*2.4. Characteristics Required for an Integrated Response to Simultaneous Challenges*

In order to be able to integrate responses, a hub must be capable of multiple interactions with various partners and collect signatures from various input pathways that can then be coherently interpreted. A high number of flexible and reversible interactions, and ability to make subtle changes to the equilibrium between the various states of control, would constitute a further advantage for such a hub. As these requirements match key characteristics of disordered regions it has been recognised for some time that such hub proteins are highly enriched for disorder [38]. Epiproteomic modification to MoRFs or neighbouring sites in a disordered platform could either block interactions, block other PTM changes at the same site (e.g., acetylation of a serine residue sometimes phosphorylated), or change the charge profile in disordered regions that then changes the dynamic of how (and/or whether) a MoRF interacts with a specific partner. While a degree of subtlety is important sometimes, a hard on-off switch will be important at other times. Disordered proteins can also undergo major conformational switches and even these can be linked to epiproteomic changes by adding larger modifying groups (e.g., glutathionylation, or AMPylation), isomerisation events around critical prolines, or by targeting the entire protein for proteolytic degradation for example.

In a recent analysis of human cells, Chavez and colleagues [39] used novel protein cross-linking methods combined with mass spectrometry to directly identify PTM decorated proteins that are physically associated with each other in complexes. New software advances have also enabled data analysis to focus on cross-linked peptides [40]. In cross-linking analyses distance constraints can be imposed by the type of chemical linker arm used, while addition of biotin groups permits enrichment for cross-linked fragments (e.g., by using avidin-mediated affinity capture technologies). Although the majority of cross-linked peptides identified were derived from homo-dimer interactions, acetylated and methylated peptides from core histone proteins participating in hetero-dimers were particularly common in this analysis. Almost half of the cross-linked histone peptides were found to contain at least one PTM event. Histones are known carry a number of highly significant PTM events. The multiple cases of linkages found between specific peptides increases the likelihood that these have biological relevance in terms of the protein interaction zones between the partners. Interestingly many of the cross-linked peptides with PTM contained modified lysine or arginine residues (residues that are also particularly enriched in disordered regions of proteins). Cross-linking sites were common in the disordered N- and C-termini of histones. In fact, it has been known for more than two decades that the histone tails are the sites where some of the most significant PTM takes place and that these modifications play key roles in the formation or dissolution of chromatin remodelling complexes. These tails serve as recognition sites for chromatin assembly as well as the assembly of the multi-component transcription machinery [41]. The largely positively charged disordered N-terminal tail also contributes to inter-nucleosome binding by contacting an acidic patch in the structured component of histone H2A/H2B dimers to influence histone stacking [42].

The cross-linking analysis allowed Chavez and colleagues [39] to build a significant interactome network map and highlights the importance of the combination of disordered regions and PTM to interactions in such networks, the hub position occupied by histones and the importance of their lysine/arginine rich disordered tails to drive their ability to organize into multi-component complexes. Other biophysical methods of experimentation can also provide indications of how closely associated proteins are in vitro or in vivo. Hydrogen-deuterium exchange (HDX) provides a measure of how exposed different parts of proteins are to PTM [43]. Changes in HDX patterns upon binding with

partners can indicate likely interaction zones in protein complexes and were initially used to map antibody binding sites [44]. Other techniques like Förster (fluorescence) resonance energy transfer (FRET) also lend themselves to analysing protein disorder. For example, Vassall et al. [45] used FRET measurements to analyse the order-to-disorder transition of the myelin basic protein (MBP). MBP is largely disordered in aqueous conditions but forms alpha helical recognition fragments upon binding to membranes and its protein partners. The MBP FRET studies, when combined with other tools used to probe structural transition in largely disordered proteins (such as circular dichroism and the membrane-mimetic solvent trifluoroethanol), yielded some surprising results. The data suggested that an intermediate conformational form between disorder and alpha helical state is in fact more compact than the alpha helical form (the latter would normally be expected to have more compactness). This longer form may provide a better bridge across to its complexing protein partners as well as facilitating faster binding to the membrane. Disorder-associated characteristics possessed by histone hubs allow them to integrate epigenetic marks with downstream modifications in mRNA expression response and transfer signals between the epigenetics and transcriptomics levels of response. The disorder properties of MBP on the other hand allow MBP to peripherally attach itself to the cytoplasmic membrane as well as interact with both cytoskeletal proteins like actin and signalling proteins that respond to Ca2+-triggered protein cascades.

