H₂O₂ augments cytosolic calcium in nucleus tractus solitarii neurons via multiple voltage-gated calcium channels

Abstract

Reactive oxygen species (ROS) play a profound role in cardiorespiratory function under normal physiological conditions and disease states. ROS can influence neuronal activity by altering various ion channels and transporters. Within the nucleus tractus solitarii (nTS), a vital brainstem area for cardiorespiratory control, hydrogen peroxide (H₂O₂) induces sustained hyperexcitability following an initial depression of neuronal activity. The mechanism(s) associated with the delayed hyperexcitability are unknown. Here we evaluate the effect(s) of H₂O₂ on cytosolic Ca²⁺ (via fura-2 imaging) and voltage-dependent calcium currents in dissociated rat nTS neurons. H₂O₂ perfusion (200 µM; 1 min) induced a delayed, slow, and moderate increase (~27%) in intracellular Ca²⁺ concentration ([Ca²⁺]_i). The H₂O₂-mediated increase in [Ca²⁺]_i prevailed during thapsigargin, excluding the endoplasmic reticulum as a Ca²⁺ source. The effect, however, was abolished by removal of extracellular Ca²⁺ or the addition of cadmium to the bath solution, suggesting voltage-gated Ca²⁺ channels (VGCCs) as targets for H₂O₂ modulation. Recording of the total voltage-dependent Ca²⁺ current confirmed H₂O₂ enhanced Ca²⁺ entry. Blocking VGCC L, N, and P/Q subtypes decreased the number of cells and their calcium currents that respond to H₂O₂. The number of responder cells to H₂O₂ also decreased in the presence of dithiothreitol, suggesting the actions of H₂O₂ were dependent on sulfhydryl oxidation. In summary, here, we have shown that H₂O₂ increases [Ca²⁺]_i and its Ca²⁺ currents, which is dependent on multiple VGCCs likely by oxidation of sulfhydryl groups. These processes presumably contribute to the previously observed delayed hyperexcitability of nTS neurons in in vitro brainstem slices.

the nucleus tractus solitarii (nTS) in the brainstem is critical for cardiorespiratory regulation. It constitutes the first step for central integration and processing of peripheral visceral information (1). Many neuromodulators alter cellular activity within the nTS (30, 33, 43) before information is conveyed to other brain centers in the cardiorespiratory reflex axis. Reactive oxygen species (ROS) are recognized neuromodulators/second messengers that effectively participate in cardiorespiratory function. For example, NADPH oxidase-derived ROS play an important role in angiotensin II signaling within the nTS (55) and during chemoreflex activation in the rostral ventrolateral medulla (40). Also, antioxidants influence baroreflex control in conscious rats, implicating active involvement of ROS (17). However, ROS are also associated with dysfunction of cardiorespiratory regions in disease states like obstructive sleep apnea, heart failure, and hypertension (8, 12, 56).

Among the members of the ROS family, H₂O₂ exhibits many features important for neuromodulation and second messenger function. It is produced upon stimulation, has specific cellular targets, is relatively long lived, easily crosses membranes, and its actions are terminated by specific antioxidants (47). Its influence on cellular activity can occur through various different types of ion channels. For example, H₂O₂ excites rat hippocampal CA1 neurons by increasing currents through voltage-activated sodium channels (38). On the other hand, H₂O₂ may inhibit spike activity by increasing currents of ATP-sensitive K⁺ channels as in substantia nigra neurons (3) or by increasing intracellular Cl⁻ through GABA_A receptor-gated channels in the cortex and hippocampus (48). ROS may also alter Ca²⁺ homeostasis via channels situated on the neuronal membrane (34, 59) or on intracellular Ca²⁺ stores (16).

With regards to the nTS, in vivo microinjection studies showed that H₂O₂ elicits bradycardia and depressor responses (7), possibly via increased output of nTS neurons in the baroreflex pathway. Using the in vitro brainstem slice, we showed that H₂O₂ exerted a biphasic effect in nTS neurons (42). Initially, H₂O₂ inhibited neuronal activity by hyperpolarizing the membrane potential via increased current of barium-sensitive K⁺ channels. Upon H₂O₂ removal, a delayed and sustained hyperexcitability became apparent. This hyperexcitability was due to a return and further depolarization of the membrane potential above resting levels and a decreased action potential threshold voltage. However, the mechanism(s) underlying the hyperexcitability were unknown. Wang et al. (54) demonstrated an angiotensin II- and ROS-related increase of L-type Ca²⁺ current in the nTS. Thus, in this study, we focused specifically on the effects of H₂O₂ on cytosolic Ca²⁺ and its entry in primary dissociated nTS neurons using the fluorescent dye fura-2 and patch-clamp recordings. Our data show that H₂O₂ increases intracellular Ca²⁺ via multiple voltage-gated Ca²⁺ channels (VGCCs) in a time-delayed fashion.

MATERIALS AND METHODS

Animals

Male Sprague-Dawley rats (Harlan; n = 54, aged 3–4 wk) were maintained in the Association for Assessment and Accreditation of Laboratory Animal Care-accredited vivarium of the Dalton Cardiovascular Research Center. Rats were held on a 12-h day-night cycle at 22°C and 40% humidity with water and food available ad libitum. All experimental protocols were conducted in accordance with National Institutes of Health guidelines (Guide for the Care and Use of Laboratory Animals) and approved by the Animal Care and Use Committee of the University of Missouri.

Cellular Imaging

Primary nTS cell culture.

Similar to our previous studies (31), the brainstem of anesthetized rats (isoflurane; VetOne, Boise, ID) was quickly removed and chilled in ice-cold low calcium-high magnesium artificial cerebral spinal fluid (in mM: 124 NaCl, 3 KCl, 1.2 NaH₂PO₄, 1.2 MgSO₄, 25 NaHCO₃, 11 d-glucose, 1 CaCl₂, and 2 MgCl₂, saturated with 95% O₂-5% CO₂, pH 7.4, ~300 mosM). Horizontal slices (~400 µm) containing the nTS were cut using a vibrating microtome (VT 1000S; Leica). Subsequently, the nTS was excised using a razor blade and placed in Hank’s balanced salt solution (HBSS; no calcium, no magnesium; GIBCO, Big Cabin, OK) with 10 U/ml activated papain (Worthington, Lakewood, NJ). The papain was activated with EDTA (0.27 mM), mercaptoethanol (0.017 mM), and cysteine-HCl (1.37 mM; Sigma, St. Louis, MO). After 30 min in the incubator (MCO-18AIC, Sanyo) at 37°C and 5% CO₂, the papain-HBSS solution was replaced with ~1 ml Dulbecco's modified Eagle’s medium (DMEM; high glucose; GIBCO) containing 10% horse serum (GIBCO) to stop enzymatic activity. The tissue was then triturated using progressively smaller fire-polished glass pipettes until cells were visibly dispersed. Cells were concentrated (700 rpm for 5 min; Centrifuge 5804 R; Eppendorf), and the supernatant was replaced with ~0.3 ml Neurobasal-A (GIBCO) with added B-27 supplement (2%; GIBCO), GlutaMAX (1%; GIBCO), insulin-transferrin (1%; GIBCO), and penicillin streptomycin (1%; GIBCO). nTS cells were plated on poly-d-lysine (100 µg/ml) coated glass coverslips and allowed to settle and attach for 2 h in the incubator. Subsequently, dishes were flooded with Neurobasal-A as culture medium. Recordings were made on the next day.

