Nucleated human cells are separated from the environment by the so-called plasma membrane and contain different subcellular compartments, called cell organelles (Figure 1). These organelles are surrounded and, thereby, separated from the aqueous, albeit gel-like, cytosol by at least one biological membrane and have to be distributed to daughter cells from the mother cell during cell division (with the exception of lipid droplets and peroxisomes). In the cytosol, the vast majority of the approximately 24,000 different polypeptides of human cells are synthesized by 80S ribosomes. Therefore, the distinct proteins of the various organelles and the plasma membrane have to, first, be targeted to the specific organelles and, subsequently, inserted into or translocated across the membrane(s) of the relevant organelles. The protein import into the organelle termed the endoplasmic reticulum (ER) is the first step in the biogenesis of about one-third of the different soluble and membrane proteins (MPs) of human cells and, therefore, represents a central cell biological research topic of the past fifty years as well as several years to come. In a second step, the non-ER proteins reach their functional location in either the extracellular space; one of the endocytotic or exocytotic organelles (ERGIC, Golgi apparatus, endosome, lysosome, or trafficking vesicles); the plasma, peroxisomal, and mitochondrial membrane or in lipid droplets by vesicular transport; direct budding of new organelles (peroxisomal precursors or lipid droplets); or the ER–SURF pathway (mitochondria) ^{[1][2][3][4][5][6][7][8][9][10]}[1,2,3,4,5,6,7,8,9,10]. The first insights into ER protein import were gained about seventy years ago. From electron microscopic images, Palade and Potter concluded that the ER represents a ‘continuous, tridimensional reticulum’ and that ‘the surface appears to be dotted with small, dense granules that cover them in part or in entirety’, i.e., cytosolic 80S ribosomes ^[11][12][11,12].

The latter observation paved the way for the ‘signal hypothesis’ by G. Blobel and colleagues ^{[1][2][13][14]}[1,2,13,14], which suggested that topogenic sequences in nascent precursor polypeptides guide the translating ribosomes to the ER membrane and that the subsequent membrane insertion or translocation occurs coupled to translation, i.e., cotranslationally (Figure 1). For membrane proteins, the beauty of this concept of cotranslational ER protein import is that their eventual hydrophobic transmembrane domain or domains do not face the problem of aggregation in the cytosol. Subsequent work in human cell-free systems and in the yeast Saccharomyces cerevisiae uncovered that the topogenic sequence, termed the amino-terminal signal peptide (SP) or SP-equivalent transmembrane helix (TMH), of a nascent precursor polypeptide is recognized and bound by the cytosolic signal recognition particle (SRP), which facilitates the association of the complex between the ribosome, nascent chain, and SRP with the heterodimeric SRP receptor in the ER membrane, termed the SR (Figure 2 and Figure 3) ^{[2][15][16][17][18][19][20][21][22][23][24][25][26][27][28]}[2,15,16,17,18,19,20,21,22,23,24,25,26,27,28]. Thus, the combined action of SRP plus SR represents an ER targeting pathway for nascent precursors of soluble and membrane proteins, as well as the corresponding mRNAs. Recently, proximity-based ribosome-profiling experiments confirmed the preference of SRP and SR for SPs and relatively amino-terminal TMHs of the nascent precursor polypeptide chains ^[29][30][29,30]. Typically, the interaction of SRP with SR leads to the cotranslational transfer of the ribosome-nascent chain complex (RNC) to the central component for both protein translocation and membrane insertion in the ER membrane, the translocon, or the heterotrimeric Sec61 complex (Figure 1) ^{[31][32][33][34][35][36][37][38][39][40][41]}[31,32,33,34,35,36,37,38,39,40,41]. SPs and TMHs of nascent precursors may spontaneously interact with and trigger the opening of the Sec61 channel; i.e., both the central aqueous channel, as well as the lateral gate (Figure 1), or the productive Sec61 interaction may be facilitated by one of the auxiliary components of the ER membrane, i.e., the heterotetrameric translocon-associated protein (TRAP) ^{[42][43][44][45][46][47][48]}[42,43,44,45,46,47,48], the heterodimeric Sec62/Sec63 complex with or without the help of the ER lumenal chaperone BiP ^{[49][50][51][52][53][54][55][56][57][58][59][60][61][62]}[49,50,51,52,53,54,55,56,57,58,59,60,61,62], and the translocating chain-associated membrane protein 1 (TRAM1) (Figure 1 and Figure 3) ^{[63][64][65][66][67]}[63,64,65,66,67]. Notably, however, there is a human paralog of the α-subunit of the Sec61 complex, termed Sec61α2, i.e., a putative alternative Sec61 complex that is more or less uncharacterized but recently addressed with respect to its substrates or client spectrum in yeast [68], and there exist alternative components for the targeting (SND, TRC/GET, and PEX3) ^{[69][70][71][72][73][74][75][76][77][78][79][80][81][82][83][84][85][86][87][88][89][90][91][92][93][94][95][96][97][98][99][100][101][102][103][104][105][106][107][108][109][110][111][112][113][114][115][116][117][118][119][120]}[69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120], as well as membrane insertion of precursors (EMC and the GEL-BOS-PAT complex) ^{[121][122][123][124][125][126][127][128][129][130][131][132]}[121,122,123,124,125,126,127,128,129,130,131,132].

