Functional Coupling Between Voltage-Dependent Sodium Channels and Activation of the Ca2+ Signaling That Mediates Endothelial Cell Migration

TL;DR

This study explores how sodium channels and calcium signaling work together to help blood vessel cells move, which is important for forming new blood vessels.

Contribution

The paper reveals a novel functional coupling between Nav1.2 and Nav1.6 channels with NCXrm and Cx43 hemichannels in endothelial cell migration.

Findings

Blocking Nav channels inhibits endothelial cell migration and calcium signaling.

Nav1.2 channels associate with caveolin-1 during cell migration.

Nav channels are essential for Cx43 hemichannel opening and tubular structure formation.

Abstract

Angiogenesis depends on Ca2+-mediated endothelial cell migration. The increase in intracellular Ca2+ concentration ([Ca2+]i) is coordinated by caveolae and the Cx43 hemichannel opening. However, the functional coupling of voltage-dependent Na+ channels (Nav) with Na+-Ca2+ exchanger reverse mode (NCXrm) activation may contribute to the response, which was evaluated using the wound-healing assay in primary cultures of rat mesenteric endothelial cells. Changes in [Ca2+]i, the hemichannel opening and the association of Nav channels with caveolin-1, a caveolae structural protein, were analyzed. Both endothelial cell migration and the associated Ca2+ signaling were inhibited by tetrodotoxin (TTX), a Nav channel blocker, lamotrigine, a preferential Nav1.2 inhibitor, or 4,9-anhydro-TTX, a specific Nav1.6 blocker. A similar result was found by disrupting caveolae organization with methyl-β-cyclodextrin or blocking NCXrm with SEA0400. TTX and SEA0400 also prevented Cx43 hemichannel opening, and tubular-like structure formation depended on Nav channels. An analysis using a proximity ligation assay showed that endothelial cell migration was paralleled by the progressive association of caveolin-1 with Nav1.2, but not Nav1.6, channels. These results suggest that the functional coupling of Nav1.2 and Nav1.6 channels with the activation of NCXrm and Cx43 hemichannels mediates the Ca2+ signaling associated with endothelial cell migration and angiogenesis, which provides new targets to modulate angiogenesis in physiological or pathological conditions.

Figures

Full text

1. Introduction

The control of blood flow distribution is a critical process that depends on the changing metabolic demand of the different tissues of the organism. Therefore, the architecture of the vascular network must be dynamically regulated to match the energetic requirements of the cells and the physiological conditions of the tissue over time [1]. In this context, angiogenesis, the growth of new blood vessels from pre-existing ones, plays an important role in the long-term regulation of tissue irrigation in different physiological and pathological conditions such as wound-healing, tissue regeneration, embryonic development, and tumor growth, which highlights the importance of this process in the control of cell function [2,3].

Endothelial cells play a central role in the control of vascular function and in the progress of angiogenesis, which is initiated by the highly regulated Ca^2+^-dependent migration of endothelial cells aimed to form tubular structures that provide the cellular basis for the formation of a new blood vessel. It is noteworthy that the increase in intracellular Ca^2+^ concentration ([Ca^2+^]i) that leads to endothelial cell migration is not a transient signal, but rather it is involved in the control of the whole process [4], and, although Ca^2+^ signaling is triggered by its release from intracellular stores, the response is sustained through the activation of a Ca^2+^ influx pathway from the extracellular compartment [5,6]. Several channels have been involved in the Ca^2+^ signaling observed during the progress of endothelial cell migration, including transient receptor potential (TRP) channels, calcium release-activated calcium modulator 1 (Orai1), cyclic nucleotide-gated (CNG) ion channels, and, most recently, connexin 43 (Cx43)-formed hemichannels [4,7,8,9].

Interestingly, tetrodotoxin (TTX)-sensitive as well as TTX-resistant isoforms of voltage-dependent Na^+^ channels (Na_v_) and the isoform Ca_v_3.2 of T-type voltage-dependent Ca^2+^ channels have been found to be expressed in endothelial cells [10,11,12,13,14]. In addition, these channels were recently shown to participate in the control of endothelial cell [Ca^2+^]i in the mesenteric resistance arteries of rats [15]. Likewise, the time course of Ca^2+^ signaling observed in response to the pro-angiogenic stimulation with VEGF was found to be associated with endothelial cell membrane depolarization and an increase in intracellular Na^+^ concentration [16,17], which is in line with findings indicating that the increase in [Ca^2+^]i involved in endothelial cell migration and tubular structure formation is mediated, at least in part, by the functional coupling of Na_v_ channels with the activation of the reverse mode of the Na^+^-Ca^2+^ exchanger (NCX) [11,18]. Although the NCX normally extrudes Ca^2+^ from the cell (forward mode), the activity of the exchanger depends on the electrochemical gradient of the transported ions (i.e., Na^+^ and Ca^2+^), and a Na_v_ channel-mediated local increase in [Na^+^]i may activate the NCX Ca^2+^ entry mode (i.e., the reverse mode). In addition, Na_v_ channels may lead to a Ca^2+^ influx triggered by the opening of Ca_v_ channels [19]. However, the molecular mechanism by which Na_v_ channels regulate endothelial cell migration during angiogenesis in the microcirculation and the isoforms of these channels involved in the process remains to be determined.

