Thermo-resistant intrinsically disordered proteins are efficient 20S proteasome substrates

Abstract

Based on software prediction, intrinsically disordered proteins (IDPs) are widely represented in animal cells where they play important instructive roles. Despite the predictive power of the available software programs we nevertheless need simple experimental tools to validate the predictions. IDPs were reported to be preferentially thermo-resistant and also are susceptible to degradation by the 20S proteasome. Analysis of a set of proteins revealed that thermo-resistant proteins are preferred 20S proteasome substrates. Positive correlations are evident between the percent of protein disorder and the level of thermal stability and 20S proteasomal susceptibility. The data obtained from these two assays do not fully overlap but in combination provide a more reliable experimental IDP definition. The correlation was more significant when the IUPred was used as the IDPs predicting software. We demonstrate in this work a simple experimental strategy to improve IDPs identification.

---

1. Introduction

It has been suggested that many regulatory processes employ protein components that lack a fixed structure, enabling them to engage in multiple interactions and executing distinct functions in the cell.^1–4 Thus over the years there is an increasing appreciation for the role of intrinsically disordered proteins (IDPs) in a multitude of biological processes, specifically in cellular signaling and regulation.⁵ IDPs as part of key regulator pathways are often involved in pathologies such as cancer and neurodegenerative disorders.^6,7 Therefore, there is a growing need to understand the biological attributes of the IDPs and their function under different cellular experimental conditions. Although there has been an increase in the abundance and accuracy of structural prediction tools and in vitro structural methods, there is little progress in understanding IDP regulation in the cellular context. The majority of intrinsic disorder prediction tools, including those utilized in this study, rely on the amino acid composition of the IDP and their chemical properties to characterize the degrees of order/disorder in a protein of interest.^8–13 However, the methods utilized to calculate intrinsic disorder vary significantly between the different predictors. Some of the algorithms predict disorder based on the intrinsic attributes of the amino acids and the calculation of the mean hydropathy and net charge (Foldindex),^12,14,15 others calculate the interresidue interaction energy (IUPred).¹⁰ Moreover, there are the machine learning predictors that were trained on datasets of disorder and order that evaluate intrinsic disorder on a per-residue basis (RONN).¹³ The different prediction methodologies are nicely reviewed by Orosz and Ovadi.¹⁶

In the context of the cell it is expected that attributes other than amino acid composition and sequence can further affect the order/disorder of a protein. These include post-translational modifications but also protein–protein interactions that can induce folding and acquisition of a specific structure for the IDP.^1,4 As a complementary approach to the bioinformatic predictions, proteomic analysis has to be performed to determine the disordered nature of the cellular proteins. As an early step toward this aim we performed a small-scale analysis to compare some of the available predictions with simple experimental tools that can be further developed in the future for large scale proteomic analyses.

The lack of a defined tertiary structure of IDPs previously led us to suggest that the degradation of a protein by the cellular proteasomal machinery, the 20S proteasome, can be utilized as an operational definition of an IDP.¹⁷ IDPs in the free form will be degraded by the 20S proteasomein vitro and in cells as shown for p53 and other IDPs.^17,18 Once the disordered domains are masked by the associated binding partners (such as Hdmx in the case of p53) their susceptibility to 20S proteasomal degradation is severely compromised.¹⁸ Thus the binding protein of the IDP was termed a “nanny”, ensuring the survival and functionality of the IDP.¹⁸ The susceptibility to 20S proteasomal degradation was analyzed on several proteins but no large-scale proteomic attempt was made.

