Chaperone Proteins

Why chaperones matter

• Crowded cytoplasm needs babysitters. Cellular protein concentration is 200 to 400 mg/mL — roughly 30 percent of the wet weight of a bacterial cell. Aggregation rates scale with the square of concentration, so a freshly translated polypeptide with exposed hydrophobic residues has only seconds to fold before it sticks to a neighbor. Chaperones bind those hydrophobic patches and isolate the substrate until folding completes.
• Co-translational folding starts on the ribosome. Trigger factor in bacteria and the NAC complex in eukaryotes sit at the ribosome exit tunnel, contacting nascent chains within their first 30 to 50 residues. Hsp70/DnaK takes over for chains longer than that. Without these handlers, ~30 percent of newly synthesized proteins misfold under fast-growth conditions.
• Heat shock is a generic stress response. Heat, oxidative stress, heavy metals, ethanol, and viral infection all trigger Hsp induction through HSF1. Chaperone levels rise 5- to 50-fold within an hour, raising the unfolding threshold and rescuing damaged proteins. Cells that fail to mount this response die at temperatures only 2 to 4 degrees Celsius above their growth optimum.
• Anfinsen's paradox in the cell. Anfinsen showed in 1961 that ribonuclease A refolds spontaneously from urea — folding is information-complete in the sequence. But that worked at micromolar concentration in clean buffer. At physiological concentrations and temperatures, kinetic traps and aggregation dominate. Chaperones convert spontaneous folding into reliable folding without changing the thermodynamic endpoint.
• Disease vector for neurodegeneration. Huntington's disease (polyQ huntingtin), Parkinson's (alpha-synuclein), Alzheimer's (Abeta and tau), prion diseases (PrP), and ALS (SOD1, TDP-43) are all proteostasis failures. Chaperone capacity declines with age, and aggregation-prone proteins outcompete chaperone supply. HSF1 induction extends lifespan in C. elegans by 30 percent, partly through chaperone upregulation.
• Cancer dependency on Hsp90. Tumor cells carry mutated, unstable kinases (BCR-ABL, EGFR variants, mutant p53) that depend on Hsp90 to fold. Hsp90 expression rises 2- to 10-fold in many cancers. Hsp90 inhibitors like 17-AAG cause coordinated degradation of dozens of oncogenic clients at once — the chaperone is a "single shot, multiple targets" oncology hub.
• The proteostasis network is quantifiable. Mammalian cells express ~330 chaperone-related genes, with abundances spanning 5 orders of magnitude. Proteomic studies put total chaperone mass at 10 to 15 percent of soluble protein in stressed cells. Models like FoldEco can predict aggregation propensity from sequence and chaperone supply with 70 to 80 percent accuracy.

Common misconceptions

• Chaperones tell proteins how to fold. No — Anfinsen showed the sequence holds the information. Chaperones increase the yield of folding by suppressing the off-pathway alternatives. They are catalysts of correct folding, not designers.
• All chaperones use ATP. Most do (Hsp60, Hsp70, Hsp90, Hsp100). The small heat shock proteins (sHsps, like Hsp27 and alphaB-crystallin) are ATP-independent holdases that bind aggregation-prone substrates and hand them off to ATP-dependent foldases. Trigger factor is also ATP-independent.
• GroEL forces substrates into a cage. The encapsulation by GroES is essentially passive — the substrate is already bound to the apical domains. ATP and GroES binding remodel the cavity into a hydrophilic, enlarged folding chamber. The substrate folds by its own thermodynamics, just shielded from neighbors.
• Hsp70 binds folded proteins. Hsp70 binds short, extended, hydrophobic stretches — about 7 residues with positive flanking charge. Folded native proteins bury those stretches, so Hsp70 cannot grip them. Hsp90 is the late-stage chaperone that interacts with near-native clients.
• One ATP equals one folding event. Many substrates require multiple cycles. GroEL clients with hard kinetic problems (e.g. mitochondrial malate dehydrogenase) iterate 5 to 20 times before reaching native state. Each cycle costs 7 ATP for the GroEL ring plus turnover on Hsp70 in solution.
• Chaperones only act on new proteins. They also rescue heat-damaged or oxidatively-modified proteins, hand mistargeted proteins to the proteasome, and disaggregate stress granules. Hsp104 (yeast) and ClpB (bacteria) actively pull out polypeptides from amorphous aggregates and feed them to Hsp70 for refolding.