In the present study, we evaluated the potential participation of Na_v_ channels in the signaling mechanism coupled to the increase in [Ca^2+^]i observed during endothelial cell migration and, furthermore, tubular-like structure formation. Our findings indicate that Na_v_1.2 and Na_v_1.6 channels play a critical role in the Ca^2+^-dependent mechanism that mediates endothelial cell migration and new vessel formation in angiogenesis. Interestingly, Na_v_1.2 is associated with caveolae-dependent signaling pathways that involve the spatial association of this isoform with caveolin-1, a structural protein of caveolae, whereas Na_v_1.6 is linked to a caveolae-independent signaling pathway.

2. Results

The analysis of cell migration in the wound-healing assay was initiated by scraping a thin strip in the middle of the monolayer of primary cultures of mesenteric endothelial cells, and the time course of cell movement from both sides of the wounded area (i.e., migration front) was monitored. The endothelial cell migration front continuously advanced into the cell-free scratched space and, in control conditions, the closure of the wounded area was attained in ~20 h (Figure 1A,B). Therefore, changes in endothelial cell migration were analyzed 15 h after starting the wound-healing assay to be able to detect a negative or positive effect in the response.

The involvement of Na_v_ channels in the migration response was confirmed by the treatment with TTX, which resulted in a clear reduction in endothelial cell migration, and, notably, the inhibitory effect was not enhanced by increasing the blocker concentration from 1 µM to 100 µM (Figure 1C), suggesting the participation of TTX-sensitive Na_v_ channels in the process. In addition, the application of 100 µM lamotrigine, a preferential blocker of Na_v_1.2-formed channels, or 500 nM 4,9-anhydro-TTX, a preferential inhibitor of Na_v_1.6-based channels, also inhibited the response in a similar magnitude to that observed with 1 µM TTX alone or in combination with these Na_v_ channel blockers (Figure 1D). Consistent with these results, the expression of the isoforms Na_v_1.2 and Na_v_1.6 of these channels was detected in mesenteric endothelial cells through immunofluorescence analysis (Figure 1E), which is in line with previous reports [15,20]. Interestingly, the analysis using the bromodeoxyuridine (BrdU) assay in cultures at 40% and 80% confluency indicated that the reduction observed in wound-healing was not associated with an inhibition in cell proliferation (Figure 2A), which was further confirmed through the direct visualization of BrdU positive cells located at the migration front or at the monolayer in control conditions or in the presence of 1 µM TTX (Figure 2B). These results highlight that the reduction observed in the closure of the scratched area in the absence of functional Na_v_ channels can only be attributed to the inhibition of endothelial cell migration and not to changes in cell proliferation.

2.1. The Ca2+ Signaling Associated with Endothelial Cell Migration Is Mediated by Nav Channels

Endothelial cell migration is commanded by the generation of a Ca^2+^ signal, and, in agreement with this, the initiation of the wound-healing assay was associated with a clear increase in [Ca^2+^]i that was restricted exclusively to the migration front (Figure 3). In line with the participation of Na_v_ channels in endothelial cell migration, the Ca^2+^ signaling activated in migrating cells was blocked by 1 µM TTX, 100 µM lamotrigine or 500 nM 4,9-anhydro-TTX (Figure 3). Likewise, the treatment with TTX also reduced the tubular structure formation in an in vitro angiogenesis assay (Figure 4). Taken together, these results indicate that Na_v_ channels play a central role in the signaling mechanism that controls endothelial cell migration and tubular structure formation during angiogenesis.

