Skype for Cells, and Valves for Mitochondria

Connexin is one of those proteins that get more roles the harder people look.

One reason genetics is hard is that genetic products (proteins, generally) can have complicated roles in the organism- i.e. phenotype. For every simple case like sickle cell anemia, there are ten complicated cases where mutations are covered by duplicative functions, or have effects only revealed under unusual circumstances, or have multiple effects that are hard to connect and understand. Today's post focuses on a protein called connexin, which joins up into hexameric (6-member) complexes to form membrane half-channels, or hemichannels. When two such hemichannels on neighboring cells meet up, then join to form a gap junction, which is a small pore, permeable to molecules up to about 1000 Daltons. So ions, water, hormones, and other small molecules get through, but not most proteins or nucleic acids. They also help align the electrical charge or electrical propagation of neighboring cells. These junctions are surprisingly common in our bodies, functioning a bit like Skype, keeping neighboring cells in close touch. They have big roles in development, adhesion, neuron activity, and even cancer. Cancer cells tend to turn their gap junctions off in order to go their own way.

Structures of some connexin complexes, as hemichannels (below), and as full gap junctions (above).  "In" denotes inside each cell, "Ex" denotes the extra-cellular space, and the lipid bilayer would be the plasma membrane in this normal case.

But recent work has shown that hemichannels by themselves have regulated conductance properties and a variety of roles, some of which do not even rely on their channel properties. One recent paper (though with precedent work going back to 2006) raised the prospect of connexin hemichannels functioning in mitochondria, on the inner membrane, as regulated potassium channels. The mitochondrial inner membrane is notoriously impermeable, in order to accumlate the proton-motive force (pmf), which is the product of respiration / catabolism of food, and used for the synthesis of ATP. Protons are pumped out of the innermost mitochondrial matrix and into the inter-membrane space, setting up a pH and charge gradient, usable as energy. The ultimate concentration of hydrogen is not terribly high, however, and the overall ionic (and osmotic) balance remains governed by more conventional gradients of ions like sodium, potassium, and chloride. Thus it is important to regulate those balances using channels that are specific for those ions and don't conduct (i.e. leak) protons. 

Aside from the connexin channel discussed here, there are numerous other potassium conducting channels in the mitochondrial inner membrane, with individual regulation.

For potassium, there are at least seven channels now known, each regulated differently. That is amazing, really, for a membrane that should be so impermeable, though most are present at low levels. There is a K+/H+ antiporter that extrudes K+ to maintain osmotic balance, and then the other channels are all leakage channels that allow K+ back in, under various regulated conditions. The logic seems to be that the electron transport chains of respiration can easily run too "hot", giving off poisonous reactive oxygen species like peroxide, instead of the coordinated O2 reduction to water and proton export, as intended. So the major transport chain stations appear to have associated potassium channels that are inducible by reactive oxygen species, among other things, in order to fine-tune the local membrane potential and respiration rate. It seems like a curious way to run things, reducing the efficiency of a system that would be better inhibited earlier in the respiration process. 

Another model is that the fine-tuning by these channels forestalls the more catastrophic activation of PTP channels (mitochondrial permeability transition pore). These are more like gap junctions, totally non-discriminating in their channel characteristics, and are induced by high levels of stress from reactive oxygen species. When induced, these can totally leak away the protonmotive force, and if sustained, can kill the mitochondrion and even the whole cell. This would lead to what the current researchers found, which was that genetic reduction of the levels of the connexin Cx43 caused all kinds of bad outcomes in cells treated with peroxide, such as lower proton motive force, lower electron transport chain coupling, and lower ATP production.

ATP production is raised by leaking K+ into the mitochondrial matrix, and lowered conversely by reduction of connexin levels. "CxKD" denotes a genetic knock-down, or reduced expression, of the Cx43 connexin protein. They note, however, that expression of one ATP synthase component is reduced as well in this setting, so there may be yet other (gene regulation) effects going on here.

 

At any rate, the finding that one connexin that participates in cell surface gap junctions, called Cx43, is also specifically transported to the inner membrane of mitochondria during reactive oxygen species stress, (a stress that has very wide-ranging occurrence and effects, at the cellular and organismal levels, not just in mitochondria), and then acts there as a regulated K+ leakage channel, is quite unexpected. While the gap junction is a promiscious, wide channel, this activity must be far more discriminating. Add to that that it associates specifically with the H+ consuming ATP synthase, at the matrix side of the inner mitochondrial membrane, and we have a protein with a double life.

• We're not a cult!
• Why not put parking lots to some use?