What Have We Learned from Design of Function in Large Proteins?

Abstract

The overarching goal of computational protein design is to gain complete control over protein structure and function. The majority of sophisticated binders and enzymes, however, are large and exhibit diverse and complex folds that defy atomistic design calculations. Encouragingly, recent strategies that combine evolutionary constraints from natural homologs with atomistic calculations have significantly improved design accuracy. In these approaches, evolutionary constraints mitigate the risk from misfolding and aggregation, focusing atomistic design calculations on a small but highly enriched sequence subspace. Such methods have dramatically optimized diverse proteins, including vaccine immunogens, enzymes for sustainable chemistry, and proteins with therapeutic potential. The new generation of deep learning-based ab initio structure predictors can be combined with these methods to extend the scope of protein design, in principle, to any natural protein of known sequence. We envision that protein engineering will come to rely on completely computational methods to efficiently discover and optimize biomolecular activities.

Designed active-site library reveals thousands of functional GFP variants

“…In fact, this single parameter exhibits an area under the ROC curve of 87%, compared to 93% 8 for the random forest. This result provides a compelling verification for the approach of combining sequence conservation with atomistic protein design, which underlies htFuncLib and other successful protein design methods developed in recent years 26 .…”

“…To conclude, unlike conventional protein design methods 25,26 , htFuncLib does not search for the most optimal mutants according to energy or structure criteria. Instead, the astronomically large space of combinatorial mutations in an active site is reduced to a tractable size through phylogenetic, structural, and energy-based analysis.…”

“…Here, we introduce a computational method called high-throughput functional libraries (htFuncLib) to design arbitrarily large libraries of active-site mutants that can be applied, in principle, to any protein. Most current atomistic design methods, including our previously described FuncLib method 24 , select designs that maximize desired energy or structure criteria 25,26 . By contrast, htFuncLib searches for a set of active-site point mutations that, when combined all-against-all, yield low-energy multipoint designs.…”

Designed high-redox potential laccases exhibit high functional diversity

“…We recently used PROSS to generate several stable and functionally diverse VPs 25 . Here, we used a different approach for diversification, demonstrating that evolution-guided atomistic design methods 48 can be used to construct a repertoire of stable and diverse HRPLs using a two-step design protocol, again with a very low experimental effort. This stabilize-and-diversify strategy dramatically increases the success rate in creating functional diversity relative to even the most advanced laboratory engineering and evolution strategies.…”

Combinatorial assembly and design of enzymes

“…By contrast, this feature is disfavored within the active-site pocket, presumably because active-site residues are selected for their impact on activity rather than stability. We also find that hydrogen-bond energies are highly discriminating, reflecting the high propensity of buried long-range hydrogen-bond networks in large proteins of a complex fold such as TIM barrels( 39 , 40 ) (fig. S5).…”