Juvenile Helix aspersa land snails were provided by a local breeder. The animals were kept in ventilated plastic boxes, at a room temperature (RT) between 20 and 25 °C, and under a 12-h light-dark cycle. They were fed lettuce and chicken feed supplemented with calcium.

The experiments were conducted under the approval of the Institutional Animal Care and Use Committee of the University of Costa Rica (authorization CICUA-026-16).

A L15 solution modified for snails was used for the maintenance of the isolated C1 neurons and as the extracellular solution for the membrane potential recordings and contained (in mM): 6.974 CaCl₂, 4.922 MgCl₂, 0.348 MgSO₄, 5.335 KCl, 0.189 KH₂PO₄, 66.520 NaCl, 0.574 Na₂HPO₄. The physiological intracellular solution used for the membrane potential recordings contained (in mM): 3 NaCl, 100 KCl, 1 MgCl₂, 5 EGTA, 10 HEPES, pH 7.4 (with KOH).

The intracellular and extracellular recording solutions used to obtain calcium macrocurrents were designed to eliminate sodium and potassium currents. The extracellular recording solution was free of sodium and potassium, low in calcium, and contained (in mM): 105 tetraethylammonium (TEA), 2 CaCl₂, 10 MgCl₂, 10 HEPES, pH 7.4 (with CsOH). The intracellular solution was potassium-free and contained (in mM): 3 NaCl, 100 CsCl, 1 MgCl₂, 10 HEPES, 5 EGTA, pH 7.4 (with CsOH).

The following final blocker concentrations were used to identify the Ca_Vs: 10 nM ω-conotoxin CVIB (ω-CnTx CVIB), 1 µM cilnidipine (first diluted with DMSO and managed under dim light during experiments), 100 nM ω-conotoxin CVIE (ω-CnTx CVIE), and 4 µM ω-conotoxin GVIA (ω-CnTx GVIA). All the blockers were purchased from Alomone Labs (Jerusalem, Israel).

Cell cultures were performed as previously described by Ghirardi et al. (199) and modified by Brenes (201b). Briefly, the snails were anesthetized with a 0.1 M MgCl₂ isotonic solution injected into the foot and euthanized by evisceration through shell removal. Cerebral ganglia were surgically isolated and incubated at 34 °C for 4 hours in protease type XIV (0.4 U/mL, Sigma Aldrich, United States) diluted with modified L15 solution. After enzymatic digestion, the ganglia were washed twice in modified L15 solution. C1 neurons were individually isolated based on their anatomical localization and morphology. Neurons were cultured in different settings according to experimental conditions.

Neurons were transferred to dishes pretreated with 5% bovine serum albumin and filled with modified L15 solution at RT. After 24 h, the neurons acquired a soma-configuration by retracting their axons. Conventional patch clamp recordings were performed in the whole-cell configuration using a Multiclamp 700B amplifier (Molecular Devices, San Jose, CA, United States), digitally converted with the Digidata 1550B (Molecular Devices, San Jose, CA, United States), and recorded using the Clampex 11.1 software. Borosilicate pipettes with a resistance of 2–3 MΩ were pulled with a Flaming/Brown Micropipette Puller (Model P-1000, Sutter Instrument, Novato, CA, United States).

All recordings were performed at RT; only 1 cell was recorded per dish. The standard I-V protocol was performed by applying 200 ms long pulses at voltages ranging from −60 to +80 mV in 10 mV increments from a holding potential of −50 mV. Data were later analyzed using Clampfit 11.1.0.23 (Molecular Devices, San Jose, CA, United States). All currents were normalized to membrane capacitance to express them as current density.

Activation curves were constructed with conductance values calculated for each cell using Equation 1, where V_rev is the experimental reversal potential of each cell, V_m is the test membrane potential, and I is the maximal current in the analyzed potential. Later, the data were normalized to each cell’s maximal conductance (G_max). The data were adjusted following a modified Boltzmann function (Equation 2), where G_max is the maximum conductance, V_rev is the reversal potential, z_act is the apparent gating valence, V_1/2 is the voltage of half-maximal activation, F is the Faraday constant, R is the gas constant, and T the absolute temperature. G = I V m − V r e v (1) I   V = G m á x   V − V r e v 1 + exp − z a c t V − V 1 / 2 F / R T (2)

The steady-state inactivation was measured at a −10 mV and 200 ms test pulse after a pre-pulse holding potential ranging from −60 to +20 mV for 200 ms. First, the values were normalized to the maximal current of each cell. Then, the inactivation curves were constructed as a function of the pre-pulse voltage and fitted according to Equation 3, where V represents the prepulse potential, I r e s is the residual current (the fraction of non-inactivating current), V 1 / 2 is the voltage of half-maximal inactivation, and k is the slope factor of the curve. I V = I r e s + 1 − I r e s 1 + exp V − V 1 / 2 k (3)

A voltage ramp was employed to determine whether the C1 neurons express LVA and HVA Ca_Vs. To do this, the cells were held at −50 mV, followed by the application of a depolarizing voltage ramp at a rate of 0.2 mV/ms, ranging from −80 mV to +50 mV. The number of peaks and the corresponding voltages at which these peaks occurred were recorded.