Several strategies were developed for enrichment of IDPs in the whole cellular proteome in E. coli, yeast and mammalian cells.^19–22 Enrichment of IDPs in E. coliprotein extract was achieved by exploiting the attribute that highly disordered proteins are resistant to acidic precipitation.¹⁹ In yeast, utilizing thermal stability followed by 2D electrophoresis analysis enabled the characterization and enrichment of IDPs.²⁰ Thermal stability properties of IDPs were further utilized to enrich for IDPs from mammalian cells, however, approximately one third of the thermo stable proteins were predicted to be completely ordered,^21,22 suggesting that thermal stability is not an exclusive property of IDPs. There are several highly structured proteins such as ubiquitin,²³ and other proteins that are structured under very high temperatures as in the case of thermophilic organisms. The best exploited example is the thermo stable DNA polymerase purified from the thermophilic bacterium, Thermus aquaticus (Taq).²⁴ Moreover, It has been proposed that heat soluble proteins can be categorized into four distinct groups, with only one being IDPs.²⁵ Given our previous operational definition of IDPs as the susceptibility to 20S proteasomal degradation, here we undertake to examine the susceptibility of thermo stable proteins from human cells, and the possible implications of the combination of the two methodologies for proteomic IDP research.

2. Experimental procedures

Cells

HeLa cells were cultured in DMEM and 10% FBS at 37 °C in a humidified incubator with 5.6% CO₂.

Protein extraction from cells

Protein extraction from different cells was performed by resuspending cells in NP40 buffer as previously described.²⁶ The extract was subjected to ultracentrifugation (13 000×g, 15 min) and the supernatant was used as the protein extract. Protein concentrations were determined by Bradford assay (Bio-Rad).

In vitro protein translation

In vitro translation was performed with the TNT quick-coupled transcription/translation system (Promega) according to the manufacturer's instructions. Plasmids used for in vitrotranslation were pCMV p53, pcDNA3.1 YAP1, TAZ, c-Jun, tau, CyclinD (kindly provided by Prof. Chaim Kahana), WW45, c-FOS, PCNA and PEFires p21.

Heat solubility assay

Purified or in vitro translated proteins were diluted in buffer A [50 mM Tris HCl, 150 mM NaCl, 5 mM MgCl₂, 2 mM DTT] to reach a final volume of 30 μl. Samples were subjected to heat treatment by incubation in a 95 °C water bath for 5 minutes or the indicated time points. Following the heat treatment the soluble fraction was separated by ultracentrifugation 13 000 rpm at 4 °C for 15 min. The supernatant is referred to as the soluble faction whereas the pellet as the insoluble fraction. The pellet was resuspended in Laemmli sample buffer [4% SDS, 20% glycerol, 10% 2-mercaptoethanol and 0.125 M Tris-HCl] and Laemmli sample buffer was added to the soluble fraction. The samples were heated at 95 °C for 5 minutes and loaded on a 12.5% polyacrylamide-SDS gel. In vitro translated ³⁵S methionine labeled proteins were detected by autoradiography.

Prediction of disorder

Prediction of disorder was done by Foldindex¹² (http://bip.weizmann.ac.il/fldbin/findex), IUPred¹⁰ (http://iupred.enzim.hu) and RONN¹³ (http://www.strubi.ox.ac.uk/RONN). We used default parameters with the protein disorder prediction servers. We believe that authors of each of the servers provided what they consider the best general conditions for a prediction, receiving in this way comparable results from the different servers.

Subcellular fractionation and protein extraction

Cell pellets were resuspended in Cyt buffer [50 mM Tris HCl pH 7.5, 5 mM MgCl₂, Protease inhibitor cocktail (Sigma), 2 mM DTT] and homogenized with a Teflon homogenizer. The nuclear pellet was pelleted by centrifugation at 1000×g for 5 min. The supernatant was regarded as the cytosolic fraction. The nuclear pellet was resuspended in Nuc buffer [50 mM Tris pH 7.5, 450 mM NaCl, 5 mM MgCl₂, Protease inhibitor cocktail (Sigma), 2 mM DTT] and sonicated at 5 intervals of 10 s. The cytosolic and nuclear fractions were further centrifuged at 13 000 rpm 4 °C for 15 min and the supernatant was taken for experimental analysis.

