Protein Targeting

Contents

General principles of targeting

General considerations:

Topologically equivalent spaces in a eukaryotic cell.

The cytosol and nucleoplasm are topologically equivalent (although items passing from one to the other must pass through a nuclear pore), because free passage is restricted to small molecules only.

Single membrane bound vesicles (the endomembranes - RER, SER, Golgi, lysosomes, peroxisomes, etc) require the items to cross one membrane.

Double membrane vesicles: (endosymbionts - chloroplasts, mitochondria) require the item to cross two.

The targeting of proteins is encoded within their structures: signal peptides are stretches (primary structure) of amino acids that target the protein/polypeptide to a particular place, because they are recognised by other proteins whose job it is to import or export proteins into particular organelles. Some signals (so-called signal patches) are formed by tertiary folding, and some signals are effected as oligosaccharides attached to the protein rather than the amino acids themselves.

Signal patches and peptides.

To investigate signals, we can create recombinant genes by adding (or changing) the DNA for signal peptides, to see where the proteins end up. We can then use radiolabelled pulse-chase experiments, i.e. expose a cell to radioactive proteins, and see where the radiation is over a period of time by autoradiography. Alternatively, we can use GFP (green fluorescent protein) fusion proteins, which show up under fluorescence microscopy.

Below is a summary diagram showing the various targets in the cell and the pathways between them.

Traffic within the cell.

Translocation into the rough endoplasmic reticulum

Why must proteins destined for the ECM pass into the RER? The reason for this is that the space in the ER cisternae is topologically identical to the extracellular space. Proteins destined for ECM, RER, SER, Golgi, lysosomes, perinuclear space, etc., must be threaded through the RER membrane.

The synthesis of all proteins begins in the cytoplasm. Proteins destined for the RER (and onwards) have a signal peptide in the first few translated codons, which is recognised by a signal recognition particle (SRP) that pauses translation until the ribosome has been attached to the RER membrane. The SRP contains six polypeptides coating an RNA chain. Prokaryotes have a very similar system to this for the export of proteins out of the cell membrane.

Translocation begins when the SRP binds the ribosome.

The SRP binds to a SRP receptor (docking protein) on the RER membrane, and directs the nascent polypeptide into a protein translocator (Sec61), which is unplugged as the ribosome is recruited.

Translocation continues as the SRP brings the ribosome to dock on the RER.

SRP then dissociates and the polypeptide continues to be synthesised, and is threaded through the translocator. Although GTP is required for correct docking, it is not required for translocation.

Translocation continues as the protein is threaded through the membrane.

Translocation continues as the protein is threaded through the membrane.

Finally, a signal peptidase cleaves off the signal peptide, freeing the protein into the ER.

Translocation is completed as the N-terminus is cleaved by signal peptidase (although some proteins retain the signal to anchor them in the membrane).

Puromycin prematurely terminates translocation into the RER, which allows ions to flow through the protein translocator Sec61, showing it has an aqueous pore.

'Stop threading' sequences can be used to thread loops of protein through the RER.

Stop threading sequences can produce membrane-spanning proteins.

Ion channels often span the membrane 7 or more times. This is achieved by having several start/stop threading signals. These signal peptides are not consensus sequences, but instead rely on a positive charge followed by hydrophobic stretch followed by polar/negative stretch. The distinction between start and stop threading signals is largely determined by their order, not by their structure.

H3N+---MMSFVSLLLVGILFWATEAEQLTKCEVFQ---…---COO−

Glycosylation and trafficking in the RER and Golgi

The protein being threaded through the protein translocator will begin to fold even as it is being translated. However, the folding is generally not perfect first time, and the 'molten globules' of protein need to be helped to fold correctly.

Misfolding can be lethal: prions (the cause of CJD and BSE) are misfolded infectious proteins.

Misfolding of PrP proteins leads to spongiform encephalopathies.