The SPs, mentioned above, have a tripartite structure (including an amino-terminal or N-region, a hydrophobic or H-region, and a carboxy-terminal or C-region) and facilitate the ER targeting of soluble proteins, GPI-anchored MPs, and single- as well as multispanning type I MPs (Figure 2) ^{[15][16][17][69][70]}[15,16,17,69,70]. In contrast, all the other MPs of the ER membrane (hairpin or HP proteins, and single- and multispanning type II or type III MPs, as well as TA proteins) depend on more or less amino-terminal or even carboxy-terminal TMHs (with a typical length of 15–25 amino acid residues) that serve as SP equivalents and facilitate membrane targeting as well as membrane insertion ^[15][16][17][15,16,17]. HP- and GPI-anchored proteins are special ^{[71][72][73][74][75]}[71,72,73,74,75]: Following ER import and the simultaneous cleavage of the SP, as well as the carboxy-terminal GPI-attachment sequence, GPI-anchored MPs become membrane-anchored via carboxy-terminal GPI attachment. HP proteins do not contain a real transmembrane domain (TMD); instead, they comprise one or more hydrophobic stretches of amino acid residues, which are typically defined by databases as TMDs and, therefore, discussed here as such. By definition, precursors of soluble proteins with SP and of single-spanning type II or III and TA MPs contain only one topogenic sequence, whereas, in all other precursors, other TMDs (i.e., other than the most amino-terminal TMHs) may serve as alternative or additional topogenic sequences. Typically, SPs are removed from the precursor polypeptides in transit by signal peptidases with ER lumenal catalytic sites ^[76][77][78][76,77,78]. Furthermore, many of the polypeptides of the secretory pathway become N-glycosylated by one of the two oligosaccharyltransferases (OST), which also have their catalytic sites in the ER lumen but act either on precursors in transit (i.e., cotranslationally, OSTA, Figure 1) or posttranslationally (OSTB) ^[79][80][81][79,80,81].

The characterization of precursors capable of SRP-independent ER targeting, such as small presecretory proteins and TA proteins, first suggested the existence of SRP-independent ER targeting pathways to the ER ^{[82][83][84][85][86][87]}[82,83,84,85,86,87]. In contrast to the SRP/SR pathway, these alternative targeting pathways can direct precursor polypeptides to the Sec61 complex co- as well as posttranslationally, and are named the TRC or GET, PEX19/PEX3, and Snd2/Snd3 pathway (Figure 3) ^{[58][71][73][74][75][87][88][89][90][91][92][93][94][95][96][97][98][99][100][101][102][103][104][105][106]}[58,71,73,74,75,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106]. Notably, there is also SRP-independent targeting of all sorts of mRNAs to the ER surface (Figure 3) ^{[29][107][108][109][110]}[29,107,108,109,110]. Typically, this mRNA targeting to the ER depends on receptors for mRNAs (such as KTN1), or RNCs with nascent polypeptide chains that are not yet long enough to allow the interaction of the topogenic sequence with SRP (such as RRBP1, LRRC59, and AEG-1) (Figure 3) ^{[29][111][112][113]}[29,111,112,113]. In the case of the ER targeting of mRNAs that code for cytosolic proteins, the nascent polypeptide-associated complex (NAC) can bind to the amino-terminus of the nascent polypeptides and trigger their release from the Sec61 complex ^{[23][114][115][116]}[23,114,115,116].