After recording the I-V curve, the inactivation protocol, and the voltage ramp in the control state, one type of calcium channel blocker was added to the extracellular medium. Twenty minutes after introducing the blocker, all the protocols were repeated to assess the blockade’s effect on calcium macrocurrents. Due to the overlapping blocking capabilities of Ca_V2.2 by ω-CnTx CVIE and cilnidipine, cilnidipine was added at the end of the recordings with ω-CnTx CVIE, and 20 min later, the recordings were repeated to determine the presence of Ca_V1 channels.

Different variables were measured to evaluate the neuronal excitability of the C1 neuron. First, the mean firing frequency (MFF) was determined as the number of action potentials (APs) fired by 500 ms depolarizing stimulations of 0.5, 1.0, and 1.5 nA of intensity. To determine the cell rheobase, 50 ms pulses of increasing amperage were applied, starting at 0.5 nA and making increments of 0.01 nA until at least three consecutive independent APs were generated at a frequency of 1 Hz. To analyze the shape of the depolarization phase, the first AP generated by a depolarizing current of 1.0 nA intensity was analyzed. The following variables were determined: AP amplitude, time elapsed from the beginning to the peak of the AP (time to peak), rise time 10–90%, rise time 0–63% (rise tau), and maximum rise slope. To analyze the shape of the repolarization phase, six APs generated by stimulating each cell with its respective rheobase value were averaged. The following variables were determined: after-hyperpolarization amplitude, decay time 100-37% (decay tau), decay slope 90-10%, and maximum decay slope. Once the Em, MFF, and rheobase were recorded in the control state, ω-CnTx CVIB was added, and all the protocols were repeated 20 minutes later to evaluate the effect of the blockade on each of these variables and the shape and behavior of the APs.

Freshly dissociated single neurons with the initial segment of their axon were placed by well on a 96-well plate; each well was pre-coated with poly-L-lysine and Aplysia hemolymph and filled with 100 µL of modified L15 medium. Cells were grown for 72 h. The calcium imaging probe was prepared according to the manufacturer’s specifications; fluo-4 was mixed with the power load and diluted in L15 medium to a final concentration of 2 µM. 50µL of this mixture was added to each well to a final concentration of 1 µM. The plate was incubated for 30 min and measured.

Cells were analyzed using the Cytation 3 Hybrid Technology™ system, selecting the corresponding well. The magnification, LED intensity, integration time, and image gain were adjusted for each neuron and, once set, remained constant throughout the experiment. The excitation wavelength was 490 nm, and the emission wavelength was 520 nm.

For both control and treated cells, a baseline recording was performed for 1 min, followed by 1 min of vehicle recording after adding medium to rule out mechanical effects on Ca²⁺ influx. The control group was then exposed to KCl. The activity was induced by a sustained exposition to KCl at a final concentration of 65 mM, since this concentration has been reported to induce the activity of C1 neurons for more than 5 min (Massobrio et al., 2013). Meanwhile, the treated group was exposed to ω-CnTx GVIA (final concentration 4 µM) and incubated for 15 min before adding KCl (65 mM, as previously reported). The recording time during excitation was 5 min. Images were captured every 3–5 s for both groups.

To quantify calcium in varicosities, images were analyzed using CellProfiler Analyst™ (Broad Institute). First, all images were aligned to ensure proper overlap, templates were manually created to select recognizable varicosities, and fluorescence was measured throughout the recording in each structure. Results were presented as the change in relative fluorescence compared to baseline fluorescence (Equation 4), where F₀ was the average fluorescence during the baseline recording. Δ F F 0 = F − F 0 F 0 (4)

The Ca²⁺ distribution along neurites was analyzed by calculating the targeting factor of the Ca²⁺ signal in the last picture of KCl treatment regarding the Ca²⁺ signal in the first picture of the baseline condition, following the protocol described by Brenes (2015b). Using the line profile tool in ImageJ (NIH, Bethesda, MD) in both pictures, a longitudinal line was traced along the neurites to measure fluorescence peaks and adjacent regions (valleys) along the neurite. An additional line was drawn in neighboring areas to subtract background fluorescence. Once background fluorescence was subtracted, the fluorescence intensity ratio of Ca²⁺ domains was calculated according to Equation 5. T a r g e t i n g   F a c t o r = p e a k / v a l l e y K C l p e a k / v a l l e y B a s e l i n e − 1 (5)

