Published On: Tue, Mar 5th, 2019

The signal peptide plus a cluster of positive charges in prion protein dictate chaperone-mediated Sec61 channel gating [RESEARCH ARTICLE]

Depletion of BiP inhibits ER import of prion protein due to the signal peptide

The PrP-derived signal peptide is believed to have a weak Sec61 channel gating capacity compared to the SP of ppl (Rutkowski et al., 2001). Therefore, the PrP precursor relies on auxiliary components of the Sec61 translocation machinery, such as the ER luminal Hsp70-chaperone BiP (Lang et al., 2012; Schäuble et al., 2012). In light of recent novel insights into the rules for engagement of BiP in translocation of small presecretory proteins (Johnson et al., 2013), we aimed to evaluate the determinants for BiP assistance in PrP transport. To address this issue, we made use of three different sets of PrP-related precursor polypeptides (Fig. 1A,C,D; Table S1) (Pfeiffer et al., 2013). They vary in the preceding SP as well as the composition of the mature region. All PrP-related precursor proteins, as well as the control model proteins ppl (SRP-dependent and Sec61-dependent) and Cyt b5-OPG (Sec61-independent), were synthesized in the presence of [35S]methionine and ER membranes and in the simultaneous presence or absence of the N-glycosylation tripeptide inhibitor NYT. For visualization, samples were subjected to SDS-PAGE and phosphorimaging. Accordingly, the comparison of the bands produced under plus or minus NYT conditions allowed the identification of N-glycosylated proteins (Fig. 1E–G). Modification occurred on either endogenous sites (PrP variants) or a C-terminal opsin-derived tag (OPG-tag of tail-anchored proteins).

Fig. 1.

Fig. 1.

Model precursor proteins. (A,C,D) Schematic representation of the constructs used in this study. (A) IDD-α2α3 variants. (C) α2α3-IDD variants. (D) PrP wt variants and control precursor polypeptides. SP, signal peptide; PM, polybasic motif (+); IDD, intrinsically disordered domain; α2α3, α-helical regions 2 and 3; β, beta-sheet; lollipops, N-glycans; TMD, transmembrane domain; GPI, glycosylphosphatidylinositol. (B) Kyte-Doolittle Hydrophobicity Plots of the signal peptides used in this study. Charged amino acid residues are indicated (-/+). Scale: −4.5 to +4.5. (E–G) The indicated PrP variants and control precursor polypeptides were synthesized in reticulocyte lysate in the absence (i.e. presence of buffer) or presence of membranes and the tripeptide NYT (-/+), respectively. (E) IDD-α2α3 variants and ppl. (F) Variants of α2α3-IDD, K4-α2α3-IDD and A4-IDD-α2α3. (G) PrP wt variants, Cyt b5-OPG and Sec61β-OPG. All samples (E–G) were subjected to SDS-PAGE and phosphorimaging. Relevant parts of the phosphorimages are shown. Filled triangle, glycosylated protein; unfilled triangle: precursor polypeptide; star, putatively ubiquitinated precursor polypeptide (Rane et al., 2008); PrP, prion protein; APP, amyloid precursor protein; Som, somatostatin; ppl, preprolactin; Cyt b5, Cytochrome b5; OPG, opsin-derived sequence with N-glycosylation site; wt, wild type. See also Tables S1 and S2 and Fig. S1. For complete phosphorimages, see Figs S4 and S5.

Having established ER transport of our model precursor proteins, we investigated the translocation requirements of the first set of PrP variants (Fig. 1A). Each precursor is equipped with a different ER signal peptide either derived of PrP, amyloid precursor protein (APP) or somatostatin (Som) (Fig. 1B). The SP precedes a minimal unit of the PrP mature region, called IDD-α2α3, composed of the intrinsically disordered domain (IDD) and C-terminal alpha-helices (α2α3). Other domains, such as the TMD or GPI anchor sequence, are lacking in favor of topological homogeneity and full import into the ER (Kim and Hegde, 2002; Miesbauer et al., 2009). The respective set of PrP-related SP-chimera (IDD-α2α3) was subjected to an established protocol for in vitro protein translocation into the ER of semi-permeabilized HeLa cells upon siRNA-mediated gene silencing of BiP (Table S3) (Haßdenteufel et al., 2018). Cells were treated for 48 h with BIP-targeting or control siRNA before digitonin-permeabilization and used as an ER membrane source in rabbit reticulocyte lysate. Precursor polypeptides were synthesized in the presence of [35S]methionine and ER membranes. For visualization, samples were subjected to SDS-PAGE and phosphorimaging. Signal peptide cleavage (ppl) or N-glycosylation (PrP variants) reported about translocation efficiency when quantified in comparison to negative control siRNA treated cells. Silencing efficiency was evaluated by western blot with established antibodies (Haßdenteufel et al., 2018). It was previously established that these depletion conditions lead to 70% BiP depletion, without substantially affecting cell growth, cell viability, ER/cell morphology, and ER protein import components (Schäuble et al., 2012; Haßdenteufel et al., 2018). After 48 h treatment with BiP siRNA, the protein content of BiP was reduced to 30% compared to control cells as expected (Fig. 2B; Fig. S2B). Although siRNA-mediated BiP depletion was rather incomplete, moderate effects on translocation of IDD-α2α3 were observed (Fig. 2A, white panel; Fig. S2A). However, glycosylation efficiency was selectively inhibited in the presence of the PrP- or APP-SP but in the presence of the Som-SP it was not. In addition, ppl transport was not affected (Fig. 2A, blue panel; Fig. S2A).