As stated above, the insertion of precursors with SP or TMH into the Sec61 channel and the concomitant gating of the Sec61 channel to the open conformation occur spontaneously or involve client-specific auxiliary components of the Sec61 channel (TRAP, Sec62/Sec63 complex) (Figure 1 and Figure 3). Typically, the orientation of SP- and TMH in the Sec61 channel follows the positive inside rule; i.e., positively charged amino acid residues in the N-region support loop insertion (N_in-C_out) and positively charged residues downstream of the SP or TMH interfere with loop insertion and, therefore, favor head-on insertion (N_out-C_in) into the Sec61 channel (Figure 2) ^{[117][118][119][120]}[117,118,119,120]. Next, TMDs can enter the phospholipid bilayer via the lateral gate of the fully open Sec61 channel by lateral movement and large hydrophilic domains (i.e., with more than 50 amino acid residues), or entire soluble proteins can be translocated into the ER lumen by vectorial movement through the channel (Figure 1 and Figure 2). Alternatively, membrane insertion of some precursors of MPs can be facilitated co- or posttranslationally by evolutionarily conserved MP insertases that comprise a hydrophilic vestibule near the cytosolic surface of the ER membrane, but cannot form an aqueous channel through the ER membrane, and, therefore, cannot translocate hydrophilic domains with more than 50 amino acid residues across the ER membrane. These Oxa1-related insertases are the Wrb/Caml complex (in co-operation with cytosolic GET3 of the abovementioned TRC pathway), or the multimeric ER membrane protein complex (EMC, which may occasionally co-operate with the Sec61 complex), or the GEL-BOS-PAT complex, which depends on the interaction with but not the channel activity of the Sec61 complex (Figure 3) ^{[121][122][123][124][125][126][127][128][129][130][131][132]}[121,122,123,124,125,126,127,128,129,130,131,132]. Notably, the first has its exclusive role in the membrane insertion of TA membrane proteins ^[92][95][98][92,95,98]. In the case of the EMC, a proteomic approach identified TA proteins, as well as multispanning MPs, as the predominant clients, the latter of which were also characterized as the main clients of the GEL-BOS-PAT-Sec61 supercomplex but proposed to have a preference for multispanning MPs with four or more TMDs, including those with marginal hydrophobicity ^[131][132][131,132]. There are two excellent reviews, plus several very recent original articles, on the structural and mechanistic details of the various MP insertases in the ER membrane ^{[69][70][81][121][122][123][124][125][126][127][128][129][130][131][132]}[69,70,81,121,122,123,124,125,126,127,128,129,130,131,132]. 

Following the pioneering work by G. Blobel and B. Dobberstein ^[13][14][13,14], protein import into the human ER was usually studied with the focus on single precursor polypeptides that were analyzed one-by-one in cell-free assays or intact cells. These classical studies led to the conclusions of whether and how the targeting and membrane insertion or translocation of a certain precursor is facilitated by a certain component. Recently, more global approaches were developed, such as the already mentioned proximity-specific ribosome-profiling ^{[29][111][126][133][134]}[29,111,126,133,134] and quantitative proteomics approaches ^{[5][106][112][126][135][136][137][138][139]}[5,106,112,126,135,136,137,138,139]. Typically, the proteomic approach employed siRNA-mediated knock-down or CRISPR/Cas9-mediated knock-out of components one-by-one in human cells, label-free quantitative proteomic analysis, and differential protein abundance analysis to characterize client specificities of components (Figure 4, left part). In contrast to the classical analyses, the quantitative proteomics approach addresses the question of which precursors use a certain pathway or component in intact cells, i.e., under in-vivo-like conditions (Figure 4, right part).

Figure 4. Components for protein targeting to and insertion into the human ER membrane with their client spectra and possible regulatory mechanisms. The first two percentages in each client spectrum refer to the percentage of precursors with SP and TMH (including HP and TA proteins), respectively, among the total number of clients, while the following two numbers refer to the percentage of precursors with HP or TA among the precursors with TMH. The red arrows highlight top scoring numbers > 10, light grey arrows represent the number 0 and are shown to indicate that fact.