As caveolae play a central role in the control of Ca^2+^ signals during cell migration, we evaluated through a proximity ligation assay the spatial relation between Na_v_ channels and caveolin-1 (Cav-1), which is a structural protein of caveolae that is involved in the spatial organization of Ca^2+^-related signaling proteins [21,22,23]. The expression of Cav-1 in endothelial cells was confirmed through an immunofluorescence analysis (Figure 1E), and, interestingly, the activation of endothelial cell migration triggered the spatial association of Na_v_1.2 channels with Cav-1, which progressively increased from 15 min to 4 h after scratching the monolayer (Figure 5A). Although the increment in the association of these two proteins was mainly achieved at the migration front, a rise in the proximity ligation assay signal was also apparent in the monolayer 4 h after the initiation of the wound-healing assay (Figure 5A). In contrast, the activation of endothelial cell migration did not lead to the association of the Na_v_1.6 channel isoform with Cav-1 (Figure 5B).

2.2. The Ca2+ Signaling Associated with Endothelial Cell Migration Depend on Caveolae Integrity

To confirm the participation of caveolae in the activation of endothelial cell migration, cholesterol-rich microdomains were disrupted through a treatment with methyl-β-cyclodextrin (MβCD), which resulted in a strong inhibition in the closure of the wounded area (Figure 6A). Consistent with the reduction in endothelial cell migration, the treatment with MβCD also prevented the increase in [Ca^2+^]i observed in the cells located at the migration front (Figure 6B), and, interestingly, a similar result was attained after blocking the reverse mode of NCX with 1 µM SEA0400 or after incubating the cells with a buffer solution containing low [Na^+^] (Low Na^+^) to disable the operational coupling of Na_v_ channels with NCX function (Figure 7). In conjunction, these results are in line with the importance of caveolae in the molecular organization of the signaling that controls endothelial cell migration and suggest that the activation of NCX reverse mode mediates the mechanism by which Na_v_ channels trigger the Ca^2+^ signal that directs the migration process.

In addition to the Na^+^-Ca^2+^ exchanger, we recently demonstrated that the S-nitrosylation-mediated opening of Cx43-formed hemichannels also provides a pathway of Ca^2+^ entry during endothelial cell migration [4], which may contribute to the Na_v_ channel-initiated mechanism of Ca^2+^ signaling. Consistent with this hypothesis, the increase in ethidium uptake observed in control conditions was abolished by the treatment with 1 µM TTX, 1 µM SEA0400 or 300 µM TAT-Gap19, a specific Cx43 hemichannel blocking peptide (Figure 8A,B), indicating that the Cx43-formed hemichannel-mediated Ca^2+^ component associated with endothelial cell migration depends on the activation of the Na_v_ channel-NCX reverse mode signaling pathway. Furthermore, in line with the effect of TTX, ethidium uptake was also blocked by lamotrigine and 4,9-anhydro-TTX (Figure 8C).

3. Discussion

Endothelial cell migration plays a central role in the formation of new vessels to keep the homeostasis of the microvascular network through angiogenesis [24]. The migration of endothelial cells is orchestrated by the interplay of different transduction pathways that coordinate localized increases in [Ca^2+^]i [25,26]. It has been shown that Na_v_ channels are involved in the endothelial cell Ca^2+^-mediated control of vasomotor tone and angiogenesis [11,15,20]. However, the precise mechanism by which these channels are implicated in endothelial cell migration remains to be determined, and the Na_v_ isoforms involved have not been clearly identified. Our findings indicate that the Ca^2+^ signaling supporting the initiation and progress of endothelial cell migration and further formation of tubular structures depends on the functional coupling of Na_v_1.2 and Na_v_1.6 channels with the activation of the reverse mode of the Na^+^-Ca^2+^ exchanger and the opening of Cx43-based hemichannels [4]. The integrity of caveolae plays a critical role in the Ca^2+^-mediated mechanism of endothelial cell migration, and, interestingly, the development of the migrating response is paralleled by an increase in the spatial association of Na_v_1.2, but not Na_v_1.6, with Cav-1, denoting the importance of caveolae in the Na_v_1.2-triggered signaling.

Na_v_ channels play a major role in the control of the electrical activity of neurons and skeletal and cardiac muscle, and, thereby, these cells are termed as electrically excitable [27]. However, several isoforms of Na_v_ channels have also been found to be expressed in different cell types traditionally considered to be non-excitable. Although the study of Na_v_ channels in non-excitable cells is just beginning, and their function in these cells has not been clearly determined, it has been proposed that these channels may be involved in the control of membrane potential [28]. In addition, it was reported that Na_v_ channels participate in the regulation of migration in astrocytes and in the invasiveness and metastatic potential of several types of cancer cells [29,30]. At least four isoforms of Na_v_ channels have been detected as being expressed in endothelial cells: Na_v_1.2, Na_v_1.6, Na_v_1.5 and Na_v_1.9. In this context, it is important to note that the sensitivity to TTX differs among Na_v_ isoforms, and Na_v_1.2 and Na_v_1.6 have been found to be “TTX-sensitive Na_v_ channels” (IC_50_ for TTX: ~10 nM), whereas the Na_v_1.5 channel has been characterized as a “TTX-insensitive Na_v_ channel” (IC_50_ for TTX: ~10 µM) and Na_v_1.9 has been characterized as a “TTX-resistant Na_v_ channel” (IC_50_ for TTX: ~100 µM) [31].