Fura-2 calcium imaging.

One fura-2 AM vial (50 µg; ThermoFisher Scientific/Invitrogen, Eugene, OR) was reconstituted with 50 µl of 20% Pluronic F-127 in DMSO (Molecular Probes). Cells were loaded for 30 min with 0.05% fura-2 AM in Neurobasal-A on a shaker at 37°C and 5% CO₂ in the dark. Then, cells were washed two times and incubated for another 30 min at room temperature with extracellular recording solution containing the following (in mM): 136 NaCl, 5.4 KCl, 0.33 NaH₂PO₄, 10 HEPES, 10 d-glucose, 1.8 CaCl₂, and 1 MgCl₂. The glass coverslips with fura-2-loaded nTS cells were placed in a superfusion chamber on an inverted microscope (IX71; Olympus) and superfused at a rate of 500–800 µl/min using a positive pressure regulator (AirPulse 7500; Fishman). Cells were visualized using a ×40 objective and a camera (Retiga Exi Fast 1394; QImaging). The fluorophore was excited at 340 and 380 nm with a Polychrome V monochromator (TILL Photonics) and visualized at ~510 nm. Images were acquired every 5–10 s using Micro-Manager software (Open Imaging).

Protocols and drugs.

Baseline periods were recorded for ≥2 min before changes were induced in cytosolic Ca²⁺ with 1-min perfusions of H₂O₂ (100–500 µM; certified ACS 30%; Fisher Scientific) or high K⁺ (55.4 mM; NaCl reduced to 86 mM). Based on our initial observations and the delayed responses to H₂O₂ in nTS slices (42), we allowed a ≥10-min window after H₂O₂ to monitor any Ca²⁺ responses. To examine the role of H₂O₂ during depolarization-induced increases in cytosolic Ca²⁺, cells were exposed to repetitive 1-min high K⁺ stimulation in the absence and presence of a 5-min perfusion of H₂O₂. Based on the Goldman-Hodgkin-Katz equation, perfusion with high K⁺ depolarizes the cell membrane potential by ~39 mV, assuming a physiological intracellular composition of the following (in mM): 15 Na⁺, 150 K⁺, and 10 Cl⁻, and permeability values relative to K⁺ of p_K:p_Na:p_Cl = 1:0.04:0.45 (23).

The following drugs were added to the extracellular solution in subsets of experiments: 1 µM thapsigargin (Sigma), 1 mM EDTA (Sigma) in extracellular solution with 0 Ca²⁺ (MgCl₂ was increased to 2.8 mM in the perfusate to substitute the omitted CaCl₂), 500 µM CdCl₂ (Sigma), 5 µM nifedipine (Sigma), 0.1 µM ω-conotoxin GVIA (Sigma), 0.1 µM ω-agatoxin IVA (Sigma), and 2 mM dithiothreitol (Sigma). For the majority of protocols, cells were exposed at least 5 min to a particular drug before H₂O₂ perfusion. Longer thapsigargin (≥ 30 min) exposure was performed to achieve greater or near complete sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) block.

Electrophysiology

nTS cell dispersion.

Rats were anesthetized with isoflurane, and the brainstem was rapidly removed and placed in ice-cold low calcium PIPES-buffered solution (in mM; 115 NaCl, 5 KCl, 20 PIPES, 25 D-glucose, 0.1 CaCl₂, and 5 MgCl₂, pH 7.0 with NaOH, ~310 mosM) (27, 32). The brainstem was cut in ice-cold PIPES-buffered solution in the coronal plane into three ~360-μm sections spanning the caudal to medial nTS using a vibratome as above. The nTS sections were allowed to recover for 20 min at ~33°C in low-calcium PIPES buffer before enzyme incubation with 20 units of activated papain (as above) in PIPES for 30–35 min at ~33°C. Subsequently, slices were rinsed three times with enzyme-free PIPES buffer and maintained at room temperature (~22°C) in a continuously oxygenated jar with PIPES buffer. When cells were needed, a single brainstem slice was placed on a Sylgard covered petri dish and the nTS region was identified and removed using two pulled glass patch pipettes that served as microscalpels. The nTS was gently triturated in ~500 μl ice-cold PIPES containing 0.1% DNAase (ThermoFisher Scientific) and 0.5% bovine serum albumin (BSA; Sigma) using progressively smaller fire-polished glass pipettes. After trituration, ~250 μl each of the 500-μl neuron suspension were placed onto poly-d-lysine (100 μg/ml)-treated 15-mm coverslips that were seated in Warner Instruments RC-42LP incubation chambers (volume of 87 μl/mm). Neurons were allowed to settle and adhere to the coverslip for at least 20 min, after which superfusion of the cells began with a normal physiological solution (see below). All neurons were studied within 2 h of dispersion from a brainstem slice and within 8 h of removal from the rat.

Calcium current protocol.

Neurons were visualized on an Olympus IMT2 inverted microscope with phase optics. Recording electrodes were pulled from borosilicate glass (B150-86-7.5; Sutter Instruments, Novato, CA) via a horizontal pipette puller (Sutter Instruments P-97) and positioned using a hydraulic micromanipulator (MX6600R; Siskiyou). Signals were filtered at 2 kHz and acquired at 10 kHz using an AxoPatch 200B amplifier controlled by Clampex 10 software by a Digidata 1320 interface (Molecular Devices, Sunnyvale, CA).

Neurons were initially bathed in a normal physiological external solution consisting of the following (in mM): 137 NaCl, 3 KCl, 10 d-glucose, 10 HEPES, 2 CaCl₂, and 1 MgCl₂ (pH 7.4 with NaOH; ~315 mosM). To record calcium currents in whole cell mode, electrodes (3–6 MΩ) were filled with a solution containing the following (in mM): 124 CsCl, 11 EGTA, 1 CaCl₂, 10 HEPES, 1 MgCl₂, 4 MgATP, 0.2 NaGTP, 20 phosphocreatine, 0.02 leupeptin, and 35 U/ml phosphocreatine kinase (pH 7.2 with CsOH; ~300 mosM). ATP and the ATP-regenerating system in the pipette solution minimized “rundown” of currents (6). Following formation of a gigaseal and membrane rupture, the external solution was switched to one designed to enhance visualization of currents through VGCCs. Barium (Ba²⁺) replaced Ca²⁺ as the charge carrier in this modified external solution, which contained the following (in mM): 132 NaCl, 5 BaCl₂, 10 tetraethylammonium chloride (TEA-Cl), 2 MgCl₂, 10 HEPES, and 10 d-glucose (30, 45, 53). Tetrodotoxin (TTX; 1 μM) was added to block sodium currents.