20S proteasomal degradation assay

In vitro translated ³⁵S Methionine labeled proteins were incubated in buffer A in the presence or absence of 1 μg of highly purified 20S proteasomes from rat liver and incubated at 37 °C for 60 minutes. For the degradation of cellular proteins, the cytoplasmic or nuclear fractions were dialyzed against buffer A overnight. 50 μg of protein extract were incubated in the presence or absence of 1 μg purified 20S proteasomes at 37 °C for 60 minutes. In all cases the degradation reaction was stopped by the addition of Laemmli sample buffer (final concentration: 4% SDS, 20% glycerol, 10% 2-mercaptoethanol and 0.125 M Tris-HC), and heating at 95 °C for 5 minutes. The samples were then analyzed by SDS-PAGE and autoradiography.

3. Results

Some IDPs are heat resistant

The loose structure of IDPs is predicted to make them resistant to severe heat shock where globular proteins will aggregate.^22,25 To examine the possible mechanisms and implications of heat resistance we utilized an in vitro reductive system whereby proteins of interest can be in vitro³⁵S methionine labeled for detection. A subset of proteins with predicted differential disorder was subjected to 95 °C for 5 minutes followed by centrifugation that separated the aggregates or insoluble fraction, representing the thermo-sensitive population, from the soluble proteins, representing the thermo-resistant fraction. As expected, different proteins displayed different levels of heat sensitivity (Fig. 1A). Whereas well-defined structured proteins such as CyclinD and NQO1 were aggregated, other well-defined IDPs such as tau remained soluble. To test the correlation we plotted the predicted percentage of disorder as a function of solubility. The plot shows that all the heat soluble proteins are highly disordered. However, some of the highly disordered proteins (such as p53) are not heat soluble (Fig. 1B). Utilizing different prediction methodologies (IUPred, Foldindex and RONN)^10,12,13 we correlated the predicted disorder with the experimental results. Surprisingly, the degree of correlation markedly varied. The square correlation coefficient (R²) of IUPred was the highest, R² = 0.63 (Fig. 1B). Foldindex exhibited the lowest level of R², R² = 0.27 (Fig. 1C) and RONN was somewhere in between, R² = 0.40 (Fig. 1D).

Fig. 1 Thermo stability of IDPs. (A) In vitro³⁵S Methionine labeled proteins of interest were incubated at 95 °C for 5 minutes. Total, soluble and insoluble fractions were analyzed by SDS-PAGE and detected by autoradiography. (B) The calculated percentage of the soluble fraction for each protein was plotted against the calculated percentage of predicted disorder by IUPred (B), Foldindex (C) and RONN (D).

Susceptibility of IDPs to 20S proteasomal degradation and the correlation with heat resistance

We have previously defined the susceptibility of a protein to 20S proteasomal degradation as an IDPs operational definition.¹⁷ Accordingly, here we subjected the in vitro translated [³⁵S] Methionine labeled proteins utilized above to degradation by the purified 20S proteasomes to evaluate the extent of their disorder. Following incubation at 37 °C for 60 minutes in the presence or absence of purified 20S proteasomes the levels of proteins were quantified (Fig. 2A). As expected, the predicted highly disordered proteins such as tau were markedly susceptible to 20S proteasomal degradation (Fig. 2A). In these analyses the 20S proteasome was partially active in digesting structured proteins as well, therefore, to differentiate between the different substrates we looked at the range of 70–100% digestion. Remarkably, plotting the percent of digestionversus the percent of disorder as predicted by the IUPred program revealed a linear regression line with the square correlation coefficient R² = 0.67 (Fig. 2B), which is very similar to that obtained with the thermal resistance regression line (Fig. 1B). Here again the R² value obtained with the FoldIndex program versus 20S proteasomal percent of degradation (Fig. 2C) is low (R² = 0.22), but similar to that obtained with thermal stability (Fig. 1C). In contrast, the obtained R² value by the RONN prediction program was much closer (R² = 0.61) to that obtained with IUPred (Fig. 2D).