How the major chaperone cycles work

Hsp70 has two domains: a 44 kDa N-terminal nucleotide-binding domain (NBD) and a 28 kDa C-terminal substrate-binding domain (SBD). In the ATP state, the SBD lid is open and substrate affinity is low (off-rate ~5 s^-1). A J-domain cochaperone (DnaJ/Hsp40) brings the substrate and stimulates ATP hydrolysis, locking the lid down and trapping the substrate with high affinity (off-rate ~0.05 s^-1). A nucleotide exchange factor — GrpE in bacteria, BAG family in eukaryotes — kicks out ADP, ATP rebinds, and the lid reopens, releasing the substrate to attempt folding. One full cycle takes 10 to 30 seconds and consumes one ATP. Roughly 30 percent of nascent chains in E. coli pass through DnaK at least once.

GroEL/GroES is a different geometry. GroEL is a homotetradecamer arranged as two stacked seven-subunit rings. A non-native substrate with exposed hydrophobic surfaces binds the apical domains of one ring (the cis ring). ATP binding to that ring is positively cooperative (Hill coefficient ~3), and ATP plus GroES binding triggers a dramatic conformational change: the apical domains rotate 90 degrees and rise 60 angstroms, the cavity expands from ~45 cubic nanometers to ~85 cubic nanometers, and the inner surface flips from hydrophobic to hydrophilic. Substrate is released into this Anfinsen cage and given 10 to 15 seconds (the GroEL ATPase half-time) to fold in isolation. The trans ring is allosterically inhibited until cis ATP hydrolysis completes; then trans ATP and substrate binding kick GroES off the cis ring, and the substrate exits — folded or to be rebound. Each cycle costs 7 ATP per ring.

Hsp90 is a constitutive dimer with N-terminal ATPase, middle, and C-terminal dimerization domains. It works with a battery of cochaperones — Cdc37 for kinases, Hop for the Hsp70-to-Hsp90 handoff, immunophilins (FKBP51/52) for steroid receptors, and Aha1 to stimulate the slow ATPase. The cycle is much slower than Hsp70, taking 1 to 5 minutes, because Hsp90 acts on near-native clients that need only the final maturation step. Hsp90 inhibitors like 17-AAG bind the N-terminal ATP pocket — a Bergerat fold shared with DNA gyrase and not with most other ATPases — making selective inhibition possible.

Hsp70 vs Hsp90 vs GroEL/GroES

Property	Hsp70 (DnaK)	Hsp90	GroEL/GroES (Hsp60)
Subunit MW	70 kDa monomer	90 kDa dimer	57 kDa × 14 + 10 kDa × 7
Quaternary state	Monomer/dimer	Obligate dimer	Double 7-ring barrel + lid
Substrate stage	Nascent / unfolded	Near-native, late maturation	Non-native, kinetic traps
Substrate motif	~7-residue hydrophobic patch	Client-specific (kinase, GR)	Exposed hydrophobic surface
Cycle time	10–30 s	1–5 min	10–15 s per ring
ATP per cycle	1	1 per protomer (2 per dimer)	7 per ring
Key cochaperones	DnaJ/Hsp40, GrpE, BAG	Cdc37, Hop, Aha1, FKBP	GroES (the lid)
Drug targeting	VER-155008 (early)	Geldanamycin, 17-AAG (clinical)	None clinical
Clients	~30% of nascent proteome	~10% — kinases, receptors	~60% of E. coli proteome

Famous examples and clinical hooks

• p53 depends on Hsp90. Wild-type p53 is borderline stable; mutant p53 (R175H, R248Q, R273H) is severely destabilized and is rescued by Hsp90. Hsp90 inhibition causes mutant p53 degradation in tumor cells but not in normal cells, providing a therapeutic window.
• CFTR delta-F508. The most common cystic fibrosis mutation creates a misfolded chloride channel that is recognized by ER quality control and degraded. Modulating Hsp70 levels and using corrector drugs (Lumacaftor, VX-661) rescues a fraction to the membrane.
• Glucocorticoid receptor maturation. The GR is held in an inactive complex with Hsp90, Hsp70, and FKBP52 in the cytoplasm. Cortisol binding triggers chaperone release and translocation to the nucleus — a textbook Hsp90 client cycle.
• Prion folding by Hsp104 and Hsp70. In yeast, the [PSI+] prion form of Sup35 is propagated by Hsp104-mediated fragmentation of amyloid fibers, supplying new seeds. Tuning Hsp104 levels can cure or potentiate the prion state — a striking demonstration that chaperone load shapes inherited protein conformations.
• Heat shock therapy. Whole-body hyperthermia and arimoclomol (an HSF1 coactivator) raise chaperone capacity. Trials in inclusion body myositis and Niemann-Pick C have shown modest benefit, validating chaperone induction as a therapeutic axis.