To investigate total Ba²⁺ current through VGCCs, depolarizing voltage steps in 10-mV increments from −80 to +60 mV for 100 ms were applied to the neuron from a holding potential of −80 mV (30). This protocol generated a current-voltage (I–V) relationship. Alternatively, a single maximally (peak) evoked current (I_peak) was evoked from −80 mV every 30 s. The voltage applied to evoke peak currents was determined by analysis of the I–V curve 3 min after cell membrane rupture and was between −10 and −20 mV (see I–V curves in results). Evoked Ba²⁺ currents were subsequently classified as Ca²⁺ currents [($I_{{\text{Ca}}^{\text{2+}}}{}^{{\text{(Ba}}^{\text{2+}}\text{)}}$)]. H₂O₂ and/or calcium channel blockers were applied for 1 min. As in our imaging experiments, ω-conotoxin GIVA, ω-agatoxin IVA, nifedipine, or CdCl₂ was either diluted in extracellular Ba²⁺ solution alone or with H₂O₂. In a subset of neurons not exposed to H₂O₂, we examined the relative proportion of L, N, and P/Q channels. In these protocols, I_peak was recorded in the absence and then following subsequent addition of VGCC antagonists to the perfusate, followed by 100 μM CdCl₂ to block the remaining current. Solutions were applied to the cell at a rate of 150 µl/min using a superfusion system with its outlet ~500 μm upstream from the cell. This system allowed switching of solutions that the cell was exposed to within a few seconds.

In seven neurons, we examined cellular resting potential as an indicator of general cell viability. For these recording, the pipette solution contained the following (in mM): 10 NaCl, 130 K⁺ gluconate, 11 EGTA, 1 CaCl₂, 10 HEPES, 1 MgCl₂, 2 MgATP, and 0.2 NaGTP (pH of 7.3, 295–300 mosM), and the neurons were bathed in normal physiological external solution detailed above. Membrane potential was −53.3 ± 3.2 mV in these dissociated neurons, comparable to our previous reports (42).

Immunocytochemistry

A subset of culture dishes with nTS cells was fixed for immunocytochemical analysis. The glass coverslips were exposed for 20 min to 2% paraformaldehyde (Fisher Scientific, Pittsburgh, PA) at room temperature and then washed three times with 0.01 M PBS and stored at 2–8°C. Following blocking using 10% normal donkey serum (Millipore, Temecula, CA) in 0.3% Triton-PBS, coverslips were incubated with primary antibody against neuronal nuclear antigen (NeuN; neuronal marker; mouse; 1:500; MAB377; Millipore) and the antioxidant catalase (rabbit; 1:300; ab1877; Abcam, Cambridge, MA) in 0.3% Triton-PBS and 1% normal donkey serum for 24 h. Then, coverslips were rinsed in 0.01 M PBS and incubated with the secondary antibodies Cy5-conjugated donkey anti-mouse IgG (1:200; 715–175–151; Jackson Immunoresearch, West Grove, PA) and Cy2-conjugated donkey anti-rabbit IgG (1:200; 711–225–152; Jackson Immunoresearch) in 0.3% Triton-PBS for 1 h in the dark. After being rinsed with 0.01 M PBS, glass coverslips were dried and mounted with ProLong Gold with 4′,6-diamidino-2-phenylindole (DAPI; nuclear marker; Invitrogen, Carlsbad, CA) on gelatin-coated slides. Fluorescence was visualized using an epifluorescent microscope (BX51; Olympus), a digital monochrome camera (ORCA-ER; Hamamatsu), a spinning disk confocal unit (Olympus), and appropriate filter sets and excitation wavelengths. Each fluorophore was imaged in the same focal plane, pseudocolored, and postprocessed using ImageJ (National Institutes of Health) to adjust contrast and brightness.

Data Analysis

For imaging experiments, postprocessing of images and calculation of the 340-to 380-nm excitation ratio were done with ImageJ (National Institutes of Health) and Excel software (Microsoft). We analyzed 517 cells and excluded those with fibroblast-like morphology or visibly moving cilia that originate from the ependymal lining of the fourth ventricle. Based on our initial experiments, we defined the following criteria to clearly identify responders to H₂O₂: 1) the Ca²⁺ response must have a clear peak with a rising and decaying phase; 2) the peak must occur within our 10-min window from the end of the stimulation; 3) no other peaks of similar or greater height should be observed before drug perfusion to separate potential random Ca²⁺ fluctuations from H₂O₂-induced responses; and 4) the peak amplitude must be greater than a threshold of 6.2% from baseline. The threshold was defined by the average maximal change (4.1 ± 0.01%) during vehicle perfusion (as time control) and the added 95% confidence interval. The baseline to measure peaks was taken directly before the response to minimize errors due to slowly rising or decaying levels of Ca²⁺ over the entire course of the protocol. The onset variance of Ca²⁺ responses between high K⁺ and H₂O₂ was calculated using Excel software. The fura-2 340/380 ratio was taken as an indicator of internal Ca²⁺ concentration ([Ca²⁺]_i) (39, 51).

Calcium current data were analyzed using Clampfit (Molecular Devices) and Excel software (Microsoft). Across all dissociated nTS cells tested, cell capacitance was 7.77 ± 0.35 pF (n = 64), comparable to other isolated studies (4, 24, 44, 52). Acquired calcium currents were first normalized to the capacitance of the cell (pA/pF). Current-voltage (I–V) graphs were then expressed relative to the peak current of the control exposure as described by Strege et al. (49). Specifically, the peak inward current of the control trace was normalized to 1 using the equation I_norm = −1 · (I–V/I_peak), where I_norm is normalized current, I–V is measured current at a given voltage, and I_peak is maximum peak inward current of the control trace. Thus peaks of all other exposures (H₂O₂, channel antagonist, or wash) per cell were expressed as a fraction of 1. Peak currents evoked from a single voltage step were also normalized to the initial control exposure.

Statistical analyses were performed using SigmaPlot 13.0 (Systat Software) and GraphPad Prism 7 (GraphPad Software). Results were tested via t-test and one- or two-way repeated-measures ANOVA followed by Student-Newman-Keuls post hoc test unless otherwise stated in the text. Results are considered significantly different at P ≤ 0.05. Group data are presented as means ± SE.

