recycling endosome compartments.^6,7 Moreover, the membrane proteins on the surface are endocytosed to sorting endosome compartments and some of them are delivered back to the cell surface directly or indirectly via recycling endosome compartments.^6,7 It is thus possible that PrP^Sc abnormally accumulated in recycling endosome might affect transport of membrane proteins from the Golgi apparatus to the recycling endosome and then to the plasma membrane. However, D177N/Met-128 mutant PrP was mainly accumulated in the ER and Golgi apparatus of transfected N2a cells and upregulated the Rab GDP dissociation inhibitor α but in contrast, downregulated Rab11, leading to impairment of vesicular trafficking of the GPI-anchored GFP to the cell surface.¹¹ PG14 PrP was also retained in the ER, causing abnormal intracellular accumulation of the voltage-gated calcium channel α₂δ-1 subunit via interaction with the subunit.¹² These results suggest that trafficking of membrane proteins to the surface is disturbed by different mechanisms in acquired and hereditary prion diseases.

Implications for the Pathogenesis of Prion Disease

We here showed that impaired vesicular trafficking of membrane proteins to the cell surface could be a common early pathogenic mechanism in prion disease. The impaired trafficking of membrane proteins caused reduced surface expression of the proteins in vitro and in vivo, suggesting that these proteins are functionally deficient in infected cells. Indeed, the downstream signal of IR was weakened in infected cells after stimulation with insulin.⁸ Moreover, it was shown that glutamate release and calcium influx evoked by K⁺-induced depolarization were impaired in the cerebellar synaptosomes of Tg(PG14) mice due to the decreased surface expression of the voltage-gated calcium channel α₂δ-1 subunit.¹² It is thus conceivable that prions impair trafficking of membrane proteins to the cell surface and consequently reduce surface expression of membrane proteins, thereby causing functional deficiency of the proteins leading to neuronal dysfunctions and cell death. On the other hand, the disturbed post-Golgi vesicular trafficking caused the aberrant accumulation of membrane proteins in the Golgi apparatus in infected cells.⁸ It has been reported that abnormal accumulation of membrane or secretory proteins in the Golgi apparatus could induce ER stress.¹⁶ It is therefore possible that prion infection might induce ER stress by disturbing post-Golgi vesicular trafficking and consequently causing aberrant accumulation of membrane proteins in the Golgi apparatus. Indeed, elevated ER stress has been reported in prion-infected cells and brains.^17,18 Other mechanisms have been proposed to explain cell death in prion diseases. Full-length PrP^Sc and its fragments were shown to be toxic to primary neurons.^17,19 Transmembrane and cytosolic PrPs, termed ^CtmPrP and ^CytPrP, respectively, were also suggested to be neurotoxic molecules in prion diseases.^20,21 Moreover, impairment of the ubiquitin-proteasome system was demonstrated in prion-infected ScN2a cells and mice.²² Further studies are required to understand the neurotoxic mechanism leading to neuronal cell death in prion diseases.

Similar abnormal phenotypes to those often observed in human prion diseases have been reported in Prnp^0/0 mice, such as demyelination in the spinal cord and peripheral nerves,²³ alteration of sleep patterns and circadian rhythms,²⁴ and defective memory and learning.^25,26 However, spongiform neurodegeneration and neuronal cell loss have not been observed in Prnp^0/0 mice.²⁷ It is therefore possible that the reduced surface expression of PrP^C that results in functional deficiency of PrP^C might be responsible for some aspects of the pathogenesis in prion diseases, such as those observed in Prnp^0/0 mice, but not in spongiform degeneration and neuronal cell death. Prion disease-like spongiform neurodegeneration is found in animals lacking functional attractin,¹⁰ suggesting that functional deficiency of attractin that is caused by its reduced surface expression might be involved in spongiform degeneration in prion diseases. However, it was recently reported that mice deficient for attractin developed prion disease with the same incubation times as control wild-type mice after inoculation with prions.²⁸ Reduced insulin signaling is suggested to be involved in the pathogenesis of neurodegenerative disorders, such as Alzheimer disease, by reducing its neuroprotective function and dysregulating synaptic plasticity and memory formation.^29,30 It remains to be elucidated whether or not insulin signal could be also involved in the pathogenesis of prion disease.

We here discussed that prions could disturb trafficking of certain types of membrane proteins to the cell surface, and that the disturbed trafficking could eventually cause neuronal dysfunctions associated with the pathogenesis of prion disease. Thus, it might be worth elucidating the molecular mechanism underlying the impaired trafficking of membrane proteins to the surface in prion-infected cells for further understanding of the pathogenesis of prion diseases and development of treatments for prion diseases.

Abbreviations:
PrPC	=	normal cellular prion protein
PrPSc	=	abnormally folded, amyloidgenic prion protein
Prnp0/0 mice	=	mice devoid of PrPC
GPI	=	glycosylphosphatidylinositol
CJD	=	Creutzfeldt-Jacob disease
FFI	=	fatal familial insomnia
ER	=	endoplasmic reticulum
VSV-G(ts045)-GFP	=	temperature-sensitive mutant of the vesicular stomatitis virus-G protein fused with green fluorescent protein
IRα	=	insulin receptor α subunit
PK	=	proteinase K

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

This study was partly supported by a Grant-in-Aid from the BSE and other Prion Disease Control Project of the Ministry of Agriculture, Forestry and Fisheries of Japan, Grants-in-Aid from the Research Committee of Prion Disease and Slow Virus infection, the Ministry of Health, Labour and Welfare of Japan, and a Grant-in-Aid for TSE research (H23-Shokuhin-Ippan-005) and Research on Measures for Intractable Diseases from the Ministry of Health, Labour and Welfare of Japan. H.M. was partly supported by a Cooperative Research Grant of the Institute for Enzyme Research, The University of Tokushima.

10.4161/pri.27381