Researchers Spot a New Code in Disordered Proteins – Discovery Institute

Scientists from the University of Washington have glimpsed possible reasons for intrinsically disordered domains in heat shock proteins. If confirmed, their observations open up phenomenal possibilities for intelligent design in these and other intrinsically disordered protein domains.

Rather than behaving as a completely 2 intrinsically disordered region, we find it to be quasi-ordered, with six sub-regions that display 3 distinct properties and binding preferences. The results reveal that, contrary to expectation, the high degree of heterogeneity and polydispersity that is a defining feature of HSPB1 (and other human sHSPs) derives not from fuzzy disorder but rather from an array of combinatorial interactions that involve discrete NTR sub-regions and specific surfaces on the structured ACD. We expect other oligomeric sHSPs are similarly defined. [Emphasis added.]

Its just a preprint in bioRxiv, but this paper could represent a giant leap in debunking the old junk-DNA paradigm. Its title, Interplay of disordered and ordered regions of a human small heat shock 1 protein yields an ensemble of quasi-ordered states, introduces quasi-ordered as a term to describe intrinsically disordered proteins, or IDPs. The work presented in this paper should not be discounted for lacking peer review at the time of publication. Six researchers in the University of Washingtons Departments of Biochemistry and Medicinal Chemistry based their models on nuclear magnetic resonance (NMR) imaging and hydrogen-deuterium exchange mass spectrometry (HDX), so the work is not armchair speculation.

Back in May, Evolution News introduced one example of an IDP with a function, and asked: Will other IDPs be found to quickly change from flopping strands into functional regulators based on environmental changes? Will the DNA sequences that produce IDPs continue to confirm the sequence hypothesis?

This paper is reminiscent of the ENCODE work that found 80 percent transcription in non-coding portions of DNA. The sound of the junk-DNA myth collapsing in 2012 (Science) set evolutionists back on their heels. What will they do if the junk-polypeptide myth falls, too? At this time, the new revelations about IDPs are a mere crack in the door to see a brightly lit room, but it looks like the same kind of epochal moment. The implications are staggering: what was dismissed as disordered could turn out to be ordered and functional at higher levels of complexity than previously imagined. IDPs may turn out to be multi-functional tools or skeleton keys able to switch on an array of processes.

Lets introduce the players in this drama.

Small heat shock proteins (sHSPs) are a class of molecular chaperones that help maintain cellular proteostasis. Like other heat shock proteins, sHSPs are believed to interact with exposed hydrophobic regions of partly unfolded or misfolded proteins to help prevent irreversible aggregation, but unlike other heat shock proteins, they perform their functions independent of ATP. sHSPs are implicated in numerous human diseases on the basis of inherited mutations in the protein sequence or upregulation in certain cancers. Cellular stressors such as oxidation and acidosis can influence their function, and stress-induced phosphorylation of sHSPs typically increases their chaperone activity. Despite their important roles in health and disease, relatively little is known about sHSP structure or structure-function relationships compared to other classes of chaperones.

The authors call HSPs natures first responders to cellular stress. Accordingly, these essential machines need a multitude of skills, just like human first responders need flexibility to handle the variety of accidents that can occur in earthquakes, fires, and floods. Until now, the disordered tails of sHSPs were difficult to interpret. In the following, NTR stands for N-Terminal Domain, the disordered portion of a heat shock protein (HSP). ACD is the ordered alpha-crystallin domain of HSP1 (heat shock protein #1), which is flanked by disordered regions:

Overall, the results from the modeling suggest that any combination of the NTR-ACD 1 interactions defined in our study is theoretically possible. While we only created models containing the maximum NTR-ACD interactions supported by our experimental data, any of the interacting motifs we have modeled could dissociate from the ACD and adopt a more disordered conformation. The results from our NMR and HDX experiments indicate that most of these NTR regions occupy both ACD-bound and ACD-unbound conformations, so it is likely that multiple combinations of NTR/ACD interactions occur in solution. Additionally, the similarity of protected regions in the HDXMS profiles of HSPB1 dimers and oligomers indicate that the interactions depicted in these models also occur within higher-order oligomers. The peptide fragments depicted in our dimeric models could conceivably be connected to other ACD dimers or monomers within an oligomer (Fig. 8D). The array of possible interactions within sHSP oligomers is depicted in Fig. 11 8F. Many regions can form intra-chain, intra-dimer, and inter-dimer interactions. The possibility for multiple combinations of interactions and connectivities contributes to the high degree of plasticity and heterogeneity observed for HSPB1 in NMR and HDX experiments.

The researchers observed disordered (i.e., non-folding) regions of this heat-shock protein combining in a variety of ways with the ordered region. Since half of this small HSP is disordered, they believe they saw only a few of the possible combinations. The more combinations, the more plastic (flexible) the proteins functionality becomes, and the more forms it can take on (heterogeneity). As they indicate, this initial glimpse of quasi-ordered states may be a general trend in IDPs. They only looked at simple combinations in a relatively small, two-part HSP. Since larger, more complex ones exist, the interactions depicted in these models also occur within higher-order oligomers.

