Mycoplasma free A7r5 (American Type Culture Collection - ATCC; VA, United States) were cultured in DMEM medium (In Vitro, Denmark) supplemented with 10% fetal calf serum, 1% L-glutamine and 0.1% KPS (kanamycin 2 g, penicillin 1 million IU, streptomycin 1 g in 20 ml phosphate-buffered saline [PBS, in mM: NaCl 137, KCl 2.7, Na₂HPO₄ 8.2, KH₂PO₄ 1.8, at pH 7.4)). Confluent A7r5 cells were detached by nonenzymatic cell dissociation solution (Sigma-Aldrich, Denmark), resuspended in DMEM medium as above and pipetted into 200 µl customer-made imaging chambers with cover glass bottom. The cells were stored for 3 h in humidified 37°C/5% CO₂ cell incubator to allow the cells to attach to the bottom prior the experiment imaging experiment.

To measure changes of intracellular Ca²⁺ transients, the cells were loaded with a mixture of Calcium Green-1/AM (3 μM) and Fura Red/AM (6 μM) for 15 min at 37°C/5% CO₂. Both dyes were initially dissolved in DMSO maintaining final concentration ≤0.01%. Calcium Green-1/AM and Fura Red/AM mixture was initially prepared in physiological salt solution (PSS, in mM: 119 NaCl, 3.0 KCl 1.18 KH₂PO₄, 1.17 MgCl₂, 25.0 NaHCO₃, 0.026 EDTA and 5.5 glucose, gassed with 5% CO₂ in air and adjusted to pH 7.4) and this solution then replaced the cell medium in the imaging chamber.

Prior the imaging protocol, the cells were washed five times with warmed to 37°C PSS that was gassed with 5% CO₂ in air. The experiments were performed using an inverted confocal laser scanning microscope (LSM 5 Pa Exciter or LSM780, Zeiss GmbH, Germany) and images were acquired with a water immersion objective (×63, 1.2 W, Zeiss GmbH, Germany).

Both Fura Red and Calcium Green-1 were excited at 488 nm, the emission from dyes was collected above 560 nm and in the interval from 505 to 530 nm, respectively. The elevation in intracellular Ca²⁺ resulted in increased fluorescence intensity of Calcium Green-1 and decreased fluorescence intensity of Fura Red. Therefore, a combination of these two Ca²⁺ indicators was used for ratiometric assessment of intracellular Ca²⁺ changes, as described previously (Matchkov et al., 2007).

PSS in the imaging chamber was kept at 37°C by thermostatically controlled chamber holder and constantly gassed with 5% CO₂ in air. The experimental protocol consisted of 5 min baseline recordings, and the pharmacological intervention by replacement of the solution in the chamber to another modified solution and followed by incubation for 15 min prior next 5 min imaging session. Up to three pharmacological interventions per single experimental protocol were done.

Changes in intracellular Ca²⁺ were analyzed using either the Zeiss LSM Image Examiner program or ZEN software (Zeiss GmbH, Germany). The local Ca²⁺ events were analyzed as a profile of fluorescence ratio (Calcium Green-1 over Fura Red) either at cell perimeter or in the center of the cell, and plotted for the profile distance, as described previously (Matchkov et al., 2007). Ca²⁺ transients were counted if elevation of intracellular Ca²⁺ was 5 or more times above averaged baseline. The number of Ca²⁺ transients was normalized to the length of profile the fluorescence was measured. The analysis was done blind with custom-made Makro in Excel Spreadsheet Software (Microsoft corp., NM, United States).

Transfection of A7r5 cells was done as described previously (Matchkov et al., 2008). The siRNA directed against the α2 isoform (Ambion; the sense sequence: 5′-GAG​AAC​ATC​TCC​GTG​TCA​tt-3′) and another control, nonrelated siRNA directed against enhanced green fluorescent protein (Ambion; the sense sequence: 5′-CCA​CUA​CCU​GAG​CAC​CCA​Gtt-3′) were used as previously described (Matchkov et al., 2012). When A7r5 cells were 80% confluent, antibiotics were omitted from DMEM medium for 2 h. The cells were then transfected by TransIT TKO transfection kit (Mirus Bio Co., United States) following the protocol given by the manufacturer. Each culture dish was transfected with 5 μl TKO and 25 nmol/L siRNA in antibiotic-free DMEM medium for 8 h, then the cells were washed out and the medium was replaced back to standard DMEM. The cells were used 72 h later for functional and expressional analyses.

