Lighting up Nobel Prize-winning studies with protein intrinsic disorder

Abstract

Intrinsically disordered proteins and regions (IDPs and IDRs) and their importance in biology are becoming increasingly recognized in biology, biochemistry, molecular biology and chemistry textbooks, as well as in current protein science and structural biology curricula. We argue that the sequence → dynamic conformational ensemble → function principle is of equal importance as the classical sequence → structure → function paradigm. To highlight this point, we describe the IDPs and/or IDRs behind the discoveries associated with 17 Nobel Prizes, 11 in Physiology or Medicine and 6 in Chemistry. The Nobel Laureates themselves did not always mention that the proteins underlying the phenomena investigated in their award-winning studies are in fact IDPs or contain IDRs. In several cases, IDP- or IDR-based molecular functions have been elucidated, while in other instances, it is recognized that the respective protein(s) contain IDRs, but the specific IDR-based molecular functions have yet to be determined. To highlight the importance of IDPs and IDRs as general principle in biology, we present here illustrative examples of IDPs/IDRs in Nobel Prize-winning mechanisms and processes.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-022-04468-y.

Introduction

Until the end of the last century, the classical sequence → structure → function paradigm and related “lock-and-key” as well as “induced fit” models have been the major theoretical basis for understanding the molecular mechanisms of protein function. The use of these ideas was extremely rewarding, and almost everything the community learned about proteins by that time was in line with these concepts. Enzymes historically have been the primary targets of scientific research, and their catalytic activities were critically depended on the presence of unique 3D-structures, at least within the active center of an enzyme. Therefore, the validity of such a “rigid” perspective on protein functionality was beyond any doubt. The situation however started to change at the turn of this century and the potential functional importance of proteins and protein regions without confined structure, which are known now as intrinsically disordered proteins (IDPs) or intrinsically disordered regions (IDRs), is now increasingly being recognized [1–8]. Typically, IDRs (which could be disordered loops or any stretch of protein sequence) are divided into long disordered regions, which include more than 30 residues, and short disordered regions containing less than 30 residues [9]. Providing a historical perspective of the developments in protein science in the light of intrinsic disorder is outside the scopes of this work, but the interested reader is referred to recent publications, e.g., [10–16].

Regardless of their lack of unique 3D structures, IDPs and IDRs are biologically active, possessing a multitude of specific functions [1–4, 17–19]. Generally speaking, protein functionality is critically dependent on both, the presence and absence of a unique structure, and based on their order–disorder status, functional proteins can be ordered, disordered, and hybrid; i.e., possessing both ordered and disordered domains/regions. Functions of IDPs/IDRs complement functions of ordered proteins and domains [20–22]. IDPs/IDRs are found in all proteomes characterized so far [2, 19, 23–25]. Dysfunction of IDPs/IDRs is commonly related to the pathogenesis of various human diseases [26]. Furthermore, these biologically active but structure-less proteins and regions have amino acid sequences and compositions, which are noticeably different from those of ordered proteins and domains [1, 2, 17, 19, 27–32], and these peculiarities allow predicting IDPs/IDRs from their amino acid sequences only [33].

IDPs and IDRs are characterized by high structural heterogeneity, where molecules can include larger or smaller amounts of flexible secondary structure elements and tertiary contacts, and therefore can be more or less compact [18, 19, 27, 29]. Based on their global disorder status, IDPs/IDRs can exist as native molten globules containing collapsed disorder or native coils or pre-molten globules with extended disorder [2, 18, 34]. Regions of semi-disorder are also common. These are protein segments showing almost equal probability to be disordered or ordered [35]. Although IDPs/IDRs exist as highly dynamic conformational ensembles [36], structures of many of them can be reasonably well described by a rather limited number, a few thousands if not a few hundred, of lower-energy conformations [37–43].

Since not only the entire protein, but its any region can be disordered to a different degree, any protein is characterized by highly complex and heterogeneous spatiotemporal structural organization. This structural mosaic includes units capable of independent folding (foldons), IDRs undergoing disorder-to-order transitions induced by binding to specific partners (inducible foldons), IDRs with the capability to fold differently being bound to different partners (morphing inducible foldons), permanently disordered regions (non-foldons), regions with perpetually semi-folded conformations (semi-foldons), and ordered regions functional activation of which requires order-to-disorder transitions (unfoldons) [10, 44–46]. Not surprisingly, this complex structural mosaic of IDPs/IDRs (and proteins in general) defines their complex molecular ‘physiology,’ multifunctionality, and binding promiscuity. This is because differently (dis)ordered structural elements of a protein or different members of its dynamic conformational ensemble can be functional, possessing characteristic and different biological activities [47]. Structural and functional complexity of functional disorder is further increased by the fact that sites of various posttranslational modifications (PTMs) are commonly located within IDPs/IDRs [48, 49].

Functional classification of IDPs/IDRs includes several broad groups, such as entropic chains (activities of which depend on the perpetual motion of highly disordered polypeptides [2]), chaperones, assemblers, display sites, scavengers, and effectors [3]. In contrast to the catalysis of various biological reactions mediated by structured enzymes, the highly dynamic spatiotemporal structural organization of IDPs/IDRs is incompatible with enzymatic catalysis, and only a few cases were reported, where molten globular proteins with collapsed disordered were shown to exhibit catalytic activity [50–53].

One of the specific functional characteristics of IDPs/IDRs is their exceptional binding promiscuity. In fact, these proteins/regions are capable of binding to other proteins, polysaccharides, nucleic acids, membrane bilayers and small molecules [1, 48]. Often, they are involved in specific interactions with multiple structurally unrelated partners [54], can be engaged in one-to-many and many-to-one binding modes [21, 22], and can form static and dynamic complexes [55–60].

Because of the lack of stable structure in the unbound state and capability to mold into sophisticated configurations at interaction with their partners, bound IDPs/IDPRs are characterized a remarkable variability of binding modes and can be found as parts of static complexes with rather unusual topology [58]. Interactions with specific binding partner(s) trigger disorder-to-order transitions in the inducible foldons of IDPs/IDRs [59, 60], often resulting in the formation of signaling complexes characterized by high specificity and low affinity. On the other hand, interaction of the inducible morphing foldons with different partners can lead to the appearance of structurally different conformations in bound states of many IDPs/IDRs [54]. There are also IDPs/IDRs that form fuzzy (dynamic or disordered) complexes [55–58], where significant levels of intrinsic disorder are preserved at least outside the binding region [61–68]. The under-folded regions of such fuzzy complexes serve as contact sites for the additional binding partners, can be regulated by PTMs [68], or can be engaged in entropic chain activities [19, 22].

In summary, IDPs/IDRs represent sophisticated systems with complex and highly heterogeneous structural and functional organization, which, due to their flexible nature, binding promiscuity, and high potential for multifunctionality, can act as hubs in protein–protein interaction networks [21, 22]. More generally, functionality of IDPs and hybrid proteins with ordered domains and IDRs can only be described within the structure–function continuum model based on the “one gene – many proteins – many functions” concept, instead of the classical “lock-and-key” theory rooted in the “one gene – one protein – one structure – one function” view [47, 69]. A cornerstone of this structure–function continuum model is an idea that the form of existence of any protein is a dynamic conformational ensemble containing multiple proteoforms with different functional potentials [13]. Proteoforms here are various protein species originating from a single gene due to mutations, mRNA editing, alternative splicing, alternative translation initiation, PTMs, intrinsic disorder, and protein functioning [69].

Despite the importance of intrinsic disorder for various biological processes, the majority of mechanistic studies are ignorant toward this phenomenon. Indeed, studies on particular protein functions that critically depend on IDPs/IDRs have received high recognition, namely gene regulation (1965 Nobel Prize, Physiology or Medicine), genetic control of early embryonic development (1995 Nobel Prize, Physiology or Medicine), prion disease (1997 Nobel Prize, Physiology or Medicine), signal transduction between nerve cells (2000 Nobel Prize, Physiology or Medicine), ubiquitin–proteasome system (UPS, 2004 Nobel Prize, Chemistry), eukaryotic transcription (2006 Nobel Prize, Chemistry), structural and functional features of the translation machinery, ribosome (2006 Nobel Prize, Chemistry), induced pluripotent stem cells (2012 Nobel Prize, Physiology or Medicine), G-protein-coupled receptors (GPCRs, 2012 Nobel Prize, Chemistry), mechanistic studies of DNA repair (2015 Nobel Prize of Chemistry), autophagy (2016 Nobel Prize, Physiology or Medicine), circadian rhythms (2017 Nobel Prize, Physiology or Medicine), oxygen sensing (2019 Nobel Prize, Physiology or Medicine), CRISPR/Cas9 (2020 Nobel Prize, Chemistry), and thermal and mechanical transducers (2021 Nobel Prize, Physiology or Medicine). However, the biological research community and even the Nobel Laureates themselves were often not aware of the key roles played by IDPs or IDRs in the molecular mechanisms underlying these just-mentioned phenomena.

The goal of this article is to fill this gap and to show the prevalence of intrinsic disorder in proteins associated with 17 Nobel Prizes in Physiology/Medicine or Chemistry that have been awarded between 1965 and 2021. We represent some illustrative examples of IDPs/IDRs in these award-winning mechanisms and processes. The reader will notice that the stories describing the individual Nobel Prize achievements have slightly different styles and foci. This is because this work originates from a joint project conducted by students of a research training group on IDPs (RTG 2467, https://rtg2467.uni-halle.de) at the Martin Luther University of Halle-Wittenberg, Germany.

Materials and methods

Among the Nobel Prizes awarded between 1965 and 2021, 17 topics were selected from the categories of Physiology or Medicine and Chemistry. This selection was solely based on the contribution of the corresponding studies to the mechanistic understanding of biological processes. For the proteins involved in the biological processes behind these selected Nobel Prizes, the content of disorder has been investigated by the database of disordered protein predictions (D²P²: https://d2p2.pro) [70] using the amino acid sequences available on UniProt databases (https://www.uniprot.org). For each biological process, one or few proteins with the highest IDR content were selected, their roles were described and their structural models were retrieved from the Alphafold database (https://alphafold.ebi.ac.uk) [71]. For those proteins that were not present in Alphafold, the model was predicted using Colab notebook, a slightly simplified version of AlphaFold v2.1.0 [71], or ColabFold advanced notebook [72]. The resulting portrait gallery of most IDPs associated with the award-winning processes is shown in Fig. 1 and Table S1.

Fig. 1.

AlphaFold-generated portrait gallery (A) and IDRs content schematic representation predicted by D²P² (B) of the most disordered proteins in datasets related to 17 Nobel Prizes

A global computational analysis of the disorder was also performed on the proteins related to the 17 selected Nobel Prizes. In particular, the analysis of charge-hydropathy-cumulative distribution function as well as per-residue disorder predisposition analysis is described in detail below.

