Cardiolipin and lipid environment of respiratory complexes

Brzezinski and colleagues review a proposal that the two phosphate groups of cardiolipin have different pKa values, one above 8.0. On that view the headgroup carries one negative charge at neutral pH and could act as a proton trap near enzymes that use proton gradients⁵. They add that more recent studies find the two pKas similar, with cardiolipin carrying two negative charges at neutral pH⁵. Preiss and colleagues note a suggestion that the high cardiolipin level of an alkaliphilic Bacillus might delay proton equilibration. They also report that mutants lacking detectable cardiolipin were not impaired in the relevant growth¹. Morelli and colleagues note that cardiolipin was tentatively identified as the proton donor at the centre of complex I in earlier work they cite³.

Letts and colleagues found that complexes dissociated over time in the absence of amphipol. They considered de-lipidation a possible cause, since lipids are thought to stabilise supercomplexes². The isolated particles held on average 122 ± 8 lipid molecules each². Assuming a ratio of about 4 phospholipids to 1 cardiolipin, they estimated about 24 cardiolipin molecules per particle². In the bovine dimeric ATP synthase, Spikes and colleagues report three cardiolipins and two phosphatidyl-glycerols bound in each wedge. They say either of the latter two could be cardiolipin⁴. Brzezinski and colleagues report that supercomplex fractions are larger when cardiolipin is present⁶.

In yeast III-IV supercomplex structures, a cardiolipin and probably a phosphocholine sit at the interface between cytochrome bc1 and cytochrome c oxidase. The lipid is suggested to act as a glue by binding sites on both complexes⁶. Rao and colleagues describe cardiolipin as a conical lipid whose four acyl chains account for the tight curvature of cristae. They link that curvature to surface area for respiratory proteins, and they say cardiolipin anchors complexes in place⁸. Farahani reports recent findings that cardiolipin and phosphatidylethanolamine segregate to the negatively curved leaflet facing the crista lumen⁷.

Brzezinski and colleagues leave cardiolipin's role in enzyme activity and apoptosis outside their review⁶. Farahani repeats the suggestion that cardiolipin headgroups may trap protons near ATP synthase. Farahani also proposes that anionic cardiolipin headgroups could form anti-electrostatic hydrogen bonds with other anionic groups, such as RNA phosphate. The author states that no direct evidence yet shows RNA associated with the inner membrane bilayer⁷. Two sources report limits on the trap idea. Preiss and colleagues report B. pseudofirmus OF4 mutants lacking cardiolipin that were not impaired in non-fermentative growth at high pH¹. Brzezinski and colleagues cite newer pKa studies that find the two pKas similar⁵.