v8a8-wagner pgmkr - CiteSeerX

Molecular Vision 2002; 8:59-66 Received 23 October 2001 | Accepted 11 March 2002 | Published 14 March 2002

© 2002 Molecular Vision

Effect of protein kinase Cγ on gap junction disassembly in lens epithelial cells and retinal cells in culture Lynn M. Wagner,1 Suha M. Saleh,1 Daniel J. Boyle,2Dolores J. Takemoto3 1

Department of Anatomy and Physiology, 2Division of Biology, 3Department of Biochemistry, Kansas State University, Manhattan, KS Purpose: To determine the effects of protein kinase Cγ(PKCγ) on phosphorylation of Cx43, the gap junction protein of lens epithelial cells, and on cell surface assembly/disassembly of Cx43-gap junction complexes. Methods: Association and phosphorylation of Cx43 by PKCγ was determined using co-immunoprecipitation and reaction with phosphoserine antisera. Activation of PKCγ was with 200 nM phorbol ester for 30 to 60 min. Effects of specific PKC isoforms was determined after overexpression of either PKCα or PKCγ for 24 h in N/N 1003A rabbit lens epithelial cells or in two retinal cell lines, WERI and Y79. Gap junction plaques were counted on the cell surface by immunolabeling of Cx43 using confocal microscopy. Results: Co-immunoprecipitation of Cx43 with PKCγ was observed only in cells over expressing PKCγ and in cells activated with phorbol ester. Both overexpression and phorbol ester produced a rapid phosphorylation of Cx43 on serine. Cx43 cell surface gap junction plaques decreased in cells over expressing PKCγ and in cells treated with phorbol ester. Similar results were observed using the retinal cell lines, WERI and Y79. The effect of PKCγoverexpression was persistent for 7 days but total cell Cx43 was not decreased. Overexpression of PKCα resulted in an increase in cell surface gap junction plaques. Conclusions: PKCγ can be co-immunoprecipitated with Cx43 from lens epithelial cells using phorbol ester activation. PKCγ phosphorylates Cx43 on serine and this causes disassembly and loss of gap junction Cx43 from the cell surface. Overexpression of PKCγconfirmed that only this PKC isoform caused the loss of cell surface Cx43. Overexpression of PKCα, the other major lens PKC isoform, caused an increase in cell surface Cx43. The presence of PKCγ and loss of surface Cx43 from two retinal cell lines, WERI and Y79, upon phorbol ester activation further suggests that activation of PKCγ may be a common mechanism for control of cell surface Cx43.

Cx43, Cx46, and Cx50 [6,7]. While Cx46 and Cx50 are found in lens fiber cells, Cx43 is found in lens epithelial cells. Many effectors of protein kinases modulate gap junctional communication, and it has been demonstrated that certain connexin proteins are phosphorylated [8,9]. Protein kinase Cγ (PKCγ) has been identified as an inhibitor of Cx43 gap junction activity in lens epithelial cells in culture [10]. The mechanism of inhibition has not been defined, however, Cx46 is also phosphorylated by PKCγ[11]. Cx43 is one of the major connexins in the body, and it is expressed in many tissues. Previous studies on lens gap junctions have identified Cx43 as epithelial cell specific [12-14]. Cx43 is a phosphoprotein [15] and its phosphorylation may be coupled to the assembly/disassembly of gap junction plaques, generation of functional channels [13], and regulation of channel pore size and open-state probability [15,16]. Cx43 has been shown to be serine phosphorylated in lens epithelial cells in situ and in primary cultures of lens cells [811,13,16-19]. Activation of PKC by phorbol ester has been found to inhibit gap junction activity [10,18-20]. The PKC isoform which alters gap junction activity is PKCγ [10]. This study indicates that PKCγ can be co-immunoprecipitated from lens epithelial cells with Cx43 and that PKCγ mediates the phosphorylation of Cx43 on serine. Furthermore, only activated PKCγ can accomplish this. Overexpression of