Results and discussion

Global analysis of disorder in proteins related to Nobel Prizes

Charge-hydropathy-cumulative distribution function (CH-CDF) analysis

A set of proteins associated with discoveries that were awarded with 17 Nobel Prizes includes 3223 proteins. The results of global disorder analysis of these proteins are presented in Fig. 2. First, we utilized charge-hydropathy (CH)—cumulative distribution function (CDF) analysis (see Fig. 2A). This approach is based on the combination of the outputs of two binary disorder predictors, CH plot [1, 73] and CDF plot [73, 74], which were designed to classify proteins as entirely ordered/compact or entirely disordered/extended based on different physico-chemical definitions of intrinsic disorder. Here, information on the absolute mean net charge and mean hydropathy is utilized by CH-plot to classify a query protein as a native coil or native pre-molten globule (i.e., a protein with substantial levels of extended disorder) or native molten globule or ordered protein (i.e., a protein with compact globular conformations) [73, 75]. In CDF analysis, proteins with any type of disorder (native coils, native pre-molten globules, and native molten globules) are differentiated from the ordered proteins [73]. Combination of the outputs of these binary predictors in a unified CH-CDF plot opens a way of predictive classification of proteins into several structurally different classes [74, 76, 77]. In the resulting CH-CDF plot, X- and Y-axes correspond to the distances of query proteins from boundaries in the CDF- and CH-plot, respectively. This combined plot contains four quadrants that provide means for the classification of proteins depending on their positions within the CH-CDF space as ordered (the lower-right quadrant, Q1), native molten globules or hybrid proteins with sizable levels of order and disorder (the lower-left quadrant, Q2), native coils and native pre-molten globules (the upper-left quadrant, Q3), and “strange” proteins predicted to be disordered by CH-plot but ordered by CDF (the upper-right quadrant, Q4) [76].

Fig. 2.

Global disorder analysis of proteins associated with 17 Nobel Prizes. A CH-CDF analysis of 3,223 proteins related to individual Nobel Prizes. Each subset of proteins is shown by specific symbol and color. Quadrants in the CH-CDF phase space group proteins based on the outputs of their disorder analysis by CH- and CDF-plots: Q1 (lower-right quadrant) contains proteins predicted as ordered by both predictors, Q2 (lower-left quadrant) represents proteins predicted to be disordered by CDF but ordered/compact by CH, Q3 (upper-left quadrant) shows proteins predicted as disordered by both predictors, and Q4 (upper-right quadrant) includes proteins predicted to be ordered by CDF but disordered by CH. B PONDR^® VSL2 output for 3223 proteins associated with individual Nobel Prizes in the form of the ADS vs PPIDR dependence. Differently colored blocks indicate regions of the ADS-PPIDR space, in which proteins are mostly disordered (red), moderately disordered (pink and light pink), or mostly ordered (blue and light blue). Intensity of the colors of the background reflects agreement between the ADS and PPIDR outputs, where dark blue or pink colors correspond to the regions of the ADS-PPIDR space where the two parameters agree, and where areas in which only one of these criteria applies are shown by light blue and light pink

Figure 2A represents the results of CH-CDF analysis of the 3223 proteins related to various Nobel Prizes. This analysis revealed that 951 and 1149 of these proteins are predicted as highly disordered or possessing a molten globular or hybrid structure, respectively. Remaining 1098 proteins are mostly ordered. Therefore, CH-CDF analysis revealed that almost 2/3 of proteins related to various Nobel Prizes (~ 65.2%) are predicted to be either globally disordered or contain sizable levels of disorder or show properties of molten globule-like conformations with high compactness, high levels of secondary structure, and lack of ordered tertiary structure (see also Table S1 and Fig. 1B).

Evaluation of the per-residue disorder predisposition

In the per-residue disorder analyses, each residue in a sequence of a query protein is assigned a disorder propensity score, values of which are used to classify residues/regions as disordered or flexible, if their disorder scores are above the threshold value of 0.5 or range between 0.15 and 0.5, respectively. For each protein, the average disorder score (ADS) can be calculated by averaging all the per-residue disorder predisposition values and the percent of predicted intrinsically disordered residues (PPIDR) can be found based on the normalized content of the disordered residues (i.e., residues with the predicted disorder scores greater than 0.5). Furthermore, ADS and PPIDR values can be used for the global classification of proteins as highly disordered (ADS ≥ 0.5; PPDR ≥ 30%), moderately disordered (0.15 ≤ PPDR < 0.5; 10% ≤ PPDR < 30%), and highly ordered (ADS < 0.15; PPDR < 10%) [78].

As there is no direct correlation between ADS and PPIDR of a given protein, one of the ways of presenting large sets is showing corresponding data in the form of the ADS vs. PPIDR plot. Figure 2B represents this plot for data generated for proteins related to various Nobel Prizes by one of the more accurate per-residue disorder predictors, PONDR^® VSL2 [79, 80]. Despite being relatively old, this commonly used computational tool is characterized by high predictive power as evidenced by the results of the recently conducted ‘Critical assessment of protein intrinsic disorder prediction’ (CAID) experiment, where PONDR^® VSL2 was recognized as predictor #3 of the 43 evaluated methods [81]. Analogous data generated for these proteins by PONDR^®-VL3, PONDR^® VLXT, PONDR^® FIT [29, 82, 83], and IUPred2A platform [84] are collected in supplementary materials. Figure 2B and Table 1 clearly show that almost all proteins associated with Nobel Prizes are expected to be moderately or highly disordered. In fact, of 3223 proteins in the analyzed dataset, only 69 (2.1%) and 7 (0.2%) were predicted as highly ordered, based on their PPIDR and ADS scores, respectively. Furthermore, none of the subsets corresponding to the proteins associated with any particular Nobel Prize was predicted as ordered, and only two subsets, namely CRISPR/Cas9 (2020 Nobel Prize, Chemistry) and thermal and mechanical transducers (2021 Nobel Prize, Physiology or Medicine), were predicted as moderately disordered, with all other subsets being predicted as highly disordered by at least one criterion (PPIDR_mean or ADS_mean; i.e., subset averaged PPIDR and ADS values).

Table 1.

Overall intrinsic disorder status of proteins related to Nobel Prizes

Rows are colored according to the global disorder status of proteins related to a give Nobel Prize. Dark red—predicted as highly disordered by both PPIDR_mean and ADS_mean; Light red—predicted as highly disordered by both at least one criterion (PPIDR_mean or ADS_mean); Pink—predicted as moderately disordered by both PPIDR_mean and ADS_mean

All these observations clearly indicate that the vast majority of proteins behind Nobel Prize-winning processes contain substantial levels of intrinsic disorder. The sections below provide a description of 17 Nobel Prizes and present illustrative examples of IDPs/IDRs in these award-winning mechanisms and processes.

Proteins associated with individual nobel prizes

1965: gene regulation

In 1965, François Jacob, André Lwoff, and Jacques Monod received the Nobel Prize in Physiology or Medicine ‘for their discoveries concerning genetic control of enzyme and virus synthesis’ [85]. Jacob discovered that the chromosomes of bacteria have a circular structure and that bacteria can have additional genetic elements called episomes, which can be expressed and replicate autonomously. He also demonstrated the existence of the messenger RNA. Lwoff proved that after bacteriophage infections, some bacteria undergo a non-infectious prophage phase, where the bacteriophage DNA integrates in a specific region of the bacterial chromosome that differs for each bacteriophage. During this phase, virions are not produced, but Lwoff observed that it was possible to induce bacteriophage production by using different inducers. Lastly, Monod discovered the “diauxy” phenomenon, in which bacteria use different carbohydrate sources displaying “two exponential phases” divided by growth arrest, one for each source. The three researchers then collaborated and worked on both lysogeny and β-galactosidase expression, demonstrating that these two systems are induced by the substrate that binds to the repressor in order to induce its dissociation from the operator and thus allows the synthesis of the operon. They also assumed that the repressor undergoes conformational changes after binding to the inducer and therefore concluded that this binding occurs at an allosteric site on the repressor. Their studies have resulted in numerous further discoveries, which, together with the development of novel technologies, have dramatically improved the study of proteins and their functions [85].

It is now known that a lot of proteins involved in gene regulation are IDPs, and in eukaryotes, these proteins have a higher disorder content compared to prokaryotes [86]. This can be explained by the fact that in more complex organisms, many different signals converge; therefore, the transcription system has to be more flexible [9]. This implies that IDPs/IDRs are related to the regulation of the most important cellular processes, and their dysregulation is linked to the development of various maladies, such as cancer. The interactions between some IDPs/IDRs and other proteins are characterized by low binding affinities, and the presence of flexible domains allows highly specific interactions [86]. As mentioned above, it has been demonstrated that this specificity can be regulated by PTMs [49].

With the use of disorder predictors, it has been found that almost all transcription factors (TFs) possess IDRs, and each TF displays different disordered contents [86]. For many TFs, the disordered region is not located in the DNA-binding domain. Instead, the IDR is located in regions, where the cofactor or other protein interaction partners bind. Examples include bHLH (basic Helix-Loop-Helix) proteins involved in differentiation and development, where their partially disordered bHLH domain is involved in dimerization with different protein partners [87].

Further studies on the lac repressor based on this Nobel prize-winning discovery resulted in a breakthrough in the understanding of fundamental mechanisms, by which most proteins recognize a specific DNA-binding site [88, 89]. Furthermore, knowledge has been acquired regarding bacteriophages and, more generally, viruses. Analysis of the amino acid sequences of different IDRs revealed that they are often characterized by the conservation of amino acid compositions (high content of charged and polar amino acids) rather than conservation of their sequences [9]. This implies a higher ability of these proteins/regions to evolve, adapt to different environments, and acquire novel functions [90].

1977: peptide hormone development

In 1977, Rosalyn Sussman Yalow, Roger Guillemin, and Andrew V. Schally shared the Nobel Prize in Physiology or Medicine for their fundamental discoveries in the field of peptide hormones [91]. By developing quantitative assays to measure peptide hormones with extremely high sensitivity and specificity, Sussman Yalow disproved with her award-winning discoveries the—at that time prevailing—assumption that insulin could not be antigenic. By establishing laborious purification procedures to identify and characterize numerous peptide hormones, Guillemin and Schally confirmed the hypothesis proposed in 1951 that hypothalamic hormones regulate pituitary function [92]. Together, all three Nobel laureates laid the foundation of the gigantic field of endocrinology, highlighted the existence of a hypothalamic–pituitary–adrenal axis, and paved the way for two Nobel Prize winners in 2012 who identified and characterized peptide hormone receptors.

In biochemical tour-de-force approaches, the teams of Schally and Guillemin independently discovered adrenocorticotropic hormone (ACTH). Both were guided by the prevailing assumption that a specific hormone must exist in hypothalamic tissue extracts that would release ACTH from the anterior pituitary gland, which was isolated much later in 1981 and named corticotropin-releasing factor (CRF) [93]. Derived from a 196-amino acid-long intrinsically disordered precursor via cleavage of the signal peptide (residues 1–24) and propeptide (residues 25–153), the mature CRF (residues 154–194) triggers its cognate receptor (CRFR), a G protein-coupled protein receptor (GPCR), to activate the adenylate cyclase signaling pathway to elicit secretion of ACTH.

