WP1, Advancement and Results:
We published two papers on QM/MM methods applied to flavin electronic structure and spectroscopic signatures. The first was an effort from the Berlinbased team. It reports a comprehensive assessment of computational tools available for the study of flavins, and identifies the strengths and weaknesses of each. This was a way of familiarizing ourselves with all our options in order to make the best choices to advance on the specific questions posed in our research. The second one detailed variations that result upon covalent substitutions at the 8 position of the flavin ring. Both are listed below.
In the summer of 2022, the master’s student submitted and defended her thesis on computation of hydrogen bonding interactions in the two flavin sites of the ETF of A. fermentans. Although she was formally enrolled in doctoral courses in the fall of 2022, the student worked with the fellow and Prof. Mroginski to prepare the work for publication. Besides documenting redox-coupled reorganization of the H-bonding network supporting the ET flavin, this work identified a particular H-bond with the unusually stable ASQ state of this site, finding that this is specific to the NHδ tautomer of the conserved His that was identified. In the spring of 2023, we submitted this work for publication (Figure 1). It was revised in April and appeared on line in June (outside the current reporting period).
Meanwhile, the second postdoc gained competence with our system and by the summer of 2022 was actively collaborating with a KY student to use Molecular Dynamics calculations performed at TU-Berlin to interpret the results of Small Angle Neutron Scattering (SANS) conducted in the U.S.A at Oak Ridge National Lab.
Challenges:
The major challenge to this project has been our difficulty in retaining postdoctoral scholars in this position. This cost us a lot of time that could have been more productive.
In the course of writing up the Masters student's manuscript, the student undertook energy minimizations of alternate conformations of the Bfflavin site wherein the aminoacid side chain forming a suspicious H-bond was reoriented to form more plausible interactions. whereas the original set of energy minimizations had all begun with the crystallographic conformation of the bifurcating site, and remained fixed there with only minute changes, the new optimizations revealed a new conformation within 1 Cal/mol in energy of the original one for the fully reduced state of this site.
The other oxidation states of this site retained energetic preference for the crystal-structure derived conformation, so the newly discovered alternative is specific to the flavin HQ state and represents a possible redox-coupled conformational change.
This result arose in the course of QM/MM calculations focussing on the immediate environs of the flavin. However we are now seeking independent assessment of the new conformation's possible significance. Extensive molecular dynamics calculations now underway would be expected to discover the same conformation, at least occasionally, if it is a likely state of the protein. We are tracking distances that distinguish this conformation from the crystallographic one, in our MD trajectories.
Relation to Original Plans:
We had proposed to calculate electronic structures for the flavins in optimized sites of A. fermentans ETF. Unfortunately the postdoc charged with doing this left without giving notice. We are now focussed on submitting publications on work that was successfully completed.
WP2, Advancement and Results:
Dynamic nuclear polarization (DNP) was deployed on a sample containing 15N-labeled flavin bound in 14N protein. DNP yields a 22-fold enhancement in signal amplitude when the protein is in a microcrystalline state, incubated with the polarizing agent. Two signals were observed in a spectral region compatible with N3 or N10. 57 mg of material were spun into 1.9 mm rotors and observed at 80 MHz for 15N. Thus, important methodological progress has occurred, although the sought-after result proved elusive. The protein we were working with initially was not well enough behaved for extensive NMR, so the fellow undertook extensive database research and identified some novel sequences representing fusions of the two canonical ETF subunits. The cloned and expressed gene for the monomeric ETF from Sulfolobus acidocaldarius produced much more protein that is moreover much better behaved than the protein described in the initial proposal. Thus we now work on the protein from S. acidocaldarius. The ETF samples made seem to retain integrity for periods of a year, and have yielded very good 1H15N correlation spectra. These are the basis for a paper that is now published, but outside the interval described by this report (Spring 2024).
Challenges:
The 15N signals of 15N-flavin in 14N protein are more difficult to obtain than they should be. One possibility is that the enormous chemical shift anisotropy of the N5 (and N1) signals makes DNP and cross polarization (CP) too inefficient at the high fields being used. We had the plan of optimizing spectrometer conditions using a commercially-available model compound such as 15N-guanine. However we cannot hold the doctoral student hostage to difficult problems, so this idea will await other hands. In order to complete a story for his thesis, the doctoral student adopted a pragmatic approach for monitoring conformational change, using 19F NMR, as suggested in the original proposal. He is advancing two complementary approaches for incorporating 19F Tyr in the ETF and will use 19F NMR chemical shift changes as indicators of conformational changes. 19F is particularly well suited to the problem because of its enormous chemical shift dispersion. Tyr is present at several strategic locations in the ETF, and the student has introduced one more at a particularly interesting spot in the interface between domains. Finally, the Masters student also screened crystallizations with the protein produced by the doctoral student for WP2, in a new collaboration with the Dobbek group (Humboldt University).
Crucially, the Master's student had considerable experience with protein crystallography and the Dobbek group has systems in place for anaerobic manipulation and characterization of proteins. Relation to Original Plans: Although use of 19F NMR was proposed, the way in which we propose to use it has evolved as we better appreciate the challenges and opportunities presented by ETFs. Obtaining NMR signatures of the domain-scale conformation change commands much more urgency now than it did when the proposal was submitted. If we can demonstrate that certain Tyr 19F chemical shifts change in response to events proposed to trigger conformational change, this will be the basis for a paper in a high impact journal.
Similarly, any successes in the crystallography will be very important, given that this new ETF is the first to our knowledge to be a monomer in which the two individual subunits are fused. Thus we have pursued it despite its absence from the original plans.