Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Evolution of dynamical networks enhances catalysis in a designer enzyme

Nature Chemistry volume 13, pages 1017–1022 (2021)Cite this article

Subjects

Abstract

Activation heat capacity is emerging as a crucial factor in enzyme thermoadaptation, as shown by the non-Arrhenius behaviour of many natural enzymes. However, its physical origin and relationship to the evolution of catalytic activity remain uncertain. Here we show that directed evolution of a computationally designed Kemp eliminase reshapes protein dynamics, which gives rise to an activation heat capacity absent in the original design. These changes buttress transition-state stabilization. Extensive molecular dynamics simulations show that evolution results in the closure of solvent-exposed loops and a better packing of the active site. Remarkably, this gives rise to a correlated dynamical network that involves the transition state and large parts of the protein. This network tightens the transition-state ensemble, which induces a negative activation heat capacity and non-linearity in the activity–temperature dependence. Our results have implications for understanding enzyme evolution and suggest that selectively targeting the conformational dynamics of the transition-state ensemble by design and evolution will expedite the creation of novel enzymes.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Activity gains during directed evolution of 1A32-2 coincided with the emergence of a negative \({\Delta}C_{\mathrm{p}}^{\ddagger}\).
Fig. 2: Global dynamics in the evolved enzyme are modulated by a local tightening of the active site associated with reaction.
Fig. 3: Directed evolution introduced an extended correlated network in the TS ensemble.

Similar content being viewed by others

Data availability

Data that support the findings of this study are available within the paper and its Supplementary Information. Input files, MD trajectories, topologies and parameters relevant to this work are available at the University of Bristol data repository at https://doi.org/10.5523/bris.l6hm9j11yil92bh9rvh27i7ge.

References

  1. Arnold, F. H. Innovation by evolution: bringing new chemistry to life (Nobel Lecture). Angew. Chem. Int. Ed. 58, 14420–14426 (2019).

    Article  CAS  Google Scholar 

  2. Bornscheuer, U. T. et al. Engineering the third wave of biocatalysis. Nature 485, 185–194 (2012).

    Article  CAS  PubMed  Google Scholar 

  3. Tournier, V. et al. An engineered PET depolymerase to break down and recycle plastic bottles. Nature 580, 216–219 (2020).

    Article  CAS  PubMed  Google Scholar 

  4. Kries, H., Blomberg, R. & Hilvert, D. De novo enzymes by computational design. Curr. Opin. Chem. Biol. 17, 221–228 (2013).

    Article  CAS  PubMed  Google Scholar 

  5. Kiss, G., Çelebi-Ölçüm, N., Moretti, R., Baker, D. & Houk, K. N. Computational enzyme design. Angew. Chem. Int. Ed. 52, 5700–5725 (2013).

    Article  CAS  Google Scholar 

  6. Bunzel, H. A., Anderson, J. L. R. & Mulholland, A. J. Designing better enzymes: insights from directed evolution. Curr. Opin. Struct. Biol. 67, 212–218 (2021).

    Article  CAS  PubMed  Google Scholar 

  7. Obexer, R. et al. Emergence of a catalytic tetrad during evolution of a highly active artificial aldolase. Nat. Chem. 9, 50–56 (2017).

    Article  CAS  PubMed  Google Scholar 

  8. Preiswerk, N. et al. Impact of scaffold rigidity on the design and evolution of an artificial Diels-Alderase. Proc. Natl Acad. Sci. USA 111, 8013–8018 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Blomberg, R. et al. Precision is essential for efficient catalysis in an evolved Kemp eliminase. Nature 503, 418–421 (2013).

    Article  CAS  PubMed  Google Scholar 

  10. Bunzel, H. A. et al. Emergence of a negative activation heat capacity during evolution of a computationally designed enzyme. J. Am. Chem. Soc. 141, 11745–11748 (2019).

    Article  CAS  PubMed  Google Scholar 

  11. Khersonsky, O. et al. Bridging the gaps in design methodologies by evolutionary optimization of the stability and proficiency of designed Kemp eliminase KE59. Proc. Natl Acad. Sci. USA 109, 10358–10363 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Fuxreiter, M. & Mones, L. The role of reorganization energy in rational enzyme design. Curr. Opin. Chem. Biol. 21, 34–41 (2014).