The generation of ACTH is especially interesting from the IDR perspective. Its 39 amino acids are proteolytically derived from the 267-amino acid-long pre-pro-opiomelanocortin (pre-POMC) precursor. However, the very same 241-amino acid-long pro-hormone precursor POMC, gives also rise to numerous other hormones, such as the endorphin β-lipotrophin and melanocyte-stimulating hormones (α- and β-MSH) [94]. Its endoproteolytic cleavage by at least five pro-hormone convertases might be aided by POMC’s fully unstructured nature. Furthermore, two of the five endopeptidases have large IDRs next to their enzymatically active center: Neuroendocrine convertase 1 has a fully unstructured C-terminal tail of approximately 150 amino acids, which represents about 20% of its entire length [71]. Peptidyl-glycine α-amidating monooxygenase exhibits a large IDR, separating the two enzymatically active domains, next to its unstructured N- and C-terminal regions, amounting in sum to ~ 30% of disorder [71].

1995: early embryonic development

The Nobel Prize in Physiology or Medicine 1995 was awarded to three pioneering scientists in the field of developmental biology working with Drosophila melanogaster, for their discoveries in the “genetic control of early embryonic development.” Edward B. Lewis first discovered the collinearity principle. He postulated that genes at the beginning of the chromosomes control anterior body segments, while the following genes control body segments [95]. Christiane Nüsslein-Volhard and Eric F. Wieschaus worked in harness and conducted a series of experimental studies, which led to the identification of 15 different development-regulating genes and their classification into three classes: gap genes, pair rule genes and segment polarity genes [96]. All their studies, although performed on Drosophila melanogaster, were found to be related to higher organisms as well. Looking into the set of genes identified and classified by the laureates, as well as the resulting encoded proteins, undeniably, disorder plays an important role. In fact, the protein Krüppel (gap gene encoded), the Hunchback protein, the segmentation protein Even-skipped (pair rule gene encoded), as well as the segmentation polarity homeobox protein engrailed (segment polarity gene encoded), are predicted to have disorder contents above 30% (Fig. 2). The following paragraphs will be focussed on the protein with the highest calculated disorder content, the segmentation protein Even-skipped (Eve).

Eve consists of 376 amino acids, contains a divergent homeobox sequence (DNA-binding domain), and is localized predominantly in the nucleus of embryonic cells, which encode the eve gene [97, 98]. It seems to play a particularly important role in the segmentation process and acts as a transcriptional regulator of other genes important for early embryonic development [99, 100]. A closer look at the predictions in the area of the homeodomain (aa 70–130) [97] reveals that although it is expected to contain 80% of disorder (VSL2b, JRONN, ESpritz-NMR), some kind of α-helical secondary structure is predicted. This is in agreement with the findings of Hirsch et al. in 1995 [98]. But from the perspective of IDPs, the N-terminal arm, bordering a region predicted to be disordered by various disorder predictors, is of particular interest. This N-terminal part is reported to lack secondary structure and having a highly mobile peptide chain in that area. However, upon complex formation, the N-terminal arm is located in the minor groove of the DNA, contributing to the HD-DNA binding. In addition to this, Arg3 is highly conserved between HDs but its interaction with DNA varies depending on how the N-terminal arm is positioned upon binding [98]. These findings suggest that a disorder-to-order transition of an IDR can be observed.

Eve homologues were studied in several species and a lot of similarities were observed from expression patterns in early embryonic development to a very high conservation of the homeodomain (varying in only two amino acids between fruit flies and humans). Therefore, it is one of many multifunctioning proteins featuring a high potential disorder content of utmost importance which, nevertheless, still has to be fully understood.

1997: prion disease

The Nobel Prize in Physiology or Medicine 1997 was awarded to Stanley B. Prusiner for his discovery of prions, a new biological principle of infection. In 1982, Stanley B. Prusiner and his colleagues successfully discovered this new genre of disease-causing agents, which could be added to the well-known list of other infectious agents including bacteria, viruses, fungi, and parasites [101]. Prusiner named the infectious agent “prion-derived” from “protein and infection.” Prion diseases are caused by the conversion of the normal cellular prion protein (PrP^C) into a conformationally altered isoform PrP^Sc (Sc refers to “scrapie”, the prion disease occurring in sheep and first discovery of the infection-causing agent) due to induced or spontaneous misfolding. The presence of the sole protein PrP^Sc abnormal isoform seems to be sufficient for stimulation of the pathological conversion since it serves as a template for the transition of other PrP^C into the infectious conformation [102]. Prions cause a variety of neurodegenerative diseases that can be infectious, inherited, or sporadic in origin. These diseases include bovine spongiform encephalopathy (BSE) in cattle, scrapie in sheep, and Creutzfeldt-Jakob disease, Gerstmann-Straussler-Scheinker disease, kuru, and fatal familial insomnia in humans [103].

The physiological form of human PrP^C (HuPrP^C) is a 210-amino acid glycoprotein, encoded by the single-copy gene PRNP. The physiological functions of HuPrP^C are highly diverse, as HuPrP^C belongs to the expanding group of multifunctional proteins. It plays an important role in neuronal development, cell adhesion, axon guidance, synapse formation, neuroprotection, regulation of circadian rhythm, and myelin maintenance. For a comprehensive review of suggested physiological roles of PrP^C, the reader is referred to [104].

The N-terminal part of HuPrP^C (NH₂-HuPrP^C) is composed of a large IDR consisting of 101 residues and bringing the disorder content of the mature prion protein to 44%. This region acts as a molecular sensor with a broad range of ligand partners that allows HuPrP^C to relay neuroprotective signals and additionally even modulates prion conversion to PrP^Sc [105, 106]. NH₂-HuPrP^C may enable interactions with small molecules (e.g., Cu²⁺, hemin) to macromolecules (e.g., phospholipids, proteins). Cu²⁺ binds to NH₂-HuPrP^C in vivo [107]. Two main functions of Cu² binding to NH₂-HuPrP^C of the prion protein can be postulated: (I) stimulation of HuPrP^C endocytosis, which requires an octapeptide repeat region [108], and (II) Cu²⁺ sensing associated to cell signalling [107]. Full binding of Cu²⁺ to the octapeptide repeat region yields a conformation, which is likely to be involved in the mechanism of endocytosis, regarding the high Cu²⁺ concentrations that are needed for this cellular process. On the other hand, Cu²⁺ binding in the non-octapeptide repeat region induces a β-sheet-like transition [109]. This interaction can have a functional purpose regarding amyloid aggregation and prion conversion to PrP^Sc.

The unstructured N-terminal region of HuPrP^C is a critical modulator of the HuPrP^C functions in neuronal survival. Under physiological conditions, NH₂-HuPrP^C displays a neuroprotective function that is modulated by a large variety of ligand partners. On the other hand, under the pathological conditions, binding of PrP^Sc to NH₂-HuPrP^C induces a switch in HuPrP^C function toward a neurotoxic pathway.

2000: signal transduction in nerve cells

The 2000 Nobel Prize in Physiology or Medicine was awarded to Arvid Carlsson, Paul Greengard, and Eric Kandel for their “pioneering discoveries concerning one type of signal transduction between nerve cells, referred to as slow synaptic transmission.” After Carlsson’s discovery of the central role of dopamine in the nervous system [110], Greengard shed light on its molecular function in the nervous system. He identified protein phosphorylation as a key event during short-term memory formation [111], while Kandel showed long-term memory to be formed by long-lasting changes in protein expression [112].

Soluble messenger compounds, such as dopamine, are crucial for neurotransmission. They are stored inside membrane-enclosed synaptic vesicles in the neuronal cytoplasm of dopaminergic neurons. Excitation of such a neuron leads to fusion of the synaptic vesicles with the presynaptic membrane and exocytosis of dopamine. Dopamine functions by binding to either activating D1 or inhibiting D2 dopamine receptors and increasing or decreasing the concentration of cyclic AMP in the postsynaptic neuron [113]. Increase in cAMP levels is sensed by the regulatory subunits of protein kinase A (PKA) leading to increased phosphorylation of a multitude of downstream targets.

One of the targets of PKA is the “dopamine and cAMP regulated phosphoprotein of 32 kDa” (DARPP-32), which is widely expressed in the neuronal soma and which plays a signal-integrating role [114]. Activated DARPP-32 inhibits protein-phosphatase 1 (PP1), thereby influencing the phosphorylation state of multiple targets in the neuron, and directly contributes to the short-term memory. One of the switches for DARPP-32 activation is dopamine-induced phosphorylation of Thr34 by PKA. On the contrary, glutamate signaling induces dephosphorylation of Thr34 by calcineurin and inactivates DARPP-32 [115]. Furthermore, dopamine and glutamate signals not only play an antagonistic role, but DARPP-32 rather acts as a temporal integrator of signals from both pathways. If the two signals occur within short time and in definite order, the increase in phosphorylation is potentiated [116]. Structural analysis of DARPP-32 by NMR spectroscopy indicates that most of the protein is unfolded in the absence of a binding partner with no effect of Thr34 phosphorylation on the secondary structure [117, 118]. Only the N-terminus was found to partly form α-helical secondary structure that was suggested to play a role in positioning Thr-34 at the active site of PP1 during inhibition similar to what was observed for the PP1-binding proteins I-2 and spinophilin [118].

In contrast to short-term memory formation through PTM-mediated regulation, long-term memory formation requires de novo synthesis of proteins. In this process, neuronal activation leads to the increase in cAMP levels and activation of PKA. In contrast to the process described for short-term memory formation, long-lasting activation leads to the recruitment of mitogen-activated protein kinase by the catalytic subunit of PKA. Both relocate to the nucleus and phosphorylate transcription factors, such as cAMP responsive element binding protein 1 (CREB-1) [119]. In its active form, CREB-1 binds to the cAMP response element in the control region of the DNA of multiple genes and activates their transcription [120]. The gene products induce the formation of novel synaptic connections, a process commonly referred to as synaptic plasticity. Initiation of transcription by CREB-1 is mediated by binding to CREB-binding protein (CBP), a histone acetyltransferase that loosens chromatin structure and allows binding of the transcription machinery. This interaction is mediated by the kinase inducible domain (KID) of CREB-1, an IDR that upon phosphorylation by PKA, folds and binds to CBP [121]. First, an unstructured encounter complex is formed that evolves into the final bound state via formation of several non-native contacts, which have been implicated in regulation of the speed of the association reaction [122, 123]. Phosphorylation elongates the life-time of the CREB-1/CBP complex and thereby eases assembly of the transcription initiation complex [124]. Interestingly, folding of the KID α-helix occurs only after binding to CBP, so that the transition state may resemble “fuzzy” complexes found in many IDPs [123]. In addition to KID, CREB-1 also contains a basic leucine zipper that forms no secondary structure when purified as isolated domain. It is required for dimerization and DNA-binding and its structural disorder was suggested to facilitate target search by remaining dynamic during diffusion along the DNA and thereby increasing the effective area the protein can explore [125].

