Contribution
The 2016 paper (LIT-1567, unread) showed that M. exilis has no mitochondrion. This one asks what the rest of the cell did about it. The answer is almost nothing. The genome is not reduced, the cellular systems are those of an ordinary eukaryote, and the only changes linked to the loss are two that came before it and one that went with it. A complete loss of the organelle did not cascade through the cell.
Key insight
The organelle went once it had no job left, and its going cost the rest of the cell nothing visible. The authors sort changes into “preadaptations, which subsequently made mitochondria dispensable”, “functions lost concomitantly with mitochondria”, and “postadaptations that evolved to compensate”. They find two preadaptations, one concomitant loss and no postadaptations: “no clear postadaptations to the amitochondriate cell organization were revealed”. The first preadaptation, SUF in place of the mitochondrial ISC pathway, is shared by “all examined Preaxostyla”, including relatives that still have mitochondrion-related organelles. So that replacement did not by itself end the organelle. The loss of the glycine cleavage system and SHMT, which “removed another essential function from the MRO”, came later, “after the split from the P. pyriformis lineage”.
Assumptions
- Absence from a draft genome is absence from the cell. The genome is in 2,092 scaffolds. Genes were found by BLAST and HMMER against reference proteins, so a highly divergent gene could be missed.
- No organelle was seen, so none is there. FIB-SEM of two cells found no mitochondria or MROs, but “the resolution of the microscopy could still allow for undetected highly reduced mitosomes”.
- The ancestor had a mitochondrion. This follows from MROs in its relatives (Figure 1c, from Leger et al. 2017), and is not re-tested.
- Only M. exilis-specific changes count as linked to the loss. Even these “may reflect adaptation to anaerobiosis or endobiotic lifestyle”.
- Complexity is measured by inventory. Gene counts, orthogroup overlap with Naegleria, intron density and pathway completeness stand in for cellular complexity. Function is inferred from homology, with no experiment, as the authors say: confirming structures “must await development of a transfection system”.
Key results
- The organism. A bacterivorous tetraflagellate, a “putative commensal” in the gut of caviomorph rodents. No conspicuous Golgi and no mitochondria by TEM and FIB-SEM. The ER sometimes forms stacks that look like Golgi but carry ribosomes.
- Genome (Table 1). 74.7 Mb assembly, 36.8% GC, 16,768 protein-coding genes, 31,693 introns, 153 tRNA genes, about 50 rDNA units. Probably monoploid. TTAGGG telomeres; about 13 telomeric signals per nucleus, “suggesting the presence of 6 or 7 chromosomes”. 39% of genes have a putative function.
- Less reduced than its relatives. Of 2,031 orthogroups present in M. exilis, 83% are shared with Naegleria, 77% with Trichomonas and 52% with each diplomonad. “M. exilis has lost fewer ancestral eukaryotic proteins than other metamonads.” Parasitic metamonads have 4 to 65 introns; M. exilis has tens of thousands, more than 30% of the analysed ones inherited from the last eukaryotic common ancestor, “a proportion similar to other eukaryotes” (Figure 3).
- Canonical systems. All four core histones and H1; all of replication protein A, which most metamonads lack or reduce; base and nucleotide excision repair “most complete” among metamonads; homologous recombination; all general transcription factors including TFIIA, with TATA-like motifs in 52% of promoters; diphthamide synthesis.
- Cytoskeleton. “Relatively canonical”, except no myosin and no dynactin, both shared with other metamonads. The kinetochore is reduced (15 of 70 orthologs) like other metamonads’.
- Endomembranes (Figure 4). “A relatively canonical complement”, with expansions: nine Rab7 and a new clade of nine divergent Rab7-like paralogs with signs of birth-and-death evolution, eight VAMP7, seven Rab32. Candidate multivesicular bodies and omegasome-like membranes in FIB-SEM images.
- Metabolism. ATP from glycolysis, the arginine deiminase pathway and other amino-acid catabolism. No oxidative phosphorylation, no TCA cycle, no oxidative pentose phosphate phase. No de novo purine or pyrimidine synthesis, so nucleotides are salvaged from the gut. A full set of essential Fe–S proteins (70 candidates), supplied by the cytosolic SUF and CIA systems. 103 genes likely acquired by lateral transfer from prokaryotes, 17 of 49 in amino-acid metabolism.
- What went with the mitochondrion.
- Peroxisomes are absent (only Pex19 found), but peroxisome loss is common in anaerobes, so it “cannot be attributed to the loss of mitochondria”.
- Mitophagy and pexophagy proteins are absent: “an example of a system likely lost together with mitochondria”.
- Among mitochondrial-dynamics proteins, M. exilis lacks a second dynamin-related protein that its MRO-bearing relatives have. “It is tempting to speculate” that it divided their MROs.
- Proteins with mitochondrial roles elsewhere are kept, presumably for other jobs: two dynamins, MSTO1, Vps13 (four paralogs), Vps39, and calcium transporters in normal numbers.
- No compensating genes. Of the genes unique to M. exilis, 1,126 have a function and none is “clearly related to the loss of mitochondrial function”.
