Proteins after translation: sorting, folding, modifications, degradation

01A protein's address

Almost every protein starts on a ribosome in the cytosol; only the 13 encoded by mitochondrial DNA are made inside mitochondria, on their own ribosomes. A protein's amino acid sequence says where it should go. A sorting signal is usually a short stretch of the chain that a receptor recognises; a protein with none stays in the cytosol. Günter Blobel proposed the idea in 1971, showed by 1975 how a signal takes a protein into the endoplasmic reticulum, and was awarded the 1999 Nobel Prize in Physiology or Medicine for showing that proteins carry such signals, which work like postcodes.

Proteins pass into the nucleus through pores, folded; the nuclear signal is not cut off, probably so that a protein can be imported again after each mitosis. They enter the ER, mitochondria and peroxisomes through protein channels in the membrane, unfolded, except into peroxisomes, which take in proteins already folded. Most mitochondrial proteins come from nuclear genes, as mitochondrial DNA codes for only 13 proteins, and they are imported after they are made. From the ER, proteins travel in vesicles to the Golgi, the lysosomes, the cell membrane and the outside: the secretory pathway.

02Into the endoplasmic reticulum

Proteins for the ER, the Golgi, the lysosomes, the cell membrane and secretion enter the ER while they are being made:

1. 1

   The signal appears

   As the signal leaves the ribosome, the signal-recognition particle (SRP), six proteins on a small RNA, binds it and slows translation.
2. 2

   SRP finds the ER

   SRP brings the ribosome to its receptor on the ER membrane and hands it over to the translocon, a protein channel called Sec61, then lets go.
3. 3

   The chain threads through

   The ribosome goes on making the protein, and the chain passes through the channel into the lumen of the ER as it grows.
4. 4

   The signal is cut off

   Signal peptidase, on the lumen side, removes the signal sequence. A soluble protein is released into the lumen once it is finished.

The ribosomes of the rough ER are the same as free ones, held there only while they make a protein with an ER signal. A simple membrane protein, with its signal at the N-terminus, also has a stop-transfer sequence, a second hydrophobic stretch that halts the chain in the channel; the channel opens sideways and leaves it in the bilayer. A protein that crosses the membrane several times has hydrophobic stretches that act in turn as start- and stop-transfer sequences. The side of a membrane protein that faced the lumen ends up on the outside of the cell, which is why the sugars of the cell membrane's proteins all face out.

03Folding and quality control

A protein's sequence decides how it folds (Proteins), but in the crowded cell the hydrophobic stretches of a new chain can stick to other chains and clump. Chaperones prevent this. Hsp70 proteins bind those stretches as the chain leaves the ribosome and, using ATP, let go as it folds. Chaperonins are barrels in which a protein folds shut off from the rest of the cell: TRiC in our cytosol, which folds actin and tubulin among others, Hsp60 in mitochondria and GroEL in bacteria. Many chaperones are heat shock proteins, made in larger amounts after heat has unfolded proteins.

In the ER, the Hsp70 BiP holds new chains, protein disulfide isomerase makes and corrects disulfide bonds, and calnexin and calreticulin hold glycoproteins by their sugars until they fold. Only proteins that have folded and assembled correctly may leave. The rest are sent back to the cytosol, given ubiquitin and destroyed by the proteasome: ER-associated degradation. If unfolded proteins pile up, the unfolded protein response makes more chaperones and slows translation; if the stress goes on, the cell dies by apoptosis.

Some misfolded proteins stack into insoluble fibres, amyloid, that build up in and around cells: amyloid β in Alzheimer's disease, α-synuclein in Parkinson's disease. A prion is a misfolded form of the prion protein that converts normal copies to its own shape, as in Creutzfeldt–Jakob disease.

added to an asparagine in one block

14sugarsadded to an asparagine in one block

in ubiquitin

76amino acidsin ubiquitin

at least, in a chain that sends a protein to the proteasome

4ubiquitinsat least, in a chain that sends a protein to the proteasome

inside a lysosome

4.5 to 5pHinside a lysosome

04What is added and taken away

Precursors. Many proteins are made inactive and cut later. Proinsulin is cut in the secretory granules of β cells into insulin and C-peptide, which are released in equal amounts, so C-peptide shows how much insulin a person's own pancreas makes, even when they inject insulin. The pancreas makes its protein-digesting enzymes as inactive zymogens: trypsinogen is normally activated only in the duodenum, by enteropeptidase, and trypsin then activates the other zymogens, so the pancreas does not digest itself. When trypsinogen is switched on too early, inside the pancreas, as in hereditary pancreatitis, trypsin damages the gland.