In summary, intrinsic disorder is a striking feature of DARPP-32 and plays a key role in CREB-1 function. Both examples show its crucial role in regulating complex networks, such as these involved in memory formation. In addition to DARPP-32 and CREB-1, six other proteins mentioned by the Nobel laureates in their lectures also show a high degree of predicted disorder (Fig. 1) and further exemplify important role of protein intrinsic disorder in learning processes.

2004: ubiquitination

The revolutionary discovery of the ubiquitin–proteasome system (UPS) awarded Aaron Ciechanover, Avram Hershko, and Irwin Rose the 2004 Nobel Prize in Chemistry. These researchers described for the first time a multi-step reaction that tag proteins with ubiquitin (Ub) by isolating and characterizing three enzymes: an Ub-activating enzyme (E1), an Ub-conjugating enzyme (E2), and an Ub ligase (E3) from reticulocyte lysates [126]. Addition of Ub to target proteins, or ubiquitylation, starts with the activation of Ub C-terminus (G76) in an adenylation followed by transesterification to generate a thioester-linked E1 ~ Ub conjugate. Activated Ub is then transferred to an E2 via transthiolation reaction. Lastly, an E3, which function as an Ub target receptor, binds both the substrate and an E2 ~ Ub conjugate, catalyzing the transfer of Ub, often to the ε-amino group of a lysine in the target protein, forming an isopeptide bond [126]. Most of ubiquitylated proteins undergo degradation by the 26S proteasome, nevertheless ubiquitylated proteins are also implicated in a plethora of cellular processes, such a membrane trafficking, cell cycle regulation and signaling cascades, since degradation is not always the final outcome of ubiquitylation.

Although the authors’ pioneering work dates back to 1978, the relevant role of intrinsic disorder in the UPS was acknowledged in late 2000, almost 30 years later! In fact, one of the earliest jobs by Catic [127] revealed that 40 ubiquitylation sites in 26 proteins in Saccharomyces cerevisiae occur within loops (26/40), followed by α-helices (10/40) and finally β-strand (4/40), highlighting a structural preference for Ub transfer. Hagai et al. [128] constructed a dataset of 42 proteins known to undergo degradation-mediated ubiquitylation, pointing out the existence of a significant (P = 2 × 10^–4) bias of ubiquitylation sites in coils regions (55.1%) of which 46% of the sites are predicted to be disordered regions. More recently, Guharoy [129] proposed that protein turnover by the UPS is regulated by a tripartite degron (i.e., signal for protein degradation) composed of: (1) a primary degron (i.e., degron motif) which allows recognition by an E3, (2) a secondary degron or ubiquitylation zone comprising one (or multiple neighboring) lysine(s), often located in IDRs, and (3) a tertiary degron, which is an IDR serving as an initiation signal for proteasome degradation. Tertiary degrons or initiation signals have been extensively studied and they locate not only at the N- or C-terminal but also at internal IDRs within a protein [129, 130]. While N- or C-terminal initiation signals have to be ~ 30 amino acids long, internal signals have to be much longer (> 40 residues) to engage the proteasome effectively. As a result, proteins with long (> 30 amino acids) disordered tails have a shorter half-life (less than 30 min) compared to proteins with short (< 30 amino acids) disordered tails (31–70 min). Not only length, but also amino acid composition of initiation signals is important as hydrophobic and nonpolar amino acids and stiffer polypeptides chains are preferred by the proteasome, while polar, acidic, and structurally flexible sequences are avoided [130, 131].

Intrinsic disorder in the UPS is not limited to ubiquitylation substrates as accumulating evidence revealed the importance of disorder in E2 and E3 [132]. Humans possess ~ 40 E2 ligases, of which 32 structures have been solved, containing a core catalytic domain named UBC domain, which adopts α/β-fold with four α-helices and four β-sheet and important loop regions that comprise the E3-binding interface and the active site [133]. Among all E2 ligases, the Ube2W family stands out for having a unique C-terminal disordered region in the UBC, which is linked to an unusual intrinsic Ub transfer reactivity, attaching Ub to the N-terminus of proteins [134]. E3 Ub ligases are a diverse family of proteins with ~ 600 members in humans that can be grouped based on accessory components as single-subunit E3, RING-finger, U-box and HECT E3, which bind E2 ~ Ub and substrate simultaneously. As IDPs can establish multiple protein–protein interactions, it has been proposed that the IDRs in E3 ligases make them a central hub for recognition of multiple substrates and/or cofactors. As such, they can establish three kind of interaction where (1) both E3 and substrate and/or cofactors remain disordered upon binding, (2) induce folding of substrate and/or cofactors by binding to a folded domain in the E3, and (3) co-folding or mutual synergistic folding [132]. The last decade has proven that intrinsic disorder in the UPS besides being a structural feature of targets that undergo ubiquitylation also plays an essential role in the regulation of E2 and E3 activities. Since protein degradation influences many cellular processes, understanding the function of intrinsic disorder in the context of the UPS may help to comprehend diseases, such as cancer and growth defects.

2006: eukaryotic transcription

In 2006, the Nobel Prize in Chemistry was awarded to the biochemist and structural biologist Roger D. Kornberg. He and colleagues uncovered the molecular basis and unexpected complexities of the eukaryotic transcription by providing the first atomic structures of the RNA polymerase II (Pol II) complex of Saccharomyces cerevisiae [135].

The Pol II comprises 12 subunits (Rpb1 to Rpb12) and plays a crucial role in a variety of cellular function by regulating the synthesis of mRNA. In 2000, a backbone model of a 10-subunit yeast Pol II, lacking the Rpb4 and Rpb7 subunits, was derived from X-ray diffraction data at 3.5 Å resolution. This model provided the basis for understanding the general architecture and led to proposals for the role of each subunit in the transcriptional process [136]. Later on, the highest resolution of Pol II structure at 2.8 Å was reported, leading to a division of this proteinaceous machine into four mobile elements named core, clamp, jaw-lobe, and shelf [137]. The core element contains the two largest subunits Rpb1 and Rpb2 forming the active center and the subunits Rpb3, Rpb10, Rpb11, and Rpb12. We analyzed the disorder content of the 12 subunits of yeast Pol II utilizing the D²P² database [138] of protein disorder and mobility annotations (Fig. 1). The catalytic subunit Rpb1 was predicted to have an overall disorder content of 14%, with the disordered region being mainly located in the C-terminal domain (CTD), spanning the amino acid 1519 to 1733. These findings correlate with the published data of Cramer et al., where missing electron density in Pol II maps was observed in the region from 1451 to 1733 amino acids, providing evidence of high structural mobility or disorder [137]. The CTD consists of tandem repeats of heptapeptide sequences (YSPTSPS). These increase in number with organism complexity (26 in yeast and 52 in humans) and undergo multiple regulatory PTMs during the transcription cycle [139, 140]. Hitherto, the most studied and well-characterized CTD modification is phosphorylation, which facilitates or inhibits the recruitment of regulatory proteins to the CTD [141]. This PTM in combination with structural flexibility could explain the various interactions of the CTD with different proteins, such as the Mediator or mRNA processing proteins during the transcription initiation and elongation [139, 141].

According to the D²P² database, the N-terminal region of Rpb6 (residues 1–81) is also predicted to be disordered. The overall predicted disorder content of Rpb6 is 52%. In support of these predictions, the absence of electron density in the N-terminal fragment (residues 1–71) was reported in Cramer et al., emphasizing an intrinsically disordered nature of this region [137]. Rpb6 is in proximity to the flexible clamp, suggesting its crucial role in the assembly and function of the Pol II [142]. Like the CTD of Rpb1, Rpb6 can be phosphorylated, but to the best of our knowledge, the function of this region remains poorly understood. It was hypothesized that this region could have an impact on the position of the clamp during DNA binding [136, 143]. It was also shown that Rpb6-Rpb7 interaction is important for assembly of the transcription initiation complex as well as for an interaction of Rpb6 with TFIIS during elongation [142, 144]. These functions could indicate why Rpb6 needs to be flexible, like the CTD.

The awarded structural analysis by Kornberg and colleagues and the mentioned unstructured regions of the described components of the Pol II provide important examples of IDRs. Future studies will have to decipher the contribution of IDRs to the function of Pol II-driven transcription in further detail.

2009: ribosome

Considering the central role of the ribosome in expressing genetic material to form viable phenotypes across all three kingdoms of life, it is not surprising that scientists strive to obtain insights into structural and functional features of the translation machinery. For their contribution in gaining significant insights in this field, Venkatraman Ramakrishnan, Thomas A. Steitz and Ada E. Yonath were awarded the Nobel Prize in Chemistry in 2009. Ramakrishnan provided mechanistic explanations for the key ribosomal tasks [145] based on high-resolution ribosomal X-ray crystallographic data, with the underlying technique mainly established by Thomas A. Steitz [146] and Ada E. Yonath [147].

The ribosome is a large ribonucleoprotein machine consisting of many different ribosomal proteins and several rRNAs. Various ribosome structures show the presence of close protein-rRNA interactions highlighting the RNA-binding character of ribosomal proteins [148]. Indeed, just like other RNA-binding proteins, many ribosomal proteins possess a high content of positively charged amino acids, counteracting the negatively charged phosphate groups in the backbone of the rRNA [149]. Furthermore, it was shown that RNA-binding motifs in proteins have the tendency to be intrinsically disordered in the unbound state, and that they undergo a transition to a more ordered state upon binding to RNA [149]. Intriguingly, according to the comprehensive analysis conducted by Peng et al., ribosomal proteins across all three kingdoms of life are expected to possess significant levels of disorder in their unbound state, in contrast to their mostly ordered structure when present in the ribosomal complex [149]. Based on these findings and the fact that the core ribosome functions are performed by the rRNA; i.e., the directed mRNA codon/tRNA anticodon interaction and the catalysis of the peptide bond formation [145, 146], it seems plausible that ribosomal proteins with IDRs play a crucial role in the proper assembly of the ribosome complex including the precise orientation of the rRNA to yield a functional ribosome. Further evidence for this presumed role is provided by studies showing that many ribosomal proteins have RNA chaperone activity [150].