Confluent A7r5 cells were washed in the culture dishes (Falcon, Becton Dickson, Denmark) 3 times in ice-cold PBS solution and collected by scrubbing. The cell suspension was centrifuged at 3,000-g and collected pellet was homogenized in lysis buffer (in mM: Tris‐HCl 10, sucrose 250, EDTA 1, EGTA 1; Triton X‐100 2%, pH 7.4; 1 tablet protease inhibitor per 10 ml and 0.01 mmol−1 of Halt’s phosphatase inhibitor cocktail) and 2x trisglycine sodium dodecyl sulphate (SDS) sample buffer (Invitrogen, MA, United States) with 1 mol/L DTT. The homogenates were ultrasonicated for 45 s and centrifuged at 10,000-g for 10 min at 4°C. The supernatants were collected, and the protein contents were determined using a bicinchoninic acid protein assay reagent kit (Pierce, United States). The sample was adjusted then with 1 M DTT and 2x Tris-glycine SDS sample buffer (Invitrogen, MA, United States) with an approximate ratio of 10:3:3.

Proteins (10 µg) were separated on 14% Tris-glycine gels and electrotransferred onto nitrocellulose membranes, which were then blocked by incubation in 5% nonfat dry milk in PBS with 0.5% vol/vol Tween 20 (PBS-T). The membrane was divided at approximately 64 kDa and the upper part of the membrane was incubated with primary α2 isoform Na,K-ATPase antibody (1:2,000; Millipore, United States [Catalog no. AB9094]) overnight at 5°C in PBS-T. The lower part of the membrane was incubated with pan-actin antibody (1:2,000, Cell Signaling Technology Inc., United States [Catalog no.4968]). Next day, the membranes were washed and incubated with horseradish peroxidase-conjugated secondary antibody (Dako, Denmark) for 1 h and bound antibody was detected by an ECL chemiluminescence kit (Amersham, United Kingdom).

Total protein lysates from rat mesenteric arteries were homogenized in 200 μl of RIPA buffer (in mM): 50 Tris pH 8.5, 150 NaCl, 1% SDS, 1% Nonidet P40, 0.5% Sodium deoxycholate and protease inhibitor cocktail (Roche, Switzerland) for 10 min at 4°C. The supernatant was collected after centrifugation at 11,000-g for 10 min at 4°C followed by protein quantitation using a bicinchoninic acid (BCA) Protein Assay kit (Thermo Fisher Scientific, MA, United States). 20 µg of protein were loaded on SDS-PAGE gels (4%–12% bis–tris; Invitrogen, MA, United States), subjected to electrophoresis, and then transferred onto a polyvinylidene fluoride (PVDF) membrane (Immobilon®-FL, Sigma-Aldrich, Denmark). The membrane was probed with antibodies against Na,K-ATPase α2 isoform (1:2,000; Millipore, United States [Catalog no. AB9094]) and GAPDH (1:10,000; Abcam, United Kingdom [Catalog no.ab181602]). Fluorescently conjugated secondary antibodies (1:10,000; Li-Cor Biosciences [Catalog no.926-32211 and 926-32210]) were used to visualize protein bands. The membrane was imaged and analysed on the Odyssey Infrared Imaging System (Li-Cor Biosciences; version 5.2.5).

The protein expression was semi-quantified as a ratio of intensity of bands for the α2 isoform Na,K-ATPase and either pan-actin or GAPDH proteins. Analysis was done with the ImageJ (National Institutes of Health, MD, United States). The normalized protein ratio is shown in percentage of mean ratio in the control group.

