Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Spontaneous, Pro-Arrhythmic Calcium Signals Disrupt Electrical Pacing in Mouse Pulmonary Vein Sleeve Cells

  • Katja Rietdorf ,

    Affiliations Department of Microbiology and Physiological Systems, University of Massachusetts, Medical School, Worcester, Massachusetts, United States of America, Department of Life, Health and Chemical Science, The Open University, Milton Keynes, United Kingdom

  • Martin D. Bootman,

    Affiliations Department of Life, Health and Chemical Science, The Open University, Milton Keynes, United Kingdom, Signalling Programme, The Babraham Institute, Cambridge, United Kingdom

  • Michael J. Sanderson

    Affiliation Department of Microbiology and Physiological Systems, University of Massachusetts, Medical School, Worcester, Massachusetts, United States of America

Spontaneous, Pro-Arrhythmic Calcium Signals Disrupt Electrical Pacing in Mouse Pulmonary Vein Sleeve Cells

  • Katja Rietdorf, 
  • Martin D. Bootman, 
  • Michael J. Sanderson


The pulmonary vein, which returns oxygenated blood to the left atrium, is ensheathed by a population of unique, myocyte-like cells called pulmonary vein sleeve cells (PVCs). These cells autonomously generate action potentials that propagate into the left atrial chamber and cause arrhythmias resulting in atrial fibrillation; the most common, often sustained, form of cardiac arrhythmia. In mice, PVCs extend along the pulmonary vein into the lungs, and are accessible in a lung slice preparation. We exploited this model to study how aberrant Ca2+ signaling alters the ability of PVC networks to follow electrical pacing. Cellular responses were investigated using real-time 2-photon imaging of lung slices loaded with a Ca2+-sensitive fluorescent indicator (Ca2+ measurements) and phase contrast microscopy (contraction measurements). PVCs displayed global Ca2+ signals and coordinated contraction in response to electrical field stimulation (EFS). The effects of EFS relied on both Ca2+ influx and Ca2+ release, and could be inhibited by nifedipine, ryanodine or caffeine. Moreover, PVCs had a high propensity to show spontaneous Ca2+ signals that arose via stochastic activation of ryanodine receptors (RyRs). The ability of electrical pacing to entrain Ca2+ signals and contractile responses was dramatically influenced by inherent spontaneous Ca2+ activity. In PVCs with relatively low spontaneous Ca2+ activity (<1 Hz), entrainment with electrical pacing was good. However, in PVCs with higher frequencies of spontaneous Ca2+ activity (>1.5 Hz), electrical pacing was less effective; PVCs became unpaced, only partially-paced or displayed alternans. Because spontaneous Ca2+ activity varied between cells, neighboring PVCs often had different responses to electrical pacing. Our data indicate that the ability of PVCs to respond to electrical stimulation depends on their intrinsic Ca2+ cycling properties. Heterogeneous spontaneous Ca2+ activity arising from stochastic RyR opening can disengage them from sinus rhythm and lead to autonomous, pro-arrhythmic activity.


Throughout a typical human lifetime, a coordinated ‘cardiac cycle’ of atrial and ventricular contraction is repeated over a billion times [1]. The cardiac cycle is initiated by the sino-atrial node (SA node) located within the right atrial wall. The SA node generates action potentials (APs) that sweep over the atrial and ventricular chambers to cause them to contract and pump blood [2]. This contraction is mediated by Ca2+ increases within cardiac myocytes via the well-known process of ‘excitation-contraction coupling’ (EC-coupling). Essentially, membrane depolarisation opens voltage-operated Ca2+ channels (VOCCs) to allow Ca2+ influx into the ‘dyadic’ cleft between the sarcolemma and the sarcoplasmic reticulum (SR; the myocyte Ca2+ store). This Ca2+ signal is amplified by Ca2+-induced Ca2+ release via ryanodine receptors (RyRs) on the SR, thus causing a global Ca2+ signal that triggers actin-myosin filament interaction and myocyte contraction [3]. After each AP, myocytes become refractory to further electrical stimulation for tens to hundreds of milliseconds, and during this time Ca2+ is returned to diastolic levels by Ca2+-ATPases on the sarcolemma and SR, and Na+/Ca2+ exchangers on the sarcolemma [4]. In humans, the normal sinus rhythm is ∼60 beats per minute, but during atrial fibrillation (AF) the atrial chambers can display activity in excess of 300 beats per minute due to aberrant electrical signals [5]. Because the ventricular chambers are largely responsible for pumping blood, AF is not immediately life threatening, but a loss of atrial function can lead to fainting and chest pain. Moreover, blood clotting resulting from stagnant or turbulent blood in the atrial chambers greatly increases the risk of stroke [6].

It is believed that increased ‘automaticity’ (spontaneous depolarisation of myocytes), acute ‘triggered activity’ (spontaneous electrical events following recovery from an action potential) or ‘re-entry circuits’ (return of an electrical impulse to cardiac cells following a refractory period) contribute to AF. The underlying causes of all these pro-arrhythmic conditions are not fully understood, but substantial evidence has implicated spurious Ca2+ signals as a likely cause [6][8]. Spontaneous Ca2+ signals occurring during the recovery from a previous AP, or during the quiescent diastolic phase, can depolarise the sarcolemma and potentially trigger an ectopic AP or alter the refractoriness of myocytes relative to their neighbours [9].

While spontaneous Ca2+ signaling and electrical events can arise within the atrial chambers themselves [10], a clinically-recognised source of pro-arrhythmic signals are pulmonary vein sleeve cells (PVCs) that form sheaths surrounding the large pulmonary veins [11]. PVCs are present in all mammalian cardiovascular systems and utilise the same EC-coupling machinery as atrial and ventricular myocytes (described above). Even though PVCs are developmentally and anatomically distinct from atrial myocytes, both cell types are in electrical continuity [12]. Thus, PVCs are entrained by sinus rhythm because APs arising in the SA node would sweep across the atrial chambers, propagate out of the heart into the PVCs and cause them to contract. However, electrical mapping has demonstrated that ectopic activity can arise within PVCs [13], and propagate into the left atrial chamber. Moreover, ablation procedures that electrically isolate pulmonary veins from the posterior wall of the left atrium (the border of the two tissues) are highly successful in treating acute and sustained AF [14], [15], supporting the notion of PVC-initiated arrhythmias. The conditions leading to the generation of arrhythmic pacemaking sites within PVC sheaths is not understood, but is likely to involve the development of spontaneous Ca2+ signals [16].

The extent of the PVC sheath varies between animal species. Here, we took advantage of the situation in mice, where the PVCs extend from the left atrium to the pulmonary veins within the lung for several branching generations [17]. In previous work, we demonstrated the utility of lung slices for studying airway smooth muscle physiology [18]; lung slices provide an intact, multi-cellular preparation that retains in situ organizational and physiological characteristics of the lung, and are viable for many days. Because the PVCs are disconnected from the left atrial chamber, they no longer receive APs arising from the SA node. This electrical isolation allows the spontaneous Ca2+ signaling capacity of the PVCs to be evident without a background of SA node-evoked events. As required, the PVCs within a lung slice can be activated by application of electric field stimulation (EFS). In the present study, we characterised spontaneous and EFS-induced Ca2+ transients in PVCs to determine whether these two processes use similar signaling mechanisms. Moreover, we explored the hypothesis that the inherent spontaneous Ca2+ signals within PVCs could corrupt their electrical entrainment, and thereby lead to pro-arrhythmic outcomes.


