Differences in gap junction channels between cardiac myocytes, fibroblasts, and heterologous pairs M. B. ROOK, A. C. G. VAN GINNEKEN, B. DE JONGE, A. EL AOUMARI, D. GROS, AND H. J. JONGSMA Department of Physiology, University of Amsterdam, 1105 AZ Amsterdam, The Netherlands; and Laboratoire de Biologie de la Differentiation Cellulaire, Centre National de la Recherche Scientifique, Unit6 Associei, 179, Marseille, France Rook, M. B., A. C. G. van Ginneken, B. de Jonge, A. El Aoumari, D. Gros, and H. J. Jongsma. Differences in gap junction channels between cardiac myocytes, fibroblasts, and heterologous pairs. Am. J. Physiol. 263 (Cell PhysioL. 32): C959-C977, 1992.-Cultures of neonatal rat heart cells contain predominantly myocytes and fibroblastic cells. Most abundant are groups of synchronously contracting myocytes, which are electrically well coupled through large gap junctions. Cardiac fibroblasts may be electrically coupled to each other and to adjacent myocytes, be it with low intercellular conductances. Nevertheless, synchronously beating myocytes interconnected via a fibroblast were present, demonstrating that nonexcitable cardiac cells are capable of passive impulse conduction. In fibroblast pairs as well as in myocyte-fibroblast cell pairs, no sensitivity to junctional voltage could be detected when transjunctional conductance was X-2 nS. However, in pairs coupled by a conductance of Cl nS, complex voltage-dependent gating was evident; gap junction channel open probability decreased with increasing junctional voltage but a nongated residual conductance remained at all voltages tested. Single gap junction channel conductance between fibroblasts was ~21 pS, very similar to an - WpS channel conductance that was found between myocytes next to the major conductance of 43 pS. Single-channel conductance in heterologous myocyte-fibroblast gap junctions was -32 pS, which matches the theoretical value of 29 pS for gap junction channels composed of a fibroblast connexon and the major myocyte connexon. A site-directed antibody against rat heart gap junction protein connexin43 recognized gap junctions between neonatal cardiomyocytes, as demonstrated by immunocytochemical labeling. In contrast, junctions between fibroblasts showed no labeling, while in myocyte-fibroblast junctions labeling occasionally was present. Our results suggest the existence of two gap junction proteins between neonatal rat cardiocytes, connexin43 and another yet unidentified connexin. An alternative explanation (cell-specific regulation of the conductance of connexin43 channels) is discussed. homologous junctions; heterologous junctions; voltage clamp; single-channel conductance; voltage sensitivity; immunofluorescence MOST TISSUES individual cells usually communicate directly with one another via gap junctions, membrane specializations providing a direct and nonselective pathway from one cell interior to another for ions and small molecules up to a size of -1 kDa (see Refs. 1 and 12 for reviews). In heart, gap junctions form the lowresistance pathways that are essential for impulse propagation and thus for coordinated contractions (see Ref. 32 for a recent review). Gap junctions are composed of a lattice of transmembrane channels. Each cell provides its own half of each of the channel structure, the hemichannel or connexon (36). By pairing in mirror symmetry with connexons in juxtaposed membranes, they form the individual gap junction channels. Connexons are composed of six proWITHIN

0363-6143/92

$2.00

Copyright

tein subunits delineating the aqueous pore of the channel proper (36). Ultrastructurally, gap junctions in all tissues can be recognized as regions of close membrane apposition between adjacent cells in which the membranes still are separated by a 2- to 4-nm gap. Despite their uniform morphology, it has been shown that gap junctional proteins isolated from various tissues differ in their apparent molecular mass, indicating tissuedependent differences. Recent molecular biological research has elucidated the amino acid composition and molecular mass of various gap junctional protein subunits in several tissues as predicted by their cDNA. In mammalian heart the predicted molecular mass of the major gap junctional protein is 43 kDa, whereas in liver two proteins coexist with predicted molecular masses of 32 and 26 kDa, respectively (see Ref. 1 for a review). Tissue and species differences of gap junctions are also reflected by differences in physiological behavior in response to changes in transcellular voltage, intracellular pH, and calcium concentrations (12). Moreover, the use of the double voltage-clamp technique on isolated cell pairs has revealed that single gap junction channel conductances may be closely related to the type of gap junctional protein. Measurements in mammalian heart cells suggest that the single-channel conductance of connexin43 (CX43) is 40-60 pS (4, 26, 28, 30, 35). Similar conductances have been reported for CX43 in other cell types, for instance in astrocytes (10). Despite the differences in molecular mass and amino acid composition, gap junctional proteins form a related family of proteins, now generally referred to as the connexins. The major tissue- and species-dependent differences are located in the cytoplasmic domain between the second and third transmembrane segments as well as in the cytoplasmic carboxy-terminal region. In contrast, the transmembrane and extracellular regions are highly conserved (see Ref. 1 for a review). The finding that the extracellular regions of different gap junctional proteins are highly similar explains the observations that different cell types in one tissue, or cells originating from different tissues, and in vitro even from different species, can be coupled to each other through gap junctions (8, 11, 13). Isolated embryonic and neonatal rat heart cells in vitro are spontaneously active and can reestablish intercellular contacts and thus electrical coupling (14,15,26). Mammalian heart tissue is composed of myocytes and nonmuscle cells, and consequently cultures of heart cells contain a mixed cell population in which, depending on isolation procedure, culture conditions, and time, myocytes and fibroblastic cells are the predominant cell

0 1992 the

American

Physiological

Society

c959

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (130.237.165.040) on December 12, 2018.

C960

GAP JUNCTIONS

BETWEEN

CARDIAC

types (31). In cultures of neonatal rat heart cells, synchronously beating myocytes frequently appear to be interconnected by nonexcitable fibroblastic cells. It has previously been demonstrated that cardiac myocytes and nonmuscle cells can be electrically coupled (13) and that gap junctions are present between the two cell types (11). In a preliminary report we showed that gap junctions are present between neonatal cardiac myocytes and fibrab!ast.s in r~~!tllrc!, that the unitary conductance of fibroblast gap junction channels was different from that of myocyte junctional channels, and that gap junction channels between heterologous pairs had an intermediate unitary conductance (27). We suggested that the differences in unitary conductances between the various cell combinations reflected different channel types. In the present study we attempt to further characterize gap junction channels between myocyte-, fibroblast-, and heterologous myocyte-fibroblast pairs by comparing their electrophysiological and immunocytochemical properties. METHODS

Preparation. Neonatal rat heart cells were isolated by collagenase dissociation and cultured for 24-48 h as described previously (14, 26). To increase the proportion of cardiac fibroblasts, the first plating step (normally used to minimize the number of these cells) was omitted. The cardiocytes were cultured in standard 60-mm plastic culture dishes (Falcon 3002). Before electrophysiological experiments, the culturing medium was replaced with serum-free N-2-hydroxyethylpiperazine-N’2-ethanesulfonic acid-buffered Ham’s FlO medium (Flow Laboratories) at room temperature, to which 1.7 mM CaCl, had been added, resulting in a final calcium concentration of 2.2 mM. The extra calcium ensured that most myocytes remained spontaneously active, even at room temperature. The cultures were observed with an inverted microscope (Nikon Diaphot TMD), fitted with phase-contrast optics, at a total magnification of X400. Electrophysiology. The electrophysiological properties of gap junction channels between fibroblast pairs (FF) and pairs consisting of a myocyte coupled to a fibroblast (MF) were studied with a double version of the whole cell patch-clamp technique as described previously (26, 28). In some experiments electrical coupling between a triplet of cells consisting of two myocytes interconnected by a fibroblast (MFM) was assessedby recording from the myocytes. Patch pipettes had tip resistances of 5-10 MQ as measured in the bath. After establishing gigaohm seals (X0 GR) and breaking the patch to obtain the whole cell recording configuration, access resistance from pipettes to cell interiors was estimated in current clamp from the initial fast-rising step in voltage displacements induced by injection of short 50-pA pulses in either cell. Access resistances ranged between 15 and 50 MR per pipette. No series-resistance compensation was applied. Junctional currents and single gap junction channel gating events were measured in voltage clamp. Each cell of a pair was clamped to identical potentials, resulting in a zero junctional current (Ij). Next, a transjunctional voltage (AVj) was applied by changing the potential in one cell in a stepwise manner to varying de- or hyperpolarized levels for different lengths of time while the voltage was kept constant in the other cell. During a AVj, a Ij will flow from the stepped cell into the neighbor cell. Under these circumstances the current recorded in the latter cell is a direct measure of Ij, because it is the current that is needed

MYOCYTES

AND FIBROBLASTS

to keep it at its holding potential. This compensating current therefore has the same amplitude but opposite polarity as Ij. Because cultured neonatal rat heart cells are small and have high membrane resistances (myocytes 2-10 GQ at resting potential, fibroblasts up to 30 GQ), the access resistance (R,) was always at least 20 times smaller than the input resistance of the cells. Thus, in poorly coupled cells (junctional conductance gj < 1 nS, or in terms of resistance Rj > 1 GQ), no correction for the R, was made and Rj was simply calculated by dividing AVj by Ij recorded in the cell held at constant potential. In well-coupled ~449, in which the mapitude of Rj approaches that of RS. the latter cannot be neglected. In these cases Rj ~8s calculated an equation we derived previously (28) R

(I,

J

‘=

l

R,,

AEJR,

-

IB

l

R&J

+

(A&

-

With

A&)

(1) -

IA

l

(1

+

R,,lR,)

in which 1A and 1B represent currents through electrodes A and B, R, and RSB denote R, from electrodes to cells A and B, RmA is membrane resistance of cell A and AE, and A& denote the change in potential of electrodes A and B. Current signals of both cells were recorded on videocassette recorder (VCR) tape using a pulse-code modulation system (Sony HF 150, which was modified to make it suitable for DC recording) in combination with a VCR recorder. The sample rate was 41 kHz/channel. The voltage recordings were fed into custom-built FM modulation-demodulation circuits and stored on tape via the stereochannels of the VCR recorder. After lowpass filtering (2-pole Butterworth filter) at 0.1-l kHz to minimize nonjunctional membrane noise, the current and voltage recordings were played back into a IBM-compatible PC/AT (Acer 910) equipped with an A/D board (PCL-718, Advantech). Digitized voltage and current recordings, current amplitude histograms and current-voltage (I-V) plots were plotted on a Hewlett-Packard ColorPro plotter. Determination of single gap junction channel conductance. In weakly coupled cell pairs, instantaneous opening and closing of gap junctional channels can be observed as steplike current transitions in Ij, which appear as current steps of equal amplitude but of opposite polarity in the current traces of each cell (23, 28). However, these fast current transitions often were superimposed on slow changes in the amplitude of Ij. This impaired an accurate determination of the single-channel conductance from amplitude histograms of Ij. We therefore adopted the following

approach:

the current

signals were played

back from

tape and recorded on paper with an ink-jet recorder (Siemens Elema Mingograph 34), which was calibrated to a sensitivity of 5 PA/cm. Next, the amplitudes of the current transitions in Ij were measured on a digitizing tablet (Morphomat 10, Zeiss) connected to a PDP 11/73 computer. From these data and the applied AVj, the conductance underlying each measured current transition was calculated and the results were compiled to stepamplitude histograms, which were fitted by Gaussian distributions using a nonlinear least-squares method. Electron microscopy. To test whether identification of the cells under the light microscope was correct, several FF pairs, MF pairs, and MFM combinations were prepared for electron microscopy. The cells were fixed in 0.15 M sodium-cacodylate buffer containing 2% glutaraldehyde (pH 7.4) at room temperature (20-22°C). Next, the preparation was rinsed with 0.15 M sodium-cacodylate buffer and postfixed in 2% 0~0, in water during 1 h at room temperature, followed by block staining in 2% magnesium-uranyl acetate in 50% ethanol. Dehydration was performed in graded ethanols, followed by embedding in Araldite D (Ciba-Geigy). During the fixation and embedding procedure the position of the cells was marked at the bottom of the culture dish as described previously (26). Serial ultrathin sections were cut on a Reichert Ultracut E microtome. The

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (130.237.165.040) on December 12, 2018.

GAP

JUNCTIONS

BETWEEN

CARDIAC

sections were mounted on Formvar coated slot grids, stained with magnesium-uranyl acetate and lead citrate, and viewed in a Philips 420 electron microscope. Western blotting of adult and cultured neonatal rat heart celZ fractions. An affinity purified site-directed antibody against residues 314-322 located in the cytoplasmic carboxy-terminal region of the rat heart gap junctional protein, connexin43 (3), was raised in rabbits as described previously (7). A comparative analysis of the connexin43 gap junction protein (CX43) in freeze-dried whole heart fraction of adult rats and in freezedried samples of neonatal heart cell cultures, were carried out by sodium dodecyl sulfate (SDS) -polyacrylamide gel electrophoresis under reducing conditions as described previously (6, 7). Molecular weights were estimated by reference to standard proteins (Bio-Rad) on gels stained with Coomassie Blue. For immunoblotting, samples fractionated by electrophoresis were transferred onto nitrocellulose membranes (0.22 pm; Schleicher & Schuell) at 25 V for 12-15 h in electrode buffer containing 0.02% SDS. Protein transfer was checked by staining the nitrocellulose membrane with Ponceau SS. These replicas were first saturated with BLOTTO (40 mM tris(hydroxymethyl)aminomethane-HCl, 0.1% Tween 20, and 4% non-fat dry milk; pH 7.5) and then incubated overnight at 4°C with the sitedirected anti-CX43 antibody (2 pg/ml in BLOTTO). Next, the immunoreplicas were treated with biotinylated goat anti-rabbit antibodies [F(ab’),; Jackson Immunoresearch Laboratories] and peroxidase-labeled streptavidin (Jackson Immunoresearch Laboratories) before detection of peroxidase activity with 4chloronaphto1, as described by DuPont et al. (6). Immunocytochemical labeling of gap junctions. Neonatal rat heart ventricles were fixed and permeabilized in a mixture of 35% methanol, 35% acetone, and 5% acetic acid in water at room temperature. The preparations were embedded in paraffin, and lo-pm microtome sections were cut according to standard methods. Sections were mounted on glass microscope slides coated with poly-L-lysine (Sigma), deparaffinized in xylene, rehydrated in graded alcohols, and thoroughly rinsed with calcium- and magnesium-free phosphate-buffered saline (PBS). Next, the preparations were incubated with anti-CX43 and labeled with goat anti-rabbit immunoglobulin (Ig) Gs conjugated to tetramethylrhodamine B isothiocyanate (TRITC) as described below. Finally, labeled sections were mounted in an antifading compound as described in the paragraph below. In neonatal heart cell cultures, the position of myocyte and fibroblast cell combinations to be investigated was marked on the bottom of the culturing dishes by means of a needle mounted on a micromanipulator. The cell cultures were rinsed with calcium- and magnesium-free PBS, pH 7.4, and were permeabilized and fixed in the mixture of 35% methanol, 35% acetone, and 5% acetic acid in water at room temperature. Alternatively, the cultures were fixed in 2% formaldehyde in PBS at 4OC. Formaldehyde-fixed cultures were permeabilized in PBS containing 0.05% saponin and 0.2% gelatin for 60 min, followed by a 30-min incubation in PBS containing 50 mM NH&l to quench free aldehyde groups. In all samples nonspecific binding sites were blocked by incubation in PBS containing 5% normal goat serum, Janssen Pharmaceutics) and 0.1% bovine serum albumin for 20 min. Next, the cultures were incubated for 1 h with anti-CX43, diluted in PBS to 6 pg/ml, and subsequently labeled with a goat anti-rabbit IgG antibody conjugated to TRITC (H + L chains, Jackson Immunoresearch Laboratories). All incubation steps were performed at room temperature, and between each step the samples were extensively rinsed with PBS. The marked regions in labeled cultures were mounted in situ under cover slips with a glycerol mounting solution containing 2% l&diazabicyclo-[2.2.2]octane (Sigma) antifading compound. The preparations were viewed

MYOCYTES

AND

FIBROBLASTS

C961

and photographed with a Nikon TMF microscope equipped for epifluorescence. Some cell cultures were prepared for immunogold electron microscopy. These specimen were fixed with formaldehyde, permeabilized, and incubated with anti-CX43 as described above. Next, the preparations were incubated with a goat anti-rabbit IgG antibody conjugated to %5-nm gold particles (Auro Probe EM, Janssen Pharmaceutics) for 3 h at room temperature. After several extensive washing steps in PBS, the preparations were briefly postfixed for 2 min in 2% 0~0, in water and treated further for electron microscopy as described in the previous paragraph. In control experiments, the incubation step with anti-CX43 was replaced by incubation with preimmune serum. RESULTS

Electrical coupling between FM combinations. Figure 1A shows a typical example of the cultures used in this study. The majority of myocytes were spontaneously active and MM pairs showed synchronized contractions. Myocytes apparently in contact with a fibroblast (MF pairs) were frequently present. Occasionally, triplets of cells were encountered consisting of two synchronously beating myocytes interconnected via a fibroblast (Fig. 1A inset, MFM). Fibroblasts (F) could be discriminated from myocytes (M) by their translucent appearance and the absence of contractile activity. Electrical coupling between myocyte (M) and fibroblast (F) cell combinations could readily be demonstrated under current-clamp conditions as shown in Fig. 1, B-E. Injection of depolarizing current pulses (50-75 pA/lO ms) in one cell of a MM pair elicited action potentials in each cell, which showed no appreciable delay (Fig. 1B). Membrane potentials measured in fibroblasts varied between -20 and -40 mV, which was considerably less negative than the resting potential in myocytes, which was -60 to -80 mV. By contrast, the membrane potential of the fibroblasts in electrically coupled MF pairs generally approached, or was identical with, the resting potential of the myocyte, a phenomenon that in itself is an indication of electrical coupling. In most MF pairs, action potentials elicited in the myocyte caused substantial but slow depolarizations in the fibroblast (not shown, but see Ref. 27). Two of seven MF cell pairs appeared to be so well coupled that action potentials of the myocyte caused voltage deflections in the fibroblast that had the appearance of low-pass filtered action potentials. This is demonstrated in Fig. lC, which shows a recording of the voltage response to a current stimulus applied to the fibroblast. In MFM triplets, action potentials recorded from the myocytes often showed an appreciable delay, which could be as large as several tens of milliseconds (Fig. ID). In apparently well-coupled MFM cells there was no delay between the action potentials (Fig. 1E). The double voltage-clamp technique allows for the determination of the intercellular conductance between ceil pairs from the relationship between the amplitude of the applied AVj step and the measured 1j. We have previously shown that the intercellular conductance between synchronized MM pairs generally is in the order of several tens of nanosiemens (26). Fibroblasts are extremely flat cells, and it proved to be difficult to perform reliable and

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (130.237.165.040) on December 12, 2018.

C962

GAP JUNCTIONS

BETWEEN

CARDIAC

MYOCYTES

AND FIBROBLASTS

B MM

-58 mV

MFM

Fig. 1. A: phase-contrast photomicrographs of 24-h-old culture of neonatal rat heart cells, showing typical examples of cell combinations used in this study. Fibroblasts (F) can be distinguished from myocytes (M) by their translucent appearance and their lack of contractile activity. Pair of myocytes is discernible in center of photograph. In top left-hand corner, 2 fibroblasts can be recognized; bottom right-hand corner shows myocyte-fibroblast cell pair. Nuclei of some cells are indicated by n. Inset: 2 myocytes (which were contracting synchronously at time of exposure) interconnected by fibroblast. B-E: examples of current-clamp recordings demonstrating electrical coupling between various cell combinations. B: pair of well-coupled myocytes (MM) exhibiting synchronous action potentials in response to 50 pA/lO ms stimulus delivered to 1 cell. C: well-coupled MF pair; action potentials of myocyte, caused by injection of 100 pA/lO ms pulse in fibroblast, were accompanied by action potential-like voltage changes of fibroblast. D: recording from myocytes of poorly coupled myocyte-fibroblast-myocyte triplet (MFM). One myocyte (top truce) was stimulated with 50 pA/lO ms depolarizing pulse. Action potentials were propagated through (inexcitable) fibroblast with delay of -50 ms. E: electrically well-coupled MFM triplet. Action potentials elicited in 1 myocyte (top trace) caused action potentials in other myocyte without appreciable delay.

stable patch recordings for longer periods of time. Nevertheless, from selected experiments in which recording conditions were satisfactory, it appeared that gj between FF and MF pairs was variable and comparatively low. In six selected experiments with FF cell pairs, the gj ranged between 175 pS and 6 nS. gj between seven MF pairs ranged between 310 pS and 8 nS. Finally, gj between six MFM triplets showing synchronous contracting myocytes (it should be noted that such triplets have two MF junctions in series) ranged between 150 pS and 3 nS. Membrane properties and input resistance of fibroblasts.

From the results presented above, it is clear that gj between pairs of fibroblasts and myocyte-fibroblast pairs generally is low; maximally -8 nS in the 24- to 48-h cultures we have studied. Nevertheless, such relatively low gj values allowed effective action potential propagation between two myocytes that were interconnected through a fibroblast (Fig. 1E). This suggests that the R, of fibroblasts must be very high. To test this, we determined the membrane characteristics of single fibroblasts under voltage-clamp conditions. A typical example is presented in Fig. 2. In Fig. 2A the membrane current in response to various de- or hyperpolarizing voltage steps from holding potential (-50 mV) is shown. From these current traces it is clear that no inward currents were present, demonstrating the passive membrane properties of these cells. At depolarizations larger than -20 mV (membrane potential) a transient outward current developed. In Fig. 2B peak- and steady-state membrane cur-

rents are plotted as function of the membrane potential. From these I,- V, relationships it is clear that fibroblasts have a very high input resistance indeed. The steadystate current indicates a R, ranging between 25 GQ at voltages negative to -20 mV and 5 GO at more depolarized potentials. Even during transient peak currents the R, still was as high as -3 GQ. Voltage sensitivity of FF and MF gap junctions. As shown before (28,35), gap junctions between weakly coupled neonatal rat cardiomyocytes reveal voltage sensitivity at AVj values exceeding 50 mV. To assess to what extent gap junctions between FF and MF cell pairs showed voltage sensitivity, Ij was measured at the beginning (instantaneous Ij) and the end (quasi-steady-state 1j) of AVj steps with a duration of 2 s and of various amplitude and polarity. From these experiments it appeared that when gj between FF and MF pairs exceeded l-2 nS, the gap junctions apparently behaved like ohmic conductors. Examples are given in Fig. 3A, in which the left panel concerns a FF pair and the right panel a MF pair. The Ij-AVj relationships were virtually linear over the entire voltage range tested. Moreover, in both cases the absence of a time- and voltage-dependent decay in Ij is clear from the fact that at all voltages the amplitudes of the instantaneous and quasi-steady-state Ij values were equal to each other. In contrast to the junctions described above, gap junctional voltage dependence was evident in weakly coupled FF and MF pairs where gj was 3O-80 s are required for the acquisition of a sufficient number of channel transitions to obtain representative distributions of Pk. Especially at higher voltages the dwell times of Ij at current levels representing the probabilities to find larger frac-

MYOCYTES

AND

FIBROBLASTS

c975

tions of the channel population to be open, which because of the voltage-dependent nature of the gating process are rare, could be underestimated. This would result in underestimation of the P, of the individual channels, which might be an explanation for some of the exceedingly low & values derived from the binomial model at AV’ values of 100 mV. Nevertheless, this study demonstrates that in preparations where the binomial model can be used, it provides a powerful method to unravel the intricacies of gap junction channel gating. From the points raised in the discussion above, it is clear that the gating behavior of cardiac gap junction channels is complex and that conventional methods to describe their voltage sensitivity might obscure important details of this complexity. Different gap junction Tj, evidence for different channel types ? Recent research has provided strong evidence that

unitary conductances of gap junction channels are closely related to the type of gap junctional protein (connexin) involved. cDNA coding for rat liver connexin32 (CX32) and human and rat heart CX43 has been expressed in coupling-incompetent hepatoma cell lines. In this expression system, unitary channel conductance measurements indicated one single “/j of 120-160 pS for liver CX32 gap junction channels (20) and two distinct “/j values of -60 and 90 pS for human and rat heart CX43 channels (9,22, 29). Rat CX43 is a phosphoprotein with three putative protein kinase C phosphorylation sites and possibly a guanosine 3’,5’-cyclic monophosphate (cGMP)-dependent phosphorylation site (9), and the macroscopic junctional conductance of CX43 gap junctions can be modulated by several agonists that affect different phosphorylation pathways (see Ref. 1 for a review). Interestingly, on the single-channel level there is experimental evidence that the 60 pS and 90 pS conductances found for CX43 channels reflect phosphorylated and dephosphorylated states of the channel protein (22, 35a). In addition, stimulation of the cGMP-dependent phosphorylation pathway shifts “/j from 43 (or 60) to 21 (or 30) pS1 in cultured neonatal rat cardiomyocytes (%a). The notion that changes in phosphorylation state of CX43 gap junction channels may shift their Tj could indicate that the --20-pS conductance class found in MM and FF cell pairs reflects a cGMP-dependent phosphorylated form of CX43. This would require that cGMP phosphorylation of CX43 in fibroblasts is extremely stable, even when fibroblasts are coupled to myocytes. On the other hand, although CX43 is still considered to be the major gap junctional protein in mammalian heart, there is increasing molecular biological evidence of other cardiac gap junction gene products. Evidence for the expression of two other connexins, CX40 and CX45, has recently been reported for canine heart (17). Considering this, our finding of a single class of 21-pS gap junction channels in fibroblastic cells (Fig. 7), and the coexistence of a similar conductance class of -18 pS with 43 pS l It should be mentioned here that the slightly different values reported for yj of CX43 gap junction channels (43 vs. 60 pS) arise from differences in experimental conditions. When instead of K-gluconate, KC1 or CsCl is used as a major constituent in the pipette filling solution, yj values in cardiac myocytes shift from -20 and 43 pS to -30 and 60 PS (28).

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (130.237.165.040) on December 12, 2018.

C976

GAP

JUNCTIONS

BETWEEN

CARDIAC

CX43 channels in myocytes, could indicate the presence of a rat isoform of one of the newly discovered canine cardiac connexins. Based on theoretical considerations (see RESULTS) two values of, respectively, -29 and 20 pS are to be expected between MF pairs. However, generally only one conductance of 32 pS was found, which is in good agreement with the theoretically expected 29 pS conductance. On the other hand, when the data were selected with respect to the applied AVj, double Gaussian distributions could be fitted through the “/j histograms constructed from the data recorded at 75 and 100 mV, although the low conductance peak was not as well defined as in the MM experiments (Fig. 7). Probably separation into two conductance classes was hampered by the small difference between the two expected values of “/j in MF pairs. Although the existence of the expected lower “/j value remains controversial, the major 32 pS “/j found in MF pairs shows that the M and F connexons composing heterologous gap junctions, regardless of their connexin type, retain their cell-specific conductance. Immunocytochemistry of myocyte and fibroblastic gap junctions; are they composed of different gap junctional proteins ? The site-directed antibody against residues 314-322 in the carboxy-terminal cytoplasmic tail of adult rat heart CX43 clearly recognizes gap junctions between neonatal cardiomyocytes. This is demonstrated by the immunoblots shown in Fig. 9A, by the extensive immunofluorescent labeling in the junctional regions between neonatal rat heart ventricle cells in situ (Fig. 9B), as well as in culture (Fig. 10, A-C), and on the ultrastructural level by the immunogold labeling of morphologically identified myocardial gap junctions (Fig. 10E). These findings are in agreement with reports on the identification of CX43 in mammalian heart (2, 6, 7, 40), as well as in cultures of neonatal rat heart cells (18). Thus CX43 is indeed an important constituent protein in neonatal rat cardiomyocyte gap junctions. Interestingly, many of the cardiomyocytes in culture also showed fluorescent spots in nonjunctional regions and the cytoplasm (Fig. 10, A, C, and D), which could indicate the presence of pools of gap junction channels or their prec lrsors or internalized gap junctional material. Although fibroblastic cells from various origin can be electrically coupled to each other (13, 24, 27; current study) and morphologically detectable gap junctions are present between these cells (25; current study), little is known about their gap junctional proteins. There are some reports on the expression of CX43 in cultured human lung fibroblasts and in a mouse fibroblastic cell line as evidenced by Northern blot analysis (19) and on immunofluorescent staining of connective tissue cells in rat ovaria and cornea (2). In contrast, no immunofluorescent labeling could be detected in the junctional zones between fibroblastic cells in our cultures (Fig. lo@. From our electron microscopic results, it can be argued that gap junctional contacts between fibroblasts are too small to yield an immunocytochemical signal above background. On the other hand, we occasionally found fluorescent labeling of MF junctions (Fig. lOC), despite the fact that on the ultrastructural level the size (50-200 nm Tj

MYOCYTES

AND

FIBROBLASTS

diam) of heterologous gap junctions between MF pairs seemed to be similar to those between FF pairs. Interestingly, Laird and Revel (18) also mentioned a low incidence of labeling between MF pairs and absence of label between cardiac fibroblasts in a study in which several site-directed antibodies were used to assess the topology of CX43 in the junctional membrane of cultured cardiocytes. In conclusion, our results clearly show that cardiac fibroblast and myocyte gap junction channels exhibit a very cell-specific distribution of 20- and 43-pS singlechannel conductances but otherwise have an identical, although complex, voltage-dependent gating behavior. Although the data suggest that the 20 pS conductance represents a connexin type that is different from CX43, a cell-specific regulation of conductance substates in an exclusive population of CX43 channels can not be ruled out at present. It remains to be seen whether functional gap junctions are present between myocytes and nonexcitable fibroblastic cells in vivo. Such interactions might be important in regions of cardiac tissue where fibroblasts are abundant, such as in the sinoatrial node, the atrioventricular node, or in cardiac tissue recovering from infarction. We thank L. Tsjernina, A. A. Meyer, and A. W. Schreurs for technical help and advice and M. S. J. Overzier, P. A. Lowie, and J. J. Nunumete for processing the photographic material. We also thank the Department of Advanced Electron Microscopy for making available their facilities for the ultrastructural part of this investigation. We are grateful to Dr. D. C. Spray for comments and suggestions on the manuscript. This work was financially supported by Grant 86.030 from the Dutch Heart Foundation (NHS) to M. B. Rook, and by Grants 90.055 from Direction des Recherches Etudes et Techniques, 88-50-09 from Institut National de la Sante et de la Recherche Medicale, and a grant from the Fondation pour la Recherche Medicinale to D. Gros. Address for reprint requests: H. J. Jongsma, Fysiologisch Laboratorium, Academisch Medisch Centrum, Merbergdreef 15, 1105 AZ Amsterdam, The Netherlands. Received

21 June

1991;

accepted

in final

form

19 May

1992.

REFERENCES 1. Bennett, M. V. L., L. C. Barrio, T. A. Bargiello, D. C. Spray, E. Hertzberg, and J. C. Saez. Gap junctions: new tools, new answers, new questions. Neuron 6: 305-320, 1991. 2. Beyer, E. C., J. Kistler, D. L. Paul, and D. A. Goodenough. Antisera directed against connexin43 peptides react with a 43 kD protein localized to gap junctions in myocardium and other tissues. J. Cell BioL. 108: 595-605, 1989. 3. Beyer, E. C., D. L. Paul, and D. A. Goodenough. Connexin43: a protein from rat heart homologous to gap junction protein from liver. J. Cell BioL. 105: 2621-2629, 1987. 4. Burt, J. M., and D. C. Spray. Single channel events and gating behavior of the cardiac gap junctional channel. Proc. Natl. Acad. Sci. USA 85: 3431-3434, 1988. 5. Colquhoun, D., and A. G. Hawkes. The principles of the stochastic interpretation of ion-channel mechanisms. In: SingleChannel Recording, edited by B. Sakmann and E. Neher. New York: Plenum, 1983, p. 135-177. 6. DuPont, E., A. El Aoumari, S. Roustiau-Severe, J. P. Briand, and D. Gros. Immunological characterization of rat cardiac gap junctions: presence of common antigenic determinants in heart of other vertebrate species and in various organs. J. Membr. Biol. 104: 119-128, 1988. 7. El Aoumari, E. DuPont, H. Reggio, A., C. Fromaget, P. Durbec, J.-P. Briand, K. Boller, B. Kreitman, and D. Gros. Conservation of a cytoplasmic carboxy-terminal domain

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (130.237.165.040) on December 12, 2018.

GAP

8. 9.

10.

11. 12. 13.

14. 15.

16. 17. 18. 19. 20. 21. 22.

23. 24. 25. 26.

JUNCTIONS

BETWEEN

CARDIAC

of connexin 43, a gap junctional protein, in mammal heart and brain. J. Membr. Biol. 115: 229-240, 1990. Epstein, M, L,, and N. B. Gilula. A study of communication specificity between cells in culture. J. Cell Biol. 75: 769-787, 1977. Fishman, G. I., D. C. Spray, and L. A. Leinwand. Molecular characterization and functional expression of the human cardiac gap junction channel. J. Cell BioL. 111: 589-598, 1990. Giaume, C., C. Fromaget, A. El Aoumari, J. Cordier, J. Glowinski, and D. Gros. Gap junctions in cultured astrocytes: single channel currents and characterization of channelforming protein. Neuron 6: 133-143, 1991. Goshima, K. Formation of nexus and electrotonic transmission between myocardial and FL cells in monolayer culture. Exp. Cell Res. 63: 124-130, 1970. Hertzberg, E. L., and R. G. Johnson. (Editors). Gap junctions. In: Modern CelZ Biology. New York: Liss, 1988, vol. 7. Hyde, A., B. Blondel, A. Matter, J. P. Cheneval, B. Filloux, and L. Girardier. Homo- and heterocellular junctions in cell cultures: an electrophysiological and morphological study. In: Progress in Brain Research, edited by K. Akert and P. G. Waser. Amsterdam: Elsevier, 1969, vol. 31, p. 283-311. Jongsma, H. J., M. Masson-P&et, and L. Tsjernina. The development of beat-rate synchronization of rat heart myocytes pairs in culture. Basic Res. Cardiol. 82: 454-464, 1987. Jongsma, H. J., L. Tsjernina, and J. De Bruijne. The establishment of regular beating in populations of pacemaker cells. A study with tissue-cultured heart cells. J. Mol. Cell. Cardiol. 15: 123-133, 1983. Jongsma, H. J., and H. E. Van Rijn. Electrotonic spread of current in monolayer cultures of neonatal rat heart cells. J. Membr. Biol. 9: 341-360, 1972. Kanter, H. L., J. E. Saffitz, and E. C. Beyer. Cardiac myocytes express multiple gap junction proteins. Circ. Res. 70: 438444, 1992. Laird, D. W., and J.-P. Revel. Biochemical and immunological analysis of the arrangement of connexin43 in rat heart gap junction membranes. J. Cell Sci. 97: 109-117, 1990. Larson, D. M., C. C. Haudenschild, and E. C. Beyer. Gap junction messenger RNA expression by vascular wall cells. Circ. Res. 66: 1074-1080, 1990. Moreno, A. P., B. Eghbali, and D. C. Spray. Connexin32 gap junction channels in stably transfected cells: unitary conductance. Biophys. J. 60: 1254-1266, 1991. Moreno, A. P., B. Eghbali, and D. C. Spray.Connexin32 gap junction channels in stably transfected cells: equilibrium and kinetic properties. Biophys. J. 60: 1267-1277, 1991. Moreno, A. P., G. I. Fishman, and D. C. Spray. Phosphorylation shifts unitary conductance and modifies voltage dependent kinetics of human connexin43 gap junction channels. Biophys. J. 62: 51-53, 1992. Neyton, J., and A. Trautmann. Single channel currents of an intercellular junction. Nature Land. 317: 331-335, 1985. O’Lague, P., and H. Dalen. Low resistance junctions between normal and between virus transformed fibroblasts in tissue culture. Exp. Cell Res. 86: 374-382, 1974. Pinto da Silva, P., and N. B. Gilula. Gap junctions in normal and transformed fibroblasts in culture. Exp. Cell Res. 71: 393-401, 1972. Rook, M. B., B. de Jonge, H. J. Jongsma, and M. A. Masson-P&et. Gap junction formation and functional interac-

MYOCYTES

27.

28.

29.

30. 31.

32. 33.

AND

FIBROBLASTS

c977

tion between neonatal rat cardiocytes in culture: a correlative physiological and ultrastructural study. J. Membr. Biol. 118: 179192, 1990. Rook, M. B., H. J. Jongsma, and B. de Jonge. Single channel currents of homo- and heterologous gap junctions between cardiac fibroblasts and myocytes. Pfluegers Arch. 414: 95-98, 1989. Rook, M. B., H. J. Jongsma, and A. C. G. Van Ginneken. Properties of single channels between isolated neonatal rat heart cells. Am. J. Physiol. 255 (Heart Circ. Physiol. 24): H770-H782, 1988. Rook, M. B., A. P. Moreno, G. I. Fishman, and D. C. Spray. Rat and human connexin43 (Cx43) gap junction channels in stably transfected cells: comparison of unitary conductance and voltage sensitivity (Abstract). Biophys. J. 61: A505, 1992. Rudisuli, A., and R. Weingart. Electrical properties of gap junction channels in guinea-pig ventricular cell pairs revealed by exposure to heptanol. Pfluegers Arch. 415: 12-21, 1989. Severs, N. J. Constituent cells of the heart and isolated cell models in cardiovascular research. In: Isolated Adult Cardiomyocytes, edited by H. M. Piper and G. Isenberg. Boca Raton FL: CRC, 1989, vol. 1, p. 3-41. Severs, N. J. The cardiac gap junction and intercalated discs. Int. J. Cardiol. 26: 137-173, 1990. Spray, D. C., A. L. Harris, and M. V. L. Bennett. Equilibrium properties of a voltage-dependent junctional conductance. J. Gen.

Physiol.

77: 77-93,

1981.

34. Spray, D. C., A. P. Moreno, B. Eghbali, M. Chanson, and G. I. Fishman. Gating of gap junction channels as revealed in cells stably transfected with wild type and mutant connexins. Biophys. J. 62: 48-50, 1992. 35. Takens-Kwak, B. R., H. J. Jongsma, M. B. Rook, and A. C. G. van Ginneken. Mechanism of heptanol-induced uncoupling of cardiac gap junctions: a perforated patch-clamp study. Am. J. Physiol. 262 (Cell Physiol. 31): C1531-C1538, 1992. uua.Takens-Kwak, B. R., and H. J. Jongsma. Cardiac gap junctions: three distinct single channel conductances and their modulation by phosphorylating treatments. Pfluegers Arch. In press. 36. Unwin, P. N. T., and G. Zampighi. Structure of the junction between communicating cells. Nature Land. 283: 545-549, 1980. 37. Veenstra, R. D. Developmental changes in regulation of embryonic chick heart gap junctions. J. Membr. Biol. 119: 253-265, 1991. 38. Weingart, R. Electrical properties of the nexal membrane studied in rat ventricular cell pairs. J. Physiol. Land. 370: 267-284, 1986. ;JBa.Wilders, R., and H. J. Jorgsma. Limitations of the dual voltage clamp method in assaying conductance and kinetics of gap junction channels. Biophys. J. In press. 39. White, R. L., D. C. Spray, A. C. Campos De Carvalho, B. A. Wittenberg, and M. V. L. Bennett. Some physiological and pharmacological properties of cardiac myocytes dissociated from adult rat. Am. J. Physiol. 249 (Cell Physiol. 18): C447-C455, 1985. 40. Yancey, S. B., S. A. John, R. Lal, B. J. Austin, and J.-P. Revel. The 43-kD polypeptide of heart gap junctions: immunolocalization, topology and functional domains. J. Cell Biol. 108: 2241-2254, 1989.

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (130.237.165.040) on December 12, 2018.