RESULTS

To study the changes of [Ca²⁺]_i in response to H₂O₂, we loaded primary dissociated nTS cells with the fluorescent dye fura-2 (Fig. 1A). Our cultures primarily consisted of neurons as identified using NeuN in paraformaldehyde-fixed cultures (Fig. 1B). These neurons colabeled to a variable degree with catalase, verifying antioxidant capabilities also found in vivo (42).

Fig. 1.

Primary dissociated cell culture and fura-2 imaging of nucleus tractus solitarii (nTS) neurons. A: example of nTS cells visualized using differential interference contrast (DIC; left) also show strong fura-2 fluorescence when excited with 380 nm (right). Scale bar = 20 µm. B: immunocytochemical verification on paraformaldehyde-fixed nTS cells that our cultures primarily consisted of neurons (NeuN, magenta) that colabeled with nuclei (DAPI, gray) and the antioxidant catalase (green). Scale bar = 20 µm. Inset: zoomed image of one nTS neuron.

H₂O₂ Causes a Delayed Increase of Cytosolic Ca²⁺

To evaluate the changes in [Ca²⁺]_i, fura-2-loaded nTS neurons were exposed to an increased K⁺ concentration of the perfusate solution (55 mM) to induce cellular depolarization. Representative traces show a prompt increase in [Ca²⁺]_i in response to 1 min of high K⁺, with a large amplitude and decay that varied among cells (Fig. 2A). In contrast, variably long latencies of intracellular Ca²⁺ peaks were observed when the cells were perfused with 200 µM H₂O₂ (Fig. 2B, dashed lines). The mean time to response peak was significantly shorter following high K⁺ (0.36 ± 0.11 min, n = 22) when compared with H₂O₂ (4.72 ± 0.37 min, n = 31; P < 0.001, t-test). The variance in peak response time is depicted in the response time histogram in Fig. 2C (bin = 30 s). Quantitatively, the variance in peak response time was significantly less in response to high K⁺ (0.26 ± 0.21, n = 22 vs. H₂O₂, 4.16 ± 1.01, n = 31; P < 0.01, t-test). Given this high variability and longer response time to H₂O₂ perfusion, for the remainder of the fura-2 study we used a 10-min window following the stimulus to allow detection of Ca²⁺ peaks in response to H₂O₂. However, the majority of responses occurred at 5 min following the stimulus.

Fig. 2.

Effects of H₂O₂ on cytosolic Ca²⁺ in nTS cells. A: typical raw traces of cytosolic Ca²⁺ levels (using fura-2) to 1 min of high K⁺ (55.4 mM). B: examples of cytosolic Ca²⁺ levels to 1 min H₂O₂ (200 µM). Note the variable timing of peak responses (dashed lines). C: peristimulus time histogram grouping response timings to high K⁺ (responders to 55.4 mM; n = 22) and H₂O₂ (responders to 200 µM; n = 31; 30-s bins are shown. D: change from baseline for vehicle (n = 21), H₂O₂ (n = 31), and high K⁺ (n = 22). *P ≤ 0.05 vs. vehicle by one-way ANOVA on ranks. E: percent responders to different H₂O₂ concentration. Total number of cells tested: vehicle = 21, 100 µM = 22, 200 µM = 58, 300 µM = 28, and 500 µM = 15. *P ≤ 0.05 vs. vehicle by one-way ANOVA.

Average changes from baseline are depicted in Fig. 2D. [Ca²⁺]_i did not appreciably change during vehicle perfusion. Of those neurons that responded to H₂O₂ with more than a 6.2% increase from baseline (see materials and methods for details), [Ca²⁺]_i augmented by 27%. In comparison, high K⁺ elevated Ca²⁺ by 114% from baseline indicating that H₂O₂ elicits only moderate responses in nTS neurons. The amount of increase between the various H₂O₂ concentrations was similar (~27% for 100–500 µM, data not shown), but the number of responders per concentration significantly differed (Fig. 2E). Most cells responded to 200 µM H₂O₂, which was used throughout the remainder of the study.

H₂O₂ Increases Baseline Ca²⁺

To determine whether H₂O₂ increases [Ca²⁺]_i during cellular depolarization, we measured the H₂O₂-dependent increase in Ca²⁺ during and following repetitive high K⁺ stimulation (1-min episodes). Repeated stimulation with high K⁺ in the absence of H₂O₂ (i.e., vehicle treatment) elicited a reliable and repeatable depolarization, which recovered between stimuli as seen by the Ca²⁺ response (Fig. 3A, left). Following H₂O₂, [Ca²⁺]_i rose, yet this was seen primarily in the baseline response (Fig. 3A, right). Ca²⁺ responses to high K⁺ were additive to the increased baseline Ca²⁺ level due to H₂O₂. These data suggest that H₂O₂ used in this study did not elicit maximal Ca²⁺ responses in nTS cells and constitutes a rather moderate stimulus.

Fig. 3.

Ca²⁺ responses to H₂O₂ during membrane depolarization using high K⁺. A: example traces for intracellular Ca²⁺ responses to high K⁺ (5 stimulations) in the absence (left) and presence of 5 min H₂O₂ (right). Uniform (red lines; aligned to lowest Ca²⁺ level after the first K⁺ response) or individual (green lines on the right; 5.5-min window aligned to beginning of rise) baselines were used to access the area under the curve (AUC) for quantification of cytosolic Ca²⁺. B and C: averaged data depicting total content of Ca²⁺ using AUC for vehicle (n = 14) and H₂O₂ (n = 21) to a uniform baseline (B) or to individual baselines (C). *P ≤ 0.05 vs. vehicle; #P ≤ 0.05 vs. own baseline by two-way repeated-measures (RM)-ANOVA.

We quantified the changes in Ca²⁺, and the influence of 200 µM H₂O₂, by measuring the area under the curve (AUC) of the K⁺ response. This was accomplished using a 5.5-min analysis window aligned to the beginning of the Ca²⁺ rise for each K⁺-induced response. The length of the analysis window ensured a return of [Ca²⁺]_i to baseline values and excluded any responses to subsequent perfusions with high K⁺. We initially examined the Ca²⁺ response from a single uniform baseline (Fig. 3B; red lines in Fig. 3A), defined as the lowest Ca²⁺ level following the first K⁺ stimulation. Such a baseline accommodated the diverse kinetics of Ca²⁺ removal (i.e., the time for [Ca²⁺]_i to return to baseline levels) and evaluated the overall increase in Ca²⁺ from the initial baseline. In contrast to vehicle, H₂O₂ significantly increased the AUC, and this effect progressively continued into our washout periods (Fig. 3B).