Fig. 2 Susceptibility of IDPs to 20S proteasomal degradation. (A) In vitro³⁵S Methionine labeled proteins of interest were incubated in the presence or absence of purified 20S proteasomes at 37 °C for 60 minutes. The levels of 20S proteasomal degradation were analyzed by SDS-PAGE and detected by autoradiography. The correlation between 20S proteasomal degradation (x-axis) and the predicted disorder by IUPred (B), Foldindex (C) and RONN (D) was plotted.

Thermo behavior and 20S proteasomal degradation

Examining the percentage of degradation of the predicted IDPs versus their thermo-stability revealed that the most highly degraded IDPs were also highly soluble following heat treatment (Fig. 3A). These included tau the well-established IDPs and some less characterized proteins such as Yap1, Taz and c-Jun. These results suggest that the combination of the two experimental strategies, heat resistance and susceptibility to 20S proteasomal degradation could be a useful approach for IDP characterization. To challenge this prediction we utilized both cellular cytosolic and nuclear extracts, the latter known to contain higher number of IDPs. Indeed the thermo-resistant nuclear fraction was much higher as compared to the cytosolic extract (Fig. 3B). Next we subjected equal amounts of thermo soluble extracts to 20S proteasomal digestion. The data revealed that the vast majority of thermo resistant proteins are efficient 20S proteasome substrates, in particular when using nuclear proteins (Fig. 3C).

Fig. 3 Thermo behavior and 20S proteasomal degradation. (A) The correlation between the solubility of the protein (y-axis) and the susceptibility to 20S proteasome degradation (x-axis) was plotted. (B) Thermo stability of cytoplasmatic protein extract (left) and nuclear (right) protein extract was examined by incubating the cellular extracts for indicated time points at 95 °C. The soluble fraction was analyzed by silver staining of SDS-PAGE gels. (C) Thermo stable (5 minutes at 95 °C) cytoplasmic and nuclear fractions were incubated in the presence or absence of purified 20S proteasomes for 1 hour at 37 °C. Total protein was detected by silver staining of SDS-PAGE gels.

Thermo behavior of protein complexes

We have reported that IDP interacting proteins protect IDPs from degradation by the 20S proteasome, a behavior that we termed nanny.²⁷ The question is whether nannies also affect the thermo stability behavior of IDPs. To this end we used the Jun–Fos complex previously reported to behave as a nanny partner in the context of 20S proteasomal digestion.²⁸ In the absence of c-Fos, a major fraction of c-Jun (45%) was soluble. Unexpectedly, c-Fos when tested alone was lost by large upon heating, possibly due to its random cleavage under this condition. However, the residual intact c-Fos was found in the insoluble fraction (Fig. 4). In the presence of c-Fos, a fraction of c-Jun became thermo-sensitive as revealed by the decrease in the soluble fraction of c-Jun (from 45% to 26%). Moreover, in the presence of c-Jun there was an elevation of insoluble c-Fos (from 11 to 23%). These results suggest that the thermo behavior of complexes of two IDP candidates might be different from their individual components. Taken together and as previously suggested,^21,22 heat resistance might be an intrinsic property of some IDPs in their free form.

Fig. 4 Thermo behavior of protein complexes. In vitro³⁵S Methionine labeled c-Jun and c-Fos were incubated at 95 °C for 5 minutes alone or mixed together. The levels of total, insoluble and soluble fraction were analyzed by autoradiography and the percentage of soluble and insoluble fraction was calculated as the percent of total. In the c-Jun and c-Fos mixed lanes the total % was normalized to 100% when calculating the soluble and insoluble fractions (3, 6 and 9).

