Carcinogenesis, Vol. 22, No. 2, 221-231,
February 2001
© 2001 Oxford University Press
CANCER BIOLOGY |
Phosphorylation of connexin43 and inhibition of gap junctional communication in 12-O-tetradecanoylphorbol-13-acetate-exposed R6 fibroblasts: minor role of protein kinase CßI and µ
Department of Environmental and Occupational Cancer, Institute for Cancer Research, The Norwegian Radium Hospital, N-0310 Oslo, Norway
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
12-O-tetradecanoylphorbol-13-acetate (TPA) inhibits gap junctional communication in many cell culture systems, but TPA-induced phosphorylation of the gap junction protein connexin43 (Cx43) varies much between systems. We have here studied whether these responses and their sensitivities can be correlated with total protein kinase C (PKC) enzyme activity and if specific PKC isoenzymes are involved. Rat R6 fibroblasts transfected with the cDNA sequence encoding PKCßI (R6-PKC3) had a total PKC activity 7- to 16-fold higher than the corresponding control cells (R6-C1), depending on the selection pressure (G418 concentration). Still, R6-PKC3 cells were no more sensitive than R6-C1 cells to TPA-induced down-regulation of communication, except at the highest selection pressure (500 µg/ml G418). Thus, total PKC activity does not indicate absolute sensitivity of a cell system to TPA-induced suppression of communication, but within a certain cell system increasing PKC activity may enhance the sensitivity to TPA in this respect. The results also suggest that PKCßI is of minor importance for TPA-induced regulation of communication. Experiments with the Lilly compound 379196, a PKCß-specific inhibitor, further supported this conclusion. Except for PKCßI in R6-PKC3 cells, both cell lines contained the TPA-responsive PKC isoenzymes
, 
, 
and µ. Long-term treatment with TPA caused strong down-regulation of PKC, and , but little down-regulation of PKCµ. Concurrently, the cells became refractory to repeated exposure to TPA, indicating that PKCµ is of minor importance. Experiments with the general PKC inhibitor GF109203X and the PKC (and ß/
) inhibitor Gö6976 suggested that both classical () and novel PKCs ( and ) might be involved in TPA-induced suppression of intercellular communication, while phosphorylation of Cx43 may mainly be mediated by PKC in the present systems.
Abbreviations: aPKC, atypical PKC; cPKC, classical PKC; Cx43, connexin43; DOPPA, 12-deoxyphorbol-13-O-phenylacetate-20-acetate; GJIC, gap junctional intercellular communication; nPKC, novel PKC; PKC, protein kinase C; TPA, 12-O-tetradecanoylphorbol-13-acetate.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Gap junctional intercellular communication (GJIC) is important in regulation of cell growth, development and differentiation (1–3). GJIC takes place between adjacent cells through channels situated in the cell membrane structures named gap junctions. Each cell contributes one half-channel. The channels are assembled from a protein family called connexins. Genes encoding 15 connexins with molecular masses from 26 to 57 kDa have been cloned from rodents (2,4). Except for the smallest connexin, connexin26, all connexins contain potential phosphorylation sites, and many of them have been shown to be phosphorylated in intact cells. One of the most studied connexins, connexin43 (Cx43), is usually found in two or three major phosphorylation variants in non-stimulated cells, as detected by western blotting (5). The phosphorylation status of Cx43 is often changed in cells exposed to agents that influence GJIC, like growth factors (see for example ref. 6) and protein-tyrosine phosphatase inhibitors (7,8).
Protein kinase C (PKC) is a serine/threonine kinase family central to cellular regulation. The potent PKC activator and tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA) was earlier shown to inhibit GJIC in several cell types (9,10), often concurrently with major changes in phosphorylation status of Cx43 (11–13). Other PKC activators, like diacylglycerol, other phorbol esters and bryostatin 1, also inhibit GJIC in many cell types (14–16). Cx43 may be a direct target for PKC (17). It has therefore been assumed that TPA-induced phosphorylation of Cx43 may close (13), or at least decrease the size (18) of the channel pores. However, the Cx43 phosphorylation response to TPA can vary widely, in spite of a strongly decreased level of GJIC (11,12).
Twelve PKC isoenzymes have been identified to date. They are divided into three main subfamilies. The classical PKC (cPKC) isoenzymes , ßI, ßII and are Ca2+-dependent and activated by diacylglycerol or TPA. The Ca2+-independent novel PKC (nPKC) isoenzymes , , 
and 
are also activated by diacylglycerol or TPA. The atypical PKC (aPKC) isoenzymes 
and 
(referred to as 
in humans) are independent of Ca2+ and cannot be activated by diacylglycerol or TPA. PKCµ (referred to as PKD in mouse) and the recently cloned PKC
(19) may constitute a fourth subfamily. TPA activates PKCµ (20) and probably also PKC. The strong response of many cell types to TPA or other PKC activating agents clearly suggests that cPKCs, nPKCs and/or PKCµ/ are involved in regulation of GJIC and/or phosphorylation of Cx43. However, it is not known whether the level of PKC enzyme activity is important for the level of sensitivity of cells to TPA and/or whether specific PKC isoenzymes are responsible for these effects.
We have here approached this problem using the rat fibroblast cell lines R6-C1 and R6-PKC3. R6-PKC3 cells are transfected with a construct encoding PKCßI generated from rat brain cDNA. They are transformed as defined by increased saturation density, colony formation in soft agar and production of tumors in nude mice (21). R6-C1 cells contain the same construct but without the PKCßI cDNA and by all criteria behave as untransformed cells (21). The PKC activity in R6-PKC3 cells is significantly higher than in R6-C1 cells (21). Thus, this gave us the opportunity to investigate the importance of total PKC enzyme activity in regulation of GJIC and phosphorylation of Cx43. Furthermore, the use of R6-PKC3 and R6-C1 cells allowed us to study whether PKCßI is specifically involved in the two processes or if other isoenzymes are more critical. R6 cells contain PKC, , and (22). Overexpression of PKCßI in R6-PKC3 cells was reported to confer a higher resistance to TPA-induced down-regulation of PKC and (22). Since PKC is not activated by TPA and PKC is down-regulated by long-term treatment with TPA, this seemingly offered us the opportunity to study the potential involvement of PKC and in regulation of GJIC and phosphorylation of Cx43. Several approaches have been used in the present study. First, a direct comparison between the PKC enzyme activities and phorbol ester sensitivities in the two systems was carried out; second, pharmacological activators and inhibitors were employed to indicate the importance of different PKC isoenzymes in the two systems; third, selection pressure was used in an attempt to alter the PKC enzyme activity within the systems. Furthermore, as GJIC is believed to be involved in growth control (1,2), it was of general interest to compare GJIC in R6-PKC3 and R6-C1 cells, as the former cells have less strict growth control (21).