Freshly dissociated neurons with the initial segment of their axon were cultured for 72 h in plates pre-coated with poly-L-lysine and Aplysia hemolymph. Plates were washed three times with 0.6% saline solution (10 min each) at RT. Subsequently, they were fixed with 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) for 45 min at RT. Following three washes with 0.01 M PBS, cells were treated with a blocking solution containing 5% bovine serum albumin (BSA) and 0.25% saponin in 0.01 M PBS for 1 hour. Cultured cells were then incubated overnight at 4 °C with rabbit polyclonal antibodies (Ab) against Ca²⁺ channels diluted in the blocking solution (Ca_V2.1: AB5152 Ab, Ca_V2.2: AB5154 Ab, Sigma-Aldrich, Germany) or with rabbit primary Ab against Helix synapsin (HelSyn) produced by InBios International.

Afterward, three washes were performed with 0.25% saponin in 0.01 M PBS, followed by incubation in a blocking solution with secondary goat anti-rabbit IgG TRICT T6778 (Sigma-Aldrich, Germany) or donkey anti-rabbit IgG Alexa A10040 (Life Technologies) for 45 min at RT. Three gentle washes with 0.01 M PBS were conducted for 10 min each, and samples were maintained in PBS at 4 °C until fluorescence microscopy visualization.

Cell images were captured using an inverted microscope (Olympus CKX53) equipped with optical epifluorescence. The varicosity fluorescence was quantified following the protocol described for Ca²⁺ fluorescence in neurites. However, autofluorescence was used to normalize volume in this case, applying the following Equation 6. T a r g e t i n g   F a c t o r = p e a k / v a l l e y a n t i b o d y p e a k / v a l l e y a u t o f l o u r e s c e n c e − 1 (6)

Due to their low volume, the fluorescence in neurites was not normalized; therefore, the peak-to-valley (domain) ratio was reported.

Data were expressed as mean values ±S.E.M. Statistical analysis was performed using GraphPad Prism 10 (GraphPad Software, Boston, MA, United States). The significance of the groups was assessed using parametric or non-parametric statistics, depending on the results obtained from the Shapiro-Wilk test for normality and the F-test for homogeneity of variances. When comparing only two groups, the Student's t-test, the Mann-Whitney U test or the Wilcoxon matched-pairs signed-rank test was used. When comparing more than two groups or groups with variations over time, two-way ANOVA was performed, followed by the Bonferroni post hoc test, repeated measures ANOVA or mixed-effects model where appropriate, or the Kruskal–Wallis test, followed by Dunn’s multiple comparisons test. Significance levels were set at p < 0.05.

The current-voltage (I-V) relationship of the whole Ca²⁺ current is illustrated in Figure 1A. The Helix Ca²⁺ current was activated at −40 mV and peaked typically between −20 mV and 0 mV (n = 49). The activation curve showed a high voltage dependence (Figure 1B, black circles) with an absolute k value of 1.2 ± 0.3, and the activation V_1/2 determined by the Boltzmann equation of individual activation curves was −27.2 ± 1.3 mV (n = 42). The analysis of the steady-state inactivation of the Ca²⁺ macrocurrents showed partial inactivation and a small voltage dependency (Figure 1B, white squares). Fitting each individual inactivation curve to the Boltzmann equation yielded a V_1/2 value of −16.9 ± 1.5 mV and a k value of 7.4 ± 0.5 (n = 37).

Using a voltage ramp from −80 mV to +50 mV (Figure 2, lower panel), it was determined that only HVA Ca_V channel populations were involved in the macrocurrents of the C1 neuron. All cells presented a single current peak at a membrane potential of −12.32 ± 0.89 mV (n = 43) (Figure 2, upper panel), suggesting the absence of LVA Ca_V3-like channels.

