Phosphorylation of connexin43 on serine368 by protein kinase C regulates gap junctional communication

Abstract

Phorbol esters (e.g., TPA) activate protein kinase C (PKC), increase connexin43 (Cx43) phosphorylation, and decrease cell-cell communication via gap junctions in many cell types. We asked whether PKC directly phosphorylates and regulates Cx43. Rat epithelial T51B cells metabolically labeled with (32)P(i) yielded two-dimensional phosphotryptic maps of Cx43 with several phosphopeptides that increased in intensity upon TPA treatment. One of these peptides comigrated with the major phosphopeptide observed after PKC phosphorylation of immunoaffinity-purified Cx43. Purification of this comigrating peptide and subsequent sequencing indicated that the phosphorylated serine was residue 368. To pursue the functional importance of phosphorylation at this site, fibroblasts from Cx43(-/-) mice were transfected with either wild-type (Cx43wt) or mutant Cx43 (Cx43-S368A). Intercellular dye transfer studies revealed different responses to TPA and were followed by single channel analyses. TPA stimulation of T51B cells or Cx43wt-transfected fibroblasts caused a large increase in the relative frequency of approximately 50-pS channel events and a concomitant loss of approximately 100-pS channel events. This change to approximately 50-pS events was absent when cells transfected with Cx43-S368A were treated with TPA. These data strongly suggest that PKC directly phosphorylates Cx43 on S368 in vivo, which results in a change in single channel behavior that contributes to a decrease in intercellular communication.

Figure 1

Phosphorylation of Cx43. (A) Cx43 was immunoprecipitated from control (CON) and TPA-treated (TPA) T51B rat liver epithelial cells metabolically labeled with ³²P_i at 0.5 mCi/ml for 3 h. TPA (100 ng/ml) was added 30 min before the end of the labeling period where indicated. Immunoprecipitates were prepared using either nonimmune (NON) or Cx43 CT368 peptide antiserum (CON & TPA) and resolved by SDS-PAGE. (B) Phosphoamino acid analysis of ³²P-labeled Cx43 purified from T51B cells, hydrolyzed and resolved by two-dimensional electrophoresis. Arrows indicate the direction of the electrophoresis in the first (pH 1.9) and the second (pH 3.5) dimension. Positions of unlabeled ninhydrin-stained phosphoamino acid standards are outlined: phosphoserine (s-p), phosphothreonine (t-p) and phosphotyrosine (y-p). C, PKC phosphorylation of immunoaffinity-purified Cx43. Cx43 was purified and phosphorylated with PKC in vitro as indicated in Materials and Methods. The sample was resolved by SDS-PAGE, and the wet gel was subjected to autoradiography for 3.5 h.

Figure 2

Two-dimensional phosphotryptic analysis of Cx43 and Cx43-derived peptides. Cx43 metabolically labeled and immunoprecipitated from T51B cells either untreated (A) or treated (B) with TPA. (C) Immunopurified and PKC-phosphorylated Cx43. (D) Peptide 360-382 of Cx43 phosphorylated by PKC. E, HPLC 8-min fraction from PKC-phosphorylated and trypsinized peptide 360-382. (F) Mixture of samples shown in B and E.

Figure 3

HPLC of PKC-phosphorylated fragments of Cx43. (A) Peptide 360-382 was phosphorylated by PKC, purified by HPLC, with trypsin, and repurified by HPLC. The smaller peak at 39 min represents undigested 360-382. (B) Cx43 from ³²P_i, metabolically labeled, TPA-treated T51B cells was trypsinized and separated via reverse phase HPLC.

Figure 4

Identification of the phosphorylated residues in peptide 360-382. The number of cpm released during each cycle of Edman degradation is shown for the corresponding residue in the sequence. Cycle 9 which corresponds to S368 had the highest release of ³²P.

Figure 5

TPA caused a decrease in event amplitude of Cx43wt but not Cx43-S368A channels. Under control conditions (A and C), ∼100-pS events predominate in wild-type–transfected (A) and S368A-transfected (C) cells. However, in the presence of TPA (B&D), ∼50-pS events predominate in the wild-type transfected cells (B), whereas the S368A-transfected cells continue to display ∼100-pS channels (D). Note the rapid voltage-dependent closure of Cx45 channels in Cx43^−/− cells and the small channel amplitudes (E) compared with Cx43-transfected cells. Scale bars: vertical = 4pA, horizontal = 2 s. Transjunctional voltage was 40 mV in all records. Zero current in each trace is marked with an asterisk.

Figure 6

Event amplitude-relative frequency histograms of single channel data. The relative frequency of 87–112-pS channels was decreased in TPA-treated Cx43wt-transfected (A) but not Cx43-S368A–transfected (B) cells. Similarly, the relative frequency of the 50–68-pS channels was enhanced in TPA-treated Cx43wt- but not Cx43-S368A–transfected cells. The data were compiled from two independent clones of the wild-type (22C-3 and 25C-2) and the S368A (pI8 and pI9) transfectants. (C) T51B cells, which express Cx43wt, also displayed a decreased relative frequency of the 87–115-pS channels and increased relative frequency of 50–62-pS channels in response to TPA. (D) Cx43^−/− cells, which express Cx45, exhibited predominantly 25–37.5-pS channels. (Cx43wt controls: 7 cell pairs, 648 events; Cx43wt TPA: 6 pairs, 742 events; Cx43-S368A controls: 8 pairs, 813 events; Cx43-S368A TPA: 7 pairs, 641 events, T51B controls: 2 pairs, 247 events; T51B TPA: 4 pairs, 102 events; Cx43^−/− cells: 5 pairs, 219 events).

Figure 7

Two-dimensional phosphotryptic analysis of Cx43 derived from Cx43wt- and Cx43-S368A–transfected cells. Cx43 metabolically labeled and immunoprecipitated from Cx43^−/− cells transfected with Cx43wt (22C-3 clone) either untreated (A) or treated (B) with TPA. (C) Peptide 360-382 of Cx43 phosphorylated by PKC. (D) Mixture of samples shown in B and C. (E) Cx43 metabolically labeled and immunoprecipitated from TPA-treated Cx43-S368A (22C-3 clone)–transfected Cx43^−/− cells. F, Mixture of samples shown in C and E.