Signal Peptide Interactions during ER Translocation

Signal Peptide Interactions during ER Translocation: Comparison

Please note this is a comparison between Version 1 by Friedrich Förster and Version 2 by Lindsay Dong.

Cleavable endoplasmic reticulum (ER) signal peptides (SPs) and other non-cleavable signal sequences target roughly a quarter of the human proteome to the ER. These short peptides, mostly located at the N-termini of proteins, are highly diverse. For most proteins targeted to the ER, it is the interactions between the signal sequences and the various ER targeting and translocation machineries such as the signal recognition particle (SRP), the protein-conducting channel Sec61, and the signal peptidase complex (SPC) that determine the proteins’ target location and provide translocation fidelity.

signal peptide
signal peptidase
ER translocon
endoplasmic reticulum

1. Introduction

The secretory pathway is a protein trafficking highway utilized by more than a quarter of the human proteome ^[1][2][1,2]. Soluble secreted proteins such as antibodies and protein hormones rely on this pathway. The pathway also delivers transmembrane proteins (TMPs) to the endoplasmic reticulum (ER), its downstream organelles such as the Golgi apparatus, and the plasma membrane.

All secretory proteins are translated by cytosolic ribosomes and must be first targeted to and then transported across (or inserted into) the ER membrane at the early stage of their life, either co- or post-translationally ^[3][4][3,4]. A complex network of cytosolic and ER membrane-resident macromolecules facilitate and assist the ER targeting and translocation.

Both ER targeting and translocation/insertion critically depend on so-called signal sequences (SSs), short hydrophobic peptide stretches in the amino acid sequence of the newly synthesized proteins that are recognized by the secretory machinery as trafficking signals. While SSs may appear exceedingly simple, they possess a remarkably versatile and complex physiology. Besides the choice of trafficking routes, SSs carry information about translocation efficiency, occurrence and timing of cleavage, and post-targeting functions.

There are four main classes of SSs: (i) cleavable signal peptides (SPs), found on secreted proteins such as insulin and type I membrane proteins such as HLA molecules; (ii) type II signal anchor sequences (SASs), found on single- and multi-pass transmembrane proteins (TMPs) such as the membrane-bound form of tumor necrosis factor (TNF); (iii) type III SASs found, e.g., on Sec61β; and (iv) tail anchors (TAs) found on proteins such as Sec61γ (Figure 1a). Signal peptides (SPs) are by far the most populous class of SSs. In humans alone, there are an estimated >3000 different SP-containing proteins, constituting >10% of the whole proteome. SPs are usually localized within the first 30 amino acids of the coding sequence but can in some cases also be found more internally. The defining trait of SPs is the capacity to be cleaved by the aptly named signal peptidase complex (SPC).

The primary sequence of SPs is only loosely defined. In fact, approximately 20% of randomized sequences can promote protein secretion in yeast ^[5][6]. SPs are characterized by a tripartite structure (Figure 1b–d): (i) an often positively charged, N-terminal ‘n-region’ that faces the cytosol; (ii) a short hydrophobic core—most commonly between 7 and 15, but not longer than 18–20 amino acids called ‘h-region’; and (iii) a polar luminal C-terminal ‘c-region’ that contains the scissile bond and must be occupied by short, hydrophobic residues at positions −1 and −3 relative to the cleavage site ^[6][7][7,8]. Initially, SPs are inserted into the ER membrane with the N-terminus facing towards the cytosol (N_in) and the mature sequence facing the organellar lumen (C_out). In the case of type I TMPs, the removal of the SP leads to an ‘inverted’ topology of the mature sequence in which the N-terminus is facing ‘outside’ (N_out), while the C-terminus is facing the cytosol (C_in) (Figure 1a).

Figure 1. Types of signal sequences. (a) Depiction of the four types of SSs with their membrane topology indicated. Hydrophobic segments are depicted in magenta. (b) Signal peptides have a tripartite structure, consisting of an n- (cyan), h- (magenta), and c-region (yellow) and are cleaved in the ER lumen by the SPC (green flash). (c) Frequency of residue types relative to the cleavage site ^[8]. (d) Relative length of the respective regions (colored as in (b)) as a function of total SP length. The bulk of the length variation stems from the n-region. Panels c–d were adapted from ^[8].