    Article  CAS  PubMed  Google Scholar 

  13. Jindal, G., Ramachandran, B., Bora, R. P. & Warshel, A. Exploring the development of ground-state destabilization and transition-state stabilization in two directed evolution paths of Kemp eliminases. ACS Catal. 7, 3301–3305 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Warshel, A. et al. Electrostatic basis for enzyme catalysis. Chem. Rev. 106, 3210–3235 (2006).

    Article  CAS  PubMed  Google Scholar 

  15. Bhowmick, A., Sharma, S. C. & Head-Gordon, T. The importance of the scaffold for de novo enzymes: a case study with Kemp eliminase. J. Am. Chem. Soc. 139, 5793–5800 (2017).

    Article  CAS  PubMed  Google Scholar 

  16. Hong, N. S. et al. The evolution of multiple active site configurations in a designed enzyme. Nat. Commun. 9, 3900 (2018).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  17. Bhowmick, A., Sharma, S. C., Honma, H. & Head-Gordon, T. The role of side chain entropy and mutual information for improving the de novo design of Kemp eliminases KE07 and KE70. Phys. Chem. Chem. Phys. 18, 19386–19396 (2016).

    Article  CAS  PubMed  Google Scholar 

  18. Petrovic, D., Risso, V. A., Kamerlin, S, C. L. & Sanchez-Ruiz, J. M. Conformational dynamics and enzyme evolution. J. R. Soc. Interface 15, 20180330 (2018).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  19. Campbell, E. C. et al. Laboratory evolution of protein conformational dynamics. Curr. Opin. Struct. Biol. 50, 49–57 (2018).

    Article  CAS  PubMed  Google Scholar 

  20. Otten, R. et al. How directed evolution reshapes energy landscapes to boost enzyme catalysis. Science 370, 1442–1446 (2020).

    Article  CAS  PubMed  Google Scholar 

  21. Broom, A. et al. Evolution of an enzyme conformational ensemble guides design of an efficient biocatalyst. Nat. Commun. 11, 4808 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Romero-Rivera, A., Garcia-Borras, M. & Osuna, S. Role of conformational dynamics in the evolution of retro-aldolase activity. ACS Catal. 7, 8524–8532 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Maria-Solano, M. A., Serrano-Hervás, E., Romero-Rivera, A., Iglesias-Fernández, J. & Osuna, S. Role of conformational dynamics in the evolution of novel enzyme function. Chem. Commun. 54, 6622–6634 (2018).

    Article  CAS  Google Scholar 

  24. Casey, M. L., Kemp, D. S., Paul, K. G. & Cox, D. D. Physical organic chemistry of benzisoxazoles. I. Mechanism of the base-catalyzed decomposition of benzisoxazoles. J. Org. Chem. 38, 2294–2301 (1973).

    Article  CAS  Google Scholar 

  25. Kemp, D. S. & Casey, M. L. Physical organic chemistry of benzisoxazoles. II. Linearity of the Broensted free energy relation for the base-catalyzed decomposition of benzisoxazoles. J. Am. Chem. Soc. 95, 6670–6680 (1973).

    Article  CAS  Google Scholar 

  26. Privett, H. K. et al. Iterative approach to computational enzyme design. Proc. Natl Acad. Sci. USA 109, 3790–3795 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Röthlisberger, D. et al. Kemp elimination catalysts by computational enzyme design. Nature 453, 190–195 (2008).

    Article  CAS  PubMed  Google Scholar 

  28. van der Kamp, M. W. et al. Dynamical origins of heat capacity changes in enzyme-catalysed reactions. Nat. Commun. 9, 1177 (2018).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  29. Arcus, V. L. et al. On the temperature dependence of enzyme-catalyzed rates. Biochemistry 55, 1681–1688 (2016).

    Article  CAS  PubMed  Google Scholar 

  30. Prabhu, N. V. & Sharp, K. A. Heat capacity in proteins. Annu. Rev. Phys. Chem. 56, 521–548 (2005).

    Article  CAS  PubMed  Google Scholar 

  31. Stourac, J. et al. Caver Web 1.0: identification of tunnels and channels in proteins and analysis of ligand transport. Nucleic Acids Res. 47, W414–W422 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Liao, Q. et al. Loop motion in triosephosphate isomerase is not a simple open and shut case. J. Am. Chem. Soc. 140, 15889–15903 (2018).