In the following, we focus on E. coli 50S ribosomal protein L2, which gained great interest for a number of reasons: L2 is a highly conserved ribosomal protein [151] that binds tightly to the 23S rRNA within the 50S subunit, where it is involved in the association to the 30S subunit to form the 70S ribosome [152]. Importantly, experimental data strongly suggest that L2 is indispensable for the catalysis of the peptide bond formation [152]. L2 from E. coli is a highly basic protein with a theoretical pI of 10.93 [153] containing 29 Arg and 25 Lys residues, already indicating its electrostatically favorable interaction with RNA. Experimental data on the unbound state of L2 suggest a dichotomous structure with a globular domain prolonged by an unstructured terminal extension [149]. This is further supported by the intrinsic disorder predisposition analysis using the MobiDB-lite algorithm. Here, a total of 28.9% (26% by D²P²) was predicted to be unstructured with most of the disorder being located at the C-terminus of this protein. The unstructured C-terminal extension of L2 is known to penetrate deep into the ribosome core structure and has multiple contact points with the 23S rRNA [146]. More specifically, in the ribosome complex, the extension of L2 fills voids between rRNA helices, neutralises RNA backbone charges and presumably stabilises the RNA fold, while approaching the peptidyl transferase center (PTC) [146]. It has to be noted though, that it cannot be directly involved in the peptidyl transfer reaction, as crystallographic data clearly show the absence of any protein within the PTC (e.g., [146]). Therefore, it has been concluded that the ribosome is a ribozyme with the 23S rRNA being the exclusive catalytic entity [145]. Yet, mutational studies indicate that the phylogenetically conserved histidine residue His²²⁹ (His²⁰⁹ in the concordant human 60S ribosomal protein L8 [154]) within the C-terminal extension of L2 is essential for efficient peptidyl transfer reactions (e.g., [152]). Since His²²⁹ is not located in the immediate vicinity of the PTC, an L2-based allosteric mechanism leading to the greatly enhanced PTC activity was proposed [152]. Further support for the general importance of the ribosomal proteins is provided by in vitro studies showing that isolated 23S rRNA was incapable of catalyzing the peptide bond formation [155] indicating that L2 and other ribosomal proteins are required to ensure formation and maintenance of a catalytically active 23S rRNA fold [156].

Although the previous paragraph is focused on L2, the disorder-based functionality of this ribosomal protein is not an exception. In fact, analysis of the InterPro database, a resource integrating eight different protein family and domain databases, revealed the presence of 3653 regions of conserved disorder prediction (CDP) in 2898 different entries [157]. The authors pointed out that many ribosomal proteins were found to contain conserved disordered regions, suggesting importance of the corresponding IDRs for the functionality of their carriers [158].

A comprehensive analysis of available structures of the ribosomal particles revealed that the basic extensions of the ribosomal proteins that penetrate deeply into the core of the ribosomal subunit play a key role in the early steps of the assembly of the eubacterial 50 S subunit through disorder-order transitions and/or co-folding mechanisms [159]. This study showed that based on their structural and dynamics properties, the basic extensions of the ribosomal proteins L3, L4, L13, L20, L22, and L24, which were experimentally shown to play crucial roles in the first steps of the ribosome assembly [160, 161], can be grouped into three categories, such as loops in L3, L4 and L13, β-hairpins in L22, and a long α-helix in L20, each playing a distinct function [159]. Furthermore, the authors emphasized that the α-helix–coil transition observed in L20 is absolutely required for the early steps of the large particle assembly [159].

A systematic bioinformatics analysis of 3,411 ribosomal proteins from 32 species revealed that the majority these proteins contain functional IDRs. In fact, in the ribosomal proteins across the three domains of life, the average disorder content (i.e., the fraction of disordered residues) was shown to range between 36 and 37.4%, which is noticeably higher than the overall disorder contents evaluated for various eukaryotic, bacterial, and archaeal proteomes, which were estimated to be 18.9%, 5.7%, and 3.8%, respectively [149]. Furthermore, most of the proteins in the S. cerevisiae ribosome are expected to undergo disorder-to-order transition; i.e., they are disordered in the unbound forms and fold upon complex formation [149]. Furthermore, the authors indicated that in the yeast ribosome crystal structure, many ribosomal proteins contained long stretches of residues with missing electron density, indicating their disordered nature [149].

In summary, L2 and many other ribosomal proteins are not only very important for ribosome assembly, but also play crucial roles in core ribosome functions, in both, prokaryotic and eukaryotic systems. Here, owing to their flexibility and co-folding capabilities, IDRs seem to be a prerequisite for the formation of a functional ribosome complex which is also nicely illustrated by the work of Timsit et al., who proposed distinct roles of IDRs during early steps of the 50S ribosomal subunit assembly [159]. Compared to prokaryotic ribosomes that are discussed here, eukaryotic systems contain ribosomal proteins that encompass even larger disordered regions, which might serve additional auxiliary functions, e.g., in regulation of the translation process [149].

2012 (Chemistry): G-protein coupled receptors

In 2012, the Noble Prize for Chemistry was awarded to Robert J. Lefkowitz and Brian K. Kobilka for their contribution to the understanding of the functionality of G-protein-coupled receptors (GPCRs). GPCRs form the largest family of membrane proteins, which act as receptors to a wide variety of ligands, e.g., hormones and neurotransmitters, making GPCRs one of the most important and attractive therapeutic drug targets. The main characteristic of GPCRs is the presence of seven α-helical membrane-spanning regions connected by alternating extracellular and intracellular loops [162]. Lefkowitz, together with his team, unraveled important questions related to the β2-adrenergic receptor (β2AR). The β2AR belongs to class A of the GPCR family and is abundantly present in the airway smooth muscles of the respiratory system. Upon the activation by the ligand epinephrine, a downstream signaling pathway is triggered which leads to bronchodilation [163]. In 2007, Kobilka et al. solved the crystal structure of inverse agonist carazolol-bound human β2AR in complex with Fab5 (antigen binding fragment of monoclonal antibody 5) in a lipid environment [164]. Later, they solved the crystal structure of β2AR in complex with a nucleotide-free Gs protein [165].

According to D²P², the consensus disorder content of β2AR is ~ 15%, with IDRs mainly located at the C-terminus, N-terminus, and the intracellular loop 3 (ICL3) of the receptor. The same regions have also been found to be most challenging for crystallization. The IDRs in GPCRs give plasticity to the receptors and allow them to bind to a wide range of ligands and other proteins. As aforementioned, IDRs are favored sites for PTMs that control receptor coupling to other cytosolic proteins during signaling [166]. IDRs found in different GPCR classes are subjects of alternative splicing generating variable forms of the receptors that are each capable of distinguishing between different intracellular ligands owing to the structural uniqueness of the ICL3 [167].

The ICL3 of β2AR connects the cytoplasmic ends of transmembrane helices TM5 and TM6, allosterically linking the extracellular region of the receptor to the intracellular region. The agonist-bound, closed form of the receptor corresponds to an open conformation of ICL3, which easily binds G-proteins. The antagonist binds to the open form of the receptor, forcing the ICL3 to adopt a closed conformation, to which G proteins fail to bind, and, as result, no signal transduction occurs. The mobility of the ICL3 allows it to dominate the interaction between the intracellular region of the receptor and other interactive cytoplasmic proteins [168]. Additional studies have shown that, besides ICL3, the disordered regions in the C-terminus of the receptor bind to cytoplasmic interaction partners, undergo a disorder-to-order transition, and adopt a structure that helps it to bind to other proteins downstream in the signaling pathway with low affinity and high specificity [169]. Out of all the cytoplasmic interaction partners of β2AR, the heterotrimeric G-protein, β-arrestin, and GPCR kinases (GRK) are the most well known. GRK phosphorylates the agonist-bound β2AR at serine and threonine residues within the 355–364 region. This phosphorylation event facilitates high affinity interaction of β2AR with β-arrestin, leading to the receptor internalization [170]. Furthermore, IDRs of β2AR play a significant role in its lysosomal degradation. Involved sites for ubiquitination have only been found within the ICL3 (K263, K270) and the C-terminal part (K348, K372 and K375) [171]. Even though the role played by IDR within the C-terminus is well understood, the extracellular N-terminus of the GPCR is most underexplored for its role in receptor functioning. Based on the analysis of intrinsic disorder predisposition as well as data from various literature sources, it can be concluded that the N-terminus of the receptor contains IDRs. The N-terminus of β2AR is N-glycosylated, which helps in receptor trafficking through the endoplasmic reticulum, the Golgi body, and membrane insertion.

It should be also emphasized that comprehensive bioinformatics analysis revealed that structural disorder is crucial for functionality of arrestins [172]. More globally, based on the bioinformatics analysis of more than 800 various human GPCRs and a large set of heterotrimeric G proteins, it was concluded that the multifunctionality of GPCR-G protein system depends on intrinsic disorder of corresponding proteins. Here, structures of the involved proteins represent a complex mosaic of differently folded/disordered regions, and the functionality of resulting highly dynamic conformational ensembles can be fine-tuned by various PTMs and alternative splicing. Furthermore, these conformational ensembles can undergo dramatic changes at interaction with their specific partners, thereby indicating that the multifunctionality of GPCR-G protein system represents an illustrative example of the protein structure–function continuum [173].

In conclusion, GPCRs are known to possess IDRs, which vary significantly among different receptor classes, giving them a dynamic and flexible nature that is important for carrying out their distinct functions.

2012 (Physiology or Medicine): induced pluripotent stem cells

In 2012, the Nobel Prize in Physiology or Medicine was awarded to John Gurdon and Shinya Yamanaka, for their work on reverting adult cells back into the embryonic state. In the 1960s, Gurdon showed that nucleus transplantation from intestinal epithelial cells of Xenopus laevis tadpoles into enucleated eggs results in normal feeding tadpoles, with a higher chance of successful development into adults, when the nucleus is taken from embryonic cells compared to adult cells [174]. These findings demonstrate the existing resistance against a developmental “step-back” that needs to be overcome in order to yield pluripotent cells. These resistance-regulating factors became later to be known as transcription factors (TFs, see above), and ~ 1600 TFs are now known to be encoded by the human genome. Over time, it was realised that more than 80% of these proteins harbor long regions of disorder [86]. In early 2000s, Yamanaka and his group showed that only a handful of TFs are needed to induce pluripotent stem cells (iPS) from adult fibroblasts [175]. The “Yamanaka factors” include Sox2, Oct3/4, Klf4 and c-Myc. At the same time, Thomson and coworkers used Sox2, Oct3, Lin28 and Nanog to transform human somatic cells into iPS cells [176]. Even though iPS have been under intense scrutiny since then, the underlying mechanism of cell reprogramming remains elusive. This might be partly due to the absence of detailed structural information of the TFs, which in turn is linked to their high disorder content [177]. Here, we focus on what is known about intrinsic disorder and how it influences the function of the prominent TF Sox2.

The name Sox is an abbreviation for “Sry-related HMG box.” A homology search, using the sequence of the TF Sry (Sex-determining gene on Y-chromosome), revealed an 80-amino acid region named High Mobility Group (HMG) that was identified as DNA-binding domain. Aside from Sry, the authors described four additional HMG-box containing proteins with an identity of 60% to Sry that were named Sox1-4 [178]. Today, 20 Sox proteins are known, all of which have the HMG domain in common.

A bioinformatics analysis of 35 proteins involved in generating iPS cells or found to be overexpressed in embryogenic stem cells has already discussed the importance of intrinsic disorder [177]. This study used disorder prediction on the sequence (PONDR^® VLXT and PONDR^® FIT) as well as engines predicting disorder status of the entire protein (CH-CDF plot; Fig. 2). Thirteen proteins were found to be mostly structured, 17 to be likely molten-globules and 5 to be mostly disordered, with Sox2 being one of the latter ones. MobiDB incorporates several disorder prediction tools and predicts the overall disorder content of Sox2 to be 27% [138], while D²P² database predicts a 53.6%.

