Review
Understanding protein non-folding

https://doi.org/10.1016/j.bbapap.2010.01.017Get rights and content

Abstract

This review describes the family of intrinsically disordered proteins, members of which fail to form rigid 3-D structures under physiological conditions, either along their entire lengths or only in localized regions. Instead, these intriguing proteins/regions exist as dynamic ensembles within which atom positions and backbone Ramachandran angles exhibit extreme temporal fluctuations without specific equilibrium values. Many of these intrinsically disordered proteins are known to carry out important biological functions which, in fact, depend on the absence of a specific 3-D structure. The existence of such proteins does not fit the prevailing structure–function paradigm, which states that a unique 3-D structure is a prerequisite to function. Thus, the protein structure–function paradigm has to be expanded to include intrinsically disordered proteins and alternative relationships among protein sequence, structure, and function. This shift in the paradigm represents a major breakthrough for biochemistry, biophysics and molecular biology, as it opens new levels of understanding with regard to the complex life of proteins. This review will try to answer the following questions: how were intrinsically disordered proteins discovered? Why don't these proteins fold? What is so special about intrinsic disorder? What are the functional advantages of disordered proteins/regions? What is the functional repertoire of these proteins? What are the relationships between intrinsically disordered proteins and human diseases?

Introduction

Proteins are the major components of the living cell. They play crucial roles in the maintenance of life, and their dysfunctions are known to cause development of different pathological conditions. Although proteins possess an almost endless variety of biological functions, one class of them, known as enzymes, biological catalysts, attracted the major attention of researchers in the early days of protein science. A catalyst is a material or substance that speeds up a chemical or biochemical reaction. Without the catalyst, such a reaction would have occurred anyway but at a much slower rate. Importantly, the catalyst is never used up in the reaction — there is always the same amount at the start and the end of the reaction.

Historically, a long-standing belief has been that the specific functionality of a given protein is determined by its unique 3-D structure. The primary origin of this structure–function paradigm is the “lock and key” hypothesis formulated in 1894 by Emil Fischer to explain the astonishing specificity of the enzymatic hydrolysis of glucoside multimers by different types of similar enzymes, where one enzyme could hydrolyze α- but not β-glycosidic bonds, and another could hydrolyze β- but not α-glycosidic bonds [1]. Based on these observations Fischer [1] wrote (as translated in [2]) “To use a picture, I would like to say that enzyme and glucoside have to fit to each other like a lock and key in order to exert a chemical effect on each other.” In this analogy, the lock is the enzyme, the key-hole is the active site of enzyme, and the key is the substrate. Similar to the situation for which only the correctly shaped key opens a particular lock, it has been hypothesized that only the correctly shaped/sized substrate (key) could fit into the key-hole (active site) of the particular lock (enzyme).

For a long period of time, the validity of “lock and key” model and its associated sequence–structure–function paradigm was unquestioned, especially after the crystal structures of proteins started to be solved by X-ray diffraction. In fact, the first determined 3-D structure of an enzyme, lysozyme, for which a bound inhibitor was co-crystallized with the protein, immediately showed that the precise locations of certain amino acid side chains is almost certainly what facilitates catalysis [3]. Since the first reports on X-ray crystallographic structures at atomic resolution for myoglobin [4], [5] and lysozyme [3], more than 61,575 protein structures have been deposited into the Protein Data Bank [6] as of November 17, 2009, most of which have been determined by X-ray diffraction but also with a small percentage of which have been determined by the newer methods based on NMR spectroscopy. These structures, especially those determined by X-ray crystallography, seemed to continue to reinforce a static view of functional protein structure, with the enzyme active site being considered to be a rigid and sturdy lock, providing an exact fit to only one substrate (key).