A pharmacological dissection was performed using channel blockers to identify the HVA Ca_V types contributing to the Ca²⁺ macrocurrents. Initially, the ω-CnTx CVIB was used to determine the presence of Ca_V2.1 and Ca_V2.2-like channels. Application of the ω-CnTx CVIB produced a strong reduction in the peak inward currents at most of the voltages tested (n = 7) (voltages F_{(12, 78)} = 6.29, p < 0.0001, treatment F_{(1, 74)} = 25.10, p < 0.0001, interaction F_{(12, 74)} = 2.51, p = 0.008, mixed-effects model) (Figure 3A). There was not a significant change in the V_rev after CnTx treatment (control = 35.0 ± 9.8 mV vs. ω-CnTx CVIB = 21.8 ± 16.0 mV, n = 6; t (5) = 1.007, p = 0.36, paired t-test) In addition, the maximum Ca²⁺ current decreased by 63.2% compared to controls (control = −1.89 ± 0.53 nA/nF vs. ω-CnTx CVIB = −0.70 ± 0.21 nA/nF, n = 6; W = 21, p = 0.031, r = 0.34, Wilcoxon matched-pairs signed-rank test) (Figure 3B).

The ω-CnTx CVIE was used to specifically identify the presence of currents mediated by Ca_v2.2-like channels. Application of this inhibitor significantly reduced the Ca²⁺ inward currents between −10 and 50 mV (n = 7) (Figure 4A, open circles). Following these recordings and assuming Ca_V2.2-like channels were blocked, cilnidipine was added to the bath to assess the contribution of Ca_v1-like channels to the remaining current. After 20 min of cilnidipine application, no further significant reduction in current was observed compared with that in the presence of ω-CnTx CVIE alone (n = 7) (voltage F_{(2.13, 12.77)} = 26.82, p < 0.0001, treatment F_{(1.11, 6.66)} = 20.95, p = 0.0026, interaction F_{(1.84, 11.04} = 3.90, p = 0.055, two-way ANOVA) (Figure 4A).

Similarly, ω-CnTx CVIE reduced the maximum Ca²⁺ current by 25.6%, changing from −5.15 ± 0.5 nA/nF to −3.83 ± 0.45 nA/nF (p = 0.006, n = 7). Since the effect of ω-CnTx CVIB is stronger than the effect of ω-CnTx CVIE, these results suggest that both Ca_V2.1-like and Ca_V2.2-like channels are present in the C1 neuron and contribute substantially to the macroscopic Ca²⁺ current. Even though, 20 min after cilnidipine exposition, an additional 12.6% reduction in the maximum Ca²⁺ current was observed (−3.19 ± 0.6 nA/nF), this change was not statistically significant (p = 0.15, n = 7) (F_{(1.12, 6.70)} = 16.86, p = 0.0044, one-way repeated measures ANOVA) (Figure 4B). These results suggest that Ca_V1-like channels, if present, do not play a significant role in the Ca²⁺ current of the C1 neuron in Helix.

In order to confirm the existence of different Ca_V2 isoforms, we used antibodies against Ca_V2.1 and Ca_V2.2. Since Ca²⁺ is the main signal evoking neurotransmitter release, we started analyzing varicosities. To quantify the percentage of mature varicosities vs. silent varicosities, we marked Helix synapsin (HelSyn) with a custom-designed antibody (Figure 5A, upper panels and middle panel insert), obtaining that in control cells 65.5 ± 4.34% (n = 15) of the varicosities have the exocytic machinery (Figure 5B).

Different cells were used to test the presence of Ca_V2.1 and Ca_V2.2 channels (Figure 5A, lower panels). When varicosities were analyzed, the percentage of varicosities positive for Ca_V2.1 (45.38 ± 8.67%, n = 8) was statistically similar to those positive for HelSyn (p = 0.09). On the other hand, the percentage of varicosities positive for Ca_V2.2 (22.40 ± 9.90%, n = 4) were smaller than for HelSyn (p = 0.002) (Figure 5B). Although there was a trend towards a higher percentage of varicosities with Ca_V2.1, no significant differences were found (p = 0.211) (F_{(2, 24)} = 8.34, p = 0.0018, one-way ANOVA).

Since both antibodies marked these structures, we analyzed the targeting factor of each channel in the positive varicosities, and no significant differences were found in varicosities regarding the presence of Ca_V2.1 and Ca_V2.2 channels. Nevertheless, a trend towards a higher localization of Ca_V2.1 can be observed with a targeting factor of 0.69 ± 0.19 (n = 8) for Ca_V2.1 and 0.24 ± 0.17 (n = 4) for Ca_V2.2 (U = 5, p = 0.07) (Figure 5C).

From these two measurements, we can derive that not all the varicosities have the exocytic machinery, and they predominantly contain Ca_V2.1, although some may also have Ca_V2.2.

