Connexin43 and Connexin45 Form Heteromeric Gap Junction Channels in Which Individual Components Determine Permeability and Regulation
Abstract— Two gap junction proteins, connexin43 (Cx43) and connexin45 (Cx45), are coexpressed in many cardiac and other cells. Homomeric channels formed by these proteins differ in unitary conductance, permeability, and regulation. We sought to determine the ability of Cx43 and Cx45 to oligomerize with each other to form heteromeric gap junction channels and to determine the functional and regulatory properties of these heteromeric channels. HeLa cells were transfected with Cx45 or (His)6-tagged Cx43 or sequentially transfected with both connexins. Immunoblots verified production of the transfected connexins, and immunofluorescence demonstrated that they were colocalized in the HeLa-Cx43(His)6/Cx45 cells. Connexons were solubilized from HeLa-Cx43(His)6/Cx45 cells by using Triton X-100 and were applied to a Ni2+-NTA column, which binds the His6 sequence. Cx45 was coeluted from the column with Cx43(His)6, demonstrating that some hemichannels contain both connexins. Single-channel recordings showed that the HeLa-Cx43(His)6/Cx45 cells exhibited single-channel conductances that were not observed in cells expressing either connexin alone. Dye-coupling experiments showed that HeLa-Cx43(His)6 cells readily passed Lucifer yellow and N-(2-aminoethyl)biotinamide hydrochloride (neurobiotin); in contrast, HeLa-Cx45 and HeLa-Cx43(His)6/Cx45 cells showed extensive intercellular passage of neurobiotin but little coupling with Lucifer yellow. Treatment with the protein kinase C activator 12-O-tetradecanoylphorbol 13-acetate reduced junctional conductance in cells expressing Cx43, Cx45, or both connexins, but it reduced the extent of neurobiotin transfer only in HeLa-Cx43(His)6 and HeLa-Cx43(His)6/Cx45 cells but not in the HeLa-Cx45 cells. Thus, biochemical and electrophysiological evidence suggests that Cx43 and Cx45 extensively mix to form heteromeric channels; however, individual connexin components dominate aspects of the physiological behavior of these channels.
Gap junctions contain intercellular channels that connect the cytoplasm of adjacent cells. They permit the passage of ions and small molecules between cells, providing a direct coordination of electrical and metabolic functions. In electrically excitable cells, such as cardiac myocytes, intercellular communication is critical for normal current flow. Derangements of cardiac gap junctions are associated with abnormal conduction and developmental defects. Gap junction channels are constructed of two hemichannels (connexons), each contributed by one of the opposing cells. These hemichannels are hexamers of protein subunits called connexins. Several connexins, including connexin37, connexin40 (Cx40), connexin43 (Cx43), and connexin45 (Cx45) form gap junctions between cells of the cardiovascular system.1
Expression of a single connexin by a cell is sufficient to establish gap junctional intercellular communication. The physiological properties of homomeric-homotypic channels, in which all 12 subunits are the same connexin, have been extensively characterized by use of various expression systems, and they differ for different connexins. For example, the unitary conductance of homomeric-homotypic Cx43 channels (100 to 120 pS) is greater than that of homomeric-homotypic Cx45 channels (30 to 40 pS).2 Cx43 channels are permeable to many different dye tracers, including Lucifer yellow, carboxyfluorescein, dichlorofluorescein, and N-(2-aminoethyl)biotinamide hydrochloride (neurobiotin), whereas Cx45 channels show less passage of Lucifer yellow and the fluorescein derivatives.3–5⇓⇓ Both Cx43 and Cx45 are phosphoproteins, and activation of kinases, such as protein kinase C (PKC), may have differential effects on junctional conductance (gj) in cells expressing these connexins.6,7⇓
Many cells contain >1 connexin. The properties of intercellular communication between such cells should be determined by the properties of channels containing either connexin alone and channels containing a mixture of connexins. If two similarly coexpressed connexins are capable of freely mixing, channels containing two connexins in the same hemichannel (heteromeric channels) will be far more abundant than those formed of homomeric hemichannels (ie, those containing only one connexin).8 Therefore, to understand cardiac cellular coupling, it is important to understand the potential ability of two coexpressed cardiac connexins to form heteromeric channels and the properties of such channels (if they do exist).