In reality, not all proteins are structured throughout their entire lengths. Instead, many proteins are in fact highly flexible or structurally disordered, and dozens of examples of functional yet disordered regions have been reported based on X-ray structure determination studies or based on the characterization of protein structure by other biophysical techniques [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21]. For example, many proteins in the Protein Data Bank (PDB) have portions of their sequences missing from the determined structures (so-called missing electron density) [22], [23]. A common reason for missing electron density is that the unobserved atom, side chain, residue, or region fails to scatter X-rays coherently due to variation in position from one protein to the next, e.g. the unobserved atoms are flexible or disordered.

For several examples, comparison of the results of the X-ray and NMR analyses of the same protein has revealed that solution and crystal structures can be quite different, with solution structures being much more flexible [24], [25]. Evidently the interactions in the crystal lattice reduce protein flexibility and, in some cases, have even been observed to induce disorder-to-order transitions in functionally important protein regions. Thus, NMR spectroscopy is able to directly confirm the flexibility of protein segments that are missing in crystallographic experiments and can sometimes indicate flexible regions that have become rigid due to crystal contacts [26], [27], [28].

Unstructured proteins and unstructured regions can be characterized by a variety of additional biophysical and biochemical methods such as small angle X-ray scattering, Raman optical activity, circular dichroism, and protease sensitivity to name a few. Indeed, more than 20 different methods have been focused on disordered protein regions with each giving different pieces of information about the unstructured state [29], [30], [31]. When time and money permit, unstructured proteins should be studied by multiple biophysical methods in order to gain a fuller understanding of their characteristics [29], [30], [31].

Some proteins represent a big challenge for protein crystallographers because of their flexible and very dynamic nature. Myelin basic protein (MBP) exemplifies these troublemakers [32]. One exhaustive series of attempts to crystallize MBP for X-ray diffraction has been reported, where the authors tried 4600 different crystallization conditions but were unable to induce crystallization of MBP [33]. Based on these observations the myelin basic protein has been suggested to belong to the category on “uncrystallizable” proteins. It can be safely assumed that many other unsuccessful crystallization attempts for numerous other proteins have not been reported, since negative results are generally assumed to be unsuitable for publication. In the case of MBP, several additional studies suggest that this protein lacks fixed 3-D structure, existing instead as in intrinsically disordered ensemble, which in turn have been suggested to provide the basis for its multifunctionality [34]. Well-structured proteins often fail to crystallize, so not every crystallization failure should be ascribed to structural disorder. Nevertheless, we wonder how many crystallization failures denote these multifunctional yet unstructured proteins.

The importance of flexible structure for some proteins emerged from studies on protein folding. In fact, it has been pointed out that partially structured folding intermediates (such as the molten globule [35], [36], [37], [38], [39], [40], [41] and the pre-molten globule [37], [42], [43], [44]), which preserve some main elements of native secondary structure and their approximate mutual positions in 3-D space, but differ from the rigid globular state by looser packing of side chains and by the dramatic increase in the mobility of loops and ends of chain, are apparently ideal for some protein functions. The pre-molten globule is much more compact than the random coil but is less compact and has less secondary structure as compared to the molten globule (see below for additional discussion). By adjusting the solvent conditions, many proteins can be made to exist as stable, artificially induced, molten globules or as stable pre-molten globules, suggesting that these forms are not always transient folding intermediates [37], [38], [42], [43], [44], [45], [46].

Molten globule formation is likely driven by hydrophobic collapse, but with insufficiently tight side chain packing to form stable structures [38]. Pre-molten globule formation, on the other hand, evidently arises due to water being a poor solvent for polypeptides [42], [47], [48], [49] (see below for further discussion of this structural form). Recent studies on model homopolymer amino acids shed additional light on the concept that collapse can be driven by water not being a good solvent for proteins. Despite their lack of hydrophobic side chains, both polyglycine and polyglutamine form collapsed forms lacking appreciable secondary structure, likely because water is a poor solvent for both of these polymers [49], [50], [51], [52]. Given the hydrophilic nature of polyglutamine, these results suggest that collapse from water being a poor solvent is likely to be a general phenomenon for proteins that lack a significant net charge. Both of these homopolymers contain dynamic, fluctuating structures that involve rapidly exchanging hydrogen bonds. While these homopolymers and the pre-molten globule state may share the property of arising from water being a poor solvent, the latter form contains much more secondary structure than the former, probably due in part to the presence of hydrophobic side chains. Further comparisons of various model homopolymers with different pre-molten globule proteins are needed to better understand their similarities and differences.

