Ionic Atmosphere Effect on the Absorption Spectrum of a Flavoprotein: A Reminder to Consider Solution Ions

Abstract

Ionizable residues and monoatomic ions in solution modulate enzyme catalysis and the structural stability of proteins; however, the delicate interplay between these short-range charges and longrange charges, and their contributions to the electrostatic environment in a protein active site, is currently not fully understood. The study presented here utilizes the FMN-dependent NADH:quinone oxidoreductase from Pseudomonas aeruginosa PAO1 (NQO, EC 1.6.5.9, UniProtKB Q9I4V0) as a model system to study the effect of introducing an active site negative charge on the flavin absorption spectrum both in the absence and presence of a long-range electrostatic potential coming from solution ions. Using pH-dependent UV-visible spectroscopy, there were no observed changes in the flavin absorption spectrum when an active site tyrosine (Y277) deprotonated in vitro. These results could only be reproduced computationally using Average Solvent Electrostatic Configuration (ASEC) hybrid quantum mechanics / molecular mechanics (QM/MM) simulations that included both positive and negative solution ions. The same calculations performed with minimal ions to neutralize protein charges predicted that deprotonating Y277 would significantly affect the flavin absorption spectrum. Analyzing the distribution of solution ions from ASEC and radial distribution functions derived from molecular dynamics indicated that the solution ions reorganize around the protein surface upon Y277 deprotonation to cancel the effect of the tyrosinate on the flavin absorption spectrum. Biochemical experiments were performed to support this hypothesis. This work highlights the importance of salt ions, which are sometimes overlooked, since they can contribute a non-uniform and anisotropic long-range potential to the electrostatic environment of an active site.

An Alternative Strategy for Spectral Tuning of Flavin-binding Fluorescent Proteins

“…In a recent joint computational and experimental study, (56) we found that solution ions may affect the outcome of ASEC-FEG calculations, especially when there are charged residues inside the active site of a flavoprotein. Therefore, we repeated the calculations for both iLOV and iLOV-Q430E after adding 4 pairs of Na+ and Cl-solution ions, approximately 1 NaCl per 2775 water solvent molecules, equivalent to 20 mM of salt used in the experiments in this work.…”

“…(61) We recently developed ASEC-FEG for flavoproteins. (56,62) With ASEC, the quantum chemical calculations are performed in the field of a time-averaged electrostatic potential environment of the protein and solution (collectively referred to as a "solvent" in the ASEC acronym). Effectively, the protein and solution are represented as a "superposition" of structures obtained from MD simulations.…”

“…Flexible atoms also have a broader and shallower Lennard Jones potential (see ref. (56) for more details). The optimization of the QM system within the ASEC configuration is done selfconsistently.…”

“…ASEC is well suited for averaging the effect of long-range charge interactions that are difficult to capture with more traditional QM/MM methods. (56) The ASEC-FEG protocol is a series of scripts building on an existing QM/MM interface between the OpenMolcas(63) and Tinker(64) software packages. (65,66) The protocol guides users through the model construction starting from the PDB file of the protein and culminating in the generation of an ASEC QM/MM model.…”

OS100: A Benchmark Set of 100 Digitized UV–Visible Spectra and Derived Experimental Oscillator Strengths

“…An important next step would be to test the accuracy of continuum or explicit solvent models in predicting solvent effects on oscillator strengths . Both implicit and explicit models appear to successfully predict trends in oscillator strengths for the same molecule in different environments, , but it is not yet clear if such predictions are quantitative. Preparing a benchmark set of experimental oscillator strengths for gas-phase UV–visible spectra would also be informative, but is more challenging because gas-phase spectra tend to have well-resolved vibronic transitions that are more difficult to integrate.…”