Studies of connexins involving live imaging of fluorescently tagged gap junction channels, electron microscopy and immunocytochemistry (Jordan et al. 2001) have shown that a considerable quantity of gap junction plaque material can be removed from the plasma membrane for degradation via the lysosomal and proteasomal pathways. The observations reveal that this process involves the internalisation of a given plaque into a single cell, forming a double-membrane bound vesicle containing gap junction holochannels (the vesicles are referred to as annular junctions or gap junction profiles depending on what you read). It is suggested that this process, in addition to gap junction synthesis (Werner, 2000) and post-translational modification of connexins (Laird et al. 1995), may have an important role in regulating the extent of gap junction mediated intercellular communication that occurs between cells. This may be so, but it may not be generally applicable as a means of regulation because annular junctions are abundantly detected only in some tissues and appear to be absent from others.
The confocal image at the top of the page and close-up annular vesicle images were obtained from fixed tissue (not dynamic, real time observation) so it was not possible to elucidate whether the large vesicles are annular junctions that were removed from the plasma membrane or transport vesicles carrying innexins to the plasma membrane for incorporation into plaques. Based on their size and location they represent a distinct population of innexins from the small transport vesicles that are observed near the cell nucleus (see webpage Inx within the cell cytoplasm). They are also the most intensely staining innexin-immunopositive structures in cells suggesting that they are double-membrane vesicles containing holochannels rather than hemichannel-containing transport vesicles. Only live imaging of cells using green fluorescent protein (GFP)-tagged innexins (for example using the inx2-GFP construct referred to by Lehmann et al. 2006 or a GFP protein-trap of the endogenous inx2 gene (...if any innexin protein-trap lines become available from Flytrap) will resolve their real origin and destination. In any event, most of the innexin-positive structures of cells appear to be associated to some extent with the actin cytoskeleton suggesting a close relationship between the cytoskeleton and innexin processing and degradation. For example, small, nuclear-associated transport vesicles are closely apposed to actin fibers and annular junction vesicles often exist in linear 'stripe' patterns suggesting an association with actin cables, and indeed, F-actin is often found next to annular vesicles. It has been reported that annular junctions are clathrin coated vesicles (Piehl et al. 2007) but we found no evidence of the GFP-tagged Drosophila clathrin light chain protein contruct (UAS-EGFP-Clc, Chang et al. 2002) staining these vesicles at higher levels than background.
Many questions remain about these structures:
- - Why do they only accumulate in certain cell types?..does it reflect the function of the cell or it's structure?... or one about to undergo apoptosis? And if so, do they have a functional role? For example storing molecules for the apoptosis phase of the salivary gland lifecycle?
- - if they are part of the degradative pathway, are all innexin-based gap junction channels disposed of this way? This is important in Drosophila particularly because most cells are so small that it is not possible to determine if punctuate patterns of innexin antibody staining reveal innexin within the cell membrane or innexin within the cell cytoplasm.
- - With so much recent emphasis on hemichannels having a biological role and comprising a significant proportion of the plasma membrane.....are hemichannels also degraded using this process?
- - Once a gap junction holochannel forms can it ever separate into hemichannels again, allowing the plasma membranes to separate?...or do they always have to be internalised and degraded for cells to dis-attach and wander elsewhere?
- -are ligand/receptor signalling molecules ever endocytosed along with these vesicles? Could annular vesicles also be a regulatory step for a ligand/receptor pathway (...in a similar fashion to the regulation of some paracrine pathways by endocytosis - Di Fiore, 2001)?...and if so, is this a process whereby aberrant innexins can disrupt ligand/receptor signalling (see section on innexins and development)?