Test Your Hypothesis

 BGPT Odds of True

Selection against high-heteroplasmy respiratory-deficient cells is well-documented mechanistically (ROS-mediated apoptosis from mtDNA depletion studies) and evolutionarily (mutation-specific transmission biases in human pedigrees; purifying selection in natural populations). The temporal bias in clonal reconstruction is a logical consequence, but no direct quantification in human tissues exists, preventing certainty. Lower bound accounts for the possibility that synonymous markers dominate in practice, minimizing the bias; upper bound reflects strong mechanistic support from multiple independent systems.

 Hypothesis Novelty

While selection on mtDNA is well-known and survivorship bias in lineage tracing is a recognized concept in other marker systems, explicitly connecting these two phenomena to identify a systematic temporal bias in mtDNA-based clonal reconstruction represents a non-trivial synthesis not directly articulated in the provided literature. The mutation-specific nature of the bias adds an additional layer of novelty.

 Quick Analysis Plan Copied

 Long Analysis Plan

Documented Selection on mtDNA Mutations

The strongest direct evidence that selection distorts mtDNA inheritance comes from Chinnery et al., who analyzed 338 transmissions across six pathogenic mutations and found that random drift alone cannot explain the patterns: A8344G is significantly selected against (O−M mean < 0), while A3243G (n=80), T8993G (n=63), and G11778A (n=56) show significant positive selection Six pathogenic mtDNA mutations show opposite selection directions. This mutation-specificity means selection is not uniform—it depends on the biochemical consequence of each variant, creating heterogeneous survivorship biases across mtDNA marker classes used in clonal reconstruction.

Complementary evidence from natural C. elegans populations (1,524 isolates) shows strong purifying selection on mtDNA polymorphism, with significantly lower median frequencies for nonsynonymous versus synonymous heteroplasmic variants Purifying selection shapes mtDNA polymorphism in natural C. elegans.

Functional Consequences of High Heteroplasmy in Human Cells

Cancer studies provide mechanistic plausibility for why high-heteroplasmy cells face selection. In HEp-2 tumor cells, deliberate reduction of mtDNA copy number (mimicking respiratory dysfunction) via TFAM knockdown or ethidium bromide elevated ROS by downregulating antioxidant genes GSR and GLRX, and increased cisplatin/doxorubicin-induced apoptosis (p < 0.01) mtDNA depletion sensitizes cells to death via ROS pathway. ROS scavengers (NAC, lipoic acid) completely rescued the apoptosis, establishing a clear causal chain: respiratory dysfunction → ROS elevation → cell death.

TFAM manipulation in aging mice showed tissue-specific responses: overexpression impaired hepatic mtDNA expression while improving spleen physiology, confirming that selection against dysfunctional mtDNA is not uniform across tissues TFAM modulation has opposite tissue-specific effects.

The Temporal Bias Mechanism

If selection eliminates cells carrying high heteroplasmy of OXPHOS-impairing mutations, mtDNA-based clonal reconstruction suffers a survivorship bias analogous to the classic "extinction of the unfit": clones carrying pathogenic variants at high heteroplasmy are disproportionately eliminated, causing their apparent expansion timeline to be compressed toward the present. A naive Wright-Fisher model assuming neutrality would date these clones as younger than they actually are, because the observed frequency decline reflects selection, not recent origin. Conversely, clones bearing selectively tolerated (synonymous or low-functional-impact) variants persist longer and appear falsely ancient relative to their true expansion time. The direction and magnitude of this bias depend on the tissue's OXPHOS dependency and the specific mutation's functional threshold.

Critical Caveats and Counterarguments

Most mtDNA clonal reconstruction studies deliberately use synonymous or control-region variants as lineage markers precisely because these are expected to be neutral Drift predominant, selection secondary in mtDNA inheritance. The bias hypothesized here would primarily affect studies using missense or tRNA/rRNA mutations where functional consequences at high heteroplasmy are known. Furthermore, no direct study has yet quantified the temporal distortion in a real human tissue reconstruction—this remains a formal prediction rather than a measured effect. The Chinnery data are from germline transmission, not somatic tissue dynamics, and the C. elegans evidence is from a different organism with different mtDNA maintenance biology.

What Would Falsify This Hypothesis

The hypothesis would be weakened if: (1) longitudinal single-cell mtDNA tracking in human tissues showed no correlation between heteroplasmy level and cell survival/expansion rate for missense variants; (2) clonal age estimates from neutral synonymous variants were indistinguishable from those from pathogenic missense variants at equivalent frequencies; (3) tissues with high OXPHOS dependency (heart, brain) showed no systematic underrepresentation of high-heteroplasmy clones relative to glycolytic tissues.