Intrinsically disordered proteins

An ensemble of NMR structures of the Thylakoid soluble phosphoprotein TSP9, which shows a largely flexible protein chain.

Intrinsically unstructured proteins, often referred to as naturally unfolded proteins or disordered proteins, are proteins characterized by lack of stable tertiary structure when the protein exists as an isolated polypeptide chain (a subunit) under physiological conditions in vitro. The discovery of intrinsically unfolded proteins challenged the traditional protein structure paradigm, which states that a specific well-defined structure was required for the correct function of a protein and that the structure defines the function of the protein. This is clearly not the case for intrinsically unfolded proteins that remain functional despite the lack of a well-defined structure. Such proteins, in some cases, can adopt a fixed three dimensional structure after binding to other macromolecules.

Biological role of intrinsic disorder

Many disordered proteins have the binding affinity with their receptors regulated by post-translational modification, thus it has been proposed that the flexibility of disordered proteins facilitates the different conformational requirements for binding the modifying enzymes as well as their receptors. Intrinsic disorder is particularly enriched in proteins implicated in cell signaling, transcription and chromatin remodeling functions.

Flexible linkers

Disordered regions are often found as flexible linkers (or loops) connecting two globular or transmembrane domains. Linker sequences vary greatly in length and amino acid sequence, but are similar in amino acid composition (rich in polar uncharged amino acids). Flexible linkers allow the connecting domains to freely twist and rotate through space to recruit their binding partners or for those binding partners to induce larger scale interdomain conformation changes.

Coupled folding and binding

Many unstructured proteins undergo transitions to more ordered states upon binding to their targets. The coupled folding and binding may be local, involving only a few interacting residues, or it might involve an entire protein domain. It was recently shown that the coupled folding and binding allows the burial of a large surface area that would be possible only for fully structured proteins if they were much larger. Moreover, certain disordered regions might serve as "molecular switches" in regulating certain biological function by switching to ordered conformation upon molecular recognition like small molecule-binding, DNA/RNA binding, ion interactions etc.

The ability of disordered proteins to bind, and thus to exert a function, shows that stability is not a required condition. Many short functional sites, for example Short Linear Motifs are over-represented in disordered proteins.

Sequence signatures of disorder

Intrinsically unstructured proteins are characterized by a low content of bulky hydrophobic amino acids and a high proportion of polar and charged amino acids. Thus disordered sequences cannot bury sufficient hydrophobic core to fold like stable globular proteins. In some cases, hydrophobic clusters in disordered sequences provide the clues for identifying the regions that undergo coupled folding and binding. Such signatures are the basis of the prediction methods below.

Many disordered proteins also reveal low complexity sequences, i.e. sequences with overrepresentation of a few residues. While low complexity sequences are a strong indication of disorder, the reverse is not necessarily true, that is, not all disordered proteins have low complexity sequences. Disordered proteins have a low content of predicted secondary structure.

Identification of intrinsically unstructured proteins

Intrinsically unfolded proteins, once purified, can be identified by various experimental methods. Folded proteins have a high density (partial specific volume of 0.72-0.74 mL/g) and commensurately small radius of gyration. Hence, unfolded proteins can be detected by methods that are sensitive to molecular size, density or hydrodynamic drag, such as size exclusion chromatography, analytical ultracentrifugation, Small angle X-ray scattering (SAXS), and measurements of the diffusion constant. Unfolded proteins are also characterized by their lack of secondary structure, as assessed by far-UV (170-250 nm) circular dichroism (esp. a pronounced minimum at ~200 nm) or infrared spectroscopy.

Unfolded proteins have exposed backbone peptide groups exposed to solvent, so that they are readily cleaved by proteases, undergo rapid hydrogen-deuterium exchange and exhibit a small dispersion (<1 ppm) in their 1H amide chemical shifts as measured by NMR. (Folded proteins typically show dispersions as large as 5 ppm for the amide protons.)

The primary method to obtain information on disordered regions of a protein is NMR spectroscopy. The lack of electron density in X-ray crystallographic studies may also be a sign of disorder.

Disorder and Disease

Intrinsically unstructured proteins have been implicated in a number of diseases. Aggregation of misfolded proteins is the cause of many synucleinopathies. The aggregation of the intrinsically unstructured protein α-Synuclein is thought to be responsible. The structural flexibility of this protein together with its susceptibility to modification in the cell leads to misfolding and aggregation.

Many key oncogenes have large intrinsically unstructured regions, for example p53 and BRCA1. These regions of the proteins are responsible for mediating many of their interactions.

De novo prediction of intrinsically unstructured proteins

Computational methods exploit the sequence signatures of disorder to predict whether a protein is disordered given its amino acid sequence. The table below, which was originally adapted from and has been recently updated, shows the main features of software for disorder prediction. Note that different software use different definitions of disorder.

