Protein Targeting and the Secretory Pathway

Protein targeting machinery is a direct target and a diagnostic readout in modern biomedicine and biotechnology. Signal-sequence engineering is central to recombinant protein production: therapeutic antibodies, insulin, growth hormone, and vaccine antigens are manufactured in mammalian, yeast, or insect cell lines using optimized signal sequences fused to the coding sequence to maximize SRP capture and ER translocation efficiency, directly increasing secreted product yield in bioreactors. Defects in signal sequence recognition or Sec61 function cause human disease: mutations in Sec61α cause a form of autosomal-dominant polycystic liver disease and some forms of diabetes by impairing ER protein translocation and calcium handling; mutations affecting the signal sequence of preproinsulin cause a rare form of neonatal diabetes by preventing correct ER entry and processing. Signal peptidase and the SRP pathway are also drug targets: small molecules that block Sec61 (such as the natural products mycolactone, produced by Mycobacterium ulcerans and responsible for Buruli ulcer immune evasion, and cotransin, a synthetic Sec61 inhibitor) selectively block translocation of certain secretory and membrane proteins, and are being explored as anticancer and immunomodulatory agents because they can selectively silence secretion of specific cytokines or growth factor receptors depending on signal sequence "gating" properties. Cryo-electron microscopy has resolved near-atomic structures of the mammalian SRP–ribosome, SRP–SR docking complex, and the Sec61 translocon engaged with translating ribosomes, turning what were once biochemically inferred intermediates into directly visualized structures and enabling structure-based design of translocation inhibitors. Bacterial homologs of this pathway (Ffh/FtsY as the SRP/SR analogs, SecYEG as the translocon) are validated antibiotic targets, since blocking cotranslational membrane insertion is lethal to actively growing bacteria, and inhibitors targeting the bacterial signal peptidase are in development as a distinct antibacterial class independent of existing resistance mechanisms.

:::quiz[In the kinetic partitioning framework, if $k_{\mathrm{SRP}}[\mathrm{SRP}] = 0.4\ \mathrm{s^{-1}}$, $k_{\mathrm{fold}} = 0.1\ \mathrm{s^{-1}}$, and $k_{\mathrm{chap}}[\mathrm{Hsp70}] = 0.1\ \mathrm{s^{-1}}$, what fraction of nascent chains is captured by SRP?]

• 0.40
• 0.50
• 0.67 ✓
• 1.00

Ask the tutor: How does the cell distinguish a genuine ER signal-anchor sequence from a transmembrane segment destined for the mitochondrial inner membrane, when both are short hydrophobic helices?

A classical N-terminal signal sequence is 15–30 residues long and has a tripartite architecture: a positively charged n-region (usually 1–5 residues, net charge +1 to +3), a hydrophobic h-region (7–15 residues, often Leu/Ala/Val-rich, capable of forming an α-helix), and a polar c-region ending near a small-residue cleavage site (Ala-X-Ala being the classic signal peptidase recognition motif, the "(-3,-1) rule": small, uncharged residues at positions -3 and -1 relative to the cut). The sequence is not conserved at the residue level — there is no consensus motif in the way a restriction site is a motif — but its hydrophobicity profile and helical propensity are what SRP reads. The signal recognition particle is a cytosolic ribonucleoprotein: in mammals, a 300-nucleotide 7SL RNA scaffold plus six polypeptides (SRP9, 14, 19, 54, 68, 72). SRP54 is the business end — it has an N-domain, a GTPase (G) domain, and an M-domain (methionine-rich) that forms a hydrophobic groove lined with flexible Met side chains, able to clamp signal sequences of very different sequence but similar hydrophobicity, the way a bag of loosely packed rubber bands can grip screws of many different shapes. SRP binds the ribosome near the exit tunnel and samples emerging nascent chains; when a sufficiently hydrophobic signal sequence appears, SRP54's M-domain captures it with nanomolar affinity, and the SRP–ribosome–nascent-chain complex pauses (in the classical model, elongation is retarded, not fully arrested, allowing time for membrane targeting without wasting the whole translation event if targeting fails). At the ER membrane, the SRP receptor (SR, a heterodimer of SRα and SRβ, both GTPases) captures the SRP–RNC complex. SRP54's GTPase domain and SRα's GTPase domain then form a composite, reciprocally activated GTPase pair — each stimulates the other's GTP hydrolysis only when properly docked, a "GAP-in-trans" arrangement essentially unique to this system. This mutual activation only fires efficiently when the ribosome has simultaneously engaged the Sec61 translocon, coupling three-way recognition (SRP, SR, translocon) to nucleotide hydrolysis, and thereby committing the complex to translocation only after the nascent chain has a channel to enter. Hydrolysis triggers release of the signal sequence from SRP directly into the Sec61 channel and dissociation of SRP from SR (each partner then hydrolyzes its own GTP and recycles). Sec61 (SecY in bacteria) is a heterotrimeric channel (Sec61α, β, γ in mammals) forming an hourglass-shaped pore through the ER membrane, gated by a "plug" domain and a hydrophobic pore ring that seals around the translocating chain to maintain the ER calcium/redox permeability barrier. The channel can open laterally between transmembrane helices 2b and 7 of Sec61α, allowing sufficiently hydrophobic transmembrane segments of the nascent chain to partition directly into the lipid bilayer — this is how multipass membrane proteins are built one helix at a time, with the ribosome, translocon, and lipid bilayer collaborating in real time as translation proceeds.