Some proteins exist as stable molten globules or as stable pre-molten globules, suggesting that for these proteins such partially folded forms can be associated with function (e.g., see [53], [54], [55], [56], [57], [58]). Indeed, molten globules have been suggested to be involved in a number of physiological processes [37], [59], [60] such as interaction with chaperones [61], protein insertion into membranes [62], [63] and interaction with ligands (summarized in [64], [65]). Although functionality has been attributed to the molten globule- or pre-molten globule-like conformations for the examples cited above, the major emphasis still remains focused on the concept that these partially folded structures represent kinetic folding intermediates trapped by chaperones just after the protein biosynthesis but before proteins become completely folded [37], [59], [60], or appear as a result of point mutations preventing polypeptides from complete folding [37], [66]. Some other proteins (such as pore-forming domains of some toxins, or proteins that act as carriers of large hydrophobic ligands) were assumed to have originally a rigid structure but were forced somehow to denature to fulfill their functions [38], [60].

Many proteins with flexible structures have been discovered one-by-one. Some of these proteins were observed as atypical cases of polyfunctional proteins (e.g., serum albumin [67]), or polypeptides with unusual amino acid compositions (e.g., prothymosine α [7], [8], [9]), or proteins involved in the binding of large partners (RNA, DNA, proteins, and heme, e.g., histones [10], ribosomal proteins [11], myoglobin [12] and cytochrome c [13], [14]) or in the binding of large numbers of small partners (e.g., osteocalcin [15]). For some of these highly flexible proteins the increased conformational flexibility was even suggested to be of functional significance, with these data indicating that sometimes proteins do not need to be rigid to be functional.

From the 1980s onwards, a number of researchers pointed out that lack of structure or flexibility can be important for biological function. Huber and Bennett [16] pointed out that missing regions of electron density of several proteins likely carried out important functions. Several papers in the late 1980s (reviewed by Sigler [17]) suggested that several important transcription factors carry out function without specific structure, requiring instead the existence of rather ill-defined “acid blobs or negative noodles.” To describe the open and relatively mobile conformation of the caseins, which allows rapid and extensive degradation of these proteins to smaller peptides by proteolytic enzymes, Holt and Sawyer suggested the term “rheomorphic protein” (meaning flowing shape) [18] and proposed later that the rheomorphism of the casein phosphoproteins is important for the protection of the mammary gland against pathological calcification during lactation by allowing the protein to combine rapidly with nuclei of calcium phosphate to form stable calcium phosphate nanoclusters [19], [20]. In a similar time frame, Pontius extended his earlier work to suggest that unstructured proteins could have an advantage for certain types of molecular interactions [21]. Based on the observations that tau-protein in solution resembled a Gaussian polymer being characterized by the lack of detectable secondary structure and compact folded conformation, together with the facts that this protein exhibited the following properties: 1. a high conformational flexibility similar to that of denatured protein; 2. a high resistance to heat and acid treatment without losing its ability to promote microtubule formation; 3. a rod-like or highly extended appearance in the electron microscope; and 4. a binding of tau to microtubules that was not defined by clearly identifiable residues, but rather was distributed over many weakly interacting sites within the C-terminal half, tau was regarded as a “natively denatured” protein [68]. In a 1995 study, Gast et al. [8] pointed out that prothymosin α, an acidic protein with an unusual amino acid composition, is characterized by a high evolutionary conservation and wide tissue distribution, yet this protein adopts a random coil-like conformation under physiological conditions in vitro. These authors also raised an important question: “whether this is a rare or a hitherto-overlooked but widespread phenomenon in the field of macromolecular polypeptides?” [8]. A year later, similar conformational behavior was described for another biologically important protein, α-synuclein (also known as the non-Aβ component of Alzheimer's disease amyloid precursor protein, NACP), which was shown to possess high stability to heat denaturation, a highly charged amino acid sequence, a “random coil” structure as demonstrated by CD, an abnormally high Stokes radius, and an abnormal SDS binding leading to unusual mobility on SDS-PAGE [69]. The authors also have pointed out that since similar diagnostic properties were earlier reported for several other proteins, all of them should be combined in a new class of “natively unfolded proteins” [69].

