Protein Targeting and Signal Sequences

A signal sequence (also called a signal peptide or targeting sequence) is a short segment of a polypeptide — typically a continuous stretch of 15 to 60 amino acids, though sometimes a three-dimensional signal patch formed by residues that are far apart in the linear sequence but folded together — that specifies where in the cell that protein belongs. Signal sequences are read by targeting receptors: soluble or membrane-bound proteins that recognize the signal with high specificity and physically deliver the polypeptide, or dock it, at the correct organelle. Most signal sequences are eventually cleaved off by a signal peptidase once the protein reaches its destination, though some (like mitochondrial internal targeting sequences and many membrane-anchor signals) are retained permanently as part of the mature protein. The system is modular in a very literal, testable sense. Each targeting pathway — ER import, mitochondrial import, chloroplast import (in plants), peroxisomal import, and nuclear import — uses a chemically distinct signal and a distinct receptor/channel apparatus, but the underlying logic is identical in every case: signal sequence → receptor recognition → energy-dependent translocation across (or insertion into) a membrane → (often) signal removal. Because the rules are combinatorial rather than exhaustive — the cell does not need a separate mechanism for every one of 10,000 proteins, just half a dozen general-purpose reading frames for half a dozen general classes of address — the system scales efficiently. A crucial conceptual distinction, first drawn clearly by Blobel, is between proteins synthesized on free ribosomes in the cytosol (cytosolic, nuclear, mitochondrial, chloroplast, and peroxisomal proteins — all imported after translation is complete, i.e., post-translationally, with mitochondrial/chloroplast/peroxisomal/nuclear import all falling in this category) versus proteins synthesized on ER-bound ribosomes (secretory and most membrane proteins, targeted to the ER during translation, i.e., co-translationally). This distinction — post-translational versus co-translational targeting — is itself set by whether the nascent signal sequence is recognized in the cytosol before or during translation, and it foreshadows the detailed molecular choreography of ER targeting covered in Cotranslational Translocation into the ER.

Step 1 — Translation begins on a free (cytosolic) ribosome. All eukaryotic protein synthesis initiates on ribosomes floating free in the cytosol, whether the final destination is the cytosol itself, the nucleus, mitochondria, chloroplasts, peroxisomes, or (via the ER) the secretory pathway. There is no separate class of "ER ribosomes" or "mitochondrial ribosomes" fixed in advance — ribosomes are recruited to the ER membrane only after a signal sequence emerges and is recognized.

Step 2 — The nascent chain displays a signal sequence. As the ribosome extends the polypeptide, the amino-terminal signal sequence (for most pathways) emerges from the ribosomal exit tunnel. Its chemical character differs sharply by destination:

Step 3 — A cytosolic or membrane-bound receptor recognizes the signal. For ER-targeted proteins, the signal recognition particle (SRP), a ribonucleoprotein complex, binds the emerging signal sequence and pauses translation, then delivers the ribosome-nascent chain complex to the SRP receptor on the ER membrane (mechanism detailed under Cotranslational Translocation into the ER). For mitochondrial proteins, cytosolic chaperones (Hsp70, Hsp90) keep the (finished) polypeptide unfolded, and the presequence is recognized by the TOM complex (translocase of the outer membrane) receptors Tom20 and Tom22. For nuclear proteins, the NLS is recognized by importin-α/β. For peroxisomal PTS1-bearing proteins, the receptor is Pex5.

Step 4 — Translocation across (or into) the target membrane, powered by energy input. Translocation is never a passive diffusion event; it requires a channel and an energy source:

• ER: translocation through the Sec61 translocon, a protein-conducting channel; energy comes from GTP hydrolysis (SRP/SRP receptor) and, for post-translational import of some proteins, the Hsp70 chaperone BiP acting as a ratchet.
• Mitochondria: sequential passage through TOM (outer membrane) and TIM23 (inner membrane) complexes, driven by the electrochemical gradient across the inner membrane (Δψ) and by ATP-driven mtHsp70 "import motor" pulling the chain in (a Brownian ratchet mechanism).
• Nucleus: passage through the nuclear pore complex, an aqueous channel large enough to pass folded proteins and even ribonucleoprotein assemblies, driven by a Ran-GTPase gradient (RanGTP in the nucleus, RanGDP in the cytosol) rather than by ATP hydrolysis at the pore itself.
• Peroxisomes: the only pathway that imports fully folded, even oligomeric, proteins intact, using the Pex5/Pex13/Pex14 docking-and-translocation system; mechanism still less completely resolved than ER or mitochondrial import.