2. SP Recognition in the Cytosol and ER Targeting

As the first step of their life cycle, secreted proteins and TMPs must be targeted to the ER membrane (Figure 2). The timing of translation, folding, and ER transport is of particular importance: on one hand, nascent chains (NCs) can only cross the ER membrane in an unfolded state, while on the other hand, this prerequisite exposes their hydrophobic segments and makes them prone to aggregation and proteolysis. Therefore, these proteins must be shielded from the hydrophilic cytosol, which is achieved by one of two separate strategies: (i) direct recognition of the nascent protein at the ribosomal exit tunnel by the SRP, leading to co-translational translocation/insertion through the recruitment of the ribosome–nascent chain complex (RNC) to the ER; or (ii) post-translational transport to the ER, which requires the involvement of chaperones to protect the clients from the aqueous environment.

Figure 2. ER delivery pathways for SP-containing proteins. The upper panel shows a schematic of each delivery pathway. The central component of each pathway is underscored. Left: SRP (blue/brown subunits) recognizes SPs emerging from the ribosomal exit tunnel and shields them through the SRP54 M-domain. SP binding triggers the heterodimerization of the SRP54 NG domain (blue) with that of SRα (green), guiding the RNC to the ER. A large conformational rearrangement partially exposes the SP for handover to Sec61 ^[9]. Middle: Cytosolic calmodulin or Hsp40-assisted Hsc70 recognize SP-containing proteins and guide them to the ER. Right: The recently discovered Snd pathway likely consists of a cytosolic component, Snd1, which might act as a chaperone, and two ER membrane-resident components, Snd2/3, which might facilitate handover to Sec61 in some unknown way ^[10][11]. Lower panels show specifics of each pathway.

2.1. Life in the Fast Lane: Co-Translational Recognition and ER Targeting by the SRP

Co-translational targeting of SP-containing proteins is accomplished by the signal recognition particle (SRP) and its cognate ER-associated SRP receptor (SR). The SRP is one of the earliest factors that scan the SPs of NCs when they emerge from the ribosomal exit tunnel. The SRP competes for NC binding with the nascent chain associating complex (NAC) ^[12][17]. Thus, whether an SP can interact efficiently with SRP is one of the main determinants of their trafficking route.

Robust targeting by the SRP pathway is estimated to require approximately 5–7 s ^[13][34]. When SRP binds to an SP at the ribosomal exit tunnel, its main functions are (i) to halt the translation to prevent cytosolic aggregation, (ii) to recruit RNCs to the ER surface via the SRP receptor (SR), and (iii) to facilitate the handover of the NC to the Sec61 translocon (Figure 2).

The SRP, a composite of one RNA and six protein molecules, binds ribosomes in a ‘scanning’ mode even before SPs emerge from the exit tunnel ^{[14][15][16][17][18]}[29,35,36,37,38]. In some cases, ribosome–mRNA complexes have been found to recruit the SRP before the SP has even been translated ^[14][29], or from within the exit tunnel through an allosteric interaction if the N-terminal SS is particularly hydrophobic, which is mostly the case for SASs ^[19][20][39,40].

One of the SRP proteins, SRP54, contains the hydrophobic SS binding groove in its C-terminal M domain. In the scanning mode, the M domain is pre-positioned in direct proximity to the exit tunnel in an auto-inhibited state ^[21][41], in which its C-terminal helix ahM6 ^[9][14] occupies the SP binding site. As soon as an appropriate SP emerges from the ribosome exit tunnel, it is embraced by the SRP54 M domain in the ‘cargo recognition’ state, providing shelter from the aqueous environment. The mammalian SRP54–SS interface in the ‘cargo recognition’ state has been resolved in a recent structure of SRPs in complex with ribosomes translating a type II SAS with a 16 AA hydrophobic core ^[9][14] (Figure 3a–e). The ahM6 helix is displaced by the SS such that it completes a ring of α-helices around the SS helix, which remains accessible at both ends. The bulk of the interaction is mediated through non-directional, hydrophobic protein–protein contacts (mostly through flexible methionine side chains) of the SS core with the SRP54 M domain. The hydrophobic surface of the SRP54 M-domain is about 20–23 Å long, matching hydrophobic sequences of ~16 AA (Figure 3e). It appears plausible that SPs with much shorter h-regions ^[8][22][5,27] produce a partial hydrophobic mismatch and are less efficient in both recognition and the induction of the necessary conformational rearrangements in SRP. Indeed, deletions in the h-region of the (already short) SP of preprolactin have drastic effects in its ability to trigger both SRP binding and ER translocation ^[23][42]. Coincidentally, these shorter hydrophobic sequences also reduce the susceptibility to aggregation in the cytosol, rationalizing their reduced dependency on the pathway ^[24][21].