Consistent with the previous findings in astrocytes and cancer cells, endothelial cell migration was inhibited by 1 µM TTX and also by the treatment with 100 µM lamotrigine or 500 nM 4,9-anhydro-TTX (Figure 1). Interestingly, the magnitude of the inhibition observed with 1 µM TTX was similar in the presence of higher TTX concentrations (10 µM and 100 µM), which strongly supports the involvement of TTX-sensitive Na_v_ channels in the control of endothelial cell migration. In consideration of the Na_v_ channel isoforms previously detected in endothelial cells and the sensitivity to TTX (low concentrations), lamotrigine and 4,9-anhydro-TTX, the most likely isoforms involved in the response are Na_v_1.2 and Na_v_1.6. In line with this proposal, the expression of these channels in mesenteric endothelial cells was confirmed through immunofluorescence analysis (Figure 1). In addition, the finding that the simultaneous application of lamotrigine with 4,9-anhydro-TTX or each of these blockers with 1 µM TTX did not affect the level of inhibition (Figure 1) suggests that the function of these two Na_v_ channels (Na_v_1.2 and Na_v_1.6) appears to be integrated within the same signaling pathway. It is important to note that the effect of Na_v_ channel blockade was not associated with a change in endothelial cell proliferation (Figure 2), confirming that the reduction attained in wound closure was exclusively the result of the inhibition of endothelial cell migration and not an effect on the potential increase in the number of cells at the migration front.

It is well accepted that TTX is a potent and highly specific blocker of Na_v_ channels. However, the availability of isoform-selective inhibitors for these channels is more disputed, and the pharmacological characteristics of the blockers depend on the Na_v_ channel subtype. In this context, 4,9-anhydro-TTX has been demonstrated to be a highly selective blocker of the Na_v_1.6 channel isoform, with a blocking efficacy ~160 times higher than for other TTX-sensitive Na_v_ channels, including those formed by the isoform Na_v_1.2 [32,33,34]. In contrast, while Na_v_1.2 channels are the main target of lamotrigine [35], this blocker also affects the function of other brain Na_v_ channels, such as Na_v_1.1 and Na_v_1.3, but is mostly ineffective in reducing the responses mediated by Na_v_1.6 channels [36,37]. Likewise, although lamotrigine can inhibit the isoform Na_v_1.5, its potency is ~10 times lower than that for Na_v_1.2 [35]. Therefore, as endothelial cells do not express Nav1.1 or Nav1.3, the inhibition of endothelial cell migration observed in the presence of lamotrigine can be mainly attributed to an effect on Na_v_1.2 function.

The progress of endothelial cell migration depends on the generation of a Ca^2+^ signal that commands the direction of the cell movement. Although the increase in [Ca^2+^]i is initiated by an inositol-1,4,5-trisphosphate (IP_3_)-triggered Ca^2+^ release from the endoplasmic reticulum, the migration mechanism is sustained by the influx of Ca^2+^ from the extracellular space [16,38,39]. The pathway of Ca^2+^ entry is controversial, but, in line with the participation of Ca^2+^ release from the endoplasmic reticulum, it has been proposed that STIM1- and Orai1-mediated signaling, a store-operated Ca^2+^ entry (SOCE) or a pathway associated with the opening of TRPC channels may be involved in the process [40,41]. However, the participation of a voltage-dependent mechanism has also been found to play a role in the control of endothelial cell migration [11], which is in agreement with the inhibition attained in the wound-healing assay (Figure 1) and the strong reduction in the increase in [Ca^2+^]i observed in the migration front after blocking Na_v_ channel function with TTX, lamotrigine or 4,9-anhydro-TTX (Figure 3). Likewise, the formation of tubular-like structures by primary cultures of mesenteric endothelial cells was also clearly reduced in the presence of TTX (Figure 4), supporting the relevance of Na_v_ channels in the development of new vessels during angiogenesis.