Step 5 — Signal removal (in most, not all, cases) and folding/maturation. A dedicated signal peptidase (ER) or matrix processing peptidase (mitochondria) cleaves the signal sequence once the protein has crossed. Chaperones in the destination compartment (BiP in the ER, mtHsp70/Hsp60 in mitochondria) assist folding. NLSs and PTS1 signals are typically not cleaved, because the same signal often continues to function each time the protein must re-enter its compartment after cell division (nuclear envelope breakdown and reformation) or because the folded protein's C-terminus stays exposed with no cost to function.

The single most important body of evidence establishing that signal sequences intrinsically direct protein targeting came from Günter Blobel, David Sabatini, and Bernhard Dobberstein at Rockefeller University in the early-to-mid 1970s, building on earlier morphological observations by George Palade (who first described the rough ER's role in secretory protein synthesis using electron microscopy and pulse-chase radioactive labeling in the 1960s, work for which Palade shared the 1974 Nobel Prize).

The problem. By the early 1970s it was known that secretory proteins are made on ribosomes bound to the ER, while cytosolic proteins are made on free ribosomes, and that secretory proteins somehow end up inside the ER lumen. But how a ribosome "knows" to dock on the ER, and how a growing chain crosses the ER membrane, was unexplained. Blobel and Sabatini proposed in 1971 the "signal hypothesis": that secretory proteins carry an N-terminal signal sequence that is translated first, is recognized by ER-membrane components, and directs the ribosome to the membrane and the nascent chain across it.

The decisive experiment (Blobel & Dobberstein, 1975, Journal of Cell Biology). This was a cell-free, reconstituted biochemistry experiment — the classic way cell biology moves from correlation to causation. The steps: 1. mRNA encoding a secretory protein (they used immunoglobulin light chain, a well-characterized secreted antibody fragment) was translated in vitro using a cell-free rabbit reticulocyte lysate system, containing ribosomes, tRNAs, and translation factors but no membranes. 2. Under these membrane-free conditions, the light chain made was slightly larger than the mature secreted protein — an extra N-terminal segment was present. This precursor was named "pre-prolactin"-type or, generically, the pre-protein. 3. When rough microsomal membrane vesicles (right-side-out ER-derived vesicles that retain their translocation machinery) were added to the translation reaction before translation began, the protein produced was the correct, mature-sized light chain, and — critically — it was found to be protected from externally added protease (protease added afterward could not digest it, because it was sequestered inside the vesicle lumen). 4. When the microsomes were added after translation was already complete, translocation and correct processing largely failed — showing that the process must occur cotranslationally, or very early during translation, for most proteins. 5. Sequencing showed the extra segment present without membranes, and absent with membranes, was a short N-terminal sequence enriched for hydrophobic residues — the signal sequence — that had been cleaved by a peptidase associated with the membrane vesicles.

What this proved. This experiment demonstrated, with a fully reconstituted and controllable system, that (1) the extra targeting information is intrinsic to the protein's own primary sequence (translated first, encoded by the mRNA itself, not added afterward by some separate tagging enzyme); (2) membranes supply the receptor and channel machinery, and are strictly required for translocation and correct processing; (3) the signal is removed by a membrane-associated peptidase after it has done its job; and (4) protection from protease is a rigorous physical proof that the polypeptide has actually crossed into an enclosed compartment, not merely bound to the outside of a membrane.

