Electron transfer from the A1A and A1B sites to a tethered Pt nanoparticle requires the FeS clusters for suppression of the recombination channel.

In this work, a previously described model of electron withdrawal from the A1A/A1B sites of Photosystem I (PS I) was tested using a dihydrogen-producing PS I-NQ(CH2)15S-Pt nanoconstruct. According to this model, the rate of electron transfer from A1A/A1B to a tethered Pt nanoparticle is kinetically unfavorable relative to the rate of forward electron transfer to the FeS clusters. Dihydrogen is produced only when an external donor rapidly reduces P700(+), thereby suppressing the recombination channel and allowing the electron in the FeS clusters to proceed via uphill electron transfer through the A1A/A1B quinones to the Pt nanoparticle. We tested this model by sequentially removing the FeS clusters, FB, FA, and FX, and determining the concentration of cytochrome c6 (Cyt c6) at which the backreaction was outcompeted and dihydrogen production was observed. P700-FA cores were generated in a menB insertionally inactivated strain by removing FB with HgCl2; P700-FX cores were generated in a menB psaC insertionally inactivated strain that lacks FA and FB, and P700-A1 cores were generated in a menB rubA insertionally inactivated strain that lacks FX, FA and FB. Quinone incorporation was measured using transient electron paramagnetic resonance spectroscopy and time resolved optical spectroscopy. Cyt c6 was titrated into each of these PS I preparations and the kinetics of P700(+) reduction were measured. A similar experiment was carried out on PS I-NQ(CH2)15S-Pt nanoconstructs assembled from these PS I preparations. This study showed that the concentration of Cyt c6 needed to produce dihydrogen was comparable to that needed to suppress the backreaction. We conclude that the FeS clusters serve to 'park' the electron and thereby extend the duration of the charge-separated state; however, in doing so, the redox advantage of removing the electron at A1A/A1B is lost.