Caveolae play a central role in the coordination of the signaling machinery involved in the control of changes in [Ca^2+^]i through the association of Cav-1 with Ca^2+^-related signaling proteins [21,22,42], and Cav-1 has been shown to be a key element for the activation and progress of endothelial cell migration and angiogenesis [43,44]. In agreement with this, our results illustrate that the initiation of endothelial cell migration by scratching the monolayer activated a progressive increase in the spatial association of Na_v_1.2 channels with Cav-1 (Figure 5), which may provide a platform for the functional coupling of these channels with the activation of signaling pathways associated with the control of Ca^2+^ signals. Consistent with this hypothesis, the disruption of the functional organization of caveolae through the treatment with MβCD produces a striking reduction in wound closure and in the increase in [Ca^2+^]i observed at the migration front (Figure 6). Interestingly, in contrast to Na_v_1.2, endothelial cell migration was not paralleled by the association of Na_v_1.6 channels with Cav-1 (Figure 5), suggesting that Na_v_1.2 and Na_v_1.6 channels are different components of the same signaling pathway. In this context, we hypothesize that Na_v_1.6 channels may connect the early migration signal with the activation of Na_v_1.2 channels, as with the functional coupling of these two channels described in neurons, where the initiation of the electrical response was found to be triggered by Na_v_1.6 channels, with the subsequent recruitment of Na_v_1.2 channels [45]. However, the specific role of these two channels in the migrating response of endothelial cells must be confirmed by direct measurements of membrane potential and an electrophysiological analysis of the cells of the migration front.

The progressive location of Na_v_1.2 channels at caveolae may provide the spatial proximity required to regulate the function of important Ca^2+^ signaling proteins, such as the NCX, which is also found in this signaling microdomain [46]. The NCX plays a central role in the regulation of [Ca^2+^]i through the exchange of 3Na^+^ for 1Ca^2+^ across the plasma membrane. Normally, NCX works in forward mode, controlling the increments in [Ca^2+^]i by extruding Ca^2+^ from the cell. However, the activity of the exchanger depends on the electrochemical gradient of the transported ions, and then, when the electrochemical gradient of Ca^2+^ overcomes that of Na^+^ (e.g., an increase in [Na^+^]i), the reverse mode of the exchanger is activated, catalyzing the influx of Ca^2+^ into the cell, as observed during the initial phase of the cardiac action potential [47,48]. Therefore, the activation of Na_v_1.2 channels may lead to an increase in [Na^+^]i in the caveolae microenvironment, which, in turn, may trigger the reverse mode function of NCX and a subsequent increment in [Ca^2+^]i. In agreement with this notion, the Ca^2+^ signaling activated after scratching the monolayer was strongly reduced by the blockade of the NCX reverse mode (Figure 7), highlighting the relevance of the exchanger in endothelial cell migration. In addition, a similar result was attained after disabling NCX function through the reduction of [Na^+^] in the buffer solution (low Na^+^ solution), which also blunts the functional coupling observed between the increase in local [Na^+^]i and the subsequent activation of NCX reverse mode, since it reduces the conducting force for Na_v_1.2 channel-mediated Na^+^ influx. Interestingly, the magnitude of the inhibition observed after disrupting caveolae organization with MβCD and after blocking the NCX reverse mode with SEA0400 or the presence of a low Na^+^ solution was similar (Figure 6 and Figure 7), supporting the pivotal role of caveolae in the molecular organization involved in the functional coupling of Na_v_1.2 with NCX.

In addition to NCX, we recently demonstrated that Cx43-formed hemichannels also contribute to the Ca^2+^ signaling associated with endothelial cell migration through a mechanism coordinated by caveolae [4]. The opening of Cx43 hemichannels was found to be triggered by a NO-dependent pathway, possibly by NO-mediated S-nitrosylation [4], and, in the present study, the analysis of dye uptake showed that the Cx43 hemichannel opening is sensitive to treatment with TTX, lamotrigine or 4,9-anhydro-TTX (Figure 8). It is important to note that the isoform of the NO-synthetizing enzyme in endothelial cells (i.e., the endothelial NO synthase, eNOS) is found in caveolae, in a close spatial association with NCX [47]. Therefore, as NO production relies on an increase in [Ca^2+^]i [49], we hypothesize that the Na_v_ channel-evoked Ca^2+^ influx through the activation of NCX reverse mode is an early signaling event that leads to the opening of Cx43 hemichannels by increasing eNOS activity (Figure 9), which is strongly supported by the reduction in dye uptake observed after the inhibition of NCX reverse mode with SEA0400 (Figure 8). Taken together, these results represents a novel mechanism mediating the Ca^2+^ signaling that commands the progress of endothelial cell migration and supports the pivotal role of caveolae in the molecular organization of the response (Figure 9).