A7r5 cells were lysed in lysis buffer as above. Ten µg of protein in Laemmli sample buffer (Bio‐Rad, Hercules, CA) was loaded onto 4–20% precast polyacrylamide stain‐free gels (Criterion TGX Stain‐free precast gel; Bio‐Rad). Total protein load was detected on the stain‐free gels using UV light in imaging system (c600; Azur Biosystems, Dublin, CA). The proteins were electrotransferred onto nitrocellulose membranes that were blocked in 3% bovine serum albumin (BSA) in Tris‐buffered solution (TBS, mmol L−1: 10 Tris‐HCl, 100 NaCl; pH 7.6) with 0.5% Tween-20 (TBST) to identify the phosphorylated Src or in 0.3% iBlock in TBS to identify total Src kinase. Membranes were incubated overnight at 4°C with primary antibodies against phosphorylated pY416 Src (1:500; Cell Signaling Technology Inc., MA, United States [Catalog no. CST-6943]) or total Src (1:200; Santa Cruz Biotechnology, TX, United States [Catalog no. sc-8056]). Next day, the membranes were incubated with horseradish peroxidase (HRP) conjugated secondary antibody (1:2000; Dako, Denmark) for 2 h at room temperature, and bound antibodies were detected by an enhanced chemiluminescence kit (ECL, Amersham, United Kingdom). Analysis was done with ImageJ (National Institutes of Health, MD, United States). Detected protein was normalized as a ratio to total protein load measured in the membrane for the same sample and expressed either as percentage of mean ratio in the control group or as a relative level of phosphorylated Src form over the total expression level of Src kinase.

PLA technique was used to determine the co-localization of the α2 isoform Na,K-ATPase with Src kinase in cultured A7r5 cells and freshly isolated rat mesenteric artery myocytes. Duolink in situ PLA detection kit 563 (Sigma-Aldrich, Denmark) was used in accordance with the manufacturer’s instructions. Cell isolation from third-order rat mesenteric arteries was conducted as described previously (Jepps et al., 2015). Briefly, cells were fixed in 4% paraformaldehyde in PBS for 20 min and permeabilized in PBS containing 0.1% of Triton X-100 for 5 min. Cells were incubated for 1 h at 37°C in Duolink blocking solution to avoid any unspecific binding. The primary antibodies used against the α2 isoform Na,K-ATPase (1:100; Millipore, United States [Catalog no. 07-674]) and Src (1:100; Santa Cruz Biotechnology, TX, United States [Catalog no. sc-8056]) were diluted in Duolink blocking solution and incubated overnight at 4°C. Combinations of secondary anti-rabbit or anti-mouse antibodies of PLA PLUS and MINUS probes were used followed by hybridization, ligation and amplification steps. Red punctae representing proteins that are located within 40 nm of each other were visualized and quantified using a Zeiss LSM710 upright laser scanning confocal microscope.

All animal experiments were performed in accordance with Directive 2010/63EU on the protection of animals used for scientific purposes and approved by the national ethics committee, Denmark. Male Wistar rats were purchased from Janvier Labs (France), group-housed in clear plastic containers and underwent at least 1 week of habituation before use. All experiments with dissected mesenteric arteries were performed using 12- to 15-week-old male Wistar rats.

Calcium Green-1/AM and Fura Red/AM were obtained from Invitrogen (MA, United States). Nocodazole was obtained from Tocris (United Kingdom). The pNaKtide peptide was obtained from HD Biosciences (China) and stock solution was prepared in water (10 mM) and stored at −20°C. All other chemicals were purchased from Sigma-Aldrich (Denmark). Stock solutions of nocodazole, genistein and PP2 were prepared in DMSO (10 mM) and stored at −20°C. Ouabain stock solution was prepared on the day of experiment (minimum 2 h prior to application) in a concentration of mM in water. Drugs were applied a minimum 15 min prior to measurements/interventions.

All statistical analysis was performed using GraphPad Prism 8 (GraphPad Software Inc., San Diego, CA, United States). Different statistical tests were performed throughout the present study, as appropriate, and are defined in the Results section for each data set analysed. The data were subjected to a normality test (Shapiro-Wilk test) that defines the use of either parametric or nonparametric statistical analysis. All data are presented as means ± standard error of the mean (SEM). A probability (P) level of <0.05 was considered significant, and n refers to number of experiments in case of cell culture. The number of culture dishes used at the day of experiment is also indicated. When mesenteric arteries were studied, n refers to number of cells and the number of rats is also provided.