Preparation of Lung Slices

For the preparation of lung slices [18] 8–12 week-old female BALB/C mice were killed by intraperitoneal Nembutal (pentobarbital sodium) injection, as approved by the Institutional Animal Care and Use Committee of the University of Massachusetts Medical School. After opening the chest cavity, the trachea was cannulated and lungs were inflated by injecting ∼1.0 ml of low melting point agarose (37°C, 1.8% in sHBSS; Life Technologies, Carlsbad, CA). Agarose was flushed out of the airways and into the alveoli by injection of 0.3 ml of air and the agarose was gelled by applying cold sHBSS to the lungs. The stiffened lungs were cut into ∼180 µm thick lung slices using a VF-300 microtome (Precisionary Instruments Inc., Greenville, NC). The presence of PVCs was confirmed by visual inspection at low magnification using the following criteria: the vein is not located next to an airway, spontaneous fibrillations and/or striations are visible in the cells surrounding the vein. Slices were kept in Dulbecco’s Modified Eagle Medium supplemented with antibiotics and antimycotics (Life Technologies) and 10% fetal bovine serum and kept in a humidified tissue culture incubator at 37°C with 10% CO2. Slices were used within 48 h of preparation, apart from experiments investigating the viability of PVCs. For imaging experiments, lung slices were mounted between 2 cover-glasses separated with silicon grease on a custom-built Perspex support, which allowed for a continuous superfusion of the lung slices with experimental solutions.

Solutions and Compounds

Experiments were performed in sHBSS (KCl 5.3 mM, KH2PO4 0.4 mM, NaCl 137.9 mM, NaH2PO4 0.3 mM, NaHCO34.2 mM, D-Glucose 5.6 mM, MgCl2 0.5 mM, MgSO4 0.4 mM, CaCl2 1.3 mM, HEPES 20 mM), pH 7.4. Ca2+-free HBSS was made by supplementing nominally Ca2+-free sHBSS with 100 µM EGTA. Ryanodine and caged InsP3/PM (Propionoxymethylester) were purchased from Enzo Life Sciences (Farmingdale, NY). Unless otherwise mentioned, all other reagents were purchased from Sigma (St. Louis, MO).

Imaging Experiments

Imaging experiments were performed on custom-built microscopes [19]. Video Savant software (IO Industries, CA) was used for recording images and controlling the timing of solution changes. Contractions were measured using a Nikon Diaphot 300 microscope with a 10x objective (N.A. 0.3) at 15, 30 or 120 images per second. Ca2+ imaging was performed on a custom-built 2-photon system (excitation wavelength of 800 nm), based on an Olympus IX71 microscope, using an oil immersion UApo 340 40x (N.A. 1.3) or PlanApo60x (N.A. 1.42) objective. Lung slices were loaded in the dark with 20 µM Oregon Green BAPTA-1 AM (Acetoxymethyl ester; Life Technologies) in presence of 0.1% Pluronic F127 (Life Technologies) and 200 µM sulfobromophthalein in sHBSS at 30°C for 1 hour, followed by a de-esterification period of at least 30 min at 30°C in the presence of sulfobromophthalein. Image acquisition was performed using Video Savant software at 30 images per second. Throughout the experiment, lung slices were perfused with sHBSS using a gravity-fed system. The solution exchange was controlled by Video Savant software. The 2-photon excitation wavelength was set by tuning the laser. The excitation light was reflected using a FF665-Di02-25×36 dichroic (Semrock, Rochester, NY), and the emitted fluorescence light was filtered using a short-pass barrier filter FF01-680/SP-25 (Semrock).

A custom-built confocal microscope, based on a Nikon Diaphot 200, with either a 40x Fluor (N.A. 1.3) or 63x Planapo (N.A. 1.4) oil-immersion objective was used for imaging immunofluorescence and di-8-ANNEPS (488 nm excitation for both). Specific staining of the PVC sarcolemma was achieved by incubation of lung slices with 25 µM di-8-ANNEPS (Life Technologies, Carlsbad, CA) for 30 min. The excitation (488 nm laser line) was reflected using a LPD01-488RU-25×36 dichroic (Semrock), and emitted fluorescence light was filtered using a long-pass LP02-488RU-25 filter (Semrock).

To avoid deterioration of the slices, or temperature gradients associated with superfusing cells, the experimental data shown in this study were obtained at room temperature (20°C). However, electrical pacing, contraction and spontaneous Ca2+ transients were evident at 20°C and 37°C (c.f. fig. S1 and Results section). The average frequencies of spontaneous Ca2+ transient activity were not significantly different at either temperature (2.3±0.2 and 2.5±0.2 Hz, 20°C and 37°C, respectively; see fig. S1C). Moreover, the effects of pharmacological reagents and experimental maneuvers (caffeine, ryanodine, KCl) were similar at both temperatures. A consequence of studying the slices at room temperature was that the electrical pacing rate was reduced from ∼8 Hz in vivo to ≤5 Hz. We speculate that the lack of effect of temperature on the frequency of spontaneous Ca2+ signals relates to the counterbalancing processes of Ca2+ release and reuptake being equally affected.


Immunostaining was performed in paraformaldehyde-fixed lung slices as previously described [20], but with the following modifications: 45 min fixation, 2 hour blocking in 10% BSA, 2 hour incubation of primary antibodies (RyR2 and RyR3, 1∶50 dilution in 0.1% triton containing PBS with 2% BSA) at 20°C. Alexa-Fluor 488 goat-anti-rabbit secondary antibody (Life Technologies) was incubated for 1 hour in the dark at 20°C (1∶500 dilution). Slices were mounted in Prolong Gold Mounting Medium (Life Technologies). Image analysis was performed with ImageJ [21].

Electric Field Stimulation (EFS)

EFS (40 V, 10 ms duration) was applied using a Grass Instruments SD9 stimulator and platinum electrodes (0.005′ wire thickness, A–M systems), placed at the opposite ends of the imaging chamber. The EFS conditions were empirically determined by finding the minimal voltage condition that would reliably pace PVCs with no appreciable run-down of the responses.


Statistical analysis was performed in GraphPad Prism 6 (GraphPad software Inc., La Jolla, CA, USA), using a Mann-Whitney U-Test (2 datasets) or a Kruskall-Wallis Test followed by a Dunn’s post-hoc analysis (>2 datasets) to test for statistically significant differences by comparing the initial values the first 30 s in sHBSS at the start of the experiment to the values at the indicated time periods. P values <0.05 were considered significant, *P<0.05, **P<0.01, ***P<0.001. Data are shown as mean ± s.e.m.


PVCs Respond to Electrical Pacing and Show Spontaneous Ca2+ Signals in the Absence of Stimulation

To explore Ca2+ signaling mechanisms within PVCs, lung slices were loaded with the Ca2+-sensitive indicator Oregon Green BAPTA-1 and examined with real-time 2-photon microscopy. PVCs displayed rapid, transient Ca2+ increases in response to electric field stimulation (EFS; Fig. 1). Most PVCs followed such pacing for long periods with consistent responses (n >50 slices; maximum observation time 17 minutes). Disruption of electrical stimulation by stopping EFS (Fig. 1A), application of the VOCC inhibitor nifedipine (100 µM; Fig. 1B) or removal of extracellular Ca2+ (Fig. 1C) inhibited the pacing-induced Ca2+ responses and provoked the occurrence of spontaneous Ca2+ transients. These data illustrate two key aspects of PVC Ca2+ signaling: the regular Ca2+ signals underpinning physiological EC-coupling, and pro-arrhythmic, spontaneous Ca2+ transients.

Figure 1. PVCs can be paced by electrical field stimulation (EFS), but are prone to showing spontaneous activity.

(A) PVCs displayed rapid increases of cytosolic Ca2+ concentration in response to EFS. Between EFS pulses, some spontaneous Ca2+ activity was observed (indicated by asterisks). After termination of EFS, the cells continued to show spontaneous Ca2+ transients. (B) 100 µM nifedipine inhibited the ability of PVCs to respond to EFS pulses, causing the spontaneous Ca2+ transients to become prominent. (C) Removal of extracellular Ca2+ stopped responses to EFS, but spontaneous Ca2+ transients were still evident [See fig. 3 for an explanation how spontaneous and paced activity was distinguished]. (D) Quantification of the spontaneous Ca2+ transients on the day of the preparation (Day 0) and the three following days, illustrating that the spontaneous activity is not significantly different between day 0 and 2 (n = 7–36 cells, 2–4 slices). (E) Normalized intensity of Oregon Green BAPTA-1 in PVCs on the day of the preparation (Day 0) and the three following days (n = 2–19 cells, 1–10 slices).