Although this study highlights the relevance of the Na_v_ channels–NCX reverse mode signaling pathway in the progress of endothelial cell migration, several limitations can be recognized. In this context, the activation mechanism of Na_v_ channels in relation with other alternative pathways of Ca^2+^ entry proposed previously, such as STIM1-Orai1 signaling, SOCE and TRPC channels [33,34], must be addressed in future investigations. Likewise, the activation, functional relation and contribution of different Na_v_ channel isoforms must be assessed through a specific electrophysiological analysis of the cells of the migration front. Furthermore, the pharmacological nature of this study has potential side effects that may have influenced the interpretation of the results, and therefore, conclusions must be confirmed through a direct molecular assessment of the participation of each signaling protein involved, not only at cellular level, but also in in vivo studies.

In summary, these results highlight the importance of Na_v_ channels in the control of microvascular network function by endothelial cells. Our data indicate that the development of endothelial cell migration in the wound-healing assay is mediated by the sequential activation of Na_v_1.2 and Na_v_1.6 channels, which triggered a Ca^2+^ signal through the operation of NCX in reversed mode and the subsequent opening of Cx43 hemichannels, potentially through NO-mediated S-nitrosylation of this Cx protein (Figure 9), as has been suggested previously [50]. Caveolae play a central role in the coordination of this signaling pathway. Although Na_v_1.6 channels were not found to be associated with Cav-1, the progress of endothelial cell migration was paralleled by a gradual increase in the spatial association of Na_v_1.2 channels with Cav-1, which is likely a critical process leading to the activation of the NCX reverse mode. Therefore, these results support the relevance of Na_v_ channel function in the control of endothelial cell Ca^2+^ signaling and confirm the relevance of caveolae in the organization of the signaling machinery involved in the control of the Ca^2+^ signal that commands the migrating response of endothelial cells observed during angiogenesis. The relevance of Na_v_ channels in the control of endothelial cell migration and angiogenesis may provide clues to the design of new therapeutic strategies to modulate angiogenesis in physiological or pathological conditions. In this context, it is interesting to note that lamotrigine is an FDA-approved medication, and it may be considered in future investigations as an alternative anti-angiogenic drug for the treatment of cancer.

4. Materials and Methods

Male Sprague Dawley rats (200–220 g) were bred and maintained in the Research Animal Facility of the Pontificia Universidad Católica de Chile. All studies were approved by the Institutional Bioethics Committee (protocol ID 170,823,033), and experiments were conducted according to the Helsinki Declaration. The National Institute of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No 8523, revised 2011) were followed. All efforts were made to minimize the suffering and number of animals used.

4.1. Primary Cultures of Mesenteric Endothelial Cells

Rats were anesthetized with xylazine and ketamine (10 and 90 mg/kg i.p., respectively), and the isolated vascular mesenteric bed was prepared as described by Figueroa et al. [51]. Briefly, the superior mesenteric artery was cannulated and perfused at 2 mL/min with a sterile 3-(N-morpholino)propanesulfonic acid (MOPS)-buffered (pH 7.4) Tyrode solution (in mM: 118 NaCl, 5.4 KCl, 2.5 CaCl_2_, 1.2 KH_2_PO_4_, 1.2 MgSO_4_, 5 MOPS, and 11.1 glucose) containing a mixture of antibiotics and antimycotics (Anti-Anti solution; Thermo Fisher Scientific, Waltham, MA, USA) to wash out blood from the vessels. After cutting the aorta to ensure a fast killing of the rats by exsanguination under deep anesthesia, mesenteries were excised from the intestinal wall to prepare primary cultures of endothelial cells as described by Ashley et al. [52]. Mesenteric vessels were incubated for 1 h in sterile Tyrode solution containing 0.2% collagenase type I (Worthington, NJ, USA) and 0.1% bovine serum albumin (BSA) for 1 h at 37 °C. Then, the solution was diluted with cold M199 medium to inactivate the collagenase. Pelleted cells were resuspended in M199 medium, centrifuged, and resuspended again in M199 medium containing 20% fetal bovine serum (FBS) and 20 µg/mL endothelial cell growth supplement (ECGS) from bovine pituitary. Thus, cells were seeded in 12 mm sterile glass coverslips located on 24-well plates. Three hours later, non-adherent cells were removed, and the remaining adherent endothelial cells were kept at 37 °C in a 5% CO_2_–95% air atmosphere at nearly 100% relative humidity. Experiments were performed using confluent cultures of endothelial cells (~2 days of culture), in which the culture media was replaced by a MOPS-buffered Tyrode saline solution (pH 7.4). Only one microscopy field was analyzed per coverslip, and a maximum of two measurements were performed per cell culture.

