Mitochondrial Heterogeneity

Abstract

Cell-to-cell heterogeneity drives a range of (patho)physiologically important phenomena, such as cell fate and chemotherapeutic resistance. The role of metabolism, and particularly of mitochondria, is increasingly being recognized as an important explanatory factor in cell-to-cell heterogeneity. Most eukaryotic cells possess a population of mitochondria, in the sense that mitochondrial DNA (mtDNA) is held in multiple copies per cell, where the sequence of each molecule can vary. Hence, intra-cellular mitochondrial heterogeneity is possible, which can induce inter-cellular mitochondrial heterogeneity, and may drive aspects of cellular noise. In this review, we discuss sources of mitochondrial heterogeneity (variations between mitochondria in the same cell, and mitochondrial variations between supposedly identical cells) from both genetic and non-genetic perspectives, and mitochondrial genotype-phenotype links. We discuss the apparent homeostasis of mtDNA copy number, the observation of pervasive intra-cellular mtDNA mutation (which is termed "microheteroplasmy"), and developments in the understanding of inter-cellular mtDNA mutation ("macroheteroplasmy"). We point to the relationship between mitochondrial supercomplexes, cristal structure, pH, and cardiolipin as a potential amplifier of the mitochondrial genotype-phenotype link. We also discuss mitochondrial membrane potential and networks as sources of mitochondrial heterogeneity, and their influence upon the mitochondrial genome. Finally, we revisit the idea of mitochondrial complementation as a means of dampening mitochondrial genotype-phenotype links in light of recent experimental developments. The diverse sources of mitochondrial heterogeneity, as well as their increasingly recognized role in contributing to cellular heterogeneity, highlights the need for future single-cell mitochondrial measurements in the context of cellular noise studies.

Keywords: cellular noise; complementation; heteroplasmy variance; macroheteroplasmy; microheteroplasmy; mitochondria.

Figure 1

Sources of mitochondrial heterogeneity from genetic and non-genetic mitochondrial sources. (A) Potential sources of mitochondrial genetic heterogeneity. (i) A proliferative cell which doubles in volume is expected to replicate its mtDNA complement by approximately ×2 to avoid dilution, which may confound measurements of copy number heterogeneity; noise at partitioning is thought to be somewhat suppressed. The extent of copy number heterogeneity in quiescent cells, and its consequences, are not fully understood. (ii) A cloud of mutations around the consensus sequence is expected given a finite mtDNA mutation rate; we term this “microheteroplasmy.” The actual mutant proportion for most mutations is expected to be very small in reality. (iii) In the more canonical case, “macroheteroplasmy,” a deleterious sequence (such as a deleterious single nucleotide polymorphism or deletion, red circles) induces a pathological phenotype above a threshold heteroplasmy. By contrast, intermediate heteroplasmy has been observed to induce large fitness disadvantages but homoplasmy does not, indicating discord between the two alleles. (B) Potential sources of mitochondrial non-genetic heterogeneity. (i) MtDNA sequence variation can induce variation in the structure of corresponding proteins. (ii) Cristal structure is variable and physiologically regulated to control respiratory output. (iii) A landscape of different supercomplex stoichiometries exists. (iv) Spatially-restricted, transient, depolarization/repolarization cycles of the mitochondrial network are observed. (v) The extent of mitochondrial fragmentation influences heteroplasmy dynamics (see Equation 1). (vi) Intercellular heterogeneity in mitochondrial membrane potential has been shown to be an important source of cellular noise. (vii) Potential sources of mitochondrial epigenetic heterogeneity. MtDNA is observed to undergo epigenetic modification by methylation, although the physiological significance of this is uncertain. Also, oxidative damage to mtDNA nucleotides may cause transcriptional errors. (viii) There are numerous other sources of non-genetic heterogeneity: from distinct spatial niches to ER-mitochondrial interactions. For further discussion, see Main Text.

Figure 2

Factors influencing neutral genetic drift of mtDNA. Heteroplasmy (h, the fraction of a particular variant allele of mtDNA per cell) is not generally constant between cells: it is a random variable and yields a distribution of cellular states. If the consensus and variant alleles experience the same instantaneous birth and death rates per cell, then the heteroplasmy distribution is subject to “neutral drift.” Neutral drift is characterized by the increase in variance of the heteroplasmy distribution with time. It is thought that when cells exceed a particular threshold heteroplasmy, a pathological phenotype may be expressed (red bars). Therefore, the number of pathological cells may increase with heteroplasmy variance. Mathematical modeling has shown that the rate of increase of heteroplasmy variance 𝕍(h) increases with mitophagy rate, as higher turnover provides more opportunities for replication of either allele and cause a change in h. 𝕍(h) changes with copy number, since large populations are more robust to fluctuations. For a given spread of heteroplasmies, 𝕍(h) changes at a maximal rate when mean heteroplasmy is 0.5, and diminishes as one allele dominates over the other. It has recently been shown that mitochondrial network fragmentation can rescale the turnover rate: larger fragmentation results in 𝕍(h) increasing faster with time—as more mitochondria are susceptible to mitophagy—independently of the absolute fusion-fission rates (see Equation 1).