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Chemicals
TPA was from Sigma (St Louis, MO). 12-Deoxyphorbol-13-O-phenylacetate-20-acetate (DOPPA) (Alexis, Läufelfingen, Switzerland) is a relatively PKCß-specific activator and is approximately equipotent with TPA in activation of this isoenzyme (23,24). In intact cells 100-fold higher concentrations of DOPPA are needed to activate other PKC isoenzymes than PKCß (25). Gö6976 (Alexis) is a staurosporine derivative with much enhanced specificity for cPKCs (26). GF109203X (Alexis) is a broad specificity PKC inhibitor of the bis-indolylmaleimide family. The Lilly compound 379196 (a kind gift from Lilly Research Laboratories, Indianapolis, IN) is a bis-indolylmaleimide compound with a good specificity for PKCß. IC50 values for PKC isoenzymes are as follows (µM):
, 0.6; ßI, 0.05; ßII, 0.03; , 0.6; , 0.7; , 5; , 0.3; , 48 (J.R.Gillig and K.Ways, personal communication). 379196 and its close relative LY333531 (27) have been used to probe the involvement of PKCß in several cellular processes (28–32).
Antibodies
A polyclonal rabbit anti-Cx43 antiserum was raised against the 20 C-terminal amino acids of Cx43. A tyrosine was added N-terminally to this peptide for coupling purposes. The antiserum recognizes Cx43 from many cell types both in western blots and in immunofluorescence experiments (6,11,12,33–36). Goat anti-rabbit and goat anti-mouse secondary antibodies conjugated to horseradish peroxidase were from Bio-Rad (Hercules, CA). For the PKC blots shown PKCµ was detected with rabbit polyclonal anti-PKCµ (C-20; Santa Cruz Biotechnology, Santa Cruz, CA) and the remaining PKC isoenzymes were detected with monoclonal anti-PKC isoenzyme antibodies from Transduction Laboratories (Lexington, KY). In addition, anti-PKC isoenzyme antibodies from several other companies were tried.
Cell cultures
R6-C1 and R6-PKC3 cells were kind gifts from I.B.Weinstein. Many of their properties have been extensively described (21,22). The cells were initially grown with a selection pressure of 200 µg/ml G418 (21). In later work the selection pressure was decreased to 50 µg/ml G418 (22) and this is the standard condition we have used in the major part of this work. With regard to saturation density and population doubling time, R6-PKC3 cells were the most TPA-sensitive clone generated by Housey et al. and had the highest PKC activity (21). The cells were grown in Dulbecco's modified Eagle's medium (BioWhittaker, Walkersville, NY) supplemented with 10% fetal calf serum (Life Technologies, Paisley, UK) at 37°C in a humidified 10% CO2 atmosphere. For the selection pressure experiments the cells were grown in mass cultures in medium containing 0, 5, 50 or 500 µg/ml G418 (Sigma) for at least 2 weeks before any experiment. Plating was done 48 h before the experiments and the same concentration of G418 was still present. In all other experiments the cells were grown in mass cultures with 50 µg/ml G418 (22) in the medium. Again, the cells were plated 48 h before the experiments, but G418 was now omitted from the medium. In all cases the cells did not later receive any fresh medium or serum to avoid any potential influence of fresh fetal calf serum. In the experiments where 24 h exposures to TPA were performed, TPA was added 24 h after seeding. Except for during handling of cells (seeding, addition of compounds, microinjection, etc.) they were always kept at 37°C in a humidified 10% CO2 atmosphere. The microinjections, sampling for western blots or PKC measurements were always done using newly confluent cells.
In all experiments where PKC inhibitors were used, the inhibitors were added 30 min before exposure to TPA and the incubation was continued for 1 h in the presence of both compounds.
Western blotting
The cells were homogenized by sonication. Approximately 10–50 µg total cell protein/sample (depending on the protein of interest) were separated by SDS–PAGE and blotted to nitrocellulose membranes. Protein was measured by the Bradford procedure (Bio-Rad). The membranes were blocked with 5% dry skimmed milk in phosphate-buffered saline and probed with the primary antibodies in suitable dilutions (between 1:250 and 1:2000). Secondary antibodies were used at dilutions of 1:1000 (development with chloronaphthol for detection of Cx43) or 1:50 000 to 1:100 000 (chemiluminescent detection of PKC isoenzymes; Pierce, Rockford, IL). Positive controls for the PKC isoenzymes were obtained from Transduction Laboratories. If the antibodies were raised against available peptides, these peptides were used to verify the specificity of the detected bands. To quantify the amount of PKC isoenzymes, the films were scanned with a Molecular Dynamics laser densitometer using ImageQuant software. Usually all lanes to be compared within a blot contained the same amount of protein. If the amounts differed, the densitometry was correspondingly corrected. All western blots were done at least three times with consistent results.
