Protonation of the Biliverdin IXα Chromophore in the Red and Far-Red Photoactive States of a Bacteriophytochrome

The tetrapyrrole chromophore biliverdin IXα (BV) in the bacteriophytochrome from Deinococcus radiodurans (DrBphP) is usually assumed to be fully protonated, but this assumption has not been systematically validated by experiments or extensive computations. Here, we use force field molecular dynamics simulations and quantum mechanics/molecular mechanics calculations with density functional theory and XMCQDPT2 methods to investigate the effect of the five most probable protonation forms of BV on structural stability, binding pocket interactions, and absorption spectra in the two photochromic states of DrBphP. While agreement with X-ray structural data and measured UV/vis spectra suggest that in both states the protonated form of the chromophore dominates, we also find that a minor population with a deprotonated D-ring could contribute to the red-shifted tail in the absorption spectra.


List of Figures
S8. Q-band and Soret-band maxima using XMCQDPT2 method S9. UV/Vis absorption spectra comparison for different sizes of the QM subsystem. S10. CASSCF (12,12) active space orbitals generated for the XMCQDPT2 excited state energy calculations. S11. Excited-state relaxed energy scans of the most important biliverdin torsions. S12. The contribution of different excited state transitions (S 0 S n ) to the absorption spectra computes with xMCQDPT2 method for the Pr and Pfr states with BV deprotonated at Dring. S13. Snapshot from MD simulation of Pr state showing the speculated proton-entry and -release wires. S14. Comparison of CAM-B3LYP and B3LYP optimized BV geometry.
S2. BV chromophore parameters generated for DOWSER program.
S3. Histidine residue tautomeric forms defined using pKa estimation for DrBphP MM simulations.
S4. Atomic charges for the BV chromophore in Pr state derived using RESP procedure for AMBER03 force fields S5. Atomic charges for BV chromophore in Pfr state derived using RESP procedure for AMBER03 force fields S6. Atomic charges used for Pr state Cys24 residue in AMBER03 force fields for MD simulations.
S7. Atomic charges used for Pfr state Cys24 residue in AMBER03 force fields for MD simulations.
S8. Comparison of DFT functionals based on the bond lengths for optimized BV chromophore for Pr and Pfr state full protonated chromophore. S6 A total of 81 and 84 water molecules were added in the Pr and Pfr phytochrome cavities. The buried water molecules placed by DOWSER were refined by energy optimization and short equilibration simulations. An environment comparable to the high resolution (1.45 Å) crystal structure of DrBphP CBD domain (PDB: 2O9C) 8 was obtained ( Figure S1). After the energy minimization and short equilibration runs the water molecules added with DOWSER form a network of non-covalent interactions with the chromophore and adjacent residues analogous to those in the experimentally resolved CBD structure.   Figure S1. A dotted sphere representation of water molecules added in the vicinity of biliverdin using DOWSER program and equilibrated using MD simulations, labelled as DOWSER water. The arrangement of DOWSER water molecules is identical to the crystal water arrangement around chromophore in high resolution (1.45 Å) structure of CBD domain from DrBphP in dark form.
The tautomeric form of Histidine (His) residues and protonation state of amino-acids with ionizable side chains are summarized in Table S3. The protonation form for other residues of photo-sensory core were assigned based on their standard pK a values in solution. Table S3. Histidine residue tautomeric forms defined using pKa estimation for DrBphP MM simulations.

Charges (e) -Pr Dark state Atom Pr
Deprot  Table S7. Atomic charges used for Pfr state Cys24 residue in the AMBER03 force field for the MD simulations.

Counts Counts
Counts Counts Figure S6. Distance distribution plots of the polar contacts between the BV A, B, and C pyrrole ring N-atoms, His260, and Asp207 in the chromophore binding pocket (shown earlier in Figure 3) for the 5 Pfr state models with different protonation form of BV. S20

Counts Counts
Counts Counts Counts Figure S7: Distance distribution plots for salt-bridge Asp207-Arg466, Tyr263-Ser468, chromophore B/C-ring side-chain carboxylate group and Tyr 216, Arg254, and Ser274 which are the conserved non-covalent interaction in active site of Pfr state. Figure S8: Inclusion of more states (nstates=10) in the xMCQDPT2 calculations reproduces the Soret-band for the Pr and Pfr fully protonated BV models.

S22
Effect of QM subsystem size and protein environment: Figure S9 compares the UV/Vis spectra for the different QM subsystem sizes for the Pr state with a fully protonated BV. The inclusion of pyrrole water or protein environment (using point charges) in the QM subsystem does not show significant change in the Q band maxima, with the corresponding peak observed at ~690 nm. This suggests that the point charge representation of the pyrrole water is sufficiently accurate. Excluding the protein environment and the propionate side chains of the chromophore do not have any effect on the Q-band maxima, which indicates that future computations could be carried out with the QMsmall subsystem to save computational effort without any qualitative loss. Figure S9. UV/Vis absorption spectra for the Pr state with fully protonated chromophore calculated using BLYP/6-31G(d,p) level of theory for the QM-full sub-system (see Figure S3) with the pyrrole water (blue) and without the pyrrole water (black). The experimental spectrum is shown as brown dotted line. S23 Figure S10: CASSCF (12,12) active space orbitals used in the XMCQDPT2 excited state energy calculations.

Proton release/uptake pathway
In our simulations we observe occasionally an exchange of the pyrrole water with bulk solvent molecules in both Pr anf Pfr. This exchange was observed by monitoring the hydrogen bond distance between BV and nearest water molecule. The pyrrole water exchange involving the residue His260 might provide a pathway for the proton from the chromophore D-ring. Figure S13: Snapshot from MD simulation of Pr state with a fully protonated chromophore illustrating a potential proton wire involving the chromophore, residue His260, pyrrole water and several solvent molecules. Colour code: carbon atoms of BV in turquoise and of the amino acid side chains involved in the proton wires in orange, oxygen atoms in red, nitrogen atoms in blue. Aliphatic hydrogen atoms are not shown.

The effect of the including Range-Separation on the chromophore force field
While the CAM-B3LYP functional has shown good performance for charge transfer excitations and excited-state dynamics in conjugated systems, it may not be the optimal choice for describing the ground state properties in such systems (M. A. Rohrdanz, J. Herbert: Simultaneous benchmarking of ground-and excited-state properties with long-range-corrected density functional theory, J. Chem. Phys. 129 (2008) 034107), we show below ( Fig. S14 and Table S8) that the most critical force field parameters, namely bond lengths, angles and charges are essentially the same, irrespective of whether the geometry has been optimized with B3LYP or CAM-B3LYP." S27 Pr Pfr (a) (d) (c) (f) Figure S14: A superimposed stick representation of the optimized BV chromophore geometry using B3LYP (cyan) and CAM-B3LYP (green) functionals and basis set 6-31G* for (a) Pr and (d) Pfr state with fully protonated chromophore. A linear regression fit of the optimized geometry bond lengths (b)-Pr, (e)-Pfr and partial charges (c)-Pr, (f)-Pfr to compare the effect of using B3LYP instead of CAM-B3LYP for geometry optimization step in the RESP procedure.