The limited structural information on Sox proteins is restricted to their DNA-binding HMG domain. It exhibits three structurally conserved α-helices, as well as extended N- and C-terminal segments, which show an L-shaped arrangement. The two “legs” of the L-shaped surface bind to the minor groove of the DNA, while significant bending of the DNA is observed upon binding of the HMG-box. The highly flexible C-terminal tail of the Sox-HMG-box builds several DNA contacts, snuggles deep within the minor groove, and undergoes a disorder-to-order transition [179]. It is important to note that DNA-sequence-specific HMG-box containing TFs, like Sox2, only harbor a single HMG-box, while non-specific TFs have two or more HMG-boxes. In their free state, specific HMG-boxes show a disordered C-terminal tail, while non-specific HMG boxes are structurally well-defined which was nicely reviewed by Weiss [180]. DNA-binding specificity is largely achieved by the ordering of the C-terminal segment that is facilitated by highly conserved amino acids [179]

It is further noteworthy that the HMG-box of Sox2 harbors both nuclear export and nuclear localization signals (NES and NLS, respectively), with the latter being found in the flexible C-terminal tail of this protein [181]. Besides these features ensuring proper subcellular localization, Sox2 is a target for secondary modifications. Interestingly, these occur mostly within the NLS stretch of the C-terminal tail of the HMG box as well as the very C-terminally located transactivation domain of the full-length Sox2 [182]. It is also of note that, apart from the HMG domain, there is no evidence for a defined secondary or tertiary structure in Sox2. The PTMs are all located within disordered regions. It was experimentally shown that these non-HMG regions have also major impact on selective protein interactions depending on cell type and developmental stage [183]. These interactions stabilize the DNA–protein complex and regulate the Sox activation, resulting in transcriptional activation or repression.

In summary, even though Sox proteins have been extensively studied due to their role in development and cancer, our understanding of their structural and topological features remain largely elusive. Analysis of Sox proteins in complex with their binding partners will reveal whether there is a disorder-to-order transition of the flexible Sox2 termini, as it is true for the flexible C-terminal part of the HMG-box. The flexible regions in the Sox proteins facilitate nuclear import, DNA-binding specificity and stabilization of complexes due to the introduced PTMs. Their further investigation in molecular detail promises to substantially advance our understanding of key mechanisms in fundamentally important transcriptional regulation events.

2013: vesicular trafficking

In 2013, Randy W. Schekman, James E. Rothman, and Thomas C. Südhof were awarded the Nobel Prize in Physiology or Medicine for unravelling the molecular principles behind vesicular trafficking. While Schekman discovered the responsible genes by analyzing several yeast mutants with defective transport machinery [184], Rothman revealed the equivalent mammalian protein machinery, identifying the system that fuses vesicles with their corresponding target membrane resulting in the transmission of cargo molecules [185]. Interested in the communication of nerve cells, Südhof uncovered that membrane fusion is a highly signal-dependent and protein-regulated process that occurs in synapses [186].

Cellular transport is mediated by a trafficking cycle of cargo-loaded vesicles. The latter are formed from a source membrane by endocytosis, followed by cargo uptake and migration to the target membrane. The vesicles then dock at the target membrane and upon activation fuse with the membrane and thereby release the transported cargos. Subsequently, the vesicles are regenerated by endocytosis [187]. Proteins involved in these secretory pathways are often structurally disordered, highlighting the importance of disorder for these processes [188].

For instance, complexins represent a family of small soluble, highly charged and evolutionary conserved neuronal proteins that co-localize with the SNARE (soluble N-ethylmaleimide-sensitive-factor attachment receptor) proteins, the key players in membrane fusion that guide vesicle and target membrane into close proximity [189]. The tight binding of complexin-1 to the core SNARE complex regulates membrane fusion by promoting or inhibiting SNARE assembly [190]. However, the underlying mechanisms are still elusive. Complexins are composed of an unstructured N-terminal domain (NTD), a flexible accessory helical domain (AH), a central α-helical domain (CH) and a highly disordered C-terminal domain (CTD) [190]. Isolated complexins are mostly unstructured although they contain the aforementioned helical region in solution [191]. This helical region binds in an antiparallel manner to the groove formed by interactions between synaptobrevin-2 and syntaxin-1 within the core SNARE complex [190, 191].

The disordered NTD of complexin-1 is located at the membrane proximal part of the SNARE core complex, consecutive to the AH domain. Two conformations for complexin-1 bound to the SNARE complex are proposed. (i) In trans-conformation, complexin-1 associates via the CH domain with the SNARE complex exposing the NTD and AH domains for other interactions [192]. This exposure enables the AH domain to weakly interact with an additional binary SNAP25:syntaxin-1 (1:1) complex, most likely stabilizing this binary acceptor complex for synaptobrevin-2 in early stages of the fusion mechanism [192, 193]. The free NTD, due to its amphipathic character, might now form a helix, thereby possibly interacting with the vesicular membrane [194]. (ii) In cis-conformation, on the contrary, the AH domain is overlaying and interacting with the SNARE complex, while the NTD clamps the C-terminal part of the SNARE complex. Accordingly, both domains induce conformational changes of the C-terminal SNARE core related to a conformation juxtaposing synaptic vesicle and presynaptic membrane, but still not able to forward the generated force [192, 195]. Consequently, spontaneous membrane fusion is captured as long as complexin-1 is not replaced by the Ca²⁺-activated sensor protein synaptotagmin-1, thereby facilitating Ca²⁺-ion triggered release of neurotransmitters [196].

Using NMR, the CTD of complexin-1 was shown to be intrinsically disordered [191]. It further contains tandem lipid-binding motifs, called amphipathic helix (AH1 and AH2) motifs, as well as a C-terminal (CT) motif. The AH motif contains an amphipathic helix including a cluster of negatively charged residues (AH2) that precedes a central serine (Ser115), followed by a second short amphipathic helix (AH1) terminated by a proline [197]. In the presence of lipids, a disorder-to-order transition of the AH motifs leads to the formation of the amphipathic helices which are then able to bind a lipid bilayer. As a result, complexin’s CTD sensitively recognizes the membrane curvature of the vesicle [197]. Membrane binding is mediated through the CT motif by insertion of its three hydrophobic phenylalanine residues, while maintaining disorder [197]. Taken together, the CTD of complexins inhibits spontaneous neurotransmitter release, while actively promoting Ca²⁺-dependent membrane fusion by localizing the ternary SNARE complex in close proximity to synaptic vesicles.

Although neither the energetic, nor the mechanical processes of membrane fusion are fully understood, it still becomes apparent that the disorder in CTD and NTD of complexin-1 and its transition to ordered structures are essential to promote a critical cellular function, such as vesicle fusion, by providing the flexibility required to facilitate opposing functions, i.e., promotion and inhibition of vesicle fusion.

2015: DNA-repair

The 2015 Nobel Prize in Chemistry was awarded for the fundamental work on the mechanistic studies of DNA repair and the potential development of novel cancer treatments in cells [198]. The prize was shared by three scientists, Tomas Lindahl, Paul Modrich, and Aziz Sancar. A common trend that is shared by DNA repair proteins is the presence of IDRs. Due to a significant percentage of disorder in these proteins, the main focus here will be on uracil DNA glycosylase (UNG), a protein that is active in base excision repair. The identification of E. coli UNG was first reported by Lindahl in 1974. Strikingly, over 70% of UNG’s amino acid sequence are conserved between E. coli, yeast, and human, demonstrating an extensive homology [199].

DNA repair is essential in all organisms, as it is estimated that more than ten thousands base lesions and single-strand breaks are induced daily in a mammalian genome [200]. With base excision repair, a misincorporated base is removed, leaving an abasic site that is then processed by short-patch or long-patch repair [201]. Deamination of cytosine can result in uracil/guanosine (U:G) mispairing that is commonly observed as mutation making the DNA strand unstable and in need of repair [202]. One of the most notable differences between mammalian and bacterial UNG is the presence of disorder at the N-terminus [203].

The amino acid sequence of human uracil DNA glycosylase (hUNG2) comprises a 227 amino acid-long, compact, catalytic domain and a 86-amino acid-long intrinsically disordered N-terminal domain [204]. Several experiments have been conducted comparing the full-length hUNG2 with the protein comprising only the catalytic domain without disordered N-terminus (hUNG2_cat). Despite showing minor effects on hUNG2’s enzymatic activity, the intrinsically disordered N-terminal domain contains a portion of the nuclear localization signal (NLS). To facilitate removal of uracil bases, hUNG2 has to be attached to the DNA strands. The DNA binding affinity proves to be crucial, as formation of the base excision repair complex can be completed and the process of uracil base excision can be initiated [205]. Sixfold and threefold decreases in the ssDNA and dsDNA binding affinities were observed in hUNG2_cat upon removal of the disordered region [206]. A regulatory role by PTMs is also mediated by the disordered N-terminal domain. The PTM sites include phosphorylation for tyrosine, serine, and threonine [207], and acetylation for lysine [204] that have a great impact on hUNG2’s activity, such as DNA translocation efficiency and selectivity of uracil excision. Proteins that act at specific DNA sequences often bind DNA randomly before they translocate to the target site, a process that is known as associative and dissociative translocation [208]. Experiments with translocation assays have shown that not only does the disordered tail boost the translocation efficiency; it is an essential component for facilitating DNA translocation, as the catalytic core alone fails to perform. The kinetic parameters show a fourfold increase in the translocation activity with the addition of the 86-residue disordered tail [204]. Moreover, in contrast to the full-length protein, a hUNG2_cat with only the catalytic core does not have the selectivity for uracil excision. Comparing the steady-state kinetic parameters, the binding affinity of hUNG2_cat shows a 20-fold decrease compared to the full-length hUNG2, confirming the enhancement of binding that is specifically aided by the presence of the N-terminal disordered tail [206].

The IDRs are responsible for various important functions of a protein. In comparison with the ordered regions, IDRs are characterized by specific biological advantages that are not achievable by a fixed structure. The function of protein is not exclusively determined by its folded domains. Instead, the IDRs have significant effects regarding the thermodynamic and kinetic properties that affect the whole protein, as shown in the case of hUNG2.

