Spc2 modulates substrate- and cleavage site-selection in the yeast signal peptidase complex

Abstract

Secretory proteins are critically dependent on the correct processing of their signal sequence by the signal peptidase complex (SPC). This step, which is essential for the proper folding and localization of proteins in eukaryotic cells, is still not fully understood. In eukaryotes, the SPC comprises four evolutionarily conserved membrane subunits (Spc1-3 and Sec11). Here, we investigated the role of Spc2, examining SPC cleavage efficiency on various models and natural signal sequences in yeast cells depleted of or with mutations in Spc2. Our data show that discrimination between substrates and identification of the cleavage site by SPC is compromised when Spc2 is absent or mutated. Molecular dynamics simulation of the yeast SPC AlphaFold2-Multimer model indicates that membrane thinning at the center of SPC is reduced without Spc2, suggesting a molecular explanation for the altered substrate recognition properties of SPC lacking Spc2. These results provide new insights into the molecular mechanisms by which SPC governs protein biogenesis.

Figure 1.

The N-length-dependent signal sequence cleavage profile in the spc2Δ strain. (A) Schematics of N#CPYt(h) constructs. The extended N-terminal sequences are from Dap2, a yeast SA protein (green). N# indicates the number of N-terminally extended residues, t indicates the C-terminal truncation after residue 323 of CPY, (h) denotes hydrophobic version of the CPY signal sequence (Table S1). N-glycan sites are indicated as Y. (B) N16CPYt(h) in the spc3-4 strain was analyzed by pulse labeling at the indicated temperatures, subjected to endoglycosidase H treatment (EH) prior to SDS‒PAGE, and analyzed by autoradiography. (C) N#CPYt(h) constructs in the spc2Δ, spc2Δ + SPC2, and spc2Δ + SPC1 strains were analyzed by pulse labeling. The relative amounts of cleaved products over total products (cleavage [%]) were plotted against the number of n-region residues (N-length). At least three independent experiments were carried out (n = 3/data point), and the average is shown with the standard deviation. P values between WT and spc2Δ and between spc2Δ and spc2Δ+SPC2 strains were calculated by multiple two-tailed t tests; n.s., P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. The cleavage profiles in the WT and spc1Δ strains (Yim et al., 2021) are shown in comparison. (D) The volcano plot of the WT and spc2Δ proteomes as quantified from mass spectrometry (Data S1). The relative abundances of Sec11, Spc3, Spc2, and Sbh2 are indicated in green circles and Pdi1 and Kar2 in red squares. (E and F) SP_Suc2-Lep (Hessa et al., 2009) and ppαF, (F) Ecm38, Kar2 in the spc3-4, WT and spc2Δ strains were analyzed by pulse-labeling. Representative gels from at least three independent experiments are shown in F. Average cleavage efficiencies with standard deviation are indicated. Filled black and red arrows indicate glycosylated full-length and cleaved products, respectively, and unfilled black and red arrows indicate deglycosylated full-length and cleaved products, respectively. Source data are available for this figure: SourceData F1.

Figure 2.

The C-terminal domain of Spc2 is important for N-length-dependent signal sequence cleavage. (A) Structures of human SPCS2 (PDB: 7P2P) and yeast Spc2 (predicted by AlphaFold2, UniProt ID Q04969) are overlaid. (B) Secondary structures of the predicted Spc2. (C) Cleavage efficiencies of N#CPYt(h) variants in spc2Δ cells with SPC2, spc2-ΔCD(58), and spc2-TM2*. Data of N#CPYt(h) variants in the spc2Δ cells in Fig. 1 (C) are shown for comparison. At least three independent experiments were carried out (n = 3/data point), and the average is shown with the standard deviation. P values between spc2Δ+SPC2 and spc2Δ+ spc2-ΔCD(58) strains were calculated by multiple two-tailed t tests; n.s., P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. (D) Cleavage efficiency of N26CPYt(h) in spc2Δ cells with EV, spc2-ΔCD(58), spc2-ΔCD(23) or SPC2 (under the GPD promoter). The expression levels of Spc2 in the indicated strains were assessed by western blotting using anti-FLAG antibodies recognizing Spc2-FLAG. At least three independent experiments were carried out (n = 3/data point), and the average is shown with the standard deviation. P values were calculated by multiple two-tailed t tests; n.s., P > 0.05; ***, P ≤ 0.001. (E) Whole-cell lysates from the spc2Δ,SPC3HA strain carrying an empty vector (EV), SPC2 (under its own promoter), spc2-ΔCD(58), and spc2-TM2* were analyzed by western blotting. PgK is a loading control. Source data are available for this figure: SourceData F2.

