Poster Presentation 21st International Conference on Biological Inorganic Chemistry 2025

Machine-learning-guided design of a dicopper phenol oxidase. (#554)

Vanessa H Eng 1 , Yiying Li 1 , Suppachai Srisantitham 1 , Angela A Shiau 2 , Mauro Gascon 1 , R. David Britt 2 , F. Akif Tezcan 1
  1. University of California, San Diego, La Jolla, CALIFORNIA, United States
  2. University of California, Davis, Davis, California, United States

Type 3 (T3) Cu proteins, such as hemocyanin, tyrosinase, and catechol oxidase, are binuclear Cu proteins that bind and activate molecular oxygen. Although these proteins share the same primary-sphere coordination motifs, they have distinct tertiary folds and exhibit notable differences in their secondary sphere environments, which account for their differential functions: hemocyanin binds dioxygen to facilitate transport and storage, whereas tyrosinase catalyzes the oxidation of monophenols and diphenols, while catechol oxidase only oxidizes diphenols.1-2 Yet, upon minimal proteolysis or treatment with detergents, hemocyanin exhibits phenol oxidation activity.3-4 To elucidate the basis of this functional differentiation among T3 Cu proteins and to investigate whether their dioxygen reactivities could be realized in minimal structural motifs, we set out to construct de novo di-Cu enzymes using machine learning (ML)-based tools recently developed in the protein design field. In this workflow, we used the tools RFdiffusion5 and ProteinMPNN6 to generate new backbones and sequences that were templated around active-site coordination motifs extracted from natural binuclear Cu proteins, followed by in-silico validation with AlphaFold3.7 This strategy enabled us to investigate the reactivity of T3 Cu active sites in entirely new structural contexts. In particular, one of our designed constructs, termed HC4, displayed high thermostability, picomolar dissociation constants for Cu(II), and oxidation of diphenolic substrates at both room and elevated temperatures (≤60 °C). Cu-HC4 was also found to oxidize L-DOPA to form melanin polymers. Current directed evolution efforts utilizing site-saturation mutagenesis are focused on evolving this enzyme’s active site to oxidize bulkier phenolic substrates. These studies provide insights into the minimal structural elements that give rise to phenol oxidation activity in T3 Cu proteins and demonstrate the efficiency of the ML-guided approaches to design artificial metalloenzymes with novel topologies.

  1. Solomon, E. I.; Heppner, D. E.; Johnston, E. M.; Ginsbach, J. W.; Cirera, J.; Qayyum, M., . . . Tian, L., Chem. Rev. 2014, 114 (7), 3659-3853.
  2. Pretzler, M.; Rompel, A. Inorg. Chim. Acta 2018, 481, 25-31.
  3. Nagai, T.; Osaki, T.; Kawabata, S.-i. J. Biol. Chem. 2001, 276 (29), 27166-27170.
  4. Decker, H.; Rimke, T. J. Biol. Chem. 1998, 273 (40), 25889-25892.
  5. Watson, J. L.; Juergens, D.; Bennett, N. R.; Trippe, B. L.; Yim, J.; Eisenach, H. E., . . . Baker, D. Nature 2023, 620 (7976), 1089-1100.
  6. Dauparas, J.; Anishchenko, I.; Bennett, N.; Bai, H.; Ragotte, R. J.; Milles, L. F., . . . Baker, D. Science 2022, 378 (6615), 49-56.
  7. Abramson, J.; Adler, J.; Dunger, J.; Evans, R.; Green, T.; Pritzel, A., . . . Jumper, J. M. Nature 2024, 630 (8016), 493-500.