The presence of domains for both Ca_V2.1 and Ca_V2.2 was also observed in neurites. Interestingly, the number of domains per cell was higher for Ca_V2.2 (12.75 ± 6.20) than for Ca_V2.1 (1.13 ± 0.52) (p = 0.033). Also, as expected, almost no domains for HelSyn were observed (0.53 ± 0.28) (Figure 5D). It is worth noting that the number of domains with HelSyn was similar to Ca_V2.1 domains (p = 0.8861) but differed from Ca_V2.2 domains (p = 0.0008) (H_3,31 = 13.27, p = 0.0013, Kruskal–Wallis). In addition, when comparing the targeting factor of the channels in the neurites, it was found that the targeting factor of Ca_V2.2 was almost twice that of Ca_V2.1 (2.22 ± 0.29 vs. 1.32 ± 0.50, respectively); however, this difference was not statistically significant (U = 13.0, p = 0.64) (Figure 5E).

All together, these data suggest that varicosities predominantly contain Ca_V2.1, which would likely be functional varicosities. In addition, Ca_V2.2 is more localized in neurites, with no neurotransmitter release sites.

The predominant expression of Ca_V2.2-like channels in neurites compared to varicosities could be related to different contributions of this channel to Ca²⁺ kinetics within these structures. To test this hypothesis, we performed Ca²⁺ imaging during KCl-induced neuronal activity in the absence and presence of ω-CnTx GVIA, a specific Ca_V2.2 blocker.

When analyzing the distribution of Ca²⁺ in varicosities during 5 min of neural activity, we observed different fluorescence patterns among varicosities over time. For this reason, the varicosities were grouped according to their behavior patterns, and the three most frequent patterns were analyzed.

In the constant behavior pattern (Figure 6A), an increase in intracellular Ca²⁺ was observed in the varicosities after adding KCl, which remained constant throughout the recording period (from minutes 2–7) (n = 5). The varicosities of the cells treated with ω-CnTx GVIA (n = 6) exhibited a behavior similar to that of the control group (time F_{(2.52, 22.1)} = 28.2, p < 0.0001; treatment F_{(1, 9)} = 0.11, p = 0.74; interaction F_{(85, 745)} = 0.41, p > 0.99, two-way ANOVA).

Figure 6B shows the ascending linear behavior pattern in which, after adding KCl, Ca²⁺ increased continuously in a linear fashion throughout the 5 min of recording (n = 4). In the same way, varicosities treated with ω-CnTx GVIA (n = 4) showed a statistically similar behavior (time F_{(1.79, 10.2)} = 17.9, p = 0.0053; treatment F_{(1, 6)} = 0.125, p = 0.73; interaction _{(85, 483)}, p = 0.99; two-way ANOVA).

In addition, some varicosities showed a fast increase in the Ca²⁺ signal after KCl addition, followed by a decrease in fluorescence intensity (Figure 6C) (n = 6). Similar to the former two behaviors, ω-CnTx GVIA treatment did not affect signal behavior (n = 4) (time F_{(1.91, 15)} = 19.4, p < 0.0001; treatment F_{(1, 8)} = 0.443, p = 0.52; interaction F_{(85, 667)} = 0.46, p > 0.99; two-way ANOVA).

These results suggest that even if some varicosities express Ca_V2.2-like channels, their contribution to the presence of Ca²⁺ in these structures is not significant.

On the other hand, when we analyzed the number of Ca²⁺ domains along neurites during neuronal activity (Figure 6D), a substantial decrease was evident after Ca_V2.2-like inhibition with ω-CnTx GVIA (t₍₁₇₎ = 2.84, p = 0.011) (Figure 6E, upper panel). In addition, in the observed domains, the targeting factor of the Ca²⁺ signal decreased after Ca_V2.2-like inhibition (t₍₁₇₎ = 2.24, p = 0.039; respectively) (Figure 6E, lower panel). This suggests that, contrary to what was observed in varicosities, Ca_V2.2-like channels play an important role in the Ca²⁺ influx in neurites, evoking Ca²⁺ microdomains.

It is known that APs in C1 neurons are dependent on both Na⁺ and Ca²⁺, and Ca_V2.2-like channels were mostly observed in sites not related to neurotransmitter release. Therefore, Ca_V2.1 and Ca_V2.2-like channels may contribute to AP firing. The blocker that had the most significant effect on Ca²⁺ currents was the ω-CnTx CVIB, and for this reason, it was used to examine the role of these channels on neuronal excitability.