4.2. Wound-Healing Assay

A confluent monolayer of endothelial cells, in control conditions or treated for 15 min with TTX (1, 10 and 100 µM), lamotrigine (100 µM), or 4,9-anhydro-TTX (500 nM) to block Na_v_ channels; SEA0400 (1 µM) to inhibit NCX reverse mode; and MβCD (5 mM) to disrupt caveolae organization were scraped using a p200 pipette tip. The monolayer was gently washed with PBS to remove cell debris and then kept in M199 medium supplemented with only 5% FBS. The treatment with TTX, lamotrigine, 4,9-anhydro-TTX, or SEA0400 was maintained during the whole experimental period. Images were captured using a Nikon Eclipse E600 FN1 microscope (Nikon Corporation, Tokyo, Japan), and changes in the size of the cell-free scratched area were evaluated using the ImageJ software (version 1.53, National Institutes of Health, Bethesda, MD, USA). Endothelial cell migration was expressed as a percentage of wound closure (%).

4.3. BrdU Incorporation Assay for Cell Proliferation

Cell proliferation was evaluated by assessing global BrdU incorporation or through direct immunofluorescence analysis of BrdU incorporation. For global BrdU incorporation, the Millipore©’s BrdU Cell Proliferation Assay Kit (Merck millipore, Carlsbad, CA, USA) was used. Endothelial cells were seeded in 96-well plates, and cell cultures with a confluence of 40 or 80% were incubated with 10 µM BrdU in control conditions or in the presence of 1 µM TTX for 12 h. BrdU incorporated into endothelial cell nuclei was recognized using an anti-BrdU peroxidase-conjugated antibody. Cell proliferation was quantified by measuring the absorbance of tetramethylbenzidine (TMB) product at 450 nm. For the direct detection of BrdU incorporation through immunofluorescence analysis, confluent endothelial cells were scraped using a p200 pipette tip and incubated for 12 h with 10 µM BrdU (Thermo Scientific, Rockford, IL, USA) in absence (control) or presence of 1 µM TTX. Then, cells were fixed with 4% paraformaldehyde (PFA); DNA was denatured using a 2 M HCl solution and blocked with 3% BSA in PBS. Coverslips were incubated overnight at 4 °C with a mouse anti-BrdU primary antibody (1:5000, Thermo Scientific, Rockford, IL, USA), and then with an Alexa Fluor 568 anti-mouse secondary antibody (1:1000, Molecular Probes, Eugene, OR, USA) for 1 h at room temperature. The fluorescent signal was examined in both the migration front and monolayer using an Olympus BX41 WI microscope associated with a CCD camera (ProgRes C5; Jenoptik, Jena, Germany). Endothelial cell proliferation was expressed as the proliferation index, which was calculated according to the following relation: BrdU+/Total cells, where BrdU+ is the number of cells that incorporated BrdU into nuclei and Total cells is the number of cells observed in the bright field.

4.4. Changes in Intracellular Ca2+ Levels

Changes in intracellular Ca^2+^ concentration were measured using the fluorescent Ca^2+^ indicator Fluo-4 (Invitrogen, Carlsbad, CA, USA). To upload the cells with Fluo-4, primary cultures of endothelial cells were incubated with 3 µM Fluo-4 acetoxymethyl ester (AM) for 1 h at room temperature. Fluo 4-AM was dissolved in DMSO and, thus, prepared in MOPS-buffered Tyrode saline solution. Ca^2+^ measurements were started 15 min after the initiation of the wound-healing assay by scraping the monolayer. The fluorescent signal was examined using an Olympus BX50 WI microscope coupled with an intensified CCD camera (Retiga Fast 1394; QImaging, Surrey, BC, Canada), and images were acquired every 3 s for 30 s in the wounded area and in the monolayer. The treatment to block Na_v_ channels (1 µM TTX, 100 µM lamotrigine, or 500 nM 4,9-anhydro-TTX), to inhibit NCX reverse mode (1 µM SEA0400) or to disrupt caveolae organization (5 mM MβCD) was started 10 min before scraping the monolayer and was maintained during the whole experimental period. In addition, a group of experiments was performed in the presence of a buffer solution containing 50 mM Na^+^ (Low Na^+^ solution) that was prepared through the equimolar substitution of Na^+^ ions by choline. Changes in [Ca^2+^]i were expressed as the variations in the fluorescence intensity observed in the migration front in relation to the monolayer, Ca^2+^F/Ca^2+^M, where Ca^2+^F is the fluorescence intensity in cells of the migration front and Ca^2+^M is the fluorescence intensity in cells of the monolayer.