Most cells can communicate with adjacent cells by gap junctions [1]. These membrane structures are clusters, called plaques, of intercellular channels that link the cytosols of adjoining cells and thereby act as direct pathways for the cellto-cell transfer of small (under 1 kD) molecules [2,3]. The importance of gap junction intercellular communication is particularly obvious in the lens since it is an avascular tissue which relies on gap junctional channels for diffusional distribution of small molecules. The lens is composed of two cell types; a monolayer of epithelial cells that overlay the anterior face and a core of elongated crystallin-rich fiber cells that are responsible for the refractive properties of the lens [3]. In this tissue, the anterior epithelial cell monolayer, which interfaces with the aqueous humor, contains many of the transporters and ion pumps and helps to control homeostasis throughout the lens via an extensive network of gap junction channels [4]. A gap junction channel is formed by two hemichannels or connexons which are contributed by each of the adjacent cells. A connexon is an oligomeric unit of six subunits termed connexins (Cxs) [5]. In the lens, gap junction proteins include Correspondence to: Dolores J. Takemoto, Department of Biochemistry, 103 Willard Hall, Kansas State University, Manhattan, Kansas, 66506; Phone: (785) 532-7009; FAX: (785) 532-7278; email: [email protected] 59

Molecular Vision 2002; 8:59-66


Transfection and overexpression of PKCγ and PKCα: N/ N1003A cells were cultured to 50% confluency, then rinsed with serum-free medium and treated with lipofectamine reagent (Life Technologies, Rockville, MD). The cells were transfected with εMTH vector which contained the empty vector, PKCα, or PKCγ. εMTH vector is inducible with 20 µM zinc, and it is selectable with Geneticin (G418), as previously described [10,11]. All cells in these experiments were treated with 20 µM zinc at the start of the experiment. The empty vector transformed cells are referred to as control cells in these experiments. Co-immunoprecipitation of PKCγ with Cx43: Control N/ N1003A cells (empty vector transfected cells) and N/N1003A cells overexpressing PKCα or PKCγ and phorbol ester-treated cells (untransfected) were cultured to 90% confluency in 75 cm2tissue culture flasks, then, the cells were harvested and lysed on ice with 0.5 ml ice cold cell lysis buffer. The cell lysis buffer contained 20 mM Tris (pH 7.5), 0.5 mM EDTA, 0.5 mM EGTA, 0.5% Triton X-100, 25 µg/ml aprotinin, and 25 µg/ml leupeptin. The lysed cells were homogenized and centrifuged at 20,000x g for 20 min, then the supernatant was collected and antibodies specific for PKCγ or PKCα (Transduction Laboratories, San Diego, CA; 1:1000 dilution), were added to the mixture which was a final concentration of 5 µg/ ml, then, incubated overnight at 4 °C with constant mixing. After overnight incubation with PKCγ or PKCα antibodies, 20 µl of protein A sepharose beads (50:50 solution in lysis

PKCγ and not PKCα caused disassembly of gap junctions in lens epithelial cells. This may be a mechanism by which PKCγ causes inhibition of gap junction activity. The phorbol esterinduced disassembly of gap junctions in two retinal cell lines suggests that PKCγ may also be involved in the control of retinal gap junction assembly. METHODS Cell culture: N/N 1003A rabbit lens epithelial cells were cultured in Minimal Essential Medium, pH 7.2, supplemented with 10% fetal bovine serum and 50 µg/ml Gentamicin. The cells were grown at 37 °C in an atmosphere of 90% air and 10% CO2 and used for experiments when they reached 90% confluency. Retinal Y79 and Weri cells were grown as suspension cultures in 15% fetal calf serum in RPMI-1640 media. Cells were a gift of Dr. Stephen Pittler, University of Alabama, Birmingham, AL. In some cases, cells were treated with 200 nM phorbol ester (Sigma, St. Louis, MO) for 60 min or 1 µM Calphostin C (Sigma) [17] for 24 h prior to harvesting.

