Sialylation of prion protein controls the rate of prion amplification, the cross-species barrier, the ratio of PrPSc glycoform and prion infectivity

Abstract

The central event underlying prion diseases involves conformational change of the cellular form of the prion protein (PrP(C)) into the disease-associated, transmissible form (PrP(Sc)). Pr(PC) is a sialoglycoprotein that contains two conserved N-glycosylation sites. Among the key parameters that control prion replication identified over the years are amino acid sequence of host PrP(C) and the strain-specific structure of PrPSc. The current work highlights the previously unappreciated role of sialylation of PrP(C) glycans in prion pathogenesis, including its role in controlling prion replication rate, infectivity, cross-species barrier and PrP(Sc) glycoform ratio. The current study demonstrates that undersialylated PrP(C) is selected during prion amplification in Protein Misfolding Cyclic Amplification (PMCAb) at the expense of oversialylated PrP(C). As a result, PMCAb-derived PrP(Sc) was less sialylated than brain-derived PrP(Sc). A decrease in PrPSc sialylation correlated with a drop in infectivity of PMCAb-derived material. Nevertheless, enzymatic de-sialylation of PrP(C) using sialidase was found to increase the rate of PrP(Sc) amplification in PMCAb from 10- to 10,000-fold in a strain-dependent manner. Moreover, de-sialylation of PrP(C) reduced or eliminated a species barrier of for prion amplification in PMCAb. These results suggest that the negative charge of sialic acid controls the energy barrier of homologous and heterologous prion replication. Surprisingly, the sialylation status of PrP(C) was also found to control PrP(Sc) glycoform ratio. A decrease in Pr(PC) sialylation levels resulted in a higher percentage of the diglycosylated glycoform in PrP(Sc). 2D analysis of charge distribution revealed that the sialylation status of brain-derived PrP(C) differed from that of spleen-derived PrP(C). Knocking out lysosomal sialidase Neu1 did not change the sialylation status of brain-derived PrP(C), suggesting that Neu1 is not responsible for desialylation of PrP(C). The current work highlights previously unappreciated role of PrP(C) sialylation in prion diseases and opens multiple new research directions, including development of new therapeutic approaches.

Figure 1. 2D analysis of PrP^C and brain- and PMCAb-derived PrP^Sc.

A. 2D analysis of Syrian hamster NBH (top panel) and NBH treated with sialidase from A. ureafaciens (bottom gel). B, C. 2D analysis of brain- and PMCAb-derived 263K material (B), and brain- and PMCAb-derived SSLOW material (C). To produce PMCAb-derived material, brain-derived 263K or SSLOW was subjected to 24 serial rounds with 10-fold dilution between rounds. Black and white triangles mark diglycosylated and monoglycosylated glycoforms, respectively, whereas arrows mark the unglycosylated form. D. Migration of PrP^C in 1D SDS-PAGE gel before and after treatment with sialidase. All blots were stained with 3F4 antibody. E. Schematic diagram of PrP^C that illustrates location of sialic acid residues (stars) on N-linked glycans and GPI anchor. Each of the two glycans can carry up to four terminal sialic acid residues.

Figure 2. De-sialylation of PrP^C increases the rate of PrP^Sc amplification in PMCAb.

A. 263K, HY, Drowsy or SSLOW scrapie brain materials were diluted 10³–10¹⁰-fold into 10% NBH, dsNBH or mock treated NBH (NBH incubated with a buffer for enzymatic de-sialylation in the absence of sialidase) as indicated and subjected to a single PMCAb round. Undigested 10% NBH was used as reference. B. Analysis of PrP^Sc amplification fold. Scrapie brain materials were diluted 10⁴-fold for 263K or 10³-fold for SSLOW into 10% NBH or dsNBH as indicated and subjected to four serial PMCAb rounds. The material amplified in each round was diluted to a specified dilution fold into 10% NBH or dsNBH for the next PMCAb round as indicated. Unamplified seeds are shown as round 0. Prior to electrophoresis, samples were treated with PK. All blots were stained with 3F4 antibody.

Figure 3. De-sialylation of PrP^C reduces species barrier in PMCAb.