Figure S1.

Presents sequences and data related to Spc2, its homologs and mutants. (A) Sequence alignment of Spc2 homologs (human, canine and S. cerevisiae). Predicted TMs are in yellow. Underlined sequences were replaced with the 4L/15A hydrophobic segment in Spc2-TM2* (shown in magenta). Orange colored sequences were truncated in Spc2-ΔCD(58) and orange colored sequences in italic were truncated in Spc2-ΔCD(23). (B and C) The volcano plots comparing the proteomes from spc2Δ cells carrying SPC2 and spc2-ΔCD(58) (B) or SPC2 and spc2-TM2* (C) as quantified from mass spectrometry (Data S2). The relative abundances of Sec11, Spc3, Spc2, and Sbh2 are indicated in green circles, and Pdi1 and Kar2 in red square. (D) Cleavage of SP_Suc2-Lep (Hessa et al., 2009) and ppαF in the spc2Δ, SPC3HA strain carrying an empty vector (EV), SPC2, spc2-ΔCD(58), and spc2-TM2* was analyzed by pulse labeling. Source data are available for this figure: SourceData FS1.

Figure 3.

Effects of Spc2 deletion on the cleavage of TM segments. (A) Schematics and sequences of LepCC variants. The Leu TM segment is in bold, and the cleavage site is indicated as ↓. LepCC(14L(P1)) carries an A to P mutation in the +1 position of the cleavage site, and the cleavage is inhibited. (B) LepCC(14L) in the spc3-4 and WT strains and LepCC(14L[P1]) in the WT strain were analyzed by 5 min pulse labeling, and Endo H (EH) was added to the samples prior to SDS‒PAGE. (C) The indicated LepCC variants in the spc3-4, WT, and spc2Δ strains were analyzed by pulse labeling. Protein samples were treated with or without Endo H (EH) prior to SDS‒PAGE. FL, full-length; C, cleaved species. (D) The relative amounts of cleaved products over total products (cleavage [%]) were plotted (n = 3). (E) Schematic of Pho8 and its SA sequence (in bold) plus 5 downstream residues. The mutated Pro54 residue is indicated in italics. A representative gel is shown. Average cleavage efficiencies from three independent experiments (n = 3/datapoint) and standard deviation are shown. Unfilled black and red arrows indicate de-glycosylated full-length and cleaved products, respectively. ^# indicates a nonspecific band. P values were calculated by two-tailed unpaired t test with Welch’s correction; P > 0.05; *. Source data are available for this figure: SourceData F3.

Figure S2.

Shows the predicted signal sequence cleavage sites of the proteins used in the study. (A) The N-terminal 60 residues of Pho8 and Pho8(P54A) are shown. Signal-anchored sequences are underlined. (B–D) SignalP (Almagro Armenteros et al., 2019) predictions for Pho8 and Pho8(P54A) (B), CPY, N#CPYt(h) CS1/2, CS1 and CS2 (C). Rrt6 (CS1-like), Rrt6_T48Q,L52A,D55P (CS2-like), Rrt6_L52A (CS1/2-like) (D).

Figure 4.

Recognition of signal sequence cleavage sites in the spc2Δ strain. (A) Signal sequences of CPYt and N#CPYt(h) cleavage site (CS) variants. (B) N16CPYt(h) CS1/2, CS1 and CS2 in the spc3-4, WT, and spc2Δ strains were pulse-labeled and subjected to Endo H treatment prior to SDS-PAGE. (C) N16CPYt(h) CS1/2 and CS0 in the spc3-4 and WT strains were assessed by pulse labeling. (D) Cleavage efficiencies of N#CPYt(h) CS2 variants in the spc2Δ strain (green). (E) CS2 variants in the spc2Δ strain (green). Cleavage profiles of CS2 or CS1 variants in the spc1Δ strain (red) (Yim et al., 2021) are shown for comparison. Three independent experiments were carried out (n = 3/data point), and the average is shown with the standard deviation. P values between CS1 variants in the WT and spc2Δ, and between CS2 variants in the WT and spc2Δ were calculated by multiple two-tailed t tests; n.s., P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. (F and G) Signal sequences and the downstream residues of Rrt6 (F) and its CS variants (G). Mutated residues are colored in red, and potential cleavage sites are indicated with an arrow (↓). The indicated Rrt6 CS variants in the spc3-4, WT, and spc2Δ strains were analyzed by pulse labeling. A representative of at least three experiments is shown. De-glycosylated full-length and cleaved products are indicated in unfilled black and red arrows, respectively. Source data are available for this figure: SourceData F4.

Figure S3.