Biochemical evidence has established the existence of heteromeric hemichannels in extracts of adult rodent or avian tissues,9–11⇓⇓ of A7r5 aortic smooth muscle cells,12,13⇓ and of baculovirus-infected or plasmid-transfected cells.8,14,15⇓⇓ Double whole-cell patch-clamp studies of connexin-transfected cells have shown that in some cases, the heteromeric channels have properties that differ from those of channels formed of either connexin alone,8,12⇓ whereas in other cases, the properties of heteromeric channels do not differ from those predicted from a combination of homomeric and heterotypic channels.15
In the present study, we focused on the possible heteromeric interactions of Cx43 and Cx45. Both Cx43 and Cx45 are found together in a number of cells of the cardiovascular system (in atrial and ventricular myocytes, the distal conducting system, and some vascular smooth muscle cells).1 Cx43 and Cx45 are also found together in astrocytes, myometrial smooth muscle cells, and osteoblasts. When coexpressed, Cx43 and Cx45 colocalize to identical distributions and gap junction structures.16 Moreover, osteoblastic cell lines expressing different amounts of Cx43 and Cx45 show differences in permeability properties.3 Therefore, we hypothesized that Cx43 and Cx45 might make heteromeric gap junction channels and that the formation of such mixed channels might have significant ramifications for understanding cardiac intercellular communication.
Materials and Methods
Cell Culture and Transfection
We have previously described the production of HeLa cells transfected with rat Cx43 containing a (His)6 epitope appended to its carboxyl terminus [HeLa-Cx43(His)6]15 or transfected with chicken Cx45 (HeLa-Cx45).2 (Previously characterized physiological properties appear identical for chicken and mammalian Cx45.2,17⇓) Clones expressing both connexins [HeLa-Cx45/Cx43(His)6 cells] were obtained by sequential transfection of HeLa-Cx45 cells with the pcDNA3.1-Cx43(His)6 construct and selection in hygromycin (1 mg/mL). Data are presented for a single clone that exhibited abundant expression of both Cx43(His)6 and Cx45, but consistent physiological data were obtained in more limited studies of other clones (containing differing ratios).
Immunohistochemistry and Immunoblots
A rabbit polyclonal antibody to Cx43 has been characterized previously.18 Mouse monoclonal antibodies to Cx43, Cx45, and the His tag were obtained from Chemicon.
Single- and double-label immunofluorescence microscopy was performed essentially as described previously3 with the use of cells grown on glass coverslips, fixed in methanol/acetone, and permeabilized with 1% Triton X-100. Cy3-conjugated goat anti-mouse IgG and Cy2-conjugated goat anti-rabbit antibodies were obtained from Jackson ImmunoResearch. Cells were examined by using a Zeiss Axiophot II microscope equipped for epifluorescence and digital microscopy.
For immunoblots, cell cultures were rinsed with PBS, scraped into cold PBS containing 2 mmol/L phenylmethylsulfonyl fluoride, and centrifuged. After the supernatant was discarded, cell pellets were frozen in liquid nitrogen and stored at −80°C. Cell pellets were lysed by sonicating in water containing protease and phosphatase inhibitors.19
Immunoblot analyses were performed as described previously19 by using 100 μg protein or fractions eluted from a Ni2+-NTA column (Novagen).
Connexon Solubilization and Affinity Purification With Use of Ni2+-NTA Column
Connexons were solubilized by incubation of a cell lysate with 1% Triton X-100, followed by centrifugation at 100 000g.11,15,20,21⇓⇓⇓ The Triton X-100–soluble supernatant contains immunoreactive material of sizes consistent with connexin monomers and hexamers (connexons).11,20,21⇓⇓ Supernatants were affinity-purified by binding to the Ni2+-NTA column, extensive washing, and eluting of specifically bound material with 1 mol/L imidazole.15 Each fraction contained 500 μL. Seventy to 100 μL was loaded in each lane, electrophoresed through 8% polyacrylamide gels containing SDS, and immunoblotted.