    Article  CAS  PubMed  Google Scholar 

  33. van der Kamp, M. W., Chaudret, R. & Mulholland, A. J. QM/MM modelling of ketosteroid isomerase reactivity indicates that active site closure is integral to catalysis. FEBS J. 280, 3120–3131 (2013).

    Article  PubMed  CAS  Google Scholar 

  34. Malabanan, M. M., Amyes, T. L. & Richard, J. P. A role for flexible loops in enzyme catalysis. Curr. Opin. Struct. Biol. 20, 702–710 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Hall, A. & Knowles, J. R. Uncatalyzed rates of enolization of dihydroxyacetone phosphate and of glyceraldehyde 3-phosphate in neutral aqueous solution. Quantitative assessment of the effectiveness of an enzyme catalyst. Biochemistry 14, 4348–4352 (1975).

    Article  CAS  PubMed  Google Scholar 

  36. Houck, W. J. & Pollack, R. M. Activation enthalpies and entropies for the microscopic rate constants of acetate-catalyzed isomerization of 5-androstene-3,17-dione. J. Am. Chem. Soc. 125, 10206–10212 (2003).

    Article  CAS  PubMed  Google Scholar 

  37. Jones, H. B. L. et al. A complete thermodynamic analysis of enzyme turnover links the free energy landscape to enzyme catalysis. FEBS J. 284, 2829–2842 (2017).

    Article  CAS  PubMed  Google Scholar 

  38. Rivoire, O., Reynolds, K. A. & Ranganathan, R. Evolution-based functional decomposition of proteins. PLoS Comput. Biol. 12, e1004817 (2016).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  39. Reynolds, K. A., McLaughlin, R. N. & Ranganathan, R. Hot spots for allosteric regulation on protein surfaces. Cell 147, 1564–1575 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Halabi, N., Rivoire, O., Leibler, S. & Ranganathan, R. Protein sectors: evolutionary units of three-dimensional structure. Cell 138, 774–786 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Lakhani, B., Thayer, K. M., Black, E. & Beveridge, D. L. Spectral analysis of molecular dynamics simulations on PDZ: MD sectors. J. Biomol. Struct. Dyn. 38, 781–790 (2020).

    Article  CAS  PubMed  Google Scholar 

  42. Sethi, A., Eargle, J., Black, A. A. & Luthey-Schulten, Z. Dynamical networks in tRNA:protein complexes. Proc. Natl Acad. Sci. USA 106, 6620–6625 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Rivalta, I. et al. Allosteric pathways in imidazole glycerol phosphate synthase. Proc. Natl Acad. Sci. USA 109, E1428–E1436 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  44. Goodey, N. M. & Benkovic, S. J. Allosteric regulation and catalysis emerge via a common route. Nat. Chem. Biol. 4, 474–482 (2008).

    Article  CAS  PubMed  Google Scholar 

  45. Melo, M. C. R., Bernardi, R. C., de la Fuente-Nunez, C. & Luthey-Schulten, Z. Generalized correlation-based dynamical network analysis: a new high-performance approach for identifying allosteric communications in molecular dynamics trajectories. J. Chem. Phys. 153, 134104 (2020).

    Article  CAS  PubMed  Google Scholar 

  46. Singh, S. & Bowman, G. R. Quantifying allosteric communication via both concerted structural changes and conformational disorder with CARDS. J. Chem. Theory Comput. 13, 1509–1517 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Åqvist, J., Kazemi, M., Isaksen, G. V. & Brandsdal, B. O. Entropy and enzyme catalysis. Acc. Chem. Res. 50, 199–207 (2017).

    Article  PubMed  CAS  Google Scholar 

  48. Williams, D. H., Stephens, E., O’Brien, D. P. & Zhou, M. Understanding noncovalent interactions: ligand binding energy and catalytic efficiency from ligand-induced reductions in motion within receptors and enzymes. Angew. Chem. Int. Ed. 43, 6596–6616 (2004).

    Article  CAS  Google Scholar 

  49. Osuna, S. The challenge of predicting distal active site mutations in computational enzyme design. WIREs Comput. Mol. Sci. 11, e1502 (2020).

    Google Scholar 

  50. Tiwari, S. P. & Reuter, N. Conservation of intrinsic dynamics in proteins—what have computational models taught us? Curr. Opin. Struct. Biol. 50, 75–81 (2018).