PrP prion protein, can isomerise from stacked α-helices, to β-sheets stacked on α-helices. These incorrectly folded PrP* proteins catalyse the misfolding of normal PrP prions, leading to a chain reaction. The PrP* aggregates in amyloid plaques, destroying brain tissue and causing the lesions in the brain characteristic of CJD, BSE, scrapie and other spongiform encephalopathies.

Protein folding is mediated by 'heat shock proteins' (chaperonins) such as hsp60 (GroEL/GroES is a well-characterised bacterial chaperonin), and hsp70 BiP. GroEL/GroES (from E. coli) is a barrel, with internal hydrophobic groups. These recognise (incorrectly) exposed hydrophobic groups on the surface of the misfolded protein, and use ATP to 'poke' them where they belong, inside the protein.

GroEL and GroES.
GroEL (the barrel) and GroES (the lid) refold proteins using ATP.

Other enzymes are required in the RER, such as protein disulfide isomerase (PDI). The reason for this is that the cytosol is reducing, but the ECM is relatively oxidising, hence an enzyme is required to allow breakage and reformation of disulfide bonds. Insulin is a simple protein with three such bonds linking two chains.

Misfolded proteins stick to BiP and are dislocated into the cytoplasm, where they are tagged with ubiquitin and sent for degradation in proteasomes.

Improperly folded proteins stick to BiP and do not pass from the RER to Golgi. Is lack of transport genuinely selective, or is it just that the proteins are degraded so quickly in the ER they never escape? We don't know. However, it is know that broken proteins are dislocated (sent back to the cytosol) through Sec61 (back the way it came, although the receptors are different, and the translocation requires energy this time). Broken proteins are tagged with ubiquitin (a small protein) on certain lysine residues and sent to proteasomes (multimeric protein 'cemeteries').

Many proteins are glycosylated in the ER and Golgi.This may either be O-linked (via serine or threonine) or N-linked (usually via asparagine. O-linked glycosylation is characteristic only of proteoglycans, but N-linked glycosylation is performed on all proteins on every asparagine (N) of every NX{ST} sequence in the protein. The core oligosaccharide is composed of two GlcNAc (N-acetylglucosamine) and three mannose residues, but the decoration on this core is very variable.

Glycosylation makes proteins more hydrophilic; it allows them to bind to ECM receptors (lectins); it allows recognition of misfolded proteins by chaperonins; and it causes reduced access to proteases (sterically: oligosaccharides are quite inflexible and restrict access to other molecules).

Glycosylation occurs at the same time as translocation (indeed, as translation). On the cytosolic face, a precursor oligosaccharide is built from UDP-GlcNAc and GDP-mannose on a membrane-bound molecule of dolichol-phosphate. Dolichol is huge terpenoid lipid, which could (and does) span the RER membrane three times. The dolichol phosphate is then flipped to the lumenal side. On the lumenal side, two oligosaccharide units are preconstructed on another molecule of dolichol from GDP-mannose and GDP-glucose, then added to the main dolichol molecule. The oligosaccharide is finally transferred wholesale from the dolichol phosphate onto the protein by an oligosaccharyl transferase.

The precursor oligo is built on dolichol phosphate and transferred onto asparagine by oligosaccharyl transferase.

Some peptides swap a C-terminus stop-threading signal for a C-terminus anchor. Glycosyl phosphatidylinositol attaches the protein loosely to the inner ER membrane. When these reach the plasma membrane, these proteins accumulate in lipid rafts, where they are readily released by hydrolases, producing extracellular signals, and also releasing DAG into the plasma membrane (where it can act as a signalling molecule).

C-terminal tether.

After initial glycosylation and quality-control in the RER lumen, most proteins pass into the Golgi body. The Golgi has stacked cisternae, which, unlike the RER, are not contiguous. Each cisterna plays a specific role:

Cisternae of the Golgi apparatus.