A. 263K or HY brain materials were diluted 10³-fold into 10% mouse dsNBH or NBH and subjected to ten serial PMCAb rounds with 5-fold dilution between rounds. Prior to electrophoresis, samples were treated with PK. Blots were stained with Ab3531 antibody. B. 22L or ME7 brain material was diluted 10³-fold into 10% hamster dsNBH or NBH and subjected to ten serial PMCAb rounds with 5-fold dilution between rounds. Prior to electrophoresis samples were treated with PK. Blots were stained with 3F4 antibody. Black and white triangles mark diglycosylated and monoglycosylated forms, respectively. C. PK-resistance profile illustrating relative representation of di-, mono-, and unglycosylated glycoforms in 263K-seeded PMCAb products produced in three reactions with mouse NBH (bold solid lines) or dsNBH (dashed lines).

Figure 4. Influence of PrP^C sialylation level on glycoform distribution in PMCAb-derived products.

A. 22L or ME7 brain material was diluted 5×10²- or 2×10³-fold, respectively, into 10% mouse dsNBH or NBH and subjected to three serial PMCAb rounds with 5-fold dilution between rounds. Prior to electrophoresis samples were treated with PK. Blot was stained with Ab3531 antibody. Undigested 1% mouse NBH was used as a reference. B. Western blot of 22L or ME7 scrapie brain material (SBH) or material produced in PMCAb using NBH or dsNBH (dsPMCAb), and stained with Ab3531 antibody. C. Western blot of 263K or SSLOW scrapie brain material (SBH) or material produced in PMCAb using NBH or dsNBH (dsPMCAb), and stained with 3F4 antibody. Black and white triangles mark di- and mono-glycosylated glycoforms, respectively, whereas arrows mark the unglycosylated form.

Figure 5. 2D analysis of brain- and spleen-derived PrP^C.

2D analysis of Syrian hamster spleen (A) and brain homogenates (B). Diglycosylated and monoglycosylated full-length PrP^C (FL) are marked by black and white triangles, respectively. Di-, mono- and unglycosylated C2 fragments (C2) in spleen or C1 fragments (C1) in brain are marked by bold, medium and thin arrows, respectively. PrP^C, C1 or C2 dimers (D) are encircled. M stands for a marker lane: brain or spleen samples were diluted 10-fold and used as references for 2D gels. Blots were stained with SAF-84 antibody.

Figure 6. Correlation between PrP^Sc sialylation status and its infectivity.

A. The following material was analyzed by 2D and animal bioassay: 1% brain-derived 263K material, PMCAb-derived 263K (products of 24th PMCAb rounds conducted in NBH with 10-fold dilution between rounds), PMCAb-derived 263K produced in dsNBH (products of 7^th PMCAb round conducted in dsNBH with 1000-fold dilution between rounds) and 263K^R+ (263K brain material subjected to 12 serial PMCAb rounds in RNA-depleted NBH and then an additional 14 PMCAb rounds in NBH as described in [46]). Prior to inoculation, all PMCAb-derived materials were diluted 10-fold. 10⁴-fold diluted 263K brain material was used for inoculating a reference group to match the amount of PK-resistant material in PMCAb-derived samples. Diglycosylated and monoglycosylated PrP^C are marked by black and white triangles, respectively. B. Western blots of 263K- or SSLOW-seeded PMCAb-derived material produced using NBH or dsNBH and treated with increasing concentrations of PK as indicated. All blots were stained with 3F4 antibody.

Figure 7. Schematic diagrams illustrating that prion polymerization and PrP^Sc glycoform ratio are controlled by sialic acid residues of glycans.

A, B. Sialylation of glycans impedes prion amplification in PMCAb due to electrostatic repulsion between sialic acid residues on the PrP^Sc surface. Desialylation of a substrate facilitates prion polymerization in dsPMCAb. Due to strain-specific differences in PrP^Sc structures, PrP^C desialylation speeds up polymerization of strains of synthetic origin (B) much stronger than strains of natural origin (A). C. Electrostatic repulsion between sialic acid residues limits percentage of diglycosylated glycoforms that can be accommodated in PrP^Sc. Desialylation of PrP^C in dsPMCAb changes the ratio of glycoforms in PrP^Sc in favor of diglycosylated glycoforms. PrP^C monomers are depicted as blue circles; glycosyls are shown as yellow circles where presence of terminal sialic acid residues is shown as a red glow. Electrostatic repulsion is represented with lightning bolts.