Hydrogen Peroxide Gates a Voltage-Dependent Cation Current in Aplysia Neuroendocrine Cells

Abstract

Nonselective cation channels promote persistent spiking in many neurons from a diversity of animals. In the hermaphroditic marine-snail, Aplysia californica, synaptic input to the neuroendocrine bag cell neurons triggers various cation channels, causing an ∼30 min afterdischarge of action potentials and the secretion of egg-laying hormone. During the afterdischarge, protein kinase C is also activated, which in turn elevates hydrogen peroxide (H₂O₂), likely by stimulating nicotinamide adenine dinucleotide phosphate oxidase. The present study investigated whether H₂O₂ regulates cation channels to drive the afterdischarge. In single, cultured bag cell neurons, H₂O₂ elicited a prolonged, concentration- and voltage-dependent inward current, associated with an increase in membrane conductance and a reversal potential of ∼+30 mV. Compared with normal saline, the presence of Ca²⁺-free, Na⁺-free, or Na⁺/Ca²⁺-free extracellular saline, lowered the current amplitude and left-shifted the reversal potential, consistent with a nonselective cationic conductance. Preventing H₂O₂ reduction with the glutathione peroxidase inhibitor, mercaptosuccinate, enhanced the H₂O₂-induced current, while boosting glutathione production with its precursor, N-acetylcysteine, or adding the reducing agent, dithiothreitol, lessened the response. Moreover, the current generated by the alkylating agent, N-ethylmaleimide, occluded the effect of H₂O₂ The H₂O₂-induced current was inhibited by tetrodotoxin as well as the cation channel blockers, 9-phenanthrol and clotrimazole. In current-clamp, H₂O₂ stimulated burst firing, but this was attenuated or prevented altogether by the channel blockers. Finally, H₂O₂ evoked an afterdischarge from whole bag cell neuron clusters recorded ex vivo by sharp-electrode. H₂O₂ may regulate a cation channel to influence long-term changes in activity and ultimately reproduction.SIGNIFICANCE STATEMENT Hydrogen peroxide (H₂O₂) is often studied in a pathological context, such as ischemia or inflammation. However, H₂O₂ also physiologically modulates synaptic transmission and gates certain transient receptor potential channels. That stated, the effect of H₂O₂ on neuronal excitability remains less well defined. Here, we examine how H₂O₂ influences Aplysia bag cell neurons, which elicit ovulation by releasing hormones during an afterdischarge. These neuroendocrine cells are uniquely identifiable and amenable to recording as individual cultured neurons or a cluster from the nervous system. In both culture and the cluster, H₂O₂ evokes prolonged, afterdischarge-like bursting by gating a nonselective voltage-dependent cationic current. Thus, H₂O₂, which is generated in response to afterdischarge-associated second messengers, may prompt the firing necessary for hormone secretion and procreation.

Keywords: H2O2; bursting; mollusk; peptidergic neuron; redox; reproduction.

Figure 1.

H₂O₂ activates a prolonged, inward, voltage-dependent current in cultured bag cell neurons. A, Left, Phase-contrast photomicrograph showing a superfusion barrel positioned near a bag cell neuron soma with neurites and the recording pipette. Neurons are whole-cell voltage-clamped at −60 mV using our standard K⁺-Asp-based intracellular saline in the pipette and nASW in the bath. Middle, Superfusion of 1 mm ACh (at bar) elicits an inward current. Arrows highlight the start of superfusion and travel time, i.e., the time required for ACh to reach the soma. Right, ACh superfusion produced a peak current (I_ACh) of −2.9 ± 1.6 nA with a latency of 38.0 ± 5.3 s following the switch from control to ACh-containing saline. B, Superfusion of 1 mm H₂O₂ (at bar) over the soma of different neurons, held at −60, −30, or 0 mV, causes increasingly larger inward current. Ordinate applies to all traces. C, Summary current/voltage relationship for the H₂O₂-induced current at −60 mV (91.0 ± 43 pA), −40 mV (51.0 ± 31.0 pA), −30 mV (230.0 ± 74.0 pA), −20 mV (288.0 ± 84.0 pA), or 0 mV (765.0 ± 138.0 pA), shows a threshold between −40 and −30 mV and an apparent maximum at 0 mV. D, Group data illustrating a 59.9 ± 12.4 s latency of the H₂O₂-induced current at −30 mV.