Microinjection of Gap Junction Tracers
Cells cultured on coverslips (80% to 100% confluent cultures) were impaled with a micropipette filled with 150 mmol/L LiCl, 4% Lucifer yellow (charge=−2, molecular weight 457; Sigma Chemical Co), and 4% neurobiotin (charge=+1, molecular weight 322.8; Vector Laboratories). Solutions were microinjected with a picospritzer (model PLI-188, Nikon Inc) by using 0.2- to 0.3-second pulses of 1 to 2 psi; cells were impaled for 0.5 to 1 minute. After the microinjection, the cells were fixed with paraformaldehyde (4% in PBS) for 30 minutes and then permeabilized with methanol/acetone (1:1) for 2 minutes at room temperature. The neurobiotin tracer was detected after staining the cells with streptavidin-Cy3 conjugate (Sigma). The extent of intercellular transfer of both tracers was determined by recording the number of adjacent cells containing the tracer after visualization by epifluorescence and digital microscopy. In some experiments, the cells were treated with 200 nmol/L 12-O-tetradecanoylphorbol 13-acetate (TPA) for 30 minutes before the microinjections.
A dual whole-cell voltage-clamp technique was applied to measure the gj between cells with the use of a CsCl pipette solution.2 In all experiments, the resistance of the electrodes had a maximum value of 5 MΩ, and extra suction was applied to reduce the series resistance. The initial current was measured at the beginning of the voltage pulses with the use of a low-pass filter of 5 kHz. Total junctional current measurements before and after TPA were performed in cell pairs from sister cultures plated 18 to 24 hours before study, which were well attached to the recording chamber and contained at least 5 μm of opposing membranes.
Single-channel currents were measured by using freshly split cells, where gj was low. Unitary junctional currents were recorded during long voltage steps of 10 seconds applied to one of the cells. Amplitudes of unitary opening or closing current events were measured by using a digitizing board (Summagraphics with Sigma- Scan software, Jandel) from the chart recorder paper (Gould Windograph), where current traces were filtered at 100 to 500 Hz. Frequency distribution histograms of the events and gaussian distribution best fits were calculated for each experiment (Origin, Microcal). Each event was defined as the current transition between channel states, where the residence time in each state was >20 ms. All-points histograms were generated by using pClamp protocols from traces filtered at 200 to 500 Hz and digitized at 1 kHz. Digitized points were grouped into 128 points per bin. Multiple gaussian functions were also obtained by following the Levenberg-Marquardt algorithm to determine the best bell-shaped curves for gaussian probability distribution functions.
Immunochemical Characterization of Transfected HeLa Cells
To study the possible hetero-oligomerization of Cx43 and Cx45, we introduced Cx45 or Cx43 containing the (His)6 sequence appended to its carboxy terminus [Cx43(His)6] or both connexins into communication-deficient HeLa cells by stable transfection. The (His)6 tag was added to facilitate the isolation of connexons. Immunoblotting and immunofluorescence confirmed the production of immunoreactive Cx45 and Cx43(His)6 in the transfected HeLa cell clones (Figures 1 and 2⇓). In immunoblots, Cx45 was detected as a single band of 48 kDa, and Cx43(His)6 was detected as multiple bands of 42 to 46 kDa, which likely correspond to the nonphosphorylated and phosphorylated forms described for Cx43 in many systems (Figure 1). The immunoblots confirmed that the HeLa-Cx43(His)6/Cx45 cells were producing substantial amounts of both connexins.
Immunofluorescence analysis showed that immunoreactive connexin proteins were produced and located in a distribution expected for gap junction proteins. Single-label immunofluorescent staining of HeLa-Cx43(His)6 (Figure 2A) and of HeLa-Cx45 (Figure 2B) cells showed intense labeling for the appropriate connexin at cellular interfaces (likely representing gap junction structures) and in a perinuclear distribution within the cytoplasm (likely representing connexin within the biosynthetic pathway). Single-label immunofluorescent staining of HeLa-Cx43(His)6/Cx45 cells showed a very similar pattern of staining (not shown). Moreover, double-label immunofluorescence staining of HeLa-Cx43(His)6/Cx45 cells showed virtually identical and superimposable distributions (colocalization) of Cx43(His)6 and Cx45 at plasma membrane appositions and in intracellular compartments (Figures 2C through 2G). Reaction of anti-Cx43 or anti-Cx45 antibodies with untransfected HeLa cells or HeLa cells transfected only with the other connexin gave no detectable staining (not shown).
