Fig. 1. Simultaneous records of membrane potential (E_m), voltage-clamp current, HCl injection current, and pH_i (from top to bottom) that enabled the discovery of voltage-gated proton channels in snail neurons (Helix aspersa) by Thomas and Meech (1982). Several HCl injections were made to lower pH_i to <6.5, then the membrane was depolarized stepwise. At −10 mV and at more positive voltages a decaying outward current is observed. The recovery of pH_i (bottom trace) from the acid load clearly is faster at more positive voltages, consistent with the outward current being carried by H⁺. The outward H⁺ current decays because the driving force (V−E_H) decreases as pH_i increases (due to the continuous H⁺ efflux). H⁺ currents in small cells like microglia decay two orders of magnitude faster because of the much smaller cell volume than in snail neurons. The neuron was pretreated with CsCl to inhibit K⁺ currents and SITS (4-acetamido-4′-isothiocyanatostilbene-2,2′-disulfonic acid) to inhibit other endogenous pH regulatory mechanisms. Taken from Fig. 3 of Thomas and Meech (1982) with the permission of Nature (http://www.nature.com)

Fig. 2. (A) The determination of reversal potential, V_rev, is illustrated in a murine microglial cell. A depolarizing prepulse opens many channels, and then when the potential is repolarized to various voltages (diagram), ‘tail currents’ are observed, which decay as H⁺ channels close. The initial tail current is outward or inward depending on the relation between the potential and V_rev; at V_rev there is no net current. (B) Mean values of V_rev at several ΔpH are plotted, along with a line which shows a slope of 40 mV/Unit. The Nernst potential for H⁺, has a slope 58 mV/Unit. The discrepancy between V_rev and E_H is likely due to imperfect control of the local pH near the membrane, and some deviation is seen in all studies of voltage gated proton channels. Taken from Fig. 6 of Eder et al. (1995a) with the permission of Springer-Verlag GmbH & Co. KG.

Fig. 3. Families of H⁺ currents in the same microglial cell at several pH_o. Identical families of voltage pulses were applied in each solution (see inset). At higher pH_o, H⁺ channels open at more negative voltages; conversely, low pH_o inhibits the current. In (E) the currents at the end of the pulses in the four solutions (A–D) are plotted: A, pH_o 8.2 (•); B, pH_o 7.4 (); C, pH_o 6.6 (▪), D; pH_o 5.8 (♦). Taken from Fig. 5 of Eder et al. (1995a) with the permission of Springer-Verlag GmbH & Co. KG.

Fig. 4. The threshold voltage for activating H⁺ currents, defined as the voltage at which clearly time-dependent outward current was detected, V_threshold, is plotted as a function of V_rev measured in the same cell and the same solution. The dotted line shows equality of the two parameters; all of the data fall above this line, indicating that V_threshold is always positive to V_rev. Open symbols indicate measurements with H₂O in the bath, filled symbols with D₂O. Bath solutions included pH_o ranging 6.5–10.0 and pD_o 7.0–10.0. Pipette solutions (pH_i) are indicated by the shape of the symbol: pD 7.0 (Δ), pD 8.0 (), pD 9.0 (hexagons), pH 7.5 (), pH 5.5 (○), 50 mM NH₄⁺ (□) The lines show the results of linear regression of the H₂O data (solid line), r=0.963, SLOPE=0.760, Y-INTERCEPT=18.1 mV. The D₂O data (dashed line) were described by r=0.926, SLOPE=0.750, INTERCEPT=22.0 mV. Data in alveolar epithelial cells, from Fig. 11 of DeCoursey and Cherny, 1997 with the permission of the publisher.