Figure 2.

A concentration-dependent H₂O₂-induced inward current. A, Current responses to bath-applied 30 μm, 100 μm, 300 μm, or 1 mm H₂O₂ (at bar) for different cultured bag cell neurons whole-cell voltage-clamped at −30 mV in nASW with Cs⁺-Asp-based intracellular saline. Scale bars apply to all traces. B, Concentration–response curve reveals increasingly higher amounts of H₂O₂ induces progressively larger currents (30 μm = 4.0 ± 7.0 pA, 100 μm = 66.0 ± 13.0 pA, 300 μm = 94.0 ± 27.0 pA, 1 mm H₂O₂ = 218.0 ± 66.0 pA). The line represents the fit of the data with a four-parameter dose–response equation, and provides an EC₅₀ of 14.2 mm with a Hill slope of 0.64. C, Upon washout of 1 mm H₂O₂ with nASW, at the height of the response, the current recovered largely back to the baseline. Inset, Summary graph shows 73.0 ± 10.6% recovery, calculated by comparing the baseline (before H₂O₂ superfusion) and the stable current at end of the trace.

Figure 3.

H₂O₂ increases membrane conductance. A, Top, Typical example of 1 mm H₂O₂ superfusion (at bar) evoking a current in a bag cell neuron whole-cell voltage-clamped at −30 mV in nASW using Cs⁺-based intracellular saline. Middle, To determine membrane conductance, a 200 ms step to −40 mV is given ∼90 s before H₂O₂ (arrow, circled 1). A second step is taken right before H₂O₂ delivery (circled 2), whereas a third step is taken at the peak of the response (circled 3). Bottom, The current evoked by the step is markedly elevated at the peak of the H₂O₂ current (3; black trace) compared with that taken immediately (2; dark gray) or ∼90 s before H₂O₂ (1; light gray). B, Summary graph shows a significant increase in fold-change conductance, calculated by obtaining the ratios of the step-current taken just before H₂O₂ superfusion versus ∼90 s earlier (control 2/1 = 1.0 ± 0.079) and during H₂O₂ versus just before superfusion (H₂O₂ 3/2 = 1.7 ± 0.33; r = 0.02381; *p = 0.0078, Wilcoxon matched-pairs signed ranks test).

Figure 4.

The H₂O₂-induced current is sensitive to change in extracellular cations. A, Different individual cultured bag cell neurons under whole-cell voltage-clamp at −30 mV using Cs⁺-based intracellular saline. Compared with the current elicited by 1 mm H₂O₂ (at bar) in Na⁺-containing nASW, the response is smaller in Ca²⁺-free or Na⁺-free saline, and mostly eliminated in Na⁺/Ca²⁺-free saline. Scale bars apply to all traces. B, Group data show that, in contrast to nASW (1.2 ± 0.2 nA), Ca²⁺-free (151.6 ± 26.4 pA), Na⁺-free (65.3 ± 15.3 pA), or Na⁺/Ca²⁺-free saline (14.1 ± 7.6 pA) all significantly reduce the current (H = 29.991; df = 2; p < 0.0001, KW-ANOVA; *p < 0.05 nASW vs Ca⁺-free; *p < 0.01 nASW vs Na⁺-free; *p < 0.001, nASW vs Na⁺/Ca²⁺-free, Dunn's multiple-comparisons test). C, Difference currents obtained by subtracting the response to a 5 s, −60 to +60 mV voltage ramp (bottom inset), taken immediately before 1 mm H₂O₂ application, from that taken at the peak of the response. For nASW (black trace), the current is voltage-dependent and reverses between +30 and +40 mV. Upper insets show magnified Ca²⁺-free (dark gray), Na⁺-free (medium gray), and Na⁺/Ca²⁺-free (light gray) difference currents from −45 to −15 mV (left) and −15 to +15 mV (right). D, Average data illustrates a significant left-shift in reversal potential from nASW (30.6 ± 4.1 mV) with Ca²⁺-free (−8.4 ± 3.4 mV), Na⁺-free (−11.3 ± 3.4 mV), or Na⁺/Ca²⁺-free (−10.3 ± 3.7 mV) saline (F_(3,37) = 27.329; p < 0.0001, ordinary ANOVA; *p < 0.01, nASW vs Ca²⁺-free, nASW vs Na⁺-free, and nASW vs Na⁺/Ca²⁺-free saline, Dunnett multiple-comparisons test). Bars as per panel B.

