We suggest that the presence of a circular array of 98 charges, with a net charge of −14, could restrict the possible pathways for association and find more dissociation between the cyt c 2 and the periplasmic surface of the RC. It seems likely that, when leaving the docking site on the RC, the cyt c 2 has to migrate out of plane (perpendicular to the membrane surface) in order to be able to evade the LH1-charged residues. This probable pathway coincides with the pulling trajectory

in our experiments (both PF-QNM and SMFS). The His-Ni2+-NTA coordination chemistry used in these experiments can achieve long-lasting attachment of proteins (several thousand force–distance cycles, up to several hours) to a functionalised AFM probe or surface, while sustaining significant force stresses (up to 500 pN) (Schmidt et al. 2002; Conti et al. 2000; Nevo et al. 2003; Berquand et al. 2005; Dupres et al. 2005) without the need of a covalent chemical linkage, thus preserving

the protein structure and functionality. The unbinding forces measured during the PF-QNM experiments, approximately 480 pN, are close to the estimated values in a theoretical study of the pathways for cyt c 2–RC-LH1-PufX interaction, with forces in the range 600–1,000 pN (Pogorelov et al. 2007). It is worth noting the fact that purified RC-LH1-PufX complexes were X-396 mouse found to retain 25–30 % of the endogenous quinone acceptor pool (Comayras et al. 2005). This quinone pool stores the electrons resulting from the photo-oxidation of the primary donor, thus maintaining the turnover of the electron transfer cycle during image acquisition. Cyt c 2 to RC electron transfer is in the low microsecond range (Overfield et al. 1979; Moser and Dutton 1988) so, having being brought into contact with a surface-bound RC, the cyt c 2 will be immediately oxidised, given the Tau-protein kinase dwell

time of ~160 μs (PF-QNM) or several milliseconds (SMFS). Thus, the interaction sampled by the unbinding events is likely to be between RC[red] and an oxidised cyt c 2 (cyt c 2[ox]). The negative control shows that a cyt c 2[red]–RC[red] interaction has a low probability. The more probable interaction arises when the initial states (prior to the tip-surface encounter) consist of reduced cyt c 2 and oxidised reaction centres. Therefore, it appears that the cyt c 2-RC association is maintained in the aftermath of the initial electron transfer event, although the oxidised cyt c 2—reduced RC control, which would have helped dissect the nature of the complex, was not performed. This study does, however, probe the long-lived forces that stabilise the cyt c 2–RC association, under controlled conditions of force and distance.