Sugars and vitamins. N-linked sugars arrive as a ready-made block of 14, moved from a lipid carrier, dolichol, onto an asparagine in the sequence Asn-X-Ser or Asn-X-Thr, where X is any amino acid but proline. The enzymes that hydroxylate collagen need vitamin C to keep their iron active, and hydroxyproline steadies collagen's triple helix at body temperature. The carboxylase that adds a second carboxyl group to glutamates of clotting factors II, VII, IX and X, and of the anticoagulant proteins C and S, uses reduced vitamin K, which vitamin K epoxide reductase then recycles. These γ-carboxyglutamates bind calcium, which lets the factors gather on the membranes of activated platelets.

05Sorting in the Golgi

COPII vesicles carry proteins from the ER to the Golgi apparatus, and COPI vesicles bring escaped ER proteins back, recognised by their KDEL. Proteins pass through the Golgi's stacked cisternae from its cis face, near the ER, to the trans-Golgi network on the far side, where they are sorted:

• To the lysosomes. In the cis Golgi, a phosphotransferase recognises a patch on the folded surface of each lysosomal enzyme and starts a tag on its mannoses, which a second enzyme finishes as mannose 6-phosphate. In the trans-Golgi network, mannose 6-phosphate receptors gather the tagged enzymes into clathrin-coated vesicles bound for endosomes, where the more acid fluid makes the receptors let go; the receptors return to the Golgi. The enzymes end up in lysosomes, whose lumen a proton pump keeps at a pH of about 4.5 to 5, where their dozens of acid hydrolases work best.
• Regulated secretion. Cells that secrete on demand, such as the β cells of the pancreas with insulin and its acinar cells with digestive enzymes, store proteins in secretory granules until a signal, often a rise in calcium, releases them.
• Constitutive secretion. Proteins that reach the trans-Golgi network with no other signal go to the cell surface all the time. This default route brings new membrane proteins to the membrane and releases antibodies from plasma cells and albumin from liver cells.

06Destroying proteins

Proteins last from minutes to a lifetime. Cyclins, which drive the cell cycle, are destroyed at a fixed point in each cycle; haemoglobin lasts as long as its red cell, about 120 days; and the crystallins in the centre of the eye lens are never replaced. Damaged, misfolded and unwanted proteins are destroyed by two routes.

The ubiquitin–proteasome system works in the cytosol and the nucleus:

1. 1

   Activate

   E1, the ubiquitin-activating enzyme, uses ATP to take up ubiquitin, a small protein of 76 amino acids.
2. 2

   Pass on

   E1 hands the ubiquitin to an E2, a conjugating enzyme.
3. 3

   Choose the target

   An E3, a ubiquitin ligase, binds both the E2 and the protein to be destroyed, and the ubiquitin is joined to one of the protein's lysines; further rounds build a chain on it. Our cells have hundreds of E3s, which is how they choose which proteins to destroy and when.
4. 4

   Destroy

   A chain of at least four ubiquitins, each linked through its lysine 48, is recognised by the 26S proteasome. Its caps take off the ubiquitins, which are used again, unfold the protein using ATP and thread it into the barrel-shaped core, whose protease sites, on the inside, cut it into short peptides.

Aaron Ciechanover, Avram Hershko and Irwin Rose were awarded the 2004 Nobel Prize in Chemistry for discovering ubiquitin-mediated protein degradation. Some of the peptides the proteasome makes are carried into the ER and loaded onto MHC class I molecules, which show them at the cell surface to T cells; this is how a cell infected by a virus is recognised (Immunity, chapter VI).

Lysosomes digest proteins brought in by endocytosis and, by autophagy, parts of the cell itself: a double membrane closes round a portion of cytoplasm, even a whole mitochondrion, and the autophagosome it forms fuses with a lysosome. Starved cells use autophagy to recycle their own material, and all cells use it to clear damaged organelles. Yoshinori Ohsumi found the first genes it needs in yeast in the early 1990s, and was awarded the 2016 Nobel Prize in Physiology or Medicine.