Coelution of Cx45 With Cx43(His)6 From Ni2+-NTA Affinity Column
To examine the possible hetero-oligomeric association of Cx45 with Cx43 in connexons, we used an affinity purification strategy similar to that used to show heteromeric association of Cx40 and Cx43.15 Triton X-100–soluble extracts containing connexons from transfected HeLa cells were applied to a Ni2+-NTA column, which has a high affinity for (His)6 and, therefore, should bind connexins containing this tag (which may bring along associated proteins). Starting material and column fractions were analyzed by immunoblotting (Figure 3). We observed that Cx45 eluted with the Cx43(His)6 when connexons from HeLa-Cx43(His)6/Cx45 cells were affinity-purified (Figure 3, top and middle panels).
In contrast, when identical procedures were applied to Triton X-100–soluble material from HeLa-Cx45 cells, all Cx45 immunoreactivity was detected in the flow-through and initial washes, suggesting that Cx45 did not bind to the Ni2+-NTA column (Figure 3, bottom panels). Similarly, we found no evidence of Cx45 binding to the column when Triton extracts were prepared from HeLa-Cx43(His)6 and HeLa-Cx45 cells that were cultured separately and the extracts were mixed before application to the column or when extracts were prepared from cocultured HeLa-Cx43(His)6 and HeLa-Cx45 cells (not shown). These data demonstrate that Cx45 association with Cx43 required coexpression in the same cell and did not occur after detergent solubilization or heterotypic interactions. The specific binding of Cx45 to Ni2+-NTA when connexons were solubilized from HeLa-Cx43(His)6/Cx45 cells suggested oligomeric interactions to form heteromeric connexons.
Unitary Conductances of Gap Junction Channels in Transfected HeLa Cells
The double whole-cell patch-clamp approach was used to obtain single-channel recordings from low-conductance pairs of HeLa cells transfected with Cx43(His)6 or Cx45 or sequentially transfected with Cx43(His)6 and Cx45. The maximal single-channel conductance was determined by using low transjunctional voltage protocols (<60 mV) in cells that had been plated for 1 to 4 hours. Examples of representative recordings from HeLa-Cx43(His)6 and HeLa Cx43(His)6/Cx45 cells are shown in Figures 4 and 5⇓, and event histograms summarizing the channel data obtained from all transfectants are presented in Figure 6.
The behavior of Cx43(His)6 (Figure 4) was rather similar to that of wild-type Cx43. Single-channel recordings showed the presence of a main state of ≈120 pS and a residual state of 30 pS (Figures 4A⇑ and Figure 6, top left panel). This value did not differ significantly from previous measurements of the unitary conductance of wild-type Cx43 as expressed in similar culture systems.2,22,23⇓⇓ Even when analyzed in multichannel recordings, small events (50 to 100 pS) were rare (<5% of events). The Cx43(His)6 channels did exhibit less voltage dependence than did the wild-type Cx43. They had remarkably high activity despite voltage pulses as great as 100 mV (Figure 4A). The voltage dependence was examined at various transjunctional voltages, and the corresponding series of inactivating currents is shown in Figure 4B. At 60 mV, the current did not inactivate at all. A comparison of the plots of normalized steady-state conductance versus voltage (Figure 4C) also suggests a somewhat lower voltage sensitivity for Cx43(His)6 than for wild-type Cx43.