Provides model validation for the MD simulations. (A and B) Comparison of the α-carbon root mean squared fluctuations (RMSFs) between all-atom (AA) and coarse-grained (CG) MD simulations in a POPC bilayer, per subunit. (A) WT SPC and (B) SPC lacking Spc2 (spc2Δ). Sec11 is shown in purple, Spc3 in red, Spc1 in orange, Spc2 in green, the atomistic reference RMSF is colored in blue and error bars are shown in grey. (C) iPTM scores for each model predicted by AlphaFold2-Multimer (V3) for the tetramer and trimer (spc2Δ) SPC. (D) RMSD time-series extracted from all-atom MD simulations of the wild-type and trimeric variants of yeast SPC embedded in a POPC bilayer. The different colors correspond to individual 1 μs long simulation runs and the black line is an average RMSD computed from the five repeats.

Figure 5.

AlphaFold2 predictions and MD simulations of the yeast SPC with and without Spc2. (A) Structural alignment of the AF2-predicted yeast SPC with the cryo-EM human SPC (Liaci et al., 2021). Spc2 in green, Sec11 in purple, Spc3 in red, Spc1 in orange. (B) Representative snapshot of the TM window membrane thickness for SPC with (top, WT) and without Spc2 (bottom, spc2Δ). The phosphate headgroups are colored in salmon (WT) or yellow (spc2Δ), and the lipid tails are represented as a transparent grey surface. (C) Model peptide illustrating that the difference in membrane thickness qualitatively fits with a hypothetical SPC substrate selectivity filter for SAs and SPs. (D) Membrane thickness computed for the ER membrane-embedded SPC using different protein models: a rigid model, elastic network models; a semi-flexible and fully flexible model. (E) Membrane thickness computed for a system composed of the yeast SPC embedded into a POPC membrane at atomistic or Martini 3 resolution.

Figure S4.

Includes MD simulation data on water density profiles and maps. (A) Water density profiles computed using gmx density. The solid line is computed from an all-atom MD simulation with SPC embedded in a POPC bilayer. The faded lines correspond to water densities extracted from five independent 20 μs MD simulations of SPC embedded in a model endoplasmic reticulum membrane. (B) Snapshot illustrating water penetration inside the SPC window without the formation of a water pore. (C) Snapshot showcasing a degree of lipid flip-flop across the ER membrane model.

Figure S5.

Illustrates structural rearrangements of the SPC upon removal of Spc2. (A) Crossing angle computed from the vector defined in the Spc3 TM helix 1, between residues 4 and 23, and the vector defined in the Spc1 TM helix 1, between residues 20 and 43. Single-point cross-angles calculated from the AF2 structures are shown in red vertical lines. (B) Representative snapshots illustrating the SPC WT (left) and spc2Δ (right) structures. The phosphate headgroups are colored in salmon (WT) or yellow (spc2Δ), Spc2 in green, Sec11 in purple, Spc3 in red, Spc1 in orange, and the lipid tails are represented as a transparent grey surface.

Figure 6.

Effects of polar residues in Spc2 TM on membrane thickness. (A) Localization of the polar (green) and charged (blue for negative, red for positive) residues of Spc2 within the TM window allow for deeper-lying phosphate headgroups within the membrane. (B) LepCC(17L) in spc2Δ cells with an empty vector (−), SPC2, spc2-Y79A, S83A were analyzed by pulse labeling for 10 and 30 min. Bottom: Expression of Spc2 (under its endogenous promoter in the CEN plasmid) and spc2-Y79A,S83A (under the GPD promoter in 2 µm plasmid) is shown. (C) Cleavage (%) was quantified and plotted as in Fig. 6 B. Three independent experiments were carried out, and the average is shown with the standard deviation. (D) Membrane thickness in the TM window for yeast SPC without Spc2 (spc2Δ), with Spc2 (WT) and with a double mutated variant of Spc2 (spc2_Y79A, S83A), embedded in a model of the yeast ER membrane, computed from Martini 3 CGMD simulations. Average values across five 20 μs simulations per system for spc2Δ and WT, and four 4 μs simulations for spc2_Y79A, S83A are shown with the standard error of the mean. Source data are available for this figure: SourceData F6.

Figure 7.

Proposed models for Spc2-mediated substrate and cleavage site selection. (A and B) The cytosolic domain of Spc2 prevents access of signal sequences with longer n-regions to the TM-window, (B) membrane thinning induced by Spc2 prevents access to signal sequences with longer h-regions to the TM-window. (C) Membrane thinning induced by Spc2 increases the exposure of proximal cleavage sites to the SPC active site (the blue star indicates a proximal cleavage site rendered inaccessible by the absence of Spc2).