Despite the significant number of important experimental results described for these unstructured proteins, the concept that these proteins form an important and novel structure–function class simply failed to take hold. Part of the problem apparently was that the information about flexible yet functional proteins was scattered in the literature, and so the concept of biological function originating from conformational flexibility was rediscovered many times and given many different names [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [53], [54], [55], [56], [57], [58], [67], [68], [69]. As a result, for a very long time each “non-traditional” protein with highly unusual structural properties and/or strange conformational behavior was typically considered to be a rare exception to the general rule that the function requires rigid 3-D structure. Also, these disordered proteins contradicted the widely-accepted protein structure–function paradigm. Perhaps especially due to this reason, the number of these proteins was assumed without evidence to be insignificantly small. Therefore, the tipping point for a concept change did not occur, and general questions about biological roles of disordered proteins were not being asked.

The situation has begun to change since the mid to late 1990s due significantly to the efforts of four research groups that came to the important conclusion that naturally flexible proteins, instead of being just rare exceptions, represent a very broad class of proteins [70], [71], [72], [73]. Interestingly, this important conclusion was reached at about the same time independently by four groups of investigators who emphasized rather different approaches, namely bioinformatics [72], [74], [75], [76], [77], [78], [79], [80], [81], [82], [83], [84], [85], [86], [87], [88], NMR spectroscopy [70], [89], [90], protein folding/misfolding [9], [64], [71], [91], [92], [93], [94], [95], and protein structural characterization [73]. The work of these four groups of course was strongly influenced by, and depended significantly upon, the many specific examples described by previous workers [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [53], [54], [55], [56], [57], [58], [67], [68], [69] but differed from previous efforts in that the lack of structure itself became the focus of attention with special efforts directed towards understanding the differences in function and mechanism between structured and unstructured proteins.

By now, many proteins have been shown to lack rigid 3-D structure under physiological conditions in vitro, existing instead as dynamic ensembles of interconverting structures. These proteins have been given various names including rheomorphic [18], intrinsically disordered [72], natively denatured [68], natively unfolded [69], [71], intrinsically unstructured [70], [73], mostly unstructured [55], and natively disordered [29]. Each of these terms has advantages and limitations. Not on this list of names is flexible, which is commonly applied to unstructured proteins but which evidently has not been suggested as a class name. Disordered proteins and regions are certainly highly flexible [96], but the word “flexible” has been used to describe many types of backbone and side chain mobility important for function [96], for example the motions in regions of high B-factor [97]. This general use of the word “flexible” does not make it such a good choice as a general descriptor for these ill-structured proteins. The term rheomorphic seems appropriate for extended random coils but perhaps not for molten globules, nor for collapsed random coils. The terms intrinsically unstructured and natively unfolded may be also be suitable for extended random coils and even those that are collapsed, but these terms don't seem to appropriately describe proteins that form transient or stable secondary structure. The term disorder suffers because of its negative connotation and its possible confusion with a pathological state, yet, on the other hand, disorder can be used for proteins like the molten globule that form substantial secondary structure but that nevertheless are highly dynamic and non-uniform. For this last reason, herein we will call these proteins “intrinsically disordered” (ID).