Figure 1. Immunoprecipitation of PKCγ and PKCα with Cx43. A: PKCγ was immunoprecipitated in control N/N1003A cells (transfected with empty vector, lane 1) and in N/N1003A which over express PKCγ (lane 2). As a control, PKCα was immunoprecipitated in N/N1003A cells which over express PKCα (lane 3). B: Immunoprecipitated proteins were analyzed on 12.5% SDS-PAGE and western blot, then the nitrocellulose membranes which contain the transferred proteins were probed with anti Cx43 antibodies and visualized using chemiluminescent secondary antisera (IgG). The higher molecular weight bands (greater than 56 kDa) represent precipitated immunoglobulins (primary antisera) reactive with the secondary antisera and they were also detected in control experiments (lanes 4 and 5). All cells recieved 20 µM zinc-acetate at the start of the experiment and were allowed to overexpress for 24 h. Lane 1: Control cells precipitated with anti-PKCγ antibodies (Transduction Labs; 1:1000 dilution) and Western blots probed with anti-Cx43 (Chemicon; 1:1000 dilution). Lane 2: Cells overexpressing PKCγwere immunoprecipitated with anti-PKCγ then probed with anti-Cx43. Lane 3: Cells overexpressing PKCα were immunoprecipitated with anti-PKCα and probed with anti-Cx43. Lane 4: PKCγ antisera and Protein A alone, probed with anti Cx43, then secondary antisera. Lane 5: Primary antisera added directly to gel lane and probed with anti-Cx43 and secondary antisera. Lane 6: Protein A alone probed with anti-Cx43. Thus, the primary antisera band at 56 kD is observed in all samples to which it was added. All lanes visualized with anti-IgG-ECL (Pierce).

Figure 2. Effects of Phorbol Ester on Interaction of PKCγ with Cx43. A: Normal, untransfected N/N1003a cells (1x106cells/ppt) were immunoprecipitated with anti PKCγ antisera (Transduction Labs; 1:1000 dilution) and Western blots were probed with anti Cx43 (Chemicon; 1:1000 dilution). TPA was added to the cells at 200 nM for 60 min prior to immunoprecipitation. Lane 1: No TPA, lane 2: TPA added. B: Western blot from A was rinsed in 1 M glycine, pH 2.0 to remove secondary antisera then, reprobed with anti-phosphoserine antibodies (Chemicon; 1:1000 dilution). Lane 1: No TPA, lane 2: TPA added. 60



buffer; Sigma) were added to the mixture and the mixture was further incubated for another two h on ice. The mixture was centrifuged at 11,900x g for 25 s and the precipitate was collected and washed four times with lysis buffer. The washed precipitate was then mixed with 20 µl Tris-glycine/SDS sample buffer and boiled for five min. Proteins were separated by 12.5% SDS-PAGE, transferred to nitrocellulose membranes, and probed with anti-Cx43 antibody (Chemicon, Temecula, CA; 1:1000 dilution), or other antisera as indicated. Immunoreactive bands were detected by chemiluminescence (ECL, Pierce, Rockford, IL). Phosphorylation of connexin 43 by PKCγ: Control N/ N1003A cells (empty vector transfected) and N/N1003A cells overexpressing PKCα or PKCγ were cultured to 90% confluency then harvested, rinsed with PBS, and lysed with lysis buffer, which contained 20 mM Tris (pH 7.5), 0.5 mM EDTA, 0.5 mM EGTA, 0.5% Triton X-100, 25 µg/ml aprotinin, and 25 µg/ml leupeptin. The cell lysate was centrifuged for 20 min at 20,000x g and the supernatant was used to immunoprecipitate Cx43 as described above. Immunoprecipitated Cx43 was analyzed by 12.5% SDS-PAGE and the separated proteins were transferred to a nitrocellulose membrane and probed with anti-phosphoserine, anti-phosphothreonine, or antiphosphotyrosine antibodies (Chemicon; 1:1000 dilution). Translocation of PKCα and PKCγ to membrane fractions: Control N/N 1003A cells (untransfected) or cells treated with either 200 nM phorbol ester (Sigma) for 1 h or 1 µM Calphostin C (Sigma, a commercially available PKC inhibitor [17]) for 24 h were harvested in 50 mM Tris, 20 mM MgCl2 (pH 7.5)

