PGT-mt Atlas

Chapter 2 · PGT-mt fundamentals

Heteroplasmy and threshold effects

A heteroplasmy measurement depends on the tissue, sampling time and analytical method. The relationship between heteroplasmy level and clinical outcome is non-linear and variant-specific.

Reading level

Essentials

What exactly is measured

The heteroplasmy level is the percentage of mtDNA molecules carrying a variant at a given position, in a given sample, measured by a given method. All three qualifiers matter. In one person, values in blood, urine, buccal cells and muscle can differ (Stewart & Chinnery, 2015; Burr & Chinnery, 2024), and different methods can yield different results for the same tissue (Treff et al., 2012; Mitalipov et al., 2014).

Different tissues, different levels

When cells divide, mtDNA molecules are shared unequally between daughter cells, and different cell types may select for or against them differently. Heteroplasmy levels in different tissues can therefore diverge over time — this is somatic segregation. For m.3243A>G, heteroplasmy in blood cells falls exponentially with age, which does not happen, for example, in skeletal muscle (Burr & Chinnery, 2024). The level in a mother’s blood does not determine the level in her oocytes.

The threshold effect

At low heteroplasmy levels, mtDNA copies without the pathogenic variant can often support sufficient mitochondrial function to prevent a measurable biochemical defect. As the proportion of mtDNA carrying the pathogenic variant increases, a biochemical defect may become apparent once a threshold is crossed; the relationship between heteroplasmy and function is non-linear (Rossignol et al., 2003). This is the threshold effect. The threshold depends on the variant, the tissue, the respiratory complex, mtDNA copy number and the nuclear background (Rossignol et al., 2003; Burr & Chinnery, 2024). The diagram below shows the idea, not real values.

Three levels of effect

Molecular
the proportion of mtDNA carrying a specific mutation in the sample being studied.
Biochemical
respiratory-complex activity and ATP synthesis in cells or tissues.
Clinical
symptoms, their severity and age at onset.

These levels are not linked by a simple one-to-one relationship. The same blood heteroplasmy level can be associated with different degrees of disease severity, because what matters clinically is the load in the affected tissues — brain, muscle, heart, retina — which are usually not accessible for testing (Gorman et al., 2016; Burr & Chinnery, 2024).

There is no universal safe level

The numbers that appear in the PGT-mt literature — for example 18% and 30% — answer different questions and come from different data (Hellebrekers et al., 2012; Hyslop et al., 2025). None of them guarantees that a child will be healthy, and none can be applied automatically to every variant. Chapter 6 discusses these reference points in detail.

The heteroplasmy threshold used to select an embryo for transfer is chosen and justified for a specific variant; it is not applied as a fixed standard.

Advanced

Formal definition and measurement artefacts

The heteroplasmy level h = n(pathogenic variant) / [n(pathogenic variant) + n(wild type)] at a specific position in mtDNA extracted from a sample. This value characterises the population of molecules in the sample. It cannot be used to reconstruct the heteroplasmy level of each individual cell. Producing a reported result involves several steps — cell selection, lysis, whole-genome amplification where required, amplification of the region containing the variant, detection and calibration. Each step can introduce bias (Mavraki et al., 2023; Spath et al., 2021).

Somatic segregation and tissue variability

mtDNA replication is not tied to the cell cycle, and at cell division mtDNA is partitioned between daughter cells at random. As a result, the proportion of mutant mtDNA can rise or fall through random segregation in subsequent cell generations (Chinnery et al., 2000; Stewart & Chinnery, 2015; Mavraki et al., 2023). Cell-type-specific selection acts in addition: rapidly renewing tissues and post-mitotic cells behave differently, which helps to explain the tissue specificity of mtDNA disease (Burr & Chinnery, 2024). For m.3243A>G, heteroplasmy in blood declines exponentially with age, whereas in skeletal muscle it does not (Burr & Chinnery, 2024). Cell-to-cell mosaicism can arise quite early in embryonic development. For some variants, individual blastomeres have been shown to differ in mtDNA as early as day 3, and this diversity may persist later (Mertens et al., 2022).

Thresholds: tissue, biochemical and variant-specific

Rossignol and colleagues showed that the biochemical threshold differs between respiratory complexes and tissues and that function depends non-linearly on the proportion of mutant mtDNA (Rossignol et al., 2003). The frequently quoted values of 60–80% are a generalised range given in reviews (Xie et al., 2026) and describe the heteroplasmy–function relationship in a particular tissue; they are not embryo selection criteria.