PKC enzyme activity
Two assays for PKC activity were used. A 32P incorporation (from [-32P]ATP) assay was supplied as a kit and was performed as suggested by the supplier (RPN77; Amersham, Little Chalfont, UK). Briefly, the cells were homogenized by sonication in a buffer containing 50 mM Tris–HCl, pH 7.5, 0.3% (w/v) ß-mercaptoethanol, 5 mM EDTA, 10 mM EGTA, 10 mM benzamidine, 2 mM phenylmethylsulfonyl fluoride, 10 µg/ml pepstatin A, 10 µg/ml aprotinin and 10 µM leupeptin. The latter four compounds were added just before use. The PKC assay component mixture contained equal volumes of calcium buffer (12 mM calcium acetate, initial concentration), lipid buffer (0.3 mg/ml L--phosphatidyl-L-serine, 24 µg/ml TPA), peptide buffer (900 µM peptide) and dithiothreitol buffer (30 mM dithiothreitol). The four buffers also contained 50 mM Tris–HCl, pH 7.5, with 0.05% (w/v) sodium azide. An aliquot of 25 µl of this component mixture was used per assay. A suitable amount of sample was diluted to 25 µl in homogenization buffer and added to the component mixture. The reaction was started by adding 5 µl magnesium ATP buffer (1.2 mM ATP, 30 mM HEPES, 72 mM MgCl2, pH 7.4) with a suitable concentration of radioactive ATP. After incubation the mixture was applied to paper discs and washed in 75 mM phosphoric acid. The remaining radioactivity was determined by scintillation counting. It was confirmed that the assay was linear with regard to time and protein amounts used. Assays without peptide buffer were used as blanks. All data shown were obtained with the radioactivity incorporation assay. The other assay was a non-radioactive ELISA kit (Calbiochem, San Diego, CA). A serine-substituted PKC pseudosubstrate was used as substrate and the phosphorylated peptide was detected with a phospho-specific antibody. This assay was also performed as suggested by the supplier.
GJIC assayed by dye transfer
GJIC was assayed by microinjection of Lucifer Yellow (Sigma) into single cells, followed by fixation in 2% formaldehyde in phosphate-buffered saline 5–8 min after microinjection and counting the number of dye-containing cells (11). For easier comparison the data are presented as means ± SEM of normalized values (i.e. controls are defined as the 100% level). Statistical significances were calculated on the raw data using Student's t-test for comparison between two means or ANOVA with Bonferroni multiple comparisons post-test for comparison of the time or dose curves for each of the cell types. For comparison of the curves between the two cell types the normalized data were used in the non-parametric Kruskal–Wallis test with Dunn's post-test (GraphPad InStat 3.0; GraphPad Software, San Diego, CA).
We also studied whether selective GJIC occurred in the two cell populations. Selective GJIC is defined as GJIC within each of two cell populations with no GJIC between the two populations (37,38). One of the cell populations was labeled with the lipophilic fluorescent dye PHK26 (Sigma). Then the two cell populations were seeded together in different ratios (1:9, 1:1 and 9:1) and after 20 h of co-culture dye transfer between the two cell populations was assayed by microinjection.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
PKC isoenzyme constitution
In accordance with previous results (22), we found R6 cells to express PKC
, and (Figure 1A
). As expected, R6-PKC3 cells also contained large amounts of PKCßI (Figure 1A). An anti-PKC antibody clearly recognized a protein in the expected area (not shown). As this antibody may show cross-reaction with PKC (according to the Transduction Laboratories catalog), we immunoprecipitated PKC from the cell lysates. No anti-PKC immunoreaction was seen in the supernatant, while the immunoprecipitate reacted strongly (not shown). Other PKC isoenzymes were minimally co-immunoprecipitated. Other anti-PKC antibodies did not recognize any bands in the area (not shown). The cells therefore probably do not express PKC. PKC and were also not found in either cell line (not shown). On the other hand, the cells expressed PKCµ (Figure 1A). In addition, we found anti-PKC to react strongly against both cell lines (Figure 1A). In rodents this isoenzyme has usually been called PKC, but we here prefer to call it PKC. R6 cells were previously reported to contain PKC (22), but we were unable to confirm this due to lack of reliable antibodies (several antibodies from different suppliers were tested). However, the 184 amino acid long peptide used to produce the anti-PKC antibody is ~80% identical to mouse PKC. Hence, it is possible that this anti-PKC antibody can detect PKC in R6 cells.
|
Generally, TPA-responsive isoenzymes, except for PKCµ (39), are sensitive to TPA-induced down-regulation during long-term exposures. It was reported that PKC
and in R6-PKC3 cells were resistant to TPA-induced down-regulation (22). However, in the present experiments most of the cPKCs and nPKCs decreased to <10% of the amount in unexposed cells after 24 h exposure to 50 or 500 nM TPA (Tables I and II). The content of PKC and ß in R6-PKC3 cells decreased to ~10%. The result for PKC may be explained by a cross-reaction of the anti-PKC antibody with remaining PKCß, as this antibody cross-reacts with PKCß (Transduction Laboratories catalog). Some down-regulation of PKC and in R6-C1 cells and PKC in R6-PKC3 cells could already be detected after 1 h exposure to TPA (Figure 1A
and Tables I and II). PKCµ was apparently somewhat down-regulated at 1 and 24 h TPA treatment (Figure 1A and Tables I and II). This is probably not the case, as the present antibody may be sensitive to certain phosphorylations occurring in its epitope (see Discussion). PKC was minimally affected by TPA treatment.
|
|
Different concentrations of G418 (0, 5, 50 or 500 µg/ml) were used to study whether expression of PKCßI or other isoenzymes was changed in the cells depending on selection pressure. There was an increase in the amount of PKCßI with increasing concentrations of G418 (Figure 1B
), but PKCßI was easily detectable even without selection pressure (Figure 1B, 0 µg/ml G418). PKC isoenzymes other than PKCßI were not appreciably affected in either of the cell types (not shown), except for PKC in R6-PKC3 cells (probably due to the antibody cross-reaction with PKCß described above).
PKC enzyme activities
The total PKC activity in R6-PKC3 cells was ~10-fold higher than in R6-C1 cells as measured by 32P incorporation (Table III). This is less than the previously published 50-fold difference (21). We therefore also measured PKC activity using a spectrophotometric ELISA kit and again found an ~15-fold higher activity in R6-PKC3 cells (not shown), although the activity in R6-C1 cells was close to the detection limit. Consequently, the results described below were obtained with the radioactivity incorporation assay. Long-term (24 h) exposure to a high concentration of TPA (500 nM) decreased PKC activity to ~10% of the control level (Table III). Western blots of the PKC isoenzymes suggested that there was already a detectable decrease in the amount of some isoenzymes after 1 h exposure to TPA (Figure 1A and Tables I and II). We therefore measured PKC activities in cells exposed to 150 nM TPA for 1 h and found a 50% decrease in both cell types (Table III). Essentially identical enzyme activities and ratios were obtained whether the blanks were defined as assays without peptide or without phosphatidylserine/TPA (not shown).
