Mitochondrial disease, a bone of contention in the church apparently, is back in the news as UK legislation makes good progress towards the first treatment. The mitochondrial genome is a second outpost of the sum human genome, but isnāt unique as far as organelles are concerned. Some protists have a nucleomorph, while plants store genomic information in their plastids such as chloroplasts (though there are unexplained exceptions, described by Smith & Lee 2014).
Amazingly, the human mitochondrial genome is thought to be encoded in a single polycistronic sequence (something more heard of among viruses, bacteria and archaea, a nod to the endosymbiotic event that spawned the organelle) that is cleaved strategically in the liberation of 22 tRNAs (ātRNA punctuationā) to give the 13 mRNAs and 2 rRNAs (12S and 16S), a model first proposed by Ojala and Attardi in 1981:
Although all are transcribed on a single polycistronic RNA, the rRNAs end up far more abundant, suggesting an entirely unexplored mode of post-transcriptional regulation at work on the mitochondrial genome.
Michael Regnier wrote for the Wellcome Trust 3 years ago that ā1 + 1 + 0.00001 ā 3ā, likening the idea of mitochondrial replacement therapy meaning a child would have three parents to declaring heart transplant recipients have four. The distinction is real though (in my opinion), insofar as transplants arenāt at the embyronic level, so donāt diffuse into every cell of the adultā¦ As Jeremy Farrar explained in a radio interview this week, the main reason we should be out in support of the technology is simply the extensive pre-clinical testing thatās gone into plans for mitochondrial medicine.
In oxidative phosphorylation (OXPHOS), ATP synthesis is coupled to electron flow through NADH or FADH2 cofactors en route to molecular oxygen. The electron carriers in the mitochondrial inner membrane respiratory assembly include quinones, flavins, FeS complexes, cytochrome protein haem groups, and copper ions. NADHās electrons hop onto the flavin mononucleotide (FMN) prosthetic group of NADH-Q oxidorectase (āComplex Iā), the first of a chain of four complexes that reduces ubiquinone (Q) as ubiquinol (QH2).
Similarly, complex II (succinate dehydrogenase, a citric acid cycle enzyme) pops electrons onto Q from FADH2 making yet more QH2, precipitating the hydrophobic electron carrier to spill them at Complex III, Q-cytochrome c oxidoreductase (which despite its name harbours cytochromes b and c1).
Cytochrome c is a more water-soluble mobile eā carrier, and a peripheral membrane protein (meaning it adheres only temporarily at the mitochondrial inner membrane). Departing from the electron transport story, cytochrome c unlocks the cellās death programme, apoptosis, in clustering with APAF1 and procaspase-9 to form an apoptosome.
The mitochondrial genome is frequently said to encode only parts of the mitochondrial bioenergetic machinery - components of the ETC. In humans, these genes reside on a single, circular, double stranded DNA molecule, but can also be linear, while cucumbers have 3 circular and individually autonomous mitochondrial chromosomes (below, an excerpt from Jennifer Machās 2011 editorial accompanying the finding).
They found that, unlike most plant mitochondrial genomes sequenced so far, which map to a single, circular chromosome, the cucumber mitochondrial genome contains three chromosomes (1556, 84, and 45 kb; see figure). These physically map as circles, although the exact structure may be more complex, as indicated by previous electron microscopy images of other plants showing branched, linear, and other structures. All of the intact, identifiable mitochondrial genes were found on the large chromosome, but all three appear to have transcribed sequences. The large chromosome is approximately twice as abundant as the smaller chromosomes (2.2 and 1.5 times as abundant), suggesting that they replicate independently. Sequence and DNA gel blot experiments showed that the two small chromosomes exist as both independent and cointegrated forms. Deep sequencing also allowed the authors to examine repeat-mediated recombination. The authors examined paired sequences from the ends of the same subclone, where the pairs do not map to the same location. Computational reconciliation of the ends showed many instances of recombination events meditated by many repeats, indicating highly active repeat-mediated recombination throughout the genome. Thus, this work explains how the cucumber mitochondrial genome got so large and provides new tools for examining the complex dynamics of these cool genomes.
mtDNA isnāt solely for ETC proteins though: the ~17,000 base pair sequence encodes 13 ācanonicalā polypeptides involved in bioenergetics, but some smaller peptides lurk amongst them, and depart not only from electron transport but from the mitochondrion altogether.
Mitochondrial-derived peptides, or MDPs were discovered in 2001 - the first being humanin (rats have a homologue called rattin). Humanin takes on a host of antistress functions in humans, as brought together in a review from Lee et al. in 2013:
Humanin was first isolated from the surviving fraction of an Alzheimerās disease patientās brain (Hashimoto 2001, Guo 2003). Itās not yet clear whether mitochondrial or cytoplasmic translation is more often used, but since mitochondrial translation uses a slightly different genetic code, the mitochondrially translated 24 amino acid peptide produced is distinct from its cytoplasmic counterpart (itās biologically effective in both forms, so the jury is out).
Mitochondrial signalling is far more elaborate than the common trope of āpowerhouseā or ābatteryā gives credit. A host of proteins: Bak, Bcl-2, BH3-only (Puma, Nox, Bim, Bid) exert a layer of pro- and anti-apoptotic regulation integrated through a small protein Bax which if unsequestered pokes pores in the mitochondrial membrane, releasing the death-inducing cyt c into the cytoplasm.
Humanin is secreted in blood plasma and bound to cell membranes (where it fiddles around in receptor signalling), but inside tissues binds to Bax, preventing it causing harm to the host cell through bursting mitochondrial membranes.
Oddly, although this question came to me upon reading the news, a paper published only 2 days ago from a trio of French, Italian, and Canadian researchers covers the same topic, registering the potential for further unknown peptides to be lying in the mitochondrial genome. After reviewing the role of humanin, they describe the gau gene:
In addition to its ubiquitous presence, the reason why the gene has been named gau for Gene Antisense Ubiquitous, strong arguments indicate that the gau ORF indeed encodes a functional protein: (i) it is evolving under purifying selection, (ii) the deduced GAU proteins share some conserved amino acid signatures and structure among different taxa, suggesting a possible conserved function, (iii) gau has been identified in sense-oriented ESTs with poly(A) tails, (iv) immunohistochemical experiments using an anti-GAU monoclonal antibody showed a mitochondrial-specific signal in human cells, and (v) BLAST analyses suggested that no part of any known human proteins exhibits a high level of amino acid identity with the peptide antigen that was used for immunization and antibody production (Faure 2011). As for humanin, potentially functional and similar but not identical gau regions have been found in the nuclear genome. However, none of the deduced protein sequence possesses a mitochondrial signal peptide (Faure 2011), a result that is not in line with the intramitochondrial localization of GAU observed using the anti-GAU antibody. According to the authors, the most parsimonious hypothesis is that gau is a mtDNA-encoded protein gene, providing evidence for antisense overlapping functional open reading frames in mitochondrial genomes.
Other than this well-studied pair, bioinformatics has brought sequence-level suggestions of extra-canonical proteins, āsmall nuclear open reading frames (sORFs) encoding biologically active peptides of 11ā32 amino acids in lengthā (Andrew & Rothnagel, 2014; Kondo et al., 2010).
Lastly, thereās suggestion that further mitochondrial proteins may be encoded through tetracodons (Seligman, 2012a,b), backed up by observation that this class of tRNA (which feature 4 interacting nucleosides at the anticodon loops rather than the regular 3 base code) is seen to evolve along with the number of sequences that would potentially form tetracoded mitochondrial open reading frames (ORFs).