We were mindful that damage to PVCs during preparation of lung slices could induce the spontaneous Ca2+ signalling that we observed. However, the slices remained viable, with no obvious visual indication of slice deterioration, in culture for 3 days. The average frequencies of spontaneous Ca2+ transients were not statistically different up to 2 days after preparing the lung slices (Fig. 1D). In addition, PVCs could be electrically paced by EFS for 2 days after slice preparation (later days were not examined). PVCs consistently loaded with the Ca2+-sensitive indicator to the same level for 3 days post preparation (Fig. 1E). Moreover, the Ca2+ indicator was stably retained within PVCs up to 3 days post preparation. Active PVCs did not label with propidium iodide (PI; fig. S2). These data indicate that the PVCs were vital and not damaged during slice preparation, and that they did not rapidly de-differentiate in culture. To ensure consistency, all data in the present study were obtained using slices within 48 hours after preparation. Atrial myocytes within slices of the right atrial chamber, prepared using the same buffer solutions and conditions as for PVCs, could also be paced using EFS, but did not show prominent spontaneous Ca2+ transients (fig. S3). These observations indicate that PVCs and atrial myocytes have a different propensity for spontaneous Ca2+ signaling.

PVC Organization, Morphology and Contraction

Pulmonary veins with a PVC myocardial sheath are found in mouse lung slices cut from the hilus region, where the pulmonary veins exit the lungs, to a distance approximately half-way through the lung lobe towards the periphery of the lungs (Fig. 2A). The myocardial sheath is absent in more peripheral lung slices. The pulmonary veins (also known as intrapulmonary veins within the lung) are easily identified by their lone appearance within the lung alveoli tissue, as compared to the consistently paired association of a pulmonary artery with an airway (Fig. 2A). The PVCs surrounding each vein could be identified using bright-field or phase-contrast optics as relatively large cells, which often displayed regular cellular striations that are typical of myocytes (Figs. 2Bi–Di). Immunostaining for the cardiac isoforms of RyR (types 2 and 3) revealed a striated distribution of these Ca2+ channels (Figs. 2Bii and Cii, respectively) within the PVCs. These striations were 1.8±0.1 µm apart (n = 5 regions per slice, 6 slices), similar to the RyR (z-line) distribution in cardiac myocytes [22]. The membrane dye di-8-ANNEPS predominantly stained only the periphery of PVCs in >98% of cells, indicating that PVCs do not express an extensive transverse-tubule system (Figs. 2Di and Dii). Collectively, these data suggest that PVCs structurally resemble rodent atrial myocytes [22] [23].

Figure 2. Characterization of PVCs and their spontaneous Ca2+ signaling in mouse lung slices.

(A) Cross-section of a lung slice showing a lone pulmonary vein, and the relative position of an airway and its associated pulmonary artery within the alveolar tissue. (Bi–Cii) Phase-contrast images of PVCs revealed the characteristic striated pattern (arrows) associated with sarcomeres of cardiac myocytes that (Bii) overlap with the striated expression pattern of RyR2 and (Cii) RyR3 detected by immunofluorescence. (Di) Light microscopic image and (Dii) membrane staining of PVCs with di-8-ANNEPS predominantly highlighted peripheral membranes with little evidence for internal transverse-tubules (n = 38 cells, 5 slices). (Ei) Differential interference contrast image of a pulmonary vein cross section that displayed (Eii) small, non-coordinated spontaneous contractions and (Eiii) larger, coordinated contractions in response to EFS (1 Hz, black tick marks; line-scan analysis along line indicated in Ei). (Fi–v) Spontaneous Ca2+ increases that either remain localized and only spread over a limited area (white arrows) or that travel as a Ca2+ wave through individual cells (red arrow). The Ca2+ waves often originated in the same location (e.g. red asterisks) and spread in the same direction with each wave. Relative fluorescence (Ca2+) increases are indicated by the pseudocolor bar (F/F0). The time interval between images is indicated in each panel. (Fvi) Grey-scale fluorescence image of the PVCs.

Contraction of Pulmonary Veins

In addition to their location and morphology, PVCs surrounding pulmonary veins could be readily identified by their prominent spontaneous contractions (Fig. 2E). The nature of this contractile activity was heterogeneous. In some veins, repetitive ripples of contraction propagated asynchronously along individual PVCs (frequency range 0.6–1.3 Hz). In other veins, multiple PVCs displayed more rapid, uncoordinated contractile activity. The numbers of PVCs involved ranged from a limited sub-population of cells, to all the PVCs surrounding the vein.

In many lung slices, asynchronous spontaneous PVC contractile activity could be coordinated into synchronous contractions by EFS, with a consequent reduction in the lumen size of the pulmonary vein during each contraction (Figs. 2Ei–iii; Video S1). During spontaneous contractions, the relative reduction in pulmonary vein lumen size was 0.2±0.1%, which increased significantly to 6.3±0.7% (P<0.01) during the EFS-induced contraction (n = 6 slices). Synchronised contraction only occurred during EFS; spontaneous contractile activity resumed with the termination of EFS.

Spontaneous Ca2+ Signaling within PVCs and Entrainment by EFS

The overwhelming majority of PVCs displayed spontaneous Ca2+ activity in the absence of EFS (n >250 slices). The spontaneous Ca2+ transients were heterogeneous, and ranged from localized Ca2+ increases (Ca2+ sparks) to Ca2+ waves that propagated along single PVCs (Figs. 2F; Video S2 and Video S3). The frequency of this spontaneous activity varied from cell to cell within a slice (range 0.3 to 3.4 Hz; mean 2.0±0.2 Hz; n = 24 cells, 5 slices). Ca2+ wave propagation occurred with a speed of 70.2±3.5 µm/s, with a mean wave duration of 388.2±18.8 ms (n = 32 waves, 20 slices). Spontaneous Ca2+ signaling could be imaged within individual PVCs for >15 minutes with little change in activity.

Typical spontaneous Ca2+ signals are illustrated by the line-scan plot in fig. 3A. This plot shows Ca2+-dependent fluorescence within a pair of PVCs (the 8-pixel-wide continuous line that was sampled is shown in fig. 3B). Within such line-scan plots, distance across the cells is represented vertically while time progresses horizontally. Propagating Ca2+ waves are evident as diagonal lines. Spontaneous Ca2+ waves can be seen to originate in the centre of cell 1 (asterisk), and propagate bi-directionally to the poles of the cell (arrowheads). The Ca2+ waves stopped at the boundary between cells 1 and 2 (dashed line). Cell 2 had its own intrinsic rhythm, and displayed Ca2+ waves that sometimes temporally coincided with those of cell 1.

Figure 3. Effect of Electrical Field Stimulation (EFS) on PVC Ca2+ signals.

(A) Line-scan analysis of spontaneous Ca2+ signals and responses to 1 Hz EFS in two neighboring cells. The cells are outlined in the 2-photon fluorescence image of PVCs in a lung slice shown in (B). In (A), the cell border is depicted by the dashed line. The spontaneous Ca2+ activity in Cell 1 shows a bidirectional wave (origin indicated by an asterisks, arrowheads indicate direction). The timing of the Ca2+ increases in both cells is independent of each other, but did occasionally coincide. EFS caused whole-cell Ca2+ increases, which are shown as vertical straight lines. EFS timing and pulse numbers are indicated by the top bar. The ability of Cell 1 or 2 to respond to EFS is indicated in the table; ‘+’ indicates full response, while ‘±’ indicates an incomplete response. (C) Summary, showing that the latency from an EFS-induced Ca2+ signal to the next spontaneous Ca2+ transient is significantly longer for PVCs with infrequent spontaneous Ca2+ activity (left). In contrast, the latency from a spontaneous Ca2+ transient to the next EFS-induced Ca2+ signal did not depend on the frequency of the spontaneous activity (right) (n >166 events in 10 cells, 3 slices).

