Intrinsically disordered proteins

The ubiquitous expression of Par-4, its post-translational modifications and a plethora of binding partners are characteristics common to many intrinsically disordered proteins (IDPs) [143] and prompted the investigation if Par-4 indeed can be classified as an IDP. A short introduction explaining the importance and characteristics of IDPs is given below. Under physiological conditions IDPs either entirely lack a well defined secondary and/or tertiary structure or contain long regions without a well defined secondary and/or tertiary structure. In contradiction to the structure-defines-function paradigm, stating that an ordered 3D structure is required for effective protein functioning, IDPs are biologically active and functional [144-146]. Proteins particularly involved in cell regulation exhibit a high prevalence for ID with approximately 75% of these proteins predicted to contain disordered regions of 40 or more consecutive residues. Cell regulation involves processes such as cell division, transcription, translation, signal transduction, protein phosphorylation, and assembly/disassembly of multi- protein complexes [146]. Additionally, cytoskeletal, ribosomal and cancer-associated proteins (e.g. Par-4) show a high abundance of disordered regions [147] implying that ID may be important for their function. The cancer-associated protein class may be considered as an extension of the cell regulatory protein class as highly regulated processes such as cell proliferation or apoptosis often become dysregulated in cancer cells [4]. A prominent example for the relation of signalling proteins with cancerous processes is the well established IDP p53. Loss of its transcriptional control due to mutations has been shown to contribute to the development of tumours [148]. This suggests an important function for ID in human diseases [149,150]. Consistent with these observations, ID is more common in the eukaryotic proteome than in the prokaryotic. Approximately 35-51% of eukaryotic proteins contain disordered regions of 40 or more consecutive residues relative to 6-33% of prokaryotic proteins. It was proposed that this is because eukaryotes utilise a more complex protein network for signalling and regulation [151,152].

IDPs possess characteristics that are distinct from those of folded globular proteins. Relative to globular proteins, the sequences of IDPs display a low complexity. Furthermore, IDPs are depleted of hydrophobic amino acids (e.g. Ile, Phe, Trp, Tyr) and hence enriched in polar and charged amino acids, and proline (e.g. Glu, Gly, Lys, Pro) [153]. The depletion of hydrophobic residues has an important consequence, it prevents the formation of a hydrophobic core. Hence, IDPs adopt rather elongated structures with increased hydrodynamic radii relative to globular proteins of similar molecular weight. The protein backbone further displays increased structural flexibility resulting in rapid interconversion between multiple conformers [146]. Multiple conformers are thought to arise as residues in disordered segments display a distribution of Ramachandran phi and psi angles. However, it was noted that IDPs tend to transiently adopt polyproline II helices in solution [154,155].

While some IDPs completely lack secondary structure, others exhibit partial secondary

structure that may be transient in nature. In most cases the residual secondary structure is of α-

helical nature predominating over β-strands [146]. Based on their structural properties such as

secondary structural content or compactness, IDPs can be classified into three non-exclusive groups: random coil, pre-molten globule and molten globule. The latter two groups are marked by varying degrees of fluctuating secondary structure and increasing compactness [145]. The Protein Quartet model proposes [145] that a protein may exist in any of the four thermodynamic states: ordered, molten globule, pre-molten globule and random coil [144]. The function of the protein may depend on any of the states or on a transition between two of the states [144,145]. Various examples for the Protein Quartet model have been reviewed in references [144-146] and some of the most prominent examples are given below. This list is not thought to be comprehensive, but to give an impression of the complexity of cellular processes IDPs are involved in.

Several cases have been described where the biological function of IDPs resides in the disordered state. Disordered segments, due to their inherent flexibility, have been described to serve as linkers between structured domains in multi-domain proteins such as p53 [156] or Calmodulin [157]. Another function of disordered segments involves the presentation of post- translational modification, or protease digestion sites as suggested for Bcl-XL and Bcl-2 [158-161]. Furthermore, disordered regions were demonstrated to function as entropic bristles necessary for the separation of neighbouring neurofilaments, or to serve as entropic springs such as Titin in muscle filaments [162]. Biologic functions for pre-molten globules, molten globules or transitions between either of the two have been described for various proteins; such as myelin

binding protein [163], calsequestrin [164] or 1,25-dihydroxyvitamin D3 receptor [165]; upon binding to metal ions or other non-protein metabolites.

A common feature for many IDPs is a disorder-to-order transition upon binding to their biologic targets, a process termed induced folding [152,166]. Examples for induced folding includes the binding of p21 to Cdk2 [146,167], formation of a functional ribosome by ribosomal proteins [168], or binding of transcription factors such as the Lac-Repressor to DNA [169,170]. However, binding to a biologic target does not necessarily induce structure and examples are known where ID regions retain disorder and a high degree of flexibility after binding. One such case has been described for p27 bound to the Cdk2-Cyclin A complex. Residual flexibility in some regions of p27 allow for phosphorylation of a buried tyrosine residue thus marking p27 for ubiquitination and proteasomal degradation [171].

These examples indicate that ID is more common then previously thought. It was therefore asked what functional advantages IDPs possess. (i) As disordered regions are solvent exposed they are easily accessible for post-translational modifications that help control function, locali- sation and turnover. (ii) Consistently, disordered segments are more susceptible to proteolysis, which may influence the lifetime of an IDP in the cell. Proteolytic degradation was shown to depend on the exposition of protease cut sites or certain regulatory sequences such as PEST sites. These sites may be buried upon complexation thus increasing the lifetime of a protein, whereas in the non-complexed state the protein is rapidly degraded. However, since protein degradation in vivo is a highly regulated process the generality of this concept has been challenged. (iii) Disordered regions confer increased structural plasticity overcoming steric restrictions that result in larger interaction surface areas per residue compared to folded globular proteins [166]. (iv) Furthermore, IDPs bind their targets with high specificity yet low affinity; (v) the latter increasing association/dissociation rates. The fly-casting mechanism has been proposed to describe how increased association rates are achieved [172]. Due to their extendedness disordered segments can sample a larger solution volume for interaction partners relative to globular proteins. Binding to the target surface is initially weak followed by folding to reel in the target thereby increasing the strength of the interaction. (vi) Another important functional advantage of IDPs is their ability to specifically bind to structurally different targets [144-146]. This promiscuous binding property sparked the idea that IDPs function as hubs in protein interaction networks [173]. However, recent studies have challenged the generality and validity of this concept [174,175]. As the number of studied IDP complexes is increasing, more definitive results can be expected.