By “intrinsic disorder” we mean that the protein exists as a structural ensemble, either at the secondary or at the tertiary level. In other words, in contrast to structured or ordered proteins whose 3-D structure is relatively stable with Ramachandran angles that vary slightly around their equilibrium positions but with occasional cooperative conformational switches, intrinsically disordered proteins or regions exist as dynamic ensembles in which the atom positions and backbone Ramachandran angles vary significantly over time with no specific equilibrium values, and these ensembles typically undergo non-cooperative conformational changes. Both extended (random coil-like) regions with perhaps some secondary structure and collapsed (partially folded or molten globule-like and pre-molten globule-like) domains with poorly packed side chains are included in our view of intrinsic disorder [72], [92].

Since publication of key studies and reviews describing this new concept [53], [54], [55], [56], [57], [58], [70], [71], [72], [73], [98], [99], the literature on these proteins is virtually exploding (see Fig. 1). Bioinformatics studies indicate that about 25 to 30% of eukaryotic proteins are mostly disordered [100], that more than half of eukaryotic proteins have long regions of disorder [85], [100], and that more than 70% of signaling proteins have long disordered regions [101].

Now it is recognized that ID is a very abundant phenomenon. In fact, many proteins were shown to contain regions of disorder or even to be entirely disordered. Uversky et al. compiled a list of 91 disordered proteins characterized by NMR, circular dichroism or other biophysical techniques [71]. A subsequent search of X-ray crystal structures and the literature have expanded this list to more than 200 proteins that contain disordered regions of 30 consecutive residues or longer as characterized by X-ray crystallography, proteolytic digestion or other physical analyses such as NMR or circular dichroism [72]. The commonness of intrinsic disorder was estimated by predicting disorder for whole genomes, including both known and putative protein sequences (see below for the discussion of the disorder predictors). Such predictions have been published for 31 genomes that span the 3 kingdoms. The percentage of sequences in each genome with segments predicted to have ≥ 40 consecutive disordered residues was used to gain an overview of proteomic disorder. For so many consecutive predictions of disorder, the false-positive error rate was estimated from ordered proteins to be less than 0.5% of the segments of 40 and less than 6% of the fully ordered proteins [72], [85]. The eukaryotes exhibited more disorder by this measures than either the prokaryotes or the archaea, with C. elegans; A. thaliana; Saccharomyces cerevisiae; and D. melanogaster predicted to have 52–67% of their proteins with such long predicted regions of disorder, while bacteria and archaea were predicted to have 16–45% and 26–51% of their proteins with such long disorder regions, respectively [85], [100]. The increased amount of disorder in the eukaryota is very likely related to the increase in cellular signaling in the eukaryota [72], [85], [100]. The functional repertoire and advantages of intrinsic disorder will be discussed below.

Section snippets

Why ID proteins do not fold

Similar to the “normal” protein for which it has been shown that the correct folding into its relatively rigid biologically active conformation is determined by its amino acid sequence, the absence of rigid structure in the “non-traditional” ID proteins may also be somehow encoded in the specific features of their amino acid sequences. In fact, some of the ID proteins have been discovered due their unusual amino acid sequence compositions and the absence of regular structure in these proteins

Intrinsic disorder and alternative splicing

Alternative splicing (AS) is a process by which two or more mature mRNAs are produced from a single precursor pre-mRNA by the inclusion or omission of different segments [277], [278]. The “exons” are joined to form the mRNA and the “introns” are left out [279]. But so far, AS of mRNA has been commonly observed only in multicellular eukaryotes [280], including plants, apicompexans, diatoms, amoebae, animals and fungi (reviewed in [281]). Genes shared among animals, fungi and plants show high

Controlled chaos: on the tight regulation of ID proteins in the living cells

ID proteins are real, abundant, diversified, and vital. The functions of ID proteins are mostly complementary to the catalytic activities of ordered proteins [70], [71], [72], [80], [81], [82], [92], [93], [101], [108], [211], [213], [215], [216], [217], [226], [298], [299], [300]. Many disorder-related functions (e.g., signaling, control, regulation and recognition) appear to be incompatible with well-defined, stable 3-D structures [70], [71], [72], [73], [82], [92], [93], [108], [211], [213],

What is the relationship between ID proteins and human diseases?