    Article  CAS  PubMed  Google Scholar 

  51. Wells, S. A., van der Kamp, M. W., McGeagh, J. D. & Mulholland, A. J. Structure and function in homodimeric enzymes: simulations of cooperative and independent functional motions. PLoS ONE 10, e0133372 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Zhang, S., Li, H., Krieger, J. M. & Bahar, I. Shared signature dynamics tempered by local fluctuations enables fold adaptability and specificity. Mol. Biol. Evol. 36, 2053–2068 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Fuglebakk, E., Echave, J. & Reuter, N. Measuring and comparing structural fluctuation patterns in large protein datasets. Bioinformatics 28, 2431–2440 (2012).

    Article  CAS  PubMed  Google Scholar 

  54. Echave, J., Spielman, S. J. & Wilke, C. O. Causes of evolutionary rate variation among protein sites. Nat. Rev. Genet. 17, 109–121 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

H.A.B. and A.J.M. thank EPSRC (EP/M013219/1 and EP/M022609/1) and, along with J.L.R.A., the BBSRC (BB/R016445/1) for funding. H.A.B. thanks the Swiss National Science Foundation (Postdoc.Mobility fellowship) for support. M.W.v.d.K. thanks BBSRC for support (David Phillips Fellowship, BB/M026280/1). V.L.A. and A.J.M. thank the Marsden Fund of New Zealand (16-UOW-027). V.L.A. is a James Cook Research Fellow (Royal Society of New Zealand). D.H. thanks the Swiss National Science Foundation. This work was conducted using the computational facilities of the Advanced Computing Research Centre, University of Bristol. We thank R. Crean and S. Osuna for help with and providing a script to perform the shortest-path analysis.

Author information

Authors and Affiliations

  1. School of Biochemistry, University of Bristol, Bristol, UK

    H. Adrian Bunzel, J. L. Ross Anderson & Marc W. van der Kamp

  2. Centre for Computational Chemistry, School of Chemistry, University of Bristol, Bristol, UK

    H. Adrian Bunzel, Marc W. van der Kamp & Adrian J. Mulholland

  3. Laboratory of Organic Chemistry, ETH Zurich, Zurich, Switzerland

    Donald Hilvert

  4. School of Science, University of Waikato, Hamilton, New Zealand

    Vickery L. Arcus

Authors
  1. H. Adrian Bunzel
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. J. L. Ross Anderson
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Donald Hilvert
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Vickery L. Arcus
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Marc W. van der Kamp
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Adrian J. Mulholland
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

H.A.B., M.W.v.d.K. and A.J.M. devised the simulation and analysis. H.A.B. performed the simulations and analysis. H.A.B., J.L.R.A., D.H., V.L.A., M.W.v.d.K. and A.J.M. wrote the manuscript.

Corresponding authors

Correspondence to H. Adrian Bunzel or Adrian J. Mulholland.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Chemistry thanks Paolo Carloni, Vicent Moliner and Sílvia Osuna for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 The emergent \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) in the Kemp eliminase variants predominantly reflects effects on the chemical step.

a, In the activity-temperature dependence of 1A53-2.9, deviations from the linear Eyring behaviour are slightly more pronounced for kcat than kcat/KM. This indicates that the \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) for kcat is of similar magnitude to that of kcat/KM, though no precise \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) has been determined for kcat. Thus, the experimentally determined \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) of −1.17 kJ∙mol−1K−1 for kcat/KM predominantly reflects thermodynamic differences between the TS and GS ensemble. b, Comparison of the activity-temperature profiles and CD melt curves of 1A53-2 (red) 1A53-2.5 (blue) and 1A53-2.9 (black) shows that the apparent \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) is not due to thermal denaturation of the protein. The melting temperatures of all variants are high, and those of the evolved variants are significantly above their catalytic temperature optima. 1A53-2.5, which evolved an activation heat capacity, has an even higher melting temperature than 1A53-2 (82 °C vs. 75 °C). Also, long incubation of all variants at the highest experimental assay temperatures does not affect activity, which shows that there is no significant unfolding at this temperature on the timescale of the experiments. Of note: the Kemp elimination only involves a single deprotonation step. A change in mechanisms is neither likely, nor did we observe any kinetic signatures signalling rate-limiting substrate binding or product release (e.g. burst phases or hysteresis under single turnover conditions). All data presented in Extended Data Fig. 1 were extracted from Bunzel et al. JACS 2019.

