Examination of abiotic cofactor assembly in photosynthetic biomimetics: site-specific stereoselectivity in the conjugation of a ruthenium(II) tris(bipyridine) photosensitizer to a multi-heme protein

Abstract

To understand design principles for assembling photosynthetic biohybrids that incorporate precisely-controlled sites for electron injection into redox enzyme cofactor arrays, we investigated the influence of chirality in assembly of the photosensitizer ruthenium(II)bis(2,2′-bipyridine)(4-bromomethyl-4′-methyl-2,2′-bipyridine), Ru(bpy) 2 (Br-bpy), when covalently conjugated to cysteine residues introduced by site-directed mutagenesis in the triheme periplasmic cytochrome A (PpcA) as a model biohybrid system. For two investigated conjugates that show ultrafast electron transfer, A23C-Ru and K29C-Ru, analysis by circular dichroism spectroscopy, CD, demonstrated site-specific chiral discrimination as a factor emerging from the close association between [Ru(bpy) 3 ] 2+ and heme cofactors. CD analysis showed the A23C-Ru and K29C-Ru conjugates to have distinct, but opposite, stereoselectivity for the Λ and Δ-Ru(bpy) 2 (Br-bpy) enantiomers, with enantiomeric excesses of 33.1% and 65.6%, respectively. In contrast, Ru(bpy) 2 (Br-bpy) conjugation to a protein site with high flexibility, represented by the E39C-Ru construct, exhibited a nearly negligible chiral selectivity, measured by an enantiomeric excess of 4.2% for the Λ enantiomer. Molecular dynamics simulations showed that site-specific stereoselectivity reflects steric constraints at the conjugating sites and that a high degree of chiral selectivity correlates to reduced structural disorder for [Ru(bpy) 3 ] 2+ in the linked assembly. This work identifies chiral discrimination as means to achieve site-specific, precise geometric positioning of introduced photosensitizers relative to the heme cofactors in manner that mimics the tuning of cofactors in photosynthesis.

Mimicking Natural Photosynthesis: Designing Ultrafast Photosensitized Electron Transfer into Multiheme Cytochrome Protein Nanowires

“…Slight deviations from linearity at low momentum transfer vector values on Figure S1 observed for K28C-Ru and G53C-Ru are consistent with a flexible linkage having the photosensitizers protruding from the protein surface as observed by MD simulations (Figure 6). In contrast, compact globular forms of K29C-Ru and K52C-Ru inferred from the Guinier X-ray scattering plots (Figure S1) are consistent with sterically restricted sites observed in MD simulations (Figure 6) and steric hindrances resulting in Ru(II)(bpy) 3 enantiomer selection (Figure S2), and as reported previously for K29C-Ru [52]. We note that K29C-Ru and K52C-Ru show a strong selectivity for enantiomers of the opposite chirality for the Ru(II)(bpy) 3 derivative, demonstrating that the chiral selectivity arises not from the covalent linkage, but from the stereo-specificity of the binding site.…”

“…To further verify structural stability of PpcA biohybrids and to test for preferential binding of Ru(II)(bpy) 2 (Br-bpy) enantiomers, we collected circular dichroism (CD) spectra at temperatures between 25 • C and 90 • C. Consistent with our recent report [52] that showed preferential binding of the ∆ enantiomer of Ru(II)(bpy) 3 to K29C and Λ enantiomer to A23C, we observed a strong preference for binding of the Λ enantiomer by K52C, and a weak preference for the ∆ enantiomer by K28C (Figure S2). Similar to the previously reported E39C-Ru biohybrid, we did not observe a significant enantiomer selection with G53C.…”

“…The falling edge is correlated with recovery of the ground state spectra and we interpret this process as charge recombination (CR). Kinetic models to describe TA kinetics in Ru-labeled multi-heme cytochromes include those which consider a multi-state scheme, and are found to favor a kinetic scheme where CR is faster than CS [60], and those which assume a simple two-state kinetic model, where CS is faster than CR [51,52]. Unfortunately, the present TA data do not provide a detectable signature of the CR kinetics for Ru(III)(bpy) 3 , so we do not have a direct, independent way to detect the CR reaction and distinguish these two kinetic models.…”

“…However, depending upon the conformation of the linked photosensitizer-PpcA construct, through-space electron transfer pathways may also exist. We note that the K29C site shows a strong selectivity for covalent linkage, selecting the ∆-enantiomer out of a racemic mixture of a Ru(II)(bpy) 3 derivative [52], implying a precise 3D geometry for the linked complex and close spatial positioning to an adjacent Heme-III [51]. A comparison of photosensitizer-to-heme ET using linkage sites within c-type heme binding motifs provides opportunities to compare the dynamics for sites that are similarly bracketed by heme cofactor thioether linkages and thus have common through-bond pathways, but can be anticipated to differ in through-space configurations.…”

“…The mutant forms of PpcA were expressed in E. coli as previously described [51,52]. We obtained protein yields comparable to wild-type PpcA.…”