The way in which the Golgi matures is not well understood: it may be by vesicular trafficking (i.e. vesicles bud from e.g. cis to medial); or it may be by cisternal maturation (i.e. the cis becomes the medial, a new cis being formed from budded-off CGN vesicles); or possibly a combination of both processes.

Cisternae are histochemically distinguishable: osmium tetroxide is reduced preferentially by the cis compartment, nucleoside diphosphatase (found in trans) cane be immunogold labelled, and acid phosphatase is most abundant in the TGN because the pH decreases through the stack.

Although most proteins traffic to the Golgi, some are to be retained in the RER, such as BiP and signal peptidase. These contain a 'ER return' signal peptide (KDEL) at their carboxy terminus.

H3N+---…---KDEL---COO−

All proteins can traffic into the CGN: BiP will be secreted if its KDEL signal is removed. However, KDEL receptors (which traffic back and forth) bring escapee proteins back to the ER. The binding of KDEL to receptors is probably pH dependent. Golgi traffic can be investigated using antibiotics: brefeldin-A blocks onward vesicular traffic, as evidence by the accumulation of Golgi proteins (such as mannosidase) in the RER. Nocodazole blocks the backwards vesicular traffic, and RER proteins can then be detected in the Golgi.

Traffic between the RER and Golgi is two-way.

An important modification to the glycosylation of proteins occurs between the RER and Golgi. Proteins that have not folded properly do not lose the three glucose residues in the 'tail' of their oligosaccharide, and are retained in the ER. This 'tail' acts something like a fuse, marking the protein for continued attention from chaperonins until it has folded correctly (or been abandoned to the proteasomes as irreparable). In the Golgi, glycosylation is not essential for continued trafficking, except to the lysosome, although some proteins will precipitate in the Golgi if they are improperly glycosylated.

Many alterations to glycosylation occur in the Golgi:

Glycosylation and modification of oligosaccharides in the RER and Golgi.

It is noteworthy that the RER and Golgi are defined by the enzymes present in them, and therefore a cell cannot create them de novo. The very least that would be required would be a vesicle with ER-importing translocators, but even this would probably not be enough. The cell therefore contains epigenetic information, not encoded in the genome, but without which it could not function.

Trafficking of vesicles to and from lysosomes

So far we have seen how vesicles traffic to and from the RER and Golgi. However, the Golgi is not the final destination for most proteins that pass into it. Many proteins are either destined for the ECM or for the endolysosome system.

Lysosomes are the principal site of cellular degradation. They contain hydrolytic enzymes, but are otherwise extremely variable in shape, role and contents. Lysosomes mainly destroy broken organelles and digest endocytosed particles and molecules. The inside of a lysosome is maintained at pH 5 by a proton ATPase and is full of hydrolytic enzymes; consequently these enzymes must be carefully targeted!

Lysosomes are large, acidic, degradative vesicles.

Different sorts of lysosome and endosome.
Endocytosis at the cell membrane produces (early) endosomes, which mature by fusion with Golgi vesicles to form late endosomes, and eventually mature to form lysosomes. Phagosomes are very large endosomes. Autophagosomes are formed in a different way, by the wrapping of organelles in need of disposal with ER.

Lysosomes are not all the same. In particular, we should distinguish lysosomes from their precursors, which are called endosomes. Early endosomes are formed by fusion of endocytosis vesicles from the plasma membrane. Recycling endosomes 'store' plasma membrane and the receptors it contains: these endosomes constantly exchange vesicles (in both directions) with the plasma membrane. Early endosomes fuse with vesicles containing hydrolytic enzymes from the trans Golgi network to form a late endosome, which matures to form a lysosome by loss of endocytosis receptors. Probably. The lysosome then fuses with other vesicles requiring digestion, and with autophagosomes containing ER-wrapped organelles destined for destruction. Lysosomes contain acid phosphatase, which can be specifically stained using lead salts and a phosphatase substrate: lead phosphate precipitates out.

