The Chaperonin Barrel and the Proteasome

GroEL-GroES Hydrophobic Cages, Ubiquitin Death Tags, and the 20S Proteolytic Shredder

Christian Anfinsen showed in 1961 that denatured ribonuclease A refolds spontaneously in a dilute glass beaker, proving that amino acid sequence dictates three-dimensional structure. Inside a living Escherichia coli cell, however, macromolecular crowding reaches 350 milligrams per milliliter. Exposed hydrophobic patches on nascent chains clump into insoluble aggregates within seconds unless shielded by chaperonins.

Paul Sigler and Arthur Horwich solved the $2.8\text{-\AA}$ X-ray crystal structure of the bacterial GroEL chaperonin in 1994. Weighing 800 kilodaltons, GroEL is a double-barreled cylinder built from two back-to-back rings of seven subunits each. Its open rims display sticky hydrophobic jaws that catch misfolded polypeptides before they can aggregate.

When seven ATP molecules and the heptameric GroES dome cap dock onto a protein-loaded GroEL ring, the apical domains twist 60 degrees upward and 90 degrees clockwise. That rearrangement doubles the cavity volume to $175,000\text{ \AA}^3$ and buries the hydrophobic jaws inside the wall. The trapped $50\text{-kDa}$ protein sits inside a hydrophilic Anfinsen cage to fold alone.

The GroEL-GroES folding chamber runs on a strict 10-second clock set by ATP hydrolysis. Once the seven ATP molecules in the capped cis ring hydrolyze to ADP, they allosterically unlock the opposite trans ring to accept a new client and ATP. Binding at the trans ring immediately pops off the cis GroES lid and ejects the protein.

Rather than acting merely as a passive isolation box, GroEL actively unfolds kinetically trapped tangles before releasing them into the cavity. Helen Saibil and Franz-Ulrich Hartl demonstrated this mechanical rescue using cryo-EM and FRET. As the seven apical domains twist open during ATP and GroES binding, they pull outward on the substrate to stretch apart wrong $\beta$-sheet misfolds.

In the late 1970s, Avram Hershko, Aaron Ciechanover, and Irwin Rose investigated a thermodynamic puzzle: why does destroying unwanted intracellular proteins require ATP when peptide bond hydrolysis is energetically downhill? Fractionating rabbit reticulocyte extracts, they solved the mystery. ATP is spent to covalently weld a 76-amino-acid polypeptide death tag—ubiquitin—onto lysine residues of target proteins.

Ubiquitin conjugation proceeds through a three-step enzymatic relay: two human E1 activating enzymes pass ubiquitin to roughly 40 E2 conjugating enzymes, which partner with over 600 distinct E3 ubiquitin ligases. Each E3 ligase recognizes a specific molecular degron on a worn-out or regulatory protein. It then strings together a polyubiquitin chain linked through lysine-48.

Once marked with a chain of four or more lysine-48-linked ubiquitins, a condemned protein is delivered to the $2.5\text{-MDa}$ $26\text{S}$ proteasome. Solved at atomic resolution by Robert Huber and Wolfgang Baumeister in 1995, its central $20\text{S}$ core particle stacks four heptameric rings ($\alpha_7\beta_7\beta_7\alpha_7$) into a hollow barrel. That sealed chamber isolates the proteolytic active sites from the cytoplasm.

If the $20\text{S}$ proteasome barrel stayed open to the cytosol, it would chew up every healthy protein in sight. To prevent accidental carnage, the N-terminal tails of the two outer $\alpha_7$ rings interlace across the ends of the cylinder. Those tails seal the entry port down to a $1.3\text{-nm}$ gated pore until a $19\text{S}$ regulatory cap unlocks it.

Mounted at either end of the $20\text{S}$ barrel, the $19\text{S}$ regulatory cap uses its Rpn11 metalloprotease subunit to slice off and recycle the ubiquitin chain for reuse. Simultaneously, a hexameric ring of six AAA+ ATPase motors (Rpt1–Rpt6) grabs an unstructured peptide tail. Pulling with $11\text{ pN}$ of force, the ring unravels folded domains into the $1.3\text{-nm}$ pore.

Inside the central cavity of the two inner $\beta_7$ rings sit six proteolytic active sites whose catalytic nucleophile is the N-terminal threonine (Thr1) of each $\beta1$, $\beta2$, and $\beta5$ subunit. Combining three distinct cleavage specificities, the chamber chops the threaded polypeptide into 7- to 9-amino-acid peptides. Those short fragments match the exact length required for MHC Class I immune display.

Synthesized by Julian Adams and colleagues and approved in 2003, the dipeptide boronic acid drug bortezomib (Velcade) locks the catalytic Thr1 hydroxyl of the proteasome's $\beta5$ subunit into a tetrahedral boron adduct. Malignant multiple myeloma plasma cells synthesize massive amounts of misfolded antibody chains. Clogging their proteasomes triggers lethal endoplasmic-reticulum stress while sparing normal tissues.