The essential physical quantities are binding constants and rate constants that convert a reversible molecular recognition event into a committed, nearly irreversible targeting outcome. The overall efficiency of cotranslational targeting can be framed as a competition between the SRP-dependent pathway and futile cytosolic folding/aggregation, and between correct GTPase-coupled release and abortive dissociation.

where:

• $K_{d,\mathrm{SRP\text{-}signal}}$ is the dissociation constant of SRP54's M-domain for a hydrophobic signal sequence (typically low nanomolar for strong signals),
• $[\mathrm{SRP}]$, $[\text{signal peptide}]$ are free concentrations of SRP and the exposed signal sequence on the ribosome,
• $[\mathrm{SRP{\cdot}signal}]$ is the bound complex concentration.

Reading: a smaller $K_d$ means tighter, more efficient capture — this is why strongly hydrophobic signal sequences (more Leu/Ala/Val, longer h-region) target more efficiently than weakly hydrophobic ones, and why some "weak" signals rely on ribosome pausing (slow elongation over rare codons) to lengthen the capture window.

where:

• $\Delta G^{\circ\prime}_{\mathrm{SRP\text{-}signal}}$ is the binding free energy of SRP for the signal sequence,
• $\Delta G^{\circ\prime}_{\mathrm{SRP\text{-}SR}}$ is the binding free energy of the SRP–SR heterodimer docking interaction (itself GTP-dependent, since both proteins must be GTP-loaded to dock with high affinity),
• $\Delta G^{\circ\prime}_{\mathrm{GTP\ hydrolysis}\times2}$ is the free energy released by hydrolyzing two GTP molecules (one on SRP54, one on SRα), each contributing roughly $-30.5\ \mathrm{kJ/mol}$ under standard cellular conditions.

Reading: the two GTP hydrolysis events (~$-61\ \mathrm{kJ/mol}$ combined) are what make the targeting and handoff step effectively irreversible — the system spends chemical energy to convert a reversible protein–protein recognition event into a directional commitment, exactly analogous to the elongation-factor GTPase cycles on the ribosome itself.

where:

• $v_{\mathrm{translocation}}$ is the rate of polypeptide chain movement through the Sec61 channel,
• $k_{\mathrm{cat}}$ is the maximal per-ribosome translocation rate, driven by continued elongation (the ribosome itself, not a separate motor, pushes the chain through in the co-translational pathway),
• $[\mathrm{RNC{\cdot}Sec61}]$ is the concentration of productively docked ribosome–nascent-chain–translocon complexes,
• $K_M$ is the concentration at which docking is half-saturating.

Reading: because the ribosome supplies the "motor" via ongoing peptide bond formation (see The Ribosome and the Elongation Cycle), translocation rate in the co-translational pathway tracks elongation rate directly — roughly 2–8 residues per second in eukaryotic cells — rather than requiring a separate ATP-driven motor as in post-translational import.

Example 1 (foundational) — identifying and interpreting a signal sequence. Consider a hypothetical secretory protein's first 26 residues: Met-Lys-Trp-Val-Thr-Phe-Leu-Leu-Leu-Leu-Phe-Ile-Ser-Gly-Ser-Ala-Phe-Ser-Arg-Gly-Val-Phe-Arg-Arg-Asp-Ala… Assign the n-, h-, and c-regions and predict the signal peptidase cleavage site.

• The n-region is Met-Lys (position 1–2): a positively charged residue (Lys) near the N-terminus, consistent with the "positive-inside rule" bias.
• The h-region spans roughly Trp-Val-Thr-Phe-Leu-Leu-Leu-Leu-Phe-Ile (positions 3–12): predominantly bulky hydrophobic residues (Trp, Val, Phe, Leu×4, Ile), long enough (10 residues) to form a membrane-spanning-length hydrophobic α-helix, exactly the length window SRP54's groove reads efficiently.
• The c-region is Ser-Gly-Ser-Ala-Phe-Ser (positions 13–18): polar, contains small residues.
• Applying the (-3,-1) rule for signal peptidase: scanning for small uncharged residues at positions -3 and -1 relative to a candidate cut site, Ala at position 16 and Ser at position 18 fit well (Ala-Phe-Ser↓Arg would place -3 = Ala, -1 = Ser, cleaving between Ser18 and Arg19).
• Predicted mature protein begins at Arg19: Arg-Gly-Val-Phe-Arg-Arg-Asp-Ala…

This illustrates that recognition depends on the physicochemical class of residues (charge, size, hydrophobicity) at each position, not a fixed sequence motif — precisely why the signal hypothesis (Blobel and Sabatini) had to be tested biochemically rather than found by simple sequence alignment.