Vacuoles are very large lysosomes found in plants and fungi; they are surrounded by a tonoplast membrane. These perform several roles, including a cheap space-filler (maximising photosynthetic surface area); and a toxic waste dump (both for toxins aimed against herbivores, and neutralised toxins taken up by roots).

Vacuoles are large lysosomes.

Signal patches mark enzymes destined for the lysosome with GlcNAc-P in the cis Golgi cisterna. The GlcNAc is cleaved off, leaving phosphate attached to mannose of an oligosaccharide. This mannose-6-phosphate is the signal for lysosomal trafficking.

Mannose-6-phosphate is the signal for lysosomal targeting.

Mannose-6-P directs vesicles to fuse with late endosomes. A clathrin coat concentrates the mannose-6-phosphate receptor.

Mannose-6-phosphate tagged proteins are concentrated by receptors and clathrin into vesicles destined for the endolysome system.

Adaptins allow clathrin to interact with and concentrate mannose-6-phosphate receptors (and through the receptors, their ligands, i.e. lysosomally targeted proteins). The low pH in endosomes causes the proteins to dissociate from the receptors, which are carried back to the TGN by vesicles.

Clathrin is a hexameric protein that forms cages. It is composed of three heavy chains and three light chains, arranged in a triskelion. The N-terminal knobs protrude into the 'cage' formed when several clathrin molecules link together.

Clathrin forms triskelia, which aggregate to form basketworks.

Clathrin forms a basketwork around small vesicles (and is found in patches on larger vesicles). Four (or more) adaptins allow different vesicles to concentrate different receptors (and their ligands). Adaptins recognise FRXY sequences or phosphorylated amino acids in the cytoplasmic tail of transmembrane proteins. Adaptins therefore bind receptors pointing into the extracellular/lumenal compartment to the clathrin which is found in the cytoplasmic compartment. Dynamin is a GTPase which pinches off budding vesicles. Vesicles rapidly lose their clathrin coat after budding: the uncoating of vesicles also requires energy, in the form of ATP (mediated by hsp70 proteins) and calcium.

Budding of vesicles requires receptors, adaptin, clathrin and dynamin.
COPI and COPII form vesicles not formed by clathrin. Note that clathrin is required where the trafficking requires physical 'clout', as at the plasmalemma, whose content of cholesterol makes it particularly stiff and difficult to bend.

Coating and uncoating of vesicles requires GTPases. Assembly of a COPII coat requires the GTPase Sar1. The donor membrane contains Sar1-GEF, which recruits cytosolic Sar1-GDP and activates it to Sar1-GTP. The GTPase acts like a timer switch: the vesicle must recruit a COPII coat and escape the donor membrane before Sar1 hydrolyses GTP and disassembles. Uncoating also requires energy - ATP, hsp70 and calcium.

Different vesicles are formed with different coating proteins.
COPII coats vesicles trafficking from ER to Golgi. COPI coats vesicles trafficking between Golgi compartments, from Golgi back to ER, from Golgi to cell surface (constitutive secretion), and between endosomes and lysosomes. Clathrin is used when physical clout is required: in concentrating enzymes to traffic to the lysosome, in creating endocytotic vesicles, and in removing membrane from Golgi vesicles as they mature to form secretory vesicles.

Vesicles are targeted to specific compartments by SNAREs and Rab. Rab proteins (which are GTPases) dock the vesicles and confer specificity.

  1. ER, Golgi.
  2. CGN.
  3. Secretory vesicles.
  4. Early endosomes.
  5. Plasma membrane.
  6. medial and trans Golgi.
  7. Late endosomes.

SNARE proteins exist in matched pairs and catalyse membrane fusion.