Predictor	What is predicted	Based on	Generates and uses multiple sequence alignment?
SPINE-D	Output long/short disorder and semi-disorder (0.4-0.7) and full disorder (0.7-1.0). Semi-disorder is semi-collapsed with some secondary structure.	A neural network based three-state predictor based on both local and global features. Ranked in Top 5 based on AUC in CASP 9.	Yes.
PONDR	All regions that are not rigid including random coils, partially unstructured regions, and molten globules	Local aa composition, flexibility, hydropathy, etc.	No
GlobPlot	Regions with high propensity for globularity on the Russell/Linding scale (propensities for secondary structures and random coils)	Russell/Linding scale of disorder	No
DisEMBL	LOOPS (regions devoid of regular secondary structure); HOT LOOPS (highly mobile loops); REMARK465 (regions lacking electron density in crystal structure)	Neural networks trained on X-ray structure data	No
SEG	Low-complexity segments that is, “simple sequences” or “compositionally biased regions”.	Locally optimized low-complexity segments are produced at defined levels of stringency and then refined according to the equations of Wootton and Federhen	No
Disopred2	Regions devoid of ordered regular secondary structure	Cascaded support vector machine classifiers trained on PSI-BLAST profiles	Yes
OnD-CRF	The transition between structurally ordered and mobile or disordered amino acids intervals under native conditions.	OnD-CRF applies Conditional Random Fields, CRFs, which rely on features generated from the amino acid sequence and from secondary structure prediction.	No
NORSp	Regions with No Ordered Regular Secondary Structure (NORS). Most, but not all, are highly flexible.	Secondary structure and solvent accessibility	Yes
FoldIndex	Regions that have a low hydrophobicity and high net charge (either loops or unstructured regions)	Charge/hydrophaty analyzed locally using a sliding window	No
Charge/hydropathy method.	Fully unstructured domains (random coils)	Global sequence composition	No
HCA (Hydrophobic Cluster Analysis)	Hydrophobic clusters, which tend to form secondary structure elements	Helical visualization of amino acid sequence	No
PreLink	Regions that are expected to be unstructured in all conditions, regardless of the presence of a binding partner	Compositional bias and low hydrophobic cluster content.	No
IUPred	Regions that lack a well-defined 3D-structure under native conditions	Energy resulting from inter-residue interactions, estimated from local amino acid composition	No
RONN	Regions that lack a well-defined 3D structure under native conditions	Bio-basis function neural network trained on disordered proteins	No
MD (Meta-Disorder predictor)	Regions of different "types"; for example, unstructured loops and regions containing few stable intra-chain contacts	A neural-network based meta-predictor that uses different sources of information predominantly obtained from orthogonal approaches	Yes
GeneSilico Metadisorder	Regions that lack a well-defined 3D structure under native conditions (REMARK-465)	Meta method, which uses other disorder predictors (like RONN, IUPred, POODLE, and many more). Based on them the consensus is calculated according method accuracy (optimized using ANN, filtering and other techniques). Currently the best available method (first 2 places in last CASP experiment (blind test))	Yes
IUPforest-L	Long disordered regions in a set of proteins	Moreau-Broto auto-correlation function of amino acid indices (AAIs)	No
MFDp	Different types of disorder including random coils, unstructured regions, molten globules, and REMARK-465-based regions.	An ensemble of 3 SVMs specialized for the prediction of short, long and generic disordered regions, which combines three complementary disorder predictors, sequence, sequence profiles, predicted secondary structure, solvent accessibility, backbone dihedral torsion angles, residue flexibility and B-factors. MFDp (unofficially) secured 3rd place in last CASP experiment)	Yes
ESpritz	Disorder definitions include: missing x-ray atoms (short), Disprot style disorder (long), and NMR flexibility. A probability of disorder is supplied with two decision thresholds which depend on a user preferred false positive rate.	Bi-directional neural networks and high quality with diverse data derived from the Protein Data Bank and Disprot. Compares extremely well with other CASP 9 servers. The method was designed to be very fast.	No
CSpritz	Disorder definitions include: missing x-ray atoms (short) and Disprot style disorder (long). A probability of disorder is supplied with two decision thresholds which depend on the false positive rate. Linear motifs within a disorder segment are determined by simple pattern matching from ELM.	Support Vector Machine and Bi-directional neural networks with high quality and diverse data derived from the Protein Data Bank and Disprot. Structural information is also supplied in the form of homologous templates. Compares extremely well with other CASP 9 servers.	Yes

Since the methods above use different definitions of disorder and they were trained on different datasets, it is difficult to estimate their relative accuracy, but disorder prediction category is a part of biannual CASP experiment that is designed to test methods according accuracy in finding regions with missing 3D structure.

See also

• DisProt database
• MobiDB: a comprehensive database or intrinsic protein disorder annotations
• IDPbyNMR