Because ID proteins play crucial roles in numerous biological processes, many of these proteins are implicated in human disease. For example, several human diseases originate from the deposition of stable, ordered, filamentous protein aggregates, commonly referred to as amyloid fibrils. In each of these pathological states, a specific protein or protein fragment changes from its natural soluble form into insoluble fibrils, which accumulate in a variety of organs and tissues [315], [316], [317],

Conclusions

Intrinsically disordered (ID) proteins are widespread and represent a distinct protein tribe, with disorder being an important structural element that exists at various levels of protein structure. Such ID proteins are commonly involved in recognition, regulation and cell signaling functions and have biophysical characteristics that are well disposed for this role. They are much more common in eukaryota in comparison to prokaryota and archaea, reflecting the greater need for disorder-associated

Acknowledgements

We thank Zoran Obradovic, Chris Oldfield, Bin Xue, Pedro Romero, Marc Cortese, and Ya-Yue Van for their continuing support and collaborations in the field of the ID protein studies. This work was supported in part by the grants R01 LM007688-01A1 (to A.K.D and V.N.U.) and GM071714-01A2 (to A.K.D and V.N.U.) from the National Institutes of Health and the Program of the Russian Academy of Sciences for the “Molecular and cellular biology” (to V.N.U.). We gratefully acknowledge the support of the

References (419)

  • V.N. Uversky et al.

    Further evidence on the equilibrium “pre-molten globule state”: four-state guanidinium chloride-induced unfolding of carbonic anhydrase B at low temperature

    J. Mol. Biol.

    (1996)
  • A. Vitalis et al.

    Atomistic simulations of the effects of polyglutamine chain length and solvent quality on conformational equilibria and spontaneous homodimerization

    J. Mol. Biol.

    (2008)
  • A. Vitalis et al.

    Quantitative characterization of intrinsic disorder in polyglutamine: insights from analysis based on polymer theories

    Biophys. J.

    (2007)
  • H. Lee et al.

    Local structural elements in the mostly unstructured transcriptional activation domain of human p53

    J. Biol. Chem.

    (2000)
  • J.A. Zitzewitz et al.

    Preformed secondary structure drives the association reaction of GCN4-p1, a model coiled-coil system

    J. Mol. Biol.

    (2000)
  • V.E. Bychkova et al.

    The ‘molten globule’ state is involved in the translocation of proteins across membranes?

    FEBS Lett.

    (1988)
  • F.G. van der Goot et al.

    The molten globule intermediate for protein insertion or translocation through membranes

    Trends Cell Biol.

    (1992)
  • V.E. Bychkova et al.

    Folding intermediates are involved in genetic diseases?

    FEBS Lett.

    (1995)
  • O. Schweers et al.

    Structural studies of tau protein and Alzheimer paired helical filaments show no evidence for beta-structure

    J. Biol. Chem.

    (1994)
  • P.E. Wright et al.

    Intrinsically unstructured proteins: re-assessing the protein structure–function paradigm

    J. Mol. Biol.

    (1999)
  • A.K. Dunker et al.

    Intrinsically disordered protein

    J. Mol. Graph Model.

    (2001)
  • P. Tompa

    Intrinsically unstructured proteins

    Trends Biochem. Sci.

    (2002)
  • H.J. Dyson et al.

    Coupling of folding and binding for unstructured proteins

    Curr. Opin. Struct. Biol.

    (2002)
  • V.N. Uversky et al.

    Zn(2+)-mediated structure formation and compaction of the “natively unfolded” human prothymosin alpha

    Biochem. Biophys. Res. Commun.

    (2000)
  • V.N. Uversky et al.

    Evidence for a partially folded intermediate in alpha-synuclein fibril formation

    J. Biol. Chem.

    (2001)
  • E. Fischer

    Einfluss der configuration auf die wirkung der enzyme

    Ber. Dt. Chem. Ges.

    (1894)
  • U.R. Lemieux et al.