Variant specificity is visible in clinical data as well. Using family data from 137 pedigrees, Hellebrekers and colleagues calculated that a muscle mutation level of 18% or less corresponds to at least a 95% chance of remaining unaffected; for m.3243A>G this chance was significantly lower (p < 0.001) (Hellebrekers et al., 2012). Ji and colleagues estimated symptomatic thresholds separately for different variants (Ji et al., 2023). In the germline, different mtDNA mutations can yield different numbers of embryos suitable for transfer: carriers of m.3243A>G had no oocytes or embryos with very high loads, whereas m.8993T>G values were predominantly extreme and, in pedigree data, offspring levels were more often higher than maternal levels (Otten et al., 2018).

Measurement methods and analytical error

Heteroplasmy is quantified using polymerase chain reaction (PCR) with restriction analysis, pyrosequencing, allele-specific quantitative PCR (ARMS-qPCR), droplet digital PCR (ddPCR), and targeted or whole-genome sequencing (Rubens et al., 2025; Mizuguchi et al., 2025; Mitalipov et al., 2014; Tytgat et al., 2021; Spath et al., 2021; Janssen et al., 2024). The methods differ in their limit of detection, linearity across the full 0–100% range, reproducibility and susceptibility to amplification bias. Their performance is established by validation in a specific laboratory, not by the name of the technology (Mavraki et al., 2023).

Analytical error can be comparable in magnitude to the differences being interpreted. Tajima and colleagues reported an error of −5 to +6 percentage points for their method (Tajima et al., 2007). An illustrative published case concerns m.3243A>G: after transfer of a blastocyst with 12% heteroplasmy, the level in the child’s buccal cells was 15% at one month; a commercial laboratory reported less than 10% in blood at 5 and 12 months; ARMS-qPCR re-analysis at 6 weeks and 18 months gave 47% and 46% in blood and 52% and 42% in urine (Treff et al., 2012; Mitalipov et al., 2014). These data appeared in Mitalipov and colleagues’ reply to Steffann and colleagues (Mitalipov et al., 2014; Steffann et al., 2014). Samples were obtained at different ages and tested by different methods, so the discrepancy cannot be attributed to the method alone.

A heteroplasmy result must be interpreted in the context of the sample, analytical method and limit of detection. The threshold depends on the variant, the tissue and the question the number is meant to answer.

Diagrams and models

Diagram

The threshold effect

thresholdthresholdvariant / tissue Avariant / tissue B0100Heteroplasmy level, %Respiratory chain function
The relationship between heteroplasmy and function is non-linear. At low proportions of mutant mtDNA, wild-type copies compensate for the defect; above the threshold, function declines steeply. The threshold differs between variants, tissues and respiratory complexes (Rossignol et al., 2003). These schematic curves do not represent measured data.

From concepts to Atlas data

  • Within individual embryos →

    Agreement in heteroplasmy levels between samples from one embryo. The grey ±5 pp band is an illustrative guide, not a validated estimate of analytical uncertainty.

  • Embryo, foetus and child →

    How embryo heteroplasmy compares with levels in foetal and child tissues, and how sample type, age and method affect the comparison.

  • Methods and limitations →

    How the Atlas handles detection limits, intervals and values digitised from figures.

Chapter references (21)

All DOIs were checked against PubMed records; exceptions are marked at the entry.