Rabs and SNAREs target vesicles to appropriate compartments.
Rab is a membrane-tethered GTPase. It allows specific docking of vesicles to target membranes, via Rab effector proteins. vSNAREs and tSNAREs conspire to fuse the docked vesicle with the target membrane (the tSNARE may require activation by Rab to do this). On fusion of the vesicle, Rab hydrolyses its GTP, and no longer binds the membrane (a hydrophobic 'tail' no longer pokes out of the molecule enough to anchor it to the membrane). Rab-GDP is ferried back to the donor membrane by GDI (GDP dissociation inhibitor), which delivers it to a GEF (guanosine exchange factor), which swaps its GDP for GTP, completing the cycle.

Endocytosis at the plasmalemma

The surface of a cell is in constant turmoil, as small and large vesicles fuse with it by exocytosis, and as it takes up items from the outside by endocytosis and pinocytosis. Pinocytosis is the endocytotic uptake of small lipid globules and fluid in <150 nm vesicles. This is a continuous process: a macrophage cell membrane turns over every 30 min. Pinocytosis also allows membranes to grow and shrink appropriately.

Phagocytosis is the uptake of larger items in larger vesicles > 250 nm, such as the endocytotic uptake of food items; sequestration of iron debris from haemoglobin degradation; or uptake of antigens by macrophages and neutrophils, which recognise complement proteins, antibodies and certain foreign oligosaccharides.

Endocytosis involves clathrin coated pits. Clathrin allows receptor mediated endocytosis, which in turn concentrates (up to 1000 times) extracellular molecules. About 25 receptor types are constitutively expressed on human cells, but clathrin coated pits can contain one (or more) of over 1000 different receptors.

Low density lipoprotein (LDL) is a 500 kDa protein which binds phospholipids and cholesterol:

LDL.

It binds to LDL receptors localised in clathrin coated pits. LDL is endocytosed by the budding off of the pit, and the vesicle fuses with an early endosome. The LDL receptors return to cell surface by exocytotic vesicular traffic. The early endosome matures, and eventually fuses with or becomes a lysosome. Mutant LDL receptors that cannot bind clathrin cause atherosclerosis.

LDL is taken up though clathrin coated pits.

Most endocytosed vesicles end up in early endosomes. The low pH (6) inside these organelles releases their ligands. The membrane and receptors are generally recycled back onto the plasma membrane, and the ligand is digested.

Recycling endosomes traffic ligands to lysosomes, but return receptors to the cell surface.

However, this is not the only fate for receptors. Transferrin acts as mobile carrier of iron, and both it and its receptor are recycled through the endosome system. Epidermal growth factor and its receptor are sent to the lysosome, so EGF down-regulates the sensitivity of its target cells. Glucose transporters are stored in the endosome, and released to the plasma membrane on stimulation by insulin.

Constitutive and regulated secretion

Exocytosis is the fusion of vesicles from the Golgi (or endosomes) with the plasma membrane. These vesicles are often targeted to particular surfaces: in polarised cells like those of epithelia, with either the basolateral membrane (towards the inside of the body), or the apical membrane (towards the outside of the body). Often, there will be two different endosome systems in such cells, each servicing a different surface.

Transcytosis is endocytosis directly coupled to exocytosis. Antibodies in mother's blood pass through milk gland cells directly into milk, and likewise through the baby's gut into its bloodstream. This avoids altering the composition of the two environments except by the exchange of a few specific molecules.

Transcytosis from apical to basolateral surface in a baby's gut.
Transcytosis of antibodies from intestinal lumen to tissue fluid. At the low pH of the intestinal lumen, antibodies associate with Fc receptors on the apical membrane. These are endocytosed and passed through the recycling endosome system to re-emerge on the basolateral surface where the higher pH causes the antibodies to dissociate.

As we can now see, there are three destinations for proteins leaving the Golgi TGN:

A sorting mechanism is obviously required in the TGN. Constitutive secretion is the default pathway (what happens if no signals are given). This pathway does not require clathrin, only COPI, and the vesicles leave the Golgi in a continuous stream.

Constitutive and regulated secretion.