The unitary conductance of the gap junction channels (γj) in HeLa-Cx45 cells was also obtained; it was best fit by a single gaussian function at γj=38±14 (5 experiments, 1045 events) (Figure 6, top middle panel). This value was indistinguishable from that obtained in previous studies of Cx45 in similar expression systems.2,24⇓
Single-channel events of many different sizes were observed in pairs of cells coexpressing Cx43(His)6 and Cx45 (Figure 5). Many of these channel events did not correspond to channel sizes detected in the HeLa-Cx43(His)6 or HeLa-Cx45 cell pairs. These data also differed dramatically from our previous analyses of heterotypic Cx43-Cx45 channels, which had a single unitary conductance of 60 pS.2 Figure 5 contains a representative trace for which multiple different unitary current transitions were recorded from a pair of HeLa-Cx43(His)6/Cx45 cells during a 60-mV voltage pulse. Multiple transitions were evident, and detailed analysis of this recording demonstrates the multiple different unitary conductances (Figures 5a through 5g). An event histogram of the channel data from HeLa-Cx43(His)6/Cx45 cell pairs (Figure 6, top right panel) required ≥8 gaussian functions for best fitting.
Our data from the HeLa-Cx43(His)6/Cx45 cell pairs are consistent with the expectation that the formation of heteromeric channels leads to a large variety of unitary channel events.8
Intercellular Passage of Tracer Dyes in Transfected HeLa Cells
To examine the possible effects of heteromeric channel formation on gap junction permeability, we examined the intercellular passage of Lucifer yellow and neurobiotin in the HeLa transfectants. Individual cells within a monolayer were microinjected with a mixture of both molecules, and the extent of gap junctional coupling was assessed by counting dye-filled neighbors. Representative examples of such experiments are shown in Figure 7 (left panels). Extensive intercellular transfer of neurobiotin (including up to second- and third-order neighbors) was observed in all of the HeLa transfectants (Figures 7B, 7D, and 7F). In contrast, extensive intercellular transfer of Lucifer yellow (primarily to first-order neighbors) was observed only for the Cx43(His)6 cells (Figure 7A). In the HeLa-Cx45 cells and HeLa-Cx43(His)6/Cx45 cells, intense Lucifer yellow fluorescence was observed in the injected cell, but there was only limited transfer to directly adjacent cells (Figures 7C and 7E).
The results of many such dye injection experiments are summarized in the bar graph in Figure 7 (right side). Under control conditions, the extent of neurobiotin transfer was similar among all of the HeLa transfectants (Figure 7, hatched bars); the number of dye-filled cells was 21.5±3.2 for HeLa-Cx43(His)6 cells, 15±2.5 for HeLa-Cx45 cells, and 22.9±3.4 for HeLa-Cx43(His)6/Cx45 cells. In contrast, extensive transfer of Lucifer yellow (Figure 7, solid bars) was observed only for HeLa-Cx43(His)6 cells (11.2±2.9 dye-filled cells), whereas the HeLa-Cx45 and HeLa-Cx43(His)6/Cx45 cells showed only limited transfer of this dye (0.8±0.2 and 1.5±0.3 dye-filled cells, respectively). Thus, the permeability properties (for these two dyes) of the HeLa-Cx43(His)6/Cx45 cells were similar to those of cells expressing only Cx45.
Effects of PKC Activation on Intercellular Communication Between Transfected HeLa Cells
In many cell types, gap junctional communication is regulated by protein kinases. One of the most studied effects is the alteration of cellular coupling after activation of PKC by treatment of cells with the phorbol ester TPA. We hypothesized that TPA treatment might have different effects on cells expressing different connexins or making heteromeric channels.
We incubated monolayers of transfected HeLa cells with 200 nmol/L TPA for 30 minutes before determining the extent of transfer of microinjected neurobiotin. In the HeLa-Cx43(His)6 cells, TPA treatment led to a dramatic reduction of dye coupling (Figure 7; compare open and hatched bars). Many investigators have previously observed reductions in dye coupling after TPA treatment of cells expressing wild-type Cx43.25–27⇓⇓ Neurobiotin dye coupling was also dramatically reduced in HeLa-Cx43(His)6/Cx45 cells after TPA treatment. However, in the HeLa-Cx45 cells, TPA treatment had no detectable effect on intercellular transfer of neurobiotin (Figure 7). Thus, the response of the coexpressing cells to PKC activation (reduced dye transfer) was similar to that of cells expressing Cx43 alone. Immunoblot analysis of HeLa-Cx43(His)6 and HeLa-Cx43(His)6/Cx45 showed an increase in the abundance of the slower mobility Cx43 forms (phospho forms) after TPA treatment but showed no effects on Cx45 (data not shown).