Fig. 2.

Fig. 2.

Engagement of BiP and Sec63 in ER import of prion protein is differentially determined. For protein depletion, HeLa cells were treated with the corresponding siRNA (Table S3) or subtilase toxin, as indicated. After digitonin-permeabilization of the harvested cells (A–E), reticulocyte lysate was programmed with the indicated precursor polypeptides and incubated in the absence or presence of depleted or control ER membranes (A,D–E). Radioactive samples were subjected to SDS-PAGE and phosphorimaging (Fig. S2A,D–I). Transport efficiencies were calculated as the proportion of N-glycosylation or signal peptide cleavage of the total amount of synthesized precursors with the individual control sample set to 100%. Individual data points and the mean of at least three individual experiments are shown. For statistical analysis (***P0.001, **P0.01, *P0.05), a Student’s t-test (upper row of stars) or ANOVA with the Dunnett’s and Newman–Keuls post hoc test, respectively, were performed (horizontal brackets). (A) BiP siRNA effects on transport efficiency of ppl and IDD-α2α3 variants with various SPs. (D,E, upper dot plots) Subtilase toxin effects on transport efficiency of IDD-α2α3 and α2α3-IDD variants (D) as well as PrP wt variants (E). (D,E, bottom dot plots) SEC63 siRNA effects on transport efficiency of IDD-α2α3 and α2α3-IDD variants (D) as well as PrP wt variants (E). (B,C) Protein content of HeLa cells depleted of BiP (B) or Sec63 (C) relative to β-actin was validated by western blot and the indicated antibodies (Fig. S2B,C). The control sample was set to 100%. Filled dots, weak SPs; unfilled dots, strong SPs and controls; red dots, charge variants; yellow panel, structural variants; blue panel, control precursor polypeptides. For complete phosphorimages, see Figs S6–S10.

Driven by this finding, we changed to an alternative strategy for highly efficient reduction of BiP content by subtilase AB (SubAB) cytotoxin (Paton et al., 2006; Schäuble et al., 2012). Strikingly, 2 h treatment of HeLa cells with SubAB before semi-permeabilization resulted in 97% knockdown of BiP and strengthened translocation defects compared with the siRNA approach (Fig. 2B,D, white panel; Fig. S2B). Transport of the negative control ppl was as efficient in SubAB treated cells as in control cells treated with inactive SubA272B toxin (Fig. 2D, blue panel; Fig. S2D). Insertion efficiency of the model tail-anchored protein Cyt b5-OPG (Sec61-independent) was assayed under posttranslational conditions, i.e. after completion of protein synthesis, demonstrating integrity of the analyzed ER membranes (Fig. 2D, blue panel; Fig. S2F).

Since BiP-dependence of PrP translocation was shown on a PrP-related precursor variant, the precursor of the wild-type (wt) protein was subjected next to the same subtilase approach (Fig. 1D; Table S1). Translocation of PrP wt showed the same perturbation upon BiP cleavage as PrP-IDD-α2α3 (Fig. 2E, white panel; Fig. S2D,F). Here, too, exchange of the signal peptide by that derived of a BiP-independent substrate, such as ppl, led to BiP-independent translocation of the PrP wt mature region (Fig. 2E, white panel; Fig. S2F).

In sum, the presented data argue for signal peptide-specific assistance of protein translocation by BiP. Having the PrP- and APP-SP identified as BiP-dependent and the Som- and ppl-SP as BiP-independent, the question arises: how do they differ and what defines BiP dependence? Consequent computational analysis of our model signal peptides indeed demonstrated differences in the overall hydrophobicity (ΔGpred) and charge load as well as the probability for loop-insertion (N-inpred) (Table S2). Furthermore, the distribution of basic and apolar amino acid residues along the sequences varied (Fig. 1B; Table S1). Both PrP and APP show accumulation of apolar residues at the N-terminus whereas positively charged amino acids are missing. In addition to the high N- in values of Som- and ppl-, we note that the positive charges at the N-terminus and the highly hydrophobic middle part may define the two SPs as ‘strong’ in terms of channel gating. To experimentally address this point, we took advantage of two chimeric signal peptides composed of the N-terminal half of either APP or PrP and the C-terminal half of either PrP (APP/PrP) or Som (PrP/Som) (Fig. 1A,B; Tables S1, S2) (Pfeiffer et al., 2013). Not much of a surprise, the SP-chimera with the two BiP-dependent halves, APP/PrP, showed unchanged requirement for BiP in translocation of IDD-α2α3 (Fig. 2D, white panel; Fig. S2D). Although total hydrophobicity was elevated by their fusion, apolar residues still accumulated at the N-terminus and charged residues were lacking. Interestingly, PrP/Som-IDD-α2α3, the SP-chimera with a BiP-dependent N-terminus and a BiP-independent C-terminus, presented an intermediate phenotype (Fig. 2D, white panel; Fig. S2D). The hydrophobic Som-SP-C-terminus indeed led to a partial rescue; however we speculate that full capacity for BiP-independent translocation may require the basic residue at the N-terminus. Of note, the two complementing SP-chimera APP/Som- and PrP/APP-IDDα2α3 completely lost their capacity for translocation into the ER along with the loss of significant hydrophobicity of their SPs required for recognition by SRP and the Sec61 channel (Fig. S1A–C, Fig. S13) (Nilsson et al., 2015).

About the Author

Leave a comment

XHTML: You can use these html tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>