Figure 3. Details of cytosolic SP recognition and ER targeting. (a–e) The SRP (orange) and its central component SRP54 (teal/cyan) recognize SPs at the ribosomal exit tunnel (a,b) and bury the hydrophobic core in cavity formed by the SRP54 M-domain (c,d). αhM5/6 are colored blue. (e) Dimensions of the hydrophobic SRP54–SS interface. Figures based on PDB 7OBR ^[9]. (f,g) Cargo recognition by Hsp70 and Hsp40 binding, exemplified by bacterial DnaK complexed with the Hsp40 DnaJ in an open conformation [PDB 5NRO] and substrate-engaged high-affinity conformation [PDB 2KHO]. ATP hydrolysis, stimulated by substrate and Hsp40 binding, leads to a high affinity state with a closed lid domain (yellow). NBD: nucleotide binding domain; SBD: substrate binding domain. (h,i) AlphaFold 2 predictions of the ER-resident membrane proteins Snd2/3 from S. cerevisiae (entries Q99382 and P38264). TM helices are numbered Membrane topologies are based on UniProt. Both predicted structured possess hydrophilic vestibules that reach into the presumed membrane environment. Helices are numbered.

2.2. With a Little Help from My Friends: Post-Translational Recognition and ER Targeting

2.2.1. Stairway to Lumen: Hsp70/40-Mediated Chaperoning in the Cytosol and ER Delivery

The Hsp70/40 system is thought to deliver its substrates mainly to Sec61/62/63. Post-translational substrates are usually poor substrates of the SRP, and their structural motives are therefore reciprocal to those of SRP substrates—the less hydrophobic an SS, and the closer it is to the C-terminus, the more likely it is to take a chaperone-mediated post-translational delivery route to the ER ^[25][26][24,62]. Hence, most substrates of the pathway feature SPs, although some type II TMPs and multi-spanning TMPs are also influenced by the depletion of both Sec62 and Sec63. Prominent substrates are weakly hydrophobic, N-terminal SPs with weak statistical trends for long h-regions and low-polarity c-regions ^[26][27][62,63]. The post-translational pathway can process more proteins per time than the SRP-dependent route ^[28][12]. Hsp70-mediated ER targeting has so far mostly been studied in yeast. Cytosolic Hsp70 (Hsc70 in mammals, Ssa1 in yeast) has broad substrate selectivity for a degenerate, hydrophobic 4–5 AA motif ^[29][64] that occurs in SPs, but also in virtually any other protein (Figure 3f,g). Despite its broad functionality as a chaperone, the protein can carry out substrate-specific ER delivery of secretory pathway proteins. Although Ssa1 can directly interact with Sec72, a part of the fungal Sec62/63 translocon that does not exist in mammals ^[30][65], the targeting is largely regulated in a substrate-dependent way by a network of Hsp40 (or J-domain protein) co-chaperones (Figure 2) such as the abundant, ER-tethered Ydj1 (Hdj2 in mammals) or Jjj3 ^{[31][32][33][34][35]}[66,67,68,69,70]. Hsp40s (i) increase the chaperoning activity of Hsp70s ^[36][71], (ii) recognize and deliver substrates to their respective Hsp70, and (iii) target their Hsp70 to a specific location in the cell ^[37][72]. It appears that these chaperones have some redundancy, as no single one of them is essential ^[32][67]. Hsc70-mediated targeting in mammals has been demonstrated for low-hydrophobicity TA proteins, but not yet for SP-containing proteins ^[38][73].

2.2.2. The Joker: The Snd Pathway

The list of ER targeting pathways was expanded when a screen in yeast revealed three proteins called Snd1-3 that can compensate for the known pathways and act as a backup ER targeting system ^[10][15]. The Snd pathway has been implicated in the biogenesis of short secretory proteins ^[39][76] as well as both single- and multi-pass transmembrane proteins ^[11][40][16,77]. The system appears capable of recognizing SSs at any position in the protein, from N- to C-terminus ^[10][15], allowing it to substitute SRP and other pathways, such as the TRC/GET pathway.

