Proteins and mechanisms regulating gap-junction assembly, internalization, and degradation

Abstract

Gap junctions (GJs) are the only known cellular structures that allow a direct cell-to-cell transfer of signaling molecules by forming densely packed arrays or "plaques" of hydrophilic channels that bridge the apposing membranes of neighboring cells. The crucial role of GJ-mediated intercellular communication (GJIC) for all aspects of multicellular life, including coordination of development, tissue function, and cell homeostasis, has been well documented. Assembly and degradation of these membrane channels is a complex process that includes biosynthesis of the connexin (Cx) subunit proteins (innexins in invertebrates) on endoplasmic reticulum (ER) membranes, oligomerization of compatible subunits into hexameric hemichannels (connexons), delivery of the connexons to the plasma membrane (PM), head-on docking of compatible connexons in the extracellular space at distinct locations, arrangement of channels into dynamic spatially and temporally organized GJ channel plaques, as well as internalization of GJs into the cytoplasm followed by their degradation. Clearly, precise modulation of GJIC, biosynthesis, and degradation are crucial for accurate function, and much research currently addresses how these fundamental processes are regulated. Here, we review posttranslational protein modifications (e.g., phosphorylation and ubiquitination) and the binding of protein partners (e.g., the scaffolding protein ZO-1) known to regulate GJ biosynthesis, internalization, and degradation. We also look closely at the atomic resolution structure of a GJ channel, since the structure harbors vital cues relevant to GJ biosynthesis and turnover.

FIGURE 1.

Location and structure of GJs A: gap junctions (GJs) are assemblies of double-membrane-spanning hydrophilic channels termed “plaques” that bridge the apposing PMs of neighboring cells to provide direct cell-to-cell (or intercellular) communication (GJIC) as shown here for epithelial cells. B: GJ channels form by the head-on docking of two hemichannels or “connexons,” each assembled and trafficked to the PM by one of the two contacting cells. Connexons are assembled from six four-pass transmembrane proteins termed “connexins” (Cxs). C: GJ plaques can be seen by immunofluorescence light microscopy when stained with fluorescence-tagged antibodies, such as the ones shown here in T51B liver cells assembled from endogenously expressed Cx43 protein. D: GJ plaques, as the one assembled from exogenously expressed Cx43-GFP in HeLa cells shown here, appear as typical, pentalaminar structures in classically stained ultrathin sections when examined by electron microscopy, revealing the PMs that are closely aligned throughout the plaque, and the 2- to 3-nm-wide, name-giving “gap” separating the membranes. Regulatory proteins, if attached to the cytoplasmic plaque surfaces, are not well resolved. E: in contrast, another cell-cell junction type, adherens junctions (AdJs), are characterized by well visible plaque proteins and attached actin-based cytoskeletal filaments when stained by similar techniques as shown here also in HeLa cells.

FIGURE 2.

X-ray structure of a Cx26 GJ channel solved by Maeda et al. A: Cx26 monomer. Transmembrane (TM) α-helical domains are shown in blue, NH₂-terminal helix in red, extracellular loops (E1 and E2) in green and yellow, respectively, and unstructured, flexible intracellular loop (I-loop) and COOH-terminal tail as dashed lines. Note the kink in the NH₂-terminal tail that positions the NH₂ terminus inside the channel, forming the funnel-shaped channel wall visible in C. B: overall structure of the Cx26 channel shown is as a side view and a top view. Note the six Cx26 monomers (shown in different colors) labeled A–E, and TM1, TM2, and NH₂ terminus (NTH) facing the hydrophilic pore, whereas TM3 and TM4 face the hydrophobic membrane environment (labeled in the green monomer in the top view). C: cross section of the Cx26 GJ channel demonstrating the electrostatic surface potential. The NH₂-terminal helices of each Cx26 monomer make up the funnel-shaped wall of the channel entrance. Figure is adapted from Ref. and used here with permission.

FIGURE 3.

Kinases involved in the regulation of the Cx43 life cycle Kinases discussed in this review are labeled red. Forward trafficking of Cx43 monomers from the ER to the Golgi apparatus may be regulated through Akt phosphorylation (and interaction with 14-3-3 protein) (left). Trafficking of connexons toward the PM is regulated through phosphorylation by PKA, whereas assembly of PM-localized connexons into newly synthesized channels in GJ plaques is regulated via CK1 phosphorylation. Src, PKC, and MAPK activation can be achieved through the activation of the EGF receptor (EGFR) in an EGF- or TPA-dependent manner (right). MAPKs (i.e., ERKs, JNKs, or p38) are activated through Ras, Raf, and MEK (pathway 1). Src can be activated directly by the EGFR, which in turn can activate ERK5 through MEKK2 (pathway 2). Src can also activate PKC through diacylglycerol (DAG) that is generated by phospholipase C (PLC)-mediated phosphatidylinositol 4,5-bisphosphate (PIP₂) cleavage (pathway 3). PKC can also be activated directly by TPA. Activated kinases (Src, MAPKs, ERK5, and PKC) regulate downregulation of GJIC, channel closure, and possibly internalization, formation of annular GJs (AJGs), and degradation through either autophagosomal and/or endo-/lysosomal pathways (middle). CDC2 phosphorylates Cx43 at the onset of mitosis, also leading to GJ internalization and degradation.

FIGURE 4.