EFS-evoked Ca2+ transients are evident in fig. 3A as vertical lines that coincide with EFS pulses. EFS restored the effect of sinus rhythm by evoking Ca2+ signals (and triggering contraction) within the PVCs. However, the efficacy of each EFS pulse varied on a pulse-to-pulse basis, and ranged from inducing a whole-cell Ca2+ signal to having no effect. For example, in fig. 3A, only on 3 occasions did EFS pulses cause a whole-cell Ca2+ signal simultaneously in both cells (at pulses 9, 11, 14; both Cell 1 and 2 are marked ‘+’). The other EFS pulses correlated with one of three types of responses; i) only one cell showed a whole-cell response (pulse 1, Cell 1 is marked ‘+’), ii) one or both cells showed an incomplete response (pulses 3–7, 12, 13 and 15, marked ‘±’) or iii) neither cell responded (pulses 2, 8 and 10). A plausible explanation for variable effects of EFS is that the PVCs enter a refractory period after a Ca2+ rise during which they cannot respond to another Ca2+-releasing stimulus. Therefore, if a cell responds to EFS with a whole-cell response, the subsequent spontaneous Ca2+ signal is inhibited or delayed. Conversely, if a spontaneous Ca2+ wave occurs just prior to an EFS pulse, the cell cannot respond with a Ca2+ transient. In essence, Ca2+ signals evoked by EFS delayed the timing of spontaneous Ca2+ activity and vice versa.

We found that the average latency between an EFS-evoked Ca2+ signal and a subsequent spontaneous Ca2+ event was correlated with the frequency of the intrinsic spontaneous Ca2+ activity. The latency was longer in cells with a sparse spontaneous Ca2+ activity (<1 Hz) as compared to cells with more frequent spontaneous Ca2+ activity (>1 Hz, fig. 3C, left). An explanation for these observations is that cells with higher frequencies of spontaneous Ca2+ signals have a greater capacity for restoring SR Ca2+ content. This would enable PVCs to recover more quickly from an EFS-induced Ca2+ pulse, as well as predispose the PVCs to show more frequent spontaneous Ca2+ signals. Interestingly, there was no similar relationship governing the interval between a spontaneous Ca2+ signal and a subsequent Ca2+ pulse induced by EFS (Fig. 3C, right). This is likely due to the fact that an EFS pulse triggers a Ca2+ influx signal to promote internal Ca2+ release. The EFS-induced Ca2+ signals may therefore be less dependent on SR Ca2+ refilling in comparison to the spontaneous Ca2+ events.

We hypothesised that increased EFS frequency might overcome the confounding effects of spontaneous Ca2+ signals and provide a more reliable entrainment of PVCs. However, elevated EFS frequencies only partially improved entrainment, and moreover caused markedly heterogeneous cell responses. For example, fig. 4 shows the responses of two adjacent PVCs to EFS applied at 2 and 3 Hz. Prior to EFS, both cells displayed independent spontaneous Ca2+ signals (Fig. 4A). Cell 1 was fully paced (without any remaining spontaneous activity, FP) with 2 Hz (Fig. 4Bii) and 3 Hz EFS. However, in response to 3 Hz, Cell 1 displayed ‘alternans’ (alternating large and small Ca2+ transients, fig. 4Cii). Conversely, Cell 2 was only partially paced (PP) with 2 Hz EFS (Fig. 4Biii), but was fully paced (FP) with 3 Hz EFS without alternans (Fig. 4Ciii). These observations indicate that while increasing the frequency of EFS could sometimes entrain spontaneous Ca2+ signals, other forms of pro-arrhythmic activity, e.g. alternans, could ensue.

Figure 4. Heterogeneity in the pulsed responses of neighboring PVCs to increasing frequencies of EFS.

At each EFS frequency (0 to 3 Hz), the responses of 2 contiguous PVCs are illustrated with a line-scan analysis (left, Ai to Ci) and a fluorescence (Ca2+) trace for Cell 1 (Aii to Cii) and Cell 2 (Aiii to Ciii). Within the line-scan image, the cell border between the PVCs is indicated by the dashed line. The timing of the EFS pulses is indicated by the top bar. (A) PVCs displayed spontaneous Ca2+ activity in the absence of EFS. (B) With 2 Hz EFS, Cell 1 displayed full pacing (FP) while Cell 2 showed partial pacing (PP). (C) With 3 Hz EFS, Cell 1 displayed alternans while Cell 2 was fully paced. (Di–iv) Summary, showing the predominate forms of response induced by different EFS frequencies (n = 25 cells, 4 slices). (E) Correlation of spontaneous Ca2+ transient frequency and the EFS stimulation frequency required for full pacing (n = 25 cells, 4 slices).

Figure 4D quantitatively summarizes the relationship between the frequency of EFS and the pattern of PVC Ca2+ signaling. It is evident that PVCs were more successfully paced as the EFS frequency increased. However, not all PVCs were paced (Fig. 4Di). PVCs that were partially paced were more evident at lower EFS frequencies (Fig. 4 Dii), whilst higher EFS frequencies were most effective for complete entrainment (Fig. 4Diii). However, the higher EFS frequencies were also capable of generating alternans (Fig. 4Div). We suggest that the heterogeneity of EFS-induced Ca2+ signaling arises from the relative timing of an EFS pulse with respect to the intrinsic spontaneous Ca2+ signaling activity; it is easier to fully pace cells which show little spontaneous activity, and more difficult to pace cells with a high frequency of spontaneous activity. In support of this conclusion, we observed that the minimal EFS frequency required for fully paced cells correlated with the intrinsic frequency of spontaneous Ca2+ signals (Fig. 4E). Those cells with more frequent spontaneous Ca2+ signals required higher EFS frequencies. Moreover, at all EFS frequencies tested, the cells that could be fully paced had a lower average frequency of spontaneous Ca2+ signals compared to those cells that could not be fully paced. With 1 Hz EFS, for example, fully paced PVCs had an average spontaneous Ca2+ signal frequency of 0.7±0.1 Hz, whereas those cells that could not be fully paced by 1 Hz EFS had an average spontaneous Ca2+ signal frequency of 1.9±0.2 Hz (n = 20, 5 slices). We suggest that all PVCs have unique periodicities of spontaneous Ca2+ signaling, and this determines the cells’ refractory periods. In some cells, the frequency of spontaneous Ca2+ signaling is so high that EFS is ineffective. It should be noted that whilst there is a clear impact of the frequency of spontaneous Ca2+ signals on the pacing of PVCs, this is not the only factor. We also observed cells with similar frequencies of spontaneous Ca2+ signaling responding differently to the same EFS pulse. This implies heterogeneity in the responsiveness of the cells to EFS in addition to the effects of spontaneous Ca2+ signaling.

Ca2+ Re-addition Induces Rapid Ca2+ Waves within PVCs and Prevents Electrical Synchronisation

Superfusion of non-paced PVCs with Ca2+-free medium caused a progressive reduction in the amplitude and frequency of spontaneous Ca2+ signals (Fig. 5A), and of the basal cytosolic Ca2+ concentration, until the PVCs became quiescent (quantified in figs. 5Fi–iv). In 58% of the PVCs, the spontaneous Ca2+ activity was fully inhibited within 5 min of Ca2+ removal (mean time to inhibition: 209.5±15.3 s). This inhibition of the spontaneous Ca2+ signals also correlated with the loss of spontaneous PVC contractions. These data indicate i) that spontaneous Ca2+ signals arise via Ca2+ release from intracellular stores, ii) that PVCs have a substantial capacity for recycling Ca2+, and iii) that PVCs require Ca2+ influx to sustain spontaneous Ca2+ activity and contraction.

Figure 5. Response of PVCs to the removal and re-addition of external Ca2+.

