Gap junction proteins are encoded by gene families. The Drosophila innexin family consists of 8 genes (Stebbings et al., 2002), humans possess 21 connexin genes (Sohl and Willecke 2003) and 3 pannexins (Panchin et al. 2000). Within a given cell, multiple gap junction subunits are often expressed and it has been known for some time (at least for connexins) that individual channels can contain different subunits; so-called heteromeric channels, versus homomeric channels that are constructed from a single type of subunit. The subunit composition of connexin-based gap junction channels may be important for determining a range of properties including; permeability (Bevans et al. 1998, Elfgang et al. 1995), mechanism of transport to the plasma membrane (Laird et al. 1995, Martin et al. 2001), post-translational regulation (Cruciani and Mikalsen, 2002) and channel lifespan - properties that are quite important for gap junction function but which also make the study of these channels quite difficult. It appears that innexins can also form hetero-multimers and are just as complicated as the connexins. The evidence that different innexin subunits can interact has mainly focused on just two innexins, Inx2 + Inx3, but is quite substantial:
- - Inx2 and Inx3 have overlapping expression patterns (Stebbings et al., 2002) and overlapping subcellular distribution (Bohrmann and Zimmermann, 2008, Image: Innexin distribution in the follicle).
- - Inx2 and Inx3 proteins co-localise in salivary gland cells (as shown in the image at the top of this page).
- - In the Xenopus oocyte cell-coupling model system, gap junctions form readily between oocytes when Inx2 and Inx3 are both present in each of the paired cells (Stebbings et al. 2000). Inx3 alone never couples oocytes. Inx2 forms channels inefficiently and they are electrophysiologically distinguishable from Inx2/Inx3 heteromeric channels.
- - Transgenically co-expressing Inx2 and Inx3 together in Drosophila led to a greater frequency of lethality (Stebbings et al. 2000).
- - The accumulation of Inx3 into gap junction plaques is dependent, not only on the presence of Inx2 (Inx3 is mislocalised in an inx2null, as shown above in embryonic gut), but also on the function of Inx2 - in a dominant mutant, Inx2TA181 (described in Inx2 summary), a single amino acid substitution results in the synthesis of a mutant Inx2 protein and again Inx3 fails to accumulate into plaque structures.
- - Double antibody staining for Inx2 and Inx3 together indicates that they co-localise. This is much better than relying on the myc-tagged Ogre co-localisation experiment shown at the top of this page. Interestingly, only some plaques displayed similar levels of co-staining for Inx2 and Inx3, so they might not always exist as heteromeric hemichannels or the relative amounts of each subunit in each channel might be variable (Lehmann et al. 2006). This paper also reported that the C-termini of Inx2 and Inx3 probably mediate the interaction.