Mitochondria

Mitochondria are double-membraned organelles in nearly all eukaryotic cells. They generate most cellular ATP through oxidative phosphorylation, and they run dozens of other processes: heat production, oxygen sensing, steroid hormone synthesis, iron–sulfur cluster assembly, calcium buffering, cell-death control, and signaling to other organelles and other cells. They descend from a free-living alphaproteobacterium taken up by a host cell more than a billion years ago.

Endosymbiotic origin

Comparative genomics traces the mitochondrial genome to the Alphaproteobacteria, the closest extant free-living relatives of mtDNA (Gray & Doolittle 1982).¹ The most bacteria-like mitochondrial genomes known are those of jakobid protists: 70+ proteins, bacteria-style multisubunit RNA polymerase, tmRNA, and operon organization — a near-complete bacterial genome still sitting inside a eukaryotic cell.²

Mitochondria originated once. Shared features, shared genome deletions, and phylogenetic reconstruction all put every eukaryote’s mitochondria downstream of a single last mitochondrial common ancestor (LMCA). The transition to organelle involved massive gene loss and endosymbiotic gene transfer (EGT) — genes moving to the nucleus, their protein products imported back into the organelle.³

The proteome complicates the genome story. Only ~10–20% of mitochondrial proteins trace to α-proteobacteria (the “α-proteobacterial component”); the other 80–90% has diverse prokaryotic or eukaryote-specific ancestry. Gray’s pre-endosymbiont hypothesis proposes that the host cell already contained a membrane-bound metabolic organelle (a premitochondrion) with a protein import system and transporters, which merged with the arriving symbiont — a synthesis of endogenous (“from within”) and xenogenous (“from outside”) origins.⁴

Which alphaproteobacterial lineage is the closest living relative of the mitochondrial genome remains unresolved. Rickettsiales keep being proposed and keep failing to win consensus; deep-branching phylogenies here are methodologically fraught (long-branch attraction, compositional bias). The jury is still out.⁵

Scale, genome, and mtDNA as signal

A single cell holds hundreds to thousands of mitochondria; metabolically active cells (heart, liver) hold more, and the population constantly turns over through recycling, biogenesis, fusion, and fission. Picard’s rough estimate for a human body is ~2 quadrillion mitochondria.⁶ Mitochondrial DNA copy number (mtDNAcn) varies roughly 100–5,000 per cell across tissues.⁷

Human mtDNA carries 37 genes, against more than 78,000 genes estimated for the nuclear genome — the small surviving remnant of the endosymbiont’s genome.⁸ mtDNA does not always stay inside: stressed or damaged mitochondria release it into the cytoplasm and beyond, where it activates immune and inflammatory pathways. Circulating mtDNA has been linked to dozens of diseases (type 2 diabetes, Parkinson’s, cardiovascular and autoimmune disease), and mtDNA fragments can insert themselves into the nuclear genome.⁹

Beyond the “powerhouse”

The textbook “powerhouse of the cell” analogy draws pushback in the mitochondrial literature. Monzel, Enríquez & Picard argue that “mitochondrial function” and “mitochondrial dysfunction” are misleading misnomers: mitochondria are multifunctional and dynamically recalibrate, so a single function/dysfunction axis cannot carry the biology. They propose a five-level hierarchy instead — cell-dependent properties (content, mtDNAcn, network topology), molecular features (protein abundance, lipids, cristae architecture), single-enzyme activities, organellar functions (ATP synthesis, protein import, steroidogenesis), and goal-driven behaviours (motility, fusion/fission, biogenesis, signaling).¹⁰

One function is non-negotiable: iron–sulfur (Fe/S) cluster synthesis. Cells can survive without oxidative phosphorylation; they cannot survive without energized mitochondria, because Fe/S cluster assembly happens there and dozens of enzymes depend on it.¹¹

Monzel et al. close with a division-of-labour framing: specialized mitochondrial phenotypes coexist across organs and tissues the way cell types and organs do — recognizable hallmarks, complementary roles, one organism. Building a systematic cartography of those phenotypes is an open task for the field.¹²

Tissue specialization

Mitochondria in different tissues are molecularly, functionally, and morphologically distinct — “mitotypes.” Immune cell subtypes occupy distinct phenotypic space (natural killer cells and neutrophils show similarly low citrate-synthase activity and mtDNAcn; CD4+ and CD8+ T cells run 2–3× more mtDNA copies). Oxidative (soleus) and glycolytic (gastrocnemius) muscle fibers differ in ADP affinity, peroxide emission (~50% lower in oxidative), and calcium retention (~60% lower). Even within one myofibre, subsarcolemmal and intermyofibrillar mitochondria differ ~2× in oxidative capacity and ~60% in cardiolipin content. In brown fat, peridroplet mitochondria run double the respiration of their cytoplasmic siblings, with more assembled supercomplexes but lower fatty-acid oxidation.¹³