|
We investigated whether selection pressure could affect the total PKC activity. Regardless of the selection pressure, R6-C1 cells kept a constant PKC activity, while R6-PKC3 cells showed increasing PKC activity with increasing concentration of G418 (Table IV
). However, even without selection pressure R6-PKC3 cells had a PKC activity seven times higher than R6-C1 cells. This increased to a 16-fold difference at the highest selection pressure (Table IV). Thus, there appeared to be a good correlation between the increasing amount of PKCßI (Figure 1B) and increasing PKC activity (Table IV) caused by selection pressure.
|
GJIC
R6-C1 cells showed a relatively high dye transfer of Lucifer Yellow, averaging 23–28 cells/injection. Somewhat surprisingly, R6-PKC3 cells had a significantly (P < 0.0001, Student's t-test) higher level of dye transfer, averaging 40–45 cells/injection (>>1000 injections performed for each of the cell lines under untreated or vehicle-treated control conditions). This suggested that expression of PKCßI did not cause a decrease in GJIC. However, as R6-PKC3 cells occupy a smaller surface area in the dish than R6-C1 cells, the total area of dye spread following microinjection was approximately similar in the two cell lines. R6-PKC3 cells are transformed according to criteria such as increased saturation density and growth in soft agar and forming tumors in nude mice (21). In cancer cells and transformed cells selective communication has been hypothesized to occur and to be involved in development of the transformed phenotype (37,38). We therefore investigated whether Lucifer Yellow could be transferred from R6-C1 to R6-PKC3 cells and vice versa. The two cell populations communicated well with each other, regardless of the population injected, the population labeled with PHK26 and the ratio of labeled cells used (data not shown). A more analytical study of GJIC, e.g. by fluorescence recovery after photobleaching (40), would give a more exact answer as to whether the rates of dye transfer are different in the homologous and heterologous conditions. However, these results suggested that the altered growth characteristics of R6-PKC3 cells have not prominently influenced the basic GJIC characteristics of normal cells.
We studied the effects of TPA and DOPPA on GJIC in the two cell lines. GJIC was dose-dependently decreased by TPA in both cell lines. When normalized (i.e. unexposed cells were defined as the 100% GJIC level) there was a minimal difference in sensitivity to TPA, with IC50 
40 nM (Figure 2A). Thus, the expression of high levels of PKCßI minimally increased the sensitivity of R6-PKC3 cells with regard to inhibition of GJIC. The suppression of GJIC in R6-C1 cells by DOPPA was relatively low (Figure 2A). In contrast, GJIC in R6-PKC3 cells was more strongly decreased, reaching 25% of control at a concentration of 1500 nM, with IC50 300 nM (Figure 2A). Thus, DOPPA was ~5- to 10-fold less potent than TPA in R6-PKC3 cells, while the difference was much larger in R6-C1 cells. This suggests that PKCßI can affect GJIC, but other phorbol ester-sensitive isoenzymes are probably more important during exposure to TPA.
|
and 
, cells exposed to TPA; 
and 
, cells exposed to DOPPA. n = 29–61 injections/point. Statistical significances (Bonferroni test) of P < 0.001 was reached at 500 nM DOPPA and 15 nM TPA for R6-C1 cells and 15 nM for both TPA and DOPPA in R6-PKC3 cells. (B) Kinetics of TPA (150 nM)-induced down-regulation of GJIC. n = 34–48 injections/point. Symbols as in (A). No statistically significant differences were found between the two cell types (Kruskal–Wallis test).We investigated whether the content of PKCßI and higher PKC activity caused R6-PKC3 cells to respond faster to TPA. We studied the level of GJIC after exposure to 150 nM TPA for 5 (injections started immediately after exposure, were carried on for ~3 min and cells were fixed 5 min after the last injection), 15, 30 and 60 min. There were only small and non-significant differences in the kinetics of suppression of GJIC (Figure 2B
).
The cells were cultured under different selection pressures with G418 (0, 5, 50 or 500 µg/ml) to study the impact of increasing PKC activity and PKCßI amounts (Table IV and Figure 1B). In these experiments G418 was present when the cells were plated for the experiments. GJIC in otherwise untreated control cultures remained at the levels described above and was thus independent of selection pressure. The sensitivity to TPA of R6-C1 cells was unchanged by high or low selection pressure, with IC50 values in the range 15–30 nM (Figure 3A). In contrast, R6-PKC3 cells increased in sensitivity to TPA with increasing concentrations of G418 (Figure 3B), with IC50 values of >150, ~150, 25 and 8 nM for 0, 5, 50 and 500 µg/ml G418, respectively. There was also a selection pressure-dependent increase in sensitivity of R6-PKC3 cells to DOPPA, with IC50 values of >>1500, >1500, ~400 and ~150 nM at 0, 5, 50 and 500 µg/ml G418, respectively (not shown). Thus, increasing PKC activity (Table IV) appeared to make R6-PKC3 cells more sensitive to phorbol esters. However, in spite of a 7- to 8-fold higher PKC activity in R6-PKC3 cells cultured at 0 and 5 µg/ml G418, they were actually significantly less sensitive to TPA than R6-C1 cells (see legend to Figure 3). Only R6-PKC3 cells grown at 500 µg/ml G418 were significantly more sensitive to TPA than R6-C1 cells (Figure 3). Thus, a high PKC activity alone did not confer increased sensitivity to TPA.