2016: autophagy

In 2016, Yoshinori Ohsumi received the Nobel Prize in Physiology or Medicine for his studies on autophagy in yeast. He showed that yeast lacking certain degradation enzymes accumulates autophagic bodies in their vacuoles under starvation conditions [209]. In further studies, using autophagy-defective yeast mutants, he discovered many genes and proteins that are essential for autophagy. Autophagy itself is a fundamental intracellular degradation process, evolutionary conserved in all eukaryotes [210]. It is responsible for the bulk turnover of proteins as well as whole cell organelles. Thereby, autophagy plays a major role in cell survival as a main response to the starvation conditions and by degradation of potential harmful macromolecules. Besides its non-selective function, different types of selective autophagy have been reported. One of the best studied autophagy-like mechanisms is the Cvt pathway in yeast, in which the hydrolase aminopeptidase 1 (Ape1) is transported into the vacuoles. Depending on the cargo, selective autophagy can be further distinguished in mitophagy, reticulophagy, pexophagy or xenophagy [211]. Hence, autophagy contributes to many physiological processes, such as intracellular quality control, cell differentiation, and development, infection response or anti-aging, and is therefore a key player of cellular and organismal homeostasis [212]. Defects in autophagy are therefore frequently linked to neurodegenerative diseases and cancer, although most of underlying mechanisms remain unknown [213]. The process of autophagy in yeast consists of five steps: induction, nucleation, expansion, fusion and cargo degradation [214]. During induction, several autophagy-related proteins (Atgs) assemble in the Atg1 kinase complex to organize the pre-autophagosomal structure (PAS) and recruit further Atgs. The PtdIns3K complex is required to start the nucleation of a vesicle structure called phagophore [214]. Expansion of the phagophore is then processed by the ubiquitin-like conjugation systems of Atg12 and Atg8 until the target of autophagy is fully enclosed by a double membrane. This mature vesicle, the autophagosome, docks and fuses with the vacuole releasing its inner part into the lumen of the vacuole where the autophagic bodies are lysed and eventually degraded by hydrolases. Despite the identification of over 30 Atgs in yeast, 18 are considered to form the core autophagy machinery [215]. One of the key players of autophagy is the protein Atg13. It is a part of the multimeric Atg1 kinase initiation complex containing Atg1, Atg13, and the Atg17-Atg29-Atg31 subcomplex [216]. Atg13 consists of a structured N-terminal HORMA domain (amino acids 1–268) and a long C-terminal IDR (amino acids 269–738) [217]. This IDR accounts for almost two-thirds of the protein and is usually highly phosphorylated by TORC1. Under starvation conditions, the serines get dephosphorylated, enabling the interaction of this protein with Atg1 and Atg17.

Atg1 is a serine/threonine kinase that binds to the MIT-interacting motif (MIM, amino acids 460–521) within the IDR of Atg13. However, Atg1 has an IDR itself, linking the N-terminal kinase domain and the C-terminal MIT domains. Furthermore, the MIT2 domain is at least partially disordered when it is not bound to Atg13 [218].

The interaction of Atg13 with Atg17 takes place via two distinct IDR binding sites, the Atg17-binding region (Atg1317BR, residues 424–436) and the Atg17-linking region (Atg1317LR, residues 359–389). These two sites are around 130 Å apart from each other and can interact independently from each other. Furthermore, they bind to two different Atg17 molecules, thereby creating the multimeric structure of the Atg1 kinase complex. It is estimated that around 28 to 56 copies of Atg1, Atg13 and Atg17 are required to build the scaffold for the PAS [219]. The multimer is also necessary for the following intermolecular autophosphorylation of Atg1, achieved by close distances of the kinase domains. This in turn leads to the assembly and phosphorylation of Atg9 and enables the start of nucleation. Although the N-terminal region of Atg9 binds to the HORMA-domain of Atg13, the IDR of Atg13 seems to be the important part of this protein, especially regarding the assembly of the Atg1 kinase complex. The flexibility provided by the IDR is crucial for the versatile interaction with the other proteins. A third Atg17-binding region (Atg1317BR2, residues 641–661) is reported within the IDR of Atg13, but seems not to be essential for autophagy. The proteins Atg29 and Atg31, parts of the sub-complex with Atg17, are highly disordered as well, showing various potential binding sites within their IDRs. They undergo disorder-to-order transition upon interaction and seem to play, besides a structural function, an important role for regulating the Atg1 kinase complex [220]. Although not all proteins and mechanisms are identical, the general process of autophagy is conserved through all eukaryotes from yeast to plants and mammals. A version of Atg13 is also present in mammalian cells, mediating the initiation of autophagy similar to yeast, by interacting with Atg1 homologue ULK1 and FIP200, an analog of Atg17 [221]. Large parts of mammalian Atg13 are disordered as well, and the binding sites lie within long IDR (amino acids 384–517). Additional studies have shown that, similar to yeast, most of the human core autophagy proteins contain IDRs [222]. That indicates that intrinsic disorder is an important feature in autophagy. This is supported by the fact that many of the protein binding sites and most of the phosphorylation sites of Atgs lie within the disordered regions, providing distinct functional advantages. The IDRs can overcome steric obstacles, possess larger interaction surfaces, with high specificity and low affinity and undergo disorder-to-order transitions. Therefore, the IDRs enable a high conformational flexibility, can bind to many various interaction partners, and are therefore able to carry out a multitude of different functions. Furthermore, they show an increased interaction speed, ensuring rapid signal transduction and reactions to cellular stress, like starvation [217]. Conclusively, IDRs play a crucial role for a complex mechanism, like autophagy, and are widely distributed in autophagy proteins throughout eukaryotes.

2017: circadian rhythm

A few decades ago, scientists started to shed light on one of the biological processes that govern the biological periodic changes in the behavior and physiology of most species, the circadian clock. In 2017, Jeffrey C. Hall, Michael Rosbash, and Michael W. Young were awarded the Nobel Prize in Medicine or Physiology for their contribution to understanding the inner mechanisms controlling the circadian rhythm in Drosophila melanogaster. They first revealed that PER (period) protein levels fluctuated during the 24-h cycle and together with TIM (timeless) formed an inhibitory feedback loop to enter the nucleus and interact with the CLK/CYC complex bound to the promoters of per and tim genes and thus repressing their own expression. Furthermore, the discovery of DBT (doubletime) provided insights on how this oscillation is regulated over time [223].

The majority of circadian clock proteins contain a basic-Helix-Loop-Helix (bLHL) PAS domain (named after Per-Arnt-Sim protein). While the N-termini of these proteins are highly similar among species, the C-termini appear to include a high variability possessing IDRs that allow these proteins to perform their diverse functions [224]. PER contains two N-terminal PAS domains (PAS-A, PAS-B) that are well-structured and crucial for homo- and heterodimer formation to enable complex interactions in the circadian feedback loop [225]. However, the well-structured part of Drosophila PER (dPER) only makes up for 50 percent of the total amino acid content that is encoded in the gene. Half of the protein is partially unstructured or disordered. Several residues in these regions are susceptible to phosphorylation, affecting the stability of the dPER monomer, promoting proteosomal degradation when phosphorylated by DBT on Ser47, or preventing protein turnover by phosphorylation of the C-terminal region of the PAS-B domain that is often referred to as PER-SM and PER-SD [226]. Although the PAS domains found in dPER are conserved in mammalian PER homologs (mPER1-3), the C-terminus of dPER contains additional disordered elements, such as Gly/Thr tandem repeats. Although the functional role of these tandem repeats remains elusive, they show high similarity with the proteoglycans in vertebrates that contain Ser-Gly tandem repeats where the serines are glycosylated [227, 228]. The mammalian circadian clock follows the same feedback mechanism with a different complexity of the protein–protein interactions displayed in D. melanogaster. Period homologs mPER1-3 cooperate with mCRY1 and mCRY2 to form heterodimers that block mCLOCK:BMAL1-mediated transcription of the PER and CRY loci. The C-terminal part of mPER2 contains a CK1 phosphorylation site as well as a domain for CRY heterodimerization [229, 230]. Complex crystals composed of mPER2/mCRY1 revealed five α-helices from residues 1138–1198 followed by a loop structure up to Glu1214. However, residues 1215 to 1252 display no electron densities hinting toward a flexible and disordered region. On the other hand, circular dichroism (CD) spectroscopy-based analysis of the mPER fragment (residues 1132–1252) revealed a predominantly unstructured nature of this segment in the absence of mCRY1 [230]. Hence, the interaction of mCRY1 with the disordered C-terminus of mPER2 protein induces a helical structure to enable complex formation and facilitate inhibition on the mCLOCK:BMAL1 transcriptional complex.

Nevertheless, the precise mechanism by which the IDRs in BMAL1-TAD and CLOCK can influence the circadian timekeeping is yet to be elucidated. Further analysis of the disordered regions and related functions could be vital to fully understand the entire mechanism of the circadian clock. What once started with research of a basic feedback loop has evolved into a complex protein interaction network that controls the daily cycle and requires structured as well as unstructured domains to facilitate the intricate functions of the circadian rhythm.

2019: oxygen sensing

William Kaelin Jr., Peter J. Radcliffe, and Greg L. Semenza shared the 2019 Nobel Prize in Physiology or Medicine for their discoveries of how cells sense and adapt to oxygen availability [231]. Their work unraveled one of the most fundamental cellular processes, the mechanisms of oxygen sensing and the cell adaptation to different oxygen levels. While the importance of oxygen sensing and its implications are widely understood, the mechanism of how cells adapt to hypoxia and the resulting physiological effects remained unclear [232]. In general, oxygen is paramount to most of the cellular critical functions, from respiration, to energy production, and to various biosynthetic pathways, etc. Thus, a direct cellular response to the fluctuating extracellular as well as intracellular concentration of oxygen would require a highly sensitive circuitry. Kaelin, Radcliffe and Semenza discovered a system that perfectly fulfils this role, with the Hypoxia Inducible Factor 1 alpha (HIF-1α) as the lynchpin.

In normal oxygen availability conditions (normoxia), the cell functions normally, utilizing the tricarboxylic acid cycle (TCA cycle) for energy production. One of the TCA cycle intermediates, α-ketoglutarate, is exported to the cytosol as it acts as a precursor for both amino acid and fatty acid production [233]. An enzyme called HIF-prolyl hydroxylase (HIF-PH) uses oxygen, iron, and α-ketoglutarate as co-substrates in order to modify the HIF-1α by adding at least one hydroxyl group (hydroxylation) to a prolyl group, more specifically either Pro564 or Pro402, both residing in oxygen-dependent degradation motifs [234]. This modification results in the binding of the Von Hippel-Lindau protein (pVHL), which in turn earmarks HIF-1α for degradation through the ubiquitin–proteasome pathway, as pVHL is able to be recognized and form a recognition complex with the E3 ubiquitin ligase. In hypoxia conditions, HIF-1α is stabilized, leading to the formation of the complete HIF complex, which in turn migrates to the nucleus and binds to Hypoxia Response Element (HRE) areas, leading to the transcription of its associated target genes. Target genes include, but are not limited to, erythropoietin, Vascular Endothelial Growth Factor (VEGF) [235] and other genes that control cell proliferation and energy metabolism [236].

Recent studies have shown that various, highly interconnected, transcription factors draw their ability to interact with multiple partners by exploiting the flexibility of intrinsically disordered regions (IDRs) [9] allowing for a variety of conformations as well as increasing interaction space [79]. pVHL is also known to interact with more than 500 partners [237], playing a critical role in multiple pathological conditions [238]. Along with the important role of HIF-1α described above, these two facts lead to the hypothesis that pVHL and HIF-1α also possess intrinsic protein disorder. Both proteins are major players of oxygen sensing and play a role in various diseases, which make them important for research in the medical field as well. Drug development for anemia, development of the brain in embryos, cancer and exercise adaptation, immune response and wound healing are just some of the physiological pathways these two play a role in [238].