Example 2 (applied) — computing targeting efficiency from a kinetic competition. Suppose a nascent secretory protein's signal sequence is captured by SRP with pseudo-first-order rate constant $k_{\mathrm{SRP}}[\mathrm{SRP}] = 0.8\ \mathrm{s^{-1}}$, while off-pathway aggregation in the cytosol proceeds with $k_{\mathrm{agg}} = 0.05\ \mathrm{s^{-1}}$ and chaperone capture (which would divert it toward a slow post-translational rescue route) proceeds with $k_{\mathrm{chap}}[\mathrm{Hsp70}] = 0.15\ \mathrm{s^{-1}}$. Using the partitioning formula,

So 80% of nascent chains are captured co-translationally by SRP before an off-pathway fate claims them; the remaining 20% are split between aggregation (5%) and chaperone-mediated rescue (15%). This is a realistic order-of-magnitude picture for a moderately strong signal sequence — strong signal sequences push $k_{\mathrm{SRP}}[\mathrm{SRP}]$ several-fold higher, driving $f_{\mathrm{SRP}}$ toward 0.95–0.99, consistent with the near-complete cotranslational capture observed for classical secretory proteins such as preproinsulin or prolactin in pulse-chase experiments.

Limiting case check: if $k_{\mathrm{SRP}}[\mathrm{SRP}] \to 0$ (signal sequence too weakly hydrophobic to be recognized at all, e.g., a point mutant replacing the Leu-rich core with polar residues), then $f_{\mathrm{SRP}} \to 0/(0+0.05+0.15) = 0$: the protein is never cotranslationally targeted and either aggregates or is entirely dependent on chaperone-based rescue — matching classic experiments (Walter and Blobel) in which deletion or scrambling of the hydrophobic core abolished SRP-dependent translocation in reconstituted systems.

Example 3 (exam/competition-level) — energetics of a proofreading cycle with a mismatched, weakly hydrophobic pseudo-signal. A cytosolic protein has, purely by chance, a short internal stretch of six hydrophobic residues that weakly resembles a signal sequence, with $K_{d,\mathrm{SRP\text{-}pseudo}} = 50\ \mu\mathrm{M}$ (compare to a genuine strong signal sequence at $K_{d,\mathrm{SRP\text{-}signal}} = 20\ \mathrm{nM}$, a 2500-fold difference in affinity). Cytosolic free SRP concentration is roughly $[\mathrm{SRP}]_{\mathrm{free}} \approx 100\ \mathrm{nM}$. (a) Estimate fractional occupancy of SRP on the pseudo-signal versus the genuine signal, assuming simple bimolecular equilibrium binding $\theta = [\mathrm{SRP}]/(K_d+[\mathrm{SRP}])$. For the genuine signal:

For the pseudo-signal (using consistent units, $50\ \mu\mathrm{M} = 50{,}000\ \mathrm{nM}$):

(b) Interpret: even though both sequences are hydrophobic in a loose sense, the ~2500-fold difference in $K_d$ produces a ~415-fold difference in fractional SRP occupancy (0.83 vs. 0.0020), which is then compounded by the reciprocal GTPase proofreading step at the ER membrane (a weakly-bound signal is far more likely to dissociate before SR docking and GTP hydrolysis complete, since a shallow binding pocket gives a fast $k_{\mathrm{off}}$). The net discrimination — binding-affinity discrimination multiplied by kinetic-proofreading discrimination at the GTPase-gated handoff — routinely exceeds $10^4$–$10^5$-fold, which is why cytosolic proteins with incidental hydrophobic stretches are essentially never mistargeted to the ER, despite SRP54's need to recognize enormous sequence diversity among bona fide signal sequences. This combination of modest per-step discrimination compounded over multiple sequential steps — binding, GTPase gating, and Sec61 channel engagement — is the same qualitative strategy used in aminoacyl-tRNA selection accuracy at the ribosome, generalized to a targeting decision.

Günter Blobel (1936–2018) formulated the signal hypothesis in the early 1970s, proposing — well before the SRP or Sec61 were biochemically identified — that nascent secretory proteins carry an N-terminal signal sequence recognized by cellular machinery that directs cotranslational translocation across the ER membrane. Working at Rockefeller University, initially alongside David Sabatini and César Milstein, Blobel built on the classic observation that secretory proteins are synthesized on membrane-bound ribosomes and pass directly into the ER lumen during translation, rather than folding in the cytosol first. Over the following two decades, his laboratory used cell-free translation and reconstitution systems — canine pancreatic microsomes providing a source of ER membranes, wheat germ or reticulocyte lysates for translation — to biochemically dissect the components: purifying signal recognition particle, demonstrating its elongation-arrest activity, identifying the SRP receptor, and characterizing the Sec61 translocon and signal peptidase. This work established that protein targeting to essentially every organelle (ER, mitochondria, nucleus, peroxisomes, chloroplasts) follows a shared logic — sequence-encoded signals decoded by dedicated receptors — for which Blobel received the Nobel Prize in Physiology or Medicine in 1999. His insistence on rigorous biochemical reconstitution, rebuilding a complex cellular process from purified components in a test tube, set the methodological standard that the field of cell biology still follows.