Morphology follows: axonal mitochondria travel far and stay small and short; dendritic ones grow long and branched and buffer local calcium. Trained investigators can identify tissue-specific mitochondrial phenotypes from electron micrographs alone.¹⁴

Dynamics, quality control, and cell-to-cell movement

Mitochondria are not static furniture. Molecular motors haul them along microtubules (long trips in neurons), and calcium signals can order them to stop. Quality control runs in parallel at several scales: mtDNA repair, protein refolding and degradation, cristae-architecture regulation, membrane permeability control, biogenesis, fusion, fission (which discards damaged segments), and full recycling of an impaired mitochondrion (mitophagy).¹⁵

They also travel between cells. Platelets release whole mitochondria into the circulation; tunneling nanotubes connect cells directly; cells with badly damaged mitochondria have been rescued by taking up healthy ones from neighbors. The same transfer can work against the body — cancer cells steal mitochondria from immune cells.¹⁶

Mitochondrial medicine

In September 2025, elamipretide (Forzinity) became the first FDA-approved treatment for a mitochondrial disease (Barth syndrome); it acts on the inner mitochondrial membrane to improve energy-transformation efficiency. Gene therapies, metabolic therapies, and mitochondrial transplantation from healthy cells are in investigation. Some of these target mitochondrial diseases specifically; others aim at conditions where energy flow is disrupted generally. The field is early.¹⁷

Open questions

• Which alphaproteobacterial lineage is closest to the mitochondrial genome, and what was the original endosymbiont like?
• Pre-endosymbiont, archezoan (host engulfs symbiont), or symbiogenesis (cell fusion)? The proteome evidence Gray assembles fits none of the pure scenarios comfortably.
• How far does mitochondrial specialization go? A systematic, high-resolution cartography of mitotypes across tissues and cell states does not yet exist.

Related

• homo-sapiens-sole-surviving-hominin — ancient mtDNA genomes are the evidence behind the Denisovan and other archaic finds
• insect-tracheal-respiration — in insect flight muscle, tracheoles contact or encircle virtually every mitochondrion, delivering oxygen straight to the consumer
• computational-abiogenesis — endosymbiosis among the major transitions in early evolution
• division-of-labor — the division-of-labour framing Monzel et al. use for tissue-specialized mitochondria
• mutual-aid — endosymbiosis as an ancient obligate partnership between once-free organisms

Sources

• Martin Picard 2026 — 10 Cool Things About Mitochondria
• Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction
• Michael W. Gray 2014 — The Pre-Endosymbiont Hypothesis: A New Perspective on the Origin and Evolution of Mitochondria

Footnotes

1. Michael W. Gray 2014 — The Pre-Endosymbiont Hypothesis: A New Perspective on the Origin and Evolution of Mitochondria ↩

2. Michael W. Gray 2014 — The Pre-Endosymbiont Hypothesis: A New Perspective on the Origin and Evolution of Mitochondria ↩

3. Michael W. Gray 2014 — The Pre-Endosymbiont Hypothesis: A New Perspective on the Origin and Evolution of Mitochondria ↩

4. Michael W. Gray 2014 — The Pre-Endosymbiont Hypothesis: A New Perspective on the Origin and Evolution of Mitochondria ↩

5. Michael W. Gray 2014 — The Pre-Endosymbiont Hypothesis: A New Perspective on the Origin and Evolution of Mitochondria ↩

6. Martin Picard 2026 — 10 Cool Things About Mitochondria ↩

7. Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction ↩

8. Martin Picard 2026 — 10 Cool Things About Mitochondria ↩

9. Martin Picard 2026 — 10 Cool Things About Mitochondria ↩

10. Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction ↩

11. Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction ↩

12. Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction ↩

13. Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction ↩

14. Anna S. Monzel, Jose Antonio Enriquez, Martin Picard 2023 — Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction ↩

15. Martin Picard 2026 — 10 Cool Things About Mitochondria ↩

16. Martin Picard 2026 — 10 Cool Things About Mitochondria ↩

17. Martin Picard 2026 — 10 Cool Things About Mitochondria ↩