and sonicated. The cell lysate was centrifuged for 1 h at 100,000x g at 4 °C. The supernatant and the pellet were saved and analyzed on a 7.5% SDS-PAGE gel with 20 µg of protein loaded per lane and probed with anti-PKCα or anti-PKCγ antibody (Transduction Laboratories; 1:1000 dilution). Confocal microscopy of gap junction plaques from phorbol ester-treated cells: N/N 1003A lens epithelial cells (untransfected, 1.0x106 total cells), and Weri or Y79 (1.0x105/ ml) retinal cells were treated with 200 nM phorbol ester for 1 h or 1 µM Calphostin C for 24 h prior to fixation. The cells were fixed for 10 min with 2.5% paraformaldehyde in phosphate buffered saline (PBS). To label lens Cx43 and PKCγ, the primary antisera was diluted in blocking buffer (3% BSA in PBS) and then added to the fixed cells and incubated for 18 h at room temperature with; anti-PKCγ (green color, mouse host, Transduction Laboratories; 1:1000 dilution) and antiCx43 (red color; rabbit host; Zymed; 1:1000 dilution). The fixed cells were then washed 3 times with blocking buffer and incubated with two types of secondary antisera (working concentration was 7 µg/ml in blocking buffer); Alexa fluor 488 (Molecular Probes, Eugene, OR) which is goat anti-mouse and has an excitation/emission wavelength of 495/519 and emits green, and Alexa fluor 568 (Molecular Probes) which is goat anti-rabbit and has an excitation/emission wavelength of 578/ 603 and emits red. The cells were then washed with blocking buffer and mounted on slides with 1% glycerol in PBS. Slides

Figure 4. PKCγ and PKCα translocation to the cell membrane. TPA (200 nM) was added to untransfected N/N 1003A cells (1.0x106) for 60 min. The membrane and cytosolic fractions were separated by centrifugation at 35,000 rpm (100,000x g) for 1 h at 4 °C. There was an increase in both PKCγ and PKCα in the membrane fraction of TPA-treated cells while the control cells showed no translocation of PKCγ or PKCα to membrane fractions compared to the cytosolic fraction (20 µg of protein per lane). There was an approximately 61% increase in membrane-bound PKCγ and an approximately 68% increase in membrane-bound PKCα in the TPA-treated cells. Percent membrane bound was calculated by density scanning using Scion Image.

Figure 3. Phosphorylation of Cx43 by PKCγ. A: Cx43 was co-immunoprecipitated with anti-PKCγantisera from cells overexpressing PKCγ. Immunoprecipitated proteins were analyzed using 12.5% SDSPAGE and western blot, then the membranes which contained the transferred proteins were probed with anti-phosphotyrosine (lane 1), anti-phosphoserine (lane 2), and anti-phosphothreonine (lane 3) antibodies. B: Western blots were probed with anti-Cx43 antibodies to determine that equal amounts of immunoprecipitated Cx43 was in each sample. All antisera were used at 1:1000 dilution. 61



were examined using a laser scanning confocal microscope (Zeiss; Thornwood, NY). The cells were examined using a 63x/1.4 oil immersion objective with an excitation filter of KP 600 and the emission filters were BP 515-540 for the green signal and LP 590 for the red signal. The dichroic beam splitter was FT 560, the pinhole was 10, and the scanning time was 8 s. The sample size was between 7 and 11 for control, TPA-treated, and Calphostin C-treated cells. The number of plaques per micrograph were determined using Scion Image software (Scion Corporation; Frederick, MD). After plaque numbers per micrograph were generated for each sample set, the average number of plaques per square micrometer of area analyzed was determined and a Student’s t-test for paired or unpaired values with a sample of less than 30 was used to determine significant differences in plaque number between sample sets. Values of p