Proteasome

Why the proteasome matters

• It is responsible for ~80% of regulated protein turnover. The remaining ~20% goes through autophagy/lysosomal degradation. Every cell cycle, every signal transduction event, every immune response depends on the proteasome destroying the right proteins at the right moment — cyclin B at anaphase, IκB at NF-κB activation, HIF1α under normoxia, β-catenin in unstimulated Wnt cells.
• It is the single best validated drug target in oncology. Bortezomib (2003), carfilzomib (2012), and ixazomib (2015) generate ~$5B/year in multiple myeloma revenue. The class doubled median myeloma survival from ~3 to >6 years. Resistance occurs via β-5 active-site mutations and autophagy upregulation but rarely via complete proteasome loss — the cell cannot live without it.
• Ciechanover, Hershko, and Rose won the 2004 Nobel. Their 1980s work in Israel and the US identified ATP-dependent proteolysis as ubiquitin-mediated, defined the E1-E2-E3 cascade, and established that ubiquitin tags substrates for the proteasome. The basic biochemistry is now textbook standard.
• It does the antigen presentation work. The immunoproteasome (in IFN-γ-stimulated cells) generates 8-9 amino acid peptides that bind MHC class I and are displayed to CD8 T cells. Every cell with a virus is destroyed via this pathway. The thymoproteasome (with β-5t) does positive selection of CD8 T cells in cortical thymus.
• It is mechanically remarkable. The 19S cap is one of the most studied AAA+ unfoldase machines. Six Rpt subunits work hand-over-hand in a sequential rotational cycle (Andreas Martin, Cell 2017) to pull substrate through their central pore at ~5-10 amino acids per ATP, generating 14 pN of pulling force.
• Proteasome activity drops with age. Aged human muscle and brain show 30-50% lower proteasome activity, contributing to accumulation of damaged and misfolded proteins. Long-lived organisms (naked mole rats, bowhead whales) maintain higher proteasome activity into old age — a candidate longevity intervention is the new class of proteasome activators (e.g., USP14 inhibitor IU1).
• Targeted protein degradation (PROTACs) hijacks it. A bifunctional small molecule that links a target-binding warhead to an E3 ligase recruiter (typically VHL or cereblon) hijacks the ubiquitin-proteasome system to destroy any target — currently extending to 'undruggable' proteins like KRAS, AR-V7, and STAT3. Arvinas's ARV-471 (estrogen-receptor PROTAC) is in Phase III for breast cancer.

Common misconceptions

• The proteasome destroys everything random. No — it requires substrates to be tagged with a K48-linked polyubiquitin chain of ≥4 ubiquitins (other linkages — K11, K63 — generally signal other fates: K63 endosome trafficking, K11 cell cycle). The 20S alone can also degrade unfolded/oxidized proteins ubiquitin-independently, but the canonical 26S is strictly tag-dependent.
• Ubiquitin-tagging is irreversible. Hundreds of deubiquitinases (DUBs — ~100 in humans) actively edit chains. Rpn11 in the 19S cap removes chains during translocation; USP14 and UCH37 trim chains pre-degradation, sometimes rescuing substrates. The system is dynamic with a steady-state on-off equilibrium.
• The 26S degrades whole proteins. It produces peptides 3-25 residues long, which are released and chewed by cytosolic peptidases (TPP2, TOP) into single amino acids within minutes. The 20S chamber holds substrate while the threonine active sites cleave repeatedly — average peptide release length is 8-9 amino acids, perfectly sized for MHC I loading.
• Only short-lived proteins go through the proteasome. Half-lives in cells span 6 orders of magnitude (seconds to weeks). Even long-lived structural proteins (collagen, myosin) eventually turn over via the proteasome at the ~weeks-to-months timescale. Median protein half-life in mammalian cells is ~36 hours.
• The 20S core is always closed. The α-ring gate is in dynamic equilibrium between closed (default, blocking entry) and open (during 19S docking or by activator binding). Free 20S has the gate closed; 19S binding triggers ATP-dependent gate opening. PA28 and PA200 caps open the gate via different mechanisms.
• Proteasome inhibition kills all dividing cells. Multiple myeloma is uniquely sensitive because of its high baseline ER stress from immunoglobulin synthesis. Other cancers (lung, breast, melanoma) tolerate bortezomib much better — the therapeutic window in myeloma reflects a rare metabolic vulnerability, not a universal anticancer effect.

How a single substrate is destroyed

The cycle starts when a substrate is poly-ubiquitinated. An E1 enzyme (UBA1 in humans, ~120 kDa) charges ubiquitin (76 amino acids, 8.5 kDa) onto its catalytic cysteine using ATP, forming a thioester. The ubiquitin transfers to an E2 (~40 in humans), then an E3 ligase recruits the substrate and catalyzes ubiquitin transfer onto a substrate lysine. RING-family E3s (~95% of human E3s) bring E2 and substrate into proximity; HECT E3s receive ubiquitin first onto their own active-site cysteine before transferring it. After the first ubiquitin is conjugated, additional ubiquitins are added — each new ubiquitin's C-terminus joining the K48 of the previous — building a chain. Chains of ≥4 ubiquitins are the standard proteasome signal, though shorter chains and even monoubiquitin can suffice for some substrates.