Fig. 5. Model of the regulation of H⁺ channel gating by the pH gradient, ΔpH. The state diagram at the bottom defines the kinetic model; the three cartoons illustrate possible physical manifestations of this model. The closed channel conformation (state 1) is stabilized by protonation of an external site, and the open conformation (state 4) is stabilized by protonation of an internal site. These hypothetical allosteric regulatory protonation sites might be the same or distinct. Only the external or internal site is accessible to protonation at any given time, not both simultaneously. A conformational change (the transition between states 2 and 3) exposes the protonation site to the internal solution. The formation of a conducting H⁺ channel requires a conformational change in each channel protomer which can occur only when the regulatory site is deprotonated. The open channel probability therefore increases at high pH_o or low pH_i. The model defined by the state diagram can be envisioned physically as (top) a ‘butterfly’ in which the protonation site on each channel protomer or ‘wing’ moves across the membrane, (middle) distinct external and internal sites which when protonated, allosterically prevent protonation at the opposite site, (lower) a protonation site in a proton well whose accessibility depends on a small conformational change, or other variants not illustrated. The voltage dependence of H⁺ channel gating could arise either from voltage-dependent binding/unbinding of protons to the regulatory protonation site, or from a voltage dependent conformational change, or some combination of the two. We assigned all of the voltage-dependence to proton binding, so that the regulatory sites behave like ‘proton wells’. The gating of voltage gated proton channels in rat alveolar epithelial cells was described by the following parameters: d_in=d_out=0.71, K_w=10, m=0, n=1.5, k₁₂=1000 s⁻¹, k₃₂=10⁶ s⁻¹, k_43,fast=3 s⁻¹, k_43,slow=0.05 s⁻¹, pK_in=pK_out =8.5. Taken from Fig. 10 in Cherny et al. (1995) with the permission of the publisher.

Fig. 6. Identical families of pulses were applied in the absence (A) or presence (B) of 50 μM ZnCl₂ in a microglial cell. Note the profound slowing of the turn-on of the H⁺ current. The currents at the end of the pulses in A and B are plotted in C. Note that the V_threshold is shifted by 60 mV toward more positive voltages. Taken from Fig. 7 of Eder et al. (1995a).

Fig. 7. The morphological changes during ACM-induced deactivation of microglia. Untreated cultured murine microglia are amoeboid cells with few processes (A). After treatment with ACM (B) the microglia become markedly ramified. Adapted from Fig. 1 of Eder et al. (1999).

Fig. 8. Families of H⁺ currents in microglial cells before (A, D) or after treatment for 1 day with 2 μM cytochalasin D (B), 20 μM phalloidin (C), 1 μM colchicine (E), or 0.5 μM taxol (F). All families recorded during the same pulse sequence (inset). Taken from Klee et al. (1998).

Fig. 9. Whole-cell H⁺ currents at +100 mV in a human neutrophil before, during and after exposure to 50 μM arachidonic acid. The current was reversibly enhanced by arachidonic acid. The pH was 7.0//6.0. Adapted from DeCoursey and Cherny (1993).

Fig. 10. Role of H⁺ channels in the ‘respiratory burst’ of phagocytes. After engulfing a bacterium into the phagocytic vacuole, the normally quiescent enzyme NADPH oxidase assembles in the phagosome membrane from five major and several other components from the membrane and cytosol. Reactive oxygen species generated by this enzyme are essential to the killing process. Superoxide anion, O⁻₂, dismutates spontaneously to form hydrogen peroxide, H₂O₂. Myeloperoxidase then combines H₂O₂ with Cl⁻ and H⁺ to form hypochlorous acid, HOCl, the active ingredient in Chlorox™ bleach as well as being a major bactericidal compound in phagocytes. The stoichiometries of the reactions shown are simplified. During O⁻₂ production, protons are released into the cell. To prevent accumulation of H⁺ in the cell, which would depolarize V_m and lower pH_i, H⁺-selective ion channels open in the cell membrane allowing passive H⁺ extrusion. The H⁺ channel activated during the respiratory burst is shown here as a distinct molecule from the oxidase complex, but it has been proposed that it is contained within the gp91^phox subunit (Henderson; Henderson and Henderson). Taken from DeCoursey and Grinstein (1999).

