What Makes a Protein Fold Amenable to Functional Innovation? Fold Polarity and Stability Trade-offs

Highlights

► Fold polarity corresponds to low connectivity of the scaffold and the active site. ► Polarity enables to stabilize the scaffold while maintaining active-site plasticity. ► Low scaffold–active-site connectivity alleviates stability–activity trade-offs. ► Polarity promotes the acceptance of active-site mutations. ► Innovable folds carry many different functions and exhibit high polarity.

Abstract

Protein evolvability includes two elements—robustness (or neutrality, mutations having no effect) and innovability (mutations readily inducing new functions). How are these two conflicting demands bridged? Does the ability to bridge them relate to the observation that certain folds, such as TIM barrels, accommodate numerous functions, whereas other folds support only one? Here, we hypothesize that the key to innovability is polarity—an active site composed of flexible, loosely packed loops alongside a well-separated, highly ordered scaffold. We show that highly stabilized variants of TEM-1 β-lactamase exhibit selective rigidification of the enzyme's scaffold while the active-site loops maintained their conformational plasticity. Polarity therefore results in stabilizing, compensatory mutations not trading off, but instead promoting the acquisition of new activities. Indeed, computational analysis indicates that in folds that accommodate only one function throughout evolution, for example, dihydrofolate reductase, ≥ 60% of the active-site residues belong to the scaffold. In contrast, folds associated with multiple functions such as the TIM barrel show high scaffold–active-site polarity (~ 20% of the active site comprises scaffold residues) and > 2-fold higher rates of sequence divergence at active-site positions. Our work suggests structural measures of fold polarity that appear to be correlated with innovability, thereby providing new insights regarding protein evolution, design, and engineering.

Introduction

In silico and experimental analyses indicate that protein stability confers tolerance to mutations and thereby promotes protein evolvability.[1], [2], [3], [4], [5], [6] Evolvability, however, has two components that are interlinked but also potentially contradictory.[7], [8] The accumulation of mutations while maintaining the original function (neutrality, or genetic robustness) is one component of protein evolvability. The acquisition of new functions, termed here innovability, is another component (the term innovability was adopted from Ref. 8; in here, innovability relates to the divergence of novel protein functions, rather than the enhancement of latent, promiscuous activities). That stability promotes robustness and therefore neutral evolution, is well established and biophysically understood. A highly ordered, well-packed protein affords a higher stability margin, or threshold, and enables more destabilizing mutations to accumulate.[1], [2], [3], [4], [6], [9] What about innovability, namely, the ability to evolve new functions? On the one hand, mutations that promote new functions tend to be destabilizing,[10], [11], [12] and thus, excess stability would promote innovability.⁵ On the other hand, increased stability coincides with reduced conformational plasticity.[13], [14] The acquisition of new functions, however, often depends on conformational plasticity—the coexistence of multiple structural conformers.[15], [16] This is certainly the case with the adaptation of β-lactamases towards new antibiotics,[17], [18] including TEM-1, the model enzyme studied here.[19], [20], [21] Thus, higher stability could also hamper innovability by reducing conformational plasticity.

It is often the case that activity and stability trade off.[10], [22] Many enzymes were, however, dramatically stabilized without compromising their activity.[23], [24] Do stability and innovability trade-off? Namely, could excess stability hamper the effect of function-modifying mutations? Does the existence or absence, of a stability–innovability trade-off relate to the protein's architecture?

Additionally, stability promotes evolvability only if stability is an additive, global parameter, whereby stabilizing mutations in one region (e.g., a protein's scaffold) readily compensate for the destabilizing effects of mutations in other locations (e.g., in the active-site region). While this is the prevailing model,[4], [5] can it be taken for granted? In some proteins, higher stability is mediated by mutations in residues that mediate function, suggesting that stability and function do trade off.²² In other cases, the compensatory mutations are in direct contact with the function-modulating mutations, that is, local, specific suppressors²⁵ and not global ones.

We thus sought to explore certain aspects of stability–evolvability. Firstly, whether, and why, the increased rigidity conferred by stabilizing mutations does not affect the active site's conformational plasticity, and/or the ability to acquire new functions. Secondly, do the lack of stability–innovability trade-offs and the enhancement of evolvability as well as innovability by stabilizing mutations relate to the protein's architecture? Protein folds seem to dramatically differ in their innovability—TIM barrels, for example, underline > 120 different enzyme families with different functions and with no apparent sequence identity. Hence, the TIM barrel fold exhibits high innovability and robustness. However, other folds, for example, dihydrofolate reductase (DHFR), are associated with only one enzymatic function. Is this a coincidence, or are certain protein architectures less prone to trade-offs and more amenable to functional innovation?