Discussion

In this work we examined two simple experimental strategies; thermo stability and susceptibility to 20S proteasomal degradation. IDPs tend to be both more thermo-stable^21,22 and susceptible to 20S proteasomal degradation.^17,29–32 Examining these two properties on a select group of proteins with varying degrees of disorder revealed that these two behaviors do not fully overlap but in combination are expected to provide a more reliable experimental IDP definition. We found that a thermo resistant protein is very likely to be a good 20S proteasome substrate, but not all the 20S substrates are thermo-resistant.

An important question is to what extent the employed experimental approaches improve the operational definition of the IDPs. To address this issue we looked for correlations between the obtained experimental data and the extent of disorder as predicted by three different software programs. Regarding the 20S proteasomal degradation, we got a quite good correlation while using the IUPred software that is based on interresidue interaction energy calculation and the machine learning predictors RONN software. In the context of thermo behavior the correlation was good only with the IUPred. We are not at the position to evaluate the software predictability, but our experimental data fit the IUPred program much better. The lack of correlation might suggest that on one hand the experimental strategies are not fully correlated with the disorder of proteins and might encompass a different protein property. On the other hand it might imply that disordered proteins can come in many different flavors^34,35 and that different predictions or experimental strategies catch unique populations that have not yet been fully characterized. Given our results that the thermo-resistant proteins are also highly disordered we believe that the second option is more reasonable to explain the correlations.

Although IDPs are expected and were shown to be thermo-stable,^21,36 it is intriguing why several examined IDPs did not exhibit thermo-stability in our work. One possible explanation is that certain IDPs might have an extended folded domain that aggregates upon heat exposure. This might be the case for p53, which exhibits a high degree of disorder at both termini but a folded domain in its core.³⁷ Here we show that p53 shows little thermo-stability, suggesting that the folded domain might be responsible for the aggregation. Moreover, the thermo-stability of IDPs can be affected by the interacting partners at a given time point. The association of an IDP with an aggregating complex can possibly induce the aggregation of the IDP itself. Although some IDPs such as p27 were shown to dissociate from their functional complex and remain soluble following heat exposure,²² other IDPs can possibly remained bound to the aggregating complex (as we see with c-Jun). The possibility that association of IDPs with protein complexes affects their thermo-stability might be further utilized in a proteomic approach to identify the IDPs that are normally not found in their unbound, disordered state in the cells. In that case the induction of dissociation of protein complexes by high salt concentration or detergents may increase the thermo-stability of these IDPs. We have preliminary evidence that this might be the case in whole cell extracts (data not shown) and we hope that additional research could lead to the optimization of a protocol that could be utilized to identify and characterize novel IDPs.

Although in this work we utilized a limited number of proteins, we were able nevertheless to correlate the experimental data with the predicting power of certain software tools. The lack of correlation could be possibly utilized to both examine and improve algorithm predictability on one hand and to improve the experimental protocols on the other hand. The combination of the two experimental strategies may result in a reduced amount of false positive results when analyzing IDPs. The proteins that are thermostable due to increased order, as observed in thermophilic organisms,³³ are not expected to be degraded by the 20S proteasome. The thermo-stability assay on the other hand enables to exclude endogenous cellular processes that can induce protein cleavage (such as endogenous proteasomes and other peptidases) enabling to analyze whole proteome samples without prior fractionation. We believe that in the future the combination of the two distinct experimental strategies, thermo-stability and 20S proteasome susceptibility, could be employed as a proteomic approach in identifying IDPs in the cellular context.

Acknowledgements

We thank N. Reuven and J. Adler for their constructive advices. This study was supported by a grant from The Israel Science Foundation (grant No. 926/07). Y.S. is the Oscar and Emma Getz Professor.