Sufficiency, the second pillar of proof. A signal sequence being present in the right proteins is necessary but not sufficient evidence that the sequence alone causes targeting — it might just correlate. The decisive sufficiency experiments came from gene-fusion studies in the late 1970s and 1980s: researchers took the DNA encoding a signal sequence (e.g., from a secreted or mitochondrial protein) and fused it, using recombinant DNA techniques, onto the front of an unrelated cytosolic protein that normally never leaves the cytosol. In every such experiment, the hybrid (chimeric) protein was now correctly redirected — secreted, or imported into mitochondria, depending on which signal was attached. Conversely, deleting or mutating the signal sequence from a normally secreted or imported protein caused it to remain, unprocessed, in the cytosol. This necessity-and-sufficiency pairing — remove the signal and targeting fails; add the signal to a foreign protein and targeting succeeds — is the gold-standard genetic logic that converts "correlated with" into "causes," and it is precisely analogous to the logic used in classical genetics to assign gene function.

Complementary evidence: the SRP and its receptor. Peter Walter and Blobel (1980–1982) purified the signal recognition particle, a cytosolic ribonucleoprotein containing a 7SL RNA and six polypeptides, and showed biochemically that it binds directly to the emerging signal sequence, arrests (or slows) elongation, and only resumes translation efficiently once the ribosome-SRP complex has docked onto the SRP receptor in the ER membrane, discharging SRP for another round. This built out the receptor half of the "signal → receptor → channel" logic and directly foreshadows the fuller molecular choreography taught under Cotranslational Translocation into the ER. For this body of work, Blobel received the 1999 Nobel Prize in Physiology or Medicine, "for the discovery that proteins have intrinsic signals that govern their transport and localization in the cell."

Example 1 — Interpreting a signal-sequence deletion experiment. A researcher studies a nuclear protein, "Protein X," which normally localizes entirely to the nucleus when a GFP-tagged version is expressed in cultured cells (fluorescence seen only over the nucleus). The protein's sequence contains a stretch, Pro-Lys-Lys-Lys-Arg-Lys-Val, near residue 130. Three constructs are made:

• Construct A: full-length Protein X–GFP → fluorescence exclusively nuclear.
• Construct B: Protein X with residues 126–136 (including the Pro-Lys-Lys-Lys-Arg-Lys-Val stretch) deleted, fused to GFP → fluorescence now spread diffusely through the cytosol, excluded from the nucleus.
• Construct C: GFP alone, with just the seven-residue Pro-Lys-Lys-Lys-Arg-Lys-Val peptide fused onto its N-terminus (GFP by itself, ~27 kDa, has no nuclear signal and stays cytosolic/diffuses passively) → fluorescence now concentrated in the nucleus.

Reasoning: Construct B is the necessity test — remove the candidate signal, lose nuclear import. Construct C is the sufficiency test — attach the candidate signal to an unrelated, normally non-nuclear reporter, and import is gained. Together, they prove the seven-residue basic stretch is both necessary and sufficient to act as a nuclear localization signal (NLS), recognized by importin-α/β and delivered through the nuclear pore complex. This is exactly the logical structure used historically to validate real signal sequences (e.g., the SV40 large T-antigen NLS, Pro-Lys-Lys-Lys-Arg-Lys-Val, the first NLS ever defined, by Kalderon, Roberts, Richardson & Smith, 1984).

Example 2 — Predicting the phenotype of a signal-peptidase-cleavage-site mutation. A secreted digestive enzyme normally has this arrangement: signal sequence (residues 1–22, cleaved by ER signal peptidase between residues 22 and 23) followed by the mature enzyme (residues 23–280). A point mutation changes residue 22 from alanine to a bulky, charged arginine, which is known biochemically to block efficient signal-peptidase cleavage at a c-region cleavage site (signal peptidase strongly prefers small, uncharged residues, especially Ala, Gly, Ser, at positions −1 and −3 relative to the cut site — the so-called "(−3,−1) rule"). Prediction: Translation still begins normally on a free ribosome; the signal sequence still emerges and is still recognized by SRP (the recognition event depends on the hydrophobic h-region, not the exact cleavage-site residue), so the ribosome-nascent chain complex still docks at the ER and the chain still translocates into the ER lumen through Sec61 — translocation itself is not blocked. But because cleavage fails, the protein remains as an uncleaved precursor, with the signal sequence still attached at its N-terminus. Depending on the specific protein, this can (a) impair folding, since many mature N-termini participate in defining a folded conformation, (b) trigger ER quality-control degradation via ER-associated degradation (ERAD), or (c) in some real disease-causing mutations of this type (e.g., certain preproalbumin or preprocollagen cleavage-site mutants), result in secretion of an abnormal precursor form with reduced or absent biological activity. The key conceptual point: recognition (by SRP) and cleavage (by signal peptidase) are separable biochemical steps, tested by separate classes of mutation — a hydrophobicity-reducing mutation in the h-region blocks recognition and prevents translocation altogether, while a cleavage-site mutation like this one permits translocation but blocks processing.

