Key aspects of the past 30 years of protein design

Abstract

Proteins are the workhorse of life. They are the building infrastructure of living systems; they are the most efficient molecular machines known, and their enzymatic activity is still unmatched in versatility by any artificial system. Perhaps proteins’ most remarkable feature is their modularity. The large amount of information required to specify each protein’s function is analogically encoded with an alphabet of just ~20 letters. The protein folding problem is how to encode all such information in a sequence of 20 letters. In this review, we go through the last 30 years of research to summarize the state of the art and highlight some applications related to fundamental problems of protein evolution.

Molecular Information Theory Meets Protein Folding

“…36 According to a random energy model for the polypeptide energy landscape, the effective alphabet size N evolved must be higher than the effective number of configurations of a residue in the unfolded state, N unfolded . 74,75 We find that N evolved = 5.0 ± 1.1 amino acids, while N unfolded = 5.4 ± 1.1 conformations (Table 1). This suggests that the effective alphabet size in evolved proteins is just enough to ensure that foldable sequences can be found for the globular structures observed in nature.…”

“…Folding stability can be regarded as a physicochemical constraint on protein evolution. In the absence of such constraints, the energy landscape of random polypeptides can be described using the random energy model. , As a result of the random interactions, the system behaves like a viscous liquid above T g . Below this threshold, the system runs out of entropy, and its kinetics exhibit glass-like behavior.…”

“…The random energy model for the energetics of protein folding has been used to analyze the problem of evolvability, i.e., under which conditions it is feasible to find sequences that fold into a given structure. For the design or natural evolution to have a chance of success, the effective alphabet size N evolved must be higher than the effective number of configurations of a residue in the unfolded state, N unfolded : , N normale normalv normalo normall normalv normale normald > N normalu normaln normalf normalo normall normald normale normald …”

“…The identity between the information required to locate the native state in a conformational search and the information present in natural sequences can be understood as a quantitative formulation of Anfinsen’s “thermodynamic hypothesis” that the information required for folding in a given environment is contained in the sequence of amino acids. ,, Because R Levinthal and N unfolded are fixed by protein chemistry and the genetic code, R sequence and N evolved must evolve toward R Levinthal and N unfolded . According to a random energy model for the polypeptide energy landscape, the effective alphabet size N evolved must be higher than the effective number of configurations of a residue in the unfolded state, N unfolded . , We find that N evolved = 5.0 ± 1.1 amino acids, while N unfolded = 5.4 ± 1.1 conformations (Table ). This suggests that the effective alphabet size in evolved proteins is just enough to ensure that foldable sequences can be found for the globular structures observed in nature.…”

“…115 The relative success of different machine learning approaches to compute a tertiary structure of a natural protein given an amino acid sequence encourages us in the search for simple descriptions for these digital to analog codes. 75 It would be fun if we are able to code an arbitrary problem in a sequence of amino acids and let the protein machine solve it by folding! Molecular information theory allows for the calculation of the energy-to-information conversion efficiency of the molecular machine's operation.…”