The MFF in response to increasing stimuli did not show significant changes after exposure to ω-CnTx CVIB (0.5 nA: 2.53 ± 0.79 Hz; 1.0 nA: 7.33 ± 1.03 Hz; 1.5 nA: 9.47 ± 1.23 Hz; n = 15 in controls; vs. 0.5 nA: 2.86 ± 1.50 Hz; 1.0 nA: 7.71 ± 1.87 Hz; 1.5 nA: 11.71 ± 2.11 Hz; n = 7 with ω-CnTx CVIB) (stimulus F_{(3, 80)} = 31.41, p < 0.0001, treatment F_{(1, 80)} = 0.76, p = 0.39, interaction F_{(3, 80)} = 0.36, p = 0.78, two-way ANOVA). However, a small not significant trend to a decreased MFF can be observed (Figure 7A).

When the AP waveform was analyzed (Figure 7B), the exposure to ω-CnTx CVIB induced a small trend on the verge of statistical significance, increasing the time to peak from 26.75 ± 4.21 ms (n = 4) in controls to 40.25 ± 7.62 ms (n = 4) after ω-CnTx CVIB exposure (t₍₃₎ = 2.47, p = 0.0897, paired t-test) (Figure 7C, left panel). Similarly, a small increase in the rise time 10–90% was observed (15.24 ± 3.79 ms, n = 4 in controls; vs. 25.84 ± 5.64 ms, n = 4 with ω-CnTx CVIB) (t₍₃₎ = 2.61, p = 0.0796, paired t-test) (Figure 7C, right panel). This slowing of depolarization could be related to the slightly smaller MFF. No changes or trends were observed in resting membrane potential, rheobase, AP amplitude, rise tau, maximum rise slope, after-hyperpolarization amplitude, decay tau, decay slope 90-10%, or maximum decay slope (data not shown).

The calcium macrocurrents of the C1 neuron of Helix aspersa were activated at −40 mV, with the current peak between −20 mV and 0 mV and a reversal potential of +53.1 ± 3.54 mV. These values are comparable to those previously reported in Helix neurons (Brenes, 2015a) and in other species; for example, mature mouse motoneurons, when low or physiological calcium concentrations were used (1–3 mM), activated at −40 mV, with the current peak between −10 and 0 mV and with a reversal potential of +40 mV (Carlin et al., 2000). In motoneurons and sensory neurons of lampreys, the current activation started between −40 and −30 mV (El Manira and Bussie, 1997). In rat chromaffin cells, the current activation started at −40 mV, peaked at 0 mV, and reverted between +50 and +60 mV (Gandía et al., 1995).

In our study, the activation curve presented a pronounced slope between −50 and 0 mV, which indicates a high voltage dependence between those potentials. The calcium macrocurrents presented a half inactivation potential of −16.9 ± 1.5 mV. This value is similar to what has been reported for LCa_V1a in Lymnaea stagnalis (V_1/2 = −15 mV) (Spafford et al., 2006), but more positive than what has been reported for LCa_V2 in the same species (V_1/2 = −39.3 mV) and in rat Ca_V2.2 (V_1/2 = −49.1 mV) (Senatore and Spafford, 2010). The inactivation slope (k = 7.4 ± 0.5) was comparable to the one observed in mouse motoneurons (k = 7.0) (Carlin et al., 2000) and in LCa_V2a in L. stagnalis (k = 8.0) (Spafford, 2003b).

The calcium macrocurrents of the C1 neurons of Helix aspersa have incomplete inactivation. This finding matches what has been reported on neurons from the suboesophageal ganglia of the same species, where it was determined that the inactivation was voltage and calcium-dependent (Brown et al., 1981). We cannot exclude the possibility that in our study, the degree of inactivation could be affected by the presence of EGTA in the intracellular medium, as this calcium chelator could avoid the possibility of calcium-dependent inactivation to take place (Hille, 2001). However, since partial inactivation was evident from the first recordings at the beginning of the experiments, it is unlikely that it had significantly affected the observed currents. It is known that the C1 neuron APs are dependent on sodium, calcium, and potassium currents (Brenes, 2015a). So, functionally, an incomplete inactivation of the Ca_V could induce an increase in the duration of the calcium currents and affect the morphology of the AP and the excitability of the cell.

A ramp protocol was used to identify the population of Ca_V according to their voltage of activation. All the cells in the study presented only one current peak at a membrane potential of −12.32 ± 0.89 mV (n = 43). This finding suggests that the C1 neuron of Helix aspersa in whole-cell configuration expresses only HVA channels in the plasma membrane. This finding differs from what has been reported in the giant cerebral neuron of L. stagnalis where, while using a similar protocol but with axotomized somata, two different peak currents were observed, suggesting the presence of LVA and HVA channels (Staras et al., 2002).

