The regulation of Cx43-hemichannel function by intramolecular interactions via the SH3-binding domain and by autophagy

The ubiquitously expressed Cx43 isoform is one of the major building blocks of gap junction channels or hemichannels in the heart, brain and bone. Gap junction channels connect the cytoplasm of two neighboring cells, thereby exchanging signaling molecules and metabolites, whereas hemichannels mediate paracrine signaling by releasing ATP and other signaling molecules into the extracellular environment. The opening of Cx43 hemichannels must be tightly controlled to avoid loss of metabolic, energetic and ionic gradients, resulting in cell demise.

The regulation of Cx43-based gap junctions has been extensively studied, revealing that their activity is controlled by intramolecular interactions between the C-terminal tail (CT) of Cx43 and the second part of the intracellular loop (L2). A “ball-and-chain” model was proposed in which binding of the CT tail to the L2 region results in closing of the Cx43-based gap junctional channels. Until recently, the mechanisms controlling Cx43-hemichannel activity remained unresolved. However, our lab found that similar loop/tail interactions exist in Cx43 hemichannels as in gap junctions but with an opposite functional outcome, whereby loop/tail interactions are essential for Cx43-hemichannel opening. Loss of loop/tail interactions, either by genetically deleting the CT of Cx43 or by physiological conditions (like increases in [Ca2+]i of 1 µM and higher) that elicit actomyosin contractility, results in loss of Cx43-hemichannel activity. This can be overcome by the last 9 amino acids of the CT (CT9), which targets the L2 region, and more specifically the Gap19 region within this sequence, in Cx43.

First, we identified the residues present in the CT9 region responsible for L2 interaction and for mediating Cx43-hemichannel activity. We found that the interaction of the CT9 region with the L2 region critically depended on the presence of Asp378 and Asp379 residues and of Pro375 and Pro377 residues. Also, altering these residues in CT9 into Ala residues rendered TAT-CT9 (a cell-permeable version of CT9) ineffective in restoring the hemichannel activity of Cx43M239-based hemichannels (CT-deleted variant) and Cx43-hemichannel activity in conditions of actomyosin contractility. These findings correlate with the highly positively charged Gap19 region within the L2 domain, the presumed target of CT9.

Second, despite the fact that only the last 9 amino acids were sufficient to restore the activity of Cx43-hemichannels lacking the complete CT, deletion of this region in the CT of Cx43 only mildly impacted the interaction with Gap19/L2. We therefore hypothesized that another region within the CT contributes to such interaction. Here, we identified the SH3-binding domain as a second region within the CT, contributing to the loop/tail interaction. TAT-SH3 peptide was able to restore the activity of CT-truncated Cx43-hemichannels or of Cx43-hemichannels in conditions of increased actomyosin contractility. Furthermore, the ability of TAT-SH3 to restore CT-truncated Cx43-hemichannels depended on the presence of Gap19, since deletion of the latter rendered the channels resistant to functional modulation by TAT-SH3. Deletion of either the SH3-binding domain or the CT9 domain in Cx43 reduced hemichannel activity, while deletion of both domains completely abolished Cx43-hemichannel activity.

Third, we know that the CT of Cx43 is not only important for channel regulation, but is also important for Cx43 trafficking. Here, we studied the contribution of autophagy, a lysosomal degradation pathway, for Cx43-hemichannel turnover by exposing cells to nutrient starvation, a condition known to trigger autophagic flux. Our data show that nutrient starvation results in a rapid decline of Cx43-protein levels, both as gap junctions and hemichannels, and of Cx43-hemichannel function, which could be partially rescued by Bafilomycin A1, an autophagy inhibitor acting at the level of the lysosomes.

In conclusion, our study (i) identified key residues present in the CT9 region responsible for functional modulation of Cx43-hemichannels; (ii) elucidated the SH3-binding domain as a novel determinant underlying loop/tail interactions critical for Cx43-hemichannel function; and (iii) proposed autophagy as a turnover pathway of Cx43-hemichannels in particular in conditions of nutrient starvation.