Günter Blobel (1936–2018) was a German-born American cell biologist whose entire career, from his PhD training under George Palade at Rockefeller University in the mid-1960s, was organized around a single deceptively simple question: how does a cell know where to send a protein? Blobel arrived at Rockefeller as a postdoctoral fellow in Palade's laboratory, which had already established, through classical pulse-chase radioautography, the basic secretory pathway (rough ER → Golgi → secretory vesicles → plasma membrane). Blobel's insight was to ask what determines a protein's entry into that pathway in the first place — why some nascent chains bind ribosomes to the ER and others do not. Working first with David Sabatini, Blobel proposed the signal hypothesis in a pair of 1971 papers, at a time when the idea that a protein could carry its own "postal code" was genuinely novel and required a leap from morphology (electron micrographs of ribosome-studded ER) to a molecular mechanism. Over the following decade and a half, Blobel's laboratory — including key collaborators Bernhard Dobberstein and Peter Walter — built the cell-free reconstitution systems that turned the hypothesis into rigorously demonstrated mechanism: in vitro translation with and without microsomal membranes, protease-protection assays, and eventually the purification of the signal recognition particle itself. Blobel extended the same logic to mitochondrial and peroxisomal targeting, arguing throughout his career that the "zip code" principle was a universal solution the eukaryotic cell had evolved independently, or adapted, for every membrane-bounded compartment.

Recombinant protein production (biotechnology and pharmaceuticals). Nearly every biologic drug manufactured today — insulin, monoclonal antibodies, clotting factors, vaccines antigens — is engineered using signal-sequence logic. Manufacturers fuse a strong, well-characterized secretory signal sequence onto the gene of interest expressed in yeast, insect, or mammalian cell culture, so that the therapeutic protein is secreted into the culture medium rather than accumulating inside cells, dramatically simplifying purification.

Gene therapy and organelle-targeted therapeutics. Because mitochondrial and peroxisomal targeting signals are modular and portable, researchers can attach mitochondrial presequences to engineered proteins (including gene-editing components and antioxidant enzymes) to direct them specifically into mitochondria, a strategy under active investigation for diseases involving mitochondrial dysfunction.

Diagnosing and understanding human disease. Mutations that destroy or create signal sequences and their cleavage sites are directly disease-causing. Zellweger syndrome and related peroxisome biogenesis disorders arise from mutations in PEX genes (including the PTS1 receptor Pex5), preventing peroxisomal matrix protein import and causing severe, usually fatal, developmental and metabolic defects. Certain forms of primary hyperoxaluria type 1 result from mutations that mistarget a liver enzyme (alanine-glyoxylate aminotransferase) that should go to peroxisomes but instead ends up in mitochondria due to altered targeting-signal recognition. Mutations affecting mitochondrial presequences or the TOM/TIM import machinery underlie some mitochondrial encephalomyopathies.

Antibiotic and antifungal drug design. Because bacterial and fungal protein-targeting machinery (e.g., the bacterial Sec system, distinct from but homologous to the eukaryotic system) differs in exploitable ways from the human system, components of these pathways are studied as potential selective drug targets.

Vaccine and antibody engineering. Signal-peptide optimization is now a standard step in designing mRNA vaccines and secreted antibody constructs, since the efficiency and fidelity of signal-sequence recognition can directly affect the yield of a properly folded, secreted antigen or therapeutic protein — a direct, present-day continuation of Blobel's founding cell-free experiments, and a bridge to the more detailed molecular mechanics covered under Cotranslational Translocation into the ER and Sorting Signals and Compartment Address Codes.

Ask the tutor: How exactly does the signal recognition particle pause translation, and what resumes it once the ribosome docks at the ER membrane?