A pharmacological dissection was used to identify the types of HVA channels present in the C1 neuron. When the cells were exposed to ω-CnTx CVIB, a broad-spectrum blocker that inhibits Ca_V2.1 and Ca_V2.2 in mammals (Lewis et al., 2000), a 63.2% reduction of the calcium macrocurrents was observed. To determine if the C1 cell expresses one or both subtypes of channels, the cells were exposed to ω-CnTx CVIE, a highly selective Ca_V2.2 channel blocker. This blocker induced a 25.6% reduction of the calcium macrocurrents. Considering that the reduction of the currents with ω-CnTx CVIE was considerably less than the reduction induced by ω-CnTx CVIB, it could be suggested that the C1 neurons express both Ca_V2.1 and Ca_V2.2-like channels.

After blocking Ca_V2.2 channels, a subset of cells were exposed to cilnidipine, an organic compound that blocks Ca_V1 and Ca_V2.2 channels (Uneyama et al., 1997). No significant differences were observed in either the I–V curves or the peak currents, suggesting that Ca_V1-like channels are not functionally expressed in the plasma membrane of C1 neurons.

Unlike mammals, it has been argued that invertebrates only have one type of Ca_V1, Ca_V2, and Ca_V3, based mainly on studies made with Lymnaea stagnalis. However, our work and studies carried out by Azanza et al. (2008), Battonyai et al. (2014), and Kiss et al. (2012) proved the existence of different subtypes of voltage-gated channels. Specifically, Azanza et al. (2008) first reported the existence of Ca_V2.1 and Ca_V2.2 channels in Helix; subsequently, our differences in current inhibition between the Ca_V2.1 and 2.2 blocker ω-CnTx CVIB and the specific Ca_V2.2 blocker ω-CnTx CVIE supported the existence of two different isoforms of Ca_V2 in Helix neurons. This was confirmed by using antibodies against these two channels, where a slightly different distribution in the labeling was evident.

Despite no significant differences, Ca_V2.1 was the predominant channel in varicosities, doubling the number of Ca_V2.2 positive varicosities, and the number of positive varicosities for Ca_V2.1 better matches the positive varicosities containing presynaptic machinery (Figure 5B). In addition, this was supported by the targeting factor analysis, which tends toward a greater presence of Ca_V2.1 over Ca_V2.2 in these structures (Figure 5C).

Since some varicosities expressed Ca_V2.2, we measured the Ca²⁺ kinetics in these structures in the presence of a Ca_V2.2 inhibitor (ω-CnTx GVIA). Following KCl stimulation, we observed Ca²⁺ increased regardless of ω-CnTx GVIA treatment (Figures 6A–C), confirming the hypothesis that Ca_V2.1 are the dominant channels in varicosities, and Ca_V2.2 channels do not play a major role in Ca²⁺ functions in these structures. In consequence, Ca_V2.1 is likely involved in synaptic functions and neurotransmitter release (Hilfiker et al., 1999).

Interestingly, the Ca²⁺ curve trends displayed different behavioral patterns after the KCl application. In some varicosities, Ca²⁺ increased and remained constant (Figure 6A); in others, it continued to rise throughout the recording period (Figure 6B), and in some of them, it decreased over time (Figure 6C). It is known that Ca²⁺ behavior over time varies depending on the stimulus, and this behavior modulates different neuronal responses. However, most studies of this type are conducted in millisecond-scale recordings (Grienberger and Konnerth, 2012; Martens et al., 2014). In this study, due to equipment limitations preventing image capture at intervals shorter than 3 s, recordings ranged from 3 to 5 s, providing a broader temporal perspective.

Under physiological conditions, depolarization through APs leads to the opening of Ca_V channels, resulting in increased intracellular Ca²⁺. This suggests that the different behavior patterns are related to distinct firing patterns in neurites and their varicosities. In varicosities where Ca²⁺ increases and remains constant throughout the recording period, the AP firing frequency is likely stable, with Ca²⁺ influx through Ca_Vs being balanced by its removal via pumps and transporters. In varicosities where intracellular Ca²⁺ increases throughout the recording, it may indicate a higher firing frequency and Ca²⁺ influx exceeding the removal capacity of pumps and transporters. Finally, in varicosities where Ca²⁺ initially increases but declines towards the end of the recording, this could be attributed to firing rate adaptation or, alternatively, to inactivation of Ca²⁺ channels facilitating intracellular accumulation, such as Ca_Vs, or increased activity of pumps and transporters promoting Ca²⁺ clearance from the cytosol (Peterson et al., 1999).