|
, •), 5 (
, 
) µg/ml G418 before the experiments and G418 was also present during the experiments. The cells were exposed to the indicated concentrations of TPA for 1 h before assessing GJIC. For each of the concentrations of TPA none of the points in (A) are significantly (Bonferroni test) different from each other. In (B) 5, 15, 50 and 150 nM TPA in the curve for 500 µg/ml G418 and 15, 50 and 150 nM TPA in the curve for 50 µg/ml G418 are significantly different (P < 0.001) from the corresponding points in the curve for 0 µg/ml G418. The curve for 50 µg/ml G418 is significantly different (P < 0.01) from the curves for 5 and 500 µg/ml G418 at 15, 50 and 150 nM TPA. Comparison of (A) and (B) (Kruskal–Wallis test): at 0 µg/ml G418, R6-PKC3 cells are significantly less responsive than R6-C1 cells at 5, 15, 50 (all at P < 0.001) and 150 (P < 0.01) nM TPA; at 5 µg/ml G418, R6-PKC3 cells are less responsive than R6-C1 cells at 15 (P < 0.001) nM TPA; there is no statistical difference between the curves for R6-PKC3 and R6-C1 cells at 50 µg/ml G418; at 500 µg/ml G418, R6-PKC3 cells are more responsive than R6-C1 cells at 15 (P < 0.001), 50 (P < 0.001) and 150 (P < 0.01) nM TPA.To further determine the relative importance of cPKCs and nPKCs, the broad specificity PKC inhibitor GF109203X, the cPKC/PKCµ inhibitor Gö6976 and the PKCß-specific inhibitor 379196 were used. GF109203X efficiently counteracted the inhibition of GJIC induced by TPA (Figure 4A
) and DOPPA (not shown) in a dose-dependent manner without affecting GJIC on its own. In contrast, Gö6976 alone was able to significantly increase GJIC when used at 3 µM or higher concentrations (Figure 4B). If this was due to the inhibition of constitutively active cPKCs we would have expected GF109203X to do the same, unless there is some kind of interaction between cPKC and nPKC isoenzymes. We therefore suspect that Gö6976 also has other actions in the cells. When 1 µM Gö6976 was used only a partial inhibition of the TPA-induced decrease in GJIC was obtained (Figure 4B). However, the relatively small effect of 500 nM DOPPA in R6-C1 cells was fully counteracted by 1 µM Gö6976 (not shown). When the concentration of Gö6976 was further increased inhibition of the TPA effect was only minimally increased. These results could suggest that both cPKCs and nPKCs participate in TPA-induced inhibition of GJIC.
|
The ability of 379196 to inhibit the TPA-induced decrease in GJIC was investigated under G418 selection pressure, as this should best disclose the potential involvement of PKCßI. 379196 alone did not affect GJIC in either R6-C1 or R6-PKC3 cells (Figure 4C
). It counteracted the TPA-induced decrease in GJIC in a dose-dependent manner at concentrations >30 nM (Figure 4C). We investigated whether 379196 (30 nM, with 30 min pre-exposure) had a more pronounced inhibitory effect during shorter TPA exposures (15 and 30 min). The effect of 379196 was still minimal and not statistically significant (not shown). Since the effect of 379196 was similar in both cell types, we conclude that PKCßI is not among the major PKC isoenzymes involved in regulation of GJIC.
cPKCs and nPKCs were strongly down-regulated by 24 h TPA exposures. Among the TPA-responsive isoenzymes only PKCµ was left in appreciable amounts (Figure 1A and Tables I and II). In parallel, GJIC was nearly normalized in both cell lines after 24 h exposure to 50 (Figure 5) and 500 nM TPA (not shown). Re-exposure to 50, 150 or 500 nM TPA did not cause any decrease in GJIC in R6-PKC3 cells, but a small, although statistically significant, decrease was seen in R6-C1 cells at 500 nM TPA (Figure 5). Thus, PKCµ appeared as a minor mediator of TPA-induced inhibition of GJIC under the present conditions.
|
Cx43
R6-C1 and R6-PKC3 cells were found to express Cx43 in an identical pattern (Figure 6
). Cx43 was mainly found in two variants, the non-phosphorylated form (NP-Cx43) and the phosphorylated form, called P1-Cx43. The P'- and P2-Cx43 forms were present in minor amounts. We investigated whether the phosphorylation status of Cx43 became altered in response to TPA or DOPPA. When exposed to TPA or DOPPA the intensity of the NP- and P1-Cx43 forms decreased and those of the P' and P2 forms increased (Figure 6). The changes could already be seen at 5 nM TPA, a concentration that minimally affected GJIC, and were well developed at 15 nM TPA in both cell lines. The clearest difference between the two cell lines was that the P1 band became slightly more diffuse in R6-PKC3 than in R6-C1 cells in most experiments. Consistent with the results for GJIC, DOPPA was more active in R6-PKC3 cells, causing the same band pattern changes as described for TPA, but starting at 50–150 nM (Figure 6B). In R6-C1 cells DOPPA caused weak Cx43 band pattern changes only at 500–1500 nM (Figure 6A).
|
As for GJIC, the kinetics of changes in Cx43 phosphorylation status were investigated. At 150 nM TPA the first changes could already be seen at 2.5 min and were evident at 5 min (Figure 6C
). The kinetics in R6-C1 cells were very similar to R6-PKC3 cells. When the concentration of TPA was lowered R6-C1 cells continued to respond as rapidly as R6-PKC3 cells. This is exemplified by 15 nM TPA (Figure 6D).
We investigated whether differential selection pressure by G418 caused differential sensitivity of the Cx43 phosphorylation changes to TPA. There were minimal differences in TPA sensitivity for R6-PKC3 (Figure 7A) and R6-C1 cells (not shown) grown at 0 and 500 µg/ml G418. Similar results were obtained for DOPPA (Figure 7B). This suggests that high PKC activity does not make Cx43 more prone to phorbol ester-induced phosphorylation and that PKCßI is not among the major isoenzymes mediating this phosphorylation.
|
To further determine the isoenzymes that mediated phosphorylation of Cx43 the PKC inhibitors GF109203X, Gö6976 and 379196 were used. When used alone, GF109203X and 379196 did not affect Cx43 phosphorylation during 1.5 h exposures (Figures 8A and B and 9A
), while Gö6976 tended to decrease the intensity of the P' band (Figure 8C and D). GF109203X (0.3–1 µM) counteracted the effects of 150 nM TPA in R6-C1 cells (Figure 8A), while 3 µM was needed in R6-PKC3 cells (Figure 8B). GF109203X also inhibited the effect of DOPPA (500 nM) in R6-PKC3 cells (not shown). Gö6976 counteracted the band pattern changes caused by 150 nM TPA at concentrations of 0.3 µM for R6-C1 (Figure 8C) and 1 µM for R6-PKC3 cells (Figure 8D), in spite of the fact that it did not fully normalize GJIC under the same conditions (Figure 4B). Gö6976 (1 µM) also inhibited the effect of DOPPA (500 nM) in R6-PKC3 cells (not shown). This could suggest the involvement of cPKCs in the phosphorylation of Cx43. We therefore investigated whether 379196 counteracted TPA. Again, cells under differential G418 selection pressure were used, since this should potentially make the involvement of PKCßI easier to detect. 379196 showed only a weak effect in R6-PKC3 cells at concentrations of 100 and 300 nM at both low and high selection pressure and the same effect was also evident in R6-C1 cells (Figure 9B and C). This suggests that PKCßI is minimally involved in TPA-induced phosphorylation of Cx43 and, accordingly, implicate PKC as a major TPA-activated Cx43 kinase in this system.