The following analogy may be strained, but it might help visualize what is going on. Think of a comic-book superhero who carries a magic chain. Depending on the crisis he faces, he can touch a link on the chain to a part of his body to transform himself into the appropriate defender. If he touches the gold link to his knee, he becomes Spiderman. If he touches the copper link to his elbow, he becomes Aquaman. If he touches the iron link to his forehead, he becomes Batman, and so on.

Something like that goes on with heat shock proteins with their disordered domains. Portions of the disordered half of HSP1 fit into certain grooves on the ordered portion, transforming the protein into the tool needed to respond to the current disaster. Even more amazing, combinations of the links in the disordered region act like codes that switch on different states.

Does the word combinatorial bring to mind concepts shared within the ID community? For example, the histone code is a combinatorial code that considerably extends the information potential of the genetic code (see Histone Code: A Challenge to Evolution, an Inference to Design). In a similar way, the disordered regions of some proteins may bind to ordered parts to extend the functional potential of these molecular machines.

The fact that multiple HSPB1 regions can bind to a given groove or surface sets up a situation in which there are more potential binding elements than there are binding sites. This, in turn, creates a large combinatorial array of possible states within a dimer, and even more states within an oligomer. Each HSPB1 dimer has a single dimer interface groove, but its potential interactions with two NTR regions creates a similarly complicated situation: a given dimer interface groove may be empty, bound by a single boundary region, a single conserved region, two boundary regions, one boundary plus one conserved, or two boundary regions plus a conserved region. Again, in the context of an oligomer, the combinatorial possibilities will be increased if the interactions can occur from neighboring dimer units.

Prior to this work, biochemists tended to downplay the functionality of these states. They called them fuzzy, meaning that they cannot be described by a single conformational state. Others disparaged the bundles of random polypeptides as molten globules without much function at all. That picture is vanishing in the junk-DNA mythology lexicon.

However, given the high degree of orientational specificity of many NTR-ACD interactions, these interactions can be described neither as fuzzy in the canonical sense, nor as molten globule-like.

Notably, ordered interactions occur for several NTR sub-regions with the ACD with varying levels of affinity, and some interactions appear to be interdependent. The high degree of heterogeneity in HSPB1 dimers and oligomers is generated not by multiple random or fuzzy states but rather by the large number of possible combinations of several specific and orientationally-defined states.

In addition, these temporary knob-and-hole states, as they describe them, exist with particular lifetimes that expand the possibilities for their usefulness.

Based on observation of multiple slowly exchanging peaks by NMR for certain residues and bimodal HDXMS at long time points, the lifetimes for these interactions range from a minimum of tens of milliseconds to several minutes. For this reason, we propose the term quasi ordered to describe the NTR of HSPB1, as it makes highly-specific long-lived (on the timescale of seconds) contacts while remaining dynamic and heterogeneous.

Perhaps super-ordered would be a better term. Theres nothing quasi about it! When the specific arrangements of IDPs are messed up with mutations, bad things can happen. Even small changes have profound effects:

Remarkably, single mutations in the NTR have profound, widespread effects on dynamics, highlighting sHSP sensitivity to mutation and modification. We find that mutations at residues only five positions apart in the NTR have distinct, almost opposite effects (G34R and P39L) while two mutations that are 50 residues apart from each other (G34R and G84R) produce highly similar effects. In particular, G34R and G84R variants in the conserved and boundary regions respectively each exhibit a coupled increase in deuterium exchange in both the conserved and boundary regions. Furthermore, the mutant G34R conserved region peptide showed a lower affinity for the dimer interface groove. Altogether the results identify an interplay between two non-local regions of the NTR, in which the location of one region affects the other. Both regions can bind at the dimer interface groove, so another way to view the interdependence is that occupancy at a given interface groove by one sub-region favors occupancy by the other.

Its time to look at IDPs with the eyes of design.

Altogether, our results show that even in a monodisperse form of HSPB1, there is substantial conformational heterogeneity, with multiple, specific contacts between regions of the NTR and the ACD. These contacts are altered in activation-mimicking and disease-associated mutated states, shedding light on the mechanisms by which perturbations such as phosphorylation or mutation can influence sHSP structure and function. The experimental approach presented here can be applied to other structurally heterogeneous systems that have proven difficult to study by traditional means, particularly those containing a mixture of ordered and disordered regions.

As Jonathan Wells suggested here back in 2014, IDPs are worth focusing on. They could be significant players in Biologys Quiet Revolution that, while undermining the old Central Dogma of biochemistry, are revealing new grand vistas of design previously unimagined. Combinatorial codes, like those found in histones, olfactory processing centers, alternative gene splicing, and other places in biology, might now be seen coming to light in intrinsically disordered proteins. Rather than viewing them as fuzzy evolving states or molten globules of little interest, biochemists are beginning to glimpse combinatorial arrangements of specific functions that may turn IDPs into the next superheroes of intelligent design.

Photo: Human first responders, by Frmatt [CC BY-SA 3.0], via Wikimedia Commons.