- The conclusion. M. exilis is “living evidence that complex amitochondriate eukaryotic cell is a viable cell type”, and such cells “could have evolved during the evolutionary process of eukaryogenesis before the acquisition of the mitochondrion” (citing Hampl et al. 2019).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | M. exilis’s genome is as complex as other eukaryotes’ and less reduced than the parasitic metamonads’ | moderate to strong | genome statistics, orthogroup overlap, intron analysis over 100 genes and 34 species; a draft assembly |
| C2 | Its cytoskeleton, endomembranes, nuclear machinery and biosynthesis are canonical | moderate | homology searches, mostly in supplementary tables not read here; no functional test |
| C3 | Replacement of ISC by SUF and loss of GCS and SHMT were preadaptations that made the mitochondrion dispensable | moderate | SUF shared by all examined Preaxostyla; GCS and SHMT in the MRO of P. pyriformis; the order is inferred from the tree |
| C4 | Mitophagy was lost with the mitochondrion; nothing else was | moderate | absence of mitophagy orthologs; the second DRP is speculation; the authors note mitophagy machinery has not been sampled deeply |
| C5 | No postadaptation compensates for the absence of the mitochondrion | moderate | a search of unique genes with functions; an absence, limited by annotation |
| C6 | A complex eukaryotic cell can exist without a mitochondrion, so such cells could have existed before it | strong for the first part, weak for the second | the first follows from the organism; the second is an inference from a secondary loss to an ancestral state |
Concepts
- Amitochondriate: lacking any mitochondrion or MRO, as opposed to having a reduced one (hydrogenosome, mitosome).
- MRO: mitochondrion-related organelle, a reduced mitochondrion of anaerobes.
- Preadaptation / concomitant loss / postadaptation: the authors’ three classes of change around an organelle’s loss: before it and making it possible, at the same time, and after it compensating.
- SUF, ISC, CIA: Fe–S cluster assembly systems. ISC is mitochondrial and otherwise “considered to be an essential house-keeping pathway”; SUF is the bacterial-type system that replaced it here; CIA is the cytosolic system.
- Preaxostyla: oxymonads plus the free-living trimastigids, such as Paratrimastix pyriformis, which have MROs.
Connections
It extends the 2016 report (LIT-1567, unread), which found the SUF system and no hallmark mitochondrial proteins. Its eukaryogenesis remark bears on the “complexity before the mitochondrion” scenario that QUESTION-277 lists with LIT-1595, unread. It does not cite the endosymbiosis literature on gene loss and dependence, and says nothing about individuality.
Bearing on the record
-
CASE-168. The case says the mitochondrion was lost “once a bacterial Fe–S system acquired by lateral transfer replaced the organelle’s one indispensable job”. The reading corrects that in two ways.
- SUF replaced ISC in “all examined Preaxostyla”, including relatives that kept MROs. So SUF was not enough by itself.
- A second essential function, the glycine cleavage system with SHMT, was still in the MRO of the closest relative and was lost later in the M. exilis line. The paper calls both preadaptations.
The “lateral transfer” origin of SUF is stated in the 2016 paper, not argued in this one. The case’s “first eukaryote known to lack any form of mitochondrion” is supported, with the paper’s caveat that “highly reduced mitosomes” below FIB-SEM resolution are not excluded, and its note that relatives may share the status.
-
QUESTION-277. The question uses this system as a negative control: “losing genes is not the same as losing function, losing autonomy or gaining integration”. The reading supports a more exact version. The organelle’s functions were moved one at a time, by a gene from bacteria (SUF) and by losing the need (GCS), until it had none. Then the organelle went, and the rest of the cell was unaffected. In the question’s terms, G was redistributed away from the partner (to the host nucleus and a third party) before D on the partner fell to zero. That is the host-compensation route, with genes from a third lineage, run to dissolution. The mapping is mine; the paper does not use these terms.
-
Dependence moved, not vanished. M. exilis is a gut commensal that salvages its nucleotides from its environment, and many of its changes are put down to “the microaerophilic and endobiotic niche”. It lost an internal partner and depends on a host gut. The inference is mine and the paper does not draw it.
-
THEORY-1100. Sørensen et al. describe organelles made by moving functions onto host-encoded proteins. This is the same move run past the end: once no function needs the compartment, it can go.
-
Filed as THEORY-1098. No instruction for machine-learning practice.
Limitations
- One draft genome and two imaged cells. The paper says so of the microscopy.
- Inventory, not function. Every system is inferred from homology. Most gene lists are in supplementary tables that were not read here.
- Absence of evidence. “No postadaptation” and “nothing else lost” are negative results bounded by annotation; 60% of gene models are hypothetical proteins.
- Comparators. The comparison set is four metamonads and Naegleria. The free-living Kipferlia genome came too late to include, and the authors say it may move where intron gains and losses are placed.
- Timing is inferred. That SUF and GCS loss came before the organelle’s loss rests on their distribution among relatives, not on dating.
- The eukaryogenesis remark is a leap. A secondary loss in a lineage that already had a complex cell shows such a cell can persist without mitochondria. It does not show one arose without them.
Open questions
- Do other oxymonads lack mitochondria too? The paper says it “may hold true” from limited cytology.
- What did the last MRO in the M. exilis line still do, after GCS and SHMT were gone? The paper names no remaining function.
- Is there a reduced mitosome below FIB-SEM resolution? Proteomics or a tagged marker would settle it, and both await a transfection system.
- Did the gut niche, with nucleotides and amino acids to hand, make the organelle cheaper to lose? The paper sets niche adaptation and mitochondrial loss side by side and does not join them.