Table I.
Dynamic properties of computationally designed proteins
| Designed protein | Design strategy | Dynamic properties | Reference(s) |
|---|---|---|---|
| Alpha-carbonic anhydrase | Insertion of strategic point mutations inspired by MD of a thermophilic homologue | Decreased RMSD/F, decreased SASA | Bharatiy et al., 2016 |
| T4 lysozyme | Proteus-designed point mutant pairs | Increased interresidue contacts; Some stabilizing (ProTherm ΔΔG) | Barroso et al., 2020 |
| Conserpin | Consensus design | More conformationally homogeneous (PCA), thermostable (Tm from CD), decreased salt bridges, decreased H-bonds, decreased SASA, decreased RMSF | Porebski et al., 2016 |
| Consensus-HD | Consensus design | Decreased backbone motion (15N NMR), more stable (ΔG from CD) | Tripp et al., 2017 |
| UVF | De novo design | Increased RMSD/F, increased unique side-chain contacts | Nguyen et al., 2019 |
| 15 scFvs and scAbs | RosettaAntibody | Some more thermostable (Tm from DSFa); Some resistant to heat deactivation at 70 °C (ELISA) | Lee et al., 2020 |
| AYEdes | Rosetta de novo | Decreased RMSD/F, decreased SASA, increased secondary structure retention, increased contacts, more stable (ΔG from CD) | Dantas et al., 2007; Gill and McCully, 2019 |
| Flu and botulism antigen-binding mini proteins | Rosetta de novo design, including backbone | Generally thermostable (Tm from CD); Antigen-binding residues were less dynamic in successful designs (backbone/side-chain RMSD) | Chevalier et al., 2017 |
| 7896 pocket proteins | Rosetta de novo design, including backbone | Stability score was correlated with total sequence hydrophobicity, Rosetta energy score, local sequence-structure agreement; Those that expressed tended to be thermostable (Tm from CD) | Basanta et al., 2020 |
| ASR, consensus EF-Tu | ASR, consensus design | ASRs were more rigid (RMSD/F); Consensus had dynamic properties unlike naturally occurring proteins (PCA) | Okafor et al., 2018 |
| AncSR1, AncSR2 steroid receptors | ASR | Older ASR had several highly dynamic regions (RMSD/F); ASRs maintained extant contact networks to mediate an allosteric conformational change | Okafor et al., 2020 |
| Ancestral glycosidase | ASR | ASR was more flexible near the active site but core was equally rigid as extant (RMSF, b-factor, proteolysis) | Gamiz-Arco et al., 2021 |
| Precambrian β-lactamases | ASR | Older ASRs were more flexible globally and in/around the catalytic pocket (RMSF, DFIb) | Zou et al., 2015; Risso et al., 2017 |
| AncHLD-RLuc | ASR | ASR was less dynamic than extant proteins, and a highly mobile helix/loop led to increased active site accessibility (RMSD, Caver) | Chaloupkova et al., 2019 |
| Nitrating P450 TxtE mutants | MD/HMMc-informed site-saturating mutagenesis | Mutants’ F/G loop stayed in the closed conformation more often (HMM,c TTN,d KD) | Dodani et al., 2016 |
| 4 LinB mutants | Caver + site-saturating mutagenesis | Mutant’s active site tunnel was open more often (MD, Caver, SAe) | Brezovsky et al., 2016 |
| 2 successful, 2 unsuccessful DIG-binding proteins (DIG10.2, DIG10.3, DIG12, DIG16) | Rosetta de novo | Successful designs had more rigid cavity entrances (RMSF), better-organized hydrophobic cores (SASA), smaller cavity volumes (POVME, RMSF), preorganized ligand-binding side chains (dihedral angles), stationary ligand in holo simulations (RMSD) | Tinberg et al., 2013; Barros et al., 2019 |
| DFSc | Rational coiled-coil design | Preorganization of SQ•-compatible Zn2+ coordination state improves binding | Reig et al., 2012; Ulas et al., 2016 |
| PS1 | Rational coiled-coil design, Rosetta | Hydrophobic core is structured and ligand-binding region is flexible (HDX, water locations from MD) | Polizzi et al., 2017 |
| ABLE | Rational coiled-coil design, COMBS, van der Mers, Rosetta | Preorganization of rotamers in the ligand-binding site except for two residues (crystal structure) | Polizzi and DeGrado, 2020 |
a Differential scanning fluorimetry.
b Dynamic Flexibility Index.
c Hidden Markov Models.
d Total turnover numbers.
e Specific activity.