Regulated secretion requires clathrin. The vesicles leave the Golgi, but are not secreted immediately, only disgorging their contents on stimulation by specific (often extracellular) signals. It is unclear how these proteins are sorted into secretory vesicles by the TGN. They may just self-aggregate, rather than using a specific signal. Exocytosis of secretory vesicles can be triggered by an increase in [Ca2+], or by action potentials (in the synapse). Synapses can exocytose vesicles 1000 times per second. The membrane added by all this vesicular traffic onto the plasma membrane is recovered by pinocytosis.

The content of secretory vesicles (mostly proteins) is concentrated by clathrin and by clathrin-mediated membrane recovery from the vesicle. Condensation of the contents often occurs by acidification, and secretory vesicles often stain densely under EM. The secretory vesicle often performs proteolytic modification:

A good example is the proteolytic processing of pro-opiomelanocortin:

Proteolytic processing of pro-opiomelanocortin to produce growth factors and endorphins.

You might wonder: why bother? The reason for this complex processing is that many hormones are only peptides - endorphins are only 5 amino acids long - and are therefore too short to be moved about directly. Processing also allows delayed activity e.g. the secretion of hydrolytic enzymes as zymogens in the digestive system.

Nuclear import

So far we have mostly discussed the secretory pathway targets in the cell. However, there are also several non-secretory pathway targets:

The cytosol is the default of defaults: where proteins go if nothing else is specified. The polypeptide synthesised on ribosome an simply diffuses into the cytoplasm on termination.

We will now see how molecules enter these other non-secretory targets.

For most macromolecules, getting in and out of the nucleus is hampered by the double membrane that surrounds it. Some small, nonpolar molecules can diffuse through the membranes (particularly important here are the steroid hormones), but much else goes via the nuclear pores.

Nuclear pore from above.

The nuclear pores, which are composed of (the imaginatively named) nucleoporin proteins, have an 8-fold symmetry, with an outer diameter of c. 100 nm, and an inner diameter in electron micrographs of c. 80 nm. The pore itself is only c. 9 nm (free diffusion), but dilates to 30 nm (active transport). Some nucleoporins (particularly the central 'plug') are sometimes lost during preparation. The cytoplasmic ring of the pore has fibrils, which are though to form queues of chemicals entering the nucleus. The nucleoplasmic ring has an attached 'basket'. Note the way the luminal nucleoporins clamp the doubled-back nuclear envelope.

Nuclear pore from the side.

If radiolabelled chemicals are microinjected into the cell, it is observed that entry into the nucleus is highly dependent on size:

However, some quite big things must be able to get in, such as RNA polymerase, DNA polymerase (300 kDa), and at least 100 histones min−1 per pore must enter the nucleus during S-phase. The two ribosomal subunits (30 nm diameter) must also get out at a rate of about 10 min−1 per pore. These proteins and RNAs are large, and fully formed: most other import mechanisms (e.g. SRP on RER) require the protein to be threaded through unfolded. These proteins must enter the nucleus complete and folded.

Proteins to be imported into the nucleus have a nuclear import peptide (nuclear localisation signal: NLS): PPKKKRKV (proline, lysine, arginine, valine) - the import signal binds FG-repeats in nuclear-import receptors (karyopherins, also known as importins), and on the fibril nucleoporins. The fibril FG repeats act as trackways. Moving chemicals against their electrochemical (concentration) gradients always requires energy (2nd law of thermodynamics). This is provided by GTP in the Ran cycle:

Ran cycle of nuclear import.

This cycle expends one GTP per cargo molecule imported into the nucleus.

Targeting to peroxisomes

Peroxisomes are small (200 nm) organelles that contain catalase for the degradation of H2O2. They provide compartmentalised detoxification of reactive oxygen species, and were probably the original oxidative organelle of the eukaryotic cell, evolving shortly after the Great Cyanobacterial Pollution of the atmosphere with oxygen. Their ability to handle oxygen was later recruited for oxidative metabolism, such as some amino acid metabolism and in specialised leaf peroxisomes that run the photorespiratory recovery pathway in plants. Glyoxysomes are another sort of specialised plant peroxisome. They oxidise fatty acids in plant seeds and feed the acetyl coenzyme-A into glucose synthesis, by a shortcut through the Krebs cycle called the glyoxylate cycle. Animals can't do this.