To determine the effect of TPA on total junctional communication, we calculated the gj of multiple cell pairs expressing only Cx43(His)6 or Cx45 or coexpressing both connexins. The total gj decreased in all cell types analyzed after the application of TPA (Figure 8). For pairs of HeLa-Cx43(His)6 cells, the total conductance was only 15% of control values after application of either concentration of TPA, whereas in HeLa-Cx45 or HeLa-Cx43(His)6/Cx45 cell pairs, TPA treatment reduced conductance to ≈30% of control values. To determine the influence of the series resistance on our measurements, we plotted all gj data against the sum of the resistance of the recording electrodes (data not shown). Extrapolation to zero resistance indicated that even in the worst scenario, we have underestimated the TPA-induced reduction of gj by <20%.
We also studied the effects of TPA treatment on the distribution of single gap junction channel conductance sizes (Figure 6, bottom panels). The event histograms appeared identical for HeLa-Cx45 cell pairs whether they were determined under control conditions or after TPA treatment. However, TPA treatment was accompanied by a reduction in the frequency of larger sized events (≥90 pS) for both HeLa-Cx43(His)6 and HeLa-Cx43(His)6/Cx45 cells (Figure 6).
The data in the present study provide strong biochemical and electrophysiological evidence indicating that coexpressed Cx43 and Cx45 form heteromeric channels. When detergent extracts containing solubilized connexons from HeLa-Cx43(His)6/Cx45 cells were subjected to affinity purification, Cx45 was copurified with Cx43(His)6, even though only the Cx43 contained the tag that would bind to the column; copurification required association of Cx45 with Cx43 (in heteromers). Control experiments excluded artifactual association. Moreover, analysis of single gap junction channel events in pairs of HeLa-Cx43(His)6/Cx45 cells yielded a large variety of events of sizes that were not seen in cells containing only homomeric connexons. In this respect, our data differ from those of Valiunas et al,15 who concluded that all of the voltage gating and most (if not all) of the single-channel events in cells coexpressing Cx43 and Cx40 could be explained by the properties of homomeric/homotypic and homomeric/heterotypic channels.
Our biochemical studies were facilitated by the incorporation of the (His)6 tag on the carboxyl terminus of Cx43. This strategy was accompanied by a theoretical risk of altering the behavior of Cx43. However, many of the functional properties of Cx43 were not altered by this modification. The unitary channel conductances observed between pairs of HeLa-Cx43(His)6 cells were indistinguishable from those previously observed in cells expressing wild-type Cx43.2,22,23⇓⇓ HeLa-Cx43(His)6 cells showed extensive intercellular passage of both Lucifer yellow and neurobiotin, like cells expressing the wild-type protein. HeLa-Cx43(His)6 cells showed an uncoupling after TPA treatment similar to that expected for HeLa cells expressing wild-type Cx43. A decrease in the voltage-dependent gating of Cx43(His)6 was observed (Figure 4C). Because this parameter is strongly influenced by the access resistance when cells express connexins abundantly, this property was not further analyzed in the coexpressing cells.
Previous studies have suggested that gap junction channels formed of different connexins differ in their permeability and selectivity. Cx43 channels have been shown to allow passage of many different tracer dyes, including Lucifer yellow, carboxyfluorescein, and neurobiotin; Cx45 channels are more selective, showing extensive coupling with neurobiotin but only limited coupling with Lucifer yellow or carboxyfluorescein.3–5⇓⇓ Our data from HeLa-Cx43(His)6 cells and HeLa-Cx45 cells are consistent with those observations. We further found that whereas neurobiotin readily passed between HeLa-Cx43(His)6/Cx45 cells, these cells showed little intercellular transfer of Lucifer yellow. These results suggest that heteromeric channels formed by Cx43/Cx45 have permeability characteristics similar to those of homomeric Cx45 channels. Previous studies have suggested that homomeric Cx45 channels are relatively selective for cations compared with anions, whereas homomeric Cx43 channels show little preference.5,28⇓ Because the two tracers that we used (Lucifer yellow and neurobiotin) differ both in molecular weight and in charge, we cannot say whether the observed difference in transfer in HeLa-Cx43(His)6/Cx45 cells is due to charge or size selectivity. Our present data may help to explain the observations of Koval et al,29 who noted that introduction of Cx45 into ROS cells, which endogenously express Cx43, led to decreased dye coupling when assayed with Lucifer yellow.