Figure 5.

The pharmacology of the H₂O₂-induced current is consistent with a cation channel. A, Whole-cell voltage-clamp recording from a cultured bag cell neuron in nASW at −30 mV using Cs⁺-Asp-based intracellular saline shows an inward current in response to 1 mm H₂O₂ (at bar). Bath-application of 0.1% (v/v) DMSO (vehicle, at second bar) at the peak of the response fails to alter the normal recovery of the current. B, Delivery of 100 μm 9-Pt (top) or 10 μm clotrimazole (bottom) at the peak of the response markedly inhibits the ongoing 1 mm H₂O₂-induced current. C, D, Summary graphs show both 9-Pt and clotrimazole significantly increase the percent recovery of the H₂O₂-induced current to 76.0 ± 11.4% and 56.5 ± 14.1%, respectively, calculated by comparing the baseline (before H₂O₂ bath-application) and the steady-state current at the end of the trace (t₍₈₎ = 2.743; *p = 0.0253, 9-Pt, t₍₆₎ = 3.183; *p = 0.0190, clotrimazole, both unpaired Student's t test). E, Group data of the H₂O₂-induced current in control vs neurons pretreated with 10 μm SKF-96365. In contrast to 1 mm H₂O₂ alone (93.4 ± 11.4 pA), the presence of 10 μm SKF-96365 (73.3 ± 11.2 pA) does not significantly reduce the response (U_(5,12) = 21.0; p = 0.3827 H₂O₂ vs H₂O₂ in SKF-96365; Mann–Whitney U test).

Figure 6.

The H₂O₂-induced current is reduced by tetrodotoxin. A, Current responses to 1 mm H₂O₂ (at bar) of separate cultured bag cell neurons whole-cell voltage-clamped at −30 mV in nASW with Cs⁺-based intracellular saline. Compared with control (top), a 30 min pretreatment with 100 μm TTX (middle) noticeably reduces the H₂O₂-induced current, while 300 μm, TTX (bottom) almost eliminates the response. Ordinate applies to all traces. B, Group data of the H₂O₂-induced current in 100 or 300 μm TTX vs control. Compared with 1 mm H₂O₂ alone (79.9 ± 7.7 pA), the presence of 100 μm (47.1 ± 9.4 pA) or 300 μm TTX (15.4 ± 3.6 pA) significantly reduces the response (F_(2,18) = 12.570; p = 0.0004 ordinary ANOVA, *p < 0.05 H₂O₂ vs H₂O₂ post 100 μm TTX; *p < 0.01 H₂O₂ vs H₂O₂ post 300 μm TTX, Dunnett multiple-comparisons test).

Figure 7.