|
|
Exposure to TPA for 24 h caused a nearly complete normalization of Cx43 band pattern and no new changes occurred when the cells were re-exposed to different concentrations of TPA (Figure 10
). As PKCµ is the only TPA-responsive PKC isoenzyme present in significant amounts, it is concluded that PKCµ is not among the major Cx43 phosphorylating PKC isoenzymes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Does high PKC enzyme activity make cells more sensitive to TPA? To answer this question we have compared PKC activity with the responsiveness of GJIC and Cx43 phosphorylation to TPA. Under non-stimulated conditions a minority of the PKC molecules are assumed to be active. After exposure to TPA there is strong activation of PKC as assayed by phosphorylation of the endogenous substrate p80, also called MARCKS (see for example ref. 25). R6-PKC3 cells have a PKC enzyme activity ~7- to 16-fold higher than R6-C1 cells, depending on the selection pressure (Table IV
). This is less than the previously reported 50-fold difference for cells cultured with 200 µg/ml G418 and is mainly caused by a 3- to 5-fold higher PKC activity in R6-C1 cells in the present work than found by Housey et al. (21). However, the difference in PKC activity between the two cell lines is still prominent and should be sufficiently large to determine whether an increased level of PKC activity in itself causes a higher sensitivity to TPA. An intersystem comparison of the GJIC results clearly shows that R6-C1 cells are as, or even more, responsive than R6-PKC3 cells, except under high selection pressure. If we consider only R6-PKC3 cells, there is an increased sensitivity of GJIC to TPA with increasing PKC activity. However, it is surprising that a doubling of the PKC activity in R6-PKC3 cells caused by differential selection pressure (0 and 500 µg/ml G418) decreases the IC50 in the GJIC assay from >150 to ~10 nM, especially when the high PKC activity in cells cultured without G418 is considered. Furthermore, there appear to be minimal changes in the response of Cx43 in R6-PKC3 cells between no selection pressure and high selection pressure. In addition, differences in the response of Cx43 from R6-C1 and R6-PKC3 cells are nearly absent. Thus, high PKC activity alone does not necessarily cause an increased sensitivity to TPA as assayed by GJIC or Cx43 phosphorylation. There are probably other, as yet unidentified, factors also involved in determining this sensitivity. Such factors could include interaction among the PKC isoenzymes expressed and other signal transduction pathways modulated by PKC isoenzymes. It may also be noted that the low correlation between changes in GJIC and changes in Cx43 phosphorylation could be due to the presence of other unidentified connexins, as has been pointed out for other systems (12), or the fact that some Cx43 phosphorylation variants may have an identical migration rate during electrophoresis (33). In the context of total PKC activity and selection pressure, it is also interesting to note that even if these cells produce tumors in nude mice (21), this must occur without G418 selection pressure. Is PKCßI involved in the two processes studied? R6-PKC3 cells were no more sensitive to TPA than R6-C1 cells, except at the highest selection pressure (500 µg/ml G418). Thus, PKCßI is probably not the isoenzyme that mediates the major effect of TPA on GJIC in R6-PKC3 cells. However, the selection pressure-dependent increase in sensitivity to TPA also suggests that when PKCßI is expressed in sufficiently high amounts it can affect GJIC. Interestingly, phorbol ester-induced Cx43 phosphorylation was minimally changed by selection pressure, suggesting that phosphorylation of Cx43 is not mediated by PKCßI, but rather by one (or more) of the isoenzymes not affected by selection pressure.
For further investigations on the importance of PKCßI we used DOPPA and 379196. DOPPA was clearly more active in R6-PKC3 than in R6-C1 cells and needed only a 5- to 10-fold higher concentration than TPA to cause similar effects, with IC50 40 nM for TPA and 300 nM for DOPPA, as measured by inhibition of dye transfer in R6-PKC3 cells. The relative insensitivity of R6-C1 cells to DOPPA indicates that DOPPA is a poor activator of other PKC isoenzymes that potentially affect GJIC and Cx43. As the two cell lines contain an identical complement of PKC isoenzymes except for PKCßI, this implies that in the absence of direct activation (or during weak activation) of other PKC isoenzymes, PKCßI may influence the two parameters studied. Whether the influence of PKCßI occurs directly or indirectly (e.g. through an interplay with other PKC isoenzymes) remains to be studied. The results obtained with 379196 support the notion that PKCßI is minimally involved in the TPA-induced decrease in GJIC, especially since the inhibition curves were similar in PKCßI-containing R6-PKC3 cells and in PKCßI non-containing R6-C1 cells (Figure 4C). These results do not exclude the possibility that PKCß can influence GJIC in other cell systems, but this may best be investigated in a system that naturally expresses one or both of these isoenzymes.
As discussed above, under most of the conditions studied here PKCßI appears to be of less importance than some other PKC isoenzymes for the present parameters. Which PKC isoenzymes then are involved in the TPA-induced decrease in GJIC? PKC, , and were previously identified in R6 cells (22). We also found PKCµ and , although the latter could be a mixture with PKC. As expected, long-term treatment with TPA induced down-regulation of PKC. However, in contrast to Borner et al. (22), we did not find an increased resistance to down-regulation of PKC and in R6-PKC3 cells. PKCµ was the only TPA-responsive PKC isoenzyme partially resistant to TPA-induced down-regulation in R6 cells. However, the present anti-PKCµ antibody is sensitive to phosphorylation of its epitope following TPA exposure (39). Thus, there is probably no down-regulation of PKCµ by long-term TPA treatment, which would be consistent with previous results (20,39). The presence of PKCµ together with the lack of responsiveness after 24 h exposure to TPA directly implies that PKCµ is not involved in regulation of the parameters under study. Moreover, PKCµ shows strong selectivity for peptides with a leucine at position –5 (N-terminal relative to the phosphorylated serine/threonine) and mainly hydrophobic amino acids at other positions, in addition to the common preference of PKC isoenzymes for a positive charge at position –3 (41). Cx43 has only one threonine (position 118) placed five amino acids from a leucine, however this sequence does not fit well with the other preferences of PKCµ. The sequence is located in the cytoplasmic loop. As yet, all are phosphorylations of Cx43 is believed to take place in the C-terminal tail.
