Spc1 regulates substrate selection for signal peptidase

Signal peptidase (SPase) cleaves the signal sequences (SSs) of secretory precursors. It contains an evolutionarily conserved membrane protein subunit, Spc1 that is dispensable for the catalytic activity of SPase, and its role remains elusive. In the present study, we investigated the function of yeast Spc1. First, we set up an in vivo SPase cleavage assay using secretory protein CPY variants with SSs modified in the n and h regions. When comparing the SS cleavage efficiencies of these variants in cells with or without Spc1, we found that signal-anchored sequences become more susceptible to cleavage by SPase without Spc1. Further, SPase-mediated processing of transmembrane (TM) segments in model membrane proteins was reduced upon overexpression of Spc1. Spc1 was co-immunoprecipitated with membrane proteins carrying uncleaved signal-anchored or TM segments. These results suggest a role of Spc1 in shielding TM segments from SPase action, thereby contributing to accurate substrate selection for SPase.


97
Proper targeting and translocation of CPY to the ER was determined by assessing the 98 glycosylation status of CPY, as it contains three N-linked glycosylation sites, which take place 99 in the ER lumen. All N#CPYt(h) variants were sensitive to treatment with endoglycosidase H 100 (Endo H), which removes N-linked glycans, indicating that they were efficiently translocated 101 into the lumen (Fig. 1B).

103
Two bands were detected for the longer N-length variants (N16CPYt(h) to N26CPYt(h)) even 104 with Endo H treatment, indicating that they are proteins of two different sizes in the ER (Fig.   105 1B). When the size of CPYt(h) and N16CPYt(h) was compared to that of CPYt(wt) (CPYt 106 possessing the original SS), CPYt(h) and the smaller size form of N16CPYt(h) migrated the 107 same as CPYt(wt), SS of which is efficiently cleaved by SPase. Thus, CPYt(h) is fully cleaved 108 and N16CPYt(h) is partially cleaved by SPase (Fig. S1B). We also prepared an N-terminal 109 SS-deleted version of CPYt(h) (mCPYt), which was expressed in vitro, and compared its size

115
To further confirm the SPase-mediated cleavage, selective N#CPYt(h) variants were 116 expressed in the spc3-4 strain, which exhibits a temperature-sensitive defect in SPase activity 117 (Fang et al., 1997). When N16CPYt(h) was radiolabeled at a permissive temperature of 24°C, 118 two forms appeared, whereas a lower band was not observed in cells radiolabeled at the 119 nonpermissive temperature of 37°C, indicating that the lower band resulted from SPase 120 activity (Fig. 1C). CPYt(h) and N9CPYt(h) variants expressed at 37°C in the spc3-4 strain 121 migrated slower than those expressed in the wild-type (WT) strain, and fast-migrated bands of 122 the N16CPYt(h) and N26CPYt(h) variants in the spc3-4 strain were no longer detected when 123 labeled at 37°C (Fig. 1C)

137
and low hydrophobicity (N#CPYt(l), l for low) were prepared and analyzed by 5 min pulse 138 labeling as above (Figs. 1E and S1E). The relative amounts of SS-cleaved species among 139 the glycosylated products were quantified (% cleavage) (Fig. 1F). The SS cleavage profiles of 140 the (N#CPYt(i)) and N#CPYt(l) variants were shifted to the right, and the estimated threshold 141 N-length (50% SS cleavage by applying a trend line on the graphs) increased as the SS 142 became less hydrophobic (~19 for CPYt(h), ~22 for CPYt(i), and >26 for CPYt(l)) ( Fig. 1F).

144
These data show that the N-length and hydrophobicity of SSs are two critical determinants 145 based on which SPase distinguishes substrates (cleavable SSs, SPs) from nonsubstrates 146 (uncleavable SSs, TMs).

148
Internal SSs are more efficiently cleaved by SPase lacking Spc1

149
The eukaryotic SPase has multiple subunits and the functions of each subunit remain poorly 150 defined. We set out to investigate the role of Spc1, a small membrane protein subunit (Fang 151 et al., 1996;Kalies and Hartmann, 1996). Spc1 spans the ER membrane twice, with both 152 termini facing the cytoplasm with a very short loop in the lumen ( Fig. 2A).

154
First, we prepared a spc1Δ strain and assessed its growth phenotype. No growth defect was 155 observed at any tested temperature, as seen in an earlier study (Fig. 2B) (Fang et al., 1996).