One of the ways that cells coordinate their response to changing situation such as stress is to form recognizable sub-cellular organelles. Examples include stress granules (SG), processing bodies (P-bodies) and nuclear stress bodies. Such organelles do not contain membranes, a factor that differentiates them from permanent cellular compartments like the ER, nucleus and mitochondria. Functional organelles must be able to keep interacting with their surrounding liquid environment and yet they must have an ability to form an interphase boundary with this environment. In a recent review Uversky [46] suggests that disorder can provide a crucial component required for forming this liquid-to-liquid interphase. Examples of this are the role that the RNA-binding protein TIA-1 plays to promote assembly of SG through its disordered domains and the disordered regions of a number of the RNA-binding proteins found in human and yeast stress granules. The latter were found to be able to undergo liquid-liquid phase transition in vitro on their own, or when combined with RNA [47]. The phase separated droplets promoted by this organisation can then also recruit other proteins with disordered regions. Furthermore mutations in the key disordered regions or PTM sites involved in regulation can then lead to aberrant fibers or granules that may then contribute to neurodegenerative conditions.

#### *2.5. How Can Multiple Diverse Signals Be Coordinated in Real-Time?*

Responses need to be organised at both the temporal and spatial levels. An important biological question is how can organisms create control points that match such elaborate requirements? Significant PTM changes can be very rapid with response times measured in minutes as opposed to hours or even days for many other types of regulation responses [48]. Rapid response makes this level of regulation ideal for responding in real time to challenges perceived by the organisms. A successful reaction to stress is dynamic and requires both sequential, temporal and spatial separation of components and the ability to be nimble in response. The high degree of sophistication required by a successful response is elegantly matched with the opportunities offered by disordered platforms to rapidly integrate PTM signals through multiple MoRFs, multiple targeted PTM sites and reversible as well as competing PTM changes at particular sites. Moreover, PTM changes can also be spatially compartmentalised by limiting where matching substrates and enzymatic functions are co-expressed. As discussed above compartmentalisation can even be aided or driven by the ability of disordered proteins to contribute to phase transition in examples like stress bodies.

Importantly many PTM changes are reversible involving balancing modifications such as phosphorylation/dephosphorylation or acetylation/de-acetylation. Pathogens in turn interfere with PTM processes by developing modifications that can compete with these changes, e.g., phospholyase reactions that break a unique phospho-threonine bond in a protein kinase activation site and make this site un-available for re-phosphorylation [15]. The very properties that make PTM changes so dynamic also make this level of response technically very demanding to illustrate. In order to capture such dynamic potential, sampling time needs to be adjusted to a much finer timescale than commonly used. In addition, techniques that can capture protein associations in real time and are not affected by their readily reversible nature (such as cross-linking techniques) will be required. This will need to be matched with detection techniques sensitive enough to identify any PTM, yet robust enough to be able to scan across complex proteomes. Physical and software enrichment strategies that can overcome the challenge of these limitations in concert with much more sensitive mass spectrometry instrumentation have recently become available. I suggest that disordered protein regions in particular have properties that indicate they are likely to feature prominently in these novel analyses in the near future. Their dynamic ability to change their binding partnerships and to be decorated by multiple PTM events, as illustrated by the examples of p53, RIN4 and histones presented above, suggest that this is one of the major reasons that disorder has become such a common feature of proteins in complex multi-cellular organisms. Indeed this fits with the proposal that disorder was a key enabler on the road to multi-cellular lifestyles [31]. The ability of disordered regions to sense the physiological milieu in which they find themselves by a combination of PTM events, charge profiles and electrostatic interactions suggests that sensing change in this milieu is a specific biological niche that disordered proteins occupy.

#### **3. Conclusions**

The animal and plant exemplars, p53 and RIN4, show some key similarities. I suggest that their disordered platform is specifically designed to integrate diverse signals that arrive via alternate post-translational changes inside (or sometimes in close proximity to) MoRFs. PTM changes have a great deal of flexibility and can be very rapid and reversible (e.g., phosphorylation and de-phosphorylation) and the term epiproteomics evokes the dynamic nature of these changes. In addition to reversibility, PTM sites can be; locked in competitive battles (e.g., phosphorylation and acetylation [24]), display competition between sites (as illustrated by the RIN4 phospho-switch concept [33]), result in more subtle shifts in equilibrium (e.g., by changing the charge profile and flexibility of the environment around a MoRF), or result in drastic conformational changes suited to acting as a molecular on/off switch (e.g., by proline isomerisation or multiple phosphorylations). I suggest that the main role of the p53 and RIN4 (and probably many other) proteins containing large disordered domains is to act as sensors and integrators of stress signals from multiple distinct sources via changes to the epiproteome. Moreover, this could explain why examples such as RIN4 and p53 play such important roles in plant and animal disease respectively.

**Acknowledgments:** This research was funded by Discovery Science Grants to EHAR on intrinsically disordered proteins (DS-1166 and DS-2002) from The New Zealand Institute for Plant & Food Research Ltd. (Auckland, New Zealand). The author thanks Joanna Bowen and Xiaolin Sun for editorial suggestions.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


© 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*International Journal of Molecular Sciences* Editorial Office E-mail: ijms@mdpi.com www.mdpi.com/journal/ijms

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18