Peroxisomes reproduce by binary fission; however, there is no evidence they are endosymbiotic.

Peroxisomes.
Proteins in peroxisomes are often so concentrated that they precipitate out on crystalloid patches.

Import into peroxisomes occurs via peroxin proteins, which recognise two import sequences:

C-terminal signal:

H3N+---…---SKL---COO−

Near N-terminal signal:

H3N+---…---RLX5HL---……---COO−

Some peroxins are cytosolic; others are embedded in the peroxisome membrane. Import is poorly understood, but proteins appear to be imported pre-folded.

Zellweger syndrome is caused by a lack of a membrane bound peroxin called Pex2. Mutations in the gene cause fatally 'empty' peroxisomes, with no catalase. This causes prenatal oxidative damage to the liver and makes nerve cells unable to make myelin sheaths (particularly an ether-linked phospholipid called plasmalogen), which causes brain damage. Babies with Zellweger syndrome rarely survive beyond a few weeks.

Plasmalogen.
Plasmalogen: note the unusual ether linkage on the upper chain.

Import into mitochondria, plastids and other endosymbionts

Genes migrate from endosymbionts to the nucleus over time. There are several selection pressures operating here, including the fact that genes are more cosseted in the nucleus (particularly from reactive oxygen species found in mitochondria and plastids). The genes also receive the benefits of sex (whatever they are): endosymbiont genes are reproduced asexually and are therefore subject to Muller's ratchet, etc.

To take this to an extreme, the hydrogenosome is an organelle that oxidise protons to hydrogen in anaerobic 'amitochondriate' eukaryotes such as Trichomonas vaginalis, which causes an STI.

Normal mitochondria:

2H+ + 2e− + ½O2 ⇌ H2O

Hydrogenosome reaction:

2H+ + 2e− ⇌ H2

Hydrogenosomes are related to (or possibly derived from) mitochondria, but seem to have lost all their genes. A few still have small genomes.

Mitochondria are derived from an endosymbiotic α-proteobacterium (a relative of Rickettsia, which causes typhus and Rocky Mountain spotted fever). Most of their genes have been lost or moved to the nucleus, and consequently, many essential proteins (e.g. cytochrome oxidase) must be imported from the cytosolic ribosomes.

Mitochondria have two compartments:

And two membranes:

Topology of a mitochondrion.

Import into the mitochondrion is mediated by the TIM and TOM translocators, which span the inner and outer membranes respectively. There is no strict consensus like KDEL or PPKKKRKV for mitochondrial import. The signal is an amphipathic N-terminal α-helix with a positive stripe down one side, and a hydrophobic stripe down the other. An α-helix turns once every 3.6 amino acids, hence the approximate spacing of the blocks below:

H3N+---MLSL RQSI RFF KPAT RTL---…---COO−

Cytochrome oxidase mitochondrial signal peptide.
Part of mitochondrial import sequence from cytochrome oxidase. Note the hydrophobic phenylalanine (hexagonal) residues mostly point towards the bottom of the diagram, and the positively charged lysine and arginine residues mostly point towards the top of the diagram.

TOM feeds the pre-protein through into the IMS. This requires ATP and the involvement of cytosolic hsp70.

TOM outer membrane translocator.

TIM23 is one of three inner membrane translocators. The signal peptide is cleaved by matrix signal peptidase.

TIM23 inner membrane translocator.

TIM23 does not require ATP directly because the electron transport chain in the inner membrane expels protons from the matrix to the IMS. This PMF is used to power translocation into the matrix.

TOM and TIM23 associate with one another, and pinch the two membranes closely together.

TIM and TOM work together.