156
To check if the deletion of Spc1 affects the stability of the other subunits in SPase, we carried 157 out mass spectrometry analysis to compare the abundance of Sec11, Spc3 and Spc2 in WT 158 and spc1Δ cells. Although the abundance of the nonessential subunit Spc2 was slightly 159 reduced in the spc1Δ strain, the abundance of Sec11 and Spc3, which are the catalytic 160 components of SPase, was unchanged (Fig. 2C)

174
Less hydrophobic N#CPYt(i) and N#CPYt(l) sets showed similar cleavage patterns: cleavage 175 efficiency increased for the longer N-length variants when Spc1 was absent and restored 176 when SPC1 was re-expressed in the spc1Δ strain (Figs. S2A and B). Thus, these data show 177 that SPase lacking Spc1 becomes more prone to cleave membrane-anchored, internal SSs.

179
Processing of Sps2

180
We wondered whether Spc1-regulated processing of internal SSs also occurs in natural 181 proteins and searched for yeast endogenous proteins possessing an internal SS. We found 182 Sps2, a protein involved in sporulation and localized to the plasma membrane and cell wall in

9
However, untargeted proteins accumulated at 37°C, thus we assessed the cleavage at 33°C, 193 a semipermissive temperature without compromising ER targeting and confirmed that the 194 cleaved product of Sps2t(L) is generated by SPase (Fig. S2C). Sps2t(L) was expressed in 195 WT and spc1Δ strains, and when the cleaved product was assessed upon Endo H digestion, 196 only the SS cleaved form was detected in spc1Δ cells whereas the full-length form was 197 detected in WT cells (Fig. 2F, lanes 7-8). Sps2t(L) was also expressed in the spc1Δ strain 198 with an EV or SPC1, and the data confirmed that the full-length protein was more readily 199 cleaved in the absence of Spc1 (Fig. 2F, lanes 10 and 12). These results show that cleavage

214
To identify which cleavage site is used, cleavage site 1 or 2 was selectively eliminated one at

241
To determine if SPase lacking Spc1 uses different cleavage sites for processing, we analyzed 242 SS processing of the cleavage site variants in spc1Δ cells. Cleavage of N20CPYt(h) and 243 N26CPYt(h) variants with Q-3 or P+1' mutations in spc1Δ cells increased compared to that in 244 WT cells (Fig. 3D). These results indicate that recognition and usage of cleavage sites for 245 SPase without Spc1 are unchanged. Next, we set out to determine whether SPase lacking 246 Spc1 uses a noncanonical SS cleavage site, thereby evading the -3, -1 rule for processing.

11
As in WT cells, no cleavage was detected for these sets of variants, indicating that SPase still 250 processes the canonical SS cleavage sites, even in the absence of Spc1 (Fig. 3E).

252
SPase-mediated cleavage of TM segments in membrane proteins is enhanced in the absence 253 of Spc1

254
Observing that SPase lacking Spc1 includes internal, membrane-anchored SSs for 255 processing, we reasoned that TM segments of membrane proteins may also be subjected to     How does overexpressed Spc1 protects TM segments from SPase-mediated processing? We 339 suspected that Spc1 might physically shield TM segments from being presented to the SPase 340 active site. If so, overexpressed Spc1 might interact with signal-anchored or membrane 341 proteins, and we carried out co-immunoprecipitation to test the idea (Fig.6). The C-terminally 342 FLAG-tagged Spc1 (Spc1FLAG) was overexpressed in the spc1Δ strain together with HA-343 tagged CPYt and LepCC model proteins. We assessed whether N9CPYt(h) Q-3/P+1' that is

351
These data show that Spc1 associates with membrane proteins having uncleaved TM 352 segments, suggesting that Spc1 interacts with TM segments of membrane proteins and shield 353 them from being processed by SPase.

356
The sequence context of SPs that are cleaved by SPase and signal-anchored segments that 357 become TM domains are similar in their hydrophobicity and overall length. Hence, both can 358 be recognized by the signal recognition particle (SRP), act as an ER targeting signal and 359 initiate protein translocation in the ER membrane. In a subsequent step, an SP is cleaved, 360 whereas a TM segment evades processing by SPase and anchors in the membrane.

···L L D K L L L T L L L C L L L L S T T L A K A I S L···
Cleavage site 1 N16CPYt