We also measured the K⁺ response (i.e., the AUC) and the influence of H₂O₂ from individual baselines for each cell depolarization (green lines in Fig. 3A, right). Individual baseline analysis provided access to any H₂O₂-provoked changes in depolarization-induced Ca²⁺ currents while excluding changes of baseline Ca²⁺ level. The averaged group data show that responses to high K⁺ were similar in the presence and absence of H₂O₂, as well as in vehicle (Fig. 3C). These data indicate that H₂O₂ does not increase K⁺-evoked intracellular Ca²⁺ but rather provides an elevated baseline that is additive to depolarizations by high K⁺.

H₂O₂ Increases [Ca²⁺]_i Through Extracellular Sources

Cytosolic Ca²⁺ may increase due to its release from major intracellular stores such as the endoplasmic reticulum, or it may enter from the extracellular fluid. To identify the Ca²⁺ source we blocked several potential pathways and examined fura-2 fluorescence before and following H₂O₂. We initially used thapsigargin (TG; 1 µM), a potent inhibitor of the endoplasmic reticulum Ca²⁺-ATPase, to explore the role of this internal store. With TG in the perfusate, the typical delayed Ca²⁺ increase to H₂O₂ was still present (Fig. 4A). The latency of mean response peaks following H₂O₂ during TG (5.05 ± 0.47 min, n = 26) and its variance (5.41 ± 1.37, n = 26) were comparable to responses to H₂O₂ in the absence of TG. Furthermore, also the percentage of H₂O₂ responders during TG was similar to that responding to H₂O₂ without TG (Fig. 4D). These data indicate that the endoplasmic reticulum is not responsible for the H₂O₂-induced Ca²⁺ rise.

Fig. 4.

H₂O₂ increases intracellular Ca²⁺ concentration ([Ca²⁺]_i) via extracellular sources. A: raw traces of intracellular Ca²⁺ to H₂O₂ in the presence of thapsigargin (1 µM; TG), an inhibitor of the endoplasmic reticulum Ca²⁺-ATPase. All examples shown stem from the same experiment with simultaneous H₂O₂ exposure. Note the high variance in peak response timing to H₂O₂. B: traces showing the typical lack of response to H₂O₂ in a zero Ca²⁺ perfusate with added EDTA (1 mM). C: Ca²⁺ levels to H₂O₂ in the presence of the voltage-gated calcium channel (VGCC) blocker CdCl₂ (500 µM). Note the lack of H₂O₂-typical Ca²⁺ responses. Inset: examples of CdCl₂-dependent rise of Ca²⁺ in the absence of H₂O₂. D: percentage of responders to H₂O₂ in the presence of TG (n = 51), zero Ca²⁺/EDTA (n = 21), and CdCl₂ (n = 17), compared with vehicle and plain H₂O₂ (displayed in light gray since the same data is shown in Fig. 2E). *P ≤ 0.05 vs. vehicle; #P ≤ 0.05 vs. H₂O₂ by one-way ANOVA.

To further explore whether Ca²⁺ enters nTS cells from the extracellular space, we removed extracellular Ca²⁺. Removing extracellular Ca²⁺ using a zero Ca²⁺ perfusate with the Ca²⁺ chelator EDTA (1 mM) reduced cellular fura-2 fluorescence and completely ablated responses to H₂O₂ (Fig. 4B). The percentage of responders strongly declined and was significantly reduced compared with H₂O₂ alone (Fig. 4D). Thus H₂O₂ triggers Ca²⁺ entry in nTS cells across the cell membrane.

To identify the site of entry, we used the general blocker CdCl₂ (500 µM) to block VGCCs. In control experiments where nTS neurons were exposed to 500 µM CdCl₂ alone, an initial decrease was followed by a steady rise in fluorescence (Fig. 4C, inset). This effect was concentration independent since it also occurred using a fivefold lower concentration of CdCl₂ (100 µM, data not shown). Apart from this CdCl₂-induced rise in fluorescence, H₂O₂ did not further increase [Ca²⁺]_i greater than threshold (Fig. 4C), as opposed to the previously observed responses shown in Figs. 2B and 4A. With CdCl₂ the number of responders to H₂O₂ significantly decreased (Fig. 4D).

Protein Oxidation by H₂O₂ Contributes to Ca²⁺ Entry

H₂O₂ is known to exert its actions via oxidizing exposed sulfhydryl groups in other tissues (20). To identify whether the same is true for nTS neurons, we used the reducing agent dithiothreitol (DTT; 2 mM) to retain sulfhydryl groups in their reduced state. Under DTT perfusion, the percentage of neurons responding with elevated [Ca²⁺]_i to H₂O₂ greatly decreased (16.8 ± 7.0%, P ≤ 0.05, n = 39, one-way ANOVA) to a value similar to vehicle perfusion and significantly different from H₂O₂ alone. These data suggest that the H₂O₂-induced increase in [Ca²⁺]_i involves oxidative modulation of sulfhydryl groups.

H₂O₂ Causes a Delayed Enhancement of Calcium Currents

We next examined the contribution of H₂O₂ to modulate inward Ca²⁺ current ($I_{{\text{Ca}}^{\text{2+}}}$) via VGCC. nTS neurons were isolated and depolarizing voltage steps were applied for 100 ms from −80 to +60 mV to evoke inward currents via VGCC (Fig. 5, A and B). Ba²⁺ served as the charge carrier and the currents evoked were classified as Ca²⁺ currents [$I_{{\text{Ca}}^{\text{2+}}}{}^{{\text{(Ba}}^{\text{2+}}\text{)}}$]. Step depolarization from −80 mV evoked $I_{{\text{Ca}}^{\text{2+}}}$ that was activated at −40 mV with maximum currents observed at approximately −10 mV (Fig. 5B). Focal H₂O₂ did not have an immediate effect on $I_{{\text{Ca}}^{\text{2+}}}$ (Fig. 5D, n = 12). However, and similar to our fura-2 experiments, removal (washout) of H₂O₂ significantly elevated $I_{{\text{Ca}}^{\text{2+}}}$ (Fig. 5, C and D, inset). The small shift in reversal potential upon H₂O₂ washout was not significant from control (P = 0.376, paired t-test). Peak currents (I_peak) increased with H₂O₂ wash by 20.7 ± 13.4% (P = 0.03 vs. control) and, following an additional 3-min washout period, returned to near baseline (3.1 ± 11.2%, P > 0.05 vs. control, ANOVA with Dunns multiple comparison, n = 11). Conversely, application of vehicle solution (instead of H₂O₂) did not increase $I_{{\text{Ca}}^{\text{2+}}}$ during its application or washout periods (Fig. 5E, n = 8). Also, peak currents during vehicle wash did not significantly change over time. Together, these data confirm that the increase in $I_{{\text{Ca}}^{\text{2+}}}$ was due to H₂O₂ exposure rather than “run-up” of the current.