(A) A fluorescence (Ca2+) trace showing the reduction of spontaneous Ca2+ transients during a period of extracellular Ca2+ removal, and increased spontaneous Ca2+ transient activity after Ca2+ re-addition. (B) A representative example of (Bi) the spontaneous Ca2+ activity before Ca2+ removal, (Bii) the restoration of the Ca2+ transients during Ca2+ re-addition and (Biii) the prolonged Ca2+ waves after Ca2+ re-addition. (Ci–iv) Quantification of the Ca2+ responses of PVCs to Ca2+ removal and re-addition. (Ci) The frequency of the spontaneous Ca2+ transients is reduced by Ca2+ removal. (Cii) The amplitude (ΔF/F0) of the spontaneous Ca2+ transients and (Ciii) the cytosolic Ca2+ concentration transients are both reduced by Ca2+ removal and increased after Ca2+ re-addition. (Civ) The duration of Ca2+ transients increases at Ca2+ re-addition (n = 35 cells, 12 slices). (D) A fluorescence (Ca2+) trace showing the effect of external Ca2+ removal while 1 Hz EFS was applied. EFS-induced pacing of the cell was lost during Ca2+ removal and was not re-established when extracellular Ca2+ was restored. (E) Summary of the (Ei) Ca2+ responses to EFS before the Ca2+ removal, (Eii) Ca2+ signals following Ca2+ re-addition and (Eiii) the prolonged Ca2+ waves established after Ca2+ re-addition. (F) shows (Fiiv) Line-scan analysis of phase-contrast images measuring pulmonary vein contraction (cf. Fig. 1). (Fi) No contraction without EFS. (Fii) Coordinated contractions with 1 Hz EFS. (Fiii) In the absence of external Ca2+, EFS did not evoke contractions. (Fiv) Strong uncoordinated contractions after the re-addition of Ca2+.

The re-addition of extracellular Ca2+ (1.3 mM) did not restore the initial pattern of spontaneous Ca2+ activity, but rather induced an entirely different form of Ca2+ signaling. Within 40 to 60 s of Ca2+ re-addition, PVCs began to display Ca2+ waves that progressively increased in amplitude, frequency and spatial propagation (Figs. 5A, Bii; quantified in figs. 5Ci–iv). Moreover, these Ca2+ waves had a longer duration than the original spontaneous activity (compare Figs. 5Biii and Bi; fig. 5Civ). Consequently, the cytosolic Ca2+ concentration did not have time to recover between wave fronts, so the diastolic Ca2+ level became elevated. Typically, all the PVCs surrounding a pulmonary vein displayed Ca2+ waves in response to re-addition of extracellular Ca2+ (100% of slices; n = 12 slices). The Ca2+ waves were associated with strong PVC contractions, but the contractions were not coordinated between cells with the result that the layer of PVCs fibrillated asynchronously. Significantly, the rapid Ca2+ waves and strong contractions persisted for some time after Ca2+ re-addition, and did not return to the original pattern of spontaneous activity within 15 minutes. The spontaneous Ca2+ waves in PVCs under control conditions were predominantly intracellular. The Ca2+ removal/re-addition experiments presented a situation in which intracellular Ca2+ waves frequently turned into intercellular Ca2+ waves (an example is shown in Fig. S4).

Application of EFS either prior to, or during, Ca2+ re-addition did not prevent the PVCs from developing long-lasting Ca2+ waves (Fig. 5D and E; compare spontaneous Ca2+ transients in Fig. 5Ei with the increased duration Ca2+ waves in Fig. 5Eiii; Video S3). Even slices showing an initial high degree of entrainment to EFS (Fig. 5 Ei) were driven to the fibrillated state that was unresponsive to EFS (Figs. 5 Eii and Eiii) by removal and re-addition of extracellular Ca2+. Similarly, EFS could not re-synchronize the contractile activity of the PVCs following re-addition of extracellular Ca2+ (Figs. 5 Fi–iv). Re-addition of a lower extracellular Ca2+ concentration (500 µM instead of 1.3 mM), slowed the onset of the rapid Ca2+ waves and contractions, but the PVCs eventually reached a similar state of fibrillation that was unresponsive to EFS.

Our data indicate that PVCs can progress to an unrecoverable, EFS-insensitive form of Ca2+ signaling that is characterized by Ca2+ waves with a longer duration and a persistent elevation of cytosolic Ca2+. Moreover, other maneuvers (see below) that promoted Ca2+ influx or altered Ca2+ homeostasis could also cause PVCs to display a rapid, EFS-resistant, form of Ca2+ signaling.

The fact that spontaneous Ca2+ signals continue for some time in the absence of extracellular Ca2+ suggests that these Ca2+ signals depend on Ca2+ release from internal stores. Consistent with this idea, we observed that caffeine (a RyR agonist; 1 mM) increased the frequency of spontaneous Ca2+ transients and typically caused a sustained elevation of cytosolic Ca2+ (Fig. 6A and B, quantified in Ei and Eii). PVCs with EFS showed a similar caffeine-evoked increase in spontaneous Ca2+ activity, and consequent loss of entrainment in 47% of the cells (Fig. 6C and D, quantified in Eiii). The effects of caffeine were reversible within 30–60 s (Fig. 6Aii and Bii), and the PVCs could be re-entrained by EFS (compare responses between 30 to 60 seconds in Fig. 6 Ci/Di with Cii/Dii). Ryanodine (a RyR antagonist; 10 µM) inhibited the spontaneous Ca2+ signals within ∼150 s (Figs. S5A and B) and completely blocked EFS-evoked Ca2+ signals within ∼100 s in an irreversible manner (Figs. S5C and D). We attempted to use a maximal caffeine concentration to provoke emptying of the SR Ca2+ store. This is a commonly-employed method for assessing the SR Ca2+ load in cardiomyocytes. However, we found that 20 mM caffeine did not produce the same large, monophasic Ca2+ responses in PVCs that it evokes in isolated cells. Rather, application of 20 mM caffeine caused an acceleration of Ca2+ transients, leading to a persistent increase of the cytosolic Ca2+ concentration (Fig. S6).

Figure 6. Caffeine stimulates Ca2+ signals in PVCs.

(Ai) A Ca2+ trace and (Bi) line-scan analysis of spontaneous Ca2+ signals, illustrating that 1 mM caffeine activated a reversible increase in the spontaneous Ca2+ activity, which was reversible (Aii and Bii). (Ci) A Ca2+ trace and (Di) line-scan analysis of Ca2+ signals paced by 1 Hz EFS illustrating that 1 mM caffeine increased the spontaneous Ca2+ activity and caused a loss of pacing, which was reversible (Cii and Dii). (E) Summary of the response to 1 mM caffeine showing (Ei) increased frequency of the spontaneous Ca2+ signals and increased basal Ca2+ levels in (Eii) the absence of EFS (n = 46 cells, 9 slices) or (Eiii) the presence of EFS (n = 35 cells, 8 slices).

Inositol 1,4,5-trisphosphate receptors (InsP3R) did not appear to play a role in spontaneous Ca2+ responses. Photolytic release of InsP3 did not increase the cytosolic Ca2+ concentration in PVCs (mean F/F0 1.0±0.01 before and 1.1±0.03 after uncaging, n = 23 cells, 8 slices). In contrast, airway smooth muscle cells within the same slices showed substantial Ca2+ signals following photolysis of caged InsP3 (mean F/F0 1.1±0.01 before and 1.7±0.01 after uncaging, n = 3 cells, 2 slices). Furthermore, 2-aminoethoxydiphenyl borate (2-APB; an InsP3R antagonist; 5 µM), did not alter the frequency or spatial properties of spontaneous Ca2+ signaling within PVCs (initial frequency in absence of 2-APB: 1.4±0.16 Hz, 180 s after 2-APB 1.3±0.19 Hz, 300 s after 2-APB 1.5±2.3 Hz,; n.s.; n = 15 cells, 4 slices).

Depolarising PVCs by elevating the extracellular KCl concentration (from 5.3 mM to 50 mM) progressively increased the frequency of spontaneous Ca2+ signals (Figs. 7Ai and Bi, quantified in fig. 7Fi) and stimulated the occurrence of spontaneous events in PVCs that were previously silent (Figs. 7Aii and Bii). KCl treatment had a similar effect to that of caffeine and induced rapid Ca2+ waves superimposed on an elevated cytosolic Ca2+ level, which were reversible upon washout (Fig. 7Fii). The effects of KCl on PVCs during EFS were similar; KCl induced the progressive development of rapid Ca2+ waves on an elevated Ca2+ baseline to the point that synchronised Ca2+ signals and contractions induced by EFS were prevented (Figs. 7C and D, Fiii; contractions in figs. 7Ei–iv). Pacing was lost after 19±1.2 s of KCl application. As with caffeine, the rapid Ca2+ waves evoked by KCl could be reversed upon washout and entrainment re-established 67.1±3.6 s after washout (Fig. 7Dii).

