15–0.42 pH units), considerably larger than the average alkalinization value of 30.7 nM reported above. Regions with predominantly

alkalinizing changes were often only ∼5 μm away from regions with predominantly acidifying responses (e.g., compare regions of interest [ROIs] #1 and #2 in Figure S3), and these local gradients remained “standing” for ∼60 s after stimulation ended. Steep cytosolic pH gradients over a distance of <5 μm have also been reported in isolated snail neurons (Schwiening and Willoughby, 2002). The Discussion considers possible functional implications of these spatially heterogeneous pH changes. Figure 2 shows that both acidifying and alkalinizing components of the stimulation-induced [H+] response selleck were eliminated by replacing bath Ca2+ with Mg2+ (Figure 2A), or by adding the nonspecific Ca2+ channel blocker Cd2+ (100 μM) in the presence of normal bath [Ca2+] (Figure 2B). Responses were also blocked by ω-agatoxin GIVA (0.5 μM), which blocks the selleck screening library P/Q-type channels (Cav2.1) that are responsible for most stimulation-evoked Ca2+ influx into mammalian motor terminals (Katz et al., 1997). Even when Ca2+ influx is blocked, Na+-dependent action potentials continue to invade mouse motor terminals (Konishi and Sears, 1984). These results thus demonstrate that stimulation-induced acidification

and alkalinization both require Ca2+ influx into motor terminals, rather than simply depolarization or Na+ entry. As reviewed in the Introduction, neuronal somata and dendrites undergo acidification during trains of action potentials, but do not display an alkalinizing component. Thus we wondered whether the prominent alkalinizing component

measured in motor terminals might be related to vesicular recycling. Figures 3Aa and 3Ab show [H+] changes recorded when exocytosis was blocked with botulinum neurotoxins (BoNTs type A and E), which cleave SNAP-25, a SNARE protein required for action potential-evoked transmitter release (Bronk et al., 2007). BoNTs do not interfere with action potential-induced Ca2+ influx (Van der Kloot and Molgó, 1994 and Schiavo et al., 2000). Both BoNTs transformed the biphasic acidification-alkalinization response recorded Adenosine triphosphate in control solution into a “pure” acidification. [H+] rapidly (∼5 s) increased to a plateau level during stimulation and recovered to baseline within ∼5 s after stimulation stopped. BoNTs did not increase the magnitude of the acidification measured at the onset of stimulation, suggesting that the lack of an alkalinization phase at the end of stimulation was not due to augmentation of an acidifying process but rather to block of an alkalinizing process. Results with both BoNTs were similar and thus were combined in the averaged values of early and late response components plotted in Figure 3Ca. The diagram in Figure 3Cb indicates the mechanism of action of all drugs applied in Figure 3.