Cx43 and Cx45 are phosphoproteins, and their phosphorylation is increased after PKC activation.18,25,30,31⇓⇓⇓ In many (but not all) Cx43-expressing cells, activation of PKC by phorbol esters is accompanied by a reduction in gap junctional communication.25–27,32⇓⇓⇓ TPA-induced effects on communication via Cx45 channels have been less extensively studied; TPA treatment has been reported to lead to an acute (<8-minute) increase in total gj,7 whereas longer treatment (10 to 45 minutes) was associated with the appearance of smaller (≈16-pS) channel events.
We found that TPA treatment led to decreased dye coupling (neurobiotin) between HeLa-Cx43(His)6 cells. Similar observations have been made in a number of cell systems expressing Cx43. In contrast, TPA treatment had no effect on dye transfer between HeLaCx45 cells. Neurobiotin coupling was also drastically reduced in HeLa-Cx43(His)6/Cx45 cells after TPA treatment, suggesting that the heteromeric channels are affected by PKC-activated pathways. These results suggest that heteromeric channels formed by Cx43/Cx45 may be regulated by PKC activation in a manner similar to that in homomeric Cx43 channels.
Total gj was reduced after TPA treatment in all of the transfected cells studied. This decrease was accompanied by a shift in unitary conductance (loss of the largest channels) for the HeLa-Cx43(His)6 and HeLa-Cx43(His)6/Cx45 cells. These observations are consistent with recent studies27 suggesting that phosphorylation of Ser368 in Cx43 is associated with decreased unitary channel conductance. The mechanism of decreased conductance in the HeLa-Cx45 cells may not be the same, because unitary conductance was not changed in these cells.
In the HeLa-Cx43(His)6 and HeLa-Cx43(His)6/Cx45 cells, gj and neurobiotin transfer decreased in parallel with TPA treatment. The shift to lower single-channel conductances suggests that the reduced channel size was concomitant with a reduction in the permeability to neurobiotin. Thus, the changes in permeability in cells coexpressing Cx45 and Cx43(His)6 may be guided by the presence of Cx43.
In the HeLa-Cx45 cells, TPA treatment led to a substantial reduction in gj, without a change in neurobiotin permeability. This suggests either that the diffusion of neurobiotin was not significantly affected by the reduction in gj or that the permeability to neurobiotin of Cx45 channels increased during a simultaneous reduction of the open probability of the channels. Regardless, as previously reported by Veenstra et al,5 the permeability and conductance properties of gap junction channels are not always directly correlated.
In summary, we have shown that coexpressed Cx43 and Cx45 can form heteromeric channels. Some characteristics of the heteromeric channels (Lucifer yellow permeability and TPA-induced reduction of neurobiotin transfer) are dominated by one of the connexin components. These properties appear to follow those of the most restrictive component.
These findings may have significant implications for the regulation of intercellular communication in the heart. Recent studies of freshly isolated myocardial and vascular cells have shown the presence of multiple single-channel conductances in their gap junctions consistent with the formation of heteromeric channels.33,34⇓ Differences in the relative expression of Cx43 and Cx45 might lead to changes in the abundances of heteromeric Cx43/Cx45 gap junction channels. Such changes might be accompanied by alterations in conduction and in the intercellular permeability/flux of signaling molecules.
This study was supported by National Institutes of Health grants HL-59199 and HD-09402 (to Dr Beyer) and HL-63969 (to Dr Moreno). Dr Martinez was supported by a postdoctoral fellowship from the American Heart Association.
Original received November 26, 2001; revision received April 12, 2002; accepted April 12, 2002.
- ↵Elfgang C, Eckert R, Lichtenberg-Frate H, Butterweck A, Traub O, Klein RA, Hulser DF, Willecke K. Specific permeability and selective formation of gap junction channels in connexin-transfected HeLa cells. J Cell Biol. 1995; 129: 805–817.