Translocated proteins transiently span both membranes: at 5°C, translocation of proteins across the mitochondrial membranes stops. Treatment with extracellular protease degrades the trailing cytosolic tail, and leaves only the peptidase-cleaved signal inside, but never anything longer. This indicates that the signal is cleaved while the protein is still being threaded, i.e. that TIM and TOM work simultaneously. If the protein spent any time in the IMS, it would be expected that large fragments with the signal peptide still attached would be retrieved when the mitochondrion is disrupted with detergents.

TIM22 is a second mitochondrial inner membrane translocator. It inserts multipass proteins into the inner membrane (of which there are many, including all the complexes of oxidative phosphorylation).

TIM22 inner membrane translocator.

TIM22 acts somewhat like the RER translocator: the signal peptide inserts into the membrane via TIM22, and start- and stop-threading signals lead to the protein being threaded through the membrane multiple times. The (TOM) signal peptide is cleaved off by the signal peptidase.

The third inner membrane translocator behaves in a similar way to TOM, but acting from the matrix to the IMS. OXA feeds the protein through the membrane: a hydrophobic sequence after the signal peptide proper causes the protein to insert into the inner membrane. OXA also exports/inserts proteins made by the mitochondrial ribosomes; and can also cleave its signal to release the protein into the IMS.

OXA inner membrane translocator.

The OXA translocator exports proteins to the IMS. These may either be synthesised in the matrix by the mitochondrion itself, of have already been imported to the matric by TIM/TOM. OXA threads proteins into the IMS: the hydrophobic signal for this may be cleaved or not, depending on whether the protein is supposed to be membrane-bound or soluble.

Hsp70 proteins and ATP are involved in mitochondrial import. Cytosolic and mitochondrial hsp70 chaperonins are involved in unfolding, import and refolding of proteins. Mitochondrial Hsp70 appears to act like a ratchet, using ATP. There are two models for its action: either a single Hsp70 changes shape as ATP is hydrolysed, letting it pull the protein in like a cog; or multiple Hsp70s bind the protein, pulling it through in a hand-over-hand fashion, as shown below. In the model shown below,  the ATP is hydrolysed as the Hsp70 dissociates from the end of the chain.

Chaperonins are required to unfold and refold proteins as they pass through TIM and TOM.

Chloroplasts have three compartments:

And three membranes:

Topology of a chloroplast.

However, don't panic: it's much the same as in mitochondria. Hydrophobic/positive stripes along a helix acts as signal peptides, although these must be distinguishable from mitochondrial signal, since if plastid genes are manipulated to have a mitochondrial import signal, they end up in mitochondria, rather than chloroplasts. However, there are two main differences: GTP is required (because there is no PMF across the inner membrane), and the third compartment adds extra complication.

Lumenal translocation requires a second signal and can use four different pathways:

Four pathways into the plastid lumen.

Basal chromists and alveolates have secondary plastids: i.e. their chloroplast are derived from green algae: a cyanobacterium within an alga within a cell. Some still retain a functional remnant of the green alga's nucleus, called the nucleomorph. This means the cell has three genomes, each directing proteins to umpteen compartments…

Chromalveolates may have up to three genomes and six compartments in the plastid system alone!

Summary

Test yourself

  1. What modifications do the ER and Golgi perform on proteins, and what is the purpose of those modifications?
  2. How do molecules enter the nucleus?
  3. Compare and contrast targeting to mitochondria and plastids.
  4. Compare and contrast targeting to peroxisomes and the nucleus.
  5. Why is the nuclear import sequence not cleaved off?
  6. Why is the RER import signal at the amino terminus?
  7. Why are proteins imported into peroxisomes pre-folded?
  8. What evidence could you collect for or against the two hypotheses for Golgi cisternae maturation?
  9. The last genes to jump ship from an endosymbiont to a nucleus are tRNAs and rRNAs. Why might this be?

Bibliography

Answers

Peer Review.
This page has been peer reviewed by 1 person.