The H₂O₂-induced current is enhanced by mercaptosuccinate and reduced by N-acetylcysteine or dithiothreitol. Cultured bag cell neurons are bathed in nASW and whole-cell voltage-clamped at −30 mV using Cs⁺-based intracellular saline. A, Bath-application (at bar) of 1 mm mercaptosuccinate alone has little effect (top), whereas delivery of 1 mm H₂O₂ to a second cell elicits a noticeable inward current (middle). Moreover, in a third neuron initially given mercaptosuccinate (mercapto), the H₂O₂-induced current is enhanced by ∼50% (bottom). Ordinate applies to all traces. B, Summary graph of peak mercaptosuccinate- and H₂O₂-induced current ± mercaptosuccinate. There is a significant difference between both the H₂O₂-induced current with (98.1 ± 9.6 pA) and without (66.4 ± 5.8 pA) mercaptosuccinate as well as mercaptosuccinate alone (7.3 ± 7.2 pA; F_(2,21) = 50.879; p < 0.0001, ordinary ANOVA; *p < 0.001 mercapto vs H₂O₂; *p < 0.05 H₂O₂ vs H₂O₂ in mercapto, Tukey–Kramer multiple-comparisons test). C, Introducing 100 μm NAC has only a nominal impact (top), whereas 1 mm H₂O₂ generates an inward current (middle). In the presence of N-acetylcysteine, the H₂O₂-induced current is reduced by ∼40% (bottom). D, Summary data showing that compared with N-acetylcysteine alone (43.3 ± 25.7 pA), the H₂O₂-induced current differs significantly with (90.3 ± 21.4 pA) and without N-acetylcysteine (148.0 ± 16.0 pA; U_(6,7) = 0; *p = 0.0034 NAC vs H₂O₂; U_(6,7) = 5; *p = 0.0221 H₂O₂ vs H₂O₂ in NAC, both Mann–Whitney U test). E, Minimal response to 1 mm DTT alone (top), whereas 1 mm H₂O₂ again elicits a clear inward current (middle); in addition, when DTT is already present, the H₂O₂-induced current is reduced by ∼70% (bottom). F, Group data showing that compared with DTT alone (11.3 ± 4.5 pA), the H₂O₂-induced current differs significantly with (43.2 ± 9.6 pA) and without DTT (155.0 ± 13.2 pA; F_(2,15) = 73.831; p < 0.0001, ordinary ANOVA; *p < 0.001 DTT vs H₂O₂; *p < 0.001 H₂O₂ vs H₂O₂ in DTT, Tukey–Kramer multiple-comparisons test).

Figure 8.

NEM- and H₂O₂-induced currents are occlusive. Representative response to 300 μm NEM or 1 mm H₂O₂ of different cultured bag cell neurons whole-cell voltage-clamped at −30 mV in nASW with Cs⁺-based intracellular saline. A, Bath-application (at bar) of NEM evokes a prominent inward current. Yet, H₂O₂ (at second bar) in the presence of NEM, yields little further change. B, Delivery of H₂O₂ elicits a typical inward current in a separate neuron, but there is no obvious current when NEM is applied in the presence of H₂O₂. C, D, Summary data demonstrates a significant difference between the NEM-elicited current (201.0 ± 65.5 pA) and the H₂O₂-induced current post-NEM (11.7 ± 13.1 pA; U_(9,9) = 7.0; *p = 0.0019, Mann–Whitney U test). Similarly, there is a significant difference between the H₂O₂-induced current (129.6 ± 24.5 pA) and the NEM-elicited current post-H₂O₂ (6.3 ± 5.9 pA; (t₍₄₎ = 4.90; *p = 0.008, unpaired Student's t test, Welch corrected). Note, there is no statistical difference between the initial current produced by NEM and H₂O₂ (white bars; t₍₉₎ = 1.022; p = 0.3337, unpaired Student's t test, Welch corrected); as well, the NEM- and the H₂O₂-induced currents in the presence of H₂O₂ and NEM, respectively, are not statistically different (black bars; U_(5,9) = 14.0; p = 0.2977, Mann–Whitney U test).

Figure 9.

