Based on what we read in the literature, I wish to make some considerations. The countless functions that IDPs show are closely related to their ability to interact. Although structurally fluctuating and disorganized, IDPs sometimes reorganize themselves by interacting with a protein partner. We don't have many details of these processes, but they depend not on two but on three molecular partners: IDP, its protein partner, and water. We have not to forget that the structure of a protein in solution is given by the balance between its internal interactions and those with the solvent molecules that surround it, all mediated by the existing rotational constrictions in the bonds of the backbone and of the lateral residues, and by the elevated intrinsic dynamics of water itself. Regarding the intrinsic interactions of liquid water and its solutes, there are still controversial positions on possible molecular models. However, with protein-protein interaction, it is necessary to consider a three-partners competition. The final equilibrium, that is the structural organization of the complex in water, whatever its dynamic properties will be, will depend on the global energy balance between enthalpy and entropy deriving from the entire interacting system. Any attempt of complex classification remains only a phenomenological description, of little practical use, because not knowing the underlying energy game, which is the real parameter to define the interaction, we risk assigning energetically similar interactions to different phenomenological classes. Besides, in the databases, there are over 70,000 pure IDPs and about forty million proteins containing different levels of disorder (the remaining 40,000,000 are globular). How do we create a classification, even by studying a thousand different cases? One thousand cases would represent only 0.000025% of the total disorder. That IDPs are dynamically disordered does not mean that they are biological objects that are not subject to the laws of thermodynamics. These aspects involve also the chemically modified forms by enzymes (PTM). When one or more residues are modified, we are dealing with a chemically different polypeptide with unconventional residues and with different chemical and physical properties in solution. If we do not characterize this polypeptide (chemically and structurally), how do we assign the biological function? It is formerly incorrect to attribute to the native protein (as often it is done) the biological functions that are instead those of different polypeptides. The molecular stiffnesses of each modified form will be different as their chemical-physical behavior in the solution. Different will also be the interaction thermodynamics with the molecular partner which is the only molecular event whose occurrence allows us to say that there is a new function. If we do not have this knowledge how do we classify the functions? When does a functional test tell us that there is a function, to which structure we have to attribute it? If we, for example, consider an IDP with six potentially modifiable sites, even more times, from the same enzyme we can calculate 63 different new chemically modified polypeptide structures. The probability of the existence of each of them will depend on the metabolic conditions of the cell, but the modified polypeptides will show different molecular dynamics and functions because each of them has different structural constrictions. This means that the structure governs the dynamics and the dynamics govern the function, but it also means that the structure governs the function. Since the native protein does not possess these properties, it is a different bio-molecule. In a few words, we assign biological functions without performing a biochemical and biophysical characterization of the polypeptide, although all the technologies exist to proceed in this direction. If we assign all these biological functions to the native form (as happens), we have that, in databases. these functions will be charged, as an annotation, to the native protein. Then, the native protein will have many more functional capacities that it does not possess. In interactomics, this means that we will have HUB nodes artificially brought to a higher degree of connectivity, and, as a result, this will direct us towards incorrect metabolic pathways. In the sixties and seventies, when the function of a protein was studied without characterizing its sequence, no article was accepted for publication. Why do we now accept speculation about a polypeptide function without its characterization? Are we sure that the dogma no longer exists? Aren't IDPs also made from the same twenty amino acid residues? It took twenty good years before we realized that thermophilic bacteria proteins had no special properties and that they followed the same thermodynamics as all the other proteins. I know well that a characterization needs much more skills in the lab; it costs time, effort and money (and fewer papers), but I think it is necessary. All this seems much more a publishing shortcut, instead of gaining new knowledge. Tell me if I'm wrong
Thanks for your attention.
Dear Professor Giovanni Colonna,
Thank you for initiating this much needed discussion on intrinsically disordered proteins. I agree that we should put more emphasis on the biochemical and biophysical characterization of these polypeptide chains. Research in IDPs indeed suffers from well-defined methodology of such characterization. Our lab is working on this for the past few year. We have tried to characterize the physico-chemical properties of disordered regions in two transcription factors. We characterized the structure and dynamics by solution NMR spectroscopy. As you have alluded to molecular stiffness, we did find did find it varies significantly within a disordered region. We referred to it as rigidity. We found 5 to 7 residues formed clusters of rigid segments with a long stretch of disordered residues. In proteins we studied, we found these rigid segments interacted with partner protein very specifically. Based on our findings we proposed a rigid-segment model of disordered regions/proteins. This work was recently published (Maiti et. al. 2019 JMB vol 431 p-1353; Article Dynamic Studies on Intrinsically Disordered Regions of Two P...
).It will be interesting to see how PTMs in disordered regions alter their stiffness resulting in new or altered biological function.
Thank you for the consideration. It is worrying that few ask questions about how we study and classify the function of IDPs. A first problem is a generalized view of proteins (including IDPs), still of the globule-centric type. I think it's a continuum of organizations. We know enough about the globular extreme, organized with fixed structures over time, little of the extreme where disorder reigns supreme and appears chameleon ally with fluctuating and different organizations and, sometimes, also in the form of a globule. It is not a problem of liquid or other physical phases. The problem is that we need to assign the correct structure, rigid or dynamic, to a function. My impression is that dogma always exists; perhaps we must review it in the terminology. Our knowledge is still very macroscopic and static. But in between, we have endless problems with about forty millions of mixed structures, in a continuum of order/disorder relationships, of which we know almost nothing. The attempts made to date are those of using static predictive methods, valid for globular proteins, to predict the disorder in systems that are instead highly dynamic and that use dynamics in a functional sense. Many dynamic studies see mainly the back-bone, but little about what happens to the lateral residues, even less the effect of water. The prevailing forces seem to be of an electrostatic nature (charge distribution on protein segments?), so the environment, i.e., water, has an important role. How can we treat with similar experimental-logic the functional properties of a globular protein with a net charge close to zero, compared to a highly dynamic polypeptide with a net charge, for example, a - 11 or + 9? It makes little sense.
What worries me most is the PTMs. This is a ruse used by evolution to expand the functionality of the product of a single gene. The modifications are of many types (around 300), however, all profoundly alter not only the functional properties of the structure but also its physical and chemical properties, i.e. the structure and its behavior in solution. A question to ask is: does a modified protein have to be classified as structurally different from the native form? I think so and for this, we must find a methodological standard that is the same for everyone that influences the experimental design of any scientific project aimed at studying one of the many functions they show. The problem is, in my opinion, very serious because its specific function must be added to a specific PTM form. Binding to the native protein the functional properties of its innumerable chemically and structurally modified forms, profoundly alters the interactive analysis (PPI) and the knowledge that is derived from the analysis of biomedical Big Data. If we do not find any form of logical standardization, we risk irreversibly polluting the biomedical big data that will gradually give less and less reliable models. I assure you I found these effects also in the genomic data of cancer patients. The diagnosis, to date, is made by comparison with healthy tissue, but what do we think is a healthy or normal tissue to refer to? We need methodological standards.
- Badges
- Science method
- Similar topics
- Methodology
- Research Methodology
- Qualitative Research