The effect of low concentrations of uncouplers on the detectability of proton deposition in thylakoids. Evidence for subcompartmentation and preexisting pH differences in the dark

Flash-induced pH changes inside thylakoids were measured with neutral red as an indicator in the presence and absence of low concentrations of uncouplers. We found that both carrier-type and pore-forming uncouplers caused the rapidly rising phase of the neutral red signal, previously attributed to proton deposition by water oxidation, to disappear. Gramicidin was particularly efficient in this respect, requiring only one molecule of uncoupler per 10⁴ chlorophyll molecules to render the rapid proton deposition undectectable. This suggests that gramicidin did not act on each water-oxidizing enzyme individually, but rather at the level of the thylakoid membrane. In contrast to the effect on water-derived protons, the appearance of protons from plastoquinol was unaffected by gramicidin. Nor did gramicidin affect the rise of the neutral red signal due to proton deposition during two Photosystem I partial reactions with artificial donors. At the low gramicidin concentrations used, its effect on the neutral red signal could not be attributed to a general increase in proton permeability of the thylakoid membrane (acceleration of half decay from 9 to 0.8 s). The extent of alkalinization of the external medium during the first few hundred milliseconds following a light flash was unaltered by gramicidin, and we did not observe a kinetic correlation between the disappearance of the water proton and the decay of the transmembrane electric field. The last two findings suggest that the undetected protons had not crossed the thylakoid membrane, but instead were buffered away by some gramicidin-induced extra buffering power. pH titration of this extra buffering power revealed an apparent pK ranging between 7.2 and 7.7 and a stoichiometry of $\text{2}\text{H}{}^{\mathrm{+}}\text{site}$. The rapid phase of the neutral red signal regained 90% of its original amplitude after seven flashes were applied at 6.7 Hz repetition rate to a sample containing gramicidin. This suggests limits to the extra buffering power. One possible interpretation of our experiments is the following: Protons derived from water oxidation are initially deposited into extended and highly buffering special domains, and only escape into the common internal phase when the buffering capacity of the domains is saturated. As an alternative one may consider that the thylakoid lumen is partitioned into at least two domains, each dominated by different photosystems and with slow proton equilibration between them. Either view requires internal subcompartmentation. The consequences of such subcompartmentation for chloroplast bioenergetics are still obscure.