Extended Data Fig. 2 MD parameters and restraints.

a, Ground state and b, transition state distances (red) and charges (black) based on gas-phase calculations with 6-nitrobenzisoxazole and acetate. To prevent artificial rigidification of the transition state by introduction of a covalent bond between the ligand and catalytic residue,1 the transition state was simulated with the transferring proton residing either on the ligand (TS) or base (TS2). Both transition state models provided similar results, supporting the significance of our findings. *, Because Glu178 has only two γ-hydrogens, their charge is adjusted to the given value according to the corresponding three hydrogens in acetate. c, Similar to previous work,1 restraints were used to avoid drifting away from reactive conformations during the extended simulations. Weak harmonic restraints for the ligand-base distance δC-C (black, >5.0 Å, 5.0 kcal mol−1 Å−1) and the ligand-base dihedral angle ΨO-C-N-O (red, >±60°, 5.0 kcal mol−1 rad2) were added to keep the ligand in the active site. Furthermore, the χ2 dihedral angle of Trp110 ΦC-C-C-C (blue, 65° > ΦC-C-C-C > 155°, 5.0 kcal mol−1 rad−2) was restrained, preventing its indole sidechain from blocking the active site. To avoid biasing the \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) calculations, the same restraints were added to all variants and ligands.

Extended Data Fig. 3 MD trajectories.

Individual (light colours) and average (dark colours) trajectories of 1A53-2 and 1A53-2.5 with GS (red), TS (blue) and TS2 (grey) bound. a, Based on the r.m.s.d, the first 50 ns of each simulations were discarded to avoid biasing of the analysis to the starting structures (dotted lines). The ligand-base distance δC-C (b), and dihedral angle ΨO-C-N-O (c), remained within the weak harmonic constraints (dotted lines) for most of the simulations. For a definition of the restraints see Extended Data Fig. 2.

Extended Data Fig. 4 Changes in energetic fluctuations give rise to an activation heat capacity \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\).

a, The energy distributions for 1A53-2 (TS: dark red; GS: light red) are almost identical, indicating a \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) close to zero. Energetic fluctuations were calculated after removal of all but 10 water molecules, though removing these only slightly affected \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\). b, The distribution for 1A53-2.5 is narrower for the TS (dark blue) than for the GS (light blue) complex, resulting in a negative \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\). Distributions in (a) and (b) are normalized to a 50 ns moving average and plotted against a log scale to emphasize changes in variance. c, \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) was calculated based on the difference in variance between the TS and GS ensemble for moving windows of various sizes, and converges after 50 ns to negative values for 1A53-2.5 (blue). For 1A53-2, \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) does not significantly deviate from zero (red). Error bars were obtained by leave-one-out cross-validation. As expected, the statistical errors increase with larger window size, because slow conformational changes that are not exhaustively sampled begin to mix into the calculations, complicating the analysis1. Clearly, the energetic fluctuations in 1A53-2 and 1A53-2.5 are different, which gives rise to the dynamical network that tightens the TS ensemble in the evolved variant1. d + e, Heat capacities calculated under various conditions qualitatively agree well with each other. Heat capacities were calculated either dry (e), or with the 10 water molecules closest to Glu178 bound (d) based on the difference in energy variance of the transition and ground state for running windows with varying size (5 ns to 80 ns) for 1A53-2 (red) 1A53-2.5 (blue) for TS (filled circles) and TS2 (open circles).

Extended Data Fig. 5 Structural fluctuations in 1A53-2 and 1A53-2.5.

Residues that rigidify (blue) or become more flexible (red) between a+b, GS and TS, and c+d, GS and TS2. In 1A53-2.5, rigidification is particularly evident in three solvent exposed loops (residues 53-65, 84-92, and 181-192), which is reproduced by clustering based on the whole protein backbone. These loops were thus employed to cluster the structures in an open and closed state (Extended Data Fig. 6).

Extended Data Fig. 6 Cluster analysis reveals an open-closed equilibrium.