Figure 7. KCl stimulates Ca2+ signals in PVCs.

(Ai, Aii) Ca2+ traces and (Bi, Bii) line-scan analysis (selected details) showing that KCl increased spontaneous Ca2+ signals in PVCs. (C) Ca2+ trace and (Di, Dii) selected line-scan analysis showing the that KCl reversibly caused a loss of pacing and increased spontaneous Ca2+ signals in PVCs during 1 Hz EFS. (Eiiv) Line-scan analysis of phase-contrast images of a pulmonary vein measuring contraction (cf. Fig. 1). (Ei) No contractions without EFS. (Eii) 1 Hz EFS induced contractions. (Eiii) KCl changed the coordinated contractions to uncoordinated fibrillations. (Eiv) EFS responses were restored after KCl wash-out. (Fi–iii) Quantification of the PVC responses to KCl addition. Summaries showing that KCl (Fi) increased the frequency of the spontaneous Ca2+ signals and (Fii) the basal Ca2+ concentration in PVCs without EFS (n = 17 cells, 4 slices). KCl also (Fiii) increased the basal Ca2+ concentration in PVCs paced with 1 Hz EFS (n = 24 cells, 4 slices).

Nifedipine (100 µM) stopped EFS-induced pacing within 63±26.1 s (Figs. S7A and B) and contraction (Fig. S7C). While inhibiting the EFS-induced Ca2+ signals, nifedipine promoted the re-occurrence of spontaneous Ca2+ events which had been previously suppressed by EFS (Figs. 1B and S7Bi–Biv). In the absence of EFS, nifedipine reduced the frequency of spontaneous Ca2+ signals in a reversible manner (Figs. S7D and E). This latter effect indicates that VOCCs provide a source of Ca2+ influx to support or trigger spontaneous Ca2+ events. Verapamil (another L-type VOCC blocker; 100 µM) had similar effects to nifedipine (data not shown).


In the present study, we characterised spontaneous and EFS-induced Ca2+ signals within PVCs, and examined the effect of spontaneous Ca2+ signaling on the ability of the PVCs to respond to sustained electrical pacing. The use of the lung slice preparation for examining PVCs has the advantages of easy preparation, a cellular architecture resembling the in situ organization, compatibility with microscopy and a viability lasting several days. Spontaneous Ca2+ signals have previously been recorded in PVCs from various species [24][26], but their influence on pacing has not been resolved. A critical aspect of PVCs that was apparent in this study was the heterogeneity between cells in terms of Ca2+ handling. PVCs display repetitive spontaneous Ca2+ release events, but the frequencies and spatial properties of those events are unique for each cell. Correspondingly, PVCs show heterogeneous responses to electrical pacing such that neighboring cells can become, and remain, desynchronised. Such heterogeneity is known to serve as a basis for cardiac arrhythmias, and this may underlie the generation of arrhythmic events within pulmonary veins.

PVCs play an important role in the genesis and maintenance of AF [15], [27][29]. Electrical mapping studies have pinpointed PVCs as a source of phase 4 depolarisations (i.e. APs during the normally quiescent diastolic period) [30] that can pervade the atria via the junction of the atrium and PV at the left atrial antrum. Ablation procedures that prevent the electrical continuity between PVCs and atrial myocytes are highly effective in treating AF, but can have serious side-effects [31]. Despite the confirmation of PVCs as a source of ectopic activity, the events that occur within these cells to generate arrhythmic signals are less clear [11]. With respect to the mechanisms generating arrhythmias, numerous studies, using a range of cardiac cell types, have demonstrated that spurious Ca2+ signals are a potent source of arrhythmic activity [32]. Moreover, 3-dimensional electrical mapping of pulmonary veins identified spontaneous activity occurring as discrete focal events, consistent with the notion that cellular Ca2+ signals may be involved [33].

Under normal conditions, contractile myocytes and conductive cells within the heart experience a Ca2+ signal during the passage of each AP. At the end of an AP, cytosolic Ca2+ must recover to diastolic levels in order for the cells to reset and be fully responsive to the next stimulation. Typically, there is a diastolic period of hundred(s) of milliseconds between each AP where Ca2+ remains at the resting level of ∼100 nM. Consequently, spontaneous Ca2+ signals may corrupt this normal activity and recovery cycle of cardiac cells in a number of ways [32]. For example, Ca2+ causes the activation of electrogenic ion transporters, such as Na+/Ca2+ exchange or Ca2+-dependent chloride flux that leads to cellular depolarisation. Ca2+ signals occurring during the recovery of an AP, or in the normally quiescent diastolic period, give rise to ‘triggered activity’ such as early- or delayed-after-depolarisations (often denoted ‘EADs’ and ‘DADs’ respectively) [34]. If EADs and DADs occur with sufficient magnitude, they can cause a cell to depolarise and trigger an AP that propagates to neighboring cells. It is not known how many PVCs would need to act in concert to be a focus for ectopic AP generation. Most likely, coordination of multiple cells is required to provide a sufficient signal for a propagating wave of depolarisation. However, even those EADs and DADs that do not reach the threshold for triggering an AP can disrupt the cyclical activation of myocytes by rendering them refractory, and thereby causing them to lose synchrony with their neighbours; a putative mechanism for generating electrical re-entry circuits. In the longer term, spontaneous Ca2+ signals can alter gene expression, and lead to deleterious phenotypic remodelling of myocytes [35].

A key finding in the present study was the propensity of PVCs to perpetually display spontaneous Ca2+ signals even in the presence of electrical pacing. It would be reasonable to expect that EFS could extinguish spontaneous Ca2+ signals by providing regular command pulses, and normalising SR Ca2+ content. Consistent with this idea, we observed that EFS could dampen spontaneous Ca2+ release, in particular in those cells with relatively low spontaneous Ca2+ activity. However, the effectiveness of electrical pacing was determined by the frequency of spontaneous Ca2+ signals. In essence, if spontaneous Ca2+ release occurred close to the EFS frequency then pacing was either partially successful or unsuccessful. Similar observations have been made using model stimulations of spontaneously active atrial myocytes [36]. The tendency of PVCs to show spontaneous Ca2+ signals would certainly contribute to heterogeneous electrical responses within a pulmonary vein [16]. Such heterogeneity could lead to the generation of focal activity that may eventually propagate to the heart and disturb the cardiac cycle. Moreover, events that change the occurrence of spontaneous Ca2+ release, such as alteration of Ca2+ influx or RyR sensitivity can further decrease the ability of APs to control PVCs and enhance the likelihood of pro-arrhythmic events. Suppressing spontaneous Ca2+ signaling in PVCs may therefore provide a plausible target to prevent the inception of AF. Other studies have also suggested clamping of RyR activity as a putative mechanism for controlling pulmonary vein-induced arrhythmogenesis [37]. Our data suggest that VOCC-mediated Ca2+ influx is necessary for EFS, but also provides a source of Ca2+ for spontaneous Ca2+ signaling. In the present study, we did not characterise all putative Ca2+ influx mechanisms, in particular those activated by removal/re-addition of extracellular Ca2+, but multiple Ca2+ influx pathways may be involved, as in smooth muscle [38].