An interesting question is whether the same isoenzymes influence GJIC and phosphorylation of Cx43? Somewhat provocatively, it is not necessarily the case in R6 cells. This is especially indicated by two observations. First, the condition with high selection pressure (500 µg/ml G418) caused a relatively strong sensitization of R6-PKC3 cells with respect to GJIC, but not Cx43 phosphorylation. Second, Gö6976 partially inhibited the TPA-induced decrease in GJIC, but fully inhibited TPA-induced changes in Cx43 band pattern. This could suggest that cPKCs (mainly ) are responsible for phosphorylation of Cx43 after TPA exposure in the present cells, while both cPKCs (mainly ) and nPKCs ( and/or ) could be involved in the TPA-induced decrease in GJIC. Consistently, lens epithelial cells express PKC and GJIC is decreased and the Cx43 phosphorylation status is changed in response to TPA (42).
Two important caveats should be mentioned. First, extension of IC50 values from cell-free systems to intact cells may be problematical. GF109203X has IC50 = 8–200 nM for cPKCs and nPKCs (43) and IC50 = 2 µM for PKCµ (44). Gö6976 inhibits cPKCs with IC50 = 2–8 nM (26), but is also a good inhibitor of PKCµ, with IC50 = 20 nM (44). In both cases counteraction of TPA was only achieved at concentrations well above the IC50 values for the relevant isoenzymes. Second, the inhibitors may affect other enzymes that could also influence GJIC and/or phosphorylation of connexins. GF109203X is generally regarded as a specific inhibitor of PKC, but is also a potent inhibitor of MAPKAP kinase-1ß, p70 S6 kinase and glycogen synthase kinase-3 (45,46). Thus, GF109203X may interfere with two other important signaling pathways, the MAP kinase pathway and the phosphatidylinositol 3-kinase/mammalian target of rapamycin pathway. PKC isoenzymes influence, or are involved in, both pathways (47–50). The MAP kinase pathway was previously shown to be involved in regulation of GJIC and phosphorylation of Cx43 after stimulation with growth factors (6,51). We recently showed that rapamycin delayed the normalization of Cx43 phosphorylation status in V79 hamster fibroblasts after a short exposure to TPA (52). It must therefore be investigated whether these pathways are involved in TPA-induced effects on GJIC and phosphorylation of Cx43. On the other hand, Gö6976 increased GJIC in a manner reminiscent of some cAMP-elevating compounds. In fact, it was recently shown that Gö6976 may affect the metabolism of cyclic nucleotides (53). Moreover, cAMP-elevating compounds have been shown to counteract the effects of TPA on GJIC (54).
In summary, total PKC enzyme activity in a cell system does not indicate the absolute sensitivity of the cells to TPA with regard to inhibition of GJIC or phosphorylation of Cx43. Although PKCßI may affect GJIC and phosphorylation of Cx43 after stimulation with TPA, other PKC isoenzymes are of more importance. In particular, PKC may be involved in phosphorylation of Cx43, while PKC, and/or may participate in the TPA-induced decrease in GJIC.
|
Notes |
|---|
1 Present address: Department of Environmental Medicine, Institute of Public Health, N-0462 Oslo, Norway
2 To whom correspondence should be addressed E-mail: s.o.mikalsen{at}labmed.uio.no
The first two authors should be regarded as joint first authors.
A preliminary report of these data was presented at the 1997 International Gap Junction Conference, Key Largo, FL, July 12–17, 1997.
|
Acknowledgments |
|---|
We thank Dr E.Rivedal for making the anti-Cx43 antiserum available to us. Dr I.B.Weinstein (Institute of Cancer Research, Colombia University, NY) is gratefully acknowledged for the gift of R6-C1 and R6-PKC3 cells. Lilly Research Laboratories are gratefully acknowledged for the gift of 379196. This work was supported by grants from the Norwegian Cancer Society (to T.H.), The Research Council of Norway (to V.C.), the Faculty of Medicine, University of Oslo (to T.H., V.C. and S.-O.M.), the Family Blix' Fund (to V.C. and S.-O.M.) and the Anders Jahre's Fund (to T.S. and S.-O.M.).
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
- Loewenstein,W.R. (1979) Junctional intercellular communication and the control of growth. Biochim. Biophys. Acta, 560, 1–65.[Medline]
-
Yamasaki,H. and Naus,C.C.G. (1996) Role of connexin genes in growth control. Carcinogenesis, 17, 1199–1213.
[Free Full Text] - Lo,C.W. (1996) The role of gap junction membrane channels in development. J. Bioenerg. Biomembr., 28, 379–385.[ISI][Medline]
-
Manthey,D., Bukauskas,F., Lee,C.G., Kozak,C.A. and Willecke,K. (1999) Molecular cloning and functional expression of the mouse gap junction gene connexin-57 in human HeLa cells. J. Biol. Chem., 274, 14716–14723.
[Abstract/Free Full Text] -
Musil,L.S., Cunningham,B.A., Edelman,G.M. and Goodenough,D.A. (1990) Differential phosphorylation of the gap junction protein connexin43 in junctional communication-competent and -deficient cell lines. J. Cell Biol., 111, 2077–2088.
[Abstract/Free Full Text] - Vikhamar,G., Rivedal,E., Mollerup,S. and Sanner,T. (1998) Role of Cx43 phosphorylation and MAP kinase activation in EGF induced enhancement of cell communication in human kidney epithelial cells. Cell Adhes. Commun., 5, 451–460.[ISI][Medline]
- Mikalsen,S.-O. and Kaalhus,O. (1996) A characterization of pervanadate, an inducer of cellular tyrosine phosphorylation and inhibitor of gap junctional intercellular communication. Biochim. Biophys. Acta, 1290, 308–318.[Medline]
- Mikalsen,S.-O. and Kaalhus,O. (1997) A characterization of permolybdate and its effect on cellular tyrosine phosphorylation, gap junctional intercellular communication and phosphorylation status of the gap junction protein, connexin43. Biochim. Biophys. Acta, 1356, 207–220.[Medline]
-
Yotti,L.P., Chang,C. and Trosko,J.E. (1979) Elimination of metabolic cooperation in Chinese hamster cells by a tumor promoter. Science, 206, 1089–1091.