Fig. 5.

H₂O₂ enhances calcium currents upon its removal. A: photomicrograph of nTS neuron with attached patch electrode. Scale bar = 20 μm. B and C: inward calcium currents ($I_{{\text{Ca}}^{\text{2+}}}$) under control and following H₂O₂ removal in a single neuron. Note the increase in currents in the wash. Currents shown were evoked between −80 and +10 mV. C, inset: current-voltage (I-V) relationship of raw traces shown in B and C. Note the increase in currents from control (blue) with H₂O₂ wash (green). D: current voltage (I–V) plots for $I_{{\text{Ca}}^{\text{2+}}}$ during control, H₂O₂, and H₂O₂ removal (i.e., wash). Currents are presented relative (Rel.) to peak control currents. Note the increase in currents in the wash period between −20 and +20 mV (green line); n = 12. E: I–V plots for vehicle (time) control experiments. Note the lack of increase in currents across the various time points; n = 8. F: relative proportion of L-, N-, and P/Q- subtype VGCC channels in nTS neurons; n = 7. Res, residual current that is blocked by Cd²⁺. *P ≤ 0.05, Veh/H₂O₂ vs. control; #P ≤ 0.05, Veh/H₂O₂ vs. wash; †P ≤ 0.05, wash vs. control by two-way RM-ANOVA.

In a separate set of neurons (n = 7), we sequentially added VGCC subtype blockers to identify their contribution to I_peak (see materials and methods for details). These experiments showed that the total calcium currents in our dissociated nTS neurons were comprised of several high-voltage VGCCs, L, N, and P/Q, as well as a residual current that was blocked by 100 μM CdCl₂ (Fig. 5F).

H₂O₂ Triggers Ca²⁺ Entry via Modulation of VGCCs

The ablation of H₂O₂-induced [Ca²⁺]_i increases with removal of external Ca²⁺ or application of CdCl₂ suggests the involvement of one or more membrane bound calcium channels. To elucidate the subtype of VGCCs involved, we repeated our fura-2 imaging with H₂O₂ stimulation in the presence of multiple antagonists to specifically block L-type (5 µM nifedipine), N-type (0.1 µM ω-conotoxin GVIA), and P/Q-type (0.1 µM ω-agatoxin IVA) VGCCs. Raw traces in Fig. 6A exemplify the absence of response to H₂O₂ after VGCC block (here: P/Q subtype). The group data show that individually blocking VGCC subtypes significantly reduced the number of responders to H₂O₂ (Fig. 6B).

Fig. 6.

Cytosolic Ca²⁺ rise by H₂O₂ via VGCCs. A: exemplary raw traces for H₂O₂ during P/Q-type VGCC blocker [0.1 µM ω-agatoxin (ω-aga) IVA]. Note the absent response to H₂O₂. B: averaged responders to H₂O₂ during different VGCC subtype-specific blockers: L type (5 µM nifedipine; n = 26), N type (0.1 µM ω-conotoxin GVIA; n = 27), and P/Q type (n = 38). Light gray bars are from data shown in Fig. 2E. #P ≤ 0.05 vs. H₂O₂ by one-way ANOVA.

We next examined the contribution of VGCCs to H₂O₂-induced changes in inward $I_{{\text{Ca}}^{\text{2+}}}$. As in our fura-2 imaging experiments, after characterization of initial $I_{{\text{Ca}}^{\text{2+}}}$ properties, L, N, or P/Q channels were blocked (1 min, control) followed by H₂O₂ (1 min) in VGCC antagonists. H₂O₂ was then removed and currents were again examined after 3–4 min in VGCC blocker (i.e., wash). As quantified, in the presence of nifedipine (Fig. 7A, n = 9), ω-conontoxin (Fig. 7B, n = 9), or ω-agatoxin (Fig. 7C, n = 9), the H₂O₂-induced increase in $I_{{\text{Ca}}^{\text{2+}}}$ was eliminated. Similar to our fura-2 imaging data, the majority of nTS neurons responded to H₂O₂ wash with an increase in $I_{{\text{Ca}}^{\text{2+}}}$ (9/12, Fig. 7D). Conversely, in the presence of an individual blocker only a small number of neurons increase peak $I_{{\text{Ca}}^{\text{2+}}}$ to H₂O₂ (nifedipine 4/9; ω-conontoxin 1/9; and ω-agatoxin 1/9). To examine the impact of residual calcium channels (presumable R and/or T type) in these responses, we simultaneously blocked L, N, and P/Q channels with a cocktail of nifedipine, ω-conontoxin, and ω-agatoxin and examined the influence of H₂O₂ on $I_{{\text{Ca}}^{\text{2+}}}$. H₂O₂ did not increase I_peak in the presence of all three blockers (−10.1 ± 2.6 pA/pF vs. in H₂O₂, −7.7 ± 2.3 pA/pF, P = 0.19, paired t-test, n = 4). H₂O₂ also did not augment $I_{{\text{Ca}}^{\text{2+}}}$ during L, N, and P/Q blockade with 100 μM CdCl₂ (I_peak, VGCC block, 1.2 ± 0.4 pA/pF vs. H₂O₂ in VGCC block, 1.3 ± 0.6 pA/pF, P = 0.53, paired t-test, n = 6). Taken together, these data suggest H₂O₂ induces calcium entry in nTS neurons through modulation of multiple VGCCs.

Fig. 7.

VGCC block attenuated H₂O₂ increases in $I_{{\text{Ca}}^{\text{2+}}}$. Relative current-voltage (I–V) plots demonstrating that block of L (A)-, N (B)-, or P/Q (C)-type calcium channels attenuates H₂O₂-induced increases in $I_{{\text{Ca}}^{\text{2+}}}$. H₂O₂ was applied in the presence of VGCC block. Data are plotted relative to control (blocker alone) for each cell; n = 9 each. D: number of cells that increase peak currents within the I–V plot during H₂O₂ and H₂O₂ in the presence of L-, N-, and P/Q-type blockers. Nif, nifedipine; Ctx, conotoxin; Agtx, agatoxin. *P ≤ 0.05, H₂O₂ vs. control; # P ≤ 0.05, H₂O₂ vs. wash; †P ≤ 0.05, wash vs. control by two-way RM ANOVA.