The reason why PVCs have a high propensity to show spontaneous Ca2+ signals is unclear. In previous studies of PVCs [39] and atrial myocytes [40] InsP3Rs were identified as triggers for spontaneous Ca2+ signaling, but these channels did not play a detectable role within the PVCs used in this study. Since SR Ca2+ content is known to regulate RyR activation [41], a plausible explanation is that PVCs have a relatively high SR Ca2+ content so that their RyRs open spontaneously. In support of this notion, experiments designed to increase SR Ca2+ content using KCl, or RyR sensitivity using caffeine, increased spontaneous Ca2+ signaling. In addition, it is evident that PVCs have sufficient SR Ca2+ to allow spontaneous Ca2+ signals to occur for several minutes in a Ca2+-free medium. All manoeuvres that provoked PVCs to show more rapid spontaneous events also made the PVCs resistant to electrical pacing. We therefore suggest that if the SR Ca2+ content, Ca2+ influx or RyR sensitivity are sufficiently enhanced, spontaneous Ca2+ activity will be accelerated and the PVCs become insensitive to EFS.

Supporting Information

Figure S1.

PVCs show spontaneous activity and can be paced by electrical field stimulation (EFS) at 37°C. The line-scans in Aiiii illustrate the correlation between EFS and contraction in PVCs at 37°C. (Ai) depicts the lack of coordinated contraction in the absence of EFS. Whereas, 1 Hz (Aii) and 5 Hz (Aiii) EFS caused obvious contraction. Panels Bi–iii illustrate the Ca2+ transients observed in absence of EFS (Bi), and during 1 Hz (Bii) and 5 Hz (Biii) EFS. (C) The frequency of the spontaneous Ca2+ transients is not significantly different at room temperature (RT; n = 53 cells, 15 slices) and at 37°C (n = 57 cells, 14 slices).


Figure S2.

Viability of PVCs within a lung slice. The images in Aiiii depict a region of a lung slice with PVCs (a portion of them are outlined) and surrounding cells. The lung slice was incubated with Oregon Green BAPTA-1 AM (Ai) and Propidium Iodide (PI) (Aii) to examine dye retention/exclusion. The images are merged in (Aiii). The images were taken during Day 1 after the preparation of the slice. The traces in B show the spontaneous Ca2+ signals in an active PVC (sampled from the region indicated by the blue circle in A) and an inactive cell (sampled from the region indicated by the magenta circle in A). C illustrates that active PVCs maintained a relatively low PI fluorescence over 4 days in culture. In contrast, dead cells (non-identified) within the slices had a relatively higher PI fluorescence. The absolute PI fluorescence sampled from labelled cells did not significantly change over the period of 4 days (One-Way ANOVA, n = 2–29 cells, 1–10 slices per day).


Figure S3.

Atrial myocytes within atrial slices do not show spontaneous Ca2+ transients. (A) A 2-photon fluorescence image of Oregon Green BAPTA-1-loaded myocytes within an atrial slice. (B) Atrial myocytes do not display Ca2+ transients without EFS. However, these atrial myocytes showed Ca2+ transients in response to 1 Hz (C), 2.5 Hz (D) and 4 Hz (E) EFS. Ca2+ signals were sampled from the region bounded by the dashed box.


Figure S4.

Intracellular Ca2+ waves change to intercellular Ca2+ waves after removal and re-addition of extracellular Ca2+. (A) A line-scan plot illustrating that intracellular Ca2+ waves (and subcellular Ca2+ signals) remain within individual cells (the cell boundaries are indicated by the dashed lines) under control conditions. (B) A line scan plot of the same cellular regions as in A following removal and re-addition of extracellular Ca2+ illustrates that the previously constrained Ca2+ intracellular waves are transformed into propagating intercellular Ca2+ waves (experiment similar to that shown in Fig. 5A).


Figure S5.

Ryanodine inhibits Ca2+ signals in PVCs. (A) Ca2+ trace and (Bi–Biii) line-scan plots showing that 10 µM ryanodine gradually inhibited spontaneous Ca2+ signals within PVCs (n = 26 cells, 6 slices). (C) Ca2+ trace and (Di–Diii) line-scan plots of PVCs during 1 Hz EFS. The EFS-evoked responses were initially enhanced, and then progressively inhibited, by 10 µM ryanodine (n = 20 cells, 3 slices).


Figure S6.

Effect of a maximal caffeine concentration on PVCs. (A) and (B) show representative traces of Ca2+ signals caused by superfusion of spontaneously active PVCs with 20 mM caffeine. Caffeine increased the spontaneous Ca2+ transient frequency, leading to a maintained plateau of elevated Ca2+.


Figure S7.

Nifedipine inhibits EFS-entrained Ca2+ signals, thus revealing spontaneous Ca2+ signals. Nifedipine also progressively reduces the frequency of spontaneous Ca2+ signals in unpaced cells. (A) Ca2+ trace and (Bi–Biv) line-scan analysis of Ca2+ signals in PVCs paced by 1 Hz EFS. The EFS responses are progressively inhibited by 100 µM nifedipine so that spontaneous Ca2+ signals become evident. (C) Line-scan analysis illustrating that EFS-induced contraction observed under control conditions (Ci) is inhibited by 100 µM nifedipine (Cii). (D) Ca2+ trace and (Ei, Eii) line-scan analysis showing a gradual reduction of spontaneous Ca2+ signals in response to 100 µM nifedipine. (F) Quantitation of the declining frequency of the spontaneous Ca2+ signals in nifedipine-treated PVCs (n = 21 cells, 5 slices).


Video S1.

Spontaneous and EFS induced contraction of a pulmonary vein. Phase contrast image of a cross section through a pulmonary vein, showing a small amount of spontaneous contraction around the whole circumference of the vein. After application of EFS a stronger contraction of the pulmonary vein is observed in response to every electric pulse. Image acquisition rate 15 frames per second.


Video S2.

Spontaneous and EFS induced Ca2+ signals in PVCs. The video shows spontaneous Ca2+ signaling in Oregon-Green BAPTA-1 loaded PVCs. The spontaneous Ca2+ waves either stay intracellular or travel through several PVCs. In response to EFS a simultaneous whole cell Ca2+ increase is seen in all PVCs. The spontaneous activity continues in-between the EFS pulses. Image acquisition rate 30 frames per second.


Video S3.

Ca2+ signals during a Ca2+ removal and re-addition experiment. The first part of the video shows the spontaneous Ca2+ signals in PVCs before Ca2+ removal from the superfusion medium. The observed Ca2+ waves are predominantly intracellular. The second part of the video shows the absence of spontaneous Ca2+ signals after 5 minutes in Ca2+ free sHBSS (100 µM EGTA). A gradual increase in Ca2+ transients after re-addition of 1.3 mM Ca2+ is shown in the third part of the video. In contrast to the spontaneous activity seen in the first part, most of the Ca2+ waves after Ca2+ re-addition are intercellular Ca2+ waves, travelling though several cells. Strong, uncoordinated contractions develop during the Ca2+ re-addition period. Image acquisition rate 30 frames per second.



We thank Prof. Vincenzo Sorrentino (San Raffaele Scientific Institute, Siena, Italy) for the kind gift of antibodies, and Prof. Sir Michael Berridge for comments and suggestions.

Author Contributions

Conceived and designed the experiments: KR MDB MJS. Performed the experiments: KR. Analyzed the data: KR. Wrote the paper: KR MDB. Discussed the results and critically commented on the manuscript: KR MDB MJS.