Fig. 11. Comparison of the density of voltage gated proton current and the ability of various phagocytic cells (•) and non-phagocytic cells (□) to produce superoxide anion in response to PMA (or LPS for THP-1 cells). The dashed line indicates the H⁺ current required to extrude all of the acid at the rate it is produced during the respiratory burst; the shaded area shows H⁺ current amplitudes that would be inadequate to serve this purpose. The intent here is to give a general impression of the relative magnitudes of these two cellular processes; these data were collected by many different groups, in a wide variety of conditions, and although we attempted to ‘standardize’ the results, they are not all directly comparable. The ‘standard’ conditions are nmoles O⁻₂/10⁶ cells collected over 30 min, and maximum I_H measured in the whole-cell configuration at pH_i 5.5 and pH_o 7.0. However, superoxide anion release was measured in different ways, and the attempt to standardize the results to a 30 min sample period ignores non-linearity in the temporal response. Similarly, some measurements of I_H were at different pH_i — in cases where multiple values were given, data at the lowest pH_i studied was used. Two values connected by a dotted line are given for HL-60 cells before and after * being induced to differentiate by DMSO (Qu et al., 1994; personal communication from S. Grinstein). Other sources of data for this figure: I_H: (DeCoursey; Kapus; DeCoursey; Eder; Schrenzel; Gordienko and Cherny) unpublished data of V.V. Cherny and T.E. DeCoursey). O⁻₂: (Petreccia; Swallow; Chao; Leibbrandt and vanunpublished measurements by T. Iastrebova, V.V. Cherny and T.E. DeCoursey). Cell types as indicated in the figure are: BASOPHIL=human basophil, type II=rat alveolar type II epithelial cell, MLS-9=rat microglial cell line, BV-2=rat microglial cell line, Mφ=mouse macrophage, PMN=human neutrophil, THP-1=human monocytic cell line, μglia=mouse microglia, EOSIN=human eosinophil. There is uncertainty about the rate of O⁻₂ release by non-phagocytes because the rates are very low and because of possible contamination by other types of cells (especially problematic is contamination by phagocytes). Several cells that express voltage-gated proton channels but do not produce superoxide anion (snail neurons, amphibian oocytes, kidney epithelial cells, mast cells) are not included on the graph. In contrast with Table 1, which expresses the density of H⁺ channels in the membrane, the values plotted here are for the total current in the whole cell membrane, and thus reflect cell size as well as current density. For example, eosinophils are small cells with ten times denser H⁺ channel expression than THP-1 cells, but because THP-1 cells have ten times larger membrane surface area, their whole-cell H⁺ current is similar.

Fig. 12. Families of H⁺ currents in a snail neuron (left panel) at pH_o 7.4 and pH_i 5.9 (Byerly et al., 1984) and in a murine microglial cell (right panel) at pH_o 7.4 and pH_i 5.8 (Fig. 5B of Eder et al., 1995a). Note the different time scales. The measurements were at similar pH and both were at room temperature (left, 20–25°C, right, 20–23°C). Modified from Byerly et al. (1984) and Eder et al. (1995a).

Fig. 13. Clues to the identity of the H⁺ channels activated during the respiratory burst in human neutrophils. Both H⁺ currents and electron currents, which reflect the electrogenic pumping of electrons across the cell membrane by NADPH oxidase, were recorded simultaneously in neutrophils studied in permeabilized-patch configuration. (A) The amplitude of the maximum g_H, g_H,max, before and after treatment with PMA was tightly correlated in human neutrophils. This result suggests that PMA modulates pre-existing H⁺ channels, rather than inducing a new type of channel. (B) In contrast, there was no correlation between the amplitudes of electron currents and H⁺ currents in individual human neutrophils. As discussed further in the text, this result is hard to reconcile with the notion that a component of the oxidase forms the H⁺ channels activated during the respiratory burst. Taken from DeCoursey et al. (2000).

Table 1. I_H density in cells reported to have H⁺ channelsc

Table 2. Varieties of H⁺ Channelse