Conversely, in neurites, although no significant differences in the targeting factor between Ca_V2.1 and Ca_V2.2 were observed (Figure 5E), results showed a trend toward greater Ca_V2.2 presence in these structures. Additionally, the number of Ca_V2.2 domains in neurites was greater than for Ca_V2.1 (Figure 5D). Aligned with this, intracellular Ca²⁺ analysis in neurites revealed aggregation sites alongside some neurites and the application the Ca_V2.2 inhibitor ω-CnTx GVIA resulted in a decrease in both the targeting factor and the number of Ca²⁺ domains per cell (Figure 6E). These results suggest that Ca²⁺ domain formation in neurites primarily involves Ca_V2.2-like channels.

Ca_V2 studies and their functions have been conducted mainly in vertebrate models, particularly mammals. In mollusks of the genera Lymnaea and Aplysia, Ca_V2-like channels have been reported, sharing high similarity with mammalian channels (Brenes, 2022). LCa_V2 also exhibit functions similar to mammalian channels in mediating neurotransmitter release (Fossier et al., 1994; Kits and Mansvelder, 1996; Senatore and Spafford, 2010) and have been studied using neuromodulators commonly employed in mammalian research, yielding similar results (Fossier et al., 1994; Hong and Lnenicka, 1997). The role of Ca_V2.1 and Ca_V2.2 in initiating neurotransmitter release in rat and mouse cells is well-documented (Catterall, 2011). These channels interact with SNARE complex proteins at the intracellular loop connecting domains II and III, facilitating vesicular exocytosis (Catterall, 2011). However, postnatal development studies indicate that cooperative Ca_V2.1 and Ca_V2.2 coupling shifts to Ca_V2.1 nanodomains initiating synaptic vesicle exocytosis (Baur et al., 2015; Cingolani et al., 2023). Furthermore, the distance between Ca_V2.1 and the Ca²⁺ sensor responsible for synaptic vesicle release (synaptotagmin) determines synaptic strength (Dittman and Ryan, 2019; Cingolani et al., 2023), and their interaction promotes synaptic facilitation, a short-term plasticity process (Mochida et al., 2008). These reinforce the idea that although both channels participate in neurotransmitter release, Ca_V2.1 plays the primary role, as our results suggest.

On the other hand, Ca_V2.2 channels and their N-type currents have been characterized in various mammalian neurons, notably dorsal root ganglion neurons involved in pain transmission (Heinke et al., 2004; Adams and Berecki, 2013). These channels regulate neurotransmitter release and synaptic plasticity and, interestingly, have been observed in postsynaptic terminals, contributing to graded potentials (Heinke et al., 2004; Adams and Berecki, 2013). In addition, Ca_V2.2 presence in neurites has been reported in Xenopus laevis embryonic neurons (Sann et al., 2008), which may explain Ca_V2.2 predominance in Helix neurites, suggesting a role in Ca²⁺ domains and APs.

The role of Ca_V2.1 and Ca_V2.2 channels on cell excitability was evaluated by taking as markers the resting potential, rheobase, and firing frequency in response to stimuli of increasing intensity. No significant effect was observed in any of the variables, suggesting that, under normal conditions, basal calcium currents do not contribute significantly to the excitability of these neurons. However, it has been reported that, under pathological conditions, an increase in these currents significantly influences neuronal excitability (Brenes, 2015b; Brenes et al., 2016).

When blocking Ca_V2.1 and Ca_V2.2 channels with the ω-CnTx CVIB, a slight tendency toward an increased duration of the depolarization was observed in C1 neurons. These results could suggest that channel blockade has a greater effect on the final third of the depolarization phase, since there was no difference between the depolarization taus between cells in the control state and cells in the presence of ω-CnTx CVIB, but a tendency toward an increase in the time to peak was observed. The partial blockade of calcium channels by ω-CnTx CVIB, which appeared to modestly influence the depolarization phase of C1 neuron APs, aligns with previous reports indicating that depolarization in Helix C1 neurons depends on both sodium and calcium currents. Furthermore, in this same study, it was reported that an increase in calcium currents was related to an increase in potassium outward currents through BK channels (Brenes, 2015b). This factor could explain why a reduction in calcium inward currents could cause repolarization to start later, since the opening of BK channels would be delayed.

Combining all the results, a picture emerges where C1 serotonergic neurons expressed Ca_V2-like channels. Two Ca_V2 isoforms were observed, with more significant Ca_V2.1 presence in the neurotransmitter release structures (varicosities) and more significant Ca_V2.2 presence in neurites. Ca²⁺ increase was evident in varicosities and neurites during cellular activity; Ca_V2.1 can be related to Ca²⁺ dynamics in the varicosities, controlling neurotransmitter release; and Ca_V2.2 can be associated with AP firing in neurites and probably in varicosities.