Lowest energy 3D solution NMR structure of the Cx43 COOH-terminal tail revealing the location of posttranslational modifications and protein-binding sites The structure of almost the entire COOH-terminal tail (residues S255–I382) as reported by Sorgen et al. (184) is depicted. Sites of Cx43 modification discussed in this review (highlighted in color) are shown in relative spatial context to each other. Sites of phosphorylation are shown as yellow side chains, with serine (S) and tyrosine (Y) residues numbered. Lysine (K) residues (possible ubiquitination sites) are labeled in red. The Nedd4 ubiquitin-E3-ligase binding site (²⁸³PPGY²⁸⁶) is labeled in blue. Tyrosine-based sorting signals (²⁶⁵YAYF²⁶⁸ and ²⁸⁶YKLV²⁸⁹) are highlighted in green. The minimal COOH-terminal ZO-1 recognition motif (³⁷⁹DLGI²⁸²) is shown in cyan. (The figure was generated from the Protein Data Bank accession file 1R5S deposited by Sorgen et al. (184), using VMD software, version 1.8.7, for Macintosh.)

FIGURE 5.

Ubiquitin and ubiquitination of Cx43 A: ubiquitin targeting cascade. E1 (ubiquitin activating enzyme) binds to ubiquitin (Ub) in an ATP-dependent process. E1 transfers ubiquitin to E2 (ubiquitin conjugating enzyme), which moves the ubiquitin moiety to the lysine (K) of the target protein. E3 (ubiquitin ligase) is bound to the target protein and aids in the transfer of ubiquitin from E2 to the target substrate and in its covalent attachment. B: lysine residues of the ubiquitin amino acid sequence involved in the formation of polyubiquitin chains. Ubiquitin has 76 amino acids, eight of which are involved in forming polyubiquitin chains. Of the seven lysines (K), K6, K11, K29, and K48 linkages lead to proteasomal degradation, whereas K63 linkages lead to trafficking, endocytosis, endo-/lysosomal and phago-/lysosomal degradation, transcription, and DNA repair. The functions of K27 and K33 linkages have yet to be elucidated. In addition to the seven lysines, methionine 1 (M1) also can link ubiquitin moieties together to form linear chains. The COOH-terminal glycine (G) residue is responsible for the covalent linkage of the ubiquitin moiety to the lysine of the target protein. C: potential ubiquitination and SUMOylation sites in the human Cx43 polypeptide sequence. Twenty-four lysine residues (K) are potentially available in the cytoplasmically located domains (NH₂ terminus, intracellular loop, and COOH-terminal tail) of Cx43 for ubiquitination or SUMOylation (highlighted in red in the intracellular loop and COOH-terminal tail). Based on the structural organization of GJ channels discussed above, with the Cx NH₂ termini being located inside the channel (see FIGURE 2), lysine-residues in the NH₂ terminus are not likely to be ubiquitinated in assembled connexons or GJ channels and to serve as signals for GJ internalization and/or degradation. The lysine present at position 243 is not conserved and is an arginine (R) in rat Cx43. Lysine 9 and 144 are known sites of SUMOylation (90); and lysine 9 and 303 have been identified as potential ubiquitination sites (205). C was adapted from Ref. and used with permission.

FIGURE 6.

ZO-1 and its interaction with Cx43 A: the known protein-binding domains of ZO-1 drawn to scale highlighting specific binding partners. The 220-kDa scaffolding protein ZO-1 consists of 1,748 amino acid residues. It includes three NH₂-terminal PDZ domains, followed by an SH3 domain, an inactive guanylate kinase (GK) domain, and two further COOH-terminal polyproline regions (shown as red boxes). Regions labeled “U1” through “U6” located between these domains denote regions that are unique to the ZO-1 protein sequence (8). Connexins (labeled red) known to bind to ZO-1 preferably bind to the PDZ2 domain, with the notable exceptions of Cx36, reported to only bind to the PDZ1 domain (117), and Cx45, reported to bind to all three PDZ domains (86) (also see Table 3). ZO-1's tight junction partners (shown in yellow) include claudins, which bind to PDZ1 (75, 217), JAMS, which attach to PDZ2 and PDZ3 (37), and occludins, which bind to the GK domain (43, 164). Adherens junction proteins (shown in orange), ARVCF, and α-catenin bind to the PDZ domains (85) and to the GK domain (131), respectively. The cytoskeletal component (light blue) Shroom2 binds to the SH3 and GK domains (39). Other cytoskeletal components, namely cortactin (84), protein 4.1R (125), and actin (44), bind to the polyproline regions. The SH3 domain is the interaction site for the transcription factor ZONAB (6) (light green), the heat shock protein Apg-2 (198) (dark green), the ZO-1-associated kinase-1 (ZAK) (5) (dark blue), the G-protein α-subunit Gα12 (127) (gray), and the Ras effector AF6/afadin(139) (brown). ZO-1 can also bind to other ZO proteins (ZO-2, ZO-3) via its PDZ2 domain (200) (white). (References 8 and 62 provide additional useful information on ZO-1 binding partners.) B: crystal structure of the ZO-1 PDZ2/Cx43 peptide-complex solved by Chen et al. (27). Each PDZ2 domain monomer is shown as a space-filling model (pink and gray), whereas the two nine-residue-long COOH-terminal Cx43 peptides (−³⁷⁴RPRPDDLEI³⁸²-COOH) (one seen in the front, whereas the other one can be partially seen in the back) are shown as stick models (green). The “domain-swapped” dimer of the two PDZ2 domains forms two specific binding clefts (or pockets) for the two Cx43 peptides, indicating that deletions within this region, as well as any additional amino acids/protein tags added to the COOH terminus of Cx43 will interfere with its ZO-1 binding. The figure was generated from the Protein Data Bank accession file 3CYY using VMD software, version 1.8.7 for Macintosh. B was adapted from Ref. and used with permission.