References

1. A. K. Dunker, C. J. Brown, J. D. Lawson, L. M. Iakoucheva and Z. Obradovic, Intrinsic disorder and protein function, Biochemistry, 2002, 41, 6573–6582.
2. A. K. Dunker, Z. Obradovic, P. Romero, E. C. Garner and C. J. Brown, Intrinsic protein disorder in complete genomes, Genome. Inform. Ser. Workshop Genome. Inform., 2000, 11, 161–171.
3. P. Tompa, Intrinsically unstructured proteins, Trends Biochem. Sci., 2002, 27, 527–533.
4. P. E. Wright and H. J. Dyson, Intrinsically unstructured proteins: re-assessing the protein structure–function paradigm, J. Mol. Biol., 1999, 293, 321–331.
5. V. N. Uversky, C. J. Oldfield and A. K. Dunker, Intrinsically disordered proteins in human diseases: introducing the D2 concept, Annu. Rev. Biophys., 2008, 37, 215–246.
6. L. M. Iakoucheva, C. J. Brown, J. D. Lawson, Z. Obradovic and A. K. Dunker, Intrinsic disorder in cell-signaling and cancer-associated proteins, J. Mol. Biol., 2002, 323, 573–584.
7. V. N. Uversky, Intrinsic disorder in proteins associated with neurodegenerative diseases, Front Biosci., 2009, 14, 5188–5238.
8. P. Romero, Z. Obradovic, X. Li, E. C. Garner, C. J. Brown and A. K. Dunker, Sequence complexity of disordered protein, Proteins, 2001, 42, 38–48.
9. R. Linding, R. B. Russell, V. Neduva and T. J. Gibson, GlobPlot: exploring protein sequences for globularity and disorder, Nucleic Acids Res., 2003, 31, 3701–3708.
10. Z. Dosztanyi, V. Csizmok, P. Tompa and I. Simon, IUPred: web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content, Bioinformatics, 2005, 21, 3433–3434.
11. A. Schlessinger, M. Punta and B. Rost, Natively unstructured regions in proteins identified from contact predictions, Bioinformatics, 2007, 23, 2376–2384.
12. J. Prilusky, C. E. Felder and T. Zeev-Ben-Mordehai, et al. FoldIndex: a simple tool to predict whether a given protein sequence is intrinsically unfolded, Bioinformatics, 2005, 21, 3435–3438.
13. Z. R. Yang, R. Thomson, P. McNeil and R. M. Esnouf, RONN: the bio-basis function neural network technique applied to the detection of natively disordered regions in proteins, Bioinformatics, 2005, 21, 3369–3376.
14. V. N. Uversky, Natively unfolded proteins: a point where biology waits for physics, Protein Sci., 2002, 11, 739–756.
15. V. N. Uversky, J. R. Gillespie and A. L. Fink, Why are “natively unfolded” proteins unstructured under physiologic conditions?, Proteins, 2000, 41, 415–427.
16. F. Orosz and J. Ovadi, Proteins without 3D structure: definition, detection and beyond, Bioinformatics, 2011, 27, 1449–1454.
17. P. Tsvetkov, G. Asher and A. Paz, et al. Operational definition of intrinsically unstructured protein sequences based on susceptibility to the 20S proteasome, Proteins, 2008, 70, 1357–1366.
18. P. Tsvetkov, N. Reuven, C. Prives and Y. Shaul, Susceptibility of p53 unstructured N terminus to 20 S proteasomal degradation programs the stress response, J. Biol. Chem., 2009, 284, 26234–26242.
19. M. S. Cortese, J. P. Baird, V. N. Uversky and A. K. Dunker, Uncovering the unfoldome: enriching cell extracts for unstructured proteins by acid treatment, J. Proteome Res., 2005, 4, 1610–1618.
20. V. Csizmok, E. Szollosi, P. Friedrich and P. Tompa, A novel two-dimensional electrophoresis technique for the identification of intrinsically unstructured proteins, Mol. Cell Proteomics, 2006, 5, 265–273.