The polyubiquitinated substrate diffuses to a 26S proteasome and docks via Rpn10 or Rpn13 ubiquitin receptors on the 19S cap. The substrate's unstructured initiation region threads into the central pore of the six Rpt AAA+ ATPase subunits. ATP hydrolysis drives sequential rotational translocation — Rpt1 grips, hydrolyzes ATP, releases; Rpt2 grips, hydrolyzes, releases; and so on around the ring at ~5-10 amino acids per ATP and ~14 pN pulling force. As the substrate moves into the chamber, the metalloprotease Rpn11 cleaves the polyubiquitin chain en bloc at the substrate-distal isopeptide bond, releasing intact ubiquitin chains for recycling. The substrate enters the 20S α-ring (gate opens upon docking) and reaches the central chamber where three pairs of catalytic β-subunits — β1 (caspase-like, after acidic residues), β2 (trypsin-like, after basic), β5 (chymotrypsin-like, after hydrophobic) — cleave it. Each cleavage event uses an N-terminal threonine as the nucleophile in an Ntn-hydrolase mechanism: the α-amino group deprotonates the Thr-OH, which attacks the carbonyl, forming a covalent acyl-enzyme intermediate, which then hydrolyzes. Peptides 3-25 amino acids long (mean ~8-9) diffuse out of the antechamber and are degraded further by cytosolic peptidases. The full cycle takes 1-2 minutes per substrate; a single 26S proteasome can degrade ~1 substrate every 30-60 seconds at saturating substrate.

Proteasome variants

Variant	Catalytic subunits	Cellular context	Function	Inhibitor / drug
Constitutive 26S	β1, β2, β5	All eukaryotic cells	General ubiquitin-tagged degradation	Bortezomib, carfilzomib, ixazomib
Immunoproteasome	β1i (LMP2), β2i (MECL1), β5i (LMP7)	IFN-γ stimulated, immune cells	MHC class I peptide generation	KZR-616 (ONX-0914 derivative, autoimmune)
Thymoproteasome	β1i, β2i, β5t	Cortical thymic epithelial cells only	Positive selection of CD8 T cells	None clinical
20S free core	β1, β2, β5	All cells (~30% of total)	Ubiquitin-independent degradation of oxidized/unfolded proteins	Same active-site inhibitors
PA28 (11S)-capped	β1, β2, β5	IFN-γ stimulated	ATP- and Ub-independent peptide trimming	None
PA200 (PI31)-capped	β1, β2, β5	Spermatogenesis, DNA damage response	Histone tail clipping, chromatin proteostasis	None
Bacterial 20S (some)	β-only homohomeromer	Mycobacteria, actinobacteria	Pup (prokaryotic ubiquitin-like) tagged degradation	Oxathiazol-2-ones (M. tuberculosis)
Archaeal HslVU / 20S	Single β subunit type, 14-mer	Archaea and some bacteria	Ancestral protease, ATP-dependent (HslU cap)	Research only

Famous experiments

• Avram Hershko, Aaron Ciechanover, and Irwin Rose, 1978-1983. Identified ATP-dependent proteolysis in reticulocyte lysate (Hershko's lab at Technion); fractionated and discovered ubiquitin (then APF-1); reconstituted the E1-E2-E3 cascade. Won the 2004 Nobel Prize in Chemistry.
• Marc Glickman & Daniel Finley, late 1990s. Genetic dissection of yeast 26S subunits established the architecture of the 19S regulatory particle and the role of Rpn11 as the chain-stripping deubiquitinase.
• Robert Huber lab, 1995 (Nature). Solved the first crystal structure of the 20S proteasome (from Thermoplasma acidophilum), revealing the barrel architecture and N-terminal threonine active site that defined the Ntn-hydrolase family. Huber had already won the Nobel for photosynthetic reaction center work in 1988.
• Andreas Martin lab, 2017 (Cell). Cryo-EM of substrate-engaged 26S proteasome at every stage of the translocation cycle. Resolved the sequential rotational mechanism of the six Rpt AAA+ subunits and showed substrate is gripped hand-over-hand at ~5-10 residues per ATP.
• Adams & Stein, late 1990s (Millennium Pharmaceuticals). Developed bortezomib (PS-341) as a boronic-acid inhibitor of β-5; clinical trials in multiple myeloma showed dramatic responses; FDA approval in 2003 launched the proteasome-inhibitor class. Carfilzomib (epoxomicin-derived) followed in 2012, and the targeted protein degradation field (PROTACs, molecular glues) bloomed in the 2010s.