- ↵Veenstra RD, Wang HZ, Beblo DA, Chilton MG, Harris AL, Beyer EC, Brink P. Selectivity of connexin-specific gap junctions does not correlate with channel conductance. Circ Res. 1995; 77: 1156–1165.
- ↵van Veen TA, van Rijen HV, Jongsma HJ. Electrical conductance of mouse connexin45 gap junction channels is modulated by phosphorylation. Cardiovasc Res. 2000; 46: 496–510.
- ↵Jiang JX, Goodenough DA. Heteromeric connexons in lens gap junction channels. Proc Natl Acad Sci U S A. 1996; 93: 1287–1291.
- ↵Bevans CG, Kordel M, Rhee SK, Harris AL. Isoform composition of connexin channels determines selectivity among second messengers and uncharged molecules. J Biol Chem. 1998; 273: 2808–2816.
- ↵He DS, Jiang JX, Taffet SM, Burt JM. Formation of heteromeric gap junction channels by connexins 40 and 43 in vascular smooth muscle cells. Proc Natl Acad Sci U S A. 1999; 96: 6495–6500.
- ↵Cottrell GT, Burt JM. Heterotypic gap junction channel formation between heteromeric and homomeric Cx40 and Cx43 connexons. Am J Physiol. 2001; 281: C1559–C1567.
- ↵Stauffer KA. The gap junction proteins 36 1-connexin (connexin-32) and 36 2-connexin (connexin-26) can form heteromeric hemichannels. J Biol Chem. 1995; 270: 6768–6772.
- ↵Valiunas V, Gemel J, Brink PR, Beyer EC. Gap junction channels formed by coexpressed connexin40 and connexin43. Am J Physiol. 2001; 281: H1675–H1689.
- ↵Kanter HL, Laing JG, Beyer EC, Green KG, Saffitz JE. Multiple connexins colocalize in canine ventricular myocyte gap junctions. Circ Res. 1993; 73: 344–350.
- ↵Valiunas V. Biophysical properties of connexin-45 gap junction hemichannels studied in vertebrate cells. J Gen Physiol. 2002; 119: 147–164.
- ↵Koval M, Harley JE, Hick E, Steinberg TH. Connexin46 is retained as monomers in a trans-Golgi compartment of osteoblastic cells. J Cell Biol. 1997; 137: 847–857.
- ↵Bukauskas FF, Jordan K, Bukauskiene A, Bennett MV, Lampe PD, Laird DW, Verselis VK. Clustering of connexin 43-enhanced green fluorescent protein gap junction channels and functional coupling in living cells. Proc Natl Acad Sci U S A. 2000; 97: 2556–2561.
- ↵Veenstra RD, Wang HZ, Westphale EM, Beyer EC. Multiple connexins confer distinct regulatory and conductance properties of gap junctions in developing heart. Circ Res. 1992; 71: 1277–1283.
- ↵Brissette JL, Kumar NM, Gilula NB, Dotto GP. The tumor promoter 12-O-tetradecanoylphorbol-13-acetate and the ras oncogene modulate expression and phosphorylation of gap junction proteins. Mol Cell Biol. 1991; 11: 5364–5371.
- ↵Lampe PD, TenBroek EM, Burt JM, Kurata WE, Johnson RG, Lau AF. Phosphorylation of connexin43 on serine368 by protein kinase C regulates gap junctional communication. J Cell Biol. 2000; 149: 1503–1512.
- ↵Veenstra RD, Wang HZ, Beyer EC, Brink PR. Selective dye and ionic permeability of gap junction channels formed by connexin45. Circ Res. 1994; 75: 483–490.
- ↵Koval M, Geist ST, Kemendy AE, Westphale EM, Civitelli R, Beyer EC, Steinberg TH. Transfected connexin45 alters gap junction permeability in cells expressing endogenous connexin43. J Cell Biol. 1995; 130: 987–995.
- ↵Li X, Simard JM. Multiple connexins form gap junction channels in rat basilar artery smooth muscle cells. Circ Res. 1999; 84: 1277–1284.