In A7r5 cells, superfusion with 10 µM ouabain induced spatially restricted [Ca²⁺]_i transients—“[Ca²⁺]_i flashes” (Figure 1A). Similar to a previous report (Matchkov et al., 2007), these Ca²⁺ flashes were only detected on the cell periphery (Figure 1B), while no ouabain-induced Ca²⁺ fluctuations were seen in the center of the cells (Figure 1C). Transfection of A7r5 cell with siRNA directed against the α₂ isoform suppressed the protein expression of the α₂ isoform Na,K-ATPase compared to cells transfected with non-targeted siRNA (Figure 1D&E). The cells transfected with non-targeted siRNA were still able to increase submembrane Ca²⁺ flashes upon administration of ouabain (Figure 1F). In contrast, A7r5 cells transfected with siRNA targeted against the Na,K-ATPase α₂ isoform did not show any ouabain-induced Ca²⁺ flashes (Figure 1G). The effect of ouabain was concentration dependent. Ca²⁺ flashes were induced by ouabain at 10⁻⁶ and 10⁻⁵ M while no significant effect of 10⁻⁷ M was seen (Figure 1H).

We have tested the possibility that Ca²⁺ flashes originated from changes in activity of the Na⁺,Ca²⁺-exchanger, because of inhibition of Na⁺ extrusion by the Na,K-ATPase with ouabain (Blaustein and Lederer, 1999). When the Na,K-ATPase ion transport was modulated by changes in the extracellular concentrations of transported K⁺ and Na⁺ cations, the ability of ouabain to elicit the intracellular Ca²⁺ flashes was still preserved (Figure 2). Omission of extracellular K⁺ suppresses the ion pumping by Na,K-ATPase (Holmgren et al., 2000). However, ouabain-induced Ca²⁺ flashes were not affected (Figure 2A). Reduction of extracellular Na⁺ to 50 mM, which is expected to potentiate the Na,K-ATPase ion transport, reduced the potency of ouabain to induce the intracellular Ca²⁺ flashes but did not prevent them (Figure 2B).

Micromolar concentrations of ouabain are known to induce auto-phosphorylation and activation of non-receptor tyrosine kinase, Src (Hangaard et al., 2017). We tested the involvement of this pathway in generation of ouabain-induced Ca²⁺ flashes. An unspecific tyrosine kinase inhibitor, genistein prevented ouabain-induced Ca²⁺ flashes in A7r5 cells (Figure 3A). Ouabain was also unable to induce Ca²⁺ flashes in the presence of relatively selective Src family kinase inhibitor, PP2 (Bain et al., 2003) (Figure 3B). Furthermore, a specific inhibitor of the Na,K-ATPase-dependent Src kinase activation with pNaKtide (Li et al., 2009) also prevented the Ca²⁺ flashes induced by ouabain (Figure 3C).

An interaction between the Na,K-ATPase and tubulin, the main component of the microtubules, has previously been reported (Casale et al., 2003; Zampar et al., 2009). As the microtubule network is considered to play an important role in signal transduction (Gundersen and Cook, 1999), we hypothesized that it might be also involved in the generation of ouabain-induced Ca²⁺ flashes and tested this hypothesis in A7r5 cells (Figure 4A). We found a close proximity (< 40 nm) of the α2 isoform Na,K-ATPase and Src kinase (Figure 4B). Pre-incubation with nocodazole disrupted microtubules (Lindman et al., 2018) and reduced the number of co-localizations between the α2 isoform Na,K-ATPase and Src kinase compared to control, while ouabain treatment was without any effect (Figures 4A,B). When analysis was done only at the cell edge, the cells pre-treated with nocodazole had reduced co-localization the α2 isoform Na,K-ATPase and Src kinase in comparison with control cells (Figure 4C). This was not associated with changes in the α2 isoform Na,K-ATPase expression, which was similar in A7r5 cells under control conditions and after pre-treatment with nocodazole (Figures 4D,E).

To address the functional importance of an intact microtubule network for generation of ouabain-induced Ca²⁺ flashes, A7r5 cells were pre-incubated with nocodazole prior to stimulation with ouabain. Ouabain failed to induce Ca²⁺ flashes after disruption of microtubules with nocodazole (Figure 4F).