3. Protein Translocation and Insertion at the ER Membrane

Following ER targeting, secretory pathway proteins must be either translocated through or inserted into the ER membrane. The main insertase/translocase for SPs is the protein-conducting channel Sec61, which is a part of the ER translocon and can associate with different auxiliary protein complexes (Figure 4) ^[41][80]. The ER translocon is a dynamic super-complex at the ER membrane that facilitates the translocation, folding, and post-translational modification of many NCs (Figure 4). In addition to its central component Sec61, the ER translocon features a portfolio of sub-complexes with specific auxiliary functions that can be proactively recruited in a sub-stoichiometric manner depending on the type of SS at hand ^[42][41][11,80]. Some, such as the translocon-associated protein complex (TRAP), are associated near-stoichiometrically to the co-translational ER translocon (Figure 4), while others such as the SPC can act in concert but are not or only transiently recruited ^[43][81]. New configurations of the translocon are still being discovered and structurally characterized, giving a plethora of new information their interactions with SSs ^{[8][44][45][46][47][48][49][50][51][52]}[5,82,83,84,85,86,87,88,89,90]. For example, a range of newly characterized ER-resident insertase complexes containing an Oxa1 family subunit can facilitate the insertion of other types of SSs, particularly type III SASs and TAs, but also multi-pass TMPs (reviewed in ^[53][54][9,10]) in concert with Sec61. The ER-resident Oxa1 family complexes comprise the ER membrane complex (EMC), WRB/CAML, which is the insertase of the TRC/GET pathway, and the TMCO1 translocon. TMCO1 proteins might be involved in the translocation of SP-containing type I membrane proteins by facilitating the insertion of downstream TMDs after initial recognition by Sec61, respectively, while the EMC is fully dispensable for the insertion of SPs and type II TMDs but may have chaperoning activities ^[55][56][75,91]. To date, it appears that Sec61 is the only insertase for SPs.

Figure 4. ER translocation and insertion machineries for SP-containing proteins. Left upper panel: The Sec61 translocon comprises, among others, the ribosome (shown in the cytosolic background), Sec61 (Sec61α blue, Sec61β dark gray, Sec61γ light gray), and the yet structurally uncharacterized TRAP (green EM map). The structure is a composite of EMDB 4315 ^[52] and PDB 3JC2 ^[57]. Middle upper panel: The yet structurally uncharacterized Snd2/3 membrane complex ^[10]. Right upper panel: S. cerevisiae Sec61 opened by Sec62 (light teal), Sec63 (dark teal), and the yeast-specific Sec71/72 (white). The structure is a composite of PDBs 7AFT and 6ZZZ ^[58].

3.1. Through the Barricades: Co-Translational Translocation/Insertion

3.1.1. Tunnel of Love: Protein Translocation by Sec61

The heterotrimer Sec61, consisting of Sec61α, Sec61β, and Sec61γ, is chiefly responsible for (i) co-translationally translocating or inserting nascent chains and for (ii) determining substrate topology ^[59][60][61][94,95,96]. Its largest subunit, Sec61α, features two pseudo-symmetrical halves (formed by TMD1-5 and 6-10, respectively) that together produce an hourglass-shaped pore across the ER membrane (Figure 5) ^{[57][60][62][63]}[92,95,97,98]. The channel must be primed and activated by a partner to enable translocation or membrane insertion. Sec61 is gated in two directions: (i) opening of the two halves in a clam-like motion reveals a lateral gate to the ER membrane; (ii) the same motion also opens the vertical channel that is initially sealed by the ‘plug’ from the lumen, allowing the passage of the nascent chain ^[64][99] (Figure 5c,d).

Figure 5. Co-translational ER translocation by Sec61 through the ER translocon. (a) Graphic description of co-translational ER insertion/translocation by the SRP-Sec61. The SP is handed over from SRP54-M and inserts head-on into primed Sec61, removing the plug, before rearranging in a hairpin with a N_in-C_out conformation. The SP is then accommodated at the lateral gate of Sec61. (b) Tomographic map of the translating ER translocon (EMDB-4315) ^[52]. TRAP is accommodated at the side of Sec61, its cytosolic portion interacting with the ribosome, its luminal portion close to the opening of the translocation channel. The oligosaccharyltransferase (OST) does not interact with SPs. (c,d) Open and closed conformations of Sec61, based on PDBs 3J7Q and 3JC2 ^[57][60]. Upon channel opening, the plug (green, seen from within the lumen facing the membrane) is removed and the vertical channel is opened. The lateral gate can now harbor the SS. (e–g) SP binding site at the lateral gate. The binding site is partially open to the ER lipid environment, partially formed by hydrophobic residues of helices α2 and α7.