Firstly, in the case of Homo sapiens pVHL, the consensus disorder prediction reveals the presence of a long IDR (corresponding to 30.5% by MobiDB, and 32.4% by D²P² of the total sequence) that covers the N-terminal segment of the protein. Although all interaction sites are located in the ordered region of the protein., a recent study has revealed that due to its inherent plasticity, the N-terminal region can adopt different conformations to differentiate between isoforms of the protein, acting as a flexible interaction interface and providing additional binding specialization that enables dimerization [239]. Interestingly, similar analysis for HIF-1α from Homo sapiens revealed that even though total disorder content is predicted to be low (12.6% by MobiDB, and 32.8% by D²P² of total sequence), critical interaction sites, such as hydroxylation site at Pro564 and pVHL interaction sites at the C-terminus are located in ordered domains that are connected to or lie between IDRs. This initial prediction, along with sequence annotation, leads to the observation that the flexible IDRs, which have also been categorized as “entropic chains” [9], can facilitate the interactions taking place at these sites by adopting multiple transient conformations, adapted to each different partner, probably governed by surface electrostatic interactions. This prediction is also supported by the fact that no complete HIF-1α structure exists in the PDB as IDRs could not be resolved due to their intrinsic flexibility.

2020: CRISPR/Cas9

The 2020 Nobel Prize in Chemistry was awarded to Emmanuelle Charpentier and Jennifer A. Doudna “for the development of a method for genome editing” [240]. The discovery of the CRISPR-Cas9 (clustered regularly interspaced short palindromic repeats—CRISPR-associated protein 9) system hallmarked the start of a new era in gene-editing technology, with numerous applications in biotechnology, agriculture [241], and medicine [242].

The studies of bacterial responses to nucleic acid invasion were instrumental in the discovery of this gene-editing system. Adaptive immune defense systems are not solely a higher-order organism trait. Bacteria and archaea have evolved their own systems to protect against external invasions by phages and other unwanted external nucleic acid elements (e.g., plasmids) called CRISPR [243]. This system, similar in function to the RNA interference (RNAi) pathways of eukaryotic organisms, uses small RNAs as guides to target and silence invading foreign nucleic acids with high specificity [244] with the help of the system’s main effectors, the CRISPR-associated proteins (Cas). In short, the system is consisted of three phases: first, the adaptive phase, where CRISPR-utilizing organisms will integrate foreign sequence fragments, called protospacers, at specific loci of the CRISPR genetic array, followed by transcription of the integrated spacer elements into what is called precursor CRISPR RNAs (pre-crRNAs) at the second step, and finally, at the third step, the pre-crRNAs are cleaved into short crRNAs that will hybridize with protospacer sequences on the foreign nucleic acids and activate targeted silencing by the Cas proteins [245]. There are three main CRISPR systems that have been widely characterized in procaryotic organisms so-far, namely type-I, type-II, and type-III, with the type-II system being the most prominent. In the type-II system, Cas9 will drive the silencing of the target by RNase III with the help of a trans-activating crRNA (tracrRNA) that is complementary to the pre-crRNA repeat sequences [246].

From the moment of its discovery, the power of the type-II CRISPR system was quickly recognized and put to use for gene-editing applications. With the use of two short RNA (crRNA and tracrRNA) sequences as templates, highly specific targeting of selected genomic regions can be achieved with the use of the type-II CRISPR-Cas9 system, making possible precise edits to the target protospacer DNA, with specificity further enhanced by the presence of a short motif introduced, adjacent to the target sequence, called PAM (protospacer adjacent motif) [247].

Cas9, the protagonist of the CRISPR-Cas9 system, is a versatile enzyme that can be adapted to multiple biological applications [248]. The crystal structure of the Streptococcus pyogenes Cas9 (SpyCas9) enzyme itself, consisting of 1,368 amino acids, has been solved [249]. Analysis of this structure reveled that SpyCas9 is comprised of two “lobes,” an RNA recognizing lobe (REC) and a nuclease (NUC) lobe, predicted to contain multiple specific domains. The NUC lobe includes two nuclease domains (HNH, which is an endonuclease domain named for characteristic histidine and asparagine residues, and RuvC, which is an endonuclease domain named for an E. coli protein involved in DNA repair) and a C-terminal domain that give the nuclease activity to this lobe forming a catalytic core and another large α-helical domain lobe preceded by an Arg-rich region, acting as a linker between the two lobes, with an additional disordered linker comprised of residues 714–717 providing overall conformational flexibility. A topoisomerase II-similar domain is also located in the C-terminus at the nuclease lobe.

Comutational analysis revealed that the overall disorder content SpyCas9 (UniProt accession number Q99ZW2 [245]) is below 3% as predicted by a 75% consensus of nine disorder predictors. This protein contains two IDR and < 2% of its amino acids participate in MoRFs (molecular recognition features, which are the disorder-based protein–protein binding sites) [9] (Fig. 2). More specifically, the analysis by multiple disorder prediction algorithms reveals two main regions that are consensus-predicted as disordered, residues 766–783 and 853–871, both located in the HNH domain, flanking the active site. In the same region, the two MoRFs are also predicted, spanning residues 809–820 and 882–890. These observations may point to a the presence of binding regions engaged in disorder-to-order transition of the HNH active site during nucleic acid binding, allowing higher adaptability during the process. It is important to highlight the importance of the disordered N-terminal loop, residues 853–871, that flanks the α-helical element of the ββα-metal fold in the HNH domain. This predicted-as-disordered region grants higher conformational flexibility to the abovementioned fold, which contains an inward-facing asparagine (N863). This residue was observed during MD simulations of the region [250] after filling out the missing flexible regions of the published crystal structures with homology-derived models. The disordered N863 coordinates the neighboring Mg²⁺ ion and additional water molecules in the area, forming a cluster that may be crucial for the activation of the catalytic center that performs tDNA cleavage. Another analysis by Uversky et al. delves deeper into additional flexible regions, such as the role of the Arg-rich region located at the N-terminus and its inherent flexibility, which may assist in nucleic acid binding [251].

2021: thermal and mechanical transducers

The Nobel Prize in Physiology or Medicine in 2021 was awarded to David Julius and Ardem Patapoutian for their discoveries of thermal and mechanical transducers. David Julius identified the molecular target of capsaicin, the spicy and hot ingredient of chili peppers. Using a cDNA library from sensory neurons, he identified a transmembrane ion channel, today known as TRPV1 [252]. It belongs to the family of transient receptor potential ion channels and was shown to be activated by temperatures perceived as painful.

Following the discovery of TRPV1, both David Julius and Ardem Patapoutia independently discovered TRPM8, a cold-sensitive receptor, which also belongs to the transient receptor potential (TRP) channel family [253]. This receptor, also known as the cold and menthol receptor (CMR1), gets activated by menthol molecules, which elicit a sensation of non-hazardous cooling. Studies in transgenic mice lacking functional TRPM8 demonstrated that the channel plays a critical role in the detection of cool, innoxious temperatures [254]. Patapoutia later also found the mechanically activated ion channels PIEZO1 and PIEZO2, using a functional screen of candidate genes expressed in a mechanosensitive cell line [255]. The impact of the discovery of PIEZO proteins as excitatory ion channels directly gated by mechanical force can hardly be understated, as it provides the first example of a molecular basis for mechanosensation. The Nobel Prize was awarded for the invaluable insights into how we sense the world through somatic sensation and the proteins (TRPV1, TRPM8, PIEZO1 and PIEZO2) involved will be described below.

The transient receptor potential cation channel subfamily V member 1 (TRPV1) exists as a homotetramer, where each subunit, whose N- and C-termini are located intracellularly, encompasses six transmembrane segments [256]. The N-terminus of TRPV1 contains six ankyrin repeats [257]. The ankyrin repeat domain (ARD) is connected with the first transmembrane segment via an N-terminal linker domain, which interacts with the last two ankyrin repeats from an adjacent TRPV1 subunit [258]. A pore-forming loop containing an outer pore ‘turret,’ a segment that protrudes out of the outer pore region of channel and includes a pore helix and a selectivity filter, is located between the transmembrane segments S5 and S6 [259]. The complex gets activated by various mediators, such as capsaicin, temperature (43–52 °C) or acidic environments (protons) [252, 260, 261]. The function of the whole channel is largely dependent on amino acids in well-folded structural elements, such as the helix in segment S4-S5, where a 1 amino acid change affects capsaicin sensitivity of different species [262]. Nonetheless, IDRs also seem to be important. It is suggested that the N- and C-terminal IDRs of TRPV1 increase the receptor’s effective reach from 3–4 nm to ~ 10 nm [263]. Also, the disordered C-terminus seems to play a major role for interaction with calmodulin, a calcium binding regulatory protein, and mutations have been linked to pathological conditions in other TRP family members [264, 265].

The PIEZOs count only two members in human, PIEZO 1 and 2, which are large proteins of over 2500 residues each. Both the proteins are gigantic channels existing as homotrimers/homotetramers with 120–160 transmembrane segments. This striking particularity of having large number of transmembrane segments make them structurally unique from other known ion channels. Further, the PIEZO monomers are predicted to span the membrane between 30 and 40 times [266]. The mechanically activated PIEZO ion channels are vital for mechanotransduction in a wide variety of cell types, since their discovery in 2010 [267]. In tissues or cell types, PIEZO1 bring out shear stress and stretch-induced transmembrane currents mainly in nonneuronal cells, such as erythrocytes, vascular endothelial cells, bladder urothelial cells, and chondrocytes. PIEZO2 is mainly expressed in sensory neurons and is responsible for proprioception, the sensing of light touch and the noxious mechanical stimuli [268, 269].

The interesting feature of PIEZO channels is their intrinsically disordered regions. PIEZO2 is made up of seven large intracellular IDRs. The IDRs that serve as linker between adjacent transmembrane helical units (THUs) include regions 1, 4, 5, 6, and 7; disordered regions 2 and 3 connect THU8 with the clasp domain and the clasp with the latch domain, respectively. To determine the role of IDRs, various mutants with deletion of IDRs were studied, and the two significant observations were made: 1) Deletion of IDR3 leads to the disruption of the adjacent beam-to-latch linker, which causes increase in the single channel conductance of PIEZO2. 2) Removal of IDR 4 accelerates the stretch-sensitivity, while removal of region 5 deaccelerates the poking-sensitivity [270].

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

AKD conceived the idea. LP, MK, AS, and VNU wrote and assembled the manuscript. LP and VNU designed figures. All involved students of the RTG 2467 contributed specific parts to the final manuscript. All authors give their consent for submission and publication of this manuscript.

Funding

All authors acknowledge financial support by the DFG (RTG 2467, project number 391498659 ‘Intrinsically Disordered Proteins—Molecular Principles, Cellular Functions, and Diseases’).

Availability of data and material

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.

Declarations

Conflict of interest

The authors declare no conflict of interest.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.