Experiments on freshly isolated rat mesenteric artery myocytes (Figure 5A) supported our findings from A7r5 cells (Figure 4). Nocodazole significantly decreased the number of co-localizations between the α2 isoform Na,K-ATPase and Src kinase in freshly isolated vascular smooth muscle cells (Figure 5B), but ouabain did not alter the number of interactions (Figure 5B). Nocodazole also decreased the co-localization at the cell edge (Figure 5C). Western blot analysis of rat mesenteric artery lysates showed that nocodazole did not change the total protein expression of the α2 isoform Na,K-ATPase (Figures 5D,E). Altogether, these results suggest that an intact microtubule network is required to maintain the α2 isoform Na,K-ATPase and Src kinase together allowing Src activation upon binding ouabain to the Na,K-ATPase.

Incubation with either nocodazole, ouabain, vehicle (DMSO), or with a combination of nocodazole and ouabain did not change the total Src content of A7r5 cells (Figures 6A,C). However, nocodazole reduced relative Src phosphorylation (Figures 6B,C). Moreover, although ouabain administration elevated relative Src phosphorylation, this was not seen in the presence of nocodazole (Figure 6C).

In this study, we show that Ca²⁺ flashes can be induced with 1 µM and higher concentrations of ouabain and suggest a primary role of Src activation for Ca²⁺ flash generation, although some modulatory role of membrane potential and trans-membrane ion translocation cannot be excluded completely. Accordingly, it has been suggested previously that ouabain mediates, at least in part, its pro-contractile and pro-hypertensive effects via ion transport independent signaling from the Na,K-ATPase Src activation (Yuan et al., 1993; Song et al., 2014).

Some controversy regarding the involvement of α1 and α2 isoforms of the Na,K-ATPase has been reported previously (Matchkov et al., 2012; Xie et al., 2015; Bouzinova et al., 2018; Staehr et al., 2018; Yu et al., 2018; Kutz et al., 2021). Although a difference in the studied tissues can contribute to the variability, previous studies demonstrated clearly that Src kinase has highest affinity to specific protein-domain on the α1 isoform (Tian et al., 2006; Li et al., 2009; Xie et al., 2015; Yu et al., 2018). However, we found in this study that a specific knock-down of the α2 isoform Na,K-ATPase abolished the ouabain-induced Ca²⁺ flashes in A7r5 cells. This cannot be because of associated changes in the expression of α1 isoform, as this siRNA knockdown protocol did not affect the expression of α1 isoform Na,K-ATPase, as we reported previously (Matchkov et al., 2012; Bouzinova et al., 2018). It has been shown previously that Src kinase binds to the Na,K-ATPase α1 isoform in E1 conformation (Ye et al., 2013). Changes in the activity of the α2 isoform Na,K-ATPase can modify localized concentration of Na⁺, which is important for E1/E2 conformation equilibrium of the α1 isoform (Jorgensen et al., 2003). However, an implication of both isoforms of the Na,K-ATPase-dependent in Src activation and Ca²⁺ flash generation is also possible. Although further analyses are needed, our results from proximity ligation assay demonstrated that the α2 isoform and Src kinase are localized in a proximity within 40 nm from each other. This is the first demonstration of an interaction between the α2 isoform and Src kinase in the cardiovascular system, although another study reported that both α1 and α2 isoforms interact with Src kinase in skeletal muscle cells (Kotova et al., 2006).

In this study, we inhibited the Na,K-ATPase-dependent Ca²⁺ flashes with two structurally unrelated tyrosine kinase inhibitors as well as with a specific inhibitor of ouabain-induced Src phosphorylation, pNaKtide (Li et al., 2009). This peptide was synthesized and validated for the specific inhibition of Src signaling from the α1 isoform Na,K-ATPase suggesting that this isoform is implicated in the signaling.