    How Emil Fischer was led to the lock and key concept for enzyme specificity

    Adv. Carbohydr. Chem. Biochem.

    (1994)
  • C.C. Blake et al.

    Structure of hen egg-white lysozyme. A three-dimensional Fourier synthesis at 2 Angstrom resolution

    Nature

    (1965)
  • J.C. Kendrew et al.

    A three-dimensional model of the myoglobin molecule obtained by X-ray analysis

    Nature

    (1958)
  • J.C. Kendrew et al.

    Structure of myoglobin: a three-dimensional Fourier synthesis at 2 Å resolution

    Nature

    (1960)
  • J.D. Watts et al.

    Thymosins: both nuclear and cytoplasmic proteins

    Eur. J. Biochem.

    (1990)
  • K. Gast et al.

    Prothymosin alpha: a biologically active protein with random coil conformation

    Biochemistry

    (1995)
  • V.N. Uversky et al.

    Natively unfolded human prothymosin alpha adopts partially folded collapsed conformation at acidic pH

    Biochemistry

    (1999)
  • M. Boublik et al.

    An investigation of the conformational changes of histone F2b by high resolution nuclear magnetic resonance

    Eur. J. Biochem.

    (1970)
  • S.Y. Venyaminov et al.

    Absorption and Circular Dichroism Spectra of Individual Proteins from Escherichia coli Ribosomes

    (1981)
  • D.T. Isbell et al.

    Metal ion binding to dog osteocalcin studied by 1H NMR spectroscopy

    Biochemistry

    (1993)
  • R. Huber et al.

    Functional significance of flexibility in proteins

    Biopolymers

    (1983)
  • P.B. Sigler

    Transcriptional activation. Acid blobs and negative noodles

    Nature

    (1988)
  • C. Holt et al.

    Caseins as rheomorphic proteins: interpretation of primary and secondary structures of the αS1-, β- and κ-caseins

    J. Chem. Soc. Faraday Trans.

    (1993)
  • C. Holt et al.

    Ability of a beta-casein phosphopeptide to modulate the precipitation of calcium phosphate by forming amorphous dicalcium phosphate nanoclusters

    Biochem. J.

    (1996)
  • C. Holt et al.

    A core-shell model of calcium phosphate nanoclusters stabilized by beta-casein phosphopeptides, derived from sedimentation equilibrium and small-angle X-ray and neutron-scattering measurements

    Eur. J. Biochem.

    (1998)
  • A.C. Bloomer et al.

    Protein disk of tobacco mosaic virus at 2.8 Å resolution showing the interactions within and between subunits

    Nature

    (1978)
  • L.Y. Lian

    NMR structural studies of glutathione S-transferase

    Cell. Mol. Life Sci.

    (1998)
  • D.J. Sidote et al.

    NMR structure of an archaeal homologue of ribonuclease P protein Rpp29

    Biochemistry

    (2003)
  • H. Mizutani et al.

    Systematic study on crystal-contact engineering of diphthine synthase: influence of mutations at crystal-packing regions on X-ray diffraction quality

    Acta Crystallogr. D Biol. Crystallogr.

    (2008)
  • B. Kobe et al.

    Crystallography and protein–protein interactions: biological interfaces and crystal contacts

    Biochem. Soc. Trans.

    (2008)
  • R.P. Bahadur et al.

    The interface of protein–protein complexes: analysis of contacts and prediction of interactions

    Cell. Mol. Life Sci.

    (2008)
  • G.W. Daughdrill et al.

    Natively disordered proteins

  • V. Receveur-Brechot et al.

    Assessing protein disorder and induced folding

    Proteins

    (2006)
  • Cited by (992)

    • Functional unfoldomics: Roles of intrinsic disorder in protein (multi)functionality

      2024, Advances in Protein Chemistry and Structural Biology
    • Detecting anomalous proteins using deep representations

      2024, NAR Genomics and Bioinformatics
    View all citing articles on Scopus
    View full text