H₂O₂ depolarizes and induces robust action potential firing. A, Separate cultured bag cell neurons are whole-cell current-clamped in nASW with K⁺-based intracellular saline and initially set to either −40 or −60 mV with bias current. Bath-application (at bar) of H₂O (0.5% v/v) at −40 (top) or −60 mV (bottom) does not affect membrane voltage. Ordinate applies to both traces. B, Exposure to 100 μm H₂O₂ depolarizes two different neurons from both −40 (top) and −60 (bottom) mV, but neither reaches threshold. Ordinate applies to both traces. C, Delivery of 1 mm H₂O₂ depolarizes the membrane and induces robust action potential firing from both −40 (top) and −60 (bottom) mV in separate cells. Ordinate applies to both traces. D, Summary graph indicating the average depolarizations from −40 (open circles) or −60 (closed circles) mV. Compared with the response produced by applying H₂O (−40 mV: 2.8 ± 0.8 mV; −60 mV: 1.1 ± 0.6 mV), the depolarization elicited by 100 μm H₂O₂ (−40 mV: 7.5 ± 0.6 mV; −60 mV: 14.0 ± 2.6 mV) or 1 mm H₂O₂ (−40 mV: 11.4 ± 2.9; −60 mV: 21.8 ± 2.1) is significantly different (−40 mV: H = 8.545; df = 1; p = 0.0070, KW-ANOVA; p > 0.05 H₂O vs 100 μm H₂O₂; *p < 0.05 H₂O vs 1 mm H₂O₂, Dunn's multiple-comparison's test; −60 mV: F_(2,12) = 29.275; p < 0.0001, ordinary ANOVA; *p < 0.01 H₂O vs 100 μm H₂O₂, *p < 0.001 H₂O vs 1 mm H₂O₂, Tukey–Kramer multiple-comparisons test). E, Group data show no significant difference in the frequency (t₍₈₎ = 0.2841; p = 0.7835, unpaired Student's t test) or duration (t₍₈₎ = 0.3884; p = 0.7079, unpaired Student's t test) of action potential firing from −40 versus −60 mV caused by 1 mm H₂O₂. In addition, there is no significant difference in the latency, i.e., the time it takes for the neuron to start firing action potentials post-H₂O₂ delivery, between responses at −40 mV (2.8 ± 0.6 mV) and −60 mV (3.0 ± 0.6 mV; t₍₈₎ = 0.3001; p = 0.7718, unpaired Student's t test).

Figure 10.

H₂O₂-evoked spiking is reduced by tetrodotoxin and prevented by 9-Pt or clotrimazole. Voltage responses to 1 mm H₂O₂ of different cultured bag cell neurons whole-cell current-clamped to −40 mV in nASW with K⁺-based intracellular saline after 20 min pretreatment with clotrimazole, 9-Pt, or TTX. A, Bath-application of 1 mm H₂O₂ (at bar) induces robust action potential firing (top). However, a 20 min pretreatment with 10 μm clotrimazole (upper middle) or 100 μm 9-phenanthrol (lower middle) virtually eliminates the depolarization and prevents firing all together. Finally, after 20 min of 300 μm TTX (bottom), introducing H₂O₂ still leads to membrane depolarization, but the duration and frequency of action potential firing is lessened. The ordinate applies to all traces. B, Summary graph indicating the average H₂O₂-evoked depolarization (top) in the presence of 300 μm TTX does not differ significantly from 1 mm H₂O₂ alone (U_(4,18) = 32.0; p = 0.7743, Mann–Whitney U test). However, spike frequency (middle) and duration (bottom) are significantly reduced when TTX is in the bath (frequency: U_(4,18) = 6.0; *p = 0.0120; duration: U_(4,18) = 5.0; *p = 0.0049, both Mann–Whitney U test).

Figure 11.

H₂O₂ depolarizes bag cell neurons and initiates an afterdischarge in desheathed clusters. A, Under sharp-electrode current-clamp, a 2 s 1 mm ACh pressure ejection (arrow) to one side of the cluster depolarizes a bag cell neuron recorded on the opposite side. B, Bath-application of 10 mm H₂O₂ (bar) post-ACh delivery initiates a prolonged burst in a bag cell neuron within a cluster. Inset, Phase-contrast photomicrograph of a desheathed bag cell neuron cluster in tcASW indicating the placement of the ACh-containing pressure ejection electrode and the intracellular current-clamp (cc) sharp electrode. C, Following 1 mm ACh pressure ejection, 10 mm H₂O₂ is introduced 8.3 min later and induces a depolarization (23.2 ± 2.5 mV; t₍₁₂₎ = 3.375; *p = 0.0055, unpaired Student's t test) that evokes a burst, whereas 10 mm H₂O₂ alone solely depolarizes the neuron, but to a significantly lesser extent (9.8 ± 3.6 mV; left). During the H₂O₂-evoked burst, the frequency of action potential firing is 0.8 ± 0.2 Hz (left), mean discharge duration is 6.4 ± 1.4 min (middle), and the latency is 3.4 ± 0.7 min (right).