Protein folding is the physical process through which a protein chain acquires an organized three-dimensional conformation. Its amino acid sequence provides the principal information specifying that structure, while surrounding conditions and cellular machinery influence whether and how it forms. Folding establishes the spatial arrangement needed for many biological activities, including catalysis and molecular recognition. It is distinct from protein synthesis and from the computational task of predicting a structure from a sequence. (ncbi.nlm.nih.gov)
Structural organization
A protein’s primary structure is its sequence of residues joined by peptide bonds. Secondary structure consists of recurring backbone arrangements, particularly the alpha helix and beta sheet, stabilized by hydrogen bonds. Tertiary structure describes the overall organization of a single chain; quaternary structure describes the arrangement of multiple chains in an assembly. These levels classify structure rather than prescribe a universal sequence of folding steps. (ncbi.nlm.nih.gov)
Many proteins contain domains, compact units that can possess considerable structural independence. Folding can therefore involve local organization, domain formation, and interactions between domains. A functional protein need not be uniformly rigid: intrinsically disordered proteins and regions lack a single stable structure in isolation. Some become more ordered when they bind another molecule, whereas others function through flexible conformational ensembles. Such disorder is not necessarily a folding defect. (ncbi.nlm.nih.gov)
Physical forces and stability
Folding reflects the combined effects of interactions within the chain and with its surroundings. For many soluble proteins, the hydrophobic effect favors burial of nonpolar side chains away from water. Hydrogen bonding, electrostatic interactions, and van der Waals forces help stabilize particular arrangements. Some proteins also contain disulfide bonds, covalent connections between cysteine residues that constrain the chain. No single interaction explains every protein fold. (ncbi.nlm.nih.gov)
In thermodynamic terms, stability depends on the difference in Gibbs free energy between folded and unfolded ensembles under specified conditions. The loss of chain conformational entropy during folding must be balanced by favorable interactions and changes in solvent organization. Temperature, solvent composition, and other environmental conditions can change that balance. Consequently, a structure favored in one environment need not remain stable in another. (ebi.ac.uk)
Sequence information and folding pathways
Experiments by Christian Anfinsen and colleagues showed that unfolded, reduced ribonuclease could regain its active structure after suitable conditions were restored. These observations supported the thermodynamic hypothesis: the sequence can specify a native conformation favored in an appropriate environment. They did not establish that every protein can refold unaided under arbitrary conditions, or that cellular folding is independent of assistance. (nobelprize.org)
The Levinthal paradox highlights the enormous number of possible chain conformations: an exhaustive random search would be incompatible with observed folding times. Folding instead proceeds through biased exploration of conformational space. The energy-landscape description represents conformations by their free energies and connectivity, often using a funnel toward native-like states. Its roughness allows intermediates, competing routes, and kinetic traps; molecules with the same sequence need not follow identical trajectories. (nature.com)
Stability and folding speed are therefore different properties. The folded state’s free energy relative to other states determines equilibrium preference, whereas barriers and accessible pathways influence the rate of reaching it. Experiments on related proteins have demonstrated that landscape frustration can slow folding, and that environmental changes can shift the relative use of alternative pathways. (nature.com)
Folding inside cells
Folding often begins during translation, while a growing chain emerges from the ribosome. The cellular environment presents challenges absent from dilute laboratory solutions: partially folded chains can expose hydrophobic surfaces and associate incorrectly with neighboring molecules. The timing of chain synthesis also limits which interactions are initially possible. (ncbi.nlm.nih.gov)
Molecular chaperones assist folding by limiting inappropriate interactions and giving chains further opportunities to reach productive conformations. Hsp70-family proteins bind exposed regions, while chaperonins can provide protected folding chambers. Many chaperone cycles use ATP. Chaperones generally do not supply a separate structural template; they influence folding efficiency and the competition between folding and aggregation. (ncbi.nlm.nih.gov)
Misfolded proteins may lose activity or form aggregates. Some aggregates have amyloid fibrillar structures. In prion diseases, an abnormal protein conformation can promote conversion and aggregation of other molecules of the same protein. These phenomena concern alternative molecular states and assemblies, not simply an absence of structure. (pdb101-east.rcsb.org)
Experimental study and computational prediction
Researchers distinguish determining a structure from observing its formation. X-ray crystallography provides structural information, while nuclear magnetic resonance can characterize proteins in solution. Time-resolved and single-molecule fluorescence measurements probe folding populations and barriers. Molecular dynamics simulations complement experiments by calculating molecular trajectories, although their interpretation depends on interaction models and adequate sampling. (ncbi.nlm.nih.gov)
Structure prediction addresses a related but narrower question: what conformation is compatible with a sequence? AlphaFold demonstrated a major improvement in prediction accuracy in work published in 2021, using deep learning, evolutionary sequence information, and experimental structures from the Protein Data Bank. A predicted structure is not a recorded folding trajectory. Accurate coordinates alone do not establish folding rates, intermediate populations, or responses to every cellular condition; these require additional experimental and computational investigation. (nature.com)