  1. 1. Bers DM (2002) Cardiac excitation-contraction coupling. Nature 415: 198–205.
  2. 2. Dobrzynski H, Anderson RH, Atkinson A, Borbas Z, D’Souza A, et al. (2013) Structure, function and clinical relevance of the cardiac conduction system, including the atrioventricular ring and outflow tract tissues. Pharmacol Ther 139: 260–288.
  3. 3. Berridge MJ (2003) Cardiac calcium signaling. Biochem Soc Trans 31: 930–933.
  4. 4. Niggli E (2011) Ryanodine receptors: waking up from refractoriness. Cardiovasc Res 91: 563–564.
  5. 5. Nattel S (2003) Atrial electrophysiology and mechanisms of atrial fibrillation. J Cardiovasc Pharmacol Ther 8 Suppl 1: S5–11.
  6. 6. Heijman J, Voigt N, Nattel S, Dobrev D (2012) Calcium handling and atrial fibrillation. Wien Med Wochenschr 162: 287–291.
  7. 7. Bers DM (2008) Calcium cycling and signaling in cardiac myocytes. Annu Rev Physiol 70: 23–49.
  8. 8. Voigt N, Nattel S, Dobrev D (2012) Proarrhythmic atrial calcium cycling in the diseased heart. Adv Exp Med Biol 740: 1175–1191.
  9. 9. Johnson DM, Heijman J, Bode EF, Greensmith DJ, van der Linde H, et al. (2013) Diastolic spontaneous calcium release from the sarcoplasmic reticulum increases beat-to-beat variability of repolarization in canine ventricular myocytes after beta-adrenergic stimulation. Circ Res 112: 246–256.
  10. 10. Yeh YH, Wakili R, Qi XY, Chartier D, Boknik P, et al. (2008) Calcium-handling abnormalities underlying atrial arrhythmogenesis and contractile dysfunction in dogs with congestive heart failure. Circ Arrhythm Electrophysiol 1: 93–102.
  11. 11. Namekata I, Tsuneoka Y, Tanaka H (2013) Electrophysiological and pharmacological properties of the pulmonary vein myocardium. Biol Pharm Bull 36: 2–7.
  12. 12. Sherif HM (2013) The developing pulmonary veins and left atrium: implications for ablation strategy for atrial fibrillation. Eur J Cardiothorac Surg.
  13. 13. Tsuneoka Y, Kobayashi Y, Honda Y, Namekata I, Tanaka H (2012) Electrical activity of the mouse pulmonary vein myocardium. J Pharmacol Sci 119: 287–292.
  14. 14. Calkins H (2013) Has the time come to abandon the concept that “pulmonary vein isolation is the cornerstone of atrial fibrillation ablation”? Circ Arrhythm Electrophysiol 6: 241–242.
  15. 15. Yoshida N, Okamoto M, Hirao H, Suenari K, Nanba K, et al. (2013) Efficacy of Pulmonary Vein Isolation on Left Atrial Function in Paroxysmal and Persistent Atrial Fibrillation and the Dependency on Its Baseline Function. Echocardiography 30: 744–750.
  16. 16. Hirose M, Laurita KR (2007) Calcium-mediated triggered activity is an underlying cellular mechanism of ectopy originating from the pulmonary vein in dogs. Am J Physiol Heart Circ Physiol 292: H1861–1867.
  17. 17. Kracklauer MP, Feng HZ, Jiang W, Lin JL, Lin JJ, et al. (2013) Discontinuous thoracic venous cardiomyocytes and heart exhibit synchronized developmental switch of troponin isoforms. FEBS J 280: 880–891.
  18. 18. Sanderson MJ (2011) Exploring lung physiology in health and disease with lung slices. Pulm Pharmacol Ther 24: 452–465.
  19. 19. Sanderson MJ, Parker I (2003) Video-rate confocal microscopy. Methods Enzymol 360: 447–481.
  20. 20. Ruas M, Rietdorf K, Arredouani A, Davis LC, Lloyd-Evans E, et al. (2010) Purified TPC Isoforms Form NAADP Receptors with Distinct Roles for Ca2+ Signaling and Endolysosomal Trafficking. Current Biology 20: 703–709.
  21. 21. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9: 671–675.
  22. 22. Smyrnias I, Mair W, Harzheim D, Walker SA, Roderick HL, et al. (2010) Comparison of the T-tubule system in adult rat ventricular and atrial myocytes, and its role in excitation-contraction coupling and inotropic stimulation. Cell Calcium 47: 210–223.
  23. 23. Bootman MD, Smyrnias I, Thul R, Coombes S, Roderick HL (2011) Atrial cardiomyocyte calcium signaling. Biochim Biophys Acta 1813: 922–934.
  24. 24. Logantha SJ, Cruickshank SF, Rowan EG, Drummond RM (2010) Spontaneous and electrically evoked Ca2+ transients in cardiomyocytes of the rat pulmonary vein. Cell Calcium 48: 150–160.
  25. 25. Chang SH, Chen YC, Chiang SJ, Higa S, Cheng CC, et al. (2008) Increased Ca2+ sparks and sarcoplasmic reticulum Ca2+ stores potentially determine the spontaneous activity of pulmonary vein cardiomyocytes. Life Sci 83: 284–292.
  26. 26. Coutu P, Chartier D, Nattel S (2006) Comparison of Ca2+-handling properties of canine pulmonary vein and left atrial cardiomyocytes. Am J Physiol Heart Circ Physiol 291: H2290–2300.
  27. 27. Moreira W, Timmermans C, Wellens HJ, Rodriguez LM (2008) Atrial tachycardia originating from the pulmonary vein: importance of the stimulation sites. Tex Heart Inst J 35: 356–358.
  28. 28. Patterson E, Jackman WM, Beckman KJ, Lazzara R, Lockwood D, et al. (2007) Spontaneous pulmonary vein firing in man: relationship to tachycardia-pause early afterdepolarizations and triggered arrhythmia in canine pulmonary veins in vitro. J Cardiovasc Electrophysiol 18: 1067–1075.
  29. 29. Huang H, Wang X, Chun J, Ernst S, Satomi K, et al. (2006) A single pulmonary vein as electrophysiological substrate of paroxysmal atrial fibrillation. J Cardiovasc Electrophysiol 17: 1193–1201.
  30. 30. Takahara A, Sugimoto T, Kitamura T, Takeda K, Tsuneoka Y, et al. (2011) Electrophysiological and pharmacological characteristics of triggered activity elicited in guinea-pig pulmonary vein myocardium. J Pharmacol Sci 115: 176–181.
  31. 31. Gonzalez MD (2005) Rate control vs. pulmonary vein isolation. Am J Geriatr Cardiol 14: 26–30.
  32. 32. Ter Keurs HE, Boyden PA (2007) Calcium and arrhythmogenesis. Physiol Rev 87: 457–506.
  33. 33. Arentz T, Haegeli L, Sanders P, Weber R, Neumann FJ, et al. (2007) High-density mapping of spontaneous pulmonary vein activity initiating atrial fibrillation in humans. J Cardiovasc Electrophysiol 18: 31–38.
  34. 34. Bers DM (2006) Altered cardiac myocyte Ca2+ regulation in heart failure. Physiology (Bethesda) 21: 380–387.
  35. 35. Roderick HL, Higazi DR, Smyrnias I, Fearnley C, Harzheim D, et al. (2007) Calcium in the heart: when it’s good, it’s very very good, but when it’s bad, it’s horrid. Biochem Soc Trans 35: 957–961.
  36. 36. Thul R, Coombes S, Bootman MD (2012) Persistence of pro-arrhythmic spatio-temporal calcium patterns in atrial myocytes: a computational study of ping waves. Front Physiol 3: 279.
  37. 37. Chen YJ, Chen YC, Wongcharoen W, Lin CI, Chen SA (2008) Effect of K201, a novel antiarrhythmic drug on calcium handling and arrhythmogenic activity of pulmonary vein cardiomyocytes. Br J Pharmacol 153: 915–925.
  38. 38. McCarron JG, Flynn ER, Bradley KN, Muir TC (2000) Two Ca2+ entry pathways mediate InsP3-sensitive store refilling in guinea-pig colonic smooth muscle. J Physiol 525 Pt 1: 113–124.
  39. 39. Okamoto Y, Takano M, Ohba T, Ono K (2012) Arrhythmogenic coupling between the Na+ -Ca2+ exchanger and inositol 1,4,5-triphosphate receptor in rat pulmonary vein cardiomyocytes. J Mol Cell Cardiol 52: 988–997.
  40. 40. Mackenzie L, Bootman MD, Laine M, Berridge MJ, Thuring J, et al. (2002) The role of inositol 1,4,5-trisphosphate receptors in Ca2+ signaling and the generation of arrhythmias in rat atrial myocytes. J Physiol 541: 395–409.
  41. 41. Eisner DA, Kashimura T, Venetucci LA, Trafford AW (2009) From the ryanodine receptor to cardiac arrhythmias. Circ J 73: 1561–1567.