21. C. A. Galea, A. A. High and J. C. Obenauer, et al. Large-scale analysis of thermostable, mammalian proteins provides insights into the intrinsically disordered proteome, J. Proteome Res., 2009, 8, 211–226.
22. C. A. Galea, V. R. Pagala, J. C. Obenauer, C. G. Park, C. A. Slaughter and R. W. Kriwacki, Proteomic studies of the intrinsically unstructured mammalian proteome, J. Proteome Res., 2006, 5, 2839–2848.
23. A. Ciehanover, Y. Hod and A. Hershko, A heat-stable polypeptide component of an ATP-dependent proteolytic system from reticulocytes, Biochem. Biophys. Res. Commun., 1978, 81, 1100–1105.
24. R. K. Saiki, D. H. Gelfand and S. Stoffel, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase, Science, 1988, 239, 487–491.
25. T. D. Kim, H. J. Ryu, H. I. Cho, C. H. Yang and J. Kim, Thermal behavior of proteins: heat-resistant proteins and their heat-induced secondary structural changes, Biochemistry, 2000, 39, 14839–14846.
26. P. Tsvetkov, Y. Adamovich, E. Elliott and Y. Shaul, E3 ligase STUB1/CHIP regulates NAD(P)H:quinone oxidoreductase 1 (NQO1) accumulation in aged brain, a process impaired in certain Alzheimer disease patients, J. Biol. Chem., 2011, 286, 8839–8845.
27. P. Tsvetkov, N. Reuven and Y. Shaul, The nanny model for IDPs, Nat. Chem. Biol., 2009, 5, 778–781.
28. J. Adler, N. Reuven, C. Kahana and Y. Shaul, c-Fos Proteasomal Degradation by Default and its Regulation by NQO1 Determines c-Fos Serum Response Kinetics, Mol. Cell Biol., 2010, 30, 3767–3778.
29. D. P. Stewart, B. Koss, M. Bathina, R. M. Perciavalle, K. Bisanz and J. T. Opferman, Ubiquitin-independent degradation of antiapoptotic MCL-1, Mol Cell Biol., 2010, 30, 3099–3110.
30. C. W. Liu, M. J. Corboy, G. N. DeMartino and P. J. Thomas, Endoproteolytic activity of the proteasome, Science, 2003, 299, 408–411.
31. Y. Shaul, P. Tsvetkov and N. Reuven, IDPs and Protein Degradation in the Cell. Instrumental Analysis of Intrinsically Disordered Proteins, John Wiley & Sons, Inc., 2010, pp. 1–36.
32. C. M. Wiggins, P. Tsvetkov and M. Johnson, et al. BIM(EL), an intrinsically disordered protein, is degraded by 20S proteasomes in the absence of poly-ubiquitylation, J. Cell Sci., 2011, 124, 969–977.
33. P. V. Burra, L. Kalmar and P. Tompa, Reduction in structural disorder and functional complexity in the thermal adaptation of prokaryotes, PLoS One, 2010, 5, e12069.
34. A. Schlessinger, C. Schaefer, E. Vicedo, M. Schmidberger, M. Punta and B. Rost, Protein disorder—a breakthrough invention of evolution?, Curr. Opin. Struct. Biol., 2011, 21, 412–418.
35. S. Vucetic, C. J. Brown, A. K. Dunker and Z. Obradovic, Flavors of protein disorder, Proteins, 2003, 52, 573–584.
36. C. A. Galea, A. Nourse, Y. Wang, S. G. Sivakolundu, W. T. Heller and R. W. Kriwacki, Role of intrinsic flexibility in signal transduction mediated by the cell cycle regulator, p27 Kip1, J. Mol. Biol., 2008, 376, 827–838.
37. S. Bell, C. Klein, L. Muller, S. Hansen and J. Buchner, p53 contains large unstructured regions in its native state, J. Mol. Biol., 2002, 322, 917–927.

---

Footnote

† Published as part of a Molecular BioSystems themed issue on Intrinsically Disordered Proteins: Guest Editor M. Madan Babu.

---