We found in this study that disruption of the microtubule network with nocodazole prevented the ouabain-induced Ca²⁺ flashes. The importance of the microtubule network for the enzymatic activity and signaling of the Na,K-ATPase has been suggested previously. That is, the interaction between the Na,K-ATPase and acetylated tubulin was shown in the rat brain (Alonso et al., 1998; Casale et al., 2003; Arce et al., 2008). Tubulin is the main component of microtubules and when acetylated, it can inhibit the enzymatic activity of Na,K-ATPase. Moreover, disruption of the microtubule network with nocodazole dissociated the complex between tubulin and Na,K-ATPase followed by increased activity in the Na,K-ATPase (Casale et al., 2005). Tubulin is able to interact in a direct manner through cytoplasmic domain 5 of the Na,K-ATPase, suggesting a role for the pump as a microtubule-plasma membrane anchorage site (Zampar et al., 2009; Santander et al., 2019). Mechanical response, i.e., membrane deformation and tension in response to osmotic stress, was suggested to regulate this interaction (Rivelli et al., 2012; Nigra et al., 2016). Moreover, the importance of the microtubule network for Src activation and trafficking was highlighted in different cell types (Abu-Amer et al., 1997; Yamada et al., 2000; Suter et al., 2004). Microtubules were also shown to play a key role in the regulation of steady-state Src activation at the plasma membrane (Wu et al., 2008).

In this study, we show a close proximity localization of the Na,K-ATPase α2 isoform and Src kinase, which was disturbed by nocodazole. Although we have not measured the expression of Src kinase in the presence of nocodazole, the unchanged expression of the Na,K-ATPase suggests that the reduced PLA signal after short incubation with nocodazole is a result of reduced interaction of these proteins. Accordingly, nocodazole decreased Src phosphorylation at resting conditions, and prevented its phosphorylation upon ouabain administration. It has been reported previously that the microtubule network is involved in Src vesicle trafficking, where nocodazole disrupted the trajectory of Src vesicles but did not inhibit their movement (Arnette et al., 2016). Surprisingly, administration of ouabain did not change proximity of the Na,K-ATPase α2 isoform and Src kinase. This may suggest that the proximity of these two proteins is important for initiation of intracellular signaling and Ca²⁺ flash generation but does not imply large intracellular movement of Src kinase. Although the importance of the microtubules for cell division, differentiation, motility, trafficking and intracellular signaling is accepted, very little is known about the role of microtubule network in vascular smooth muscle cells (Brozovich et al., 2016; Jepps, 2021). The importance of intact microtubule network for interaction between the Na,K-ATPase α2 isoform and Src kinase was validated in both cultured A7r5 cells and freshly isolated vascular smooth muscle cells suggesting the importance of this signaling downstream from the Na,K-ATPase for the regulation of vascular function in vivo. Based on this and previous studies on the Kv7.4 channel trafficking (Lindman et al., 2018; van der Horst et al., 2021), we suggest that the intact microtubule network in vascular smooth muscle cells is a key element to maintain vascular function.

The transient changes in intracellular Ca²⁺ induced by micromolar ouabain were shown to be spatially restricted to submembrane regions, whereas no fluctuation in Ca²⁺ was seen in the center of the cells. In accordance with previous reports (Mulvany et al., 1984; Aalkjaer and Mulvany, 1985; Matchkov et al., 2007; Matchkov et al., 2012; Hangaard et al., 2017), ouabain did not induce global changes in intracellular Ca²⁺ concentration but elicited spatially restricted Ca²⁺ transients, i.e., Ca²⁺ flashes. Spatial and temporal changes in intracellular Ca²⁺ are critical for vascular smooth muscle cells where Ca²⁺ has been shown to act as a second messenger and initiate contraction (Wamhoff et al., 2006). Moreover, intracellular Ca²⁺ transients were shown to modulate a phenotype of vascular smooth muscle cells via regulation of transcription factors (Kudryavtseva et al., 2013). Although, functional significance of the ouabain-induced Ca²⁺ flashes was beyond the scope of this study (Glavind-Kristensen et al., 2004; Matchkov et al., 2007; Matchkov et al., 2012; Hangaard et al., 2015; Hangaard et al., 2017; Bouzinova et al., 2018; Staehr et al., 2018; Kutz et al., 2021), we can suggest based on previous reports that this spatial and temporal Ca²⁺ signaling could play an important role in vascular remodeling and contraction, intracellular signaling and smooth muscle cell metabolism. This study adds further complexity in the signaling by showing the key role of microtubule network in the Na,K-ATPase-dependent intracellular signaling mediated by Src kinase.