DISCUSSION

In the present study, we show that short exposure of nTS neurons to H₂O₂ elicits a moderate and reversible increase in cytosolic Ca²⁺ ([Ca²⁺]_i). This rise in [Ca²⁺]_i affects baseline levels independent of cell depolarization. The response to H₂O₂ occurs within 10 min following stimulation and its timing is highly variable among cells. Eliminating the contribution of endoplasmic reticulum stores to [Ca²⁺]_i changes did not alter H₂O₂-induced Ca²⁺ entry. On the other hand, removing extracellular Ca²⁺ or blocking general Ca²⁺ channels with CdCl₂ prevented responses to H₂O₂. This increase on [Ca²⁺]_i is most likely dependent on sulfhydryl oxidation. Elevation of [Ca²⁺]_i occurs via multiple subtypes of VGCCs as shown by fura-2 imaging and patch-clamp electrophysiology.

The nTS is the first central synapse for integration and processing of several viscerosensory inputs before its relay to other central nuclei. This neurophysiological profile is reflected in the heterogeneous nature of its neuronal population due to well-known input and output connections, as well as its numerous interneurons. Our cultures likely represent this heterogeneity. While we do not know whether the neurons we recorded or imaged receive sensory information from baro-, chemo-, or other inputs, nor do we know their projections, it is likely that each neuron has its own distinct response to ROS and ionic channel composition and density. In addition, while our fura-2 imaging represents spontaneous influx of Ca²⁺ (under physiological conditions) due to ROS, our studies under patch-clamp depolarization-induced Ca²⁺ currents ($I_{{\text{Ca}}^{\text{2+}}}$) represent the maximal changes in Ca²⁺ entry and allow the biophysical cell properties to be evaluated. While they cannot be directly compared due to these distinct differences, each may lend insight into the other on the modulatory role of H₂O₂.

We have previously shown in nTS slices that micromolar concentrations of H₂O₂ induce delayed hyperexcitability in postsynaptic nTS neurons (42). Here we examined the influence of H₂O₂ on [Ca²⁺]_i as a potential contributor to these excitatory responses. This was tested in dissociated nTS neurons to eliminate the influence of synaptic activity and to examine the direct impact of H₂O₂ on cellular Ca²⁺. H₂O₂ increased [Ca²⁺]_i with the greatest percentage of responders occurring at 200 µM H₂O₂. Higher concentration of H₂O₂ did not increase the number of responders or the magnitude of the response, which could possibly be attributed to cytotoxicity and/or inactivation of the Ca²⁺ channels. The augmentation of [Ca²⁺]_i was relatively modest (27% across all concentrations using Ca²⁺ imaging) and readily reversed. A similar ~21% increase to 200 µM H₂O₂ was observed with the Ca²⁺ currents of nTS cells in our patch-clamp analysis and in [Ca²⁺]_i of rat hippocampal cells (16). The H₂O₂ concentrations examined encompass those expected during endogenous ROS production (i.e., 15–150 µM H₂O₂) considering the predicted 10- to 100-fold decrease with cell entry (47). Given the moderate and reversible response, our data suggest that H₂O₂ did not produce permanent oxidative damage but rather modulated Ca²⁺ via one or more mechanisms.

In contrast to the rapid, time-invariant, and robust increase in [Ca²⁺]_i to K⁺ depolarization, H₂O₂ produced effects that varied greatly in their onset among cells. This variability is consistent with our previous study and others (16, 20, 26, 42) that demonstrated H₂O₂ induces functional cellular responses of up to 6 min following its application. Several possibilities may contribute to the delayed response in nTS neurons. For instance, we have shown that H₂O₂ initially hyperpolarizes the membrane potential of nTS cells by increasing barium-sensitive potassium currents (42). This effect is reversible with removal of H₂O₂ and may thus contribute to the longer response latency of [Ca²⁺]_i, which may differ among cells. Another possibility is that that concentration or location of endogenous antioxidants protecting ROS-sensitive targets may vary among cells. We have shown that the antioxidant catalase is widely distributed in nTS neurons in a dense perinuclear network (42); a similar expression was seen in dissociated nTS neurons. Responses may then be delayed when target sites of H₂O₂ are particularly close to ROS-scavenging antioxidants. Another possibility for increased variability is the site of modulation for the Ca²⁺ increase. If the effect of exogenous H₂O₂ requires passage through the cellular membrane to reach intracellular targets, it is likely to require a longer onset for responses. We have previously shown that H₂O₂ exerts its effects via modulation of intracellular sites (42).

Several potential sites by which H₂O₂ affects Ca²⁺ were examined. The endoplasmic reticulum is a major intracellular Ca²⁺ store in neurons and may contribute to [Ca²⁺]_i rises upon ROS exposure. Thapsigargin is the prototypical inhibitor of SERCA pumps that control release and reuptake of Ca²⁺ in the endoplasmic reticulum (37). When oxidized, SERCA pumps are inhibited and augment [Ca²⁺]_i (11). In the present study, thapsigargin did not prevent the H₂O₂-induced augmentation of Ca²⁺, suggesting that intracellular endoplasmic reticulum stores did not substantially contribute to its elevation. However, depletion of intracellular stores by thapsigargin may also induce extracellular calcium entry via activation of store-operated (capacitive) channels, and H₂O₂ may contribute to this effect (19). In the absence of extracellular Ca²⁺, H₂O₂-induced responses were eliminated and similar to vehicle. Exogenous cadmium, a nonspecific VGCC blocker, also effectively eliminated H₂O₂-induced Ca²⁺ rises, suggesting the involvement of VGCCs in the Ca²⁺ influx across the membrane. Unexpectedly, CdCl₂ alone induced a steady rise in fura-2 fluorescence independent from our H₂O₂ stimulation. This increase may be due to Cd²⁺ augmenting [Ca²⁺]_i by block of membrane-bound Ca²⁺-ATPase or Na⁺-Ca²⁺ exchange (28). Cadmium is also able to increase ROS by depleting intracellular antioxidants (15) to elevate intracellular Ca²⁺, as shown in other cell cultures (57). Alternatively, Cd²⁺ is able to directly bind to fura-2 and increase the fluorescent signal (22, 41), which may contribute in the present study. Taken together, these data indicate H₂O₂ enhances [Ca²⁺]_i via influx across the plasma membrane and not intracellular stores.