Average structures of the a, 1A53-2 and b, 1A53-2.5 ensembles in the open (light blue) and closed (dark blue) state in complex with the TS. Distance-based clustering was performed using the k-means algorithm based on the three solvent exposed loops (bold tubes, residues 53-65, 84-92, and 181-192). Loop closure is indicated by the Cα distance of residues 58 and 188 (spheres). c + d, Histograms and e + f, distributions of the open (light colours) and closed (dark colours) states bound to GS (red), TS (blue) and TS2 (grey). g + h, Loop closure decreases the radius of gyration which is smallest in the closed TS of 1A53-2.5 (blue line). i + j, The open and closed state interconvert on the timescale of the simulation. Error bars represent the standard deviation of 10 independent simulations.

Extended Data Fig. 7 Principal component analysis reproduces the two state model.

Principal component analysis of the mobile loops (residues 53-65, 84-92, and 181-192) was performed for the a, GS (red), b, TS (blue), and c, TS2 (grey) ensembles of 1A53-2 (left) and 1A53-2.5 (right). The histogram of the first principal component after partition of the trajectory according to the cluster analysis (Extended Data Fig. 6) into open (light) and closed (dark) states reproduces the two-state model.

Extended Data Fig. 8 Local structural effects at the active site.

a+b, The distance-distribution of the ten water molecules closest to the Oε of Glu178 is shown for the GS (red), TS (blue), and TS2 (grey) in the closed (dark colours) and open (light colours) state. The active site becomes less hydrated in the transition states compared to the ground state for both enzymes. Furthermore, the A157Y mutation in 1A53-2.5 displaces a water molecule in the evolved variant at approximately 3 Å. Remarkably, loop closure expels water from the active site in 1A53-2.5 (arrows) which was not observed for 1A53-2. The c+d, SASA of the ligand and e+f, RMSF of the ligand and base are furthermore lower in 1A53-2.5 than 1A53-2, indicating tighter packing of the ligand. Error bars represent the standard deviation of 10 independent simulations.

Extended Data Fig. 9 Emerging dynamical networks in 1A53-2 and 1A53-2.5.

Only dynamical changes between the same states are comparable. The comparison of 1A53-2 with 1A53-2.5 is either performed between GS to TS (left) or between GS to TS2 (right). a, Backbone movements become highly correlated in the closed TS ensemble of 1A53-2.5, indicating the presence of global vibrations comprising most of the protein. b, Cross correlations that increase between GS and TS by ≥20% are indicated as lines on the structures. c, Based on the cross-correlations, shortest path maps were calculated2. These show that evolution introduced a communication network in 1A53-2.5 that originates from the TS and comprises large parts of the closed state. The size of the edge (black sticks) and vertices (blue spheres: protein, orange sphere: ligand) indicate the increasing weight of the network between GS and TS.

Extended Data Fig. 10 QM/MM modelling of the Kemp eliminases.

Free energy surfaces of 1A53-2 (top) and 1A53-2.5 (bottom) at the a, PM6/CHARMM36, b, AM1/CHARMM36 and the c, PM6/AMBERff99SB level of theory. The ligand as well as the carboxylate of Glu178 were part of the QM region during umbrella sampling of the C-H and N-O bonds. The transition state (black dot) reproduces the asynchronous Kemp elimination with C-H cleavage preceding N-O cleavage observed for similar systems.3-5 PM6/CHARMM36 gave activation energies of 101 kJ∙mol-1 and 90 kJ∙mol-1, which reproduce the improvements achieved during evolution (86 and 68 kJ∙mol-1 for kcat). Similarly, simulations with AM1/CHARMM36 preserved that trend, but gave higher energies (117 and 105 kJ∙mol-1). In contrast, simulations with PM6/AMBERff99SB resulted in an inverse trend for the activation energies (111 and 116 kJ∙mol-1). This observation for AMBERff99SB mimics the classical MD, in which the ligand was expelled from the active site within 100 ns.

Supplementary information

Supplementary Information

Supplementary Methods.

Supplementary Table 1

Variances and \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) of 1A53-2.

Supplementary Table 2

Variances and \({{{\Delta}C}}_{\mathrm{p}}^{\ddagger}\) for 1A53-2.5.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bunzel, H.A., Anderson, J.L.R., Hilvert, D. et al. Evolution of dynamical networks enhances catalysis in a designer enzyme. Nat. Chem. 13, 1017–1022 (2021). https://doi.org/10.1038/s41557-021-00763-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41557-021-00763-6

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing