Measurement of single channel currents from cardiac gap junctions.

Cardiac gap junctions consist of arrays of integral membrane proteins joined across the intercellular cleft at points of cell-to-cell contact. These junctional proteins are thought to form pores through which ions can diffuse from cytosol to cytosol. By monitoring whole-cell currents in pairs of embryonic heart cells with two independent patch-clamp circuits, the properties of single gap junction channels have been investigated. These channels had a conductance of about 165 picosiemens and underwent spontaneous openings and closings that were independent of voltage. Channel activity and macroscopic junctional conductance were both decreased by the uncoupling agent 1-octanol.

Molecular cloning and functional expression of human connexin37, an endothelial cell gap junction protein.

Gap junctions allow direct intercellular coupling between many cells including those in the blood vessel wall. They are formed by a group of related proteins called connexins, containing conserved transmembrane and extracellular domains, but unique cytoplasmic regions that may confer connexin-specific physiological properties. We used polymerase chain reaction amplification and cDNA library screening to clone DNA encoding a human gap junction protein, connexin37 (Cx37). The derived human Cx37 polypeptide contains 333 amino acids, with a predicted molecular mass of 37,238 D. RNA blots demonstrate that Cx37 is expressed in multiple organs and tissues (including heart, uterus, ovary, and blood vessel endothelium) and in primary cultures of vascular endothelial cells. Cx37 mRNA is coexpressed with connexin43 at similar levels in some endothelial cells, but at much lower levels in others. To demonstrate that Cx37 could form functional channels, we stably transfected communication-deficient Neuro2A cells with the Cx37 cDNA. The induced intercellular channels were studied by the double whole cell patch clamp technique. These channels were reversibly inhibited by the uncoupling agent, heptanol (2 mM). The expressed Cx37 channels exhibited multiple conductance levels and showed a pronounced voltage dependence. These electrophysiological characteristics are similar to, but distinct from, those of previously characterized connexins.

Monovalent ion selectivity sequences of the rat connexin43 gap junction channel.

The relative permeability sequences of the rat connexin 43 (rCx43) gap junction channel to seven cations and chloride were examined by double whole cell patch clamp recording of single gap junction channel currents in rCx43 transfected neuroblastoma 2A (N2A) cell pairs. The measured maximal single channel slope conductances (gammaj, in pS) of the junctional current-voltage relationships in 115 mM XCI were RbC1 (103) > or = CsC1 (102) > KC1 (97) > NaC1 (79) > or = LiC1 (78) > TMAC1 (65) > TEAC1 (53) and for 115 mM KY were KBr (105) > KC1 (97) > Kacetate (77) > Kglutamate (61). The single channel conductance- aqueous mobility relationships for the test cations and anions were linear. However, the predicted minimum anionic and cationic conductances of these plots did not accurately predict the rCx43 channel conductance in 115 mM KC1. Instead, the conductance of the rCx43 channel in 115 mM KC1 was accurately predicted from cationic and anionic conductance-mobility plots by applying a mobility scaling factor Dx/Do, which depends upon the relative radii of the permeant ions to an estimated pore radius. Relative permeabilities were determined for all of the monovalent catious and anions tested from asymmetric salt reversal potential measurements and the Goldman-Hodgkin-Katz voltage equation. These experiments estimate the relative chloride to potassium permeability to be 0.13. The relationship between the relative cation permeability and hydrated radius was modeled using the hydrodynamic equation assuming a pore radius of 6.3 +/- 0.4 A. Our data quantitatively demonstrate that the rCx43 gap junction channel is permeable to monovalent atomic and organic cations and anions and the relative permeability sequences are consistent with an Eisenman sequence II or I, respectively. These predictions about the rCx43 channel pore provide a useful basis for future investigations into the structural determinants of the conductance and permeability properties of the connexin channel pore.

Monovalent cation permeation through the connexin40 gap junction channel. Cs, Rb, K, Na, Li, TEA, TMA, TBA, and effects of anions Br, Cl, F, acetate, aspartate, glutamate, and NO3.

The unitary conductances and permeability sequences of the rat connexin40 (rCx40) gap junction channels to seven monovalent cations and anions were studied in rCx40-transfected neuroblastoma 2A (N2A) cell pairs using the dual whole cell recording technique. Chloride salt cation substitutions (115 mM principal salt) resulted in the following junctional maximal single channel current-voltage relationship slope conductances (gamma 1 in pS): CsC1 (153), RbC1 (148), KC1 (142), NaC1 (115), LiC1 (86), TMAC1 (71), TEAC1 (63). Reversible block of the rCx40 channel was observed with TBA. Potassium anion salt gamma j are: Kglutamate (160), Kacetate (160), Kaspartate (158), KNO3 (157), KF (148), KC1 (142), and KBr (132). Ion selectivity was verified by measuring reversal potentials for current in rCx40 gap junction channels with asymmetric salt solutions in the two electrodes and using the Goldman-Hodgkin-Katz equation to calculate relative permeabilities. The permeabilities relative to Li+ are: Cs+ (1.38), Rb+ (1.32), K+ (1.31), Na+ (1.16), TMA+ (0.53), TEA+ (0.45), TBA+ (0.03), Cl- (0.19), glutamate+ (0.04), and NO(3)- (0.14), assuming that the monovalent anions permeate the channel by forming ion pairs with permeant monovalent cations within the pore thereby causing proportionate decreases in the channel conductance. This hypothesis can account for why the predicted increasing conductances with increasing ion mobilities in an essentially aqueous channel were not observed for anions in the rCx40 channel. The rCx40 effective channel radius is estimated to be 6.6 A from a theoretical fit of the relationship of relative permeability and cation radius.

Localization and distribution of gap junctions in normal and cardiomyopathic hamster heart.

Gap junctions in mammalian heart function to provide low-resistance channels between adjacent cells for passage of ions and small molecules. It is clear that the almost unrestricted passage of ions between cells, ionic coupling, is required for coordinate and synchronous contraction. This knowledge of gap junction function has made it important to study their properties in normal and abnormal tissues. In the present study, we analyzed gap junction distribution in normal and cardiomyopathic heart tissue utilizing immunofluorescent and electron microscopy techniques. Frozen, unfixed sections of age-matched normal and cardiomyopathic cardiac tissues were immunofluorescently stained using an antibody directed against a specific peptide sequence of the connexin-43 gap junction protein. These studies revealed a characteristic punctate staining pattern for the intercalated discs in normal tissues. Some of the intercalated discs in cardiomyopathic hearts appeared to stain normally; however, others stained diffusely. The pixel intensity distribution of the confocal images demonstrated a marked difference of up to 90% increase in the number of pixels in cardiomyopathic myocardium (CM), yet the pixel intensity of gap junctions had a decrease of approximately 60%. This suggests the possibility that connexin-43 is present in CM cells in significant quantity; however, it does not become localized on the membranes as in normal cells. Electron-microscopic findings corroborate these observations on CM cells by showing an irregular distribution of intercalated discs relatively smaller in size with abnormal orientation and distribution.

Cardiac intercellular communication: consequences of connexin distribution and diversity.

Gap junctions contain channels which allow the exchange of ions and small molecules between adjacent cells. In the heart, these channels are crucial for normal intercellular current flow and the propagation of action potentials throughout the myocardium. Molecular cloning studies have demonstrated that these channels are formed by members of a family of related proteins called connexins each containing conserved and unique regions. There are several consequences of this multiplicity of connexins. Multiple connexins are expressed in differing, but sometimes overlapping, distributions within cardiovascular and other tissues. Connexin40, connexin43, and connexin45 are all found in cardiac myocytes, but their abundance differs in specialized cardiac regions with disparate conductive properties. Individual connexins form channels with differing voltage-dependence, conductance, and permeability properties, as demonstrated by functional expression of the cloned sequences. Connexins differ in their modification by phosphorylation, which may contribute to physiological regulation of intercellular communication. Expression of multiple connexins may lead to the formation of multiple channel types in a single tissue or cell and potentially allows mixing to form heterotypic and/or heteromeric channels. Thus, multiple connexins may contribute to the differences in intercellular resistance in cardiac regions with differing conductive properties and possibly may allow differences in the signalling molecules that pass between cells.

Voltage clamp limitations of dual whole-cell gap junction current and voltage recordings. I. Conductance measurements.

Previous correction methods for series access resistance errors in the dual whole-cell configuration did not take into account the effect of nonzero resting potentials (E(rest)) and junctional reversal potentials (E(rev)). Dual whole-cell currents were modeled according to resistor-circuit analysis and two correction formulas for the measurement of junctional currents (I(j)) were assessed. The equations for I(j), derived from Kirchoff's law before and after baseline subtraction of the nonjunctional current, were assessed for accuracy under a variety of whole-cell patch-clamp recording conditions. Both equations accurately correct for dual whole-cell voltage-clamp errors provided that the cellular parameters are included in the nonbaseline subtracted I(j) derivations. Junctional conductance (g(j)) estimates are most reliable at high junctional resistance (R(j)) values and minimize the need for corrective methods based on electrode series and cellular input resistances (R(el) and R(in)). In the "open-cell" configuration, low R(j) values relative to R(in) are required for accurate g(j) estimates. These methods provide the basis for accurate quantitative measurements of junctional resistance (or conductance) of gap junction channels or connexin hemichannels in the dual whole-cell or open-cell configurations. Revaluation of V(j)-dependent gating of rat connexin40 g(j) produced nearly identical Boltzmann fits to previously published data. Continuous g(j)-V(j) curves generated by variable slope V(j) ramps provide for more accurate fits and assessment of the time-dependence of the half-inactivation voltage and net gating charge movement.

Size and selectivity of gap junction channels formed from different connexins.

Gap junction channels have long been viewed as static structures containing a large-diameter, aqueous pore. This pore has a high permeability to hydrophilic molecules of approximately 900 daltons in molecular weight and a weak ionic selectivity. The evidence leading to these conclusions is reviewed in the context of more recent observations primarily coming from unitary channel recordings from transfected connexin channels expressed in communication-deficient cell lines. What is emerging is a more diverse view of connexin-specific gap junction channel structure and function where electrical conductance, ionic selectivity, and dye permeability vary by one full order of magnitude or more. furthermore, the often held contention that channel conductance and ionic or molecular selectivity are inversely proportional is refuted by recent evidence from five distinct connexin channels. The molecular basis for this diversity of channel function remains to be identified for the connexin family of gap junction proteins.

Voltage-dependent gating of gap junction channels in embryonic chick ventricular cell pairs.

The dependence of macroscopic gap junctional conductance (Gj) on transjunctional voltage (Vj) was studied in paired myocytes after enzymatic dissociation of 7-day-old embryonic chick ventricles. The membrane voltage of both cells was independently controlled by separate patch-clamp circuits in the whole cell configuration. Two distinctive unitary junctional conductances were identified in recordings from seven different cell pairs. The larger channel had a mean conductance of 166 +/- 37 pS (n = 6 pairs), whereas a second channel averaged 58 +/- 10 pS (n = 3). Instantaneous Gj remained linear over a Vj range of -100 to +100 mV, whereas the steady-state Gj declined when voltages exceeded +/- 30 mV. Both decay and recovery phases of Gj follow exponential time courses, with the recovery time constant being four times slower than inactivation, requiring 1.1 s at 80 mV. The normalized steady-state Gj-Vj curve could be defined by a two-state Boltzmann distribution, assuming an effective gating charge of 1.72, a half-inactivation voltage of 45 mV, and a residual voltage-insensitive Gj of 27% of maximum. Single-channel recordings revealed closure of 160-pS channels on a Vj step to 80 mV, and the ensemble average of five such records produced an exponentially decaying junctional current with a time constant of 184 ms. The single-channel current-voltage relationship remains linear with a slope of 145 pS over the entire Vj range. The results support the hypothesis that a population of 160-pS gap junction channels is gated by transjunctional potentials.

Developmental changes in regulation of embryonic chick heart gap junctions.

Embryonic chick myocyte pairs were isolated from ventricular tissue of 4-day, 14-day, and 18-day heart for the purpose of examining the relationship between macroscopic junctional conductance and transjunctional voltage during cardiac development. The double whole-cell patch-clamp technique was employed to directly measure junctional conductance over a transjunctional voltage range of +/- 100 mV. At all ages, the instantaneous junctional current (or conductance = current/voltage) varied linearly with respect to transjunctional voltage. This initial response was followed by a time- and voltage-dependent decline in junctional current to new steady-state values. For every experiment, the steady-state junctional conductance was normalized to the instantaneous value obtained at each potential and the data was pooled according to developmental age. The mean steady-state junctional conductance-voltage relationship for each age group was fit using a two-state Boltzmann distribution described previously for other voltage-dependent gap junctions. From this model, it was revealed that half-inactivation voltage for the transjunctional voltage-sensitive conductance shifted towards larger potentials by 10 mV, the equivalent gating charge increased by approximately 1 electron, and the minimal voltage-insensitive conductance exactly doubled (increased from 18 to 36%) between 4 and 18 days of development. Decay time constants were similar at all ages examined as rate increased with increasing transjunctional potential. This data provides the first direct experimental evidence for developmental changes in the regulation of intercellular communication within a given tissue. This information is consistent with the hypothesis that developmental expression of multiple gap junction proteins (connexins) may confer different regulatory mechanisms on intercellular communication pathways within a given cell or tissue.

Electrotonic interactions between aggregates of chick embryo cardiac pacemaker cells.

Synchronization of spontaneously active heart cell aggregates occurs shortly after they are brought into contact. The synchronous rate is determined by pacemaker phase resetting and passive subthreshold electrotonic interactions. To further study the effects of passive electrical interactions, we have used 150-microns diameter aggregates prepared from cells of 4d (4-day ventricle + 1 day in vitro), 7d, and 14d embryonic chick ventricle as models of primary, latent, and nonpacemaker tissues, respectively. Coupling of 4d and 7d aggregates (4d/7d pairs) leads to intermediate synchronous rates. We show here that elevating external K+ from 1.3 to 2.8 mM, which has no effect on 4d/4d pairs but selectively reduces the beat rate of 7d/7d pairs by 42%, slows the synchronous beat rate of 4d/7d pairs by 23%. Increases in electrical coupling in newly joined 4d/14d pairs cause the 4d rate to slow to a minimum value (16 +/- 13 beats/min, n = 16) just prior to the onset of synchronous activity. The rate slowly recovers to a final value of 40 +/- 12 beat/min. We conclude that the spontaneous beat rate of a primary pacemaker is modulated by both active and passive interactions with latent or nonpacemaker tissues.

Cardiac gap junction channel activity in embryonic chick ventricle cells.

We have recorded single-gap junction-channel currents from pairs of 7-day chick embryo ventricle cells, using the double whole cell patch-clamp technique. Junctional conductance (Gj) was variable from one preparation to the next, ranging from 0.15 to 35.0 nS. Single-channel conductance (gamma j) of the main junctional channel was 166 +/- 51 pS and was independent of Gj; a second conductance level of 60-80 pS was also seen in favorable records. The transition time from the closed to the open state was 285 +/- 153 microseconds, with some slow transitions lasting 1-5 ms. Channels opened and closed stochastically; Gj could be defined by the product of the number of active channels in the junction (N), the mean open-state probability (Po) of the channels, and gamma j. Channel activity was unaffected by cell membrane potential or by transjunctional potential. Po and Gj were reversibly reduced to low levels by 1-octanol or by elevated [Cai], whereas gamma j was unchanged by these agents. The 60-80 pS conductance mechanism was octanol- and Ca-resistant, but it is not clear whether this represents a subconductance level of the main channel or a separate class of smaller channels.

Purkinje and ventricular activation sequences of canine papillary muscle. Effects of quinidine and calcium on the Purkinje-ventricular conduction delay.

We have studied in vitro preparations of canine right and left papillary muscles to determine the excitation sequences of the Purkinje and ventricular cells, using both monopolar surface electrodes and intracellular microelectrodes. Our results show, for right papillary muscles, that the Purkinje layer covers the basal part of the muscle, and that activation from the right bundle branch propagates over all of the Purkinje layer, but directly activates the underlying ventricular layer only at specific junctional sites. Left papillary muscles have attachments to both apical and basal Purkinje strands and the Purkinje layer covers the entire muscle, but, as for right papillary muscles, activation from the Purkinje layer to the ventricular layer occurs only at basal junctional sites. Antidromic conduction in papillary muscles (propagation from the ventricular layer to the Purkinje layer) can occur at regions other than the specific sites through which the Purkinje layer activates the ventricular layer. At the identified junctional sites, the Purkinje cell action potential duration is significantly shorter than in the free-running strand, but it remains longer than that of the ventricular cells. The time delay at the junctional sites is increased by quinidine, increased calcium concentration, and increased pacing frequency.

Selective dye and ionic permeability of gap junction channels formed by connexin45.

Gap junctions are thought to mediate the direct intercellular coupling of adjacent cells by the gating of an aqueous pore permeable to ions and molecules of up to 1 kD or 8 to 14 A in diameter. We performed ion-substitution and dye-transfer experiments to determine the relative Cl-/K+ conductance and dye permeability of anionic fluorescein derivatives in chick connexin45 (Cx45) channels. We demonstrate that Cx45 forms a 26 +/- 6-picosiemen (pS) channel with a maximum detectable Cl- permeability of 0.2 relative to K+ or Cs+. Although homogeneous channel conductances were observed in multichannel recordings, the open probability estimates were indicative of nonhomogeneous gating behavior and occasional cooperativity. A second conductance state of 19 +/- 4 pS begins to predominate at higher voltages. Cx45 gap junctions are permeable to 2',7'-dichlorofluorescein but are not permeable to the more polar 6-carboxyfluorescein dye. These observations suggest that the Cx45 pore diameter is approximately 10 A and is associated with a fixed negative charge within the junctional channel.

Gating of mammalian cardiac gap junction channels by transjunctional voltage.

Numerous two-cell voltage-clamp studies have concluded that the electrical conductance of mammalian cardiac gap junctions is not modulated by the transjunctional voltage (Vj) profile, although gap junction channels between low conductance pairs of neonatal rat ventricular myocytes are reported to exhibit Vj-dependent behavior. In this study, the dependence of macroscopic gap junctional conductance (gj) on transjunctional voltage was quantitatively examined in paired 3-d neonatal hamster ventricular myocytes using the double whole-cell patch-clamp technique. Immunolocalization with a site-specific antiserum directed against amino acids 252-271 of rat connexin43, a 43-kD gap junction protein as predicted from its cDNA sequence, specifically stained zones of contact between cultured myocytes. Instantaneous current-voltage (Ij-Vj) relationships of neonatal hamster myocyte pairs were linear over the entire voltage range examined (0 less than or equal to Vj less than or equal to +/- 100 mV). However, the steady-state Ij-Vj relationship was nonlinear for Vj greater than +/- 50 mV. Both inactivation and recovery processes followed single exponential time courses (tau inactivation = 100-1,000 ms, tau recovery approximately equal to 300 ms). However, Ij recovered rapidly upon polarity reversal. The normalized steady-state junctional conductance-voltage relationship (Gss-Vj) was a bell-shaped curve that could be adequately described by a two-state Boltzmann equation with a minimum Gj of 0.32-0.34, a half-inactivation voltage of -69 and +61 mV and an effective valence of 2.4-2.8. Recordings of gap junction channel currents (ij) yielded linear ij-Vj relationships with slope conductances of approximately 20-30 and 45-50 pS. A kinetic model, based on the Boltzmann relationship and the polarity reversal data, suggests that the opening (alpha) and closing (beta) rate constants have nearly identical voltage sensitivities with a Vo of +/- 62 mV. The data presented in this study are not consistent with the contingent gating scheme (for two identical gates in series) proposed for other more Vj-dependent gap junctions and alternatively suggest that each gate responds to the applied Vj independently of the state (open or closed) of the other gate.

Ionic blockade of the rat connexin40 gap junction channel by large tetraalkylammonium ions.

The rat connexin40 gap junction channel is permeable to monovalent cations including tetramethylammonium and tetraethylammonium ions. Larger tetraalkyammonium (TAA(+)) ions beginning with tetrabutylammonium (TBA(+)) reduced KCl junctional currents disproportionately. Ionic blockade by tetrapentylammonium (TPeA(+)) and tetrahexylammonium (THxA(+)) ions were concentration- and voltage-dependent and occurred only when TAA(+) ions were on the same side as net K(+) efflux across the junction, indicative of block of the ionic permeation pathway. The voltage-dependent dissociation constants (K(m)(V(j))) were lower for THxA(+) than TPeA(+), consistent with steric effects within the pore. The K(m)-V(j) relationships for TPeA(+) and THxA(+) were fit with different reaction rate models for a symmetrical (homotypic) connexin gap junction channel and were described by either a one- or two-site model that assumed each ion traversed the entire V(j) field. Bilateral addition of TPeA(+) ions confirmed a common site of interaction within the pore that possessed identical K(m)(V(j)) values for cis-trans concentrations of TPeA(+) ions as indicated by the modeled I-V relations and rapid channel block that precluded unitary current measurements. The TAA(+) block of K(+) currents and bilateral TPeA(+) interactions did not alter V(j)-gating of Cx40 gap junctions. N-octyl-tributylammonium and -triethylammonium also blocked rCx40 channels with higher affinity and faster kinetics than TBA(+) or TPeA(+), indicative of a hydrophobic site within the pore near the site of block.

Multiple connexins confer distinct regulatory and conductance properties of gap junctions in developing heart.

Multiple gap junction proteins (connexins) and channels have been identified in developing and adult heart. Functional expression of the three connexins found in chick heart (connexin42, connexin43, and connexin45) by stable transfection of communication-deficient neuro2A (N2A) cells revealed that all three connexin cDNAs are capable of forming physiologically distinct gap junctions that differ in their transjunctional voltage dependence and unitary channel conductances. The transjunctional voltage dependences of connexin45 and connexin42 closely resembled those of 4-day and 18-day embryonic chick heart gap junctions, respectively. The multiple channel conductances between 80 and 240 pS, including the predominant 160 pS channel, observed in embryonic chick heart were also common to connexin42. The expression of multiple gap junction channels with distinct conductance and regulatory properties within a given tissue may account for developmental changes in intercellular communication.