  1. Stewart JB, Chinnery PF (2015). The dynamics of mitochondrial DNA heteroplasmy: implications for human health and disease. Nat Rev Genet 16:530–542. doi:10.1038/nrg3966 · PMID 26281784
  2. Burr SP, Chinnery PF (2024). Origins of tissue and cell-type specificity in mitochondrial DNA (mtDNA) disease. Hum Mol Genet 33(R1):R3–R11. doi:10.1093/hmg/ddae059 · PMID 38779777
  3. Treff NR, Campos J, Tao X, et al. (2012). Blastocyst preimplantation genetic diagnosis (PGD) of a mitochondrial DNA disorder. Fertil Steril 98:1236–1240. doi:10.1016/j.fertnstert.2012.07.1119 · PMID 22921075
  4. Mitalipov S, Amato P, Parry S, Falk MJ (2014). Limitations of preimplantation genetic diagnosis for mitochondrial DNA diseases. Cell Rep 7:935–937. doi:10.1016/j.celrep.2014.05.004 · PMID 24856294
  5. Rossignol R, Faustin B, Rocher C, et al. (2003). Mitochondrial threshold effects. Biochem J 370:751–762. doi:10.1042/BJ20021594 · PMID 12467494
  6. Gorman GS, Chinnery PF, DiMauro S, et al. (2016). Mitochondrial diseases. Nat Rev Dis Primers 2:16080. doi:10.1038/nrdp.2016.80 · PMID 27775730
  7. Hellebrekers DMEI, Wolfe R, Hendrickx ATM, et al. (2012). PGD and heteroplasmic mitochondrial DNA point mutations: a systematic review estimating the chance of healthy offspring. Hum Reprod Update 18:341–349. doi:10.1093/humupd/dms008 · PMID 22456975
  8. Hyslop LA, Blakely EL, Aushev M, et al. (2025). Mitochondrial donation and preimplantation genetic testing for mtDNA disease. N Engl J Med 393:438–449. doi:10.1056/NEJMoa2415539 · PMID 40673696
  9. Mavraki E, Labrum R, Sergeant K, et al. (2023). Genetic testing for mitochondrial disease: the United Kingdom best practice guidelines. Eur J Hum Genet 31:148–163. doi:10.1038/s41431-022-01249-w · PMID 36513735
  10. Spath K, Babariya D, Konstantinidis M, et al. (2021). Clinical application of sequencing-based methods for parallel preimplantation genetic testing for mitochondrial DNA disease and aneuploidy. Fertil Steril 115:1521–1532. doi:10.1016/j.fertnstert.2021.01.026 · PMID 33745725
  11. Chinnery PF, Thorburn DR, Samuels DC, et al. (2000). The inheritance of mitochondrial DNA heteroplasmy: random drift, selection or both?. Trends Genet 16:500–505. doi:10.1016/S0168-9525(00)02120-X · PMID 11074292
  12. Mertens J, Regin M, De Munck N, et al. (2022). Mitochondrial DNA variants segregate during human preimplantation development into genetically different cell lineages that are maintained postnatally. Hum Mol Genet 31:3629–3642. doi:10.1093/hmg/ddac059 · PMID 35285472
  13. Xie Q, Walker HE, Minczuk M, Chiaratti MR, Chinnery PF (2026). Inheritance and selection of the mitochondrial genome. Nat Rev Genet (online ahead of print). doi:10.1038/s41576-026-01019-0 · PMID 42834120
  14. Ji D, Zhang N, Zou W, et al. (2023). Modeling-based prediction tools for preimplantation genetic testing of mitochondrial DNA diseases: estimating symptomatic thresholds, risk, and chance of success. J Assist Reprod Genet 40:2185–2196. doi:10.1007/s10815-023-02880-2 · PMID 37439868
  15. Otten ABC, Sallevelt SCEH, Carling PJ, et al. (2018). Mutation-specific effects in germline transmission of pathogenic mtDNA variants. Hum Reprod 33:1331–1341. doi:10.1093/humrep/dey114 · PMID 29850888
  16. Rubens P, Mayeur A, Chatzovoulou K, et al. (2025). Profiling mitochondrial DNA variant segregation during human preimplantation development: a prerequisite to preimplantation genetic testing for mitochondrial DNA-related disorders. Hum Reprod 40:956–961. doi:10.1093/humrep/deaf050 · PMID 40174913
  17. Mizuguchi Y, Sueoka K, Sato S, Tanaka M (2025). Successful birth after preimplantation genetic testing for rare mitochondrial DNA mutation m.10197G>A. Hum Genome Var 12:7. doi:10.1038/s41439-025-00311-5 · PMID 40164591
  18. Tytgat O, Tang MX, van Snippenberg W, et al. (2021). Digital polymerase chain reaction for assessment of mutant mitochondrial carry-over after nuclear transfer for in vitro fertilization. Clin Chem 67:968–976. doi:10.1093/clinchem/hvab021 · PMID 33822904
  19. Janssen AEJ, Koeck RM, Essers R, et al. (2024). Clinical-grade whole genome sequencing-based haplarithmisis enables all forms of preimplantation genetic testing. Nat Commun 15:7164. doi:10.1038/s41467-024-51508-1 · PMID 39223156
  20. Tajima H, Sueoka K, Moon SY, et al. (2007). The development of novel quantification assay for mitochondrial DNA heteroplasmy aimed at preimplantation genetic diagnosis of Leigh encephalopathy. J Assist Reprod Genet 24:227–232. doi:10.1007/s10815-007-9114-0 · PMID 17342424
  21. Steffann J, Gigarel N, Samuels DC, et al. (2014). Data from artificial models of mitochondrial DNA disorders are not always applicable to humans. Cell Rep 7:933–934. doi:10.1016/j.celrep.2014.05.005 · PMID 24856293

Bibliography →