3.1.2. Smooth Operator: TRAP Is a Sec61 Assistant

SRP delivers SSs to Sec61 irrespective of their ‘gating‘ efficiency ^[65][66][100,107]. When encountering substrates with ineffective or slow channel gating abilities, Sec61 requires assisting factors that help it to exert second pulling event that occurs during translocation. The most stoichiometrically present of several auxiliary complexes is TRAP, which specializes in co-translational enhancement of Sec61 translocation/insertion.

3.2. Insane in the Membrane: Post-Translational Translocation/Insertion

Much like the co-translational pathways, post-translational translocation/insertion of proteins with N-terminal SPs and SSs is carried out through Sec61 (aided by auxiliary factors) while Oxa1 family complexes such as EMC and Get1/2 complexes likely do not interact with SPs.

4. SP Removal and Post-Targeting Functions

4.1. Time to Say Goodbye: SP Removal by the Signal Peptidase Complex

Regardless of how the specific characteristics of SSs influence the routes taken, the ultimate choice between SPs and SASs is—per definition—made after ER targeting and translocation. Completed translocation of about 80 amino acids through Sec61 marks the earliest possible time point for co-translational SP removal ^[67][132]. The SPC, a hetero-tetrameric serine protease, which exists in two distinct paralogs in higher eukaryotes ^[8][5], is responsible for this process.

By definition, the SPC only cleaves SPs, although TAs can also be engineered sufficiently short to be cleaved in the ER lumen ^[68][133]. SPC substrates require an h-region below 18–20 amino acids and a c-region with short, hydrophobic residues at the positions −1 and −3 relative to the cleavage site ^[8][68][69][5,133,134].

Although it is assumed that SP removal typically occurs co-translationally, post-translational SP cleavage is well within the SPCs repertoire ^[70][135]. SPs of co-translationally translocated NCs need to transfer from Sec61 to the SPC. However, the SPC is not resolved as part of native ribosome–translocon complexes by cryo-electron tomography ^[71][72][106,136], indicating that the SPC likely associates with the ER translocon transiently or in a structurally flexible manner ^[43][81].

The molecular workings of eukaryotic SP removal, and the distinction between SPs and SASs by the SPC, have eluded mechanistic explanation for decades. Recent structural insights reveal that the enzyme complex employs two structural motifs to make this distinction ^[8][5]: (i) a specialized transmembrane window (TM window) that locally thins the ER bilayer directly above (ii) a conserved shallow hydrophobic binding pocket that contains the cleavage site, located in the lumen about 17 Å from the luminal membrane interface (Figure 6).

Figure 6. SP removal by the signal peptidase complex. (a) Atomic model of an SP fitted into the binding pocket of SPC-C (shown as EM map, catalytic subunit SEC11C purple, SEC22/23 dark gray, SPC25 gray, SPC12 light gray; EMDB-13172). The micelle (show in the background in pale yellow) is thinned inside the TM window and measures the length of the h-region. The blowup shows the binding groove for with the SP c-region and the catalytic residues of SEC11C (orange). Residues −1 and −3 point towards the bottom of the pocket and therefore need to be short and hydrophobic. (b) Coarse-grained molecular dynamics simulations show the mean distance of an SP with 11 residues (L11) and 20 residues (L20) from the binding pocket. Long h-regions are excluded by the thinned membrane in the TM window. Figure based on ^[8].

4.2. The Show Must Go On: Post-Targeting Functions of SPs

While SASs and TAs remain associated with their proteins, SPs can have a range of interesting and versatile functions after their cleavage that are unrelated to their original protein. While many are recycled by signal peptide peptidases, specific SPs can, for example, modulate NK-cell mediated immune responses via HLA-E, become a part of mature virions, or associate with cytosolic factors such as calmodulin as feedback loops. For further reading, we refer to ^[28][73][12,142].