4.5. Dye Uptake Assay

The hemichannel opening was analyzed by measuring the ethidium uptake, as described by Figueroa et al. (2013) [53]. The wound-healing assay was initiated in control conditions and during the inhibition of Na_v_ channels with 1 µM TTX, 100 µM lamotrigine, or 500 nM 4,9-anhydro-TTX or the blockade of NCX reverse mode with SEA0400 and Cx43-formed hemichannels with 300 µM TAT-Gap19, a specific Cx43 hemichannel blocking peptide. After an equilibrium period of 10 min, 5 µM ethidium bromide was added, and the recording of ethidium uptake was started 10 min later. The fluorescent signal (excitation 530–550 nm and emission 590 nm) was examined using an Olympus BX50 WI microscope and an intensified CCD camera (Retiga Fast 1394; QImaging, Surrey, BC, Canada). Images were acquired every 30 s for 15 min in the migration front and in the monolayer. Changes in ethidium uptake were expressed as arbitrary units.

4.6. Formation of Tubular Structures

The analysis of tubular-like structure formation by endothelial cells was performed using 12 mm coverslips covered with 100 µL Matrigel^®^ (Corning Incorporated, Corning, NY, USA), as described in the manufacturer’s protocol. Matrigel^®^ solution was added to coverslips located in a 96-well plate and allowed to solidify and polymerize at 37 °C. Then, endothelial cells were seeded on top of the Matrigel, and the tubular-like structure formation was evaluated for 6 and 12 h in control conditions or in the presence of 1 µM TTX. Seven fields per coverslip were examined using a Nikon Eclipse E600 FN1 microscope (Nikon Corporation, Tokyo, Japan), and the results were expressed as the angiogenic index according to the following relation: (Total cells + connected cells)/total cells × (1-non-connected cells), where Total cells is the number of total cells in the field, connected cells is the number of the cells that form tubular structures and non-connected cells is the number of the cells outside of tubular structures.

4.7. Immunofluorescence Analysis

Endothelial cells were fixed with 4% PFA, blocked with 3% BSA in PBS and incubated overnight at 4 °C with a mouse primary antibody directed against Cav-1 (1:100; Thermo Scientific, Rockford, IL, USA), rabbit primary antibodies directed against Na_v_1.2 (1:350; Alomone Laboratories, Jerusalem, Israel) or Na_v_1.6 (1:350; Alomone Laboratories, Jerusalem, Israel), and then with Alexa 488-labeled goat anti-mouse or Alexa 568-labeled goat anti-rabbit secondary antibodies (1:2000; Molecular Probes, Eugene, OR, USA) for 1 h at room temperature, as appropriate. The fluorescence signal was examined using a Nikon spectral C2si confocal microscope (Nikon Instruments, Melville, NY, USA).

4.8. Proximity Ligation Assay

Spatial interaction between Cav-1 with Na_v_1.2 or Na_v_1.6 was performed using the technique for a proximity ligation assay (PLA, Millipore Sigma, Carlsbad, CA, USA). The endothelial cell monolayer was scraped in control conditions and fixed with 4% PFA at room temperature 15 min or 4 h thereafter. Cells were incubated overnight at 4 °C with rabbit primary antibodies directed against either Na_v_1.2 or Na_v_1.6 and with a mouse primary antibody directed against Cav-1. Primary antibodies were then detected with oligonucleotide-conjugated secondary antibodies, as described in the manufacturer’s protocols. If the target proteins are closer than 20 nm, the oligonucleotides provide a template for DNA ligase-mediated joining of additional oligonucleotides to form a circular DNA molecule, which was amplified using hybridizing, fluorophore-labeled oligonucleotides. The fluorescence signal was examined using a Nikon spectral C2si confocal microscope and analyzed using ImageJ software (version 1.53, National Institutes of Health, Bethesda, MD, USA).

4.9. Chemicals

All chemicals of analytical grade were obtained from Merck (Darmstadt, Germany). M199 medium and FBS were purchased from Gibco (New York, NY, USA). Ethidium bromide, BSA, HEPES, MOPS, ECGS and MβCD were purchased from Sigma-Aldrich (Missouri, MO, USA). TTX, lamotrigine and 4,9-anhydro-TTX were obtained from Alomone Laboratories (Jerusalem, Israel) and SEA0400 from Bio-Techne (Minneapolis, MN, USA). SEA0400 was dissolved in DMSO (final DMSO concentration <0.1%) and was then diluted in buffer solution to reach the final working concentration. DMSO did not have an effect per se.

4.10. Statistical Analysis

Results are expressed as mean ± SEM. All values represent data from at least three independent cultures. Comparison between groups was performed using unpaired or paired Student t-tests, one-way ANOVA followed by Bonferroni post hoc test or two-way ANOVA as appropriate.