[Abstract/Free Full Text] - Murray,A.W. and Fitzgerald,D.J. (1979) Tumour promoters inhibit metabolic cooperation in cocultures of epidermal and 3T3 cells. Biochem. Biophys. Res. Commun., 91, 395–401.[ISI][Medline]
-
Husøy,T., Mikalsen,S.-O. and Sanner,T. (1993) Phosphatase inhibitors, gap junctional intercellular communication and [125I]EGF binding in hamster fibroblasts. Carcinogenesis, 14, 2257–2265.
[Abstract/Free Full Text] - Cruciani,V., Mikalsen,S.-O., Vasseur,P. and Sanner,T. (1997) Effects of peroxisome proliferators and 12-O-tetradecanoyl phorbol-13-acetate on intercellular communication and connexin43 in two hamster fibroblast systems. Int. J. Cancer, 73, 240–248.[ISI][Medline]
- Oh,S.Y., Grupen,C.G. and Murray,W. (1991) Phorbol ester induces phosphorylation and down-regulation of connexin 43 in WB cells. Biochim. Biophys. Acta, 1094, 243–245.[Medline]
-
Husøy,T., Mikalsen,S.-O. and Sanner,T. (1993) Effects of five phorbol esters on gap junctional intercellular communication, morphological transformation and epidermal growth factor binding in Syrian hamster embryo cells. Carcinogenesis, 14, 73–77.
[Abstract/Free Full Text] -
Enomoto,T. and Yamasaki,H. (1985) Rapid inhibition of intercellular communication between BALB/c 3T3 cells by diacylglycerol, a possible endogenous functional analogue of phorbol esters. Cancer Res., 45, 3706–3710.
[Abstract/Free Full Text] -
Pasti,G., Rivedal,E., Yuspa,S.H., Herald,C.L., Pettit,G.R. and Blumberg,P.M. (1988) Contrasting duration of inhibition of cell–cell communication in primary mouse epidermal cells by phorbol 12,13-dibutyrate and by bryostatin 1. Cancer Res., 48, 447–451.
[Abstract/Free Full Text] - Sáez,J.C., Nairn,A.C., Czernik,A.J., Spray,D.C. and Hertzberg,E.L. (1993) Rat connexin43: regulation by phosphorylation in heart. Prog. Cell Res., 3, 275–281.
- Kwak,B.R., Van Veen,T.A.B., Analbers,L.J.S. and Jongsma,H.J. (1995) TPA increases conductance but decreases permeability in neonatal rat cardiomyocyte gap junction channels. Exp. Cell Res., 220, 456–463.[ISI][Medline]
-
Hayashi,A., Seki,N., Hattori,A., Kozuma,S. and Saito,T. (1999) PKC, a new member of the protein kinase C family, composes a fourth subfamily with PKCµ. Biochim. Biophys. Acta, 1450, 99–106.[Medline]
- Johannes,F.-J., Prestle,J., Dieterich,S., Oberhagemann,P., Link,G. and Pfizenmaier,K. (1995) Characterization of activators and inhibitors of protein kinase Cµ. Eur. J. Biochem., 227, 303–307.[ISI][Medline]
- Housey,G.M., Johnson,M.D., Hsiao,W.L.W., O'Brian,C.A., Murphy,J.P., Kirschmeier,P. and Weinstein,I.B. (1988) Overproduction of protein kinase C causes disordered growth control in rat fibroblasts. Cell, 52, 343–354.[ISI][Medline]
-
Borner,C., Guadagno,S.N., Fabbro,D. and Weinstein,I.B. (1992) Expression of four protein kinase C isoforms in rat fibroblasts. Distinct subcellular distribution and regulation by calcium and phorbol esters. J. Biol. Chem., 267, 12892–12899.
[Abstract/Free Full Text] - Evans,F.J., Parker,P.J., Olivier,A.E., Thomas,S., Ryves,W.J., Evans,A.T., Gordge,P. and Sharma,P. (1991) Phorbol ester activation of the isotypes of protein kinase C from bovine and rat brain. Biochem. Soc. Trans., 19, 397–402.[ISI][Medline]
-
Ryves,W.J., Evans,A.T., Olivier,A.E., Parker,P.J. and Evans,F.J. (1991) Activation of the PKC-isotypes , ß1, , and by phorbol esters of different biological activities. FEBS Lett., 288, 5–9.[ISI][Medline]
-
Kiley,S.C., Olivier,A.E., Gordge,P.C., Ryves,W.J., Evans,F.J., Ways,D.K. and Parker,P.J. (1994) 12-Deoxyphorbol-13-O-phenylacetate-20-acetate is not protein kinase C-ß isoenzyme-selective in vivo. Carcinogenesis, 15, 319–324.
[Abstract/Free Full Text] -
Martiny-Baron,G., Kazanietz,M.G., Mischak,H., Blumberg,P.M., Kochs,G., Hug,H., Marmé,D. and Schächtele,C. (1993) Selective inhibition of protein kinase C isoenzymes by the indolocarbazole Gö 6976. J. Biol. Chem., 268, 9194–9197.
[Abstract/Free Full Text] - Jirousek,M.R., Gillig,J.R., Gonzalez,C.M., Heath,W.F., McDonald,J.H., Neel,D.A., Rito,C.J., Singh,U., Stramm,L.E., Melikian-Badalian,A., Baevsky,M., Ballas,L.M., Hall,S.E., Winneroski,L.L. and Faul,M.M. (1996) (S)-13-[(dimethylamino)methyl]-10,11,14,15-tetrahydro-4,9:16, 21-dimetheno-1H, 13H-dibenzo[e,k]pyrrolo[3,4-h][1,4,13]oxadiazacyclohexadecene-1,3(2H)-dione (LY333531) and related analogues: isozyme selective inhibitors of protein kinase Cß. J. Med