nTS neurons express the high-voltage-activated L, N, and P/Q VGCCs (2, 14). Direct exogenous application of H₂O₂ to isolated nTS neurons augments L-type Ca²⁺ currents (54). Likewise, NADPH oxidase-derived ROS potentiate L-type Ca²⁺ currents in nTS neurons during angiotensin II signaling (55). Our electrophysiological recordings further support the idea that H₂O₂ enhances $I_{{\text{Ca}}^{\text{2+}}}$. Interestingly, block of a single VGCC attenuated H₂O₂-induced responders as well and the increase in $I_{{\text{Ca}}^{\text{2+}}}$. This was not anticipated as while one VGCC was blocked (e.g., L-type), the other channels should remain functional and thus continue to provide a H₂O₂ response. While we do not know the exact mechanism by which this did not occur, several possibilities may exist. For instance, the antagonists used may not specifically block the subtype desired. This possibility is unlikely as we used concentrations similar to other studies (13, 18, 30, 35), the neurons were only exposed to a given antagonist for a short period, and each blocker when given sequentially provided a proportional block (Fig 5F). Another possibility is that each neuron likely expresses a unique proportion of VGCCs on its membrane, with its threshold for activation determined by the membrane potential and intracellular signals. Thus, in the neurons examined, the influence of a particular VGCC blocker on H₂O₂ responses may not be equally distributed across channels or neurons. The modest 27% increase in fura-2 fluorescence and 21% increase in $I_{{\text{Ca}}^{\text{2+}}}$ by H₂O₂ is likely spread across multiple subtypes, which is dependent on their expression pattern on individual neurons. In addition, the H₂O₂ increase in fluorescence does not occur in every neuron at the concentrations we tested; this may be due, in part, to our conservative threshold in our fura-2 analysis. H₂O₂ may induce a response less than this threshold after an individual VGCC blocker. Nevertheless, although not significant, the continued number of responders (fura-2, Fig 6) and increases in peak current (Fig 7D) in the presence of VGCC blockers suggest that other channels are responding to H₂O₂. Additional future studies are required to completely delineate these possible mechanisms.

While H₂O₂ augmented basal [Ca²⁺]_i, and our studies show a prominent role of VGCCs in this response, there was only a small, nonsignificant increase in the [Ca²⁺]_i (fura-2) response during concurrent K⁺-induced depolarization. The H₂O₂-induced augmentation in [Ca²⁺]_i rather seemed independent and additive to K⁺-induced increases in [Ca²⁺]_i. As noted above, depolarization with high K⁺ in the absence of H₂O₂ induced a robust increase in [Ca²⁺]_i. The kinetics and time course of these increases were comparable to previous studies in isolated nTS neurons (21, 46), although with the observed time course fura-2 acting as a Ca²⁺ buffer cannot be completely ruled out (5). Although we did not isolate the influence of each Ca²⁺ channel subtype in our imaging studies, our results suggest minimal ROS effects during more pronounced K⁺-induced Ca²⁺ entry. This may be due to reduced Ca²⁺ drive during K⁺-induced elevation of [Ca²⁺]_i, which would subsequently minimize the influx produced by H₂O₂. Alternatively, H₂O₂ may differentially affect VGCC subtype currents, perhaps having opposite and offsetting effects. However, given that all VGCC antagonists tested in the present study reduced ROS effects on [Ca²⁺]_i and $I_{{\text{Ca}}^{\text{2+}}}$ of VGCC subtypes, this mechanism is less likely. Additional membrane sources or ion channels activated by depolarization may also temper H₂O₂-mediated Ca²⁺ responses. For instance, activation of voltage-gated K⁺ channels may limit membrane depolarization and thus may not further increase Ca²⁺ influx. Washout of K⁺ and the resulting repolarization may permit prolonged VGCC activation as the membrane voltage may be within the range for its activation and thus induce the observed increases in baseline Ca²⁺. This may especially be prevalent in intact cells. The lack of changes in $I_{{\text{Ca}}^{\text{2+}}}$ immediately during H₂O₂ in our patch-clamp studies is consistent with the fura-2 data suggesting minimal initial influence of ROS but rather a prolonged effect that is observed during the wash period.

Exposed cysteine residues (sulfhydryl groups) are primary targets for oxidation by ROS (10), and Ca²⁺ currents are directly altered when associated with the channel itself (9). We demonstrated the necessity of sulfhydryl oxidation in the augmentation of [Ca²⁺]_i after H₂O₂ in nTS cells. Specifically, in the presence of the reducing-agent DTT, H₂O₂ did not augment [Ca²⁺]_i. Our results are consistent with H₂O₂ augmenting Ca²⁺ influx via sulfhydryl oxidation of P/Q channels resulting in facilitation of channel opening (34). Also, oxidative modulation by H₂O₂ increases cardiac L-type current that is effectively blocked in the presence of DTT (50). Ca²⁺ channel activity can also increase due to direct glutathionylation by oxidized glutathione (25), an antioxidant for H₂O₂. Glutathione, in turn, is oxidized at its exposed sulfhydryl groups (36).

While our data suggest that H₂O₂ alters Ca²⁺ entry via oxidation and modulation of membrane-bound VGCC channels, other possibilities may exist. H₂O₂ may inhibit Ca²⁺-ATPase and Na⁺-Ca²⁺ exchanger pumps on the plasma membrane to elevate [Ca²⁺]_i (29, 58). However, our electrophysiological data clearly show that H₂O₂ enhances $I_{{\text{Ca}}^{\text{2+}}}$ across the membrane via VGCCs, although both mechanisms are not necessarily exclusive. Regardless of the mechanism, our studies demonstrate that H₂O₂ enhances [Ca²⁺]_i through modulation of one or more ion channel functions. Ongoing experiments may help to elucidate these possibilities in the near future, as well as to identify the phenotype of nTS neurons and the extent to which they are sensitive to H₂O₂.

In summary, we have shown that physiological concentrations of H₂O₂ activate VGCCs via oxidation of sulfhydryl groups resulting in raised [Ca²⁺]_i and $I_{{\text{Ca}}^{\text{2+}}}$ in the majority of nTS cells. This increase in Ca²⁺ occurs in a delayed fashion and likely explains the delayed hyperexcitability of nTS neurons in response to H₂O₂ (42). The H₂O₂-induced increased Ca²⁺ in nTS neurons likely influences physiological responses and may constitute the underlying mechanisms for the bradycardia and hypotension observed with H₂O₂ microinjections in the nTS (7). Since ROS are readily produced under normal physiological conditions within the nTS (54), these data widen our understanding of the modulatory actions H₂O₂ has in this critical cardiorespiratory brainstem region. It may also be part of the central processes involved in diseases associated with excess ROS along the cardiorespiratory neuroaxis.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants R01-HL-098602 (to D. D. Kline) and HL-128454 (to D. D. Kline) and American Heart Association Grant 12POST11670002 (to T. D. Ostrowski).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

T.D.O., L.P.-P., and D.D.K. conceived and designed the research; T.D.O., H.A.D., L.P.-P., and D.D.K. performed experiments; T.D.O., H.A.D., L.P.-P., and D.D.K. analyzed data; T.D.O., L.P.-P., and D.D.K. interpreted results of experiments; T.D.O. and D.D.K. prepared figures; T.D.O. and D.D.K. drafted manuscript; T.D.O., L.P.-P., and D.D.K. edited